Economic Development Institute of The World Bank A Supply-Demand Model of Health Care Financing with an Application to Zaire A Training Tool Ricardo A. Bitran EDI TECHNICAL MATERIALS EDI TECHNICAL MATERIALS A Supply-Demand Model of Health Care Financing with an Application to Zaire A Training Tool Ricardo A. Bitran The World Bank Washington, D.C. O 1994 The International Bank for Reconstruction and Development / THE WORLD BANK 1818 H Street, N.W. Washington, D.C. 20433, U.S.A. All rights reserved Manufactured in the United States of America First printing May 1994 The Economic Development Institute (EDI) was established by the World Bank in 1955 to train officials concerned with development planning, policymaking, investment analysis, and project implementation in member developing countries. At present the substance of the EDI's work emphasizes macroeconomic and sectoral economic policy analysis. Through a variety of courses, seminars, and workshops, most of which are given overseas in cooperation with local institutions, the EDI seeks to sharpen analytical skills used in policy analysis and to broaden understanding of the experience of individual countries with economic development. Although the EDI's publications are designed to support its training activities, many are of interest to a much broader audience. EDI materials, including any findings, interpretations, and conclusions, are entirely those of the authors and should not be attributed in any manner to the World Bank, to its affiliated organizations, or to members of its Board of Executive Directors or the countries they represent. Because of the informality of this series and to make the publication available with the least possible delay, the manuscript has not been edited as fully as would be the case with a more formal document, and the World Bank accepts no responsibility for errors. Some sources cited in this book may be informal documents that are not readily available. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center Inc., Suite 910, Rosewood Drive, Danvers, Massachusetts 01923, U. S. A. The backlist of publications by the World Bank is shown in the annual Index of Publications, which is available from Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, Banque mondiale, 66, avenue d'Ina, 75116 Paris, France. Ricardo A. Bitran, an economist, is senior scientist in the International Health Group of Abt Associates Inc., Cambridge, Massachusetts, U. S. A. Library of Congress Cataloging-in-Publication Data Bitran, Ricardo A., 1958- A supply-demand model of health care financing with an application to Zaire : a training tool / Ricardo A. Bitran. p. cm.-(EDI technical materials) Includes bibliographical references. ISBN 0-8213-2342-3 1. Medical economics-Computer simulation. 2. Medical economics- Zaire-Computer simulation. 3. Medical care-Utilization--Computer simulation. 4. Medical care-Utilization-Zaire-Computer simulation. 5. Health facilities-Cost effectiveness-Computer simulation. 6. Health facilities-Zaire--Cost effectiveness- Computer simulation. 7. Public health-Econometric models. 1. Title. II. Series. RA410.5.B58 1994 338.4'33621'01 13-dc20 93-20079 CIP EDI Catalog No. 480/014 CONTENTS FOREWORD Vii PREFACE ix ACKNOWLEDGMENTS xi 1. INTRODUCTION 1 2. OVERVIEW OF THE MODEL 3 Pricing 6 Break-Even Analysis 7 Health Facility Loss and External Subsidization 8 Comparative Statics 9 A Change in the Price of the Focus Health Facility 9 A Change in a Competitor's Price 10 A Change in the Number of Competitors 10 A Change in the Illness Incidence Rate 10 A Change in Population Size 11 A Change in Household Income 11 A Change in the Geographic Distribution of the Population 11 Changes in Costs 11 3. AMBULATORY CARE 13 Structure of the Market 13 Health Care Demand 14 Curative Care 14 Preventive Care 16 Consumer Categories 16 Health Center Finances 18 Health Center Revenue 18 Health Center Costs 18 iii iv Contents 4. DEMAND, COSTS, AND HEALTH CENTER FINANCIAL PERFORMANCE: SIMULATIONS 21 Structure of the Health Center Component of the Model 21 Simulations 21 The Base Scenario 21 An Increase in the Price of Curative Care at the Health Center 24 A Change in a Competitor's Price 26 Price Cross-Subsidization between Curative and Preventive Care 27 Price Discrimination by Income 28 Price Discrimination by Distance to the Health Center 29 A Change in Insurance Enrollment and Premium 30 An Increase in the Insurance Copayment 32 An Increase in the Amount of Free Care 33 A Devaluation of the Country's Currency 34 A Change in the Population Distribution 34 5. HOSPITAL CARE 35 Main Features of the Hospital Component of the Model 35 Inpatient Services 35 Bed-Days as a Measure of Utilization 35 Hospital Beds Needed 36 Administrative and Support Personnel 37 Medical Personnel 37 6. AGGREGATE PROJECTIONS AND CONCLUSIONS 39 Conclusions 39 APPENDIXES Technical Appendix A: Computation of Weighted Average Distance between the Population in Ring r and a Competitor 41 Technical Appendix B: Demand Equations and Elasticities 42 Appendix C: Menu Structure of Computer Model 44 Appendix D: Computer Model, Health Center Component 45 Appendix E: Computer Model, Hospital Component 58 Appendix F: User's Manual 66 Appendix G: Additional Simulation Results 73 BIBLIOGRAPHY 78 EXHIBITS Exhibit 4-1 Health Center, Reduced Menu, Base Scenario 22 Exhibit 4-2 Price, Income, and Distance Elasticities of Demand, Health Center and Competitors 24 Exhibit 4-3 An Increase in the Price of Curative Visits at the Health Center 25 Exhibit 4-4a Health Center's Expanded Menu---A Reduction in a Competitor's Price of Curative Care 26 Exhibit 4-4b A Reduction in a Competitor's Price of Curative Care 27 Exhibit 4-5 Cross-Subsidization of Preventive by Curative Care-Prices and Quantity Demanded 28 Exhibit 4-6 Comparison between a Single-Price Policy and a Policy of Price Discrimination Based on Income, Curative Ambulatory Care 29 Exhibit 4-7 Comparison of a Policy of a Single Price and a Policy of Differential Prices According to Distance to the Health Center---Curative Ambulatory Care 30 Exhibit 4-8 A Drop in the Insurance Premium Coupled with a Rise in the Proportion of People Insured 31 Exhibit 4-9 An Increase in the Insurance Copayment 32 Exhibit 4-10 A Rise in the Proportion of People Receiving Free Care at the Health Center 33 Exhibit 5-1 Hospital, Reduced Menu, Base Scenario 36 Exhibit 6-1 Aggregate Projections 39 FIGURES Figure 2-1 Interaction between Demand and Supply and the Effect on a Health Facility's Financial Performance 4 vi Contents Figure 2-2a Public Facility Pricing with Break-Even Constraint and Declining Average Cost 7 Figure 2-2b Case 1: Single-Product Facility--Two Break-Even Solutions Exist 7 Figure 2-2c Case 2: Single-Product Facility---Break-Even Not Possible, External Subsidy is Needed 7 Figure 3-1 Market for Ambulatory Care 14 Figure 3-2 Consumer Decision Tree 14 Figure 3-3 Demand by the Insured Who Pay a Copayment C (C< 1) and the Uninsured Who Pay the Full Price 17 Figure 3-4 Demand by the Nonpaying with and without Rationing 17 Figure 3-5 Labor Costs Modeled as Step-Fixed Costs 19 Figure A-1 Computation of Distance between Population in Ring r and a Competitor 41 FOREWORD Developing countries are turning increasingly to cost recovery to pay for health care services; but the decisionmakers who must develop and implement systems for cost recovery are often unfamiliar with health care financing issues. In recent years, economists have begun to develop computer models to help decisionmakers formulate policy in this field. The model discussed in this manual is a training tool that uses information about the demand and supply of health services to help make projections about the effects of health care financing reform on the utilization and the financial performance of the health system. It is intended to help teach the broad interrelationships of certain health policy variables and to assess their impact. In itself the model does not pretend to be complete enough to be used as a decisionmaking tool. The model may be used to explore the implications of policies such as charging uniform user fees to all; charging differential fees based on income; implementing cross-price subsidization between curative and preventive services; exempting certain patients from payment; and establishing different levels of insurance premium and co-payment. It can also be used to assess the effects on utilization and performance of changes in a number of nonfinancial parameters such as the geographic distribution of the population relative to the providers; the distribution of income; the crude birth rate; population size; the incidence of acute illnesses; and the prevalence of chronic conditions. This user-friendly computer model is flexible enough to allow the user to enter data on demand and costs from various developing country settings and to simulate various changes in health care financing under a broad range of circumstances. It has been used as a training tool in several seminars organized by the World Bank's Economic Development Institute. Amnon Golan Director Economic Development Institute vii a PREFACE In developing this model I have tried to maintain a balance between completeness and simplicity. A complete model is one that addresses all the important aspects of the phenomenon being studied and thus simulates reality as closely as possible. Greater completeness often implies greater complexity; models grow bigger and more intricate. Yet this complexity comes at the cost of making the model inaccessible to the intended users. In the interest of keeping the model as transparent as possible, I have made several simplifying assumptions in the present version. About half of this paper is devoted to explaining in detail the assumptions and theory behind the model. The second half presents examples to illustrate how the model can be used to simulate alternative policy options and scenarios. Space limitations have allowed us to provide only a handful of examples. Users are, of course, encouraged to perform additional simulations. The most important omission from the model is the qualitative side of health care services and its effect on demand. Leaving health care quality out of the model, however, is consistent with our current lack of empirical knowledge about how consumers view health care quality in poor countries and how their perceptions affect their health-care-seeking behavior. I have implemented the model in Lotus 1-2-3 using that program's macrolanguage. While testing the model, several programming bugs have surfaced. Although the current version appears to be problem-free, it is possible that some defects still remain. If so, please contact me and I will try to correct them. Ricardo Bitran ix ACKNOWLEDGMENTS The development of this model has been made possible thanks to the financial support of the Resources for Child Health Project (REACH), the Zaire-based Basic Rural Health Project (SANRU), and the Economic Development Institute (EDI) of the World Bank. I wish to thank David Dunlop, from the World Bank, and Gerald Rosenthal, from REACH, for their encouragement and for being patient enough to sit through numerous presentations of the model. I thank Frank Baer, Manunga Mapele, Steven Brewster, and Munkatu Mpese, from SANRU, for their helpful ideas and suggestions. The field data used to develop the model were collected in Zaire from the health zones of Kisantu, in Lower Zaire, and Bokoro in the Bandundu Region. I am grateful to Dr. Kamba, from Kisantu, and Dr. Yves Heyligers, from Bokoro, for their enthusiastic collaboration. Helpful ideas were provided by Elca Rosenberg, Margaret Saunders, Mead Over, Jacques Baudouy, Ed Elmendorf, and Jean Louis Lamboray from the World Bank; by Drs. Paul Cartier and Mark de Freyder, from the Belgian Cooperation in Zaire; and by Glenn Post, Rhonda Smith, and William Bertrand, from USAID/Kinshasa. Leonardo Bitran helped me with valuable mathematical insight. David Deal, of Abt Associates, did an excellent job making an earlier version of this model user-friendly. xi s 1 INTRODUCTION Developing countries are turning increasingly toward cost recovery, particularly user fees, to pay for health care services. Decisionmakers in those countries face the difficult task of developing and implementing cost-recovery systems. Unfortunately, those who decide are often unfamiliar with health care financing issues and lack the necessary skills to design pricing systems and determine price levels. In recent years health economists have begun to develop computer models of health care financing as an analytic tool for decisionmakers. Most models developed focus on either demand or supply, but few take into account both sides of the market. Mwabu (1984) and Gertler, Locay, and Sanderson (1987), for example, use demand information to project demand as a function of price, income, and other variables that affect demand. Their models, however, incorporate no information about the providers' cost structure, and thus say nothing about the implications of demand and price projections on the providers' financial performance. Other models use information about the providers' costs in their model but do not include explicitly any behavioral model of the consumer. Cost and revenue projections are thus blind to possible demand responses to prices and other variables. Inspiration for the present model came from earlier modeling efforts by Makinen and Block (1986) and Block, Donaldson, and Foster (1988). Although their models were mainly supply-side-based, they included a single, constant-elasticity demand function to account for price responses. The main property of the present model is its ability to combine information about demand and supply. Some of the model's innovative features include: the interaction between the demand for health care and the financial performance of health facilities; the explicit inclusion of competition in the health care market and its effects on demand and cost recovery; the ability to model demand by different groups of individuals, including the insured, the uninsured, and the indigent; the inclusion of several kinds of health care services as different health facility products, including curative, preventive, ambulatory, and inpatient care; and the inclusion of a wide range of parameters that enable the user to represent a vast set of situations. Model parameters include prices of health care services; population income and geographic distribution; epidemiological and demographic data; input costs; exchange rate; and size of the market. Users of the model can assess the effect on health care use and health facility financial performance of health financing policies such as: cross-subsidizing prices among any set of services in general and between curative and preventive care in particular; extending health insurance coverage; modifying the premium and copayment of the insurance plan; changing the free-care policy; changing price levels; and price discrimination (for example, based on income or distance to the facility). The model can also be used to study the effects of changes in other variables on demand and cost recovery performance, including: a devaluation or revaluation of the country's currency; changes in the degree of competition in the market for a particular health service; and labor substitution within the facility among different categories of health professionals. The intended readership of this paper and model includes health economists and health financing analysts with a background in basic microeconomic theory. It is hoped that those individuals will use the model as a pedagogical tool with students, technicians, and decisionmakers in developing countries. 1 2 A Supply-Demand Model of Health Care Financing The model has been developed in Lotus 1-2-3, is contained in four spreadsheet files, can be run on any IBM-compatible microcomputer with at least 640 kilobytes of RAM memory, and has been made user-friendly through a menu system written in Lotus' macroprogramming language. Chapter 2 presents an overview of the model, outlining its assumptions, capabilities, and limitations. Chapter 3 provides an in-depth description of the model's assumptions for the case of ambulatory care, both curative and preventive. Chapter 4 describes the computer implementation of the model and illustrates its use through a series of simulation exercises for the case of ambulatory care using field data collected from Zaire. Chapter 5 describes the model's inpatient component. Chapter 6 explains the model's aggregate projections feature, provides a summary and conclusions, and outlines areas for research. 2 OVERVIEW OF THE MODEL The market for health care services comprises consumers and producers. Consumers are characterized by their number, income, demographic traits, epidemiological data, and geographic location in relation to all providers. Providers are classified into two groups. Group one contains a single provider or focus health facility, which can be thought of as being owned by the government or private, and whose utilization levels and financial performance are studied in detail through the model. The second provider group contains all the other providers that compete with the focus health facility; these are referred to as competitors. Each provider can supply several kinds of health services (for example, curative ambulatory care, obstetric care, preschool care, inpatient care, and the like). The model computes demand for care from each competitor but does not assess the competitors' financial performance. This model is concerned with two categories of indicators: the quantity of health services demanded by the population and the financial performance of the focus health facility. It is assumed that neither the focus health facility nor its competitors are supply constrained. This means that each individual provider can meet any demand level by increasing production. In the case of the focus health facility, the model automatically computes, and shows explicitly, the production inputs required to meet demand. Chapter 6 discusses possible strategies for relaxing this assumption and their implications. The absence of supply constraints implies that, for each individual provider, demand is always equal to utilization. Utilization is the health system's output and is a proxy for the system's health outcome that consists of the deaths and illnesses averted or cured. The focus facility's financial performance is measured by its revenue from cost recovery, its costs, and the associated net income (or profit), which is the difference between total revenue and costs. The model has three additional limitations. First, it does not provide any direct measures of welfare. Users are expected to judge on their own the merits of alternative scenarios. Second, because of the lack of empirical knowledge about the way health care quality is perceived by consumers and how it affects demand, the model does not include any explicit measures of health care quality and its effect on demand.' Third, although the model automatically adjusts inputs to meet demand, it does not have any other built-in rules of provider behavior. This means, for example, that the model does not have cost-minimizing routines or endogenous rules about the response of a given provider to changes in its competitors' prices. The model's main feature is its ability to take account simultaneously of the relationship among: Demand determinants (for example, out-of-pocket price, household income, population size and distribution) and demand 1. Strategies for relaxing this assumption are discussed in chapter 6. 