IMPROVING EQUITABLE ACCESS TO CATARACT SURGERY IN RURAL SOUTHERN CHINA USING WILLINGNESS TO PAY

IMPROVING EQUITABLE ACCESS TO CATARACT SURGERY IN RURAL SOUTHERN CHINA: USING WILLINGNESS TO PAY DATA TO ASSESS THE FEASIBILITY OF A TIERED PRICING MODEL TO SUBSIDIZE SURGERIES TO THE PO

IMPROVING EQUITABLE ACCESS TO CATARACT SURGERY IN RURAL SOUTHERN CHINA: USING WILLINGNESS TO PAY DATA TO ASSESS THE FEASIBILITY OF A TIERED PRICING MODEL TO SUBSIDIZE SURGERIES TO

THE POOREST

by Elaine M Baruwa

A dissertation submitted to Johns Hopkins University in conformity with the

requirements for the degree of Doctor of Philosophy

Baltimore, Maryland June 2007

© Elaine M Baruwa 2007 All Rights Reserved

UMI Number: 3288601

Copyright 2007 by Baruwa, Elaine M

All rights reserved

INFORMATION TO USERS

The quality of this reproduction is dependent upon the quality of the copy submitted Broken or indistinct print, colored or poor quality illustrations and photographs, print bleed-through, substandard margins, and improper alignment can adversely affect reproduction

In the unlikely event that the author did not send a complete manuscript and there are missing pages, these will be noted Also, if unauthorized copyright material had to be removed, a note will indicate the deletion

®

UMI

UMI Microform 3288601 Copyright 2008 by ProQuest Information and Learning Company All rights reserved This microform edition is protected against unauthorized copying under Title 17, United States Code

ProQuest Information and Learning Company

300 North Zeeb Road P.O Box 1346

Abstract

Title: Improving Equitable Access to Cataract Surgery in Rural Southern China: Using Willingness to Pay Data to Assess the Feasibility of a Tiered Pricing Model to Subsidize Surgeries to the Poorest

Aim: To assess the equity of financial access to cataract surgery given willingness to pay (WTP) for cataract surgery at the current price of surgery and for added amenities such as surgery by a senior surgeon, an improved intraocular lens, transport and food To determine the feasibility of a tiered pricing and cross-subsidization model using these estimates

Methods: A WTP survey was administered at community screenings and hopsital cataract surgery clinics in rural Guangzhou WTP was estimated using interval regression and then compared to the price of surgery to determine access A further equity analysis was

conducted using concentration indices and curves The WTP for amenities was similarly analyzed to determine potential demand

Results: WTP surveys were conducted with 656 patients and 342 of their caregivers The mean WTP for the community screening patients was 371RMB (S.D 114RMB) and

570RMB (S.D 69RMB) for the hospital patients (8RMB =US$1) For caregivers the mean was 619 RMB (S.D 77 RMB) At the two prices charged by HKI, 500RMB and 630RMB, the estimated concentration indices were 0.18 and 0.36 for patients, which implies that financial access is inequitably concentrated amongst the wealthier patients However, the respective index measures were 0.01 and 0.10, for caregivers indicating lower inequity at 630RMB and no inequity at 500RMB The WTP for amenities was low, only 78RMB for a

senior surgeon and 42RMB for an improved IOL

Conclusion: Access to cataract surgery is inequitably distributed between the poor and the poorest in this population even at cost, 500RMB We determined that not enough patients would be able to purchase surgery at higher, tiered prices for additional amenities in order to subsidize any significant number of surgeries at a lower price While WTP for cataract surgery was significantly higher when assessed by patient's caregivers, adjusting for this did not change the finding that access is inequitable for this population and creative ways must be found to lower prices

Thesis Committee:

Kevin Frick, PhD, Department of Health Policy and Management, JHSPH

David Bishai, MD PhD, Department of Population and Family Health, JHSPH

Emily West Gower, PhD, Department of Ophthalmology, JHMI

Damian Walker, PhD, Department of International Health, JHSPH

Laura Morlock, PhD, Department of Health Policy and Management, JHSPH

ACKNOWLEDGEMENTS

I would like to thank:

The Department of International Health, JHSPH for the excellent teaching and support that they gave me during my doctoral studies In particular, Carol Buckley for her all her help, making sure that I never got lost administratively

My colleagues at the PneumoADIP for their encouragement and my director Angeline Nanni, for understanding my priorities and accomodating them with such empathy

My fellow doctoral students were an invaluable source of encouragement and friendship, particularly during both of my pregnancies Arantxa Colchera, Nhan Tran, Marjorie Opuni, Rebekah Heinzen and Tram Lam studied with me, baby-sat for me, pondered the pros/cons

of doctoral studies (mostly the cons), and attended my defense

My family: Chiadi, Ketandu and Omenka for being so patient with a wife and mother who seemed to always have too much to do at the same time I love them so much Their smiles and laughter kept me going on the rare occasions when I did feel as though 24 hours

in a day and a single brain were not quite enough to get throught this journey

Finally my advisor, Kevin Frick, who is extremely bright, seems to have 36 hours in work day and possesses a bizarre affection for econometrics but his patience, his work ethic and his generosity have been inspirational to me I aspire to be the type of mentor, teacher

and friend that he has been to me and consider myself truly blessed to shared this experience

