Report No. 27005-BD
Bangladesh
Private Sector Assessment for Health, Nutrition
and Population (HNP) in Bangladesh
November 18, 2003
South Asia Human Development Sector Unit &
HD Network Health, Nutrition and Population Team
Document of the World Bank
CURRENCYEQUIVALENTS
CurrencyUnit - Taka
US$l - 58.45 BDT
GOVERNMENTFISCALYEAR
July 1-June30
ACRONYMS
AAA Analyticaland AdvisoryAssistance
ADB Asian Development Bank
AIDS Acquired Immune Deficiency Syndrome
ALWARI Acute (Lower)RespiratoryInfection
APP AlternativePrivatePractitioner
BBS BangladeshBureau of Statistics
BCC Behavior Change Communication
BCG Tuberculosisvaccine
BINP BangladeshIntegratedNutrition Project
BMA BangladeshMedical Association
BMDC BangladeshMedical and Dental Council
BMMS BangladeshMaternal Health Servicesand Mortality Survey
BNC BangladeshNursing Council
BPCDOA BangladeshPrivate Clinic and DiagnosticOwners' Association
BPHC BangladeshPopulation and Health Consortium
BPMPA BangladeshPrivateMedicalPractitioners' Association
BRAC BRAC (formerly,BangladeshRural Advancement Committee)
CBR CrudeBirth Rate
CDR CrudeDeath Rate
CIDA CanadianInternational DevelopmentAgency
CIET CommunityInformation & EpidemiologicalTechnologies
CPR Contraceptive PrevalenceRate
DFID Department for InternationalDevelopment(UK)
DHS Demographic and Health Survey
EC European Commission
EIS Epidemiological Information System
EOC Essential Obstetric Care
EPI Expanded Program on Immunization
ESP Essential ServicesPackage
,GAVI GlobalAlliance for Vaccination and Immunization
GDP GrossDomestic Product
GOB Government of Bangladesh
HA Health Assistant
HDS Health and DemographicSurvey(BBS)
HEU Health Economics Unit
HDS Health andDemographic Survey (BBS)
HEU HealthEconomics Unit
HKJ HealthInformation Unit
HIV HumanImmunodeficiency Virus
HNP Health, Nutrition, andPopulation
HPSP Health and Population Sector Program
HRDU HumanResources Development Unit
ICDDR,B Center for Health and Population Research
IDD Iodine Deficiency Disorder
IHE Institute o fHealthEconomics (Dhaka University)
I M C I IntegratedManagement o f Childhood Illness
IMR Infant MortalityRate
I-PRSP Interim Poverty Reduction Strategy Paper
MBBS Bachelor ofMedicine and Bachelor o f Surgery
M C H Maternal and ChildHealth
MDG MillenniumDevelopment Goal(s)
MMR Maternal Mortality Ratio (No. of maternal deaths/l00,000 live births)
MOHFW MinistryofHealth& FamilyWelfare
NGO NonGovernmental Organization
ORS Oral Rehydration Solution
ORT Oral Rehydration Therapy
PHC Primary Health Care
PPP Public-Private Partnership
PSA Private Sector Assessment
PSB Pharmaceutical Society o f Bangladesh
QA Quality Assurance
R T I Reproductive Tract Infection
SD Standard Deviation
SDS Service Delivery Survey
Sida Swedish International Development Cooperation Agency
SMC Social Marketing Company
STD Sexually Transmitted Disease(s)
TB Tuberculosis
TBA Traditional BirthAttendant
TFR Total Fertility Rate
THC ThanaLJpazilaHealth Complex
Tk Taka
UNICEF UnitedNations Children's Fund
USAID United States Agency for InternationalDevelopment
WHO World Health Organization
Vice President Praful Pate1
Country Director Christine Wallich
Sector Manager Anabela Abreu
Task Team Leader Sundararajan Gopalan
Table of Contents
Chapter 1. TheProvisionOfHNPServices in thePrivate Sector........................................................................ i
1.1 Introduction..................................................................................................................... 1
1.2 The Actors....................................................................................................................... 2
1.3 Physicians ....................................................................................................................... 4
1.4 Nurses ............................................................................................................................. 6
1.5 Trained Paramedicals...................................................................................................... 8
1.6 Alternative Private Practitioners ..................................................................................... 9
1.7 HealthFacilities ............................................................................................................ 12
1.8 Other Private HNP InputMarkets................................................................................. 15
1.9 Conclusions................................................................................................................... 15
Chapter2. Demandfor Consumption$Pkuate SectorHNP Services
/ ......................................................... 17
2.1 Introduction................................................................................................................... 17
2.2 Health Service Consumption ........................................................................................ 17
2.3 Public/Private Shares o f Services ................................................................................. 19
2.4 Determinants of Consumption / Health Care Seeking Behavior .................................. 22
2.5 Conclusions................................................................................................................... 31
Chapter3. Interaction between the Publicand the Private Sectorsin HNP ...................................................... 33
3.1 Introduction................................................................................................................... 33
3.2 Current Situation........................................................................................................... 33
3.3 Perceptionso fthe Public Sector towards Working with the Private Sector.................40
3.4 Perceptionso fthe Private Sector.................................................................................. 41
3.5 Conclusions................................................................................................................... 42
Chapter4. Main messages andpolip options ................................................................................................... 44
4.1 Findings......................................................................................................................... 44
4.2 PolicyFonnulation: A newparadigm........................................................................... 45
4.3 Policy Options............................................................................................................... 46
4.4 The Way Forward......................................................................................................... 55
Executive Summary
Bangladeshhas seen remarkable improvements in health indicators over the last 30 years. The
infant mortality rate (IMR) stood at 153 deaths per 1000 live births in the mid-1970s while the
latest data suggest an IMR o f 62. In a similar way under-fivemortality has declined by two
thirds inthirty years andnow stands at 83. Life expectancy at birthis now estimated at 61 years.
Two-thirds o f the deaths in children under five are due to acute respiratory infections (ARI,
27%), perinatal causes (24%) and diarrhea (16%). Neonatal mortality (Le., deaths in the first
month of life) currently accounts for about two thirds o f infant deaths and almost half o f under-
five deaths. More than halfthe neonatal deaths occur inthe first week, many on the first day o f
life. This points clearly to the vital importance of safe delivery and post-natal care to improve
child survival. The data conceal significant socioeconomic disparities. For instance, in 1997,
children in the poorest households suffered 83% higher mortality than children in the richest
households. The matemal mortality rate (MMR) is still high by all standards, with recent
estimates rangingfrom 320-400 deaths per 100,000 live births.
Malnutrition i s an underlying cause of many childhood deaths. Though nutritional status has
improved over the years, still an estimated 35% of children are moderately and 13% severely
underweight. The data show a wide economic differential in malnutrition, with children in the
poorest households beingtwice as likely to be moderately malnourished, and four times as likely
to be severely malnourished as children in the richest homes. Malnutrition has declined in all
economic quintiles, but faster amongthe richest quintile.
The number o f women dying from causes related to pregnancy and delivery i s around 17,000
each year. Thus roughly 45 young Bangladeshi women and nearly 1,000 under-five children die
every day; almost all o f these deaths in women and children could be prevented if appropriate
care services could be provided.
Beyond childhood, a major cause o f morbidity and mortality i s tuberculosis. The number o f
deaths due to the disease was estimated at about 70,000 in 2000-01. HIV/AIDS is an emerging
challenge, still largely limited to certain high risk groups, but with significant factors which
threaten its spread to the general population. Among non-communicable diseases, cancer and
cardiovascular diseases are the leading causes o f morbidity and mortality. Projections show that
as early as 2010, non-communicable diseases will increase their share as cause o f mortality to
59% from 40% in 1990. Injuries are expected to increase their share from 9% to 11%.
Communicable diseases are expected to decrease their contribution to mortality inthe same time
period fi-om 51% to 30%. However, set-backs in the control o f communicable diseases may
occur iffull vigilance i s not maintained.
Thus far, the government's focus inthe health sector has largely been on the establishment and
operation o f facilities and services in the public sector. A majority of contacts between people
seeking health care and providers, however, takes place in the private sector'. It would seem
1 The private sector i s defined to include all actors outsideof the govemment. This broad definition encompasses both
the commercial sector and the not-for-profit sector, non-governmental organizations (YGOs),care providers with or
without formal qualifications, practicing allopathy, homeopathy, ayurveda, or other systems of medxine, fachties o f
various sizes, hospitals, chits, pharmacists and drug vendors, and suppliers o f health-sector related goods and services.
11
..
obvious that addressing the sector's problems requires the institution o f appropriate public
policies to enhance the effectiveness o f the private sector's contribution to public health goals.
Inorderto take full advantage ofthe potential-and address the challenges -ofworkingwiththe
private sector, governments need to gain a better understanding o f private actors. Who are the
private sector actors? What goods and services do they provide? What are the different kinds o f
incentives influencing the behavior o f the private sector? In which areas i s the private sector
well placed to complement public sector efforts? What are the most effective strategies for
engaging the private sector? Although some evidence is emerging regarding these questions,
there is a clear need for further information and knowledge.
The objectives of this Private Sector Assessment (PSA) are to gain a better understanding
of the private health care markets in Bangladesh and to identify areas for increased
collaboration betweenthe government and the private sector.
While the study analyzes private health care markets ingeneral, it uses maternal and child health
(MCH) as an area o f special focus to illustrate general principles and/or draw lessons for the
broader HNP sector. M C H was chosen for this emphasis in view o f its great importance to
Bangladesh, and because M C H outcomes constitute a significant part o f the Millennium
Development Goals (MDG).
Conscious choices had to be made to limit the focus o f the study, as the private HNP sector is too
complex to be covered in full detail under one study. Inparticular, the study does not include a
comprehensive treatment of pharmaceuticals, medical equipment, hygiene products, infant feed
formulations or other such health-related commodities, all o f which constitute important private
sector inputs impacting on health.
The broader context of the Analytical and Advisory Assistance (AM)
This PSA is just one o f several studies initiated under the umbrella o f the HNP Policy Options
AAA work. Notable among the other closely related studies are: a Labor Market Assessment, a
Study o f Health Care Financing Options, a Decentralization Study, a Governance Study, a
comparative analysis of efficiency among private and public providers, and a pilot design for
pro-poor targeting. The AAA work i s a collaborative venture, with the active involvement o f the
Government and various development partners, who are financing specific studies. Broad-based
consultationhas been an integral part o fthe AAA process.
Itis hopedthat this PSA would, along with other existing andnew evidence, inform future public
policy discussions in the HNP sector o f Bangladesh, in the context o f the Poverty Reduction
Strategy, currently under development. It i s also contributing to the preparation o f the HNP
Sector Program (HNPSP), to be implemented in the next three years. The AAA benefited from
the divisional level consultations conducted for the preparation o fHNPSP.
The AAA i s particularly relevant to the new reform agenda being envisioned for the HNP sector
by the Government, with a focus on areas such as public-private partnerships, rather than solely
onthe improvement o fthe publiclyprovided services.
...
111
Main Findings
TheProvision o fHNPServices in thePrivate Sector (Chapter I)
0 Private service delivery sector i s dominant
0 Among the private providers, the major proportion is those with low levels o f formal
training (Alternative Private Practitioners), raising serious concems about quality o f care
0 Lack o f quality standards
0 Inequitable distribution o fhealth care
0 Lack o f competition between public andprivate providers
The PSA analysis confirmed the findings from other studies that the private sector dominates the
provision o f basic care, nursing homes, laboratory and ambulatory diagnostic services in
Bangladesh. The public sector, however, remains the main provider o f inpatient care. Private
sector providers are a heterogeneous group, differing in their training, legal status, system o f
medicine used, type o f organization and on whether or not they held a public sector employment
as well, Altemative private practitioners (APPs) are by far the largest group o fproviders. These
include partially qualified or unqualified allopathic practitioners, drug vendors, and practitioners
of non-allopathic or mixed systems o fmedicine.
Interms o fhumanresources, the private sector predominates (innearly every category o fhealth
professionals, a greater proportion o f them work in the private sector). However, Bangladesh
has one o f the lowest nurse to population ratios in the world: 11nurses per 100 000 population,
compared to 132 in low income countries, and 750 inhigh income countries. The shortages are
accompanied paradoxically by a significant problem o f nurse unemployment because many
private facilities make do with unqualified and unregistered nurses, which i s a cause for concern
interms ofquality.
Gender issues are very relevant when analyzing the human resources capacity in the private
sector in Bangladesh. Other than traditional birth attendants and nurses, male private health
providers far outnumber females: by about 4 to 1 among qualified doctors and by about 9 to 1
among APPs. This has a deleterious effect on women's access to care.
Findings indicate important deficiencies in the technical quality of care delivered even by
formally trained practitioners in both private and public sectors. For example only 10% o f
private providers in hospitals used medical protocols to treat tuberculosis patients. In the
absence o f appropriate mechanisms and institutions to promote quality o f care in a systematic
manner it i s impossible to monitor and assure quality o f care; in such a scenario, the quality i s
unlikely to be high. These problems are even greater inthe case o f APPs, the largest and least
measured group o f providers in the country. Traditional providers' knowledge was particularly
poor on M C H issues, for example, the management o f a newborn with pneumonia and the
complications o f delivery.
Consumption ofprivate HNPServices
The following bullet points capture the main demand-side issues, raised inchapter 2:
0 Overall low consumption o f essential care
0 Poor populations demand / consume basic ambulatory care services from private
providers; private services are not merely for the rich
iv
0 Gender disparities inaccess
0 Financial barriers to access
0 Information and others barriers
Overall health care consumption in Bangladesh in both public and private sectors is low
compared with other countries and relative to need. The private sector is used for the
overwhelming majority o f outpatient curative care, while the public sector i s used for a larger
proportion o f hospital deliveries and preventive care. The higher proportion o f institutional
deliveries in the public sector should be understood in the backdrop o f the fact that overall
proportion o f institutional deliveries i s only 8%. About 90% o f medical care for children with
acute respiratory infection (ARI)or diarrhea i s obtained from the private sector. This indicates
the importance o f the private sector in terms o f access and signals the need for effective quality
o f care measures.
The dependence on the private sector for curative care i s also true for the poor in Bangladesh.
The poorest 20 percent o f Bangladesh children have a higher dependence on the private sector
for the management o f ARI and diarrhea than the richest quintile. The largest differences
between the rich and the poor are for medically trained deliveries, antenatal care, treatment for
ARI, and immunizations. In contrast, the use o f modem contraceptives and oral rehydration
therapy for diarrhea, two commodities where there has been extensive social marketing, do not
show such disparities between the rich and the poor. This suggests that social marketing may
help inreducing some o f the inequities inthe consumption o f certain health-related commodities
across income quintiles. Women and girls tend to receive less medical care than their male
counterparts, with gender bias resulting from cultural norms that require women to obtain
permissionprior to seeking medical care, andneeding to find someone to accompany them when
they do. The situation is made worse bythe lack o f female healthproviders.
Perceptions o f provider's experience and familiarity with the provider are important reasons for
selecting private health providers. Further studies are needed to examine what specific factors
influence care-seeking behaviors inBangladesh.
While expectedly the richest quintile spends more than the poorest quintile (by a factor of 6) on
health care, the proportion o f the spending that goes to the private sector is higher among the
poor than among the rich. The private providers are often closer to the clients and more
conveniently located than public facilities.
coverage - public or private - appear to be major constraints to access to care for the poor.
Financial barriers and lack o f basic insurance
Serious efforts need to be made to address the financial, physical and social barriers to access,
especially for the women and the poorer population groups. Operations research is needed to see
how consumers can influence quality o f care -by being empoweredto demand better quality.
Interaction betweenthe public and private sector in HNP
Chapter 3 brings out the following findings:
0 Weak regulatory framework; ineffective enforcement
0 Several pilot initiatives include government-NGO partnerships, but most are donor-
financed and have not been scaled up
0 MOHFWneeds the fiscal space andgreater and different kinds o fcapacity to enhance
engagement with the private providers.
V
0 Misperception o f size and scope o f private sector by government and lack o f capacity to
play any role other than service provision.
The range and magnitude o f government engagement with private providers i s not congruent
with their importance. The bulk o f interaction takes place interms o fregulation, andwith regard
to private clinics and hospitals. Less formal, less organized providers, such as non-allopathic
practitioners, including traditional birth attendants, and drug vendors and retail pharmacists have
very little interaction with government. Thus whatever little public-private engagement has
occurred inBangladesh has mostly excludedthe providers o f greatest importance to the poor.
There are, however, positive experiences in the area o f public-private engagement, including a
number of pilot initiatives to work with private, mostly non-profit, service providers. Notably,
the very successfil NGO-contracting experiences onnutrition andurbanprimary health care, and
other forms o f partnership in areas o f family planning, TB control and immunization, bear
important lessons for the rest o f the HNP sector. In addition, involvement of, and collaboration
with, some private sector actors has occurred sometimes inpolicy discussions and formulation,
though this has not been a consistent feature. Recently the government has been consideringthe
possibility o f contracting NGOs to better manage several hundred public facilities at the union2
level with a view to expanding the coverage and improving the quality o f essential 3" services.
Such an initiative i s a very welcome step inthe right direction and should be supported actively
by the development partners, so that it quickly matures into a well-designed large-scale pilot
with the potential for scaling up iffound successful.
Secondly, the fulfillment o f government's stewardship responsibilities in the HNP sector could
be enhanced. Health services regulation currently appears to be a fairly low priority issue. There
i s little collaboration with professional and providers organization, nor support for self
regulation. Currently professional and provider organizations are primarily playing the role o f
trade unions. Neither consumer nor patients' organizations have yet emerged to play an
advocacy role, nor to engage in monitoring o f service quality and outcomes. Instruments to
engage private actors require government officials to perform tasks very distinct from their
traditional activities. There is currently very little capacity to implement such instrumentsinthe
MOHFW or inlocal government bodies.
Thirdly, misperception and low capacity underlie weak public-private engagement. The policy-
makers' interviews reveal that there is limitedunderstanding o fthe private sector size androle in
provision o f care, especially for M C H services in rural areas. Most policy makers - especially
those at the national level - believe that private providers mainly cater to tertiary care needs o f
the rich inthe capital and other urban areas; while in fact, it i s the poor who are more dependent
on the private providers, especiallythe APPs.
A union i s the lowest administrative division with fixed public facilities for the provision of health care and
consists o f around 20 villages, with around 20,000 population on average. 8-10 unions generally make a sub-district
(Upazila or Thana), which has a population o f around 200,000 on average and 6-8 sub-districts make a district.
There are about 4,700 unions, 470 Upazilas and 64 districts inBangladesh.
vi
Key issues
Chapter 4 highlightsthe following conclusions:
Public sector is not strategically usingthe scarce resources that are available inthe
private sector
The low level ofpublic expenditure on health care and the fact that all public spendingon
health goes to public providers leaves little head room for contractingwith private
providers
Low level care provided byAPPs andpersistent shortages o f formally trained staff
Uneven quality andproblems with access
Poor needbetter capacity to make informed decisions about the quality o f care provided
byprivate practitioners
The poor are more likely to forego medical treatment due to financial constraints
Otherbarriers prevent appropriate health seeking behavior
Lack o f competition between public andprivate providers leads to inefficient use o f
resources
Poor coordination betweenpublic and private sector, and lack o f complementarity, which
contributes to gaps incoverage
PolicyImplications
The central policy implication from these conclusions i s to revisit the role o f government in
H",given the realities ofresource andcapacity constraints inthe public sector, the already
dominant place held by private actors in the financing and delivery o f HNP services and the
serious concerns about quality, access, accountability and governance with regard to both private
and public services. A policy shift from an approach o f fixing the public sector problems to one
of greater engagement with the private sector appears to be warranted. In particular the
following three broad areas would appear to deserve priority ingovernment actions:
-Under-consamption0fsemz;CeS thepoor and women
-Service quality and outcomes
-The knowledgebase
PolicyOptions
The following policy options were discussed during stakeholder consultations in early M a y
2003, and at the policy retreat and dissemination workshop in July 2003. These consultations
included government officials, private sector actors, civil society, academia, and development
partners. While there was broad agreement on the need to increase the engagement with the
private sector and on the value o f the options presented here, it was felt that further debate,
consultations, pilot tests and studies are neededbefore policy decisions are taken.
a Develop a clear public policy towards the private sector that harness the valuable
resources that are available inthis sector.