3 4 A Supply-Demand Model of Health Care Financing * Demand and focus health facility revenue * Demand and focus health facility costs. Figure 2-1 provides a simplified description of the model's logic. At the top of the figure are a series of exogenous, or user-defined, variables that have been shown to influence demand for health care (M = population size; G = population geographic distribution; I = illness incidence; Y = population income; and P = price).2 With this information the model computes demand for the focus health facility (Q) and for each of its competitors. Once Q is known the model calculates the focus health facility's total revenue (TR) by multiplying demand (Q) times price (P). At the same time, demand, which is equal to the amount of services produced by the facility, determines the facility's total cost (TC). As shown in the figure, however, cost not only depends on Q, but is also a function of other factors such as the exchange rate (FX), which directly affects the costs of drugs and fuel; the cost of labor (L); and other variables. The difference between total revenue and total cost is the facility's net income (NI), or profit. Figure 2-1. Interaction between Demand and Supply and the Effect on a Health Facility's Financial Performance Q = f(M,G,1,Y,P) = quantity demanded Total revenue Total cost TR = P.Q TC = C(Q,FX,L,.) Net income Price (P) NI = TR-TC increase NI <0 NI=0 N> (loss) (break-even) (rft To understand the model's logic, suppose that it predicts a negative net income (NI < 0 in the figure). If the aim of the user is to achieve break-even, he or she might consider, say, a price increase.3 Everything else being constant, a higher price will translate into a lower quantity demanded and, given the existence of variable costs, into a lower total cost. A higher price times a lower quantity demanded may result in either a higher or lower revenue, depending upon the magnitude of the demand response to the price change, that is, consumers' price elasticity of demand, ep. Thus, the effect of a price change 2. Recent empirical studies of health care demand in developing countries include Gertler, Locay, and Sanderson 1987 and Dor, Gertler, and van der Gaag 1987. 3. Other measures are also possible. Overview of the Model 5 on focus facility financial performance cannot easily be anticipated. A complete discussion of the effects of a price change on total revenue, total cost, and net income is provided in the section entitled "Comparative Statics, " below. Demand is modeled through a two-stage probabilistic demand equation using McFadden's nested logit specification (McFadden 1981). This specification of demand has been widely used in the empirical literature on health care demand in recent years.4 According to the nested logit specification, in the first stage the model computes the proportion of people who seek care from the focus health facility and from each competitor among those who decide to seek care outside the home. In the second stage, the model uses the first-stage information to compute the proportion of people with a medical problem who decide to seek care outside the home. Thus, the demand for health care from provider i for each unit of time (Qi) includes the product of two probabilities or proportions as follows: (2-1) Qi = H - Prs Pri where H is the number of people with a health problem for each unit of time (for example, one month), Prs denotes the proportion of people seeking care outside the home among those with a health problem, and Pri is the proportion of people who choose provider i among those who seek care outside the home. Both Prs and Pri are assumed to be functions of the out-of-pocket price faced by the consumer, the travel distance to the provider, and the consumer's income.5 Linear (in utilization) total cost functions have been used to model the costs of producing and providing health services. The model includes both fixed costs (such as depreciation of facilities and equipment and salaries) and variable costs (for example, pharmaceutical products and fuel). The total cost (TC of the health facility can therefore be written as: (2-2) TC = FC + MC*Q where FC is fixed costs (those that do not vary with the level of production Q) and MC is marginal cost or variable cost for each unit of output. For simplicity, MC is assumed to be constant. The model's logic can be presented more formally as follows: let Q (P, Z) be the quantity of services demanded by the population from the focus health facility where P is the facility's price and Z represents all the other model-exogenous variables affecting demand, such as competitors' prices and location, size of the population, illness incidence, and so on. The total revenue of the facility, denoted by TR, can be expressed as follows: 4. See, for example, Gertler, Locay, and Sanderson 1987, Dor, Gertler, and van der Gaag 1987, and Bitran 1990. 5. Prs and Pri are obviously a function of many other variables, such as education, type of health problem, and others. In order to keep the model simple, however, only the three variables mentioned in the main text have been included explicitly in the demand function. 6 A Supply-Demand Model of Health Care Financing (2-3) TR = P Q (P, Z). The model's main relationship can thus be represented through the following reduced-form expression: (2-4) NI = TR - TC or NI=P-Q(P,Z)-FC-MC Q(P,Z). Expression 2-4 allows one to understand the main forces underlying the model. The next section is a discussion of possible pricing strategies for the focus health facility; it is followed by a discussion about strategies for and feasibility of breaking-even. The next section comments on the size of the facility's loss, if any, and external subsidization; we then briefly discuss the model's comparative statics. Pricing How much should the focus health facility charge for its services? A well-known tenet of welfare economics is that to maximize welfare in a perfectly competitive market selling private goods and with well-informed consumers, price should be equal to marginal cost. In the health care market price may depart optimally from marginal cost for several reasons. First, the presence of externalities in the consumption of some health care services (for example, immunizations) may warrant an optimal price that is below marginal cost. Second, if the service produced is a merit good, a price below marginal cost may be desirable. Third, that price equals marginal cost is a result that is oblivious to equity considerations. If price is set at marginal cost for all, the poor will demand less (per capita) than the rich, a situation that is inequitable and that may be morally unacceptable. Given different income groups and a fixed subsidy, the price of a product may be set differentially by income above or below marginal cost. Third, in the absence of external subsidies, or in the presence of a small (that is, less than fixed costs) subsidy, a facility with constant marginal cost will be unable to break even if it sets its price equal to marginal cost. As is shown in figure 2-2a, this facility will operate under declining average costs and marginal cost will be less than average cost. The need to break even thus forces the facility to depart from the marginal cost pricing rule, as is discussed below. A popular welfare economics result is that in such a case the facility will charge a price that is proportional to the marginal cost and inversely proportional to the price elasticity of demand. Thus, if there are, say, two population groups, the rich and the poor, in order to break even the facility may charge a higher price to the rich and a lower one to the poor.6 6. This result for pricing parallels the Ramsey principle for taxation. It is assumed that the price elasticity of demand for the focus facility services by the rich will be lower than that of the poor since the focus facility services are assumed to be normal goods. For a discussion of optimal pricing under a budget constraint see Atkinson and Stiglitz 1980. Overview of the Model 7 Figure 2-2a. Public Facility Pricing with Break-Even Constraint and Declining Average Cost7 P AC Marginal ost MC pIricing Demand QA OB Break-Even Analysis Given a fixed external subsidy, or no subsidy at all, will the facility be able to break even?8 Consider, for simplicity, a facility that produces a single service such as curative ambulatory visits. Two possible cases arise. In the first case, illustrated in figure 2-2b, the demand curve for visits crosses the facility's average cost curve at two points, (Qj, Pj) and (Q2, P2)- Figure 2-2c. Case 2: Single-Product Figure 2-2b. Case 1: Single-Product Facility---Break-Even Not Possible, Facility---Two Break-Even Solutions Exist External Subsidy is Needed Price, Price Average Cost Average Cost P1P PI ~AC " P Subsidy A P2 - AC \D AC Q1 Q2 Quantity Q Quantity 7. From Atkinson and Stiglitz 1980, p. 463. 8. With an external subsidy, breaking even is defined here as the ability to generate sufficient revenue to cover total costs net of subsidy. 8 A Supply-Demand Model of Health Care Financing Both points meet the break-even condition because total revenue (price times quantity demanded) equals total cost (average cost times quantity demanded). The facility chooses point (Q2, P2) because it prefers to deliver more (Q2) rather than less (Qj) services, as long as it breaks even.9 In the second case, depicted in figure 2-2c, the demand curve lies below the average cost curve, and thus break-even cannot be achieved. At any price P the facility needs an external subsidy, equal to the shaded area, to remain in business. This analysis can be extended to the case of a multiservice facility with two differences: (1) the search for a break-even point may involve changes in the prices of several services and (2) there may be a multitude of price combinations that satisfy the break-even constraint. More formally, suppose that the facility produces and sells n services at prices (PI, P2, . . . Pn). A break-even solution is a set of prices (P P2*, ,P*) that satisfies the condition NI = 0. Dependin on supply and demand conditions, there may be S such solutions, each denoted by (P,*, P2* . P, )s, for s = 1 to S. To illustrate the practical relevance of this discussion, consider the example of a facility that chooses to price preventive services slightly below average cost to promote consumption of preventive care (because of positive externalities or misinformation about consumption benefits) and that, to compensate for this subsidy while breaking even, it prices curative services above average cost. This same facility could also break even by providing preventive services at no charge while charging prices much higher than average cost for curative care. Depending on the health goals of the provider, it could choose between these two options; each would allow the facility to break even. As in the single-product case, it is possible that a break-even solution may not exist; that is, there is no combination of prices (PI, P2, . . . ,P,) that will satisfy NI = 0. Nevertheless, as already mentioned, it is likely that if break-even is at all possible, there may be multiple solutions. Welfare considerations should determine which set of prices is the most desirable. The model does not automatically calculate break-even points but allows the user to search for solutions with ease through successive iterations. In many cases the user will fix a few prices to reflect policy goals (for example, zero or low prices for certain chronic and preventive care) and will search for prices for the other services (such as curative care) that may allow it to break even. Health Facility Loss and External Subsidization Should the health facility be allowed to take a loss if it is a government-owned facility? Suppose that losses are allowed for two reasons. First, for equity reasons the government decides to subsidize the facility so that the poor can be charged subsidized prices, thereby increasing the financial accessibility of services. Second, the government recognizes the existence of positive externalities in the consumption of certain preventive services, such as immunizations. These externalities may accrue to society beyond the facility's market, and thus a subsidy is warranted. To simplify, suppose that the government has a fixed subsidy to allocate among several facilities; that is, the government has already allocated subsidies 9. The implicit assumption that leads to preference for point (Q2, P2) over (Qj, Pj) is that consumption of service at level Q1 is below the socially desired level of consumption; that is, at level Q, there is underconsumption of the service. Overview of the Model 9 among competing sectors (health, education, and so forth). How should it go about deciding how much to allocate to each facility? A precise answer to this question is beyond the scope of this study, but it is safe to say that subsidy allocation criteria will depend on equity considerations, externalities, and size of the market. Comparative Statics The remainder of this section explores how changes in model-exogenous variables affect demand and focus health facility financial performance. A Change in the Price of the Focus Health Facility The effect of a change in the price of the focus facility on its total revenue depends on the size and the direction of the price change, the magnitude of the price elasticity of demand, and the relationship between the facility's price and its marginal cost (MC), as shown in table 2-1. Table 2-1. Effect of a Focus Facility Price Change on the Facility's Total Revenue, Total Costs, and Net Income Price Elasticity of Demand for Focus Facility Care, EP EP < -1 eP = -1 -1 < EP < 0 Ep = 0 Meaning Demand is elastic: quantity Demand is unit- Demand is in- Demand is demanded changes propor- eLastic: elastic: perfectLy tionateLy by more than price quantity quantity ineLastic: demanded demanded quantity changes in the changes demanded same proportion proportionately does not as price less than price change with price Price is TR goes DOWN TR UNCHANGED TR goes UP TR goes UP increased TC goes DOWN TC goes DOWN TC goes DOWN TC UNCHANGED NI UNCHANGED if MC/P = 1+1/EP NI goes UP NI goes UP NI goes UP NI goes UP if MC/P > 1+1/Ep NI goes DOWN if MC/P < 1+1/E_ Price is TR goes UP TR UNCHANGED TR goes DOWN TR goes DOWN decreased TC goes UP TC goes UP TC goes UP TC UNCHANGED NI UNCHANGED if MCIP = 1+1/6p NI goes DOWN NI goes DOWN NI goes DOWN NI goes DOWN if MC/P > 1+1/cp NI goes UP if MC/P < 1+1/ep Note: TR = totaL revenue; TC = total costs; NI = net income. If demand is unit-elastic (e = -1) the price increase will result in an unchanged total revenue, lower total costs, and higher net income. If demand is inelastic (c-P > -1), the price increase will bring 10 A Supply-Demand Model of Health Care Financing about an increase in total revenue, a drop in total costs, and a higher net income. If demand is elastic (c- < -1), the price increase could result in either an increase or a decrease depending on both the magnitude of the elasticity and the relationship between the facility's marginal cost and its price.o Finally, if demand is perfectly inelastic with respect to price, a price increase (decrease) will result in a higher (lower) revenue. A Change in a Competitor's Price A competitor's health care services are assumed to be substitutes for those of the other competitors as well as for those of the focus health facility. Changes in the price of one competitor affect the demand for and the financial performance of all providers." For example, if a competitor raises its price, two simultaneous effects take place. First, fewer people seek care in the market (Prs in expression 2-1 goes down). Second, among those seeking care, a smaller proportion chooses the provider that raised its price. If the drop in the proportion of people seeking care outside the home is small, the providers other than the one who raised its price, including the focus health facility, may find themselves facing a higher demand. If the focus health facility's price is greater than its marginal cost, this higher demand will result in an improvement in the facility's net income. Otherwise, if P < MC, the facility will need a larger external subsidy (or will need to raise the price of other services, or take other measures) to meet this greater demand while balancing its finances. A Change in the Number of Competitors If a new competitor enters the market of health care services, some share of the demand for the services of the focus facility may be diverted toward the new producer. The effect on utilization and NI of the focus health facility will parallel the effect of a price reduction on the part of a competitor. Conversely, if one or more competitors exit the market, some patients may be diverted toward the focus facility, increasing Q, TR, and TC. This effect will be similar to that of a price increase by the competitors in the direction of the changes in TR, TC, and NI. A Change in the Illness Incidence Rate In the case of curative care, illness incidence is obviously an important determinant of demand. Changes in illness incidence will result in proportionally identical changes in demand for all providers. 10. Using expression 2-4 for NI, and the definition of price elasticity of demand, Ep = (aQ/P) . (P/Q), one can easily show that the expression for a change in net income as a result of a small price change is 8NI MC =p Q -[1+ E*(10- )]A. Using this last relationship, it is easy to show that net income will move in the same direction as price if eP 2 -1. If ep < -1 the change in net income induced by a small change in price will depend on the magnitude of eP as well as on the relationship between MC and P as is summarized in table 2-1. 11. This is true, of course, if the demand for that provider's services is not perfectly inelastic. Overview of the Model 11 If the focus health facility's price exceeds its marginal cost, a higher illness incidence will translate into a higher net income; otherwise it will result in a greater loss. A Change in Population Size A higher (lower) population will bring about a proportionally higher (lower) demand from all providers. Its effect on the direction of the change in Q, TR, and TC will be equivalent to that of a change in the illness incidence rate. A Change in Household Income Health services of all providers are assumed to be normal goods, which means that, other thins being equal, changes in household income will result in changes in the same direction in demand. As in the preceding cases, the effect on the focus facility's NI will depend on the relative magnitude of MC and P. A Change in the Geographic Distribution of the Population Travel time to a health facility has been shown to have a negative effect on demand.13 If the average distance from a representative household to the focus health facility increases, so will travel time. Demand for care from that facility will thus go down, affecting both TR and TC.14 In contrast, if a higher number of people is located in the vicinity of the focus facility, demand will be higher. The model parametrizes the population distribution to study its effect on demand and health facility performance. For example, through the model it can be seen that, everything else being equal, health providers located in areas of high population density have a greater ability to be financially self-sustaining than those located in sparsely populated areas. Changes in Costs A devaluation (revaluation) of the currency, manifested by a higher exchange rate, translates into higher (lower) fixed and marginal costs expressed in local currency. Everything else being constant, a devaluation unambiguously increases total cost and reduces NI, while a revaluation has the opposite effect. Changes in other fixed or variable costs will affect total costs and net income. 12. Normal goods are those that are demanded in greater quantity as people's incomes increase. 13. For example, see Dor, Gertler, and van der Gaag 1987. 