with him

TABLE OF CONTENTS

1 STUDY AIM AND OBJECTIVES 1

1.1 OBJECTIVE 1 3

1.2 OBJECTIVE 2 3 1.3 OBJECTIVE 3 4

2 BACKGROUND 5

2.1 EPIDEMIOLOGY OF CATARACT AND CATARACT SURGERY 5

2.2 RURAL HEALTH CARE IN CHINA 6

2.3 HKI, CHINA AND TIERED PRICING 10

2.4 THE ARAVIND EYE HOSPITAL, INDIA 11

2.5 HKI, CHINA AND CATARACT SURGERY 13

3 CONCEPTUAL FRAMEWORK 14

3.1 DEFINING EQUITY IN TERMS OF WILLINGNESS TO PAY 14

3.2 SOCIAL WELFARE AND THE EQUITY-EFFICIENCY TRADE-OFF 16

3.3 THE ECONOMICS OF TIERED PRICING 18

4 CONTINGENT VALUATION AND WILLINGNESS TO PAY 20

5 THE USE OF WTP IN DEVELOPING COUNTRY RESEARCH 31

5.1 WTP FOR INSECTICIDE TREATED BEDNETS IN EASTERN NIGERIA 31

5.2 WTP FOR COMMUNITY-BASED INSURANCE IN BURKINA FASO 32

5.3 WTP FOR CATARACT SURGERY IN NEPAL 33

5.4 WTP FOR CATARACT SURGERY IN TANZANIA 33

5.5 FINDINGS AND IMPLICATIONS 34

5.6 BEST PRACTICE FOR WTP SURVEY ADMINISTRATION 37

6 DATA COLLECTION 42

6.1 SAMPLING FRAMEWORK 42

6.2 SAMPLE SIZE 43 6.3 SURVEY DESIGN 44 6.4 SURVEY ADMINISTRATION 50

7 STATISTICAL METHODS 53

7.1 CATEGORICAL OUTCOMES - INTERVAL REGRESSION 53

7.2 CONCENTRATION CURVE AND INDEX ESTIMATION 58

8 RESULTS 63

8.1 SAMPLE SIZE AND RESPONSE RATE 63

8.2 SAMPLE CHARACTERISTICS 65

8.3 BIVARIATE ASSOCIATIONS WITH WTP ANYTHING FOR CATARACT SURGERY 75

8.4 MAXIMUM WILLINGNESS TO PAY FOR CATARACT SURGERY 79

8.5 OBJECTIVE 1 PATIENTS WILLINGNESS TO PAY 86

8.6 OBJECTIVE 2 CAREGIVERS WILLINGNESS TO PAY 99

8.7 EQUITY OF ACCESS USING CAREGIVER'S WTP 104

8.8 OBJECTIVE 3 WILLINGNESS TO PAY FOR AMENITIES 105

9 DISCUSSION 109

9.1 FACTORS AFFECTING PATIENT'S WTP 109

9.2 PATIENT'S WILLINGNESS TO PAY 113

9.3 FACTORS AFFECTING CAREGIVER'S WTP 115

9.4 HOUSEHOLD CHARACTERISTICS' IMPACT ON WTP ON PAIRED RESPONDENTS 116

9.5 CAREGIVERS' PREDICTED WILLINGNESS TO PAY 118

9.6 EQUITY OF ACCESS 122

9.7 POLICY IMPLICATIONS FOR HKI 124

9.9 WAS THE METHODOLOGY APPROPRIATE FOR OUR OBJECTIVES? 128

9.10 WAS THE METHODOLOGY APPROPRIATE FOR THIS POPULATION? 128

9.11 BEST PRACTICE FN PRACTICE 133

9.12 STUDY LIMITATIONS 137 9.13 CONCLUSION 140

10 APPENDICES 151

10.1 WTP SURVEY FOR PATIENTS 151

10.2 WTP SURVEY FOR CAREGIVERS 171

11 CURRICULUM VITAE- ELAINE MONISOLA BARUWA 190

TABLE OF TABLES

TABLE 1 NEW COMMUNITY MEDICAL SCHEME - PREMIUMS, CO-PAYMENTS AND DEDUCTIBLES 9

TABLE 2 SURVEY STRUCTURE 45

TABLE 3 SAMPLE SIZE BY SITE AND TYPE 63

TABLE 4 SAMPLE SOCIODEMOGRAPHICS 66

TABLE 5 WORK STATUS AND CARE REQUIREMENTS 67

TABLE 6 VISUAL ACUITY CLASSIFICATION 69

TABLE 7 SAMPLE VISUAL ACUITY 69

TABLE 8 SAMPLE HOUSEHOLD INCOME 71

TABLE 9 REASONS FOR NOT WANTING TO PAY FOR SURGERY 72

TABLE 10 FIRST PAYMENT CARD AS A DETERMINANT OF MAXIMUM WTP 74

TABLE 11 BIVARIATE ASSOCIATIONS WITH WILLINGNESS TO PAY ANYTHING FOR CATARACT

TABLE 12 MAXIMUM WTP ANYTHING FOR CATARACT SURGERY, (N=656) 79

TABLE 13 NUMBER OF RESPONDENTS AND THEIR MAXIMUM EXPRESSED WTP BY PAYMENT CARD

AND BY SITE 84 TABLE 14 CHECKING THE CONSISTENCY OF IMPUTED INCOME VARIABLES 86

TABLE 15 MAXIMUM WTP - FINAL PATIENT MULTIVARIATE MODEL (N=656) 89

TABLE 16 PREDICTED WTP AND ACCESS FOR PATIENTS 92

TABLE 17 CONCENTRATION INDICES FOR PATIENTS 96

TABLE 18 CAREGIVER MAXIMUM WTP MODEL 101

TABLE 19 PREDICTED MAXIMUM WTP FOR PAIRS 103

TABLE 20 CONCENTRATION INDEX FOR CAREGIVERS 105

TABLE 21 WILLINGNESS TO PAY FOR AMENITIES 106

TABLE 22 MAXIMUM WTP FOR A SENIOR SURGEON FROM PATIENTS 107

TABLE 23 PREDICTED WILLINGNESS TO PAY FOR AMENITIES 107

TABLE 24 HOUSEHOLD SIZE AND NUMBER OF CHILDREN 117 TABLE 27 CONCENTRATION CURVE FOR 250RMB SURGERY FOR COM SCREENING PATIENTS 125