0 The government needs to create "head room" in its public expenditure envelope so that
some public resources will become available for influencing the behavior o f private
providers through contracting with private providers and subsidizing care for the poor.
a BringAPPs into the service provider systembyworking with them instrengthening skills
and increase the number o f formally trained staff through training.
vii
Increase quality benchmarking, performance based competitive pressures and incentives
to attract private practitioners to work in low coverage areas in addition to traditional
regulatory and quality assurancetechniques.
Make information about the quality and price o f private providers readily available to
consumers, especially for the poor.
Introduce targeted subsidies and community level insurance for the poor and social
insurance mechanisms for civil servants and formal sector workers.
Use financial incentives (Le., fees for vaccinations) and social marketing techniques to
overcome other barriersto appropriate health seekingbehavior.
Increase competition between public and private sector through competitive and selective
contracting andperformance benchmarking.
Introduce internal markets (make public providers compete for public funding on a
performance basis) and new public sector management techniques (ie. contracting out,
contracting in, management contracts etc).
Redefine the role o f the M O H and strengthen its core stewardship capacity in areas such
as strategic planning, monitoring and evaluation, coordination, regulation, quality control
and enforcement.
It is clear that the PSA has not coveredthe whole ground on the subject and further studies and
analytical work are needed. It must be stressed however, these suggestions for further
studies are not to be misconstrued as a reason to delay policy actions for which
considerable evidential basis already exists. A distinction must be made between operations
research to pilot-test the policy options and the other research activities aiming to generate new
knowledge.
The Way Forward
The authors do not wish to be prescriptive about the solutions for the issues emerging from their
study. Rather, public policy should evolve through a participatory process in Bangladesh, with
the active and broad-based involvement o f all stakeholders. Therefore, the policy options
presented here - boththe "what" and the "how" - are merely a startingpoint for national debate.
As the Government i s preparing its new Health, Nutrition and Population Sector Program
(HNPSP), and has outlined a new reform agenda for the future, this study and the other related
studies under the AAA work could meaningfully inform policy dialogue, taking a fresh
perspective on sector reforms needed to achieve HNP outcomes as part o f the MDGs.
In order to advance such national policy debate and enable the government of Bangladesh to
better harness the potential o f the private sector for the achievement o f health outcomes, the
following are possible next steps:
Set up a Public-Private Task Force inthe MOHFW.
0 Create the necessary fiscal space or "head room" inthe public resource envelope.
Capacity development inthe MOHFW to enhance its engagement with the private health
sector.
Participatory policy-making and more inclusive planning andprogramming.
0 Pilot activities to test the selected policy options.
...
Vlll
Completion o f other related studies (demand-side financing, May 2004
governance, pro-poor targeting, comparative study of efficiency)
Disseminationo f existing evidence, multi-pronged communication Julyto December 2003
exercise, consultations across the country
Development o fbroad-based HNP Policy Options July 2003 to June 2004
Initiation o fPilot Interventions (e.g., vouchers, micro-insurance, January 2004
results-based contracting with private sector, demand-side
subsidies, pro-poor targeting)
Chapter 1. The Provision of HNP Services in the Private Sector
By far the main suppliers of clinical care to Bangladeshis are the alternative private
practitioners (APPs), who includepartially qualiped or unqualiJied allopathic and non-
allopathic practitioners, and village pharmacists. The best estimate is that APPs
outnumber all qualiJied allopathic physicians by about 12:l.
Bangladesh has one of the lowest nurse topopulation ratios in the world. The challenge
is not only to increase their numbers, but to enhance the role and quality of nursing care.
The widespread use of unqualijied workers as "nurses" is a serious causefor concern.
Although physicians are also in short supply, they outnumber nurses by 1.7 to I,
compared to an average ratio of 0.6 to I among low-income countries. The government
invests relatively heavily in the education of physicians compared to other providers.
Current plans to expand the numbers of physicians ought first to address the existing
problems in theproduction and retention of physicians. Newfindings in this study show
a signipcant "brain drain" of doctors migrating to high-income countries.
Other than traditional birth attendants and nurses, male health providers far outnumber
females: by 4 to 1 among qualijied doctors and by 9 to I among APPs, which results in
signijicant gender disparities in access.
New findings indicate that there are important deficiencies in the technical quality of
care, especially though not only by APPs. Institutions and mechanisms topromote quality
in the health sector are lacking, in the public as well as private sectors. There is little
experience in Bangladesh with strategies to influence where or how private providers
practice medicine. Not surprisingly, the largest group of providers, the APPs, is also the
least influenced by currentpublic policies.
Thegovernment could make the biggest gains in ensuring theprovision of health services
by vigorously and systematically tackling quality issues, and by beginning large
experiments to learn how to monitor and influence the most signiJcant group of health
providers, the APPs.
1.1 Introduction
This chapter deals with the issues related to the supply side o f private health services,
specifically examining the health labor market3 and private health infrastructure. We begin by
examining the different types o f private sector providers o f health services. Inthe last section o f
the chapter, we outline what i s known about the types o f private health facilities, particularly the
different types o finpatient andoutpatient facilities, pharmacies, andlaboratory services.
This chapter does not fully deal with the supply o f health goods, notably the commodities such
as contraceptives, bed-nets, oral re-hydration fluids, and infant formula and feeding supplements.
Also, we do not examine other private key input markets such as those for manufacturers and
distributors o fpharmaceuticals, medical equipment, and construction o fbuildings.
Greater details on the labor market o f HNP sector inBangladesh can be found inthe related study on the health
labor market, which was fiianced by CIDA. This chapter's treatment o f the labor market discusses key aspects
about the supply o f services.
2
1.2 The Actors
In this report, we categorize the health providers as qualified allopathic providers who are
formally recognized through qualifications that allow them to be registered with a public agency
(e.g. doctors, nurses, and trained paramedicals who practice "westem" medicine), and providers
without formal training and recognized qualifications in allopathic medicine, whom we have
labeled as alternative private practitioners (APP); dais OY traditional birthattendants (TBA) form
a specialized type o f APP, but we have treated them as a separate group o f providers inview o f
the largenumbers andthe very limitedscope o fservices they provide.
The main sub-categories o fAPPs include:
Non-qualified allopathic providers, who have not received formal education to practice
allopathic medicine.
Pharmacists who supply allopathic or other medicines, but who do not have formal
qualifications to diagnose illness or prescribe medicines.
Kabivaj, who practice an ayurvedic system o f medicine, based on ancient Hindi systems
of medicine and commonly involving diet, herbs, and exercise. Some o f these providers
have been formally trained inayurvedic colleges.
Totka, who combine ayurvedic, unani (a traditional Muslim system o f medicine), and
shamanistic systems. They may use allopathic medicines as well.
Spiritual healers, who often rely on chants or sacred readings intheir treatment.
Homeopaths, who follow the homeopathic system o f medicine, which involves treatment
through minute quantities o f the presumed cause o f disease. In some cases, these
providers have been formally trained and recognizedinhomeopathy.
The distinction between the public and private sector providers is not always clear. A
considerable proportion o f qualified allopathic doctors and other formal sector providers work
entirely in the private sector (see below), but many o f the public sector doctors also practice
privately -- either by working at private clinics and hospitals after public hours, or by charging
private fees while practicing at public facilities. In this chapter, we consider this type of
practitioner as first a public sector health worker, even if their participation inthe private sector
dominates their time and income.
Private providers are clearly a heterogeneous group, differing in their training, legal status,
nature o f service, mix o f public and private practice, and type o f organization (Table 1.1). The
consequence o f this is that some types o f providers are less visible for study or regulation,
making it difficult to ascertain the true size or nature o f the private sector. This is particularly
the case for those that are not formally trained and arepracticing illegally.
- 3
Table 1.1-Visibility of Private Health Providers According to Selected Characteristics
Characteristic More Visible Less Visible
Legal status and Formally trained and Informal -untrained and illegal
training operating legally (e.g. (e.g. shopkeepers, itinerant vendors)
doctors, nurses
pharmacists)
Organization IIncorporated for-profit or Unincorporated solo practitioners
non-profit organization
Size o f facilities ILarge hospitals, networks Solo practitioners
o f clinics
Nature o f service Comprehensive clinical Single product or service (e.g. drugs)
care
Public-private mix Fulltime public practice or Illegal dual practice
legal dual practice
Alternative private 284,000 231 100% (Ali,77 473
2001) (Sarder& Chen, 1981)
Allopathic providers 110,000 90 100% 38 145
(Claquin, 1981) (ORQ-MargQuest, 2000)
~
Traditional providers 173,000 50 239
Dais (TraditionalBirth
Attendants) 119,000
Sources: Midrange estimates derived from weighting provideripopulationratios from Peters et al, 2003; Ali et al,
2001; ORQ-Marg Quest Ltd, 2000ab; BBS, 1998; Sarder and Chen, 1981; Claquin, 1981
Doctors and nurses are here considered as public sector providers ifthey are employed by the public sector, even if
they also practice privately. This leads to an under-estimate of the proportion o f these categories in the private
sector. The more appropriate approach would be to count dual practitioners under both the public and the private
sectors. While that would be double-counting those providers, it would lead to a more accurate reflection o f the
respective share o f the market held by the public and private sectors.
4
International comparisons on health personnel are fraught with difficulties, because o f the
differences indefinitions, and the poor quality o f data. Compared with other countries, there are
fewer qualified physicians and nurses per capita in Bangladesh (Table 1.3). Another striking
finding is that the physician to nurse ratio is very highinBangladesh. The implications are that
Bangladesh needs to consider ways o f dealing with the shortage o f physicians, but even more
urgentlythe shortage ofpracticingnurses.
Table 1.3 - International Comparisons of Physician and Nurse to Population Ratios
(Around 1998)
Country Physiciansper 100,000
population Nursesper 100,000
population Physician/Nurse Ratio
Bangladesh 19 11 1.8
India 106 94 1.1
Pakistan 57 34 1.7
Sri Lanka 37 103 0.4
Global Average 146 334 0.4
Global Median 114 233 0.5
Low Income Countries 73 132 0.6
Middle Income Countries 142 278 0.5
High Income Countries 286 750 0.4
1.3 Physicians
There i s no precise count o f the physicians actively practicing inBangladeshtoday. As of March
2003, the Bangladesh Medical and Dental Council (BMDC) reported a cumulative total of
34,541 physicians registered. But according to BMDC, until 2002 Bangladesh had 28,537
doctors (HRDU, 2003). However, a survey of their membership registered through the year
2000, returnedonly 9,988 responses (BMDC, 2003). In 1998 an HEUreport made reference to a
study, based upon a 1993 census, reporting there to be approximately 13,200 practicing
physicians with an additional 2,800 abroad and 2,000 unemployed. The same report also cited a
Bureau o f Statistics Survey o f Professional and Miscellaneous Services Personnel, which said
that there should have been22,356 doctors in 1993-1994based upon the same 1993 census.
Registereddoctors work inboth the public and private sectors, but estimates o f those working
only in the public sector are difficult to come by. In the B M D C survey, 27% of respondents
reported they were only in the public sector (BMDC, 2003). However 39% reported mixed
public employment with private practice. Another study reported that 69% o f practicing
physicians were in the public sector, but the percentage of those who also had a private
component was not given (HRDU,2003). The estimated percent o f those working exclusively in
the private sector varies from 22% (BBS, 1998) to 31% (Health Economics Unit, 1998b) to 34%
(BMDC, 2003).
"Brain drain" i s a serious problem for Bangladesh, with many physicians leaving the country to
work elsewhere (See Box 1.1).
5
Box 1.1-The BrainDraininthe HealthSector inBangladesh
Ina background study examining graduatingMBBS classes from three Bangladeshmedical colleges, it
was found that more than 20% of the 1975 graduating class had emigrated, comparednearly 28% of the
1985 cohort, with an annual loss of nearly 1.5% of doctors graduating between 1975 and 1995. The
USA, Saudi Arabia, and the UK are the most common countries of emigration. Further details about the
physiciansandtheir characteristicsmay be found inthe Labor Market Assessment (Peters et al, 2003).
The total cost of losing a doctor cannot be easily measured. However, the costs of medical educationcan
be reasonablydetermined. There are currently 13 medical collegesinthe public sector (of which five are
new) and 20 nongovernmental medical colleges (HRDU, 2003). Estimates for the cost of medical
education for an MBBS degree varies between establishedpublic schools, newly opened public schools
and private schools. Based upon a weighted average of five establishedpublic medical colleges, it i s
estimatedthat Tk 250,000 (US$4,000) i s spent over the average of 5.8 years it takes to produce anMBBS
graduate (Health Economics Unit, 1998a). A comparable figure for a five-year program at Dinajpur
Medical College, a new public school (including costs of establishing the school and annual recurring
costs) i s Tk 491,000 (USSl0,OOO). Private schools are the most expensive with total costs per graduate
estimatedto be between Tk 800,000 and 1,000,000 (US$l6,000-20,000) (HRDU,2003).
1.3.1 IncomeLncentives/Disincerztivesfor Physicians
Estimates o f income of physicians vary by type o f practice and level o f experience. Ina survey
o f private health service establishments, monthly income varied from Tk 7,500 to Tk 150,000
(US$ 150 to US$ 3,000) with the majority o f the practitioners earning incomes at the lower end.
However, 95% o f those sampled also had some public component to their income (BBS, 1998).
In a separate study o f public employees who also had a component o f private practice, 79%
reported a monthly government salary between Tk 5,000-10,000 (US$ 100-200), 16% had
salaries up to Tk 15,000 per month, and 5% earned Tk 5,000 or less. As for their private salary
component, 19% earned less than their government salary, 21% earned an amount similar to their
government salary, and 56% earned more. The overall average total income was Tk 27,500
(US$ 550) per month. In the absence of third party payments and reliable income tax
information, these numbers are likely to be underestimates even though efforts were made to
verify them (Gruen et al, 2002). It i s estimated that only 12% o f doctors are listed by the
National Board o fRevenue, which i s the agency responsible for tax collection (UNB, 2002).
Disincentives for private practice in the rural areas include lower purchasing power o f patients,
competition with alternative private providers, and a weak infrastructure. Another concern i s
harassment from local mafia-like structures (Gruen et al, 2002). Another difficulty with
attracting doctors for rural areas i s that most physicians come from urban backgrounds. They are
reluctant to give up that life and do not want their children to be deprived o f the opportunities
available incities (Chaudhury and Hammer, 2002). Inthis study o f absenteeism, determinants o f
government physicians' likelihood o f attending their rural postings included living in the same
locality as that o fthe clinic, access to roads, and electrification.
Inan extensive surveydone for the MOHFW on incentive schemes for public sector doctors and
other health workers, the most frequently mentioned source o f dissatisfaction was "lack o f
promotion", followed by "low salaries". Figure 1.3 shows reasons for discontent among medical
practitioners: non-recognition o f good work was also a cause o f discontent, along with
6
inadequate residential and clinical facilities, lack o f access to quality health care for the
physicians and their families, and physical and social insecurity (SRGB, 2002).
1.3.2 QualiQ of CareProvided by Physicians
There i s little assessment or reporting on the quality o f physician care in Bangladesh, in the
public as well as the private sector. A background study on private practitioners found that about
90% o f private hospitals maintained patient records, and nearly 60% used standard treatment
protocols (HEU/IHE/NIPORT, 2003) (Figure 1.1). Nearly all facilities claimed to use safe
disposable syringes. Ninety-eight percent o f private physicians appropriately recommended ORS
for treatment o f acute diarrhea, but only two-thirds correctly identified the need to do a sputum
test to assess a suspected case o f tuberculosis, and only 61% would correctly do an examination
for a woman with post-partum bleeding. Though limited inscope, these findings show that there
i s significant room to improve the technical quality o f care among private practitioners. In
Chapter 2, quality o f care i s examined from the patient's perspective.
Figure 1.1- Percentage of Private Hospitals using MedicalProtocols to Treat Patients
I I
C h i l d h o o d d i s e a s e s I I I
I I I
I I I I
II
M a t e r n i t y c a r e I I II I
I
S e x u a l l y t r a n s m i t t e d II I
I I I
d i s e a s e s I
H I V I A I D S ,I I I
I
T u b e r c u lo si5 I I
1 I
0 % 5 % 1 0 % 1 5 % 2 0 % 2 5 % 3 0 % 3 5 % 4 0 % 4 5 % 5 0 `
Source: HEU/IHE/NIPORT, 2003 [Study commissioned specifically for this PSA] and authors' calculations
1.4 Nurses
Eventhough the estimates on the number ofnurses working inBangladeshare not consistent, by
any standard the number o f qualified nurses i s very low and the number actually employed i s
even lower. As o f February 2003, the Bangladesh Nursing Council reported a total o f 19,066
nurses (93% female) in their registry. Distribution by year o f registration and sex i s shown in
Figure 1.2. But based on a national census o f professional services personnel, there were just
over 12,000 nurses working in 1993-94, a ratio o f about 11nurses per 100,000 population (CSIP
1995). Ninety-five percent o f the nurses work inhospitals and clinics inurban settings.
7
Figure 1.2 - Registration of Nurses by Year and Sex
Registered Nurses by Year and Sex
I2O0
I 1000 11
I Fl
1 .v,
I 2cn 600
5
400
I Z
I 200
I
I 0
Source: BangladeshNursing Council, 2003
Data on immigration o f nurses from Bangladesh into other countries are not available, and a
survey o f Bangladesh nursinggraduates did not produce meaningful results on the magnitude o f
the problem. Giventhe longstanding global shortage o f nurses, it i s likely that many nurses that
are qualified to work inother countries will emigrate if employment opportunities remainpoor in
Bangladesh. However, nursing training facilities are inpoor condition, and the vast majority o f
students and teachers are not proficient in English, which is a requirement for work in many
other countries (Peters et al, 2003).
The nursing market in Bangladesh is further undermined by a lack o f standards and regulation.
To obtain work in private clinics, qualified nurses must compete with unregistered individuals
who act as nurses, even though they may have minimal private training to give injections and
provide unskilledcare to patients. The average monthly salary o f registered nurses i s Tk 8,700-
10,000 (US$200-230) inthe public sector, though unqualified nurses may work for much lower
wages inprivate clinics (Begum, 1998). Low salaries and lack o f promotion appear as some o f
the major concerns o f public sector nurses, who share many o f the same concerns about
incentives as physicians (Figure 1.3).
The challenges o f enhancing the role o f nursing, improving the quality o f nursing education, and
better regulating the nursingmarket are at least as important as increasing the number of nurses
in Bangladesh. In order to deal with a growing population and maintain the ratio with
physicians, it i s estimated that by 2020 an additional 20,567 nurses would need to be trained. If
one wanted to improve the doctor:nurse ratio to 1:1, an additional 45,649 nurses would be
required (HRDU, 2003). Before accounting for losses due to retirement, emigration, and leave,
this would requiremore than doubling the current output o fnurses.
8
Figure1.3 - Causes of DissatisfactionAmong DoctorsandNursesin the Public Sector
Percent Ranking Either First or Second as Cause of ,
Dissatisfaction
40%
5 30%
2 20%
I I
I 10%
I
0%
I Cause of Dissatisfaction I
I
Source: SGRB, 2002
1.5 TrainedParamedicals
There are many types o f paramedical professionals practicing in Bangladesh, though there i s
little reliable information on their numbers or types o f practices. Most o f the paramedical
schools require a grade 10 or secondary school certificate (SSC) to enter their three-year training
programs. As can be seen from Table 1.4, the public sector employs some types o f specialized
paramedical personnel trained specifically for public health functions, such as the Health or
Family Planning Inspectors. On the other hand, the private sector has the dominant number o f
laboratory technicians (94%), medical assistants (89%), and pharmacists (69%); though in each
case, it i s doubtful that the private sector practitioners actually have the full educational
qualifications as those employed in the public sector. The vast numbers o f private sector
personnel in these categories suggest that there i s a considerable private market for the types o f
services they can provide. Inthe case o f the medical assistants and pharmacists, they are likely
also acting as physician providers, i.e. making diagnosis and prescribing treatment to patients.