14. Some may argue that a greater distance may not necessarily translate into a higher travel time because more distant consumers choose faster means of transportation. While that is possible, it is assumed here that people's travel speed is constant regardless of their location. Hence a greater distance is assumed to result in a longer travel time. 8 3 AMBULATORY CARE The last section provided an overview of the model's logic. This segment describes in detail the ambulatory care component of the model. Structure of the Market The market for ambulatory health care services, both curative and preventive, is composed of consumers and producers. Consumers are characterized by their number, income, demographic traits, epidemiological data, and geographic location in relation to the providers. Providers consist of a health center (focus health facility), with utilization levels and financial performance that are studied in detail through the model, and its competitors. The health center provides five kinds of ambulatory services: * curative care for people with acute health problems * curative care for people with chronic problems * preschool preventive care for children under five years of age, including immunizations * prenatal preventive services for pregnant women * obstetric services. Competitors can also provide all or some of these services. Figure 3-1 depicts the market for ambulatory care. The health center is located at the center of an imaginary circular region (referred to as the health center's market). Its competitors, as well as the consumers, are located within the region. The health center and the competitors are characterized by their prices. In addition, each competitor is characterized by its distance from the health center. The user of the model must specify the radius of the market, the providers' prices, and the distance from each competitor to the health center. The population is assumed to be distributed uniformly within five population rings, as shown in figure 3-1. The user of the model must specify the total population living in the market as well as the percentage living within each of the five population rings in order to allow simulations based on different assumptions about the population density and the travel distance. Based on the location of the competitors 13 14 A Supply-Demand Model of Health Care Financing and the distribution of the population, the model computes the average distance between the population living in each population ring and each competitor.1 Figure 3-1. Market for Ambulatory Care Population Market rings DIK PK Radius * = Health Center P = Price m = Competitor D = Distance Health Care Demand Curative Care People with a health problem must decide whether or not to seek care outside the home; those who decide to seek outside care must choose a provider. The consumer's decision tree is illustrated in figure 3-2. Figure 3-2. Consumer Decision Tree - Health center Seek - Competitor 1 care _ Competitor 2 Person with a Competitor J health problem Do not seek care The number of people in population ring r choosing provider j for a given health service (for example, curative care) (Qrj) can be computed by multiplying the number of people with a health problem within a given time period living in r (Hr) times the proportion (or probability of) seeking care (ProbSeek)r times the proportion seeking care that chooses that particular provider (ProbChoose)d. Q4 1. See Technical Appendix A for a derivation of an expression to compute the travel distance. Ambulatory Care 15 can be viewed as the demand for first visits in the case of curative care, or as the number of deliveries in the case of obstetric care. (3-1) Qrj = Hr -(ProbSeek)r -(Probchoose)rj. The parameter Hr, which denotes the number of people in r who experience a health problem within a certain time period (for example, one month), is in turn equal to the number of people living in r (Mr) times the annual illness plus injury incidence rate. The total demand for first visits from provider j is therefore equal to the sum of the demand across all five population rings. 5 (3-2) Q. = EiQr r=1 Once the patient chooses a provider, the quantity of visits demanded from provider j by the people in ring r is assumed to be a linear function of the out-of-pocket price of a visit (P), the distance to the provider (Dj), and the income of the patient (Y'j).2 According to figure 3-2, each consumer faces a set of discrete choices (no care, care from the health center, care from competitor 1, and so forth). It is assumed that individuals rank the choices according to the amount of utility that they can obtain from consuming the services of each provider. Utility, as understood by economists, is an individual-specific measure that reflects the amount of satisfaction that individuals draw from the consumption of various goods and services. This model assumes that individual utility is a function of the option's price (Pj), the individual's travel distance (Drj), and his or her household income (yr)-3 An interaction term equal to the product of price times income, P. Yr, has been included to permit different price responses by income group. Each consumer chooses the option that brings him or her the highest utility. Both P. and Drj reduce utility, and thus have negative coefficients. The coefficient on P. Yr is positive to be consistent with the assumption that health care is a normal good, as well as with the lower sensitivity of higher- income people to price changes. If VrJ denotes utility of a representative individual living in population ring r associated with optionj, then 2. Dor, Gertler, and van der Gaag (1987) have used a similar specification for the quantity of visits demanded in their study of demand in CMte d'Ivoire. These authors first model the decision to enter the market by the people with a health problem. Then, conditional on seeking care, they model the choice of provider. Finally, among those choosing a particular provider, they estimate a linear demand equation for the quantity of visits. 3. This is indirect utility. 16 A Supply-Demand Model of Health Care Financing (3-3) Vj = 0oj + 013'j + 02'pj'yr + 03 -Dr where P = consumer's out-of-pocket price at provider j's facility Yr = household income of people living in population ring r Drj = weighted average distance between the people in population ring r and provider j 0j = constant that captures the effect on consumer utility of all the variables affecting utility other than price, income, and distance (quality perceptions, for example, are imbedded in this coefficient) 01 = coefficient associated with the variable price (1 < 0) 02 = coefficient associated with the product of the variables price and income (02 > 0) 03 = coefficient associated with the variable travel distance (03 < 0). Technical Appendix B shows how health care demand is derived as a function of consumer utility. Briefly, in order to compute demand, the model expresses both (ProbSeek)r and (ProbChoose),, as a function of utility (Vrj), as defined above.4 Preventive Care The demand for preventive care services is similarly specified, except that the relevant population groups are either children under five (for preschool care) or pregnant women (for prenatal care) instead of people with a health problem. Consumer Categories Three categories of consumers are included in the model: the uninsured people who pay the providers' full price; the people who are insured at the health center through the payment of an annual insurance premium and who pay a copayment (that is, a proportion of the full health center price) when obtaining care at the health center; and the uninsured people who are given free care at the health center, referred to as the nonpaying. The user must specify the percentages of the population that fall within each of these three groups. It is assumed in the model that those percentages are the same across population rings. Using equations 3-1 through 3-3 and the equations for ProbSeek and ProbChoose of Appendix B, the model computes demand for each service and provider by each category and population ring. The model allows for household income to differ between the insured and the uninsured. Generally, the former have higher income than the latter. 4. See Technical Appendix B for a derivation of the expressions for the demand equations. Ambulatory Care 17 The demand by the uninsured who pay the full price is computed through the model by plugging into the demand formulas the full price of each provider, the household income, and the average household travel distance. The demand by the insured people is computed similarly, although a lower out-of-pocket price, the copayment, is entered as the health center price. Thus, at the same household income level, the demand for health center care by an insured person will be greater than that of someone without health insurance, as shown in figure 3-3. Finally, demand by the nonpaying is assumed to be rationed by the health center personnel (for example, through longer waiting time) to a level equal to that of the paying, on a per capita basis (see figure 3-4). Figure 3-3. Demand by the Insured Who Pay a Copayment C (C< 1) and the Uninsured Who Pay the Full Price Price P D insured C-P uninsured Quantity Figure 3-4. Demand by the Nonpaying with and without Rationing Price (P) Demand by paying Demand by non-paying P with without rationing rationing / Quant. (P=P) (P=O) Dem. 18 A Supply-Demand Model of Health Care Financing Health Center Finances This section explains how health center revenue and costs are computed in the model. Health Center Revenue Monthly health center total revenue (TR) is computed by adding up annual health insurance premium payments; copayments, if any, made by the insured; and direct payments made by the paying uninsured. Thus: (3-4) TR = Prem + P-QU + P*C*QI 12 where Prem is annual premium payments by the insured (Prem = M . i . R, where M is the total population of the market, i is the proportion of people insured, and R is the annual insurance premium), P is health center full price, QU is monthly demand for health center care by the uninsured, C is the copayment rate (for example, C = 0.2 means that the insured must pay 20 percent of the health center's full price when obtaining care), and QI denotes monthly demand for health center care by the insured. Health Center Costs Total monthly health center costs are the sum of fixed and variable costs. Fixed costs include labor, depreciation of buildings, equipment, vehicles, office supplies, refrigerator fuel, and other miscellaneous fixed costs. Variable costs consist mainly of drugs and medical supplies and supervision fees. FIXED COSTs: LABOR, DEPRECIATION, AND FUEL Labor costs are modeled as step-fixed costs; that is, costs that are fixed within a certain range of output but that increase or decrease to a new plateau as output expands or contracts. Health center labor costs are depicted in figure 3-5 as a function of the amount of health services produced by that provider. Labor costs are computed with the model by type of health service (deliveries, for example) and by type of medical provider (such as a nurse). For example, to compute monthly nurse costs associated with deliveries, the model proceeds as follows. First, it computes the daily demand for deliveries. Second, based on the amount of time spent by a nurse per delivery, it computes the total number of nurse-minutes required daily to satisfy demand. Third, based on the number of hours worked daily by a nurse, it computes the number of nurses required daily to satisfy demand. If the number of nurses is not an integer, the figure is rounded up (for example, if 4.4 nurses are required the model assumes that 5 nurses are needed). Finally, by multiplying the results times the monthly salary of a nurse, the model obtains the monthly nurse labor costs for deliveries. This type of calculation is done by the model for all five services provided by the health center and for the following personnel categories: medical doctor, nurse, midwife, and laboratory technician. Ambulatory Care 19 More formally, monthly health center labor costs of health personnel type m (nurse) for health care service i (LCim) (deliveries) are computed according to the following formula: (3-5) LCim = INTG( ) -Sm Tm 6 where INTG denotes the integer function, Qi is daily demand for health center service i, tim is the time (in minutes) spent by employee type m producing one unit of service i, Tm is the total number of hours worked by employee type m in one day, 60 is the number of minutes in one hour, and Sm denotes the monthly salary of personnel m, expressed in local currency. Figure 3-5. Labor Costs Modeled as Step-Fixed Costs Labor cost Quantity produced Total monthly labor costs of the health center are computed by calculating the sum of labor costs across the five health care services and the four types of medical personnel: 54 (3-6) LC = LCim. i=1 m=1 Although labor is treated as a fixed cost in the above specification, it could easily be converted into a variable cost by eliminating the integer function from the above formula. In such a case, the number of minutes of each category of employee would continually be adjusted up or down with demand. The model also requires specification of the types of investments made at the health center, including buildings, medical equipment, and vehicles, as well as their years of useful life and the initial 20 A Supply-Demand Model of Health Care Financing investment, specified in U.S. dollars.5 The monthly depreciation cost of asset a (DCa) is computed in local currency, using straight-line depreciation, according to the following expression: (3-7) DC F a L 12 where la is the investment cost of asset a, expressed in U.S. dollars, FX denotes the current exchange rate (units of local currency per US$), L is the useful life of the asset (in years), and 12 is the number of months in a year. Fuel is used to power refrigerators and to sterilize vaccination equipment. For simplicity, these fuel costs are treated as fixed in the model. The user is simply required to specify the number of liters of fuel used monthly and the price of a liter, expressed in U.S. dollars. VARIABLE COSTS: DRUGS AND MEDICAL SUPPLIES In order to compute variable costs for drugs and medical supplies, the user of the model must specify the average cost to the health center of a typical prescription, laboratory exam, and other pharmaceutical products associated with each type of health service. The user must also specify the amounts of prescriptions, exams, and other supplies used in each type of intervention. The monthly variable labor costs of drugs and medical supplies for service i (for example, curative visits), ECi, is computed by multiplying the average cost of each production input times the quantity used for each category of service or: (3-8) ECi = (Gi - ACG + Xi * ACX + Oi * ACO) * Qi where Gi, X,, and Of are the number of prescriptions, laboratory exams, and other supplies, respectively, used to produce one unit of service i; ACk is the average cost to the health center (expressed in local currency) of a prescription (k = G), laboratory exam (k = X), and other medical supplies (k = 0); and Qi is monthly demand for service i. 5. The market value of many investments is tied to the rate of foreign exchange. Investments must be specified in U.S. dollars even if they were paid for in local currency. This is done to adjust upward or downward the value of the investments in order to reflect their real replacement value. 4 DEMAND, COSTS, AND HEALTH CENTER FINANCIAL PERFORMANCE: SIMULATIONS Chapter 3 was an explanation of the model's logic and basic assumptions. This chapter briefly describes the structure of the computer model and provides examples, for the case of ambulatory care, to illustrate some of the model's capabilities. The examples use case study data collected in 1987 by the author and a team of researchers (sponsored by the REACH and SANRU projects) in the Kisantu and Bokoro health zones of Zaire. Structure of the Health Center Component of the Model The health center component of the model offers two menus: the reduced menu, which contains only two tables, one with basic input data and one with basic output data, and the expanded menu, made of about two dozen tables. The input table of the reduced menu contains key exogenous variables that can be modified by the user to analyze a wide range of policy issues. The output table of the reduced menu provides basic utilization indicators and a condensed health center income statement. The expanded menu contains more detailed input and output information. The examples of this section deal mainly with the reduced menu. The complete menu structures of the health center and hospital components of the model can be found in Appendix C. The full set of tables contained in the health center component can be found in Appendix D, and the hospital components of the model are contained in Appendix E. Finally, a user's manual for the model is in Appendix F. Exhibit 4-1 shows the input table (table B. 1) and the output table (table B.2) of the health center's reduced menu. A description of the input information required by table B. 1 as well as the output information shown in table B.2 follows exhibit 4-1. Simulations The Base Scenario Exhibit 4-1 constitutes the base scenario. As can be seen from table B. 1, the base consists of a market of 10,000 people, with 60 percent living within 2 kilometers of the health center. The radius of the market is 10 kilometers (area = 314 Km2). The currency of the country is the zaire and the current exchange rate is 350 zaires per U.S. dollar. Health center total prices range from a low of Z150 for prenatal and preschool care to a high of Z1,019 for deliveries. Health center total price for prenatal care, preschool care, and chronic care is below the actual cost to the health center of drugs used for those services. Ten percent of the population (1,000 people) is insured at the health center through an annual premium payment of Z1,500 and a direct copayment of 20 percent. The monthly household income of the uninsured of Z20,000 is equal to that of the insured. Five percent of the population is entitled to free care at the health center. The health center competes with two other providers in the market (see list of 21 22 A Supply-Demand Model of Health Care Financing Exhibit 4-1. Health Center, Reduced Menu, Base Scenario (M) -(P) - (L) - (0) Table B.1_ Population distribution _ _ 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5Km Total B H -- Cot o o | -A E Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100% drugs, jS A !Number 3500 2500 2000 1500 500 10000 ab.exams| I L I C T First Repeat Total & other I H PRICE visit visits Drugs expend phar.prdl ------ ------- ------- - (N) jCurative Care 200 0 558 758 558 1 C |DeLiveries 200 --- 819 1019 819 N E !Pre-Natal Care 0 --- 150 150 183 P N |Pre-SchooL Care 0 --- 150 150 441 U T !Chronic Care 0 --- 200 200 456 T E j R (A) *Radius health area(Km): 10 *Ins.premiumi/yr.(Zaires) 1500 (G) (B) *Population: 10000 *Percentage insured: 10% (H) (C) *Currency: Zaire *Copayment insured: 20% (I) (D) *Exchge.rete(Zaire/USS): 350 *Percent non-paying: 5% (J) (E) *HousehoLd income/month: 20000 *Superv.exp.(% rev.HC): 10% (K) (F) *Had.inc.insured/non-ins 1.00 -TABLE B.2 ALL INFORMATION IS MONTHLY_ (R) ---------New cases------ ------Total visits--...% mar- Non-inar Insured Total Non-inard. Insrd. Total k t HC B H A E Cur.Ce 505 101 606 1285 437 1722 68% *S A Deliv. 