TABLE OF FIGURES

FIGURE 1 CONCENTRATION CURVE - EQUITABLE ACCESS 59

FIGURE 2 FIRST PAYMENT CARD ASKED BY INTERVIEWER 74

FIGURE 3 UNADJUSTED MAXIMUM WTP BY SITE 85

FIGURE 4 PREDICTED WTP BY SITE 94 FIGURE 5 ACCESS TO CATARACT SURGERY AT 500RMB SHOWING THE % IN NUMBERS OF

RESPONDENTS BY SITE 95 FIGURE 6 ACCESS TO CATARACT SURGERY BY INCOME AT 630RMB SHOWING THE % IN NUMBERS

OF RESPONDENTS BY SITE 96 FIGURE 7 CONCENTRATION CURVES FOR PATIENTS AT 500RMB BY SITE 98

FIGURE 8 CONCENTRATION CURVES FOR PATIENTS A T 6 3 0 R M B BY SITE 99

FIGURE 9 COMPARISON OF UNADJUSTED MAXIMUM WTP RESPONSES BY PAIRS 100

FIGURE 10 ACCESS TO CATARACT SURGERY BY CAREGIVERS 104

FIGURE 11 CONCENTRATION CURVE FOR PATIENTS AND CAREGIVERS AT 500RMB 105

FIGURE 12 SAMPLE GENDER (A) AND ACCESS BY GENDER (B) 111

FIGURE 13 SAMPLE EDUCATION AND ACCESS BY EDUCATION 112

FIGURE 14 WTP FOR CATARACT SURGERY FOR MATCHED PAIRS (CHILD IS PAYMENT SOURCE) 120

FIGURE 16 WTP RESPONSES BY SITE 131

FIGURE 17 PREDICTED MAXIMUM WTP AS A % OF ANNUAL HH INCOME 132

1 Study Aim and Objectives

This study was designed to evaluate whether access to cataract surgery is equitable in the Guangdong Province of the People's Republic of China (PRC) and to explore the feasibility

of using a tiered pricing model to increase uptake by the poorest, using data from a

willingness to pay survey administered to a rural population in this region

China and Cataract

Cataract is the leading cause of blindness in the PRC in people aged 50 and over

Prevalence rates of cataract blindness have been estimated to be up to 4.37%, with rates of low vision being even higher in this age group (Hsu, Cheng, Liu, Tsai, & Chou, 2004; Li, Xu,

He, Wu, Munoz, & Ellwein, 1999a) Combined with a low cataract surgery rate of 230 per million per year the result is that China has a severe burden of curable blindness and low vision (Apple, Ram, Foster, & Peng, 2000)

Helen Keller International, China

Helen Keller International (HK.I), in conjunction with the privately owned Guangming Eye Hospital (GEH) and the Yang Jiang local government health department set up a cataract screening and surgery program in 2001 The program now provides about 1800 surgeries a year which translates roughly to a rate of at least 720 per million if we do not include the surgeries performed by other providers A cross sectional willingness to pay study conducted three months after the program began, suggested that income would be a limiting factor for

access to cataract surgery even with the service priced at cost (He M et al., 2007) Now the

program would like to determine whether or not it is feasible to use a tiered pricing structure

to increase its revenues in order for it to provide cataract surgery at a lower price to those

unable to pay the current fee of 500 - 630 Renminbi (RMB) where 1 US$=8RMB

Access: Inequality and Inequity

In the 2001 study it was found that there were significant differences in the amount that respondents were willing to pay across income groups, specifically, those in higher income groups were willing to pay higher amounts This finding highlights an inequality in

willingness to pay that is not necessarily inequitable - there is nothing 'unfair' about

individuals with a higher income being willing to spend more than individuals with lower income However it was also found that only 37% of the respondents would be willing to pay 500RMB or more to obtain cataract surgery This result suggests that even though the service

is now available to this population, there may remain access limitations for some individuals due to the pricing and this outcome is inequitable The combination of these findings suggests that, with enough income variation, it might be possible to induce those with higher incomes who may be willing to pay more for surgery, to actually do so and then to use the increased revenue to subsidize a lower price that improves access for those with lower incomes and willingness to pay In other words we could take advantage of an existing income inequality and provide somewhat unequal services to reduce an access inequity for the most basic level

of service

This study will utilize data from a willingness to pay survey to obtain a valuation of

cataract surgery by respondents and their caregivers which, when combined with the known prices, will determine whether or not access is equitable It will then determine whether or

not a large enough number of respondents value additional amenities highly enough to enable

higher pricing Such amenities could include having a senior surgeon perform their surgery, having an improved intra-ocular lens implanted or having food and transport provided for

them With estimates of revenue, a range of possible subsidized prices can be determined and used to predict the impact of the model on the equity of access to cataract surgery

1.1 Objective 1

To determine whether access to cataract surgery is equitable in this population using willingness to pay survey data from respondents with cataract