There i s very little information on paramedics. Basic questions concerning how paramedical
professionals are trained or supervised, how they practice their professions, what their concerns
are, or what contributions they are making to the health system have not been seriously
addressed. Given the lower cost o f training, and the relative ease o f selecting paramedical
trainees from rural populations, the main strategic issue for Bangladesh i s to consider whether
paramedicals could take up more o f the responsibilities in the health care system. This question
i s particularly relevant in considering options in remote areas where it i s difficult to get MBBS
doctors and registered nurses to be stationed in either public or private sectors.
9
Table 1.4 -Estimatesof HealthWorkers inBangladeshOther than PhysiciansandNurses
Provider Public Sector Private Sector
I Medical Assistant I 2002 1996-97
5.598 I 45.603' I
Pharmacists 7,622 1,789
Licensedpharmacists (without universityor technology 0 15,477
Laboratory Technicians 1,840 29,085
Radiographers 1,054 ?
Health Inspectors 1,401 0
Family PlanningInspectors 4,110 0
Dentists (& Dental Surgeons) 1,740 1,247
Other trained paramedical6 3,574 ?
Total 28.941 1.247
Sources: Public sector FromHRDData Sheet 2002 (HRDU, 2003). Figures do not include other para-professionals,
including Family Welfare Assistants (22,350), Health Assistants (2 1,016), Assistant Health Inspectors (4,202), and
Family Welfare Visitors (5,248). Private sector Most categories are from BBS (1998), except for pharmacists (ORQ,
2000), which are likely under-estimates. The estimates do not include another 45,820 health related workers
estimated to be working in the private health sector, and i s intended to exclude those working as village doctors
(BBS, 1998).
There are also a large number o f auxiliary health workers who have shorter periods o f training
and lower entry requirements than the paramedical workers considered above. Many o f these
workers have been specifically trained for the public health workforce, including some 50,000
Family Welfare Assistants, Health Assistants, and Family Welfare Visitors, and an estimated
25,000 traditional birth attendants who have been involved in public sector programs (PRU,
2002). NGOs and government have also trained various types o f community health volunteers,
such as BRAC's Shasthyo Shebika (Khan et al, 1998). While the drop-out rate of community
volunteers can be quite high, their success seems to be dependent on carehl selection o f
volunteers, involvement o f communities, supportive supervision, and good training (Islam et al,
2002; Khan et al, 1998; Amhold, 1979). In the case o f tuberculosis treatment, BRAC's use o f
illiterate community volunteers turned out to be 50% more cost-effective than the comparison
government program (Islam et al, 2002). Around the world, the recurrent lesson has been that it
i s easier to initiate these programs than to sustain them (Walt, 1988). Further opportunities for
usingcommunity volunteers inBangladesh should be considered only when it is clear that proper
attention can be paid to the conditions that make community volunteers successful beyond an
initialperiod, and where adequate monitoring can be sustained.
1.6 AlternativePrivate Practitioners
It has long been recognized that APPs provide the majority o f health care in Bangladesh,
particularly inrural areas (Claquin, 1981;Sarder and Chen, 1981). These providers are very well
embedded into the culture and society o f villages (Bhuiya, 1992; Feldman, 1983; Ashraf et al,
1982; Leslie, 1976). APPs are also becoming more organized, forming their own professional
'Probably includes many healthproviders who do not have a three-year Medical Assistant training
Other trained paramedicals includes Physiotherapists, Family Planning Officers andrelated professionals
10
associations in local areas. Despite the dominance of these types o f providers, relatively little is
known about the actual number of the different types o f alternative private providers, the types o f
practices they have, or how their behavior can be influenced. Other than a brief period when
government sponsored the palli chikitshak training program in the early 1 9 8 0 ~policymakers ~
have largely ignoredthe informal sector.
1.61 Typesof Qualifications andPractices
Formal training in traditional systems o f medicine now exists in Bangladesh, with govemment
recognizing nine unanicolleges and six ayurvedic colleges with each havinga four-year diploma
course (BBS, 2002ab). However, the vast majority o f APPs practicing in Bangladesh have not
received formal education in their system of medicine, though a substantial proportion have
received some semi-fonnal training. For example, in a study in Brahmanpura in 2000, 61% o f
APPs had some kind o f certificate o f health training (ORG-Marg Quest Ltd, 2000b). Most
traditional providers have hadtrainingthrough apprenticeship (Feldman, 1983; Sarder and Chen,
1981). The general education levels among the APPs tend to be higher in the allopathic
providers than in the traditional practitioners, with the majority o f allopathic providers having
completedjunior secondary school, and many having completed high school (ORQ-Marg Quest
Ltd, 2000ab; Bhuiya, 1992; Sarder and Chen 1981). The gender distribution o f private
providers, which has important consequences for women's access to health care, i s described in
Box 1.2.
Box 1.2 - Gender Distributionof Private Providers
With the exception o fthe traditional midwives, who are female, the APPs are largely male, particularly
the allopathic practitioners. In 1976-77, a nationwide survey found that 99% o f alternative providers
were male (excluding traditional midwives). Ina census taken inMatlab thana in 1978, similar
proportions o funqualified allopathic and homeopathic providers were male, whereas the kabiruj and
totkas were more evenly distributedbetweenmale and female (Sarder and Chen, 1981). Inthe
Brahmanpara study mentioned above, 98% o f village doctors were male (ORG-Marg Quest Ltd, 2000b).
The obvious implication o f such a male dominance o fproviders ina traditional m a l society i s that it i s
more difficult for women to access health care even through the APPs. This issue i s discussed further in
Chapter 2 where the demand for and use o f health services are considered.
Most o f what i s known about the practice patterns o f APPs comes from small area studies,
making it difficult to generalize results to the entire country. APPs are known to provide
services for a wide range o f health conditions, though almost exclusively on an outpatient basis.
Village doctors nearly always sell medicines (94% o f village doctors in Brahmanpara; 95% o f
allopaths in Matlab), and the majority provides dressings. Kabivaj and totka providers are less
likely to sell medicines, and many o f the spiritual healers provide a much narrower range o f
services for a more limited set o f conditions. Some providers, such as bonesetters, provide a
very specific set o f services. Similarly, most traditional birthattendants tend to provide services
only for childbirth.
Most APPs provide services close to their population base. The allopaths and homeopaths tend
to have a small building for their business in a local market or in the village (ORG-Marg Quest
Ltd, 2000b; Bhuiya, 1992), as do those who are pharmacists (ORG-Marg Quest, 2000a).
Traditional healers tend to operate fi-om their homes, and are more likely to make home visits
(Bhuiya 1992). Allopaths are more likely to work on a full-time basis than other providers
11
(ORG-Marg Quest Ltd, 2000b; Sarder and Chen, 1983). InClaquin's (1981) study, only 35 YOof
APPs worked on a fulltime basis, seeing an average o f 17 to 52 patients per week. In
Brahmanpara, the village doctors nearly all worked fulltime, and averaged 18 patients a day
(ORQ 2000). Incontrast to these providers, traditional midwives averaged only two patients a
week (Claquin, 1981).
1.62 Incentive and Disincentives of APPs
There are a few reports on the incomes and fees charged by APPs, though both appear to be
generally modest. Claquin (1981) reported that APPs charged between two to four Taka for
consultation fees on average, which was equivalent to $0.12 to $0.25 at the time o f the survey.
However, payments from medicines could average 10 times this amount, with unqualified
allopaths charging more than ayurvedics, followed by homeopaths and spiritual healers. In
Brahmanpara, village doctors claimed to eam about Tk. 1,600 (US$ 32) per month from
prescribing medicines to their patients, and another Tk. 2,200 (US$ 44) per month through other
pharmacy sales (ORG-Marg Quest Ltd, 2000b).
There i s little other information on the motivations o f APPs, nor about their aspirations or
practice constraints. The Brahmanpara study reported a near unanimous interest among village
doctors in working in partnership with the public sector (ORG-Marg Quest Ltd, 2000b).
Although there have been a number o f projects that attempted to train APPs, these are not well
documented, and little i s known about what strategies would be effective in improving their
skills, preventing them from practicing illegally, or reducing the potential harm caused by poor
quality services.
1.63 Quality of ServicesProvided by Alternative Private Practitioners
Poor quality o f services is one o f the major concems regarding informal sector providers. In a
detailed anthropological study on village practitioners in three villages, Ashraf and colleagues
(1983) emphasized that a large variety o f drugs are being prescribed, usually inappropriately.
Because o f a fatalistic attitude o f the villagers, the health providers are rarely blamed or held
accountable for poor practice. Bhuiya (1992) pointed out that understanding o f the causes o f
diarrhea was quite limited among providers, and that only 60% o f the providers used oral
rehydration solution (ORS) for treatment, despite working in an area where ORS had been
actively promoted for decades. H e also pointed out that the allopathic providers were more
likely to provide appropriate care for diarrhea than the traditional ones. Ina similar vein, Ali and
colleagues (2001) found that access to allopathic providers (both qualified and unqualified) was
related to lower childhood deaths due to pneumonia, whereas access to traditional providers was
related to higher mortality.
In a background survey for this study, different types of private health providers were asked
about how they would handle specific medical conditions (Figure 1.4) (HEU/IHE/NIPORT,
2003). With one exception (a newbom with pneumonia should be treated with antibiotics at a
hospital, may not require a referral from a doctor who practices at a hospital), the correct medical
care would include an affirmative answer to each o f the responses shown at the bottom o f the
figure. The results indicate a very low level o f quality of care by the APPs in absolute levels,
and in comparison to private hospital doctors. Among the APPs, the allopathic drug vendors
performed better than the homeopaths and traditional providers in cases of childhood diarrhea
12
and newborn pneumonia, but similarly badly for the appropriate investigations for suspected TB
or management of a women with post-partum bleeding.
Figure 1.4 - Percent of Different Types of Providers Offering Appropriate Medical Care
for SpecificMedicalConditions
120%
100%
80%
60%
40%
20%
0%
ORS for childhood Referralfor newborn Sputum test for PhysicaVpelvic exam Ergometrinefor post-
diarrhea pneumonia suspected TB for post-partum partum bleeding
bleeding
1 Hospitaldoctors Allopathic drug vendors 0 Homeopaths 0Traditional practitioners 1
Source: HEU/IHE/NIPORT, 2003 and authors' calculations
1.7 HealthFacilities
Information on the numbers and types o fprivate health facilities inBangladesh i s quite weak. In
the public sector, there are precise counts on the locations and types o f health infrastructure,
though little i s reported on how functional they are (Le. staffed, equipped, and seeing a full
complement o f patients). Because most private practitioners operate out o f small clinics or their
homes, these types o f doctor's chambers are hard to enumerate, monitor, or regulate.
The last census o f private health facilities was undertaken in 1997-98 (BBS, 1998). It focused
only on facilities that had inpatient beds or provided laboratory services (Table 1.5). The data
suggest that government inpatient facilities comprise about half (51 %) o f the total inpatient
facilities, but a much larger majority o f the hospital beds (72%). Just examining the number o f
registered private facilities would also seriously underestimate the total number o f private
facilities: only about 70% o f the private inpatient clinics that were enumerated had been
registered with the government. The majority o f the other facilities had no type o f registration,
though some had obtained various types o f licenses from local authorities. Inanother analysis o f
13
private clinics and hospitals in Bangladesh, the Health Economics Unit (1998a) found that only
27% o fthe 252 clinics sampled hadbeen registered.
Table 1.5 -Estimated Number of Government and Private Health Facilities in 1997-98
Source: BBS, 1998. The Health Economics Unit extrapolated from their sample to estimate that there were 584
private hospitals nationally (HEU, 1998a), compared to 613 reported here. NA not available - laboratories inthe
-
public sector are nearly always part ofpublic hospitals
The largest gap concerning private health facilities is the absence o f estimates on private
outpatient clinics (or doctor's chambers). As noted above, APPs tend to work out o f a building,
often their own home or place in the market. In many cases, it may be difficult to distinguish
their places o f practice from a pharmacy. But many public sector physicians and private MBBS
doctors also have outpatient clinics, about which very little i s known.
Giventhe limited information on even the number of health facilities inBangladesh, particularly
for outpatient clinics, it i s obvious that there will be even less information available about their
quality, efficiency, or pricing.
The HEU study (1998a) on 252 private medical hospitals provides a rare insight into their
operation. Bedoccupancy rates averaged only 56%, rangingfrom 12.5% to 97.5%. The average
length o f stay was 5 days, compared to about 7.2 days in the comparison government owned
hospital. However, information on the case-mix o f patients i s not known, so the value o f
comparison i s limited. The study estimated that average return to capital was 38%, suggesting
that private hospitals were making profits much larger than would be available in other sectors.
The HEU study concluded that because facilities with a bed size o f 11-20 beds had the lowest
average costs, this was resultingina large number o fprivate facilities o fthis size.
A background study o f private hospital and clinic managers reported on some o f their main
constraints (HEU/IHE/NIPORT, 2003). As shown in Figure 1.5, their main concerns are with
motivating and recruiting qualified staff, problems with infrastructure, ability to purchase
supplies, and uncertainty over government policies and regulations. Problems with electricity
supply were universally reported as the main infrastructure problem. Incontrast, issues related
to obtaining credit, or with having a low demand are far less o f a concern to them. It may be that
barriers to entry related to obtaining high quality personnel and supplies, and the difficulty o f
managing infrastructure problems, allows the existing private hospitals to make substantial
'Includes facilities that have obtained licenses from local authorities, comprising 8% o fthe total
Nursing homes have inpatient services but no outpatient services
14
profits in spite o f the relatively high number o f beds unoccupied. Caution i s needed in
interpreting these findings, as the surveywas conducted on existing facilities. Ifit were possible
to study cases inwhich private hospitals were sought to be established by investors but failed to
materialize, or had started but failed thereafter, the list o fbarriers to entry mightbe different.
Figure1.5 -PrivateHospitalManagers' Opinionson the BusinessEnvironment
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Difficult to motivate staff to do well
Problems of infrastruct
Difficult getting qualified s
Availability of suppl
Regulations (non-bank
Uncertainty about govt poli
Availability or cost of spare p
Getting space for hospital-cl
Taxes and tax admini
Too little demand for s
Problems with getting
Level of interest rates to
Too little financing from Gove
Problems with money transfer
Source: HEU/IHE/NIPORT, 2003 and authors' calculations
Turning to laboratory facilities, it i s clear from the number o f people working as laboratory
technicians in the private sector (over 29,000 in 1996-97) that the number o f registered
laboratories under-estimates the number o f facilities providing laboratory services by a large
margin. Given that the average laboratory on the average employs somewhat less than two
laboratory technicians, there would be about 16,000 laboratories inBangladesh, compared to the
1,042 identified inthe 1998 survey (BBS, 1998).
Data from the private health service establishment survey found that on average, laboratories
performed about 28 tests per day, and had an annual operating profit (total income - total
expenditure) o f Tk. 404,000 (US$ 8,600) (BBS, 1998). The market i s currently allowing for
highprofit margins on nearly all types o ftests. The operatingprofits onnineteen laboratory tests
examined in the same study was 175%, ranging from a low o f 45% for plain x-rays o f the
abdomen, to 400% for mammography. Overall, the largest profits were being derived from
blood sugar tests and hemoglobin examinations, and the least on x-ray procedures. It i s possible
that equipment costs and technical ability to operate the machines are creating the conditions
where laboratories can charge highprices relative to their costs.
15
In contrast to provider and hospital medical services, the quality of laboratory services is
relatively more measurable, using standard sample selection techniques andreference laboratory
standards. Quality assurance procedures should be easier to establish to assess the quality o f
laboratory test results. However, systems to monitor the quality o f laboratory services do not
exist inBangladesh for all but a few laboratories that do so on their own, and largely for research
purposes. The result i s that the validity o f most laboratory test results i s unknown.
1.8 Other PrivateHNPInputMarkets
Whereas a detailed assessment o f the other input markets i s beyond the scope o f this study, it i s
important to recognize the scale and influence o f these components o f the private health sector.
For example, inthe pharmaceuticals market, there were 767 licensed drugmanufacturingunits in
1997, 1,353 drug wholesale trading firms, and 33,975 retail trading firms (BBS, 2002a). Locally
produced drugs were valued at US$ 312 million in 1997, accounting for 1.0% o f GDP. About
40% o f non-clinical family planning methods were distributed through private pharmacies in
1989 (BBS, 2002a). Spurious drugs are a serious problem.
The Social Marketing Company (SMC) is the dominant player insocial marketing inBangladesh
and distributes health related products to 200,000 retail outlets. SMC distributes 170 million
condoms (70% o f estimated use in Bangladesh) and 37 million cycles o f oral pills (30% o f
estimated use) annually. In addition to reproductive health products, it distributes 105 million
sachets of oral rehydration solution (ORs), which accounts for about 60% o f estimated use
(SMC, 2003). ORS was first marketed inBangladeshin 1986, andhas been an importantpart o f
health initiatives to reduce child mortality from diarrhea. The ORS market in Bangladesh has
grown substantially through consumer education and advertising, and now encompasses 15
brands. In addition to providing ORs, SMC also provides training to ApPs on diarrhea
management. SMC i s considering to market zinc as a supplementary dose to persons with
diarrhea, andbelieves that there is social marketingpotential for safe deliverykits (SMC, 2003).
Currently SMC i s subsidizedby the government, which provides the commodities free o f charge
for distribution, and SMC charges the consumer a nominal price. SMC i s 40% self-financed.
Bed-nets, another important health commodity, are imported and sold in the private sector.
However, the market for bed-nets i s currently not very large. Soap i s currently too expensive for
the poor population, who mainly uses ash for hand washing (SMC, 2003).
More detailed studies are requiredto understand the private markets for goods and commodities
relevant to HNP, e.g., bed-nets, infant food formulas, hygiene products, pharmaceuticals,
vaccines, and medical equipment.
1.9 Conclusions
There are clearly large information gaps concerning the supply o f private health services in
Bangladesh. However, enough i s already known to conclude that the biggest portion o f health
providers i s also the most neglectedbypublic policy-makers.
For policy-makers, there are a number o f key recurring themes that run through this analysis o f
the private health sector. One o f these themes is the challenge to improve the quality o f care.
There i s no single best way to address this in Bangladesh, but what i s known from experience
elsewhere is that high level leadership and commitment i s needed. Also critical i s active
16
participation from key stakeholders, including provider groups, government agencies,
community and consumer representatives, and often some independent monitoring agencies. In
the short term, some o f the key steps may include pulling together leaders, testing quality
improvement tools, and assessing performance and developing standards and benchmarks for
different types o f providers. Inthe medium to long term, credible institutions will be needed to
promote professional self-regulation, consumer protection, targeted regulation, and use o f
payment mechanisms and information disclosure techniques to continually improve health
services performance.
Another urgent issue i s to correct the imbalances in the production o f health personnel. There
are too few nurses and doctors, and an over-abundance o f unqualified providers, over whom the
government has little influence. There are also too few female health care providers. Ambitious
plans are proposed to increase the number o f physicians, but given the resource constraints,
alternative sources for the production o f qualified medical care need to be considered.
A number o f strategies could be pursuedto try to retainhealthprofessionals inthe country. One
relatively simple measure to reduce losses to emigration i s to demand reimbursement or a bond
covering the costs o f medical expenses from physicians who obtain visas to work overseas.
Developing more nurses andparamedicals that could take on more o f the primarycare diagnostic
and therapeutic responsibilities that physicians now assume i s likely to be a more cost-effective
strategy than focusing on training more physicians. This could be coupled with admission
strategies that target rural communities. Engaging with the existing APPs also has potential to
reduce the needs for new formally trained healthpersonnel.
Finally, the question o f private practice among public practitioners i s a persistent and pervasive
issue that affects the credibility o f government health services. Partial approaches to this issue,
such as formalizing private fees in government hospitals are likely to exacerbate the problem.