23 3 26 --- ... --- 78% I L (0) P-N Ca 13 2 15 124 16 140 72% C T P-S Ca 14 2 16 237 30 267 72% N Chr.Ca 0 0 0 27 4 31 67% 0 000s Zair U C (1) REVENUE Services 229 HC utiliz.rates T E Drugs 300 528 --------------- P N (2) EXPENSES Personnel 205 Non-inar Insrd U T (T) Drugs 370 ------T ----- T E Other 31 605 17% 30% R (3) PROFIT C(1)-(2)] (77) 64% 82% T (4) SUPERVISION EXPENSES 23 37% 43% A (5) DEPRECIATION 79 41% 47% B (6) PROFIT [(3)-(4)-(5)3 (178) 36% 44% L E (S) Demand, Costs, and Health Center Financial Performance: Simulations 23 TABLE B.1: INPUT DATA TO BE SPECIFIED BY (0) Total patient expenditure at the health THE USER center: (i) per episode of acute illness; (ii) per delivery; (iii) per prenatal (A) Radius of the health center's market (in care registration; (iv) per preschool kilometers) care registration; and (v) per year's worth of treatment for chronic illnesses (B) Total population of the health center's market (P) Number of people Living in each population ring. (C) Name of the country's currency () Country's current exchange rate TABLE B.2: OUTPUT DATA CALCULATED BY THE (E) Average monthly household revenue for the MODEL, INCLUDING HEALTH CENTER unns)e Averag e UTILIZATION. MARKET SHARE, AND uninsured peopleFIACLPEORNE (F) Ratio between the average monthly FINANCIAL PERFORMANCE household revenue of the insured and the (0) Utilization at the health center by the uninsured (for example, a ratio of 1.2 uninsured (both the paying and the means that those who are insured have a nonpaying), the insured, and total household revenue 20 percent higher than utilization. The top three columns at the uninsured) the left of table B.2 show new patients (those who registered during the month) (G) Annual per capita insurance premium treated at the health center, whereas the (H) Percentage of the population insured at following three columns show the number the health center of visits originating from the new (1) Copayment paid directly by the insured to patients as well as from old patients the health center (those who registered in previous (J) Percentage of the total market population montns). In the case of acute curative that is given free care at the health care and deliveries, illness episodes are center assumed to last a month or less; therefore, alt acute care and delivery (K) Supervision fees, expressed in patients treated at the health center in percentage, paid by the health center to any given month are new patients. For a central administrative office the three other categories of service, care is assumed to last more than a month (1) Health center's prices for the first (six months for prenatal care, five years for preschool care, and one year for visit, for each repeat visit, and for chronic illnesses); thus, the patients drugs, and totaL patient expenditure for for these categories include both new and each iLLness episode old patients (M) Percentage of the market population (R) Percentage of the ambulatory care market Living within each population ring. captured by the health center. For in example, the 78 percent figure for With these, and other data specified ndeliveries indicates that 78 percent of the expanded menu (see Appendix F), the model aLL deliveries taking place outside the calculates the information of (N)-(P) of table home are assisted at the health center, B.1 and (0)-(T) of table 2. whiLe the remaining 42 percent take place TABLE B.1: OUTPUT DATA CALCULATED BY THE in competing facilities MODEL (S) Percentage utilization by the uninsured (N) Marginal cost (drugs and medical and the insured. These percentages show, for both curative and preventive care, supplies): (i) per episode of acute the proportion of individuals who choose illness; (ii) per delivery; (iii) per to seek care outside the home and who go woman registered in the prenatal to the health center (referring back to preventive program (prenatal care) expression 2-4 of these percentages throughout her pregnancy; (iv) per child correspond to the product registered in the preschool preventive (ProbSeek)-(ProbChoose). For example, program (preschool care) throughout the the 37 percent for prenatal care child's first five years of Life; and (v) indicates that 37 percent of the per chronic patient, annually. The health uninsured pregnant women seek prenatal center's cost of drugs and medical care from the health center supplies is provided to help the user of the model to price the services, although (T) Health center's income statement. The P = MC is not implied. In the example, revenue is broken down into revenue from drugs are priced at marginal cost for payments other than drugs and payments acute illnesses and deliveries and below for drugs. Costs are broken down into cost for prenatal care, preschool care, labor costs, drugs, and other costs, and chronic care including fuel, office supplies, and supervision fees. 24 A Supply-Demand Model of Health Care Financing providers with their prices and location in tables A.3.1 through A.3.5 of Appendix D). One competitor is located at 2 kilometers from the health center and charges the same prices as the health center; a second competitor is situated at 5 kilometers and charges prices 30 percent lower than those of the health center. As shown in table B.2 of exhibit 4-1, with the above-specified input variables, the health center has a negative monthly net income of -Z77,000 before paying for supervision fees and depreciation, and negative net income of -Z178,000 after supervision fees and depreciation. According to the base scenario, this implies that the health center needs external subsidies not only to pay for supervision and depreciation but also to cover part of its recurrent costs. This was a common situation in Zaire in 1987. Exhibit 4-2 shows the baseline elasticities of demand with respect to price, income, and distance, all computed at a distance of 500 meters to the respective provider (table A. 10 of the model's expanded menu). The curative care elasticities resemble those obtained empirically in a separate, household-based study of curative ambulatory health care demand conducted in Bokoro and Kisantu (Bitran 1990). Exhibit 4-2. Price, Income, and Distance Elasticities of Demand, Health Center and Competitorsa A.10- Price, Income, and Distance ELastic"ties of Demand Curativ DeLiv. P-N CarP-S Care Chronic H.Center Price -0.79 -0.33 -0.17 -0.16 -0.24 Income 0.26 0.06 0.03 0.02 0.04 Distan -0.18 -0.07 -0.13 -0.12 -0.14 Compet.1 Price -1.30 -1.33 -0.37 -0.36 -0.49 Income 0.26 0.06 0.03 0.02 0.04 Distan -0.30 -0.30 -0.28 -0.28 -0.29 Compet.2 Price -1.00 -1.18 -0.36 -0.36 -0.46 Income 0.03 -0.15 0.00 -0.01 0.00 Distan -0.33 -0.34 -0.35 -0.35 -0.34 Compet.3 Price -- -- -- Income -- -- -- Distan -- -- -- a. ALL eLasticities computed at a distance of 500 meters from the provider. In the remainder of this section a series of simulation exercises is presented to illustrate how utilization and financial performance variables are affected by changes in selected exogenous variables. The outcome of each simulation exercise must be compared with that of the base scenario. An Increase in the Price of Curative Care at the Health Center The first simulation exercise, shown in exhibit 4-3, consists of a Z100 price increase for curative care on the part of the health center, from Z758 to Z858. The price increase translates into a new situation that differs from the base scenario in several ways. Demand, Costs, and Health Center Financial Performance: Simulations 25 Exhibit 4-3. An Increase in the Price of Curative Visits at the Health Center Table B.1 _ Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total B H ------- ------- ------- ------- ------- -------- |Cost of A E !Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%j drugs, S A jNunber 3500 2500 2000 1500 500 10000 Lab.exams I L First Repeat Total |& other C T visit visits Drugs expend.|phar.prdi H PRICE ------- ------- -------- ------ -------- iI !Curative care 300 0 558 858 | 558 | N C !Deliveries 200 --- 819 1019 1 819 | P E Pre-nataL Care 0 --- 150 150 | 183 U N !Pre-SchooL Care 0 --- 150 150 1 441 1 T T :Chronic Care 0 --- 200 200 456 | E T R *Radius health area(Km): 10 -Ins.premium/yr.(Zaires) 1500 A -Population: 10000 -Percentage insured: 10% B -Currency: Zaire -Copayment insured: 20% L *Exchge.rate(Zaire/US$): 350 -Percent non-paying: 5% E -Household income/month: 20000 -Superv.exp.(% rev.HC): 10% *Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY --------- New cases------ ------Total visits------ % mar-' Non-insr Insured Total Non-insrd. Insrd. Total ket HC: B H -------- ------- ------- -------- ------- ---- ------ | A E !Cur.Ca 449 99 548 1125 428 1553 64%| S A ,Deliv. 23 3 26 --- --- --- 78%| 1 L 'P-N Ca 13 2 15 124 16 140 72%1 C T IP-S Ca 14 2 16 237 30 267 72%: H jChr.Ca 0 0 0 27 4 31 67%| 0 000s Zaires | U C (1) REVENUE Services 263 | HC utiliz.rates| T E Drugs 270 532 N---------------| PN 1(2) EXPENSES Person[. 205 | Non-insr Insrd| U T Drugs 337 -------- ------ | T E Other 34 577 | Cur.Ca 15% 30%| R 1(3) PROFIT 1(1)-(2)] (44) Detiv. 64% 82%1 T |(4) SUPERVISION EXPENS. 26 1 P-N Ca 37% 43%| A |(5) DEPRECIATION 79 | P-S Ca 41% 47%: B 1(6) PROFIT[(3)-(4)-(5)1 (149)| Chr.Ca 36% 44%| L I |E First, utilization by the uninsured drops in an important way. For example, the number of new curative care acute patients goes down from 505 to 449, or an 11 percent fall. The implied arc-price elasticity of demand for this service is -0.84.1 Second, utilization by the insured also goes down, although by a much smaller percentage because the out-of-pocket price of the insured only increases by 20 percent of the price increase to the uninsured. Third, the financial performance of the health center improves---its recurrent cost gap drops to -Z44,000 from -Z77,000 in the base scenario. Both drug revenue and costs go down in relation to the base scenario because of lower demand. Also, despite the reduction in demand, the nondrug revenue goes up because of the price increase. Fourth, while the health center's financial performance improves after a price increase, its market share for acute curative care drops from 68 percent to 64 percent as a larger proportion of those seeking care outside the home choose the cheaper care supplied by the competitors. Fifth, the health center price increase has a negative 1. The elasticity is computed by dividing the percentage reduction in utilization (11.1 percent) by the percentage increase in price (13.2 percent). 26 A Supply-Demand Model of Health Care Financing effect on total market demand for curative care. At the base price of Z758, the demand for health center curative care was 606 episodes a month and the total market demand was 891 episodes (this last figure is computed by dividing the health center's demand of 606 episodes by the health center's market share of 68 percent). In contrast, at the higher price of Z858, total market demand for curative care is 856 episodes (548/64 percent). Thus, in spite of substitution among providers on the part of consumers, a price increase by one provider translates into lower overall demand for that service because some consumers drop out of the market. Finally, labor costs remain unchanged at Z205,000. A reduction in the price has the opposite effect of a price increase. For example, if the health center lowered its price of curative care, health center utilization would go up and the facility's financial performance would deteriorate: its net income would drop, and its market share would rise. A Change in a Competitor's Price Price changes on the part of a competitor mirror price variations by the health center. Exhibit 4-4a, drawn from the expanded menu (see table A.3. 1 in Appendix D), shows how the first competitor's price for curative care is changed from Z758 in the base scenario to Z500 in the new scenario. Exhibit 4-4a. Health Center's Expanded Menu---A Reduction in a Competitor's Price of Curative Care A.3.1- Curative care: Prices and distances HC and competitors (Zaires) Price Price Price of Average first repeat drugs distance visit visit per Total to HC episode price (metrs) Base scenario IHealth Center 200 0 558 758 0 lCompetitor 1 758 2,000 lCompetitor 2 526 5,000 lCompetitor 3 price change A.3.1- Curative care: Prices and distances HC and competitors (Zaires) Price Price Price of Average first repeat drugs distance visit visit per Total to HC episode price (metrs) New scenario IHealth Center 200 0 558 758 0 Competitor 1 500 2,000 lCompetitor 2 526 5,000 Competitor 3 Demand, Costs, and Health Center Financial Performance: Simulations 27 Exhibit 4-4b shows the impact that this change has on health center utilization and financial performance. As can be seen, the center's finances deteriorate (although slightly) as its level of utilization decreases along with its market share. This exercise illustrates how non-governmental competition, which in countries like Zaire is often important, can capture an important part of the market, especially for curative care, and affect negatively the financial performance of the health center. Exhibit 4-4b. A Reduction in a Competitor's Price of Curative Care Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total I B H ------- ------- ------- ------- ------- -------- Cost of IA E ,Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%| drugs, I S A lNumber 3500 2500 2000 1500 500 10000 Lab.examsl I L First Repeat Total I& other I C T visit visits Drugs expend.lphar.prd| H PRICE ------- ------- -------- ------ -------- | I lCurative care 200 0 558 758 558 1 N C IDeliveries 200 --- 819 1019 819 jP E Pre-natal Care 0 --- 150 1501 183 U N Pre-School Care 0 --- 150 1501 441 IT T Chronic Care 0 --- 200 200 456 1 E I T R -Radius health area(Km): 10 *Ins.premium/yr.(Zaires) 1500 A -Population: 10000 *Percentage insured: 10% B -Currency: Zaire -Copayment insured: 20% L *Exchge.rate(Zaire/USS): 350 -Percent non-paying: 5% E -Household income/month: 20000 -Superv.exp.(% rev.HC): 10% -Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY --------- New cases------ ------Total visits------ % mar-: Non-insr Insured Total Non-insrd. Insrd. Total ket HCI B H -------- ------- ------- -------- ------- ---- ------ A E ,Cur.Ca 457 96 554 1165 411 1,576 59% S A :Deliv. 23 3 26 --- --- --- 78%: I L P-N Ca 13 2 15 124 16 140 72% C T IP-S Ca 14 2 16 237 30 267 72%I H lChr.Ca 0 0 0 27 4 31 67% 0 000s Zaires 1 U C (1) REVENUE Services 220 HC utiliz.ratesl T E Drugs 274 494 --------------- PN 1(2) EXPENSES Persont. 205 ' Non-insr Insrdl U T Drugs 341 -------- ------| T E Other 30 576 1 Cur.Ca 15% 29%| R I(3) PROFIT [1)-(2)] (82)1 Deliv. 64% 82% T (4) SUPERVISION EXPENS. 22 | P-N Ca 37% 43%| A (5) DEPRECIATION 79 | P-S Ca 41% 47%| B 1(6) PROFIT[(3)-(4)-(5)] (182)1 Chr.Ca 36% 44%| L i_ IE Price Cross-Subsidization between Curative and Preventive Care In this simulation (exhibit 4-5), the price of preventive care (prenatal and preschool care) is varied from a price equal to marginal cost (MC) to a price of zero. The purpose of lowering the price below MC is to promote consumption of preventive care. In order to finance this policy, the price of curative care is increased with the goal of maintaining a recurrent cost gap of Z77,000, as in the base scenario, 28 A Supply-Demand Model of Health Care Financing and thus keep the need for external subsidies unchanged. Partial or full price cross-subsidization of preventive care is a common policy in Zaire's health zones. Exhibit 4-5. Cross-Subsidization of Preventive by Curative Care---Prices and Quantity Demandeda Price Level Prenatal and Preschool Care P =MC P =A.Mc P= 12.MC P = 4.Mc P = 0 PRICE OF PricepN care 183 137 92 46 0 PREVENTIVE CARE Priceps care 441 331 221 110 0 PRICE OF Pricecur. care 745 751 758 764 774 CURATIVE CARE REQUIRED TO BREAK-EVEN QUANTITY Q.Dem.PN care 119 126 133 140 147 DEMANDED (NUMBER OF Q.Dem.pS care 155 185 217 249 279 VISITS PER MONTH) Q.Dem-Cur.care 1306 1296 1285 1275 1258 PRICE EPN care -0.21 -0.15 -0.09 -0.04 0 ELASTICITY OF DEMAND EPS care -0.70 -0.46 -0.26 -0.11 0 ECur.care -0.77 -0.78 -0.79 -0.80 -1.00 (a) Prices in Zaires of March 1987; quantity demanded is monthly visits to the health center. The above simulation shows that, based on the assumed coefficients of the demand equations for curative and preventive care, only a modest price increase of curative care (3.9 percent from Z745 to Z775) is sufficient to subsidize preventive services fully. Curative care demand would obviously drop as a result of the price increase, from 1,306 visits a month when preventive care is priced at MC to 1,250 visits when preventive care is provided free of charge. Policymakers should judge whether the drop in curative care demand would be worth the increase in demand for preventive services. Price Discrimination by Income Other things being equal, a single price will result in higher per capita demand by the rich and lower demand by the poor. A policy of price discrimination, common in Zaire's health zones, can help promote equity by charging prices that are inversely related to people's income. Further, price discrimination, the strategy used by the profit-maximizing private monopolist, can allow the public health center to improve its financial performance while fostering equity, with little harm to the higher income groups. In this simulation (exhibit 4-6), the model is used first to show how a strategy of a single price Demand, Costs, and Health Center Financial Performance: Simulations 29 results in an inequitable demand configuration among income groups. For example, the fifth of the population that belongs to the highest income quintile would demand 136 new episodes of curative care per month, or 46 percent more than those in the lowest income group. Second, the model is used to calculate, for each income quintile, the price P (first row at the bottom section of the table) necessary to achieve a demand equal to that of the third quintile.2 Exhibit 4-6. Comparison between a Single-Price Policy and a Policy of Price Discrimination Based on Income, Curative Ambulatory Care Income QuintiLe (Zaires 3/87) 0-7000 7001- 14001- 20001- 27001- TotaL 14000 20000 27000 34000 PopuLation 2000 2000 2000 2000 2000 8000 POLICY OF Price (P) 758 758 758 758 758 -- A SINGLE PRICE Demand (Q) 93 103 112 124 136 489 Revenue = 70,494 78,074 84,896 93,992 103,088 430,544 P. Q VC=MC. Q 43,314 57,474 62,496 69,192 75,888 308,364 Revenue-VC 27,780 20,600 22,400 24,800 27,200 122,180 Ep -1.09 -0.93 -0.79 -0.63 -0.49 -- CY 0.10 0.19 0.26 0.34 0.40 -- POLICY OF Price (P ) 635 695 758 865 1,010 -- PRICE DISCRIMIN- Demand (Q) 112 112 112 112 112 560 ATION BASED ON Revenue 71,120 77,840 84,896 96,880 113,120 443,856 INCOME p* Q VC=MC.Q 62,496 62,496 62,496 62,496 62,496 312,480 Revenue-VC 8,624 15,344 22,400 34,384 50,624 131,376 The above exercise shows that, based on the implicit price and income elasticities of demand, a strategy of price discrimination by income can promote equity, improve financial performance, and have only a modest negative effect on demand by those in the highest population group. Price Discrimination by Distance to the Health Center This simulation is similar to the one above, except that distance to the facility is used as the criterion to discriminate by price. With a uniform price, those living within one kilometer of the facility represent 35 percent of the population in the market yet account for 50 percent of total demand (exhibit 4-7); those living beyond 5 kilometers represent 5 percent of the population but account for only 1 2. To simplify, it is assumed in this simulation that there are no people insured and that no free care is available at the health center. 30 A Supply-Demand Model of Health Care Financing percent of total demand. The bottom part of the table shows the price necessary to ensure that the demand by the population within each distance group is equal to the group's share of the total population, and equal, on a per capita basis, to the demand by those living within 1 kilometer of the health center. The table shows that, for those living beyond 3 kilometers, the price would have to be negative (that is, those living beyond 3 kilometers would have to be paid to come to the health center) in order to equate their per capita demand with that of those living closer to the center. A total monthly subsidy greater than Z140,377 (see bottom-right cell of the table) would be necessary to sustain this policy! Exhibit 4-7. Comparison of a Policy of a Single Price and a Policy of Differential Prices According to Distance to the Health Center---Curative Ambulatory Care Distance Household - Health Center (Km) Total 0-1 1-2 2-3 3-5 5-10 - Population 3,500 2,500 2,000 1,500 500 10,000 % of population 35% 25% 20% 15% 5% 100% SINGLE Price (P) 758 758 758 758 758 - PRICE Demand(Q)a 304 159 94 44 5 606 % 50% 26% 15% 7% 1% 100% DIFFE- Price (P) 758 580 265 <0 <0 - RENTIAL PRICE Demand 304 217 174 130 43 868 (Q)a % 35% 25% 20% 15% 5% 100% P-MC 200 -123 -448 <-558 <-558 - Subsidy needed b -60,800 26,691 77,952 >72,540 >23,994 >140,377 a. Quantity demanded (first visits). b. Calculated as Q.(P-MC), or utilization times the price minus the marginal cost. A surplus of 60,800 zaires would be generated from those living within 1 Km. of the health center. ALL other distance groups would require a subsidy, as shown. A Change in Insurance Enrollment and Premium In this simulation, the model is used to study the effect of (1) expanding the insurance coverage; (2) reducing the insurance premium; and (3) both (1) and (2) simultaneously. The insurance coverage is increased from 10 percent to 20 percent of the population (simulation shown in exhibit G1 of Appendix G). The facility almost breaks even before supervision fees and depreciation, compared with a loss of Z77,000 in the base scenario. This indicates that the annual cost of providing care to the insured is less than the premium. The percentage insured is then restored to 10 percent and the premium reduced from Z1,500 to Z800 (exhibit G.2). As expected, the facility's finances worsen in relation to the base, with a monthly loss of Z135,000. Suppose that, as could be expected, the reduction in the insurance premium brings about a rise in enrollment from 10 percent to 20 percent (exhibit 4-8).3 3. The assumed elasticity of insurance with respect to the premium is approximately -1.0. Demand, Costs, and Health Center Financial Performance: Simulations 31 Exhibit 4-8. A Drop in the Insurance Premium Coupled with a Rise in the Proportion of People Insured Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total | B H ------- ------- ------- ------- ------- -------- Cost of IA E !Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100% drugs, I S A Number 3500 2500 2000 1500 500 10000 Lab.examsl I L First Repeat Total I& other C T visit visits Drugs expend.|phar.prdl H PRICE ------- ------- -------- ------ -------- I Curative care 200 0 558 758 | 558 N C Deliveries 200 --- 819 1019 1 819 1 P E Pre-natal Care 0 --- 150 150| 183 U N ,Pre-School Care 0 --- 150 150 441 j T T IChronic Care 0 --- 200 200 | 456 E T R Radius health area(Km): 10 lIns.premium/yr.(Zaires) 1075 A -Population: 10000 'Percentage insured: 20% B -Currency: Zaire 'Copayment insured: 20% L *Exchge.rate(Zaire/USS): 350 'Percent non-paying: 5% E 'Household income/month: 20000 -Superv.exp.(% rev.HC): 10% -Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY S ---------New cases------ ------Total visits------ % mar-I Non-insr Insured Total Non-insrd. Insrd. Total ket HCI B H -------- ------- ------- -------- ------- ---- ------ A E 1Cur.Ca 449 202 650 1142 704 1846 70%1 S A Deliv. 20 6 27 --- --- 80% I L IP-N Ca 12 3 15 110 32 142 T3%1 C T IP-S Ca 13 4 16 211 60 271 73%| H jChr.Ca 0 0 0 24 7 31 68%| 0 000s Zaires | U C (1) REVENUE Services 275 HC utiliz.rates| T E i Drugs 278 553 ---------------| P N (2) EXPENSES Persont. 205 | Non-insr Insrdj U T Drugs 395 ' -------- ---- T E Other 30 630 | Cur.Ca 17% 30%| R (3) PROFIT [1)-(2)] (77): DeLiv. 64% 82%| T 1(4) SUPERVISION EXPENS. 28 1 P-N Ca 37% 43%| A :(5) DEPRECIATION 79 i P-S Ca 41% 47% B 1(6) PROFIT[(3)-(4)-(5)] (183): Chr.Ca 36% 44%| L IE Exhibit 4-8 shows that the health center's profit remains unchanged after cutting in half its insurance premium and doubling insurance enrollment. At the same time, total utilization of curative care goes up at the health center because the insured face a lower out-of-pocket price. The above simulation illustrates the advantages of health insurance. First, by lowering its insurance premium the facility stimulates the demand for health insurance, enabling more people to benefit from the health, psychological, and financial advantages of insurance. Second, the increase in enrollment and utilization do not affect fixed costs.4 The health center can thus benefit from economies of scale and can provide more care at a lower price, while maintaining its financial position. 4. Notice, however, that the model does not take into account the administrative costs of health insurance. 32 A Supply-Demand Model of Health Care Financing An Increase in the Insurance Copayment Exhibit 4-9 shows the effect of an increase in the copayment from 20 percent of the full health center price in the base scenario to 40 percent. Demand by the insured goes down as their out-of-pocket price is doubled. At the same time, the health center enjoys a higher net income in the new scenario because its revenue arising from the copayments by the insured goes up while its costs drop. A higher out-of-pocket price to the insured reduces moral hazard to the benefit of the insurer, although it makes insurance a less attractive option to the client. Exhibit 4-9. An Increase in the Insurance Copayment Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total B H ------- ------- ------- ------- ------- --------Cost of A E |Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%j drugs, S A Number 3500 2500 2000 1500 500 10000 Lab.exams I L First Repeat Total & other C T visit visits Drugs expend.lphar.prdl H PRICE ------- ------- -------- ------ I--------i I Curative care 200 0 558 758 558 | N C Deliveries 200 --- 819 1019 819 1 P E Pre-natat Care 0 --- 150 150 183 | U N Pre-School Care 0 --- 150 150 441 1 T T !Chronic Care 0 --- 200 200 456 1 E _ T R -Radius health area(Km): 10 -Ins.premium/yr.(Zaires) 1500 A *Population: 10000 -Percentage insured: 10% B -Currency: Zaire -Copayment insured: 30% L -Exchge.rate(Zaire/US$): 350 -Percent non-paying: 5% E -Household income/month: 20000 *Superv.exp.(% rev.HC): 10% -Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY --------- New cases------ ------Total visits------ % mar-: Non-insr Insured Total Non-insrd. Insrd. Total ket HC1 B H -------- ------- ------- -------- ------- ---- ------ IA E ,Cur.Ca 505 95 600 1285 394 1678 67%1 S A Deliv. 23 3 26 --- --- --- 78%1 1 L P-N Ca 13 2 15 124 16 140 72%: C T IP-S Ca 14 2 16 237 30 267 72%j H Chr.Ca 0 0 0 27 4 31 67% 0 i _ _ _000s Zaires I U C (1) REVENUE Services 231 HC utiliz.rates T E i Drugs 304 535 --------------- IP N (2) EXPENSES PersonL. 205 i Non-insr Insrdl U T Drugs 366 -------- ------ T E Other 31 602 | Cur.Ca 17% 28%1 R 1(3) PROFIT E(1)-(2)] (67) Deliv. 64% 80% T (4) SUPERVISION EXPENS. 23 | P-N Ca 37% 43% A (5) DEPRECIATION 79 1 P-S Ca 41% 46% B (6) PROFITE(3)-(4)-(5)] (169); Chr.Ca 36% 43%| L Demand, Costs, and Health Center Financial Performance: Simulations 33 An Increase in the Amount of Free Care Exhibit 4-10 shows the effect of a more generous free care policy (10 percent of the population is entitled to free care, as opposed to 5 percent in the base scenario). Exhibit 4-10. A Rise in the Proportion of People Receiving Free Care at the Health Center Table B.1 ______PopuLation distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total B H ------- ------- ------- ------- ------- --------cost of A E 'Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%: drugs, S A 'Number 3500 2500 2000 1500 500 10000 Lab.examsi I L First Repeat Total |& other C T visit visits Drugs expend.|phar.prdl H PRICE ------- ------- -------- ------ -------- iI Curative care 200 0 558 758 : 558 N C Deliveries 200 --- 819 1019 819 | P E |Pre-natal Care 0 --- 150 150 183 | U N |Pre-School Care 0 --- 150 150 441 T T |Chronic Care 0 --- 200 200 456 E |,T R -Radius health area(Km): 10 -Ins.premium/yr.(Zaires) 1500 A *PopuLation: 10000 -Percentage insured: 10% B *Currency: Zaire -Copayment insured: 20% L -Exchge.rate(Zaire/USS): 350 *Percent non-paying: 10% E -Household income/month: 20000 *Superv.exp.(% rev.HC): 10% -Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY --------- New cases------ ------TotaL visits------ % mar-i Non-insr Insured Total Non-insrd. Insrd. Total ket HCI B H -------- ------- ------- -------- ------- ---- ------ A E Cur.Ca 505 101 606 1285 437 1722 68% S A Deliv. 23 3 26 --- --- --- 78% I L P-N Ca 13 2 15 124 16 140 72% C T 'P-S Ca 14 2 16 237 30 267 72%| H Chr.Ca 0 0 0 27 4 31 67% 0 i _ _ 000s Zaires I U C (1) REVENUE Services 223 HC utiLiz.rates| T E Drugs 283 506 |' --------------- P N (2) EXPENSES PersonL. 205 Non-insr Insrd: U T Drugs 370 -------- ------| T E Other 30 605 Cur.Ca 17% 30% R |(3) PROFIT [(1)-(2)] (99)1 DeLiv. 64% 82%j T |(4) SUPERVISION EXPENS. 22 P-N Ca 37% 43%| A |(5) DEPRECIATION 79 : P-S Ca 41% 47%| B '(6) PROFIT[(3)-(4)-(5) (200) Chr.Ca 36% 44%| L i I E As shown above, utilization remains unchanged because it is assumed that the indigents' per capita utilization is equal to that of the paying uninsured as a result of nonprice rationing (see chapter 3). Total revenue drops, however, as the number of people who are given free care doubles. Monthly net income drops by Z22,000 in light of lower revenue and constant costs. Thus, in order to finance this more generous free-care policy, the health center would have to receive additional external subsidies worth Z22,000 a month. 34 A Supply-Demand Model of Health Care Financing A Devaluation of the Country's Currency A devaluation of the country's currency has a perverse effect on the providers' financial performance. This simulation assumes an increase in the value of the dollar from Z350 to Z400 (simulation shown in exhibit G.3). As long as the price of care is not increased, utilization and revenue remain unchanged. Costs do increase, however, particularly for drugs, which rise in price in proportion to the devaluation. Net income drops to -Z131,000 a month compared with a loss of Z77,000 in the base scenario. Suppose that, in order to offset this loss, the health center raises the price of curative care. Exhibit G.4 shows that the facility would have to increase the price per episode of curative care to Z918 in order to offset the effect of the devaluation. This higher price would result in a 15 percent drop in demand for curative care, from 606 episodes a month before the devaluation to 515 episodes after both the devaluation and the price increase. A Change in the Population Distribution Changes in the population's geographic distribution also affect demand and the financial performance of providers, although the effect on each individual provider will depend upon its geographic location within the market. In this final simulation (exhibit G.5) it is assumed that all the population lives within 2 kilometers of the health center, with three-quarters living within 1 kilometer. With more people living in the vicinity of the health center, its utilization increases substantially. For example, the monthly number of curative illness episodes treated at the health center goes up from 606 to 810, a 34 percent increase. The center's net income deteriorates, however, as the marginal cost of providing more services exceeds the marginal revenue from cost recovery. 5 HOSPITAL CARE Main Features of the Hospital Component of the Model This chapter presents the hospital component of the model. In general, the health center and the hospital components are very similar in their logic: demand and costs are modeled using the same kinds of functions; the hospital also serves a circular market with the inhabitants uniformly distributed around population rings and with competition from other providers. Because of the conceptual similarities of both model components, this section describes the features that are unique to the hospital module. They are: (a) the kinds of services produced; (b) the use of bed-days as a measure of utilization; (c) the need for hospital beds; (d) the inclusion of administrative and support personnel; and (e) the category of medical personnel considered. These features are described below. Inpatient Services The hospital is assumed to produce three categories of service: hospitalizations that do not require surgery, hospitalizations with surgery, and deliveries. Each of these services can consume any of the following production inputs: drugs, laboratory exams, x-rays, other pharmaceutical products, and labor. As in the case of the health center, the user of the model must specify the amounts of inputs used for each service as well as the cost of the inputs to the hospital. Exhibit 5-1 shows the reduced hospital tables (table E. 1 for basic input and table E. 2 for basic output). Bed-Days as a Measure of Utilization As shown in the sixth column of table E. 1, the user must specify the patients' average length of stay (ALOS), in days, for each type of service. For example, an assumption made in the table is that people who get hospitalized and undergo surgery spend an average of ten days in the hospital. The model's direct measure of demand for inpatient care is the number of people who become hospitalized to obtain each of the three categories of service offered by the hospital. Such number is computed through three separate demand equations, similar to the ones used by the health center component of the model, and varies according to the prices and location of providers and to people's income. The total number of bed-days utilized is computed by multiplying the number of hospitalizations demanded (calculated by the model) times the ALOS, which is a fixed quantity and is user-defined. In mathematical notation: (5-1) DAYSi = Q*.ALOS 35 36 A Supply-Demand Model of Health Care Financing where DAYSi, is the number of bed-days of service i demanded, Qj is the number of service i hospitalizations demanded, and ALOS is average length of stay (in days) for service i. Exhibit 5-1. Hospital, Reduced Menu, Base Scenario Table E.1 Population distribution 0-1 Km 1-5 Km 5-20 Km 20-50 Km >50 Km Total !Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%lCost of A Number 35000 25000 20000 15000 5000 100000 drugs, S Drugs, Total Average ILab.exmsl I Medical exams, patient Length & other C fees & other expend. stay(ds)lphar.prdl H PRICE ------- ------- ------ -------- -------- I 0 IHosp. without surgery 3000 3500 6500 6.0 4403 N S lHosp. with surgery 4000 5500 9500 10.0 6136 1P P DeLiveries 2500 2500 5000 6.0 3707 j U I _T T -Radius health area(Km): 10 -Ins.premium/yr.(Zaires) 1000 A -Population: 100000 -Percentage insured: 10% T L -Currency: Zaire -Copayment insured: 10% A -Exchge.rate(Zaire/US$): 350 -Percent non-paying: 5% B -Household income/month: 20000 -Superv.exp.(% rev.HC): 5% L -Hsd.inc.insured/non-ins 1.00 -Other pers./100 beds: 30 E *Percent occupancy beds: 80% TABLE E.2 ALL INFORMATION IS MONTHLY ------- Utilization------ Nbr.bds % mar-I Non-insr Insured Total Bed-days necessa ket Ho B -------- ------- ------- -------- ----- ------ A lHosp. without s 80 14 95 570 23 74%1 S IHosp. with surg 69 15 84 838 34 79%j 1 Deliveries 103 17 120 720 30 75%j C 1 H 000s Zaires- 87 1 0 0 (1) REVENUE Services 1,582 U S Drugs 887 2,469 T P (2) EXPENSES Person[. 2,190 Utiliz'n rates P I Drugs 1,381 --------------- UT other 10 3,581 Non-insr Insrdl T A 1(3) PROFIT [(1)-(2)J (1,112) -------- ------ L 1(4) SUPERVISION EXPENS. 79 IH.w/o s. 27% 43% T 1(5) DEPRECIATION 326 IH.wth s. 23% 45% A (6) PROFIT[(3)-(4)-(5)J (1,517)lDeliver. 29% 43%I B L E The first three columns at the top of table E.2 show model-predicted demand (number of people hospitalized) by the uninsured, the insured, and the total. The fourth column provides the number of bed- days, which is the product of column three of this table times the ALOS column of table E. 1. Hospital Beds Needed Based on the number of bed-days provided by the hospital, the model computes the total number of beds required, assuming that the actual occupancy of the beds is that specified at the bottom of table Hospital Care 37 E. 1. For example, with an 80 percent overall bed occupancy, the hospital would need 87 beds to accommodate all of its patients. The number of beds needed is computed as follows: DAYS,-12 (5-2) BEDSi -DYS-1 'OCC-365 where BEDSi is the number of beds required in the hospital to satisfy demand for service type i at the occupancy level OCC, DAYSi is the monthly number of bed-days required to satisfy demand for service type i (see computation of DAYSi above), the number 12 is the number of months a year, OCC is the overall hospital occupancy rate, and 365 is the number of days in a year. Administrative and Support Personnel Hospitals usually employ a number of administrative personnel, such as accountants, cashiers, pharmacists, and managers. In addition, hospitals hire auxiliary personnel and other people who clean and maintain the premises. The model assumes that the number of administrative and support personnel needed by the hospital is proportional to the hospital's number of beds and, therefore, to demand. The bottom section of table E. 1 in exhibit 5-1, shows that the number of administrative and support personnel has been set by the user at 30 people for each 100 hospital beds. Based on the actual number of beds required, as predicted by the model, and on the peoples' salaries, specified in a separate table of the expanded hospital menu (see tables D.5.1 and D.5.2 of Appendix E), the model computes the monthly labor cost for this category of personnel using the following relationship: (5-3) OTHPERS = BEDS OP 100 where OTHPERS is the administrative and support personnel needed in the hospital, BEDS is the total number of hospital beds needed to meet demand, and OP is the number of administrative and support personnel needed per 100 beds in the hospital. Medical Personnel The health center component of the model considered four categories of medical or paramedical personnel. The hospital module considers additional personnel categories as follows: medical doctor, nurse, midwife, laboratory technician, x-ray technician, and other medical or paramedical personnel (see tables D.5.1 and D.5.2 of Appendix E). As in the case of the health center, the number of personnel of each kind required is determined based on the model-predicted demand for each service and the amount of time spent by each kind of professional on each service category. Because of the correspondence between the health center and the hospital components of the model, a series of simulation exercises that illustrate the use of the hospital module is not warranted. The reader is encouraged to do similar exercises with the attached version of the model. 6 AGGREGATE PROJECTIONS AND CONCLUSIONS The model can be used to obtain aggregate projections of resource needs at the district or country level. In order to do this, the model combines health center and hospital output information on projected resource use in a table. The procedure followed for obtaining aggregate projections consists of specifying in the model the number of health centers and hospitals, as characterized by the user in both components of the model, that exist in the region or country. The model then extracts selected output information from the two model components, multiplies this information by the number of facilities of each type, and combines this information into a table. Aggregate information corresponding to the health center and hospital models, as defined in the examples of this paper, are shown below. Exhibit 6-1. Aggregate Projections District of country monthly projections (US$ 000's & Zaires 000,000's) Number of hospitals 50 Number of health centers 1000 i ---HOSPITALS--- -HEALTH CENTERS-' -----TOTAL---- LABOR -Number- -Cost----Number- -Cost--- -Number- -Cost----' Doctors 100 15.0 0 0.0 100 15.0 Nurses 600 27.0 2,000 100.0 2,600 127.0 Midwives 50 2.3 1,000 80.0 1,050 82.3 Lab.technicia 50 2.3 1,000 25.0 | 1,050 27.3 X-Ray technic 50 2.3 0 0.0 | 50 2.3 Other personn 1,350 60.8 0 0.0 1,350 60.8 | TOTAL (Zaire) 2,200 109.5 4,000 205.0 6,200 314.5 TOTAL (US$) 312.9 585.7 898.6 |TOT.RECUR.COSTS(Zaires) 183.0 628.3 1 811.3 | |TOT.REVENUES (Zaires) 123.4 528.3 651.8 |DEFICIT/SURPLUS(Zaires) (59.6) (100.0)1 (159.5) |FOREIGN.EXCH.NEEDS(USS) 126.7 1,055.8 1,182.5 |Number of hospital beds 4373 --- 4,373 The aggregate projections table shows: the total health-facility-level (hospital and health center levels) manpower requirements at the district or country level; the total manpower costs, in both local currency and in U.S. dollars; the total recurrent costs (which include all costs except depreciation of buildings and equipment); the total health facility revenues from user fees, insurance premiums, and insurance copayments; the recurrent cost gap, calculated as the difference between the two preceding figures; the total foreign exchange needs (calculated as the cost of all pharmaceutical products, expressed in U.S. dollars); and the hospital beds required to satisfy demand. Conclusions This paper has presented a computer-implemented model of health care financing. The model is a useful, user-friendly analytical tool for studying the effects of policy decisions (such as pricing and subsidization), management strategies (such as the use of personnel), and external shocks (a devaluation 39 40 A Supply-Demand Model of Health Care Financing of the currency, for example) on both health facility financial performance and utilization of health services.1 The analysis illustrates the importance of knowledge about health care demand information for decisionmaking. Specifically, the model demonstrates how information about the elasticity of demand with respect to price, income, and time costs can be used to assess the likely impact on the market of alternative health care financing policies. The model has also conveyed the importance of knowledge about the supply side of the market, including the number, types, location, and price levels of competitors as well as the cost structure of facilities. Further, it has been shown that in most cases, a responsible assessment of health care financing policies requires the simultaneous use of information about supply and demand. The use of information on either side of the market in isolation provides a partial, sometimes unfeasible, outlook of the effects of policy. For example, price decisions that are based solely on demand information may result in fees that imply health facility losses that are inconsistent with existing government budgets for health. As discussed earlier, the model has several limitations. First, it assumes an absence of supply constraints; that is, providers always meet demand. Second, the model does not provide direct measures of welfare. Third, explicit measures of health care quality and their effect on demand are not included in the model. Fourth, the model does not incorporate endogenous information about provider behavior, including production (or cost) functions, and pricing rules. Some of these constraints can be overcome with relative ease and this will be done in future versions of the model. For example, supply constraint could easily be built in the model by entering maximum, short-term capacity information for each provider. This would bound utilization to an upper limit equal to each provider's capacity. The existence of capacity constraints would imply a rationing of demand. Rationing effects could also be incorporated, although this would involve more complex assumptions and programming. For example, waiting time could increase as a function of demand, which in turn would depress demand to an equilibrium point. Welfare measures could also be added in the model by adopting welfare measures such as compensating variations, derived for the discrete choice case by Small and Rosen (1981) and used by Gertler, Locay, and Sanderson (1987) in their study of health care demand in Peru. The inclusion of endogenous information about provider behavior poses more difficult challenges. Although pricing rules could be incorporated without to much difficulty, these would not be substantiated by any empirical evidence because of the lack of studies in this area. Similarly, production rules could be incorporated endogenously (for example, cost-minimizing rules), although such rules would be arbitrary given the absence of empirical information on this subject. Finally, adding explicit measures of quality and their effect on demand poses the most difficult challenge, again because of the lack of empirical data on this aspect of the problem. Refinement of the model on this front may have to wait until better data on both quality and user behavior become available. 