Empirical Analysis: The results from a survey administered to respondents who are

cataract blind in at least one eye will be used to explore how willingness to pay for cataract surgery may differ by demographic and socioeconomic characteristics, vision status and potential sources of payment Following this exploration, an appropriate model to estimate willingness to pay will be proposed and tested From these results an 'incidence rate' for cataract surgery at current pricing levels will be determined and combined with the income data to construct a concentration index that describes the equity of access

1.2 Objective 2

To determine whether willingness to pay differs between respondents with cataract and their households/caregivers and what impact this has upon the willingness to pay estimates

Empirical Analysis: The results from a survey administered to the caregivers who

accompany respondents will be used to explore how willingness to pay for cataract surgery may differ between patients and another member of their household and to determine how much care the patient needs because of their impaired vision Specifically this comparison will be used:

l) To determine why there may be differences in WTP from respondents who come from

the same household and are subject to the same income constraint It could be important if sources other than own savings and insurance are used to pay for surgery

2) To determine if there are intergenerational differences in the perceived need for surgery

3) To determine whether there are differences in perceived control of household resources

4) To determine whether a societal valuation of cataract surgery might be significantly higher or lower than the patients' valuation of cataract surgery

1.3 Objective 3

To estimate the revenue that can be expected from a tiered pricing model and

to determine the potential of the model to improve equity of access to surgery

Empirical Analysis: The willingness to pay data will be used to assess the potential

demand for the additional amenities that GEH/HKI could provide at minimal cost These amenities are having a senior surgeon perform the surgery, an improved intraocular lens, transport to/from the clinic and the provision of meals Subsequently, the projected revenue from the provision of such services will be estimated and used to determine the feasibility of a tiered pricing and cross subsidization model To avoid having to determine patients' choices between amenities, revenues will be determined from the provision of a single amenity at a time Finally a concentration index will be estimated at each feasible subsidized price to see what impact this model may have on the equity of access

2 Background

2.1 Epidemiology of Cataract and Cataract Surgery

Cataract is the opacification of the lens and its major risk factor is aging Other risk factors postulated include diabetes, smoking, alcohol and UVB exposure (Cataract: Epidemiology and service delivery.2000) Cataract can occur unilaterally but is more often bilateral, with the cataract in each eye likely to develop and worsen vision at differing rates Surgical removal of the lens is the only treatment There are different methods for cataract extraction and costs are very dependent upon which method is used and whether the lens is replaced or the patient is given aphakic spectacles to improve their post surgery vision Global estimates

of visual impairment by the WHO are that I6l million people were blind or had low vision in

2002 and 90% of these people live in developing countries Despite being treatable, cataract

is the leading cause of visual impairment by far and causes 48% of global blindness compared

to the next largest cause glaucoma which causes 12%( World Health Organization, November 2004) Blindness is defined as having visual acuity of less than 0.05 in the better eye with best possible correction That is, being able to see at 3 meters what a person with normal vision can see at 60 meters or 3/60 in Snellen Visual Acuity in meters Low vision is defined

as visual acuity of less than 0.33 (20/60), but equal to or better than 3/60' In terms of

' By the 10th Revision of the WHO International Statistical Classification of Disease, Injuries and Causes of Death

functional vision , consider that in the state of Maryland a driver can obtain an unrestricted license with a minimum visual acuity of 10/20 or 0.5 in either eye and the legal definition of blindness in the United States is 20/200

Cataract is the leading cause of blindness in people over the age of 50 in China where there are approximately 90 million people over the age of 60 (Li, Xu, He, Wu, Munoz, & Ellwein, 1999a; Zhang et al., 1992; Zhang, Zou, Gao, Di, & Wang, 1992) The estimated prevalence

of blindness in China is estimated to be between 2.94% and 4.37% and estimates of low vision are even higher (Hsu et al., 2004; Li, Xu, He, Wu, Munoz, & Ellwein, 1999a) The cataract surgery rate (CSR) varies across China and was estimated to be 138 per million per year in Guangdong circa 1992 It was as low as 28 in Hebei and as high as 1500 in Xizang The current average of 230-320 across China is very low when compared to India where the CSR is 3650 despite having a cataract surgeon to total population ratio that is comparable to China's (Foster, 2001; Li, Xu, He, Wu, Munoz, & Ellwein, 1999a; Zhao, Sui, Jia, Fletcher, & Ellwein, 1998)

2.2 Rural Health Care in China

Yang Jiang is a county of the Guangdong Province comprised of 1 urban city and 3 rural sub-counties, located on the South East coast of China The Guangdong Province is one of the wealthiest in China by virtue of its economically beneficial coastline, which has given rise

Functional vision will be described in more detail in the Results section in order to

illustrate the level of visual impairment in the population sample

to massive swathes of industrialization However, there is severe inequality between the urban and rural economies Income estimates for all the Central Provinces, including

Guangdong are 7,900 RMB annual per capita in urban areas and 2,652 RMB annual per capita

in rural areas (Sicular T., Yue X., Gustafsson B., & Li S., 2006) Rural Guangdong is as poor

as rural regions across China

The Old Rural Cooperative Medical System

Until the late 1970s about 90% of China's rural villages were well served by the Rural Cooperative Medical System (RCMS) that was a pre-payment plan financed by household premiums, village level collective welfare funds and a small amount of higher level

government funding This system, famous for its 'barefoot doctors,' is widely acknowledged

as being a tremendous success It is credited with making a major contribution to China's first 'health care revolution' in which life expectancy increased from 38 in 1949 to 68 in 1978 With the move from central planning toward a market economy, the commune system moved

to a 'household responsibility' system With the removal of the risk-sharing benefit of the collective welfare fund, the majority of RCMS funds collapsed and by 1998 only 9.5% of the rural population was insured The government maintained its level of very limited investment

in rural health care services, which combined with a high level of financial decentralization, left the rural population with severely limited access to care Local county, township and village health posts are available, but they have little impact on access to healthcare for the poor because they receive such limited government subsidization Yang Jiang General