To help the government in taking steps further work i s needed to test the feasibility o f different
options to improve the governance o f public hospitals, change labor relations, develop
contracting o fprivate physicians, and strengtheningmonitoring andreporting o f hospitals.
17
Chapter 2. Demandfor / Consumptionof PrivateSector HNP Services
Overall health service consumption in Bangladesh is low in comparison to other
countries and to levels of need. The use of maternity services is particularly low: e.g.
only 8% of deliveries occur in a healthfacility.
The private sector is used for the overwhelming majority of outpatient curative care,
while the public sector is used for a larger proportion of hospital deliveries and
preventive care. The dependence on the private sector for curative care is even more
true for the poor in Bangladesh; the poorest quintile of Bangladeshi children have a
higher dependenceon theprivate sectorfor acute respiratory infection and diarrhea care
than the richest quintile.
Poverty is a signijkant constraint to health care access and hence, use. The largest
differences between the rich and thepoor arefor medically trained deliveries, antenatal
care, treatmentfor acute respiratory infection, and immunization. Women and girls tend
to receive less medical care than males, with gender bias resultingfrom cultural factors
and the relative lack of female health providers.
New analyses show that nearly all private health spending is at privatefacilities (88%).
I n absolute terms, the richest quintile spends about six times as much as the poorest
quintile on health care, presumably purchasing a higher quality of health service.
The cost of health care often results in foregone medical treatment. The cost of drugs
and transport, and distance to the provider are some of the most important barriers to
health care. Theprivate providers are generally closer and more conveniently located
than public facilities. New studies show that perceptions of provider's experience and
familiarity with the provider are also important reasonsfor selecting private providers.
Further studies are needed on determinants of care seeking behaviors.
More attention is needed to overcome the social,financial, and physical barriers to care.
There are indications that social marketing can help to reduce some of the disparities in
use of services between the rich andpoor. Further experimentation is needed to see how
consumers can influence the quality of care, how health decisions in the home can be
improved, and how thefinancial impact of health care costs can be reduced.
2.1 Introduction
This chapter addresses the issues related to the demand for health services. The chapter begins
by an analysis o f consumption levels for various types of health services, and the differences
between public and private sector consumption. We then tum to the question o f barriers to
access, andwhy people choose to use certain health providers.
2.2 HealthService Consumption
Although data on overall outpatient clinic visits or hospital utilization are not available in
Bangladesh, it i s possible to use Demographic and Health Survey (DHS) data to compare
selected services with other countries (Table 2.1). These data suggest that consumption o f
maternal and child health services in Bangladesh i s quite low. Bangladesh has lower rates of
institutional deliveries and use o f medical services for antenatal care and treatment o f childhood
18
diarrhea. Among 45 countries with comparable DHS data, Bangladesh has the highest rate o f
home delivery (World Bank, 2003). Hospital deliveries and births attended by a medically
trained person are remarkably low in Bangladesh - only 5% o f all deliveries were in a health
facility in 1996/97, compared to an average of 49% for all other developing countries. By the
time o f the 2000-01 DHS survey, only 8% o f deliveries occurred in a health facility, and a
medically trained person attended 22% of all deliveries. In contrast, untrained traditional birth
attendants were at 54% o f deliveries, and a relative (or no person) was at 24% o f deliveries
(NPORT et al, 2001).
Table 2.1 InternationalComparisonsof HealthService Consumption
Average for 45
Indicator Bangladesh India Nepal Pakistan developing
countries
% of children with diarrhea seen 22 61 14 48 39
medically
% of childrenreceiving full 54 35 43 35 51
immunization schedule
% ofbirths where antenatal care is from 26 49 38 26 71
a medically trained person
Of all deliveries, infacility 5 26 8 15 49
Of all deliveries, % attended by 8 34 10 19 53
medically trained person
19
Ensor et al (2002) found that overall levels o f per capita consumption o f the essential service
package (ESP) - which i s targetedtoward the poor, women, and young children - would have to
increase by 40% in order to achieve desired consumption levels. Consumption o f the child
health component o f the ESP would have to increase by 12% to reach desired levels. The same
study found that consumption of maternal health services was particularly low relative to
desirable levels; consumption would have to increase by 122% to reach the desired per capita
consumption level. Two critical questions remain: (1) how can demand for the essential service
package be raised to match the healthneeds? (2) can access to quality essential health services be
met by relyingentirely on the public sector, or should the much larger private sector be used?
2.3 PublidPrivate Shares of Services
Bangladeshis are much more likely to use private providers than public providers for most o f
their ambulatory care services. In a background study on household use of health services
conducted in2003, the preliminaryresults show that o f those who sought care outside the home
for an illness, 87% o f urban residents used private providers, compared to 75% o f rural residents
(HEU/IHE/NIPORT, 2003). Private allopathic providers were used as the first source o f care for
32% of the cases, while pharmacies were used 26% of the time. Another 10% of the cases were
seen at private hospitals, private non-allopathic providers saw 9%, and the public sector was used
21% of the time. The qualified private providers were used twice as often as unqualified
allopathic providers (excluding pharmacies). Since the unqualified allopathic providers vastly
outnumber the qualified providers, and since the outpatient workloads between the two appear
comparable (Chapter l), it is likely that the public are more likely to view their allopathic
provider as qualified to practice medicine, even when they do not have an MBBS degree.
Other surveys show that consumption of outpatient curative services are largely in the private
sector. The 2000 Service Delivery Survey (SDS) found that the public sector accounted for only
21% o f visits in the last month, whereas NGO and for-profit providers covered 30%, and
alternative private practitioners (APPs) had 49% o f the visits (CIET Canada and MOHFW,
2001). The same survey also found that visits to a private provider were more likely to be for
curative care (90%), compared to visits to a public provider (71%). Another study reported that
75% of the first point o f contact for care occurs in the private sector (Sen, 2001). Levin et al
(2001) also found that people used primarily village doctors and traditional practitioners for
health care, and together they accounted for almost two-thirds of the care utilized. A study o f
infant mortality and health seeking behavior in a rural area of Bangladesh found that 90% of
parents sought treatment for their sick children from private providers (Bhardwaj and Paul,
1986). The same study reported that 53% o f sick children received treatment from ApPs; the
majority o fthese were kabiraj, followed by non-qualified allopathic providers.
International comparisons o f the distribution o f the use of public and private health services are
difficult to come by, but data from comparable DHS surveys provide some insights. These
surveys show that Bangladesh stands out for its high level o f dependence on the private sector
for treatment o f childhood diarrhea, ARI, and institutional deliveries (Figure 2.1). Although
other South Asian countries also have a high dependence on the private sector for childhood
curative services, Bangladesh leads the region, and i s second only to Haiti among all 45
countries with comparable data. O f children with diarrhea inBangladesh, 22% were brought to a
health facility. O f those children, 92% were seen in a private sector health facility. Of children
20
with ARI, 33% were brought to a health facility. O f those children, a private provider saw 89%
(Gwatkin et al, 2000).
Bangladesh's position i s even more striking when making intemational comparisons o f obstetric
services. The most significant point i s that Bangladesh leads the world indeliveries that occur at
home - about 91% of all deliveries inthe three years preceding the1999-2000 DHS (NIPORT et
al, 2001). Of those few deliveries that occur in a health facility, Bangladesh still has a
substantial share occurring in private facilities (40%). Public facilities provided more treatment
for obstetric complications (73%) than private qualified allopathic providers (27%), yet private
providers performed a higher proportion o f caesarean sections (56%) than public providers
(44%) (ACPR and UNICEF, 2001). The higher proportion o f caesarean sections in the private
sector may suggest that there i s a problem with supplier-induced demand: patients using private
providers have higher caesarian section rates than would be neededbecause these providers have
an incentive to give more services, a common issue where the private sector i s poorly regulated
(Peters, 2002).
The use of private providers is not restrictedto the rich, as the poor also have a high dependence
on private providers for childhood curative care (Figure 2.2). However, as noted above, there
may be differences in the type o f private provider used, with the poor more likely to rely on
unqualified practitioners, and the better off able to afford qualified physicians.
Figure 2.1 - International Comparison of the Share of Medical Care Used in the Private
Sector For Selected Services
10
0
Bangladesh India Nepal Pakistan Average for 45
developing
countries
Source: World Bank, 2003; based on DHS inthe 1990s (1996197 for Bangladesh)
21
Figure 2.2 - Proportionof Richest and Poorest QuintilesUsingPublic and Private Health
Services
Institutional Deliveries
richest 20%
Institutional Deliveries
poorest 20%
Childhood ARI treatment
richest 20%
Childhood ARI treatment
poorest 20%
Childhood diarrhea
treatment richest 20%
Childhood diarrhea
treatment poorest 20%
I I I I I I II I II I
0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%
Source: World Bank, 2003; based o n Demographic and Health Surveys inthe 1990s (1996197 for Bangladesh)
Figure 2.3 - InstitutionalDeliveries in Private and Public Facilities in Bangladesh and 44
OtherDevelopingCountriesAccordingto Level of Wealth
70
I
60
10
0
Poorest 20% 20-40% 40-60% 60-80% Richest 20%
1~ I
Source: World Bank, 2003; based on Demographic and Health Surveys inthe 1990s (1996197 for Bangladesh)
22
Figure 2.3 shows that there are enormous differences inthe use o fprivate institutional deliveries
between the richest quintile (8.7% o f deliveries) and poorest quintile (0.1% o f deliveries) in
Bangladesh; a rkhipoor ratio of 87. O f all countries with comparable data, Bangladesh has the
greatest inequality in private institutional deliveries, but also ranks behind only Nepal and
Pakistan in terms of inequality of public sector institutional deliveries (World Bank, 2003). In
sum, women have very little access to any institutional delivery care inBangladesh, and the rich
predominantly consume the little that i s usedinboth public and private sectors.
2.4 Determinantsof Consumption/ HealthCare SeekingBehavior
Inthis section, we summarize the data available inBangladesh on determinants ofhealthservice
use, focusing on physical, financial, and social barriers, and users' perceptions o fhealth services.
In a background household survey about health services use/consumption (HEU/IHE/NIPORT,
2003), the preliminary results on 342 children who received care for an illness showed a familiar
dependence on private providers: 43% were seen by a qualified private physician, 30% by an
APP, and 27% were seen at a government facility. The main reasons for their choices are
illustrated inFigure 2.4. These results suggest that knowledge about the provider's experience i s
particularly important in choosing private practitioners, whereas cost and physical proximity are
more important reasons for those who chose a public facility. For those who chose an APP, the
second most common reason was familiarity with the provider. The provider's behavior, and the
availability of medicines were less frequent reasons for selecting a particular provider.
Figure 2.4 - Main Reason for Choosing a Specific Provider for Treatment of the Most
RecentChildIllness
Experience of provioer
c o s t
Proximity
Fam iliantywith provider
Behaviour of provider I
Availabilityof medicine
0 10 20 30 40 50 60 70
Percent
Source: HEUIIHEINIPORT,2003
23
2.4.1 PhysicalAccess
Numerous studies have shown that physical access to health services i s an important determinant
o f consumption in Bangladesh. Location is one o f the most important factors to determine the
access to health services in Bangladesh as documented in the CIET baseline survey (CIET
Canada and MOHFW, 1999). Geographic access at least partially explains why consumption
rates are higher inurban areas compared to rural areas (NIPORT et al, 2001).
Physical access i s a barrier to maternal and child health services inparticular. Inthe 1999-2000
DHS, 79% o f women reported that the lack of a health facility nearby was a constraint to
consumption (Streatfield et al, 2001). Inthe same survey, 50% o f women respondedthat getting
to the health facility was a problem to them. Levin and colleagues (2001) confirmed the
significant negative association betweenboth distance to the provider and travel time and the use
o f health services. A child was less likely to be taken to a qualified allopathic provider or a
traditional practitioner than a village doctor, ifthe travel time was 40 minute or greater compared
with travel time o f 15 minutes or less. Other research has shown that a majority (74%) o f sick
children in a rural area o f Bangladesh were taken less than two miles for treatment; and that a
majority o f those children were seen by APPs. In contrast, children who were taken more than
two miles for treatment received health care from qualified allopathic providers (Bhardwaj and
Paul, 1986).
These findings strongly suggest that the distance from the household to the health provider i s an
important factor in determining whether or not to use a health provider. Because o f the
importance o f physical access, the large number and close proximity o f APPs to most rural
Bangladeshis give them an advantage over other providers inthe public andprivate sectors.
2.4.2 Financia1Access
The cost o f health care can be a strong determinant of health care use, as well as a cause o f
poverty. Ability to pay i s a particularly important determinant o f access when a highproportion
o f health care is financed privately, and without any type o f financial risk protection from health
insurance. In Bangladesh, 60% o f total health expenditure in 2000 was in the form o f out-of-
pocket payments by individuals (64% o f total health expenditure was from private sources), so
that households' ability to pay for care is important (WHO, 2003a). There i s essentially no
social security or private health insurance, although public hospitals are intended to provide a
form o f insurance incase o f serious illness.
Different types o f cost items can be barriers to the use of health care. Health care costs can be
divided among direct medical costs (e.g. medicines and service fees), direct non-medical costs
(e.g. transportation costs) and indirect costs (e.g. traveling and waiting time, lost earnings). In
the SDS, the cost o f medicines was the most important cost element that prevented people from
using health services, followed by transportation costs (CIET Canada and MOHFW, 2001).
Even ifthe cost o f medicines could be reduced, such as through non-profit organizations that are
able to purchase drugs in bulk and distribute essential drugs more efficiently, other approaches
may be required to deal with patient transport costs, and with gender bias inhousehold decision-
making.
A background study analyzing the BangladeshHousehold Income Expenditure Survey (Peters et
al, 2003), not surprisingly, found that 88% of private health expenditures are for services from
24
private providers, and that the rich pay far more for medical care than the poor (the
richestlpoorest quintile ratio was 6.2) (Figure 2.5). As a proportion o f total income, health
spending by households was found to comprise 6.8% o f their total income overall, with the
richer groups spendinga slightly higher proportionthan poorer groups.
It is noteworthy that even though the poorer people spend less per episode o f illness, a greater
proportion o f that expenditure goes to private providers. Thus the poor are in fact more
dependent on private care, contrary to a common beliefthat private care is for the rich.
Figure 2.5 - Household Per Capita Payments to Private and Public Health Providers in
Bangladeshby IncomeQuintile
1407
Poorest 2040% 40.60% 60-80% Richest 20%
Income Quintiles
Source: BBS, 2001 and authors' calculations
Total household payments to health providers i s estimated according to the results o f the HIES
(2000) in 1.5 billion dollars which includes fees, hospitalization expenses, medicines, laboratory
and tests, transportation costs, tips and others charges.
Gwatkin et a1 (2000) disaggregated the Bangladesh 1996-97 DHS data by wealth quintiles based
on assets o f the household, and then reported the use o f health services (Table 2.2). The largest
differences between rich and poor were found for medically trained deliveries, antenatal care,
treatment for ARI, and absence o f any childhood immunizations. For example, they found that
the poorest 20% o f children with ARI are less likely (23%) than richest (51%) to be taken to any
medical facility for treatment. There were far smaller differences in the treatment o f childhood
diarrhea and use o f contraceptives. This may be due to the fact that in Bangladesh there have
been decades o f social marketing and public programs to support family planning and oral
rehydration therapy, and to some extent childhood immunizations.
25
Table 2.2 - Wealth Differences in Health Service Consumption
Indicator Poorest Richest Population Rich/Poor
20% 20% Average Ratio
Proportion o f surviving children (12-23 months)
who received all vaccinations (BCG, DPT, oral 47% 67% 54% 1.4
polio, measles)
Proportion o f survivingchildren (12-23 months)
who receivedno vaccinations 18% 5% 12% 0.3
Proportion o f children (under five) with diarrhea
inlasttwo weeks who receivedORT 62% 68% 61% 1.1
Proportion o f children (under five) with diarrhea
inlast two weeks who were taken to any medical 22% 24% 22% 1.1
facility for treatment
Proportion o f children (under five) with AFU in
lasftwo weeks who were taken to any medical 23% 51% 33% 1 2.2
facilitv for treatment II II II
Proportion o f birthsfor which a woman received
at least one antenatal care consultations by a 14% 59% 26% 4.2
medically trained person (doctor, nurse/ midwife)
Proportion o f births for which a woman received
two or more antenatal care consultations by a 9% 51% 20% 5.7
medicallv trained person
Proportion o f births attended by a medically
trainedperson 2Yo 30% 8% 15.0
Proportion o f married women who report use of
modem means o f contraceDtion 39% 49% 42% 1.3
Proportion o f married men who report use o f
modern means o f contraception 46% 54% 49% 1.2
Source: Gwatkin et al, 2000; Bangladesh data based on DHS 1996197
Figure 2.6 - Proportion o f Household Expenditure Spent on Healthby Income Quintiles
7.4% , I
7.2%
6.6%
L
$
sf 6.4%
-
P
Y 6.2%
2
.d
6.0%
L
E 5.8%
0
n 5.6%
5.4%
5.2%
Poorest 20% 20.40% 40-60% 60-80% Richest 20%
Income Quintlle
Source: BBS, 2001 and authors' calculations
fry pri~sttcattd public pro
27
Figure2.9. Healthconsumptionof the poorestthe quintileof income.
NGOs and others-
6% Governmentproviders
Traditionalproviders
3%
Private providers
37%
Pharniacyidisp
45%
Figure2.10. HealthConsumptionof the richestquintile of income.
NGOs and others-
4% ,
Traditional DrOvlderS ~ Government providers
Pharmacyidisi
23%
28
Table 2.3 Total householdpaymentsto healthprovidersinBangladeshby incomequintiles
Quintiles Fees Hospitalization Medicines Lab. And Transport Tips Other TOTAL
test charges Monthly total
household
payments
(millions tks and
percentage).
Poorest quintile 5.9% 2.5% 9.4% 3.6% 5.2% 4.0% 4.7% 562 (7.5%)
Second 10.4% 2.1% 16.0% 5.5% 9.7% 22.2% 8.1% 958 (12.8)
Third 13.2% 7.8% 19.3% 14.1% 18.3% 5.3% 14.2% 1,274 (17%)
Fourth 23.3% 9.8% 22.2% 16.3% 24.1% 26.8% 17.3% 1,550 (20.7%)
Richest quintile 47.2% 77.7% 33.1% 60.5% 42.7% 41.7% 55.6% 3,126 (41.9%)
Total
Monthly total 680 524 4,712 681 419 50 405 7,476 (100%)
household
payments (100%) (100%) (100%) (100%) (100%) (100%) (100%)
(millions tks and
percentage).
29
qualified allopathic providers and homeopaths provided most o f the health services for female
children (Bhardwaj andPaul, 1986).
The information on gender differences inhealthspending is not as clear-cut. One study o fpublic
sector health expenditure found that on average, more money was spent on males compared to
females (Ensor et al, 2002). However, analysis o f the Bangladesh HIES (Peters et al, 2003)
shows that innearly all income quintiles, spendingon women i s greater than on men or children
(Figure 2.11).
Figure2.11 -Average AnnualAmount (in Taka) of Private Out Of PocketSpendingon the
Health of Women, Men, andChildrenAccordingto IncomeQuintile,Bangladesh,2000
'O 1
60
50
40
Y
m
I-
m
30
20
10
0
Poorest 20% 2040% 40-60% 60-8O% Richest 20%
Income Quintile
1 a Women Men 0 Children
Source: BBS, 2001 and authors' calculations
2.4.4 Patient Perceptions of Providers
Although there is relatively little data on patient or client satisfaction in health care in
Bangladesh, what i s available suggests that clients are more satisfied with private than public
providers. In the 2000 service delivery survey, 41% of households reported that they think
public sector health and family planning services are "bad", whereas only 10% of households
thought the same about private providers. Conversely, only 10% o f households thought that
public health services were "good", while 25% thought the same about private providers (CIET
and MOHFW, 2001).
Table 2.4 summarizes the main reasons for dissatisfaction with public or private health and
family planning services (CIET Canada and MOHFW, 2001). Poor staff attitudes and problems
with the lack of medicines or their quality were the biggest problems with the public sector.
while costly medicines, and lack o f doctorshurses were cited inthe private sector.