1. The model allows the user to substitute health care personnel, such as nurses and physicians, and to assess the effect of such changes on health facility costs, given demand. For space considerations, examples about this feature have not been provided. TECHNICAL APPENDIX A Computation of Weighted Average Distance between the Population in Ring r and a Competitor Under the assumption that the population is uniformly distributed within a population ring of radius R, the weighted average distance between that population and a given competitor located at a distance D from the health center can be expressed through the following integral (see figure A-1): 7r2 = () VR2 + D2 - 2 * R - D * cosO - dO This integral cannot be solved analytically and has been approximated numerically in the model. Figure A-1. Computation of Distance between Population in Ring r and a Competitor Household 1 Household n dn di D Health Competitor Center Population ring r 41 TECHNICAL APPENDIX B Demand Equations and Elasticities Using McFadden's nested logit formulation (McFadden 1981), ProbSeek and ProbChoose can be expressed as follows: N (ProbSeek)r k=1N eA+( Ee Va k=1 and V (ProbChoose)r Ee Vk k=1 where A is indirect utility associated with the no-care option, sigma is 1 minus the correlation coefficient of the provider options, N represents the total number of providers, and e is the base of the natural exponential function. The proportion of people who choose not to seek care outside the home is equal to 1-(ProbSeek)r, or one minus the proportion of people who seek outside care. Among those who decide to seek care outside the home, the sum of the probabilities of choosing a given provider r equals one, or: N E (ProbChoose)rk = 1, for all r. k=1 All the coefficients and parameters included on the right-hand side of the above demand equations are user-specified, while demand is computed by the model. 42 Technical Appendix B 43 With the above specification of demand, the price elasticity of demand is aProb. P. 1 e Vk EP = -a = [Pj. -(01 + 02 *Y)] e ( + Ee Vk- e ). P1 pTo1W, Ee Vk eIA + (EeVk )y k k k The distance elasticity of demand has a similar expression, except that the term in the square bracket becomes [D * 031. The income elasticity of demand is ( Ee k)o Y Vk k + V E e Vk (eA+( EeVk k k k These three elasticities are computed automatically by the model and shown in the model's expanded menu (table A. 10). When a= 1, nested logit reduces to conditional logit. The expression for the price elasticity becomes ep = [P (1 + 02 - 1 1-Probj) and that for income elasticity ey1 = 02 k - Probk + Pj) k Expressions for cross-elasticities are more involved and have not been derived here. The current specification of demand implies that demand becomes less price- and distance-elastic as income goes up, and more price- and distance-elastic as price and income increase. APPENDIX C Menu Structure of Computer Model Initial Menu HealJth Center Hospita Expanded Reduced Exadd Reduced MeuI en Menu Menu EiDemogra cal Basic Input Basi Out0- pu t &Demiolaoica B Bsic Input BscOtu Table Table Table Table T ableTal 2. Cost of 2. Cost of Investments Table Investments Table 3. Price of Care 3 Price of Care a,nd Location of -and Location of - Providers TalsProviders Tables 4. Cost and Use of 4. Cost and Use of Pharnaceutical -Pharmaceutical Products Tables Prod ucts Tables 5. Cost and Use 5 Cost and Use of Labor Tables of Labor Tables 6. Utilization 6 Utilization Tables Tables 7. Utilization and 7. Utilhzation and Revenue Tables Revenue Tables 8. Income 8. Income Statement Statement 9 Demand 9Dmn Equation - qato. Coefficets Coefficients 44 APPENDIX D Computer Model, Health Center Component Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total | B H ------- ------- ------- ------- ------- --------jCost of A E Percent 35.0% 25.0% 20.0% 15.0% 5.f)% 100% drugs, S A !Number 3500 2500 2000 1500 500 10000 lab.examsj I L First Repeat Total & other C T visit visits Drugs expend.lphar.prd H PRICE ------- ------- -------- ------ -------- I Curative care 200 0 558 758 558 N C Deliveries 200 --- 819 1019 1 819 P E Pre-natal Care 0 --- 150 150 183 U N !Pre-School Care 0 --- 150 150 1 441 T T |Chronic Care 0 --- 200 200 456 1 E | T R -Radius health area(Km): 10 *Ins.premium/yr.(Zaires) 1500 A -Population: 10000 -Percentage insured: 10% B -Currency: Zaire -Copayment insured: 20% L *Exchge.rate(Zaire/US$): 350 *Percent non-paying: 5% E *Household income/month: 20000 -Superv.exp.(% rev.HC): 10% -Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY ---------New cases------ ------Total visits------ % mar-' Non-insr Insured Total Non-insrd. Insrd. Total ket HCI B H -------- ------- ------- -------- ------- ---- ------ AE !Cur.Ca 505 101 606 1285 437 1722 68%1 S A ,Deliv. 23 3 26 --- --- --- 78%j 1 L P-N Ca 13 2 15 124 16 140 72% C T P-S Ca 14 2 16 237 30 267 72% H Chr.Ca 0 0 0 27 4 31 67%1 0 i _ _ _ _ 000s Zaires I U C (1) REVENUE Services 229 HC utiliz.ratesl T E Drugs 300 528 --------------- PN (2) DEPENSES Persont. 205 Non-insr Insrd U T Drugs 370 -------- ------ T E Other 31 605 Cur.Ca 17% 30%1 R 1(3) PROFIT C(1)-(2)] (77) Deliv. 64% 82%1 T 1(4) SUPERVISION EXPENS. 23 P-N Ca 37% 43% A 1(5) DEPRECIATION 79 P-S Ca 41% 47%1 B 1(6) PROFIT[(3)-(4)-(5)] (178)1 Chr.Ca 36% 44% L E 45 46 A Supply-Demand Model of Health Care Financing A.1- Epidemiological and demographic data |Prevalence of chronic illnesses (cases per 1000 peopLe.yr.) 50 lCurative illness episodes per person per year 4 lNet population growth 3.0% |Crude birth rate (per 1000 inhabitants) 47 Percentage of the population between 0-5 years 20%| A.2- Investment and some recurrent costs at health center Invest- Invest- (These Labels can be Useful ment ment changed by the user) Life (USS) (000 Zai) *Building 10 10,000 3,500 -Medical equipment 5 2,500 875 !Refrigerator 0 0 0 -X-Ray Machine 2 1,000 350 -Bicycles 0 0 0 :-Furniture 5 960 336 :-Sterilization kits 0 0 0 -Cold chain equipment 5 2,500 875 - 0| TOTAL 16,960 5,936 *Monthly depreciation (Zaires): 78,517 1*Petrol cons./month (Lt): 30 *USS/Lt: $1.00 Appendix D 47 A.3.1- Curative care: Prices and distances HC and competitors (Zaires) Price Price Price of Average first repeat drugs distance consul- consul- per Total to HC tation tation episode price (metrs) Health Center 200 0 558 758 0 |Competitor 1 758 2,000 |Competitor 2 526 5,000 |Competitor 3 I- ICost of drugs to HC 503 ;Cost of tab.exams to HC 50 1 :Cost of o.ph.prds.to HC 6 1 TOTAL 558 1 i i A.3.2- Deliveries: Prices and distances HC and competitors (Zaires) Average Price Price distance of of Total to HC delivery drugs price (metrs) Health Center 200 819 1,019 0 Competitor 1 1,019 2,000 Competitor 2 815 5,0001 |Cost of drugs to HC 756 |Cost of Lab.exams to HC 10 |Cost of o.ph.prds.to HC 53 1 |TOTAL 819 | 48 A Supply-Demand Model of Health Care Financing A.3.3- Pre-natal care: Prices and distances HC and competitors (Zaires) Average Price Price distance of regis- of Total to HC tration drugs price (metrs) Health Center 0 150 150 0 Competitor 1 150 2,000 ICompetitor 2 120 5,000 jCost of drugs to HC 100 ICost of lab.exams to HC 30 |Cost of o.ph.prds.to HC 53 TOTAL 183 i i A.3.4- Preschool care: Prices and distances HC and competitors (Zaire) Average Price Price distance of regis- of Total to HC tration drugs price (metrs) Heaith Center 0 150 150 0 lCompetitor 1 150 2,000 ICompetitor 2 120 5,000 ICost of drugs to HC 425 1 Cost of Lab.exams to HC 10 Cost of o.ph.prds.to HC 6 TOTAL 441 A.3.5- Chronic illnesses: Prices and distances HC and compets.(Zaires) Price Price Average per of drugs distance consul- per Total to HC tation episode price (metrs) IHealth Center 0 200 200 0 lCompetitor 1 200 2,000 Competitor 2 160 50,000 Competitor 3 |Cost of drugs to HC 400 ICost of Lab.exams to HC 50 lCost of o.ph.prds.to HC 6 ITOTAL 456 1 Appendix D 49 A.4.1- Pharmac. products and tab. exams: Avge. purchase price (Zaires) -EXCHANGE RATE: 350 EXCH.RATE: 350 :1. DRUGS Curative cons. 251.4 251.4 Delivery 378.0 378.0 Pre-Natal Cons. 100.0 100.0 Pre-School Cons. 425.2 425.2 Chronic ILLness 200.0 200.0 :2. LAB. EXAMS Curative cons. 100.0 100.0 Delivery 100.0 100.0 Pre-Natal Cons. 100.0 100.0 Pre-School Cons. 100.0 100.0 Chronic ILlness 100.0 100.0 13. OTHER PRODS.Curative cons. 5.7 5.7 Delivery 52.9 52.9 Pre-NataL Cons. 52.9 52.9 Pre-SchooL Cons. 5.7 5.7 Chronic ILLness 5.7 5.7 A.4.2- Number of laboratory exams and prescriptions per service at HC Average Average | number number of of J Laboratory prescriptions | exams (*) !Curative consuLtation(per episode) 0.5 2.0 'Delivery 0.1 2.0 j |Pre-NataL Consultation 0.3 1.0 |Pre-SchooL Consultation 0.1 1.0 |Chronic ILLness Consultation 0.5 2.0 | (*): One prescription = 1 type of drug prescribed 50 A Supply-Demand Model of Health Care Financing A.5.1- Minutes of personnel required by service, by personnel category Minutes of pers. required per service Birth Laborat. Doctor Nurse Attendant Technic. ICurative Consultation First Consultation 15 Repeat Consultation 5 'Delivery 150 :Pre-Natal Consultation 10 :Pre-School Consultation 10 Chronic Illness Consultation 20 |Laboratory Exam 15 | A.5.2- HC: Personnel necessary to meet demand, salaries, Labor costs Hours Monthly salary per day ------------------------------------- required By employee I Total to --------------------------- --------- satisfy Personnel % of doctor's 000s demand necess. Zaires salary US$ Zair - - - ---- - ---- - - - - - - - - - - - - - - - - - - Doctor 0.0 0 100,000 100% 286 0 'Nurse 11.5 2 50,000 50% 143 I 100 I :Birth A 3.4 1 80,000 80% 229 | 80 ILab.Tec 3.1 1 25,000 25% 71 | 25 | I I II I |Work hours per day: |Doctor 8 1 INurse 8 | Birth Attendant 8 1 ILab.Technician 8 1 Total salaries: I 205 Appendix D 51 A.6.1- Monthly demand at health center: Summary table Non- Non- Insured Insured Paying Totall Curative Consultations Number of Episodes 477 101 28 606 j First Consultations 477 101 28 606 1 Repeat Consultations 736 336 43 1116 Total consultations 1213 437 71 1722 1 jDeliveries 21 3 1 26 j IPre-NataL Care Registrations 12 2 1 15 1 IPre-Natal Care Consultations 117 16 7 140 Pre-SchooL Care Registrations 14 2 1 16 1 |Pre-School Care Consultations 224 30 13 267 1 IMaladies chroniques nouveaux cas 0.4 0.1 0.0 0.5 1 jMaladies chroniques consultations 25 4 1 31 | jLaboratory Exams 258 53 15 327 1 jPrescriptions 1,073 219 63 1,356 | A.6.2.1- Curative Episodes Demanded Per Month |Demand (episodes) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km from HC from HC from HC from HC from HC I TOTAL % :HCenter 304 159 94 44 5 | 606 67.7%1 jComp.1 81 59 41 21 3 1 205 22.9%1 |Comp.2 27 22 19 13 3 84 9.4%1 Comp.3 -- - -- - ----- - ----- ---- -- - -- - - - - ---I 411 241 154 78 11 I 894 100.0%11 A.6.2.2- Consultations Demanded per Curative Episode IDemand (consultations) as a function of price and j |distance, originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km from HC from HC from HC from HC from HC iHCenter 3.26 2.64 2.29 1.98 1.66 Comp.1 2.84 2.41 2.14 1.89 1.61 Comp.2 2.98 2.49 2.19 1.92 1.63 Comp. 3 I _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _I 52 A Supply-Demand Model of Health Care Financing A.6.2.3- TotaL Curative Consultations Demanded per Month IDemand (consuLtations) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km from HC from HC from HC from HC from HC TOTAL % -- - -- - ----- - ---- -- - -- - -- ----I- - - IHCenter 991 422 214 86 9 1,722 70.8%1 1Comp.1 229 143 88 40 5 504 20.7% |Comp.2 80 55 41 26 4 206 8.5%1 |Comp.3 I 1,300 619 344 152 18 2,432 100.0%1 A.6.3- DeLiveries Demanded per Month IDemand (deLiveries) originating between: I I 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km from HC from HC from HC from HC from HC TOTAL % - - - -- - - ---- -- - - ------ - - -- --I- - - HCenter 10 7 5 3 1 26 78.0%, Comp.1 2 1 1 1 0 5 15.9%1 jComp.2 0 0 0 0 0 2 6.1%1 12 9 7 4 1| 33 100.0%1 Appendix D 53 A.6.4.1- New Cases of Pre-Natal Care Registered per Month :Demand (number of new clients) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km | from HC from HC from HC from HC from HC TOTAL % - - - - - - - - - - - - - - - - - ---- - ---- - ---- - - HCenter 7 4 3 1 0 15 72.4% :Comp.1 2 2 1 1 0 6 27.6% Comp.2 0 0 0 0 0 0 0.0% - - - - - - - ---- - - - - - ---- - - - - - ---- - - 9 6 4 2 0 21100.0%| A.6.4.2- Number of Pre-Natal Care Consultations per Pregnancy per woman registered :Demand (consultations) as a function of distance: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km from HC from HC from HC from HC from HC ConsuLt 2.4 1.7 1.3 1.0 0.7 A.6.4.3- Pre-Natal Consultations Demanded per Month Demand (consultations) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km from HC from HC from HC from HC from HC TOTAL % IHCenter 82 34 17 7 1 | 140 73.8% IComp.1 24 14 8 4 0 50 26.2%: Comp.2 0 0 0 0 0| 0 0.0%| 105 47 25 11 1 190 100.0%| 54 A Supply-Demand Model of Health Care Financing A.6.5.1- Pre-Natal care: Number of new cases (children registering) during the month 'Demand (children registering) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km from HC from HC from HC from HC from HC TOTAL % IHCenter 7 4 3 2 0 16 72.3%1 Comp.1 2 2 1 1 0 6 27.7% Comp.2 0 0 0 0 0 0 0.0% 9 6 4 2 0 22 100.0%1 A.6.5.2- Pre-Natal Care: Total number of children registered ,Demand (children registered) originating between: I I 0-1 Km 1-2 Km 2-3 Kmn 3-5 Km > 5 KmI from HC from HC from HC from HC from HC I TOTAL % HCenter 411 241 159 88 14 | 913 72.3%1 Comp.1 118 98 77 48 8 349 27.7%j Comp.2 0 0 0 0 01 0 0.0% I I 529 339 237 136 22 1262 100.011' A.6.5.3- Pre-Natal Care: Annual nutber of consultations demanded at the HC as a function of distance and age of the child 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km | Age from HC from HC from HC from HC from HC I 1 an 10.4 8.2 6.4 4.4 1.9 |2 ens 5.9 4.6 3.5 2.3 1.0 13 ans 3.2 2.4 1.8 1.2 0.5 14 ans 1.6 1.2 0.9 0.6 0.2 |5 ans 0.8 0.6 0.5 0.3 0.1 Appendix D 55 A.6.5.4- Pre-Natal care: Total number of consuLtations demanded per month :Demand (consultations) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km : from HC from HC from HC from HC from HC : TOTAL % HCenter 150 69 35 13 1 1 267 73.7% lComp.1 43 28 17 7 0 95 26.3% Comp.2 0 0 0 0 0 1 0 0.0% 194 96 52 20 1 363 100.0%j A.6.6.1- Chronic ILLnesses: Number of new cases demanded per month IDemand (new cases) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km : from HC from HC from HC from HC from HC TOTAL % - - - - - - - - - - - - - - - - - - - - I - - - -I-- - :HCenter 0.2 0.1 0.1 0.0 0.0 i 0.5 67.2%: Comp.1 0.1 0.0 0.0 0.0 0.0 0.2 25.2%: Comp.2 0.0 0.0 0.0 0.0 0.0 : 0.1 7.5% :Conc.3 0.3 0.2 0.1 0.1 0.0 1 0.7 100.0%: A.6.6.2- Chronic Illnesses: TotaL number of patients registered :Demand (patients registered) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km from HC from HC from HC from HC from HC TOTAL % :HCenter 85 49 31 17 2 1 184 67.2%: Comp.1 24 20 15 9 1 69 25.2% lComp.2 6 5 5 4 1 21 7.5%1 Comp.3 i I I I 11I3 5 9 5 ~ 230.% 56 A Supply-Demand Model of Health Care Financing A.6.6.3- Chronic ILLnesses: Total number of consuLtations demanded per month (*) |Demand (consultations) originating between: * 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km I from HC from HC from HC from HC from HC TOTAL % jHCenter 14 8 5 3 0 31 67.2%1 lComp.1 4 3 2 1 0 11 25.221 Comp.2 1 1 1 1 0 3 7.5%1 IComp.3 - - - -- - - ---- - ---- - - - - - - - - - - ---- - - 19 12 8 5 1: 46 100.0%1 (*) Assumption: two consultations per patient registered per year. A.7.1- Monthly revenue from services HC -- Summary table (000s Zaires) 11.SERVICES-----DEMAND -------- (5) (6) (7) 1 (1) (2) (3) (4) Revenue Revenue Percent Non- Non- Price non-insr insured monthly Insurd Paying Insured paying(1)x(4) co-pmt. revenue jCur.Ca,lv 477 28 101 200 95 4 18.8% !Cur.Ca,rv 736 43 336 0 0 0 0.0% iDeliverie 21 1 3 200 4 0 0.8% 1CPN regis 12 1 2 0 0 0 0.0%i 1CPN visit 117 7 16 0 0 0 0.0% !CPS regis 14 1 2 0 0 0 0.0%i !CPS visit 224 13 30 0 0 0 00%1 Ch.ca.reg 0 0 0 0 0 0 0.0% Ch.ca.vis 25 1 4 0 0 0 0.0%j SUB-TOTAL DIRECT PAYMENTS 100 4 19.7% SUB-TOTAL PREMIUM PAYMENTS 125 23.7%1 !TOTAL REVENUE SERVICES 229 43.3% Note: Cur Ca 1v = curative care, first visits; Cur Ca rv = curative care repeat visits; PNS = prenatal care registration; CPN visit = prenatal care visits; CPS regis = preschool registration; CPS visit = preschool visits; Ch. ca. reg. = chronic care registration; Ch. ca. vis = chronic care visits. A.7.2- Monthly revenue from drugs HC -- Summary table (000s Zaires) |2.DRUGS ------DEMAND -------- (5) (6) (7) 1 (1) (2) (3) (4) Revenue Revenue Percent Non- Non- Price non-insr insured monthly Insurd Paying Insured paying(1)x(4) co-pmt. revenue Cur.visit 477 28 101 558 266 11 52.5%1 IDeLiverie 21 1 3 819 18 1 3.4%j ,P-N care 12 1 2 150 2 0 0.4% IP-S care 14 1 2 150 2 0 0.4%1 jChr.iLLn. 0 0 0 200 0 0 0.0%j SUB-TOTAL DRUGS 288 12 ITOTAL REVENUE DRUGS 300 56.7%1 TOTAL REVENUE (TabLe A.7.1 + Table A.7.2) 528 100.0%1 Appendix D 57 A.8- Health center monthly income statement (000s Zaires) |Revenue Services (direct payments) 104 Drugs 300 Premiums from the insured 125 528 lExpenses Personnel 205 Drugs 370 PetroL 11 Office supplies 10 Supervision expenses 10 605 IProfit (Loss) (77) Depreciation 79 jProfit (Loss) after depreciation (156) A.9- Demand Equations: Coefficients CONSTANT -------------- Price Health Compe- times center titors Price Distance income Sigma lCurative Care 1.00 0.60 -3E-03 -7E-04 5.OE-08 0.50 iDeliveries 5.00 4.00 -3E-03 -7E-04 5.OE-08 0.50 jPre-NataL Care 1.00 0.60 -4E-03 -7E-04 5.OE-08 0.50 1Pre-SchooL Care 1.00 0.60 -4E-03 -7E-04 5.OE-08 0.50 Chronic Care 1.00 0.60 -4E-03 -7E-04 5.OE-08 0.50 A.10- Price, Income, and Distance Elasticities of Demand Curativ Detiv. P-N CarP-S Care Chronic: IHCenter Price -0.79 -0.33 -0.17 -0.16 -0.24 Income 0.26 0.06 0.03 0.02 0.04 Distan -0.18 -0.07 -0.13 -0.12 -0.14 jCompet.1 Price -1.30 -1.33 -0.37 -0.36 -0.49 Income 0.26 0.06 0.03 0.02 0.04 Distan -0.30 -0.30 -0.28 -0.28 -0.29 jCompet.2 Price -1.00 -1.18 -0.36 -0.36 -0.46 * Income 0.03 -0.15 0.00 -0.01 0.00 1 | Distan -0.33 -0.34 -0.35 -0.35 -0.34 | jCompet.3 Price -- -- -- Income -- -- -- Distan -- -- -- APPENDIX E Computer Model, Hospital Component Table E.1 Population distribution 0-1 Km 1-5 Km 5-20 Km 20-50 Km >50 Km Total :Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%lCost of I A Number 35000 25000 20000 15000 5000 100000 drugs, S Drugs, Total Average lLab.exms I Medical exams, patient length I& other I C fees & other expend. stay(ds)lphar.prdl H PRICE ------- ------- ------ --------|-------- 1 0 Hosp. without surgery 3000 3500 6500 6.0 4403 N S Hosp. with surgery 4000 5500 9500 10.0 6136 P P DeLiveries 2500 2500 5000 6.0 | 3707 U I i_ |T T Radius health area(Km): 10 -Ins.premium/yr.