Hospital for example, has a turnover of 170 million RMB but receives only 300,000 RMB in

subsidy from the government3 In order to ensure access to basic care and equity, government 'mandates' prices with little regard to the feasibility of maintaining services Consequently there is little incentive to provide basic services The need to remain financially stable

encourages the practice of lucrative drug over-prescription and unnecessary/avoidable

expensive procedures At the village level there is no government subsidy at all and so the providers of basic services are essentially private providers of care, working on a fee for service basis

The New RCMS

The scheme that replaced the RCMS proved to be largely unsustainable due to the lack of central government financial participation It was modified into the New Community Medical System (NCMS) that now incorporates a matched financing model to encourage enrollment and increase sustainability Specifically, enrollees pay 10 RMB per capita per year, which is matched by 20 RMB by the central government and at least 20 RMB by the local government (40 RMB in the wealthier provinces) The matching is entirely reliant on the enrollee's contribution Uptake has been estimated at 70% in the pilot schemes that cover about 65 million of China's 450 million rural dwellers (Wagstaff A, Lindelow M, Jun G, Ling X, & Juncheng Q, 2007) The program is set to begin expansion to cover 80% of the country's rural population in 2007 When paid, the premium entitles the rural enrollee to discounts at health service institutions, both private and public, in the form of co-pays for services, with a maximum of 3000 RMB per year beyond which fees are out of pocket

Personal communication, Dr He, Director, Guangming Eye Hospital

Table 1 shows how the NCMS operates in rural Yang Jiang and how the urban health insurance works for those seeking cataract surgery (there are no 'rural' reimbursable providers

of cataract surgery)

Table 1 New Community Medical Scheme - Premiums, Co-payments and Deductibles

Annual

Premium per member per household

(maximum Annual Coverage 3000 RMB)

Co pay at a local clinic (n/a)

Co-pay at a local hospital

Co-pay at a private hospital

Deductible at a public facility followed by 30% co-pay

(e.g Peoples (General) Hospital)

Deductible at a private facility followed b\ a 30%

A review of the program conducted and reported by the World Bank found that enrollment

is lower in poorer households and that the scheme has had no impact on out-of-pocket

spending or on utilization among the poor (Wagstaff A et al., 2007) although it has improved access for the poor in urban areas (Liu GG, Zhao Z, Cai R, Yamada T, & Yamada T, 2002) Wagstaff et al's review also found that coverage is mainly limited to in-patient, curative care with very high deductible requirements The deductible requirement, as illustrated in Table 1,

is not trivial For cataract surgery in Yangjiang City, patients with insurance still have to pay 700RMB for surgery at the People's Hospital This level of deductible, which must be paid for out-of-pocket is regressive and unfavorable for equity, resulting in either household resource depletion or continued visual impairment

2.3 HKI, China and Tiered Pricing

This environment, combined with the high burden of disease, prompted the initiation of the cataract screening and surgery program by HKI and GEH This has increased the CSR in this county to 780 per million annual (compared with the 280 million per year quoted in (Apple et al., 2000)) However, there is still a huge backlog of cataract caused visual impairment and there remain concerns about increasing the uptake of surgery equitably, while maintaining sustainability The program tries to improve equity by offering free surgeries to the small number of patients that have government provided 'proof of need' documents One of these is

an official card that confirms that the patient has no children and is not a dependant The other type of card is the 'Di bao' or Minimum Living Standards Program card, which provides transfers in cash or in kind to those living below the 'di bao' line for their municipality, which lies at or below RMB 637 per year, the national poverty line These cards do not actually entitle the holder to free health services however, HKI/GEH have chosen to provide free services to those who have them even though there is no reimbursement for the services provided from government In practice this amounts to a very small number of surgeries

In order to ensure the support of the local government, a critical success factor for any NGO wishing to operate in China (Hsia R & White L., 2002), the community screening and surgery program was designed in consultation with local health department officials This body agreed to mandate its community health workers to organize the screenings and in return HKI/GEH agreed to provide cataract surgery at a health department mandated price of

500RMB HKI/GEH worked towards ensuring that the program could be sustainable at that price as there is no actual financial contribution from local government However, the 2001 study suggested that even at this price, uptake from the poorest might be limited Unable to

raise the price of basic surgery or to reduce the cost of providing each surgery presently, they have decided to look at a different model to address this situation The other HKI site at the Yang Chun Hospital charges 630RMB, but both these prices are extremely low compared to the standard prices at the GEH, which start at 1,500RMB

2.4 The Aravind Eye Hospital, India

Since its creation in 1976, the Aravind Eye Hospital has grown from 20 to 3,590 beds in

2005 (Aravind Eye Hospital) and has provided free eye care to two-thirds of its patients from the revenue generated by the paying patients who make up the other third (V Kasturi Rangan, 1993) Using a tiered pricing model, Aravind was able to provide three quarters of the

247,235 eye surgeries it performed in 2005 for free (Aravind Eye Hospital) Aravind utilizes revenues from customers who are willing to pay for certain amenities along with the cataract surgery service to subsidize the provision of a 'no frills' service to those unable to pay

anything at all In other words, it practices a form of price discrimination by using tiered pricing to maximize its output conditional upon offering as many as possible of the surgeries for free and subject to breaking even