30
Table 2.4 -CommonlyIdentifiedProblemsin Public and PrivateServices
Proportionof Respondents
Problem PublicFrivate
Public Private Ratio
Lack of, and poor quality of, medicines 58% 18% 3.2
Bad service 40% 33% 1.2
Bad staff attitude 25% 6% 4.2
Have to pay for medicines, expensive
medicines 17% 41% 0.4
'Lack
Lack o f doctors, nurses, specialists 14% 32% 0.4
o f different services 14% 24% 0.6
Dirtyandpoor equipmentand facility 13% 8% 1.6
Doctors not available 13% No data --
Extrapayments to doctors and other
workers 12% 11% 1.1
Too few beds, lack o f facilities 7% 7% 1.o
Source: CIET Canada and MOHFW, 2001
In the background study 'on household use of health services, satisfaction with child health
services was also assessed for 342 children who were treated for an illness outside the home
(HEU/IHE/NIPORT, 2003). As showninFigure2.12, there was little difference inoverall levels
o f satisfaction between users o f different types o f providers.
Figure2.12 Levels of Satisfactionwith ChildHealthServices, by Type of Provider
. . . .. . . , . .. . . . . . 1" " "
P u b l i c Facility
Private Qualified D o c t o r
Source: HEUIIHEMIPORT, 2003
31
Figure2.13 Satisfactionwith ChildHealthServices: MostImportantFactor Accordingto
Type of Provider
1' . . . , . .. . . . , ,
. .. ....... .... ...I......,,.. .. .... . .. . . ... .
I
I
Courtesy of provider
Perce veo nealtn
0-tcom e
Explanation about
treatment
Cost for services I
I I
Cleanliness II I
I
I I
Consultation time I I
I I
I
0 5 10 15 20 25 30 35
Percent
Source: HEUIIHEINIPORT,2003
Other studies also provide insights into how satisfaction with health services affects
consumption. For example, 54% o f female respondents in the 1999-2000 DHS reported lack o f
confidence inhealth services as a reason for not using those services (Streatfield et al, 2001). In
the SDS there were differences in levels o f satisfaction between types of providers. The
proportion o f users satisfied with overall services was 62% for public sector services, 88% for
qualified allopathic providers (for-profit and NGO), and 88% for APPs (CIET Canada and
MOHFW, 2001).
2.5 Conclusions
This chapter confirmed that Bangladeshis overwhelmingly use the private sector for their health
care, though much o f this care i s obtained from unqualified providers and other APPs. In the
case o f maternal health services, there are remarkably low levels o f access, particularly for
delivery and antenatal care.
The data also show that there are considerable gaps in information concerning the demand for
health care that policy makers need to be aware for the design and implementation of policy
options. There i s little experience in how to influence demand for higher quality o f health
services. Family planning and use o f oral rehydration therapy, which have benefitedfrom long
public health education campaigns, are associated with relatively small disparities between rich
and poor in Bangladesh. But it i s not known how behavior change communications can be used
to influence the type o f care offered by health providers. Interventions are needed to test how to
empower people to demand better quality o f health services from both private and public
providers. The biggest need i s in matemity services and general health services for girls and
women.
32
The dependence on out-of-pocket financing o f health care puts Bangladeshis at risk o f poverty
and foregone medical treatment. This is more so for the poor. Interventions to reduce the
financial impact o f drug and transport costs have not been tested in Bangladesh, and the
effectiveness o f interventions such as subsidies, demand-side financing, or health insurance on
financial protection i s a challenge ina large scale. These ought to be tested.
Taken together, the findings in this chapter emphasize the importance o f people's reliance on
different aspects o f the private sector for their health care. More than ever, there is a need in
Bangladesh to implement actions that take advantage o f the existing private sector to address
basic public health goals of improving quality, access, and affordability o f health services. The
identification o f activities where the private sector has comparative advantage for the delivery o f
health services is crucial to scale-up the process o fpurchasingprivate services with public funds.
33
Chapter 3. Interactionbetweenthe Publicandthe Private Sectors inHNP
Reviewing the interaction between government and theprivate sector in health generates
several key conclusions. First, the range and magnitude of government engagement with
private providers is not congruent with their presence serving the poor. The bulk of
interaction takes place in terms of regulation and with regard to private clinics and
hospitals. APPs have very little interaction with government. Thus, public-private
engagement has largely excluded service providers of greatest importance to the poor.
There are, however, positive experiences, including a number of pilot initiatives, where
thepublic sector haspartnered withprivate, mostly non-profit, health serviceproviders -
- e.g. in nutrition, TBAeprosy, family planning, urban primary health care and
immunizations. I n addition, the involvement of private actors has occurred sometimes in
policy discussions andformulation, though this has not been a consistentfeature.
Second, the government stewardship responsibilities to the health sector could be
enhanced. There is little collaboration withprofessional and provider organizations, or
support for self-regulation. Currently professional and provider organizations are
primarily playing the role of trade unions. Neither consumer nor patients ' organizations
have yet emerged strongly to play an advocacy role, nor to engage in monitoring of
service quality and outcomes. Instruments to engageprivate actors require government
officials toperform tasks very differentfrom their traditional activities.
Third, misperception and low capacity underlie weak public-private engagement. The
policy-makers ' interviews reveal that there is a limited understanding of the private
sector size and role in provision of care, especiallyfor maternal and child health services
in rural areas. Most policy makers believe that the bulk of private provision is tertiary
carefor the rich.
Broad-based efforts are needed to improve quality ofprivate health services and specijk
initiativesfocusing on APPs. I n addition, there is a need for an increased capacity of
the MOHFW and strong engagement of the key stakeholders, through the setting up of a
Public-Private TaskForce in the MOHFW
3.1 Introduction
The previous chapters described the size and characteristics o f the private heath sector supply
and the demand and consumption of private health services by the population. This chapter will
focus on the interactionbetween the government o f Bangladesh and the private health sector. It
will attempt to shed some light on the current situation and the existing experiences in the
country, but also draw on some o f the perspectives that governmental actors and private actors
have on the issue of collaboration. It aims at providing and informing the debate on the two
sides and it will conclude by suggesting some options to improve the interaction between public
andprivate sector inH" inBangladesh.
3.2 CurrentSituation
As described inthe previous chapters, it is clear that the private sector, inits many, varied forms,
is the predominant source of health services for the people o f Bangladesh. Health policy in
Bangladesh, as inmany developing countries, has traditionally focused on the public sector, and,
in particular, on administration of public facilities. Notwithstanding the public-sector focus,
34
however, there are a number o f areas in which public and private actors interact. Below, this
interaction i s presented according to the following categories: governmental regulation; self- and
non-governmental regulation; service and facilities planning; information dissemination; disease
surveillance; contracting; and, grants and subsidies.
3.2.1 GovernmentalRegulation
Regulation o f health services i s a central role for governments in health systems with private
delivery. While virtually all such health systems have the basic components o f health service
regulation in place, in many developing countries, there i s a huge gap between the legal
provisions and implementation (Afifi et al, 2003). This i s also the case in Bangladesh.
Regulations are in place relating to most key inputs for health services, including: premises,
equipment, and education and licensing o f medical and health workers. Regulations also exist
with regardto pricing o f goods and services. The Directorate o f Health Services, headed by the
Director-General o f Health Services (DGHS) i s responsible for implementation o f most
regulations. An autonomous government agency, the Bangladesh Medical and Dental Council
(BMDC), is responsible for undertaking and enforcing registration, as well as approval o f
curriculum for medical education programs. Another autonomous agency, the Bangladesh
Nursing Council (BNC) recognizes nursing and allied health traininginstitutions.
The standards specified in the laws are significantly out o f date, and therefore not applicable to
the current status and state o f development in the health sector. For example the provisions for
maximum fees for surgical operation, normal deliveries and diagnostic tests were established in
1984, andhave neitherbeenupdated nor revoked since then.
The effectiveness o f regulation is very limited. At the most fimdamental level, major constraints
appear to be insufficient capacity and attention to regulatory issues. There are too few human
and financial resources devoted to regulation. It is evident that the capacity o f the govemment
bodies to implement this legislation i s weak, or non-existent insome cases. While registration i s
formally required to practice, the vast majority o f providers practice without being registered
(Health Economics Unit, 1998ab). Inspections for monitoring o f service quality are not possible
with available resources. There are nearly 85,000 healthpersonnel under the Director General o f
Health Services, hence it i s not an absolute shortage o f humanresources that explains the weak
capacity. Rather too few staff are assigned to these activities, and there i s a shortage o f
personnel with relevant training.
Even a substantial portion (30%) o f inpatient facilities, which are fewer innumber and easier to
identify, operate without registration (BBS, 1998). Inspection o f health facilities occurs only
when a license i s initially given. Subsequent re-licensing i s done based on self-reported
information.
Out-of-date standards and rates established in regulation seriously undermine compliance. The
fact that the standards or values set inthe law are so incongruent with reality inprivate practices
andclinics, diminishes the leverage andpotential influence o fthe regulations.
35
3.2.2 Sew- and Non-governmental Regulation
The role o f professional and provider groups is another critical element o f a well-hctioning
regulatory framework. Invirtually all well-performing health systems, these non-governmental
bodies undertake extensive quality assurance activities, with varying degrees o f oversight by
government (Afifi et al, 2003). The technical expertise and credibility o f these bodies makes
them invaluable partners indeveloping standards for education andpractice. They often play a
substantial role inimplementationo fregulation as well.
There i s a substantial number o f professional and provider organizations inthe health care field
in Bangladesh (see Box 3.1). Currently, these bodies undertake few self-regulation activities.
To date, the organizations function primarily as trade bodies, that is, they focus their activities on
protecting the interests o f their provider-members as businesses. The government does involve
representatives from these organizations in some policy initiatives, for instance consultation on
proposed legislation. However, it generally does not involve professional or provider
organizations in implementation o f policies. The degree o f governmental support and oversight
for self-regulationbyprofessional or provider bodies i s likewise low.
Consumer or patient organizations are not developed, nor i s a role established for them in
regulation or government activities. Victims o f malpractice or mistreatment can only make an
appeal to the Director General o f Health Services, an event that i s fairly uncommon. When
asked, private providers confirmed that no other non-governmental groups strongly influence
their environment (HEU/ME/NIPORT,2003).
36
Box 3.1 -PotentialSelf-Regulatory Bodies in Bangladesh
The Bangladesh Medical Association (BMA) is the main representative organization o f the medical
profession. Its formal objectives include improvement o f medical sciences and enhancing the status and
honor of physicians. Inpractice, its main concerns relate to protecting the interests, rights and privileges
of its members. Physicians working both in the public and private sectors are members o f the BMA.
Many other professional societies of special medicallsurgical disciplines are affiliated with the BMA.
Theprofessionaljournal ofBMAi s publishedirregularly.
Private medical practitioners have an association of their own, namely the Bangladesh Private Medical
Practitioners 'Association (BPMPA). Many of its membersare also members ofthe BMA.
Theprofessional association o fregisterednurses is theBangladeshNursing Association.
Similarly, the Pharmaceutical Society of Bangladesh (PSB) is the apex professional association of
graduatepharmacists. Diplomapharmacists have a separateprofessional society.
Private pharmaceutical manufacturers (mainly the larger ones) have their own organization, the
BangladeshAushad Shilpa Samity.
The owners of retail pharmacies have a society of their own, the Bangladesh Chemists and Druggists
Samity.
The owners o f private clinics and laboratories have recently formed their own association inthe name of
Bangladesh Private Clinic and Diagnostic Owners' Association (BPCDOA). It has a nine-member
committee to fix more or less uniform charges and to address quality of services issues. Like the other
professional associations mentionedhere, this entity functions more or less as a trade organization.
Source: Hve. 2003
3.2.3 Planning
In mixed-delivery health systems, government service and facility planning must take into
account the existing private sector capacity, as well as their plans for development. Otherwise,
the impact o f public sector construction and service development is reduced through duplication
and"crowding out". Service and facilityplanning inBangladeshis undertakenby the MOHFW.
Currently, these planning activities are based only on public facilities and services. Some private
hospitals and clinics evidently receive information from the government as to planned
development (see Table 3.1). However, information about private capacity appears not to be
taken into account ingovernment facility andinvestment plans.
37
Table 3.1 -Public/PrivateSector Interactionby Type of Provider
Source: HEU/ME/NIPORT, 2003 and authors' cal~ulations.~
3.2.4 Policy Making
Inmixed-delivery health systems, private providers are typically involved, through a range of
mechanisms, inthe policy makingprocess. Most often this involvement i s formal, and provider
representative bodies have an official "seat at the table" in deliberations regarding policy
development. InBangladesh, representatives from the more formal and organized professional
and provider groups (allopathic qualified physicians; private clinics/ hospitals) are sometimes
involved inpolicy development and implementation in the health sector, especially where such
policies affect the private sector. Typically, these representatives serve as members o f formal
government appointed bodies or committees. Many components o f the service delivery sub-
sector, such as traditional healers, homeopaths, pharmacists, are less organized and more
atomistic. Not surprisingly, neither they nor their representatives are involved in the policy
making process. Uncertainty about govemment policies towards the private sector was
mentioned as an important constraint to their operations by a large majority o f the private
providers interviewed(see Table 3.1).
3.2.5 Information Dissemination
Information dissemination activities are a critical part o f a govemment's stewardship in the
health sector. Such activities constitute a substantial part o f government activities in the health
sector in successful mixed-delivery health systems. Governments usually partner with self-
regulatory and consumer or patients' organizations inimplementing these activities. Information
disseminated to consumers usually relates to appropriate health-seeking behavior and what
constitutes high-quality health care. This knowledge guidespatient consumption andpreferences
toward higher quality care and providers. Better informed patients are also less likely to pressure
providers for unneeded, even harmhl interventions - pressure to which private providers are
known to be responsive. Information dissemination to providers about policies, regulatory and
planning activities promotes compliance and reduces uncertainty in private investment and
development. Information dissemination to providers increases knowledge, and improves
coordination and clinical practice.
The percentages were calculated among the private actors who provided an answer to the question.
38
InBangladesh, government information dissemination activities appearto reachalimitedportion
o f health providers. From the provider survey, 40.6% o f hospital or clinic managers receive
some information about health sector planning for example. A smaller proportion o f clinicians
(26.8%) indicated that they receive similar information. Beyond the formal provider group
however, virtually no information is received from government (see Table 3.1 above). With
regard to information dissemination activities to ,consumers, though very little information is
available it appears that such efforts are limited to special programs (family planning; child
health; TB) and, with the exception o f family planning, are associated with donor-driven
initiatives.
3.2.6 DiseaseSurveillance
Disease surveillance i s a central responsibility o f government in the health sector. For
surveillance systems to be effective, it i s essential that all providers o f health care actively
contribute information. It is equally important that all providers, both public and private,
participate in programs to appropriately identify, treat and/or refer patients. Currently in
Bangladesh there i s no systeminplace for reporting o f infectious diseases from private sector in
the main surveillance system. Private providers are not legally required to report disease patterns
nor the number o f cases treated. The exception to this situation is the reporting o f tuberculosis
cases inthe context o f the government collaboration with NGOs (see Box 3.2).
39
Box 3.2 -Government-NGOCollaborationto Implement the NationalTB Program
In 1994, the Government signed two Memoranda of Understanding separately with BRAC and the
Leprosy Coordinating Committee, the latter being an umbrella of NGOs working on leprosy. Both
agreements, renewed on annual basis, state the principles o f collaboration in implementing the National
Tuberculosis Program. NGOs' resources are channeled to the provision of standardized TB services, to
rural and less accessible areas, through the use o f community health workers. The results are very
encouraging. The BRACprogram achieved highrates of case detection and treatment compliance, with a
cure rate of at least 85% (Chowdury et al, 1997). Inaddition, when compared to the government, the use
of CHWs was found to be more cost-effective in rural Bangladesh. With the same budget the BRAC
programcould cure three TB patients for every two cured inthe government program (Islam et al, 2002).
The successful collaboration in TB control in Bangladesh has been achowledged internationally
(Kumaresan et al, 2000) and could serve as a model for other programs in the country. Crucial to the
success seems to be the clear identification of roles and responsibilities o f the different actors. The
division o f responsibilities i s presented in the following table, and reporting o f tuberculosis cases i s
clearly mentionedunder the NGOs' responsibilities.
Area ofCoIIaboration Government 1NGOs
Source: Barkat et al, 2003
3.2.7 Contracting
Engaging in long-tenn purchasing arrangements with private providers i s the most influential
instrument used to guide independent service providers in mixed-delivery health systems. The
interest o f private providers in a predictable revenue flow yields the government purchaser
strong leverage over their operation. Health services contracts are typically longer-tenn (relative
to other public services), and subject to on-going consultation and coordination. Therefore,
health services contracting both requires and generates frequent communication and interaction
with private providers, and, as such constitutes an important component of a government's
overall interaction with the private sector (Taylor, 2003).
InBangladesh, a number ofpilots have beeninitiated to engage NGO providers via contracting
to provide health care services. These experiences constitute a rich resource for evaluating and
improving the interaction between the GOB and private service providers. To date, most of
these public-private partnership initiatives have been implemented in conjunction with donor-
supported initiatives. Recently, the MOHFW has declared its intention to contract out NGOs in
350 unions to provide essential service delivery.
40
3.2.8 Grants and Subsidies
In some health systems, governments support private service providers via grants or subsidies.
Since the hnds are not formally linked to any service or output, such arrangements are usually
applied to NGO service providers, whose activities are linked to government's sector objectives.
InBangladesh, there are some instances of such support to NGO providers (see Box 3.3). As
with contracting however, these experiences are taking place mainly in connection with donor
supported programs and pilot activities.
Indicators Numbers Percent National
Women who received at least two
antenatal care (ANC) consultations out o f 28959 74.0 30.3
total deliveries
Deliveries conducted by qualified person
out of total deliveries 14439 36.9 NIA
Women receiving at least one postnatal
care (PNC) consultations within 42 days 19886 62.1 NIA
o f delivery
1 Contraceptive Acceptance Rate (CAR) N / A 63.O 53.8
Neonatal death I I I 1
I 968 1 25.8/1000 I 50.4/1000 I
death I 1.9/1000 4.34000
Source: Barkat et al. 2003
3.3 Perceptions of the Public Sector towards Working with the Private Sector
The information described below was collected during the semi-structured interviews conducted
with a total of 21 policy makers and managers inthe government system at central level to elicit
their views o fpolicy making inthe health sector, especially withregard to public/private mix o f
services and the overall role o f the private health care sector in achieving national health goals.
41
Mid-level government officials -- those directly in charge o f implementation o f health sector
policies -- were interviewed to obtain an understanding o f their views on private provision
(Forsberg andAxelsson, 2003).
Many o f the persons interviewed felt that there i s not a clear andwell-communicated government
policy towards the private health sector. However, it was said in some o f the interviews that the
government has recently decided to contract out some health services to private actors. More
active supervision o f the private for-profit providers in Dhaka district has also taken place
recently following an initiative taken by the new government to improve services in the private
sector.
Senior officials tended to describe the private health sector as providing tertiary care to well-off
people. The widespread use o f private providers at other levels and by all socioeconomic
groups, as documented in surveys, was not put forward by most respondents. The fact that
publicly employed doctors often work in the private sector and therefore lose focus on their
duties inthe public sector was often described as a problembythe persons interviewed.
3.4 Perceptionsof the Private Sector
As part o f the policy-makers interviews, four privatehealth care providers were also interviewed
to provide a private sector perspective on government health sector policies and the interaction
between the private and the public sector. In addition, during the provider survey several
questions elicited their perspective o f the interaction with the government (HEU/IHE/NIPORT,
2003).
Private providers in general perceived that there was little interaction between themselves and
government programs and activities. Again, the informal andnon-allopathic providers perceived
the least interaction. Nevertheless, all groups of providers expressed strong willingness to work
with the government and the MOHFW. The majority expressed an interest inworking with the
government to expand their involvement in health promotion activities - an area where private
providers are typically perceived as weak. Inaddition, the private clinic managers indicated that
training o fprivate staff would be a usefil support from the government (see Table 3.2).