(Zaires) 1000 A -Population: 100000 'Percentage insured: 10% T L -Currency: Zaire -Copayment insured: 10% A -Exchge.rate(Zaire/USS): 350 -Percent non-paying: 5% B *Household income/month: 20000 -Superv.exp.(% rev.HC): 5% L -Hsd.inc.insured/non-ins 1.00 -Other pers./100 beds: 30 E -Percent occupancy beds: 80% TABLE E.2 ALL INFORMATION IS MONTHLY -------Utilization------ Nbr.bds % mar-' Non-insr Insured Total Bed-days necessa ket Ho| B -------- ------- ------- -------- ----- ------ | A !Hosp. without s 80 14 95 570 23 74%1 S Hosp. with surg 69 15 84 838 34 79%1 I Deliveries 103 17 120 720 30 75%1 C I H 000s Zaires_ 87 1 0 0 1(1) REVENUE Services 1,582 U S Drugs 887 2,469 | T P 1(2) EXPENSES Persont. 2,190 Utiliz'n rates P I Drugs 1,381 '--------------- UT Other 10 3,581 1 Non-insr Insrdl T A (3) PROFIT [(1)-(2)] (1,112) -------- ------ L (4) SUPERVISION EXPENS. 79 |H.w/o s. 27% 43%1 T :(5) DEPRECIATION 326 |H.wth s. 23% 45%| A (6) PROFIT[(3)-(4)-(5)] (1,517):Deliver. 29% 43%| B IL E 58 Appendix E 59 D.1- Epidemiologic and demographic data IILLness episodes that require hospitalization without surgery, I(cases per 100 inhabitants per year) 4.0 II |ILLness episodes that require hospitalization with surgery, :(cases per 100 inhabitants per year) 4.0 |Crude birth rate (per 1000 inhabitants) 47 D.2- Investment and some recurrent costs at hospital Invest- Invest- (These Labels can be Useful ment ment changed by the user) Life (US$) (000 Zai) -BuiLding 30 80,000 28,000 j-Medical equipment 5 20,000 7,000 j*Refrigerator 0 5,000 1,750 |*X-Ray Machine 2 3,000 1,050 |*BicycLes 0 3,000 1,050 |*Furniture 5 15,000 5,250 ISterilization kits 0 5,000 1,750 |- 0j 1.0 TOTAL 131,000 45,850 [-Monthly depreciation (Zaires): 325,694 60 A Supply-Demand Model of Health Care Financing D.3.1- Hosp. w/o surgery: Prices and distances Ho and compet's (Zaires) Average Drugs, distance Medical exams, and Total to Hosp. fees other expend. (metrs) Hospital 3,000 3,500 6,500 0 Competitor 1 5,000 5,000 Competitor 2 4,000 20,000 :Cost of drugs to Hosp. 503 | :Cost of lab.exams to Ho 150 ICost of X Rays Hosp. 1750 ICost of o.ph.prds.to Ho 2,000 ITOTAL 4,403 | D.3.2- Hosp.with surgery: Prices and distances Ho and compet's (Zaires) Average Drugs, distance Medical exams, and Total to Hosp. fees other expend. (metrs) |Hospital 4,000 5,500 9,500 0 |Competitor 1 9,000 5,000 |Competitor 2 7,000 20,000 ICost of drugs to Hosp. 756 | :Cost of Lab.exams to Ho 30 | :Cost of X Rays Hosp. 350 | |Cost of a.ph.prds.to Ho 5,000 | |TOTAL 6,136 Appendix E 61 D.3.3- Deliveries: Prices and distances Hosp. and competitors (Zaires) Average Drugs, distance MedicaL exams, and TotaL to Hosp. fees other expend. (metrs) lHospitaL 2,500 2,500 5,000 0 lCompetitor 1 4,000 5,000 lCompetitor 2 2,000 20,000 |Cost of drugs to Hosp. 567 |Cost of Lab.exams to Ho 90 jCost of X Rays Hosp. 1050 lCost of o.ph.prds.to Ho 2,000 | 1TOTAL 3,707 | i i D.4.1- Pharmac. products and Lab. exams: Avge. purchase price (Zaires) EXCHANGE RATE: 350 EXCH.RATE: 350 :1. DRUGS Hosp. w/o surgery 251.4 251.4 Hosp. with surgery 378.0 378.0 Deliveries 567.0 567.0 2. LAB. EXAMS Hosp. w/o surgery 300.0 300.0 Hosp. with surgery 300.0 300.0 Deliveries 300.0 300.0 13. X RAY EXAMS Hosp. w/o surgery 3,500.0 3,500.0 Hosp. with surgery 3,500.0 3,500.0 DeLiveries 3,500.0 3,500.0 14. OTHER PRODS.Hosp. w/o surgery 2,000.0 2,000.0 Hosp. with surgery 5,000.0 5,000.0 Deliveries 2,000.0 2,000.0 62 A Supply-Demand Model of Health Care Financing D.4.2- Nbr.of tab.exams, X rays, and prescriptions per service at Hosp. Average Average Average nunber number number of of of Laboratory X ray prescriptions exams exams (*) H-ospitalization without surgery 0.5 0.5 2.0 lHospitalization with surgery 0.1 0.1 2.0 jDelivery 0.3 0.3 1.0 (*): One prescription = 1 type of drug prescribed D.5.1- Minutes of personnel required by service, by personnel category Minutes of pers. required per service Laborat. X Ray Doctor Nurse Midwife Technic. Technic. Otherl jHosp.w/o surgery 60 360 10 !Hosp.with surgery 180 720 10 ( jDelivery 360 60 20 10 jLaboratory exam 20 10 jX Ray exam 20 10| D.5.2- Hosp.: Personnel necessary to meet demand, salaries,labor costs Hours Monthly salary per day ------------------------------------- required By employee I Total (to --------------------------- --------- satisfy Personnel % of doctor's j 000s demand necess. Zaires salary US$ Zair IDoctor 13.3 2 150,000 100% 429 300 |Nurse 88.3 12 45,000 30% 129 540 'Midwife 4.6 1 45,000 30% 129 45 1 !Lab.Tec 2.7 1 45,000 30% 129 45 j 1XRey T. 1.2 1 45,000 30% 129 45 1 !Other 3.1 1 45,000 30% 129 45 | 'Additional Pers 26 45,000 30% 129 1,170 II I I_ _ _ _ _ __ _ _ _ _ _I I ,Work hours per day: j j 'Hospital 8 Total salaries: 1 2,190 | 1 I__ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ I Appendix E 63 D.6.1- Monthly demand at Hospital: Summary table Non- Non- Insured Insured Paying Total -- :Hospitalization without surgery 76 14 4 95 Hospitalization with surgery 64 15 4 84 ,Deliveries 99 17 4 120 Laboratory exams 74 14 4 92 X Ray exams 74 14 4 92: Prescriptions 380 76 22 478 D.6.2- Hospitalizations without surgery demanded per month |Demand (hospitalizations) originating between: 0-1 Km 1-5 Km 5-20 Km 20-50 Km >50 Km from Ho from Ho from Ho from Ho from Ho Total % IHosp. 39 25 18 11 3 95 74.3%| Comp.1 10 8 7 5 1 31 24.6%1 jComp.2 0 0 0 0 0 1 1.1%j 50 33 25 16 4 128 100.0%1 D.6.3- Hospitalizations with surgery demanded per month |Demand (hospitalizations) originating between: 0-1 Km 1-5 Km 5-20 Km 20-50 Km >50 Km from Ho from Ho from Ho from Ho from Ho Total % Hosp. 34 22 15 10 2 83 78.6% comp.1 7 5 5 3 1 21 20.3%1 Comp.2 0 0 0 0 0 1 1.1%1 42 27 20 13 3 | 106 100.0%| 64 A Supply-Demand Model of Health Care Financing D.6.4- Deliveries demanded per month IDemand (number of deliveries) originating between: 0-1 Km 1-2 Km 2-3 Km 3-5 Kn > 5 Kn from HC from HC from HC from HC from HC | TOTAL % -- - -- - ----- - ---- -- - -- - -- ---- - - - |HCenter 50 32 23 14 3 122 75.5%1 |Comp.1 12 10 8 6 2 1 38 23.2%1 |Comp.2 1 0 0 0 0| 2 1.3%1 i --- ---I- --- 63 42 31 21 5 162 100.0%1 D.7.1- Monthly revenue from services Hos.---Summary table (000s Zaires 11.SERVICES-----DEMAND -------- (5) (6) (7) 1 (1) (2) (3) (4) Revenue Revenue Percent Non- Non- Price non-insr insured monthly insurd paying Insured paying(1)x(4) co-pat. revenue -- - - -- - - - -- - - - -- - - - - - - --I- - - - 1H.w/o sur 76 4 14 3000 228 4 9.4%1 1H.wth sur 64 4 15 4000 258 6 10.7% IDeliverie 99 6 17 2500 249 4 10.2%1 ISUB-TOTAL DIRECT PAYMENTS 734 14 30.3% ISUB-TOTAL PREMIUM PAYMENTS 833 33.8%1 ITOTAL REVENUE SERVICES 1,582 64.1%1 D.7.2- Monthly revenue from drugs Hos.---Summary table (000s Zaires) 12.DRUGS ------DEMAND -------- (5) (6) (7) 1 (1) (2) (3) (4) Revenue Revenue Percent Non- Non- Price non-insr insured monthly insurd paying Insured paying(1)x(4) co-pmt. revenue 1H.w/o sur 76 4 14 3500 266 5 11.0%1 IH.wth sur 64 4 15 5500 355 8 14.7%1 IDeliverie 99 6 17 2500 249 4 10.2%1 SUB-TOTAL DRUGS 869 17 ITOTAL REVENUE DRUGS 887 35.9%8 ITOTAL REVENUE (Table D.7.1 + Table D.7.2) 2,469 100.0%1 Appendix E 65 D.8- Hospital monthly income statement (000s Zaires) IRevenue Services (direct payments) 749 Drugs 887 Premiums from the insured 833 2,469 lExpenses PersonneL 2,190 Drugs 1,381 Office supplies 10 Supervision expenses 79 3,660 IProfit (Loss) (1,191) Depreciation 326 jProfit (loss) after depreciation (1,517) D.9- Demand Equations: Coefficients CONSTANT ------------- Price Hosp- Compe- times itaL titors Price Distance income Sigma Mu -Hsp /sr 15 . --0- -3E- 2.5E- 0.50 .50 (Hosp. w/o surg. 1.50 1.00 -7E-04 -3E-04 2.5E-08 0.50 0.50 jHosp. wth surg. 1.50 1.00 -7E-04 -3E-04 2.5E-08 0.50 0.50 |DeLiveries 1.50 1.00 -7E-04 -3E-04 2.5E-08 0.50 0.50 D.10- Price, Income, and Distance ELasticities of Demand | H.w/o s H.wth s Deliv. jHospitaPrice -0.60 -0.89 -0.43 Income 1.08 1.61 0.78 1 Distance -0.07 -0.07 -0.07 jCompet.Price -0.84 -1.57 -0.68 | Income 0.33 1.36 0.28 | Distance -0.13 -0.13 -0.13 ICompet.Price -0.79 -1.39 -0.40 | Income -0.17 0.36 -0.72 Distance -0.15 -0.15 -0.15 | APPENDIX F User's Manual The Program The model has been developed using Lotus 1-2-3 release 2.01. The model is interactive and is run through a system of menus written in Lotus' macroprogramming language; it is contained on four Lotus 1-2-3 worksheets. Software and Hardware Requirements To run the model you need the to have an IBM-compatible microcomputer with at least 640 kilobytes of RAM memory, Lotus 1-2-3, release 2.01 or higher, and a hard disk with about 1 megabyte of free space. Lotus 1-2-3 and the model occupy about 87 percent of a 640-K RAM memory. Thus, if you have a computer with 640-K RAM you should make sure that the RAM memory does not have any resident program prior to loading 1-2-3 and the model. Any program residing in the RAM memory, however small, may leave you with insufficient RAM space to load the model. If you have a printer connected to your computer you can print the model's tables through the program. The print defaults of 1-2-3 have been set to the following: left margin: 20; right margin: 240; top margin: 0; and page length: 80. It is convenient (but not necessary) to print in portrait condensed mode by appropriately defining the setup string. Check the appropriate setup for your printer. If you wish to change any of the above setups, you need to interrupt the execution of the model (see Interrupting and Restarting the Model) and change the setups using the usual 1-2-3 commands. Installing the Program The diskette containing the program has four files. Their original names and content are as follows: File Name File Content OPEN.WK1 Opening menu that provides access to the health center model, the hospital model, and the aggregate projections COMBINE.WK1 Aggregate projections spreadsheet HCMODEL.WK1 Health center model INMODEL.WK1 Hospital model To install the program follow the steps below. 66 (1) On the C partition of the hard drive create a directory called "model". To create the directory, proceed as follows: (a) Get the DOS prompt (b) Go to the root directory of the C partition of your hard disk (c) At the prompt (which will look like C:\>), type md model and then press the Enter key (make sure that you leave a space between md and model) (2) Go to the model directory by typing cd model followed by Enter (make sure that you leave a space between cd and model) (3) Copy all four program files on the model directory of drive C. To do this, insert the program diskette in drive A of your computer, type copy a:*.* and then press the Enter key (make sure that you leave a space between COPY and a:*. *) (4) Wait until all four files are copied. You will know that they have been copied when you see the following DOS message on the screen: 4 file(s) copied Executing the Program To execute the program proceed as follows: (1) Invoke Lotus 1-2-3 by typing 123 followed by Enter 67 (2) Make model on the C drive your default directory for Lotus 1-2-3. To do this, type the Lotus commands Worksheet Gobal Default Directory , or /WGDD If a different directory is the default, press the Escape (ESC) key until you see in the command line the following message: Default directory: Then type c:\model\ followed by Enter. Then press the command Quit (Q) to return to the main command menu. (3) Detach all Lotus 1-2-3 Add-Ins, if any, such as WYSIWYG. Consult your Lotus manual on how to do this. (4) Load the model by using the File Retrieve command (OR) and by typing the file name open. The command line should look as follows: Name of file to retrieve: c:\model\open then press Enter. Execution will start automatically. Saving your Work After having worked with the health center or the hospital model, you may want to save your work. Both HCMODEL and INMODEL offer you this option. For example, if you are working with INMODEL and want to save your worksheet, choose option (0) of the opening hospital menu, below, by entering the number 0. (1) Hospital (2) Summary table (0) QUIT or SAVE You will then be presented with the following menu: (1) END session WITHOUT saving worksheet (2) SAVE worksheet (0) PREVIOUS MENU 68 Enter the number 2 to save your file as it is. Beware that the file will be saved with the same name INMODEL on top of the previous version of INMODEL. That is, the new INMODEL file with the updates you made during the session will replace the old INMODEL file. If you want to conserve the old INMODEL file, you may want to save it elsewhere, for example on a diskette, prior to starting the current session. Moving around Menus and Submenus The model provides you with an extended system of menus and submenus. The model's menu structure is shown in Appendix C. There are two types of menu in the model: those that appear in a rectangular box in the middle of the screen and those that appear on a single line at the top of the screen. Rectangular menus always appear alone on the screen. Single-line menus always accompany a model table (see description of tables in Input, Output, and Input-Output Tables, below). Rectangular menus offer a series of numbered options. To go to the desired menu, you must hit the number shown in parentheses on the left of the option's name. If you hit a number that falls outside the range of the menu or type any non-numerical character, the program will beep while giving you an error message and will then return to the same menu. The number zero (0) always takes you to the preceding menu. Single-line menus offer a series of options that are not numbered. To execute the desired option, you have two possibilities: (1) hit the first character of the option (for example hit the letter "C" if you wish to calculate); or (2) highlight the desired option using the right and left arrows of your keyboard and hit Enter. Input, Output, and Input-Output Tables The model's information is presented in tables. Most tables fit on one 24-line screen, although a few are bigger (vertically). Tables are always accompanied by a single-line menu that allows you to enter data into the table (if applicable, see below), return to the previous menu, recalculate the spreadsheet, or execute other options. There are three types of tables: (1) input tables; (2) output tables; and (3) input-output tables. Input tables are those that contain only information given by the user and do not provide any model-calculated result. Output tables are the ones that contain only information calculated by the model. Finally, input-output tables are the ones which contain some input and some output information. Input and input-output tables allow the user to move the cursor into the table to add or modify any input variable. To do this, you must hit the letter "E" or highlight the "ENTER DATA" option followed by Enter. To move around the table use the four arrows of your keyboard. To prevent problems, the cursor is only allowed to move within the input cells. If you cannot move into a desired cell it means that it is an output cell and you should not try to get in there. If you have a color monitor, Lotus will display input cells in a different color. 69 Output tables are also accompanied by a single-line menu, although the menu does not allow you to move the cursor into the table. Entering Data in the Model To input data into a given input or input-output table proceed as follows: (1) Identify the appropriate model table. If you are unfamiliar with the model structure, refer to Appendix C, Appendix D, or Appendix E to learn what types of tables are in the model, where they are, and what information they contain; (2) using the appropriate sequence of menus, go to the desired table; (3) once the table appears on the screen, move the cursor into the table by hitting the letter "E" or by highlighting the "ENTER DATA" option followed by Enter; (4) position the cursor in the cell where you want to enter information; (5) type the entry normally; and (6) hit the Enter key twice to return to the table's menu. If you did not enter any information hit Enter only once to return to the menu. AVOID HITTING THE ESCAPE KEY AT ALL TIMES. Under certain circumstances, hitting the Escape key may interrupt the operation of the model. Recalculating the Model THE MODEL DOES NOT DO AN AUTOMATIC RECALCULATION AFTER YOU INPUT DATA. The model consists of hundreds of interrelated mathematical formulas. If you change an entry of the model all the formulas that are directly or indirectly related to that entry will be affected. This, however, will not occur automatically. After you modify an input entry and return to the single-line menu associated with that table (see Entering Data in the Model), you must ask the model to recalculate itself. To do this, hit the letter "C" or highlight the option "CALCULATE" with the cursor and hit Enter. While the model is recalculating, the single-line menu will disappear. When the model has finished recalculating, the menu will reappear at the top of the screen. If any of the values contained in the table are affected by the change(s) you made, you will see changes occurring while the recalculation takes place. DO NOT HIT ANY KEY WHILE RECALCULATION TAKES PLACE. Saving Your Work or Quitting After using the model and doing some simulations you may want to save your work, especially if you have spent some time modifying several of the model variables to characterize a particular situation (for example, health financing at the district level in rural Mali). To save the spreadsheet go to the opening menu and choose option (0) ("QUIT or SAVE"). To save your file, select option (2) ("SAVE worksheet"). The program will automatically save the current worksheet on top of the older version on your disk. This means that you will lose what was on the older version. Make sure that you really want to replace the old version before choosing option (2). Notice that you can also quit the program without saving. To do so, choose option (1) ("END session WITHOUT saving worksheet"). The program will ask you to confirm that choice in order to make sure that you really want to quit without saving your work. 70 What to do if ERROR Messages Appear Errors can occur for several reasons, including bugs in the program or user mistakes when entering data. If one or more ERROR message appears on the tables or at the top-right or bottom-left of the screen, hit the keys Control and Break simultaneously, followed by the Enter key. This will allow you to interrupt the execution of the model. You may then want to restart model execution by hitting the Alternate (Alt) key and the letter "A" simultaneously. Go back to the last table where you entered data and see if you made data entry mistakes there (for example, you entered a letter instead of a number). If you find such an error, correct it by entering the right value in the appropriate cell and ask the model to recalculate. If the ERROR message continues to appear, interrupt the model by hitting the Control and Break keys simultaneously, followed by Enter and exit the model. To exit, use Lotus' command /QY. DO NOT SAVE A MODEL WORKSHEET THAT HAS ERRORS ON TOP OF THE PREVIOUS VERSION. Retrieve the previous version of the model and try again. If errors reappear, it may be an indication of a programming error. In such cases, please let me know. In the case of programming errors and other types of error, the Lotus macro routines may provide an error message at the bottom of the screen. Please write down the message, if any. This will help me debug the program. Interrupting and Restarting the Program Sometimes you may want to interrupt the execution of the program to change the printer defaults or for other reasons. In that case, proceed as specified above: press the Control and Break keys simultaneously and the press the Enter key. This will take you to the Lotus-Ready mode. To return to the program execution mode, hit the Alternate (Alt) key and the letter "A" simultaneously to go back to the opening menu. Printing the Model Tables To print one or more model tables, select option (3) ("PRINT") from the main health center or hospital menus. Once on the print table, move up or down with the keyboard arrows to select the table or group of tables you wish to print. To make the selection, enter the letter "I" in the appropriate box and then hit Enter twice. The program will print the chosen table(s) automatically. Prior to printing, make sure the printer is connected and on-line and also that you have enough paper. If an error occurs at this point interrupt the program (see Interrupting and Restarting the Program) and check your printer and the computer-printer connection. Try to print again by restarting execution and returning to the corresponding print menu. Entering Data: Where to Start? The original copy of the model that you will receive will already have data in it. These data were entered by me and correspond to the examples shown in the text. You will most likely want to change some of these data to adequately represent your desired scenario. You can enter new data in any order; that is, you can start with any table and, through the menu system, follow any desired path. It is usually 71 helpful to know in advance which entries you are going to modify. To do this, you may want to print the entire set of tables of the model or study the sets of tables of Appendixes D and E. That will allow you to identify the entries that need to be changed. Also, become familiarized with the structure of the menu system shown in figure C. 1 to know the adequate menu path that you must follow to reach the required tables. Special Features of the Model (A) ERASING COMPETITORS As you will see, the menus of the prices-distances tables (tables A.3.1 ---A.3.5 for the health center and tables D.3. 