The 'tiers' of service are the availability of three different classes of rooms 'A' class rooms include a bed and a private toilet for which the patient is charged $3 a night 'B' class rooms include a bed and a shared toilet for which a patient is charged $1.50 a night ' C class rooms are basically rooms with a mat for which the patient pays $1 a night All fee paying patients pay a $1 per outpatient visit Other chargeable services include the choice of surgical techniques (more expensive surgical techniques have shorter recovery times) and different brands of IOLs The 'free' patients occupy a separate building with 'wards' that are large airy rooms with rows of mats for patients and their caregivers The quality of clinical service

however is the same in terms of surgical outcomes whether the patient is fee paying or not because the same surgeons perform them with the same equipment

Essential to its ongoing viability as a self-sustaining firm and its plans for growth, Aravind places emphasis on minimizing costs by purchasing high quality capital such as appropriate modern equipment, research into improved techniques and using both to their maximum capacity In addition, since 1992 when it set up Aurolab, it has been manufacturing and selling its own brand of intraocular lenses (IOL) for lens replacement after extra capsular cataract extraction (ECCE) This has brought the cost of IOLs down from $30 (Rs 800) to $7 (Rs 200) This vertical integration is part of an ongoing mission to minimize the cost of increasing the quality of its services by offering all its free patients IOLs, an improvement over the 'coke-bottle' aphakic glasses that were previously provided for post surgery vision correction by intracapsular cataract extraction (ICCE) and increase revenues by selling the lenses to other facilities By constantly seeking ways to minimize costs while producing a high quality service that is demanded by both fee paying customers and the 'poorest' of the poor, Aravind is able to maintain a ratio of paying to non-paying customers of 1:2 with the firm maintaining an expenditure to income ratio of between 48-51% (Kumar Nirmalya & Brian Rogers, 2000)

The core principle of the Aravind System is that the hospital must provide services to the poorest and to the non-poor alike, yet be financially self-supporting This principle is

achieved through high quality, large volume care and a well-organized system Beyond the

economic model however, Aravind's success is as much due to the social role that it sees itself

fulfilling - both its founder and its staff are completely dedicated to providing services for free

HKI/GEH would like to determine if there might be a similar demand for amenities from some of the users of their program and would like to determine whether there is enough income and preference variation in the population they serve to utilize this model and improve access to cataract surgery while remaining sustainable

2.5 HKI, China and Cataract Surgery

Support for the idea that tiered pricing might be a feasible way to apply cross-subsidization

to HKI, China's operation came from a study conducted at HKI's community screenings in

2001 (He M et al., 2007) It was found that the amount that respondents were willing to pay differed substantially across income groups, suggesting that those in higher income groups may be willing to pay more for cataract surgery In addition, it was found that those with less visual impairment were willing to pay more for surgery than those with greater levels of visual impairment He et al suggest that this may be because younger patients place a higher value on the benefits of cataract surgery as well as the fact that they may still be working and

so better able to afford the service These two findings suggest that is a section of the

population who could benefit from cataract surgery that may consider paying more than the current HKI price for surgery and that the offer of additonal amenities may provide the

incentive for them to actually do so

3 Conceptual Framework

3.1 Defining Equity in terms of Willingness to Pay

Equity researchers choose to focus on different and not always mutually supporting issues

in health care, but generally an 'equitable' situation is one in which people who are in need of

a health service, usually one considered a basic necessity, are able to obtain that service regardless of their ability to pay - i.e equal access for equal need This definition, used for this analysis, is also referred to as 'horizontal equity' (Culyer & Wagstaff, 1993; Kawachi, Subramanian, & Almeida-Filho, 2002)

The determination of equitable access using willingness to pay for cataract sugary data requires two assumptions First we assume cataract surgery is a basic health necessity Given the wide range of other health services that are under utilized across the developing world it is worth supporting this assumption Cataract is not preventable so there are no measures that can be implemented to avoid the eventual need for surgery However, cataract is treatable and

so cataract blindness is preventable Cataract surgery has also been found to be cost-effective

to treat in low income countries (Baltussen, Sylla, & Mariotti, 2004) Finally, cataract surgery has a high success rate for restoring vision, even in resource constrained environments

(Venkatesh, Muralikrishnan, Balent L., Prakash, & Prajna, 2005) Despite all of these

findings, cataract causes an increasing level of disability, particularly in China's aging

population For these reasons we find it reasonable to consider cataract surgery a basic and

essential service In this population it is available but not always accessible, and we have

reason to believe that this is due to the price relative to local per capita income This leads us

to hypothesize that access is inequitable

Assessing Equity

Equity research is a vast body of literature comprised of both philosophical treatments of equity and the practical measurement and assessment issues of interest here Several

measures have been developed to describe the extent of inequity, particularly as defined here

- equal need, equal access The goal of these measures is to see if the observed distribution of health care utilization by income is different from the distribution that would be observed if utilization were distributed by need A commonly used measure is the concentration index, which will be described in more detail in the section on methodology The concentration index is a modification of the Gini coefficient The Gini coefficient is used to measure the degree of inequality in a variable, e.g income, by comparing the levels of income in groups to the total distribution of income The concentration index incorporates the comparison of distribution to another variable e.g health care utilization It is standardized to take a range of values from -1 to +1 where 0 implies that the distribution of utilization is perfectly equitable Values further away from 0 imply increasing inequality across income levels and the

direction, negative or positive, shows who is 'benefiting' from the inequality, positive - the rich and negative values - the poor The concentration index has been used to measure

immunization inequality and the impact of health insurance on the equity of access to health services in developing countries amongst other things (Wagstaff A, 2005; Waters, 2000/8/15)