While there was a surprisingly large amount o f interest inworking with the Government, private
sector representatives expressed some reservations, which they attributed to corruption and lack
of accountability infinancial management. "Support by donors to health care provision could go
directly to private actors. When finds go through the government there is corruption and too
much diversiodleaking o f funds." Or "One major reason why the public sector is not a good
choice for provision o f care i s that the accountability o f the government is poor. The quality o f
public services therefore becomes very poor."
42
Table 3.2 -Public/PrivateSector Interaction:the Perceptionby Type of Provider
wouldbe usefbl support from NIA NIA NIA (87.5%
govemment
lo
The percentages were calculated among the private actors who provided an answer to the question.
43
access to quality services for these population groups. Much has been learned from these
experiences about the opportunities and challenges o f "hamessing" the private sector to meet
health goals in the Bangladeshi context. For example the collaboration between the national
tuberculosis program and large NGOs could serve as an example for other programmatic areas.
To date, however, few o f these pilot initiatives have been expandedor "scaled up". Buildingon
these initiatives, to apply "lessons learned" and to reach greater numbers o f people remains an
important unused opportunity. Encouragingly, recent statements by the GOB indicate increasing
awareness o f and openness to working with the private sector and in particular with NGOs to
provide service delivery at the unionlevel.
3.52 Public-Private Engagement Excludes Providers of GreatestImportance to thePoor
Ingeneral, the bulkofinteractionbetweenthe government andprivateproviders takes placewith
regard to private clinics and hospitals. Less formal, less organized providers, such as non-
allopathic practitioners, including traditional birth attendants, and drug vendors and retail
pharmacists on the other hand have very little interactionwith government. This i s unfortunate,
as the poorer and more rural populations tend to utilize the latter. While there are multiple
factors that contribute to this situation, a stakeholder analysis o f the relevant political economy
may be helpful to identifyresistance points anddevelop appropriate solutions.
3.53 Misperception and Low Capacity Underlie WeakPublic-Private Engagement
A number o f factors undoubtedly contribute to the current poor state o fpublic policy toward the
private health sector.
0 There appears to be limited knowledge o f the role (instruments) o f the government in
guidingthe private sector towards contributing more towards social objectives related to
access and service quality, inparticular.
0 The policy-makers' interviews revealed that there is a limited understanding, indeed,
even misunderstanding o f the private sector size and role in provision o f health care, in
particular to the poor in rural areas. Most policy makers believe the bulk o f private
provision i s concentrated in tertiary care in the capital, overlooking the significant role
altemative private providers play in serving the poor and especially for maternal and
child health services inrural areas.
Instruments to engage private actors (providers and representative organizations) are inherently
complex to operate. They also require government officials to perform tasks very different from
their traditional ones related to administration o f public facilities. There i s currently very little
capacity to operate such instrumentsinthe MOHFW or inlocal government bodies.
The bulk o fhealth services are provided by altemative private practitioners. This segment o fthe
health services sector is consistently the most difficult to monitor, engage and influence.
Experience from many countries, both developed and developing, reveal that initiatives to
influence service provision inthis segment i s particularly challenging.
44
Chapter 4. Main messagesandpolicyoptions
4.1 Findings
(a) Overall health service consumption inBangladesh (from any source, public or private) i s low
compared to levels o freported illnesses andto levels inother countries.
(b) The poor are far more likely to forego medical treatment (30% o f poor sick people are not
treated), with differences between rich and poor households being the largest for medically
trained deliveries, antenatal care, treatment for ARI, and immunizations. The differences nearly
disappear when comparing the use o f modern contraceptives or oral re-hydration therapy for
diarrhea, two services where there has been extensive social marketing.
(c) Women and girls tend to receive less medical care than men and boys despite the fact that
females show a greater disease burden than males. HIES 2000 indicates that 58% of curative
services are utilized bymen.
(d) Theprivate sector is used for the overwhelming majority o f outpatient curative care, while
the public sector is used for a larger proportion o f hospital and preventive care. This broad
division o f roles cuts across economic strata of the consumers, contrary to a common perception
that private sector caters mainly to the tertiary care needs o fricher populations.
(e) The poor also use private curative health care services and pay for those services with higher
proportion o f their income than the richest households. However, both the poor and the rich
choose private providers insimilar proportion. Inthe poorest households, 81% o f the health care
services are provided by the private sector, and 88% inthe richest households. Also 40% of total
out-of-pocket expenditures are made by the 40% bottom poorest households.
(f) There are major gaps inknowledge concerning the private health sector - the actual numbers
o f providers, the services they provide, the conditions under which they practice their trade, their
incentives and disincentives, etc.
(g) One very well established fact is that a majority o f private providers o f health services in
Bangladesh (referred to in this report as alternative private practitioners, APPs) do not have
formal training and recognized qualifications inallopathic medicine. The poor, especially, make
heavy use o f these APPs. Implementing training programs to drugs shop workers and birth
attendants i s needed to improve quality o f services to the poor.
(h) The bulk ofprivate health service providers are males (the exception beingtraditional birth
attendants), which poses a major problem o f access to their services bywomen.
(i)Although there isrelativelylittleknownabout the quality ofcare ofindividualprivate
providers, or the health facilities they work at, the available information suggests that assuring
technical quality is a signzjkantproblem, particularly among alternative private practitioners.
('j)Amajorityofconsumers, however, reporttobesatisfiedwiththeprivateservicestheyhave
received, and rate them superior to government-provided services. Alternative private
practitioners are given preference over qualified doctors mostly on account o f their easy access.
Availability o f drugs i s also reported to be a key factor in choosing private practitioners
(qualified or not) over government facilities.
45
(k) Inspite ofthe obvious importance o ftheprivate sector, healthpolicy inBangladesh thus far
has focused on the public sector and, inparticular, on administration o f public facilities. There
has been insufficient attention paid to the potential o f using the know how and resources o f the
private sector more systematically to achieve societal goals in health. There have been some
instances such as the collaboration with NGOs (e.g., National Tuberculosis Program, Integrated
Nutrition Project, social marketing o f contraceptives), and a number o f pilot initiatives as in
urban primary health care. Such initiatives have yet to be scaled up and lessons from these
experiences yet to be evaluated andbe applied to HNP program.
(1) Government regulations are in place for many aspects o f health service provision inprivate
facilities but enforcement has been uneven. The main constraints appear to be insufficient
capacity and attention. There has been little collaboration between the government and
professional and provider organizations in ensuring adequate standards, and there is no wide
support for self-regulation. Neither consumer nor patient organizations have yet emerged to play
anadvocacy role nor to engage inmonitoring o f service quality andoutcomes.
Inmanyways, boththe public andthe private sectors have failedto meetthe essentialhealthcare
needs o f the people in terms of both access and quality. Near half o f the users of the publicly
financed essential service package are non-poor people. The relevant question i s whether the
performance o f the private sector can be improved to meet the needs o fthe people and to provide
the value for the money they spend for their health care.
4.2 Policy Formulation: A new paradigm
The above messages suggest that there are several areas where a reformulation o f government
policies would be desirable, with a view to helping the attainment o f the social goals in health
included in the government's December 2002 interim PRSP. Given the lack o f resources and
capacity constraints inthe public sector, there is a need to shift the role o f the Government from
provision o f health care services to the purchase o f health services. This new paradigm is
justified by the current dominant place held by the private sector inthe financing and delivery o f
HNP services, and the serious concerns about quality, access, accountability and govemance
with regard to both private and public services. Inparticular, the following policy areas would
appear to deserve priority in government thinking and action to strength the implementation o f
the next Health, Nutrition andPopulation Program.
4.2.1 Under-ConsumptionofHeaJth Careby thePoor and Women
The fact that many patients have a preference for private providers suggests that the observed
under-consumption o f certain essential health services, especially by poor households and by
women, cannot be remedied without increasing the access o f under-served populations to private
providers. The problem is particularly important for matemal health services, especially assisted
delivery by a skilled attendant. The more traditional government approach o f expanding the
supplyo f services provided by government employees out o fpublic facilities is unlikelyto reach
the poor. The gender distributiono fthe providers poses an additional disincentive for women.
4.2.2 Service Quality and Outcomes
Private health services (mostly clinical services) appear to be o f good quality in the eyes o f
consumers. Thisjudgment i s likelybased on those characteristics o f private services that can be
46
easily assessed by patients, such as ease o f access, degree o f courtesyhespect, and means to
obtain both advice and medicines in one place. Much more problematic for consumers i s the
ability to assess correctly the technical quality o f private treatment received or to relate such
treatment to outcomes o f illness episodes (good or bad). To make up for this important
deficiency, and ensure that in most cases consumers derive good value from their purchases o f
privately provided services, deliberate and well-conceived collective action will be required.
Issues o f quality need to be looked at separately for the formal and informal sectors as public
policy interventions would be different for these two groups o f providers. The multitude o f
provider types inthe private sector warrants a mix of differentpolicy options.
4.2.3 ThekhowledgeBase
While various public interventions could be conceived based on what i s now known about the
private health sector in Bangladesh, the large knowledge gaps that exist would magnify the
uncertainty always associated with new policies and courses o f public action. The knowledge
base about private health services needs to be widened to enable the progressive refinement o f
policies and programs, and also to help users to make informed decisions related to the
utilization o fhealthproviders.
4.3 Policy Options
The followingpolicy options were discussed during stakeholder consultations inearly May 2003
with government officials, private sector actors, civil society, the academia, and development
partners. While there was broad agreement on the need to increase the engagement with the
private sector and on the value o f the options presented here, it was agreed that firther debate,
consultations, pilot tests and studies are neededbefore policy decisions are taken.
Provision
Develop a clear public policy towards the private sector that harness the valuable
resources that are available inthis sector
The government needs to create "head room" in its public expenditure envelope so that
some public resources will become available for influencing the behavior o f private
providers through different measures such as contracting with private providers and
subsidizing care for the poor.
BringAlternative PrivatePractitioners (APPs) into to service provider system byworking
with them in strengthening skills and increase the number o f formally trained staff
throughtraining.
In addition to traditional regulatory and quality assurance techniques, it is necessary to
promote health services standards, performance based competitive pressures, and
incentives to attract private practitioners to work inlow coverage areas
Consumption
0 Make information about the quality and price o f private providers readily available to
consumers, especially for the poor
0 Introduce targeted subsidies and community level insurance for the poor and social
insurance mechanisms for civil servants and formal sector workers
47
Publicprivate interaction
0 Increase competition between public andprivate sector through competitive and selective
contracting andperformance benchmarking and service standards.
0 Introduce intemal markets (make public providers compete for public fbnding on a
performance basis) and new public sector management techniques (ie contracting out,
contracting in,management contracts etc).
0 Redefine the role o f the MOH and strengthenits core stewardship capacity in areas such
as strategic planning, monitoring and evaluation, coordination, regulation, quality control
andenforcement
The options described below are initial ideas to address the three main sets o f issues raised, Le.,
under-consumption, quality concems and knowledge base. These options need to be pilot-tested
and evaluated first, before being scaled up nationally, but such pilots should be large enough to
yield meaningful lessons.
4.3.1 ToAddress Under-ConsumptionofServices by thePoor and Women
Pilot contracting private providers with government finding: the government has already
had some experience in contracting out some of its services out to non-profit organizations, but
most o f these models have been through donor-financed projects and have yet to be scaled up.
Serious consideration needs to be given to the option o f larger scale contracting o f HNP service
provision to the private providers, financed by public finds. The recent government
commitment to contract out the management o f 350 community clinics and Union Health and
Family Welfare Centers is a positive step. Such contracting should, however, follow a
transparent and fair process for the selection o f firms/NGOs and should include rigorous
monitoring and accountability procedures, to ensure boththe quality of care and the efficient use
o fpublic resources. The contracting and the monitoring should be performed by an independent
body such as an NGO or private company. Health services should initially be focused on child
andmatemal services. Contracting arrangements should be performance-based. Contracts fees
should be linked to agreed outputs and health outcomes. Box 4.1 provides some pointers on
"contracting in" and "contracting out", based on experiences inother countries.
Expand social marketing: Bangladesh has a positive record o f social marketing in
contraceptives and oral rehydration therapy. Social marketing has been shown to reduce
inequalities inaccess to such commodities. This experience could be expanded to other essential
health-promoting commodities, such as bed nets and soap, which i s known to have a positive
impact on the health o f the poor.
48
Box 4.1:Contracting-Inand Contracting-Outto ImproveHealthServices Utilizationand
QualityinCambodia
Inresponseto poor healthoutcomesandapublic infrastructure ravagedbya quarter-century of
conflict, Cambodia has recentlyundertaken innovative approaches to collaborating with the private
sector for health services delivery. The Ministryo f Health i s currently piloting health financing
reforms inselected districts, through its Accelerated District Development (ADD)program. With
the assistance o f the Asian Development Bank (ADB), Cambodiahas piloted two models o f
contracting for primary healthcare services inan eight-district interventioncovering a total o f one
millionpeople: 1) contracting-out, where contractors (typically NGOs) have full responsibility for
service delivery and directly employ their staffi and 2) contracting-in, where contractors provide
only management support to civil service health staff. The contracting model was intendedto
address one o f the primary causes o fpoor health system performance: poor staff morale and
inadequate management.
Preliminary results show that those districts with contractingmodels outperformed control districts
on several health service coverage indicators, with the contracting-out model yielding the greatest
gains. For example, districts with contracting out increased annual per-capita health care contacts
among the populationto 1.7, compared with 1.2 for contracting-in districts and 0.8 for control
districts, with lower recurrent costs (Bhushan et al., 2002). Contracting-out also proved equitable,
as utilization o f government-financed services increased disproportionately among the poor, whose
out-of-pocket payments were also reduced, up to 70% incontracting-out districts and 40% in
contracting-in districts.
Several factors o fthe Cambodian contractingexperience have contributedtoward its success,
including: establishment o f contract management capacity inthe Project CoordinatingUnit (PCU)
o f the MOH; a transparent competitive biddingprocess; and an agreed-upon monitoring process.
With assistance from ABD, the Ministryof Health specified objectives and solicitedcontract bids
from for-profit and not-for-profit internationalagencies, and subsequently monitored awarded
contracts through the PCU.
For more details on the contracting-inmodel, please refer to an in-depth case study o f the Pereang
District in: Soeters, R, and Griffiths, F. 2003. "Improving Government Health Services Through
ContractManagement: a Case from Cambodia." Health Policy andPlanning, 18: 74-83.
Sources: Bhushan, I., S., and Schwartz, B.March 2002. "Achieving the Twin Objectives o f Efficiency
Keller,
and Equity: Contracting Health Services in Cambodia." Asian Development Bank. ERD Policy Brief No. 8;
Loevinsohn, B. "Contracting for the Delivery o f Primary Health Care in Cambodia: Design and Initial
Experience o f a Large Pilot-Test." The World Bank; Soeters, R, and Griffiths, F. 2003. "Improving
Government Health Services Through Contract Management: a Case from Cambodia." Health Policy and
Planning, 18: 74-83.
Explore insurance / risk-poolinn and mepayment mechanisms: Since people are paying
for private sector HNP services from their pockets, well-designed community insurance schemes
could provide a feasible option for better delivering o f a benefit package with acceptable quality.
Such schemes would not only pool risks and resources for curative services for minor illnesses,
but also provide for catastrophic coverage (if public subsidy for such coverage is considered
appropriate) and could include incentives for seeking preventive services by buildingdifferential
49
co-payments or deductibles. The Bangladesh experience such as Grameen Bank's schemes for
their members may prove to b e h e l p f d in designing locally appropriate community insurance
mechanisms. Box 4.2 provides a brief account o f the Grameen Insurance system. The inclusion
of maternal and child health services in benefit packages is crucial, particularly for assisted
delivery and complications requiring hospital care. Insurance schemes could b e o f different
types in terms of the benefits package, beneficiary pool, and other aspects, ranging from micro-
insurance schemes at the community level to the social insurance programs covering large
populations employed inthe organized sector.
Box 4.2 GrameenKallyan(a communityhealthinsurancescheme)
Grameen Kallyan has been operating since the year 1996. The major actors in this organisation are:
Grameen Kallyan, Grameen Bank members, and non-members o f Grameen Bank. The target
beneficiaries are the rural poor families (within / outside Grameen Bank members). It i s Grameen
Bank's long-term experience flowing from its involvement in micro credit operations that the rural
poor in Bangladesh are prone to serious setbacks emerging from natural calamities as well as from
personal or family-based misfortunes from time to time. This type o f vulnerability deprives them of
the opportunity to continue their thrift savings sustainably. Inaddition, the level and quality o f social
security, especially in the area o f medical care being one o f the lowest, Grameen Bank thought of
introducingmicrofinancing coupled with quality health care services toward improvedhealth status of
the rural poor.
The goal o f Grameen Kallyan i s to provide sustainable quality primary health care services. The
existing interventions o f Grameen Kallyan spread through the districts o f Tangail, Dhaka, and
Comilla. This organisation does not receive any donor aid. In this sense, it i s limited to being a
partnership o f the nature o f Private (NG0)-Community Venture. The major activities included in
Grameen Kallyan are: collecting premium, consultation services, selling essential medicines, selling
pathological services, referral o f pathological cases to outside laboratories, referral o f patient cases for
prescriptionto outside doctors, payment o f certain fees to a patient in case o f admission to an outside
hospital on referral,.
Grameen Kallyan's services are two-fold: (a) management o f microfinancing through health
insurance, and (b) provision o f health care services. Grameen Kallyan, from time to time, also
arranges for health care camps especially inthe area o f care for cataract. angladesh.
Expand direct information Campaigns: Such campaigns should address under-served
households to enhance their appreciation o f the importance of the health services they are not
seeking - and the risk linked to the care obtained from unqualified practitioners. Strategic
planningo fthe Government would guide the private sector inthe provisiono fpublic information
campaigns on individual and household behaviors, with attention to b e paid to changing health-
care seeking behaviors, especially of mothers with children. Bangladesh does have some
successfil experience in modifying health-seeking behaviors through information campaigns,
e.g., family planning, immunizations and oral rehydration therapy. Areas where information
campaigns could have been more effective include the use o f iodized salt, maternal health care,
and tuberculosis control programs. Clearly thought out communication strategies, based on
formative research which study not only health-seeking behaviors but also their socio-cultural
50
determinants are a key pre-requisite for successful information campaigns. Other conditions for
success would be effective leadership and political will to carry out such campaigns, backed up
by highquality services that the population can rely on. An example o f information campaigns
not backed up by appropriate services might be found in the manner that Bangladesh has been
dealing with the problem o f arsenic contamination o f drinking water; while information
campaigns warned people not to drink contaminated water, no viable alternatives have yet been
made available inan affordable and acceptable manner.
Explore demand-side subsidies: Public resources may be usedto provide health coupons
or such similar instruments to poorer population groups, giving them the necessary purchasing
power to consume essential services from the private sector; this approach puts the choice o f
providers in the hands o f the consumers, and empowers them. The success o f demand-side
subsidies depends on the capacity for identifying the potential beneficiaries, and the availability
o f appropriate services o f acceptable quality; hence, this option needs to be used in conjunction
with other policy measures geared towards quality improvement, e.g., mechanisms for
identification o f the poor and accreditation systems .
4.3.2 ToImprove Service Quality and Outcomesin thePrivate Sector
A recurring theme o f this study is how to improve the quality o f health care services.
There are extremely weak and insufficient systems for assuring the quality o f health care in
Bangladesh, and as a result, there i s little information about the quality of care. Weak
performance o f the civil service andpersistent governance problem underpinthe need to design
and implement adequate legal and economic incentives and to establish autonomous regulatory
agencies financially and administratively independent o f the MOHFW. Such agencies must
ensure, promote and disseminate best practices in quality services. Quality o f services will only
be achieved as result o f consistent set o f actions. If Bangladesh's focus on quality becomes
merely a program to regulate the health sector, it is unlikely to change the quality o f health care
significantly inthe near future. It will be necessary to tackle the quality issues on several fronts,
using strategies that may be characterized as "top-down"; "bottom-up", "leading edge", and
"driving force" (Table 4.1). The different strategies for addressing quality involve different
actors, have complementary objectives, andhave their own limitations and timefiames.