1---D.3.3 for the hospital) allow you to add new or erase existing competitors. To add competitors, simply use the "ENTER DATA" option and enter their price and distance from the health center (in meters). To erase a competitor, select the option "DELETE" of the menu associated with then simply specify the sequential number of the competitor that you wish to erase. (B) THE REDUCED AND EXPANDED MENUS The reduced menu of the health center and hospital components was designed to help you to quickly perform simulations without the use of multiple tables. As can be seen from the input and output tables of the reduced menu, only basic input and output information has been included. The input table contains primarily the prices of the health center (or hospital), the population size and distribution, the market size, the insurance parameters, and so on. Those variables can only be modified through the input table of the reduced menu and do not reappear as input variables elsewhere in the expanded menu. Conversely, none of the exogenous variables of the expanded menu are repeated in the input table of the reduced menu. Thus, in most cases, you will have to switch from the reduced menu to the expanded menu to change certain exogenous variables. Also, you will typically have to switch from the reduced menu to the expanded menu if you wish to see more aggregated or disaggregated output data. (c) SPECIFYING DEMAND COEFFICIENTS AND ASSOCIATED DEMAND ELAsTICrrIEs Tables A.9 and D.9, of the health center and hospital components of the model, respectively, allow the user to enter the coefficients of the nested logit demand equations. Because of the highly nonlinear nature of logit, it is difficult to interpret both the sign and the magnitude of the coefficients. The implied elasticities of demand have more intuitive meaning, however. Unfortunately, because of the nonlinear nature of the demand equations, the elasticities cannot be entered by the user directly. Instead, the user must enter the demand coefficients. After doing so, the user can move to table A. 10 (for the health center, or table D. 10 for the hospital) and ask the model to recalculate the elasticities, using the recently entered demand coefficients. If the implied elasticites of demand do not match the user needs, he or she can go back to the coefficients table, change the coefficients, and reiterate until the desired price, income, and distance elasticities are obtained. This process usually will take only a few minutes. 72 APPENDIX G Additional Simulation Results Exhibit G.1. An Increase in Insurance Coverage Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total B H ------- ------- ------- ------- ------- -------- Cost of AE :Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%1 drugs, 1 S A Number 3500 2500 2000 1500 500 10000 tab.examsl IL First Repeat Total & other I C T visit visits Drugs expend. pher.prd, H PRICE ------- ------- -------- ------ -------- |Curative care 200 0 558 758 558 1 N C ,Deliveries 200 --- 819 1019 819 1P E jPre-natal Care 0 --- 150 150 183 U N lPre-Schoot Care 0 --- 150 150 1 441 1 T T jChronic Care 0 --- 200 200 j 456 1 E i IT R -Radius health area(Km): 10 *Ins.premium/yr.(Zaires) 1500 A -Population: 10000 -Percentage insured: 20% B -Currency: Zaire -Copayment insured: 20% L -Exchge.rate(Zaire/USS): 350 -Percent non-paying: 5% E *Household income/month: 20000 'Superv.exp.(% rev.HC): 10% -Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY ---------New cases------ ------Total visits------ % mar-' Non-insr Insured Total Non-insrd. Insrd. Total ket HCj B H -------- ------- ------- -------- ------- ---- ------ A E ICur.Ca 449 202 650 1142 704 1846 70%1 S A IDeLiv. 20 6 27 --- --- --- 80%j I L IP-N Ca 12 3 15 110 32 142 73%1 C T IP-S Ca 13 4 16 211 60 271 73%1 H lChr.Ca 0 0 0 24 7 31 68%1 0 000s Zaires U C 1() REVENUE Services 346 HC utiliz.rates| T E Drugs 278 624 | ---------------P N 1(2) DEPENSES PersonL. 205 1 Non-insr Insrdl U T Drugs 395 -------- ------ I T E Other 30 630 1 Cur.Ca 17% 30%1 R 1(3) PROFIT [(1)-(2)] (7)1 Deliv. 64% 82%1 T 1(4) SUPERVISION EXPENS. 35 1 P-N Ca 37% 43%1 A (5) DEPRECIATION 79 ' P-S Ca 41% 47%1 B 1(6) PROFIT[(3)-(4)-(5)] (120)1 Chr.Ca 36% 44%| L jE 73 74 A Supply-Demand Model of Health Care Financing Exhibit G.2. A Drop in Insurance Premium Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total B H ------- ------- ------- ------- ------- ---------Cost of IA E Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%1 drugs, I S A INumber 3500 2500 2000 1500 500 10000 lab.examsj I L First Repeat Total I& other C T visit visits Drugs expend.Iphar.prdl H PRICE ------- ------- -------- ------ -------- I lCurative care 200 0 558 758 558 1 N C ,Deliveries 200 --- 819 1019 819 P E IPre-natal Care 0 --- 150 150 183 U N IPre-School Care 0 --- 150 150 441 T T IChronic Care 0 --- 200 200 456 E T R Radius health area(Km): 10 *Ins.premium/yr.(Zaires) 800 A Population: 10000 -Percentage insured: 10% B -Currency: Zaire -Copayment insured: 20% L -Exchge.rate(Zaire/USS): 350 -Percent non-paying: 5% E -Household income/month: 20000 -Superv.exp.(% rev.HC): 10% -Hsd.inc.insured/non-ins 1.00 TABLE 8.2 ALL INFORMATION IS MONTHLY ---------New cases------ ------Total visits------ % mar-I Non-insr Insured Total Non-insrd. Insrd. Total ket HCI B H -------- ------- ------- -------- ------- ---- ------ A E iCur.Ca 505 101 606 1285 437 1722 68%1 S A IDeLiv. 23 3 26 --- --- --- 78%j I L IP-N Ca 13 2 15 124 16 140 72%1 C T IP-S Ca 14 2 16 237 30 267 72%j H jChr.Ca 0 0 0 27 4 31 67%.10 i _ _ 000s Zaires I U C j(1) REVENUE Services 170 HC utiliz.rates! T E Drugs 300 470 ---------------P N 1(2) DEPENSES Person[. 205 Non-insr Insrdl U T Drugs 370 -------- ------ T E Other 31 605 Cur.Ca 17% 30%, R I(3) PROFIT [(1)-(2)] (135)1 Deliv. 64% 82%1 T 1(4) SUPERVISION EXPENS. 17 | P-N Ca 37% 43% A I(5) DEPRECIATION 79 1 P-S Ca 41% 47%1 B ,(6) PROFIT[(3)-(4)-(5)] (231)1 Chr.Ca 36% 44% L E Appendix G 75 Exhibit G.3. A Devaluation of the Country's Currency Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total B H ------- ------- ------- ------- ------- -------- Cost of | A E IPercent 35.0% 25.0% 20.0% 15.0% 5.0% 100% drugs, IS A Number 3500 2500 2000 1500 500 10000 tab.exams I L First Repeat Total & other C T visit visits Drugs expend.lphar.prd H PRICE ------- ------- -------- ------ -------- I Curative care 200 0 558 758 638 1 N C IDeliveries 200 --- 819 1019 936 P E jPre-natal Care 0 --- 150 150 209 U N jPre-School Care 0 --- 150 150j 504 T T Chronic Care 0 --- 200 200 I 521 1 E i_ I T R Radius health area(Km): 10 -Ins.premium/yr.(Zaires) 1500 A -Population: 10000 *Percentage insured: 10% B -Currency: Zaire *Copayment insured: 20% L -Exchge.rate(Zaire/USS): 400 -Percent non-paying: 5% E -Household income/month: 20000 *Superv.exp.(% rev.HC): 10% *Hsd.inc.insured/non-ins 1.00 TABLE 6.2 ALL INFORMATION IS MONTHLY ---------New cases------ ------Total visits------ % mar-i Non-insr Insured Total Non-insrd. Insrd. Total ket HC B H -------- ------- ------- -------- ------- ---- ------ A E ICur.Ca 505 101 606 1285 437 1722 68%1 S A IDeLiv. 23 3 26 --- --- --- 78% I L P-N Ca 13 2 15 124 16 140 72%1 C T P-S Ca 14 2 16 237 30 267 72% H Chr.Ca 0 0 0 27 4 31 67% 0 i _ _ _ _ 000s Zaires I U C (1) REVENUE Services 229 HC utiliz.rates T E i Drugs 300 528 --------------- PN 1(2) DEPENSES Persont. 205 i Non-insr Insrd U T Drugs 422 -------- ------ IT E Other 32 660 1 Cur.Ca 17% 30%j R (3) PROFIT [(1)-(2)] (131): DeLiv. 64% 82% T I(4) SUPERVISION EXPENS. 23 1 P-N Ca 37% 43%1 A (5) DEPRECIATION 90 1 P-S Ca 41% 47% B (6) PROFIT[(3)-(4)-(5)] (244)1 Chr.Ca 36% 44% L E 76 A Supply-Demand Model of Health Care Financing Exhibit G.4. A Devaluation of the Country's Currency Coupled with a Curative Care Price Increase Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Km 3-5 Km > 5 Km Total B H ------- ------- ------- ------- ------- -------- |Cost of | A E |Percent 35.0% 25.0% 20.0% 15.0% 5.0% 100%| drugs, S A INumber 3500 2500 2000 1500 500 10000 Lab.exams| I L First Repeat Total |& other | C T visit visits Drugs expend.jphar.prdl H PRICE ------- ------- -------- ------ -------- I Curative care 360 0 558 918 638 I N C DeLiveries 200 --- 819 1019 936 P E Pre-nataL Care 0 --- 150 150 209 U N IPre-SchooL Care 0 --- 150 150 504 T T Chronic Care 0 --- 200 200 I 521 E i IT R Radius health area(Km): 10 -Ins.premium/yr.(Zaires) 1500 A Population: 10000 -Percentage insured: 10% B Currency: Zaire -Copayment insured: 20% L *Exchge.rate(Zaire/USS): 400 -Percent non-paying: 5% E Household income/month: 20000 -Superv.exp.(% rev.HC): 10% *Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY ---------New cases------ ------Total visits------ % mar-' Non-insr Insured TotaL Non-insrd. Insrd. Total ket HCI B H -------- ------- ------- -------- ------- ---- ------ | A E ICur.Ca 417 98 515 1036 421 1457 61%: S A IDeliv. 23 3 26 --- --- --- 78% I L P-N Ca 13 2 15 124 16 140 72%: C T 'P-S Ca 14 2 16 237 30 267 72%1 H IChr.Ca 0 0 0 27 4 31 67% 0 i _ _ _ 000s Zaires 1 U C (1) REVENUE Services 278 HC utiliz.ratesi T E Drugs 253 531 ---------------| P N 1(2) DEPENSES Persont. 205 | Non-insr Insrd| U T Drugs 364 | -------- ------| T E Other 37 607 1 Cur.Ca 14% 29%| R 1(3) PROFIT [(1)-(2)] (76)1 DeLiv. 64% 82% T I(4) SUPERVISION EXPENS. 28 | P-N Ca 37% 43% A 1(5) DEPRECIATION 90 1 P-S Ca 41% 47% B :(6) PROFIT[(3)-(4)-(5)] (193)1 Chr.Ca 36% 44% L |E Appendix G 77 Exhibit G.5. A Change in Population Distribution Table B.1 Population distribution 0-1 Km 1-2 Km 2-3 Kin 3-5 Km > 5 Km Total B H ------- ------- ------- ------- ------- --------ACost of AE Percent 75.0% 25.0% 0.0% 0.0% 0.0% 100%1 drugs, jS A INuiber 7500 2500 0 0 0 10000 Lab.examsl I L First Repeat Total I& other | C T visit visits Drugs expend.|phar.prdl H PRICE ------- ------- -------- ------ -------- lCurative care 200 0 558 758 | 558 N C 'Deliveries 200 --- 819 1019 | 819 1 P E Pre-nataL Care 0 --- 150 150 183 1 U N jPre-SchooL Care 0 --- 150 150 1 441 1 T T lChronic Care 0 --- 200 200 | 456 I E i I T R *Radius health area(Kan): 10 *Ins.premium/yr.(Zaires) 1500 A *Population: 10000 *Percentage insured: 10% B *Currency: Zaire *Copayment insured: 20% L *Exchge.rate(Zaire/USS): 350 *Percent non-paying: 5% E *Household income/month: 20000 *Superv.exp.(% rev.HC): 10% *Hsd.inc.insured/non-ins 1.00 TABLE B.2 ALL INFORMATION IS MONTHLY ---------New cases------ ------Total visits------ % mar-' Non-insr Insured Total Non-insrd. Insrd. Total ket HCI B H -------- ------- ------- -------- ------- ---- ------ AE |Cur.Ca 681 129 810 1877 668 2544 72%| S A 'DeLiv. 26 3 29 --- --- --- 82%j I L IP-N Ca 16 2 18 185 23 208 76%| C T jP-S Ca 18 2 20 347 44 390 76%| H 'Chr.Ca 1 0 1 34 4 38 72%| 0 000s Zaires U C (1) REVENUE Services 264 HC utiliz.rates! T E Drugs 399 662 | --------------- P N (2) DEPENSES PersonL. 255 Non-insr Insrd U T Drugs 489 |-------- ------ T E Other 34 778 | Cur.Ca 23% 39%j R (3) PROFIT [(1)-(2)] (116)1 DeLiv. 73% 88%: T (4) SUPERVISION EXPENS. 26 1 P-N Ca 47% 53%j A (5) DEPRECIATION 79 P-S Ca 50% 56%| B (6) PROFITE(3)-(4)-(5)] (221) Chr.Ca 45% 54%| L I__IE BIBLIOGRAPHY Atkinson, A., and J. Stiglitz. 1980. Lectures on Public Economics. New York: McGraw-Hill. Bitran, R. 1988. Health Care Demand Studies in Developing Countries: A Critical Review and Agenda for Research. Arlington, Va: Resources for Child Health Project. .. 1990. A Household Health Care Demand Study in the Bokoro and Kisantu Zones of Zaire, vol. 3, Determinants of Health Care Demand. Arlington, Va.: Resources for Child Health Project. Block, S., D. Donaldson, and S. Foster. 1988. "Cost Recovery for Recurrent Costs of Public Health Care in Guinea-Bissau." Cambridge, Mass.: Abt Associates. Dor, A., P. Gertler, and J. van der Gaag. 1987. "Non-Price Rationing and the Choice of Medical Care Providers in Rural Cote d'Ivoire." Journal of Health Economics 6: 291-304. Gertler, P., L. Locay, and W. Sanderson. 1987. "Are User Fees Regressive? The Welfare Implications of Health Care Financing Proposals in Peru." Journal of Econometrics 36: 67- 88. Makinen, M., and S. Block. 1986. Pricing for Cost Recovery in Primary Health Care in Guinea. CCCD/Guinea: Resources for Child Health Project. McFadden, D. 1981. "Econometric Models of Probabilistic Choice." In C. Manski and D. McFadden, eds., Structural Analysis of Discrete Data with Econometric Applications. Cambridge, Mass.: MIT Press. Mwabu, G. 1984. "A Model of Household Choice Among Medical Treatment Alternatives in Rural Kenya." Unpublished Ph.D. diss., Department of Economics, Boston University. Rosenthal, G., A. Arrazola, G. M. Crowley, C. J. Cuellar, and A. Solari. 1988. "Toward Self- Financing of Primary Health Services: A Market Study of PROSALUD in Santa Cruz, Bolivia." HCF/LAC Research Report No. 6, Stony Brook, New York. Small, K., and H. Rosen. 1981. "Applied Welfare Economics with Discrete Choice Models." Econometrica 1: 105-30. 78 Distributors of World Bank Publications ARGENTINA The Middle East Observer ITALY PORTUGAL Carlos Hirsch, SRL 41, Sherif Street Licosa Commissionaria Sansoni SPA Livraria Portugal Caleria Cuemes Cairo Via Duca Di Calabria, 1/1 Rua Do Carmo 70-74 Florida 165 4th Floor-Ofc. 453/465 Casella Postale 552 1200 Lisbon 1333 Buenos Aires FINLAND 50125 Firenze Akateeminen Kirjakauppa SAUDI ARABIA, QATAR AUSTRALIA, PAPUA NEW GUINEA. P.O. Box 128 JAPAN Jarir Book Store FIJI, SOLOMON ISLANDS, SF-00101 Helsinki 10 Eastern Book Service P.O. Box 3196 VANUATU, AND WESTERN SAMOA Hongo 3-Chome, Bunkyo-ku 113 Riyadh 11471 D.A. Information Services FRANCE Tokyo 648 Whitehorse Road World Bank Publications SINGAPORE, TAIWAN, Mitcham3132 66, avenue d'Idna KENYA MYANMAR,BRUNEI Victoria 75116 Paris Africa Book Service (E.A.) Ltd. Gower Asia Pacific Pte Ltd. Quaran House, Mfangano Street Golden Wheel Building AUSTRIA GERMANY P.O. Box 45245 41, Kallang Pudding, #04-03 Gerold and Co. UNO-Verlag Nairobi Singapore 1334 Graben 31 Poppelsdorfer Allee 55 A-1011 Wien 53115 Bonn KOREA, REPUBLIC OF SOUTH AFRICA, BOTSWANA Pan Korea Book Corporation For single titles: BANGLADESH HONG KONG, MACAO P.O. Box 101, Kwangwhamun Oxford University Press Micro Industries Development Asia 2000 Ltd. Seoul Southern Africa Assistance Society (MIDAS) 46-48 Wyndham Street P.O. Box 1141 House 5, Road 16 Winning Centre Korean Stock Book Centre Cape Town 8000 Dhanmondi R/Area 7th Floor P.O. Box 34 Dhaka 1209 Central Hong Kong Yeoeido For suscription orders: Seoul International Subscription Service Branch offices: HUNGARY P.O. Box 41095 Pine View, 1st Floor Foundation for Market Economy MALAYSIA Craighall 100 Agrabad Commercial Area Dombovari Ut 17-19 University of Malaya Cooperative Johannesburg 2024 Chittagong 4100 H-1117 Budapest Bookshop, Limited P.O. Box 11.27, Jalan Pantai Baru SPAIN BELGIUM INDIA 59700 Kuala Lumpur Mundi-Prensa Libros, S.A. Jean De Lannoy Allied Publishers Private Ltd. Castello37 Av. du Rol 202 751 Mount Road MEXICO 28001 Madrid 1060 Brussels Madras - 600 002 INFOTEC Apartado Postal 22-860 Librerla Internacional AEDOS CANADA Branch offices: 14060 Talpan, Mexico D.F. Consell de Cent. 391 Le Diffuseur 15 J.N. Heredia Marg 08009 Barcelona ISIA Boul. de Mortagne Ballard Estate NETHERLANDS Boucherville, QuAbec Bombay - 400 038 De Lindeboom/InOr-Publikaties SRI LANKA AND THE MALDIVES J4B 5E6 P.O. Box 202 Lake House Bookshop 13/14 Asaf Ali Road 7480 AE Haaksbergen P.O. Box 244 Renouf Publishing Co. New Delhi - 110 002 100, Sir Chittampalam A. 1294 Algoma Road NEW ZEALAND Gardiner Mawatha Ottawa, Ontario 17 Chittaranjan Avenue EBSCO NZ Ltd. Colombo 2 KIB 3W8 Calcutta -700 072 Private Mail Bag 99914 New Market SWEDEN CHILE Jayadeva Hostel Building Auckland For single titles: Invertec ICT S.A. 5th Main Road, Candhinagar Fritzes Fackboksforetaget Av. Santa Maria 6400 Bangalore - 560 009 NIGERIA Regeringsgatan 12, Box 16356 Edificio INTEC, Of. 201 University Press Limited S-103 27 Stockholm Santiago 3-5-1129 Kachiguda Three Crowns Building Jericho Cross Road Private Mail Bag 5095 For subscription orders: CHINA Hyderabad - 500 027 Ibadan Wennergren-WilliamsAB China Financial & Economic P. O. Box 130S Publishing House Prarthana Flats, 2nd Floor NORWAY S-171 25 Solna 8, Da Fo Si Dong Jie Near Thakore Baug, Navrangpura Narvesen Information Center Beijing Ahmedabad - 380 009 Book Department SWITZERLAND P.O. Box 6125 Etterstad For single titles: COLOMBIA Patiala House N-0602 Oslo 6 Librairie Payot Infoenlace Ltda. 16-A Ashok Marg Case postale 3212 ApartadoAereo34270 Lucknow -226001 PAKISTAN CH 1002 Lausanne Bogota D.E- Mirza Book Agency Central Bazaar Road 65, Shahrah-e-Quaid-e-Azam For sulcrtion orders: COTE D'IVOIRE 60 Bajaj Nagar P.O. Box No. 729 Librairie Payot Centre d'Edition et de Diffusion Nagpur 440 010 Lahore 54000 Service des Abonnements Africaines (CEDA) Case postale 3312 04 B.P. 541 INDONESIA PERU CH 1002 Lausanne Abidjan 04 Plateau Pt. Indira Limited Editorial Desarrollo SA Jalan Borobudur20 Apartado3824 THAILAND CYPRUS P.O. Box 181 Lima I Central Department Store Center of Applied Research Jakarta 10320 306 Silom Road Cyprus College PHILIPPINES Bangkok 6, Diogenes Street, Engomi IRAN International Book Center P.O. Box 2006 Kowkab Publishers Suite 1703, Cityland 10 TRINIDAD & TOBAGO, ANTIGUA Nicosia P.O. Box 19575-511 Condominium Tower 1 BARBUDA, BARBADOS, Tehran Ayala Avenue, H.V. dela DOMINICA, GRENADA, GUYANA, DENMARK Costa Extension JAMAICA, MONTSERRAT, ST. SamfundsLitteratur IRELAND Makati, Metro Manila KITTS & NEVIS, ST. LUCIA, Rosenoerns Allk 11 Government Supplies Agency ST. VINCENT & GRENADINES DK-1970 Frederiksberg C 4-5 Harcourt Road POLAND Systematics Studies Unit Dublin 2 International Publishing Service #9 Watts Street DOMINICAN REPUBLIC Ul. Piekna 31/37 Curepe Editora Taller, C. por A- ISRAEL 00-677 Warzawa Trinidad, West Indies Restauraci6n e Isabel la Cat6lica 309 Yozmot Literature Ltd. Apartado de Correos 2190 Z-1 P.O. Box 56055 For subscrtin orders: UNITED KINGDOM Santo Domingo Tel Aviv 61560 IPS Journals Microinfo Ltd. Ul. Okrezna 3 P.O. Box 3 EGYPT, ARAB REPUBLIC OF 02-916 Warszawa Alton, Hampshire GU34 2PC Al Ahram England Al Calaa Street Cairo u F w g The World Bank Headquarters European Office fokyo Office 1818 H Street, N.W. 66, avenue d'lena Koku-,ai BuliI, Washington, D.C. 20433, U.S.A. 75116 Paris, France 1-1, Marunoud i -Chome Chivda-ku, T,)k% I0, Telephone: (202) 477-1234 Telephone: (1) 40. b9. 3). l lapan Facsimile: (202) 477-6391 Facsimile: (1) 40. t)0. 30. hi Telex: WUI 64145 WORLDBANK Telex: 640651 Telephone: 3) 3214-5001 RCA 248423 WORLDBK Facsinile: (3) 3214-3t57 Cable Address: INTBAFRAD Telex: 26838 WASHINCTONDC SUPPLY & DEMAND MODEL HE 400000011967 - $23 .95