For the purposes of this study the consumption variable accounts for need incompletely since we are using data from people already known to need/be capable of benefiting from cataract surgery as opposed to a population-based measure However, it is reasonable to assume that while access may differ, since the community screenings are free and randomly

across income levels in the rural community In addition, we make no judgment about the distribution of income but take it to be what it is since we need the variation to implement our intervention/tiered pricing model This reasoning follows from the focus of this analysis, which is on a direct health policy issue as opposed to a broader economic policy For the same reason we will not be 'weighting' the income distribution in the concentration index when measuring the equity of access

3.2 Social Welfare and the Equity-Efficiency Trade-off

The Multi-dimensional Nature of Equity

Researchers in the field of health inequalities debate whether or not equity research can be placed in a welfare economics framework so that we can discuss the actual impact of reducing health inequities, by reducing socioeconomic inequality, on overall social welfare (Bleichrodt

& van Doorslaer, 2006) The reasoning is as follows Inequality of socioeconomic status, for which income is a proxy, does impact the distribution of health, but it is not the only welfare dimension that impacts the distribution of health Therefore, any analysis that treats it as such will be conceptually flawed In practical terms this implies that if we attempted to solve the issues of inequitable access by instituting some transfer of income based on our particular equity goal, those transfers are likely to be at odds with, or even cancelled out by, other similarly proposed income transfers based on preferences with regards to other equity goals such as educational attainment or gender equality Since education gender and health are all reasonable components of a societal social welfare function, income transfer appears to be a

complex solution that would likely fail to address the inequity to anyone's satisfaction An

illustration of this approach would be to charge wealthier people higher prices for the same service because of the need to subsidize the free surgeries given to those who pass some means test that declares them too poor to pay Intuitively this seems inequitable - we

would be asking the wealthier households to pay more, just because they can, when they have the same health need Should some of them choose not to get surgery because of this higher price then clearly equal need, equal access would not hold In addition, this higher price would limit access to all the 'borderline' WTP patients who will now be non-users, noting that 'wealthy' is relative term in this extremely poor rural population, not an absolute one Thus the multi-dimensional nature of the problem becomes intuitive So, the dilemma is the

following: how to effect an improved distribution of health without just transferring the benefits from one group to another This analysis proposes a specific type of tiered

pricing/cross subsidization model that can do precisely this in theory

Clearly the burden of disease is high in this population and any affordable cost-effective measure that decreases the burden of disease is appropriate and should be vigorously applied The subsequent question would be whether or not there is a reduction in the maximal impact

on the burden of disease, by not producing the maximal number of surgeries because we are concerned about the distribution of the reduction in disease burden In terms of individuals' welfare since the demand for the higher priced surgeries represents voluntary self-selection into this category there is no reduction in welfare for these people as their choice set is 'no surgery', 'basic surgery' or 'basic surgery with amenities' However there is an increase in welfare for those who obtain the lower-priced surgery because currently, their ability to take advantage of the entire choice set is limited So the improvement in equity by this proposed model occurs by taking advantage of the inequality of the income distribution that gives wealthier people a larger choice set with respect to purchasing cataract surgery without potential effecting a mere 'transfer' of benefits to the those with lower incomes and a limited choice set We can hypothesize an increase in social welfare due to the increased individual welfare

Equal access to unequal services - an improvement in equity?

While striving for it's objective of improving equity, it is in HKI's interests to produce both amenities and basic surgeries as efficiently as possible to minimize the equity efficiency trade off Increasing its program scale to reduce per surgery fixed costs would be one method

of doing this as Aravind has done so successfully A program on the scale of Aravind's, that provides both improved access to a basic surgery for the poorest and a higher quality

experience to the wealthier patients, is clearly successful However, is it equitable to

implement the same program when the ratio of paying to free/subsidize patients is inverted, that is, we are required to 'sell' many more 'luxury' surgeries in order to provide only a few subsidized surgeries in order to make the model work? Should we just focus on scaling the program up to supply as many we can to those who can afford them? We would have to convince ourselves first that the distribution is indeed inequitable and then that we can

improve it (addressed in more detail in WTP and Social Welfare) This analysis will help to determine what scale we can expect to achieve and we can make an assessment at that point

on whether or not we can really improve equity with this model in this population A follow

on from this discussion is whether a tiered model could be used to provide certain health services in the public sector where the pursuit of efficiency and equity is on going across low-income countries It would seem so, provided there is no aversion to a public sector

institution providing a service with tiered pricing

3.3 The Economics of Tiered Pricing

Tiered pricing systems rely on a basic microeconomic principle referred to as price

discrimination In the model proposed here, there is self-selection by the consumer into different categories of price This is not unlike the self selection induced by airline fare

pricing schedules where some consumers choose to pay more for flexibility to change travel dates and waiting lounges etc but receive what is essentially the same service as those who buy the cheapest tickets Generally the microeconomic model of price discrimination requires that there be a monopoly provider of a service that cannot be resold and so medical care, particularly procedures, is a quintessential example of a service that provides an opportunity for price discrimination In this context the potential competition would come from other public and private hospitals While the price of cataract surgery in this area has decreased somewhat, no one provides a service near the 500 - 630 RMB range, so HKI/GEH effectively have a monopoly However, a limitation of this model is that the total price of surgery after the additional charge for an amenity has to remain low enough for HKI/GEH to maintain their 'monopoly' status Otherwise these consumers might go elsewhere and the model collapses

In terms of non-profit competition at similar/lower prices, such a 'competitor' would be welcome as the backlog of needed surgeries is massive and any sharing of technology that leads to a further lowering of marginal costs would be beneficial to HKI/GEH's objectives of increasing access to surgery