51
Tble 4.1 QualityImprovementStrategiesRelevantto Bangladesh: MultipleFronts
Strategy Examples Objectives Limitations
Top-down Licensing, egistration Set minimum Limited impact in informal sector;
accreditation standards effectiveness linked to good
governance
Bottom-up (i) Consumer education & (i) Raise demand and (i) Long run solution, but limited
advocacy expectations for experience in countries where
quality education levels are Idw, modest
(ii) Pilot quality improvement successes have been achieve in
projects in leading (ii) Strengthen specific areas
organizations capacity for quality
(ii) Limited scale
Leading edge Collaborative professional Improve performance Little experience in low-income
learning networks and change patterns countries
of practice
Driving force Financing incentive Influence provider Medium to long term solution;
mechanisms linked to behavior to improve requires group purchasing of health
clinical practices quality services through insurance or pre-
payment
The top-down strategies usually involve government agencies setting standards of care,
most often with an objective o f demanding a minimum level o f quality and safety. In
Bangladesh, most standards inhealth care concem the qualifications of staff and physical inputs
at health facilities, rather than quality assurance processes, such as clinical guidelines, standard
diagnosis, treatment guidelines and continuing medical education linked to certification and
accreditation . Most top-down approaches are mandatory, though accreditation by professional
bodies i s more often a voluntary process. In many other countries accreditation i s mandatory
because large payers (e.g. governments and insurance companies) will require accreditation for
an organization to be eligible for finding. Top-down approaches work best when the criteria can
be easily measured and enforcement is straightforward; this is far from being the case in
Bangladesh. While improvements and additions could be made to top-down strategies in
Bangladesh, concentrating efforts on these approaches i s not likely to make substantial
improvements in the quality o f care in the short and medium term, since most providers work
outside the formal sector, and the govemance environment i s weak.
Bottom-up approaches involve both demand and supply sides o f the equation. On the
demand side, such strategies involve educating and empowering consumers and consumer
organizations. The main purpose i s to enable the public to expect and demand better health care,
or insome cases to redress harm caused to patients. Consumer ratings o fproviders, facilities, and
products (e.g. pharmaceuticals) can lead to changes in the behaviors of providers or the quality
of products. Although there i s a tremendous theoretical appeal o f approaches that reduce
information asymmetries between patients and providers, in practice, they have had limited
success. Where successful, as inthe case o f reducingthe demand for antibiotics for sore throats,
the campaigns were quite targeted, and occurred in conditions where levels o f education and
access to health services are much higher. Supply side strategies involve individual pilot
projects by innovative and leading health care institutions. The limitation with these approaches
i s that they often depend on the individual circumstances o f the hospital or organization
52
undergoing the changes, and are not taken to a larger scale. Yet both types o f bottom-up
strategies are lacking in Bangladesh, and are worth initiating soon, even if their impact i s more
likelyto befelt inthe mediumto longterm.
Probably the most gains inthe short and mediumterm would be made through so-called
leading edge approaches, characterized by collaborative leaming networks. These strategies
involve pullingtogether networks o f providers and facilities to establish priorities for services, in
a framework that uses up-to-date quality improvement methods and information sharing. In
contrast to top-down strategies that try to set minimum standards, these approaches seek to
improve care, and to continually develop best practices, while buildinghuman and institutional
capacity. They are particularly effective inconditions where health systems are fragmented, as i s
clearly the case inBangladesh.
Finally, financing mechanisms can be considered as a driving force for quality
improvement. Financing tools can be linked to the demonstration o f good practice or good
outcomes though subsidies, quality-linked payment rates, or contracting that is partly based on
quality provisions. Their purpose i s to influence provider behavior in a way that improves the
quality o f care and increases accountability. One limitation with these strategies in Bangladesh
i s that people make most health care payments individually from out-of-pocket, so that the
ability to use the power o f group purchasing to influence provider behavior i s lacking. These
payments are essentially in the nature o f "fee for service", and there i s little scope to use other
mechanisms o f compensating the provider (e.g., capitation payments for general practice), as
levers to influence provider behavior. Risk-pooling mechanisms, e.g., community health
insurance schemes, might help develop different kinds o f incentive mechanisms for providers
and consumers alike and could help improve quality o f health services in Bangladesh. Inthe
short to medium term, pursuing public sector contracting for non-public health services would
help to build up experience with financing strategies.
Whatever sets o f strategies are chosen to improve quality, leadership and consensus
building will form important components o f the first steps. The strategies and technical
components will need to be seen as being endorsed by influential decision-makers in the
political, professional, and civic arenas, and the clinical aspects to be developedendorsed by the
top clinicians. Government should play a leading role in catalyzing these initial steps. For
example, govemment may establish a task force or commission to develop and oversee its
quality improvement strategies. Another way govemment can immediately support a movement
to improve quality inhealth care i s to support research and assessments o f quality o f health care.
Providing information on the quality o f current practices i s important to raise awareness and
support for improving quality, and i s also needed to provide a basis for setting standards and
benchmarking, and to get healthworkers involved inquality improvement activities.
Based on the above considerations, the following policy options aimed at improving
quality are targeted at both the formal sector, Le., the qualified allopathic providers and facilities
andthe informal sector (APPs):
Institute specific initiatives to improve aualitv o f services provided by altemative private
practitioners: Encourage the development o f APP representative organizations, both local and
national, through formal involvement in consultation and policy dialogue. Expand existing
initiatives (contracting, training, information dissemination) to include APPs where viable; of
53
particular relevance are the informal providers o f maternal and child health services, including
the traditional birthattendants.
Creation o f a National accreditation agency: Both public and private providers and
facilities should be accredited, by a established independent body with a reliable system o f
regular monitoring and maintenance o f standards, and legal capacity to provide incentives and
enforce sanctions. This agency would provide information to consumers for judging the quality
of services. The work o f this agency should be linked to direct information campaigns to
households.
Upgrade the regulatory framework: This should include the revision o f outdated
regulations and identification o f appropriate mechanisms and resources for enforcement o f
existing, appropriate regulations. Considering the existing, limited capacity to monitor and
enforce regulations, this may be a long-term goal. Professional/provider associations need to be
identified andtheir capacity to play the role o f a self-regulatory bodyneeds to be strengthened.
To promote the dissemination o f franchising: This involves brandname development o f
health services, to give the consumers o f health care a way to choose providers with an assured
standardo f services. Franchising could also provide credit andmanagement support to providers
andstaff and a sense o fbelonging andpride
Direct information campaigns to households: A direct information campaign, as
mentioned above, can increase the consumption o f services. It can also influence the demand for
increased quality o f services, resulting in pressure on private providers to improve their
practices. Such strategy could initially be targeted to the poorest districts and the poor areas o f
selected urbancities.
Promote the formation o f one or more consumer organizations in health: This would
supplement and reinforce the ability o f individual consumers to demand better quality services
and to represent their concerns and even negotiate more competitive prices. Consumer
organizations can also play a role in the monitoring and improving o f quality o f care in the
private sector.
Promoting training of private providers: This option i s particularly usefbl for UPS,
whose qualifications and skills are variable and inadequate in most cases. Recognizing the
ineffective enforcement o f the laws that ban their illegal practice, an alternative or
complementary public policy might be to equip them with acceptable levels o f knowledge and
skills, both to improve their effectiveness and reduce the potential harmthey cause to the public.
Special focus should be given to the training o f local midwives, birth attendants and drug shop
workers through the provision o f scholarships to be trained and further monitoring o f
performance.
A.3.3. To Improve the Knowledge Basefor Policy-Making
Subsidize further research, including operational research and pilot initiatives: This was
emphasized in the stakeholder consultations as a pre-requisite for making policy decisions on
public-private interaction. Policy decisions must have a solid evidence-base. Filling information
gaps and testing new approaches was recognized as essential for enhancing the private sector
contribution to health outcomes.
54
To enrich our understanding o f the dynamics o f the private sector, further work should
include a more in-depth analysis of: (a) factors market (pharmaceuticals, medical equipment,
consumables, etc); (b) labor market dynamics; (c) capital markets; and (d) potential for insurance
markets. A broadeningo fthe knowledge base is also necessary inother related subject areas. It
must be stressed however, these suggestions for further studies are not to be misconstrued
as a reason to delay policy actions for which considerable evidential basis already exists. A
distinction must be made between operations research to pilot-test the policy options and the
other research activities aiming to generate new knowledge.
While the specific areas o f research would need to be determined through future
consultations, some suggestions are listedhere:
e Conduct a more comprehensive, nationally representative survey o f private sector
providers, their consumers (geographical and socio-economic strata), types o f
services provided, prices o f these services, quality, client satisfaction, level o f
training o f providers, types o f medicine practiced. Such a follow-on survey and
training should include all types o f alternative providers as well, particularly
traditional birth attendants, and drug store managers; and should include a
comparative analysis o f strengths and weaknesses o f private providers, including
NGOs.
e Carry out econometric analysis o f the private health care market, including an
analysis o f the supply-demand curves, price-elasticity, income elasticity, unit-costs
and such other aspects that could enlighten the policy-makers more about the
economic drivers o f supply of, and demand for, private services.
e Conduct a further analysis o f HNP-related commodities (pharmaceuticals, vaccines,
baby food formula, hygiene products, bed nets, etc.), to study both supply and
demand side factors relevant to these markets and look at ways o f buildingpublic-
private partnerships inthese areas.
e Conduct a labor market assessment, including the market dynamics o f supply and
demand in relation to human resources, and the incentive mechanisms influencing
providers inthe private sector.
e Conduct a full-fledged study o f governance issues as they apply to private as well
as public sector, including the issue o f dual practice.
e Good research and development programs (such as those developed by HMOs in
the US)might contribute significantly to policy options aimed at improvingquality
o f services.
Based on a thorough analysis o f the various markets relevant to HNP services, it i s
essential to develop a strategy for creating a more competitive environment among the
private providers and also between the public and private providers, using public policy as
a facilitating tool.
Develop better information systems: Expand the health information system to collect
more reliable data from the private sector, especially with regard to disease burden but also in
private sector resources (practitioners, infrastructure, equipment etc.).
Information sharing: Such efforts would aim at disseminating information to policy-
makers, managers, and other key actors inthe public sector to enhance their understanding o f the
55
private sector and vice versa. Key data to be shared are: public budgets and resource allocation
across districts, health indicators, consumer preferences on health, market analysis on health,
best management practices on healthprovision etc.
4.3.3 Promotingprivateinvestmentin heaIth
The scaling-up o f private health care providers requires inthe long term o f additional resources
from local and foreign investors. Investment will provide financial and physical resources,
"known how" and best practices that can be rapidly spread across existing private and public
providers with a positive impact in quality and unit cost. The Government should initiate an
aggressive policy to encourage private investment in the curative health care sector. Some
specific actions are suggested as follows:
To install a Suecial Task-Force Unit in the Office of the Primer Minister to help local and
foreign investors to seek potential business inthe health sector in Bangladesh. This unit would
have to define goals interms o fthe value o fthe investment, andwill coordinate all neededpublic
actions to accomplish those objectives. The Unit will have to develop the appropriate legal
conditions and to monitor the dissemination o f lease management, concessions as options to
contract agreements and also to evaluate the transfer o f ownership o f public health facilities to
the private sector.
Bonuses to NGOs or private providers that serve inunderserved areas.
To encourage alliances between the MOHFW and the private sector to achieve specific public
health goals such as immunization, tuberculosis and malaria control etc. Creation o f Health
Management Organizations
Implementing social insurance for civil servants and private employees. This implies a
mandatory earnings contribution (about 4% o f total earnings) that could go either to the Ministry
o f Health and Family Welfare providing a right to receive medical attention without fees or to a
health maintenance organizations (HMO). This strategy may allow to obtainnot less than 200
milliondollars a year into the health sector to strengthenthe developmentof a competitive
marketwith "incentive schemes" and "minimum standards" in the search of better quality
and coverage. This option requires the design of a new legislation that promotes the formation
o fHMO.
The accreditation o f activities and health plans offered by these HMO will be under the
responsibility o f the NAAH. The Health observatory/Consumer alliance will inform consumers
about the best choices and disseminate user's rights.
4.4 The Way Forward
This section provides some options on how the above policy options might be realized into
action. The original intention o f the authors was not to be prescriptive about the solutions for the
issues emerging from their study. However since the Government is preparing its new Health,
Nutrition and Population Sector Program (HNPSP), the findings o f this study and the suggested
policy options presented here - both the "what" and the "how"- could serve as a useful vehicle
for policy dialogue inBangladeshand also as starting point for the design and implementation o f
the sector reforms required to achieve the desired HNP outcomes as specified in the MDGs, in
the context o fa choice competitive model. The following are possiblenextsteps:
56
0 Set up a Public-Private Task Force in the MOHFW to provide a focal point for
developing processes and activities towards a fruitful relationship between public and
private actors. This task force would promote, institutionalize, and coordinate public-
private interactions. This would be a temporary measure until durable processes,
systems, financing mechanisms, and regulations relating to public/private partnershipin
the health sector are well-established and absorbed by the relevant agencies.
0 Create the necessary fiscal space or "head room" inthe public resource envelope, so as to
ensure the availability o f the substantial additional resources needed to finance increased
engagement with the private sector and the required pilots on a large enough scale to
makethem replicable (over and above the current health sector budget, most o f which is
already committed to public sector provision).
0 Capacity development inthe MOHFW to enhance its engagement with the private health
sector. Such capacity requires the development o f new types o f skills, in areas that have
not traditionally been among the functions o f MOHFW such as negotiation, social
marketing, monitoring performance based on results etc. The capacity o f other relevant
actors, such as civil society, development partners, and private sector actors should also
be considered. This would include substantial education for public sector staff on the
size and distribution o f private HNP sector in Bangladesh, and ways o f building
partnerships with private actors; similarly, training and confidence-building activities
targeted at the private sector would also be critical.
0 Participatory policy-making and more inclusive planning and programming: The
findings o f this and other related studies should be actively and widely disseminated, as
input for the for national policy debate, involving stakeholders at all levels. Private
sector actors should be included in such process and a full consideration given to
strengthen private sector capacity in service and facility planning. Participatory policy
debate should take advantage o f the preparation o f the full PRSP from the I-PRSP.
0 Establishment o f anAccreditation System
An Accreditation body would work with consumers, health care purchasers, legislators and the
providers in developing standards for both public and private providers in Bangladesh.
Participation in accreditation and certification programs would be voluntary. This would create
incentives to good providers to distinguishthemselves from badproviders.
The NAAHshould evaluate health care inthree different ways;
a. Through accreditation (a rigorous on -site assessment o f key clinical and administrative
processes)
b. Through healthplanemployer data ifthey exist.
c. Througha comprehensive user's andprovider's surveys.
Since regulation i s usually costly and often ineffective, accreditation is recommended to
"protect a minimal set of standards on health service provision". Non-regulatory
interventions or incentives would also be essential for improving private sector activity.
57
The NAAHmay use several factors for the elaboration o fthe rankingo fproviders such as:
a. Access and service
-Are there enough doctors and specialists to serve the number o fpatients?
-Which are the services o f the health facility?
-What i s the access to training o f the healthworkers and practitioners inthe health facility?
-Sufficient and adequate resources: human (health personnel), equipment, and infrastructure
o f health facilities.
-Availability o f diagnostic tests and drugs
Qualifications o fproviders
-Ensuring each doctor is licensed and trained to practice medicine and that the users are
satisfied with the services received bypractitioners.
-How do the providers rate their own doctors?
-How do the providers deal with bad practices and complaints against their practitioners or
healthworkers?
-Which i s the level o f cleanliness o f the health facility?
-Which i s the level o f access to water, electricity and other services inthe health facility?
-What i s the quality o f the food providedby the health facility?
-What i s the ratio o fpractitioners/administrativeworkers inthe health facility?
-What i s the status o f financial procedures?
-Does the hospital have clear and written standards and protocols for treatment and
diagnosis?
-What i s the average waiting time o f a patient before he/she receives treatment?
b. Healthoutcomes
-Are the health activities o f the provider helping people to stay healthy?
-Are preventivehealth activities appropriate?
-Evaluate whether people recover effectively from illness.
c. User's satisfaction
Measure by the level o f comfort, promptness of service, waiting time to receive the medical
attention, treatment by the doctors and health workers, privacy, availability o f services and
food receivedby the patient.
To be eligibly for accreditation, a health provider would need to be in operation for a
minimumo fthree years. The rankingo f providers should be simple and easy to understand
for the public such as: excellent, good, satisfactory, below average, fail.
The NAAHwill use two options for accrediting healthcare providers:
a. D o the inspections directly with own personnel.
b. Contractinginspectors from other NGOsandprivate institutions.
Both options will demand special training o fpersonnel.
- Creation of a Health observatory/Consumer Coalition
The mainactivities ofthe HealthObservatory/Consumer Coalitionwould be the following:
58
a. Advocacy
Help to educate, assist and protect the rights o f individuals through consumer information,
consumer participation, consumer advocacy programs, data collection and independent quality
oversight. Draft model policies or legislation on the areas o f interest. Help consumers know
about options o f coverage, provision and treatment.
b. Grassroots organizing
Broadencoalitions by creating worker/consumers partnerships at the local andnational level.
Implement a quality watch-line (toll free number) which collects individual experiences o f poor
quality o f care services from consumers or health workers in the country. These would be real
life stories very useful for the design ofpro-consumer strategies inthe health sector.
c. Participation inquality measurement
Ensure that the consumer voice is listened to inall forums andwork groups relatedto legislation
or decisions on the health sector.
d. Accountability
The observatory should be independent of providers and financers o f health care and free of
conflict o f interests. The agency should collect and disseminate information to strengthen the
national data system.
A National Contracting Agency
The formulation o f a strategy framework for purchasing services from NGOs and other non-
public providersshould be defined clearly. The themes for further definition andwork are:
a. The services to be purchased.
b. Criteria for choosing the geographical areas o f interventions. The proposal is to start in
the unions were there are already NGOs delivering health services.
c. The procedures to monitor andto supervise the performance o f the NGOs.
d. The establishment o fthe contract agreements andbiddingprocess.
e. Types o f partnership between the GOB andNGOs.
f. The payment mechanisms to providers. One option is to initiate the process with a
simpleper capita allocation.
g. To give the contracted NGOs the possibility ofretention o fuser fees.
h. Accreditation shouldbemanagedbyanindependent agency.
A private institution (The National Contracting Agency) would be contracted by the GOB to
manage the bidding process, and the supervision and monitoring o f performance, together with
the training o fhealthworkers inthe NGOs.
A permanent GOB-Civil Society committee lead by the MOHFW would guide and provide
inputs to the National Contracting Agency on the national goals o f targeting the poor. The
National Contracting Agency would beresponsible for:
59
1. The decision o fwhat to contract out.
2. The decision of from whom to purchase services.
3. The definition o fthe contract payment option.
4. To negotiate the terms o f the contract with the selected provider.
5. To supervise andto monitor the performance of the contracted provider.
6. To modify contracts base on performance.
7. To promote the formation o fpublic andprivate healthnetworks at the unionlevel.
The NCA will need also to develop indicators to measure contracting outcomes such as:
Improvements in average health indicators with significant reduction in disparities
between the rich and the poor, betweenmale and female.
Infant Mortality Rates reduced to a target rate with a significant decrease inrich/ poor ,
male/female ratios.
Under-5 Mortality Rates reducedto a target rate with a significant decrease in rich/poor,
male/female ratios.
Maternal Mortality Ratio (or suitable proxy) reduced to a target rate with a significant
decrease inricldpoor ratio.
Percentage o f fblly immunized children against 6 diseases within the first year o f life
increased to x %.
Proportion o fwomen with obstetric complications treated at facilities increased to x %.
Use o f curative essential services by women, children andthe poor increased to x %.
Discontinuation rate o f contraceptionreduced to x %.
Proportion o fwomen who receive antenatal care.
Proportion o fwomen who receive post-natal care.
Fertility rates.
Pilot activities to test the selected policy options. Selected policy options need to be
tested through operations research to determine their feasibility and measure their impact
in the Bangladeshi context. The design o f such pilots should build on previous
experience, and evidence available about public-private partnerships such as demand-side
financing, micro-insurance schemes, and contracting with results-based financing.
A summary o fissues and suggested policy options is given intable 4.2.
60
A preliminary timeline for the next steps is suggested subject to previous agreement with the
Government:
Completion o f other related studies: BetweenAugust
- Governance issues inthe health sector. 2003 and May 2004
- Pro-poor targeting mechanisms.
- NGO contracting evaluation.
- Comparative study of cost-effectiveness among public andprivate
health care providers.
Dissemination o f existing evidence, multi-pronged communication From July to
exercise, consultations across the country December 2003
Development o f broad-based HNP Policy Options dialogue May 2004
~ ~ ~
Initiation of Pilot Interventions (e.g., vouchers, micro-insurance, methods Starting January
for contracting-out non-public providers, demand-side subsidies such as 2004
cash transfers to the poor )
"1 i
r
c
fi
.I
x
4
4
a
cl
s
75
e,
2
0
0
N
r-
N
e
29 .3
F F B2
References
ACPR and UNICEF (2001) Review of availability and use of emergency obstetric care (EmOC)
services in Bangladesh. Dhaka: ACPR andUNICEF.
Afifi, N.H,, R Busse, Harding, A.L. (2003) Regulation of Health Services, inHarding, A.L. and
Preker, A.S. (eds.) Private Participation in Health Services. Washington, DC: World Bank.
Ahmed S. (2002) The Politics of Reforms in South Asia: Bangladesh and Pakistan. South Asia
Region Internal Discussion Paper. Washington, DC: World Bank.
Ali M., Emch, M.,Tofail, F. and Baqui, A.H. (2001) Implications of health care provisionon
acute lower respiratory infectionmortalityinBangladeshi children. Social Science and Medicine,
52: 267-277.
Ali, Q.L.(2001) Calculation oftotal unit cost for diarrhoeal management at district hospital and
Thanahealthcomplex. Dhaka: HealthEconomics Unit,Ministry o fHealthandFamily Welfare.
Alisjahbana A., Williams, C., Dharmayanti, R., Hermawan, D., Kwast, B.E., and Koblinsky, M.
(1995) An Integrated Village Matemity Service to Improve Referral Pattems in a Rural area in
West-Java. InternationalJournal of Gynecology & Obstetrics, 48(Suppl): 83-94.
Amin, R., Chowdhury, S.A., Kamal, G.M. and Chowdhury, J. (1989) Community Health
Services and Health Care Utilization inRural Bangladesh. Social Science and Medicine, 29(12):
1343-1349.
Amhold, R.G.(1979) Paramedicalprograms inruralBangladesh. SouthMed J, 72(8): 992-996.
Ashraf, A., Chowdhury, S. and Streefland, P. (1982) Health, Disease and Health-Care inRural
Bangladesh. Social Scienceand Medicine, 16: 2041-2054.
Axelsson, H.,Bustreo, F. and Harding, A. (2003) Private sectorparticipation in child health: A
review of WorldBankprojects 1993-2002. Washington, DC: World Bank.Inpress.
BangladeshBureau o f Statistics (1999a) Statistical Yearbook of Bangladesh. Twentieth Edition.
Dhaka: PlanningDivision, MinistryofPlanning.
Bangladesh Bureau o f Statistics (1999b) Health situation and health care expenditures in
Bangladesh: evidencesfrom nationally representative surveys.
BangladeshBureau of Statistics (2001) Preliminary report of household income and expenditure
survey -2000.
BangladeshBureau of Statistics (2002) 2000 Statistical Yearbook of Bangladesh, 2IsfEdition.
Dhaka: PlanningDivision, MinistryofPlanning.
BangladeshBureau of Statistics (2002) Report of the Labour Force Survey Bangladesh 1999-
2000. Dhaka: Planning Division, Ministryof Planning.
BangladeshBureau of Statistics andUNICEF (2000) Progotir Pathey: On the road toprogress -
Achieving the goals for children in Bangladesh. Report on Multiple Indicator Cluster Survey
(MICS) 2000.
BangladeshBureau o f Statistics (1998). Report on survey of private health service establishment
1997-1998.Dhaka: Statistics Division, Ministry of Planning.
Bangladesh Medical Association (2002) Proposed policy framework for "Health Sector
Program (HSP: 2004-2006) ". Draft submittedto The MinistryofHealth and FamilyWelfare.
BangladeshMedical andDental Council (2003) Personalcommunication.
BangladeshNursing Council (2003) Personalcommunication.
Baqui, A.H., Sabir, A.A., Begum, N.,Arifeen, S.E., Mitra, S.N. and Black, R.E. (2001) Causes
of childhood deaths inBangladesh: an update.Acta Paediatrica, 90:682-690.
Barkat, A., Ara, R., Huque, M. and Syeduzzaman, K.M. (2003) Case Studies in privatehublie
partnership in the HNP sector in Bangladesh: Preliminary report. Background study to the
Private Sector Assessment. Unpublished.
Begum, H. (1998). Health care, ethics and nursing in Bangladesh: a personal perspective.
Nursing Ethics, 5(6), 535-541.
Bhardwaj, S.M. andPaul, B.K. (1986) Medical Pluralismand Infant Mortality ina Rural Area of
Bangladesh. Social Science and Medicine, 23(10): 1003-1010.
Bhuiya, A. (1992) Village Health Care Providers in Matlab, Bangladesh: A Study of Their
Knowledge inthe Management o f Childhood Diarrhoea. JDiarrhoeal Dis Res, 1O( 1): 10-15.
Bhuiya, A. and Streatfield, K. (1995) Feeding, home-remedy practices, and consultation with
health care providers during childhood illness in rural Bangladesh. Journal of Diarrhoeal Dis
Res, 13(2): 106-112.
Bhushan, I. et al(2002) Achieving the twin objectives of efjciency and equity: contracting health
services in Cambodia. Manila: Asian Development Bank, Economics and ResearchDepartment
PolicyBriefNumber6.
Buchan, J. (2000). Nursingbrain drain. Nurse Stand, 15(4), 22-23.
Cash, K.et al(2001). Telling them their own stories: legitimizing sexual andreproductive health
educationinruralBangladesh.Sex Education, 1:43-57.
Centre for Software and Info-processing (1995) Report of the Survey on Professional and
Miscellaneous ServicesPersonnel. BangladeshBureauo f Statistics.
Chakraborty, S. et al (2000) Improving private practitioner care o f sick children: testing new
approachesinruralBihar.Health Policy and Planning, 15(4): 400-407.
Chanda, R. (2002) Trade in health services. Bulletin of the World Health Organization, 80(2),
158-163.
Chaudhury, N. and Hammer, J.S. (2002) Ghost Doctors: Absenteeism in Bangladeshi Health
Facilities. Washington: World Bank, draft mimeo.
CIET Canada and MOHFW (1999) Health and Population Sector Program 1998-2003
Bangladesh, Baseline Service Delivery Survey: FinalReport.
Claquin, P. (1981) Private Health Care Providers in Rural Bangladesh. Social Science and
Medicine, 15: 153-157.
Data International (1998) Bangladesh National Health Accounts 1996/97:Final Report. Dhaka:
Prepared for the HealthEconomics Unit,Ministryof Health and FamilyWelfare.
DKT Intemational(2003) Accessed at www.dktintemational.org.
Dmytraczenko, T. et a1 (1999a) Bolivia: a mid course assessment of the National Mother Child
Health InsuranceProgram. Health Reform Prior Sen, Summer-Fall: 19-20.
Dmytraczenko, T., Scribner, S., Leighton, C., Novak, K. (1999b) Reducing maternal and child
mortality in Bolivia. Bethesda, MD:Partnershipsfor Health Reform.
Ensor, T. (2001) Projecting the cost of the Essential Service Package. Health Economics Unit,
Dhaka. ResearchPaper26.
Ensor, T., Dave-Sen, P., Ali, L., Hossain, A., Begum, S.A. and Moral, H. (2002) Do essential
service packages benefit the poor? Preliminary evidence from Bangladesh. Health Policy and
Planning, 17(3): 247-256.
Feldman, S. (1983) The Use of Private HealthCare Providers inRuralBangladesh: A Response
to Claquin. Social Scienceand Medicine, 17(23): 1887-1896.
Forsberg, B.C. and Axelsson, H. (2003). Interviews with Senior Managers in the Public and
Private Health Sector on Public Policies versus Private Care and Experience on Private Care.
Background study to the Private Sector Assessment.
Government of Bangladesh and UNICEF (1999) Situation assessment and analysis of children
and women in Bangladesh.
Greenwood, A.M., Bradley, A.K., Bypass, P., Greenwood, R.W. et a1 (1990) Evaluation o f a
primaryhealth care program inthe Gambia. The impact o ftrained traditional birthattendantson
the outcome ofpregnancy.Journal of TropicalMedicine and Hygiene, 93: 58-66.
Gruen, R., Anwar, R., Begum, T., Killingsworth, J.R. and Normand, C. (2002) Dual Job
holdingpractitioners inBangladesh: an exploration. Social Scienceand Medicine, 54: 267-279.
Gwatkin, D.R., Rustein, S., Johnson, K., Pande, R.P. and Wagstaff, A. (2000) Socio-economic
differences in Health, Nutrition, and Population in Bangladesh. Washington, DC: World Bank.
Harding, A.L. and Preker, A.S. (2003) Private Participation in Health Services. Washington,
DC: World Bank.
Hay R., et al. (2002) Health Futures in Bangladesh: Some Key Issues and Options. Report to
The World Bank (Unpublished).
Health Economics Unit (HEU) (1998a) Private Sector Medical Clinics and Hospitals Survey.
Research PaperNo. 15. Dhaka: MinistryofHealth and FamilyWelfare.
HEU(1998b). Economic aspects of human resource development in health andfamily planning
in Bangladesh: Costs of education and training of health workers in Bangladesh. ResearchPaper
No. 14.Dhaka :Ministryof Health andFamilyWelfare.
HEU, Institute o f Health Economics (ME), National Institute for Population Research and
Training (NIPORT) (2003) Private HNP sector assessment: Jindings of surveys on private
providers and consumers. Background study to Private Sector Assessment. Unpublished.
Hossain, S.M.,Bhuiya, A., and S. Rasheed(2001) Correlates of perceived malarial episodes and
treatment-seeking behavior in a malaria-endemic rural area in Bangladesh. Southeast Asian
Journal of TropicalMedicine and Public Health, 32(4): 707-719.
Human Resources Development Unit. (2003). Task forces reports on human resources
development in health, nutrition and population sector. Dhaka: Ministry of Health and Family
Welfare.
Huang, P (1995) "An overview of hospital accreditation in Taiwan, Republic of China", The
international Journal o fHealthPlanning andManagement. 10(3), 183-192.
Hye, H.K.M.A. (2003) Health Regulation Review. Background study to Private Sector
Assessment. Unpublished.
Islam, M.A., Wakai, S,Ishikawa N., Chowdhury, A.M., Vaughan, J.P. (2002) Cost-effectiveness
of community healthworkers intuberculosis control inBangladesh.Bulletin of the World Health
Organization, SO(6):445-50.
Khan, S.H., Chowdhury, A.M., Karim, F., and Barua, M.K. (1998) Training and retaining
Shasthyo Shebika: reasons for turnover of community health workers in Bangladesh. Health
Care Superv, 17(1):37-47
Kumaresan, J.A., de Colombani, P., and Karim, E. (2000) Tuberculosis and health sector reform
inBangladesh.IntJTuberc LungDis, 4(7):615-21.
Leslie, C. (ed.) (1976) Asian Medical Systems. University o f CaliforniaPress.
Levin, A., Rahman, M.A., Quayyum, Z., Routh, S. and Barkat-e-Khuda (2001) The demand for
child curative care in two rural thanas of Bangladesh: effect of income and women's
employment. International Journal of Health Planning and Management, 16: 179-194.
Loevinsohn, B. (2000) Contractingfor the delivery of primay health care in Cambodia: design
and initial experience of a large pilot-test. Washington, DC: World Bank Institute Flagship
Program Online Journal.
Luby, S. et al (2002) Improving private practitioner sick-child case management in two urban
communities inPakistan. TropMed Int Health, 7(3): 210-219.
Newbrander William (1999). `Accreditation of providers for the National Health InsuranceFund
of Tanzania". Management Sciences for Health, 2000.
Marek, T. et a1 (1999) Successfbl contracting o f prevention services: fighting malnutrition in
SenegalandMadagascar.Health Policy and Planning, 14(4): 382-389.
Miller, L.C., Jami-Imam, F., Timouri, M., and Wijnker, J. (1995) Trained traditional birth
attendants as educators ofrefbgeemothers. WorldHealth Forum, 16(2): 151-156.
Ministry of Finance, Government of Bangladesh (2002) A national strategy for economic
growth, poverty seduction and social development.
Ministry of Health and Family Welfare, Government of Bangladesh(2001) Public Expenditure
Review (2000/01) of the Health and Population Sector Program.
Ministry of Health and Family Welfare, Government of Bangladesh (2002) Conceptual
Framework on Health, Nutrition and Population Sector Program (2003-2006): Interim Draft.
Planning Wing.
MinistryofHealthandFamilyWelfare, Government ofBangladesh(2001)Bangladesh National
Strategyfor Maternal Health.
National Institute of Population Research and Training (NIPORT) (2002) Bangladesh Maternal
Health Services and Maternal Mortality Survey 2001: Preliminay Report.
Nieves, I., Forgia, G.M., and Ribera, J. (2000) Large-scale government contracting of NGOs
La
to extend basic health services topoor populations in Guatemala. Washington, DC: IESE, Inter-
American Development Bank, World Bank.
NIPORT, Mitra and Associates, and ORC Macro (2001) Bangladesh Demographic and Health
Survey 1999-2000. Dhaka, Bangladeshand Calverton, USA.
ORC Macro (2003) Demographic and Health Surveys - STAT Compiler. Available at
www.measuredhs.com. Accessed April 13,2003.
ORG-Marg Quest Ltd. (2000a) Report on Medical Practitioners and Pharmacists in
Bangladesh. Dhaka: Report for BritishCouncil andNICARE.
ORG-Marg Quest Ltd. (2000b) Survey on Village Doctors Practicing in Brahmanpara. Dhaka:
Report for BritishCouncil andNICARE.
O'Rourke (1994) An Evaluation of a TBA Training Program: Its Effect on TBA Practices and
Perinatal Mortality. PhD Thesis. Amherst, MA: University o f Massachusetts, School o f Public
Health.
Pang, T., Lansang, M.A., and Haines, A. (2002). Brain drain and health professionals: a global
problem needs global solutions. British Medical Journal, 324: 499-500.
Peters, D.H. (2002) Oversight o f Health Sector Policy and Performance in Low-Income
Countries: The Example of Sexual and Reproductive Health Services in India. Reproductive
Health Matters, lO(20): 82-94.
Peters, D.H., Yazbeck, A.S., Sharma, R.R., Ramana, G.N.V., Pritchett, L.H. and Wagstaff, A.
(2002) Better health systemsfos India 's poor: findings, analysis, and options. Health, Nutrition,
andPopulation Series. Washington, DC: HumanDevelopment Network, World Bank.
Peters, D.H. et a1 (2003) Labor market assessment for the HNP sector in Bangladesh.
Unpublished.
Policy andResearchUnit (2002) HRDData Sheet 2001. Ministryof Healthand Family Welfare
Rosen, James "Contrating for Reproductive Health Care: A Guide". The World Bank,
December 2000.
Saade, C. et a1 (2001) The stoy of a successful public-privatepartnership in Central America:
handwashing for diarrheal disease prevention. Arlington, VA: BASICS 11, EHP, UNICEF,
USAID, WorldBank.
Sarder, A.M. and Chen, L.C. (1981) Distributionand Characteristicso fNon-Government Health
Practitioners is a RuralArea of Bangladesh. Social Scienceand Medicine, 15: 543-550.
Schellenberg, J.R. et a1 (1999) KINET: a social marketing program o f treated nets and net
treatment for malaria control inTanzania, with evaluation of childhealth and long-term survival.
TransR Soc TropMed Hyg, 93(3): 225-231.
Schellenberg, J.R. et a1 (2001) Effect of large-scale social marketing o f insecticide-treated nets
on child survival inrural Tanzania. Lancet, 357(9264): 1241-1247.
Schuler, S.R., Bates, L.M. and Islam, M.K. (2002) Paying for reproductive health services in
Bangladesh: intersections between cost, quality and culture. Health Policy and Planning, 17(3):
273-280.
Sen, B. (2001) The health divide: analysis of inequalities in health in Bangladesh. Dhaka:
BangladeshInstitute for Development Studies.
Smith, E., Brugha, R. and Zwi, A. (2001). Working with Private Sector Providers for Better
Health Care, An Introductoy Guide. London: Options and London School o f Tropical Hygiene
andMedicine.
Smith, E. (2002) Social Franchising Reproductive Health Services: Can It Work - A Review of
the Experience. Working PaperNo. 5. London: Marie Stopes International.
Social Marketing Company (2003) Personalcommunication.
Soeters, R. and Griffiths, F. (2003) Improving government health services through contract
management: a case from Cambodia. Health Policy and Planning, 18(1):74-83.
Stenson, B. et a1 (2001a) Real world pharmacy: assessing the quality of private pharmacy
practice inthe Lao People'sDemocratic Republic. Social Science and Medicine, 52(3): 393-404.
Stenson, B. et a1 (2001b) Private pharmacy practice and regulation. A randomized trial in Lao
P.D.R. Int JTechnol Assess Health Care, 17(4): 579-589.
Streatfield, P.K., Hadley, M. and Chakraborty, N. (2001) Status of program indicators: annual
program review 2001 of Health and Population Sector Program 1998-2003. Dhaka: World
Bank.
Survey Research Group o f Bangladesh. (2003) Research for Development of Alternative
Incentive Schemefor HRD in Health & Population Sector. Dhaka: Draft Report to the Policy and
ResearchUnit, Ministry of Health & Family Welfare.
SustainableHealthcare Enterprise Foundation(2003) Accessed at
http://www. shefoundation.org/projects/index.html
Taylor, R. (2003) Contracting for Health Services, in Harding, A.L. and Preker, AS. (eds.)
Private Participation in Health Services. Washington, DC: World Bank.
UNB(2002) NBRplansfresh huntfor new taxpayers.The Independent,December12.
UNICEF (1999) Situation Assessment and Analysis of the Children and Women in Bangladesh.
Dhaka: UNICEF.
Walt, G. (1988) Community Health Workers: Are National Programs in Crisis? Health Policy
and Planning, 3(a): 1-21.
Winch, P. (2003) Personalcommunication.
World Bank and Asian Development Bank (2002) Poverty in Bangladesh: Building on Progress,
joint report by the WorldBank and Asian Development Bank.
World Bank (1998) Health and Population Program Project. Project Appraisal Document.
World Bank (2000a) Bangladesh: Financial Sector Distress and Lost Economic Growth.
Washington, DC: World Bank.
World Bank (2000b) Corruption in Bangladesh, Costs and Cures. Dhaka: WorldBank
World Bank (2001) BangladeshCountry Assistance Strategy. Dhaka: World Bank.
World Bank (2003) HNP Poverty Net.
World Health Organization (WHO) and John Snow Inc.(2002) County ProJile on Child Health
and Development in Bangladesh. Dhaka: WHO.
WHO (2003a) SelectedIndicatorsfor Bangladesh.
Accessed at http://www3 .who.int/whosis/country
WHO (2003b) Report of county visit to Bangladesh, B. Daelmans.
WHO (2003~)WHO EstimatesofHealthPersonnel.
Accessed at: http://www3.who.int/whosis/health Dersonnel/health personnel.cfm
Young, A. (1983) The relevance of traditional medical cultures to modern primary health care.
Social Scienceand Medicine, 17: 1205-1211.