To develop a model of tiered pricing based on the economics of price discrimination, a potential monopoly provider of a service must be able to identify whether and by how much, the service is valued differently by different groups within the population One economic evaluation methodology that estimates the value of a service to an individual is willingness to pay

4 Contingent Valuation and Willingness to Pay

4.1 Contingent Valuation

Contingent valuation has its origins in environmental economics where research is

conducted to value non-market goods that improve the environment for society as a whole and are paid for by society through some public institution These studies use surveys to

determine individuals' willingness to pay for an intervention/good in order to make some estimation of whether society actually values the proposed intervention enough to make the project 'worth' implementing It is referred to as 'contingent' valuation because the value of the benefit as given by the survey respondent is contingent on the actual provision of the good That is, the respondent has to conceptualize a hypothetical market for the good and determine at what price he or she would enter the market, because no such market actually exists

WTP and the Health Sector

Willingness to pay (WTP) surveys, as they are now commonly referred to in the health sector, were adapted for use in the cost-benefit analysis of health care interventions While the costs of most health care interventions are relatively easy to assess, the benefits are much more complex to quantify and the methods used are much more contentious The first point

of contention is whether or not it is appropriate to view benefits in monetary terms This may not be the case if it is considered 'unpalatable' to place a monetary value on 'life' itself

because it follows that assessing differences in quality of life i.e diseased states, in monetary

terms for comparison with each other would be equally distasteful However, since resources are limited, some judgments do have to be made about how to prioritize their distribution amongst the different forms of healthcare services and the service's benefit seems to be the

most appropriate way to do so Without such an evaluation, services will be prioritized anyway but possibly on less appropriate/logical criteria, which do not necessarily result in an outcome that benefits society Consequently the evaluation of benefits, termed economic evaluation is accepted as a necessary component of health services research However when evaluated by a non-monetary measure the comparison between services is complex and the net benefit to cost assessment can be impossible or uninformative to decision makers A net benefit to cost assessment in monetary values is easily understandable particularly to policy makers and allows for comparison across services The next point of contention lies in the way in which the monetary valuation is obtained Contingent valuation has been proposed as

an appropriate method of valuation because it allows respondents to assess the value of

services in terms of their use, non-use and the option of their use (Richard D Smith, Jan Abel Olsen, & Anthony Harris, 1999a) In addition, because it is a monetary valuation, if

appropriately incorporated into a welfare function by asking respondents to consider their WTP in the context of their budget constraints, it allows us to obtain figures that can be aggregated To illustrate, consider the quality adjusted life year (QALY) another measure of benefit but non-monetary in nature The QALY can only provide the valuation of a health state by an individual It provides no information as to the patient's preferences for that health state controlling for other aspects of the patient's life or components of their social welfare function such as their budget or externalities But with WTP values, we can incorporate a broader spectrum of preferences into the valuation, which means that it can be used to assess feasibility in terms of price and access It is precisely for this reason that it is being used

increasingly in developing countries

WTP and its use in Developing Countries

WTP surveys are increasingly being used to evaluate non-market services/goods by their

potential recipients to provide data on the impact that price might have on their uptake Developing country examples include WTP for community-based health insurance bednets for malaria prevention and vaccines for HIV prevention (Bishai D, Pariyo G, Ainsworth M, & Hill K, 2004; Dong, Kouyate, Cairns, Mugisha, & Sauerborn, 2003; Onwujekwe, Chima, Shu, Nwagbo, & Okonkwo, 2001) Techniques continue to evolve to make it more applicable for different goods, populations and objectives (Onwujekwe, 2001; Shiell A & Gold L, 2003; Smith RD, 2005; Smith RD, 2007; Yeung RY, Smith RD, Ho LM, Johnston JM, & Leung

GM, 2006) Reviews that have been published over the last 10 years show an increasing acceptance of the methodology as guidelines and methods have been have developed to address the key issues of the validity of the data and more studies are conducted with results that confirm the reliability of the technique (Diener, O'Brien, & Gafni, 1998; Morrison & Gyldmark, 1992; Portney, 1994; Sach TH, Smith RD, & Whynes DK, 2007; Smith, 2003; Yeung & Smith, 2005) These reviews also show that a shift has occurred in the objectives and use of WTP data In developing countries, where the methodology is increasingly being used, its purpose is not usually for the valuation of goods/services with a view to determining the benefit for cost benefit analysis Instead it is used as a tool to determine the potential impact of prices on the demand for health services This use is largely attributable to the differences in household/individual financing of health care between developed and

developing countries The price of a service has less impact on health services demand when there is insurance and low or no payments are made at the point of service by users such as in developed countries However price is a critical component of demand for health care in the developing world where the size of out-of-pocket payments for health care is not trivial relative to household/individual resources (Newbrander, Collins, & Gilson, 2000)

Choice of WTP as the survey tool to study equity

For the purposes of this study, WTP data were chosen over ex post demand data for

cataract surgery because the latter does not conveniently afford us the opportunity to observe the following:

1) Who is not getting surgery and why Although we could attempt to describe the income distribution of surgery recipients and compare that distribution of the population with the inclusion of prevalence data, we would still not be able to control for other factors that may influence whether or not surgery is obtained such as the visual acuity of the non-user who needs it

2) Accounting for indirect costs from ex post data is cumbersome as it would require patient follow up and in addition could be subject to recall bias However the impact of indirect costs are implicit in the WTP valuation given by a respondent who is yet to purchase the service and so their effect is captured indirectly For example, if the costs of time and travel to obtain a service are large, then they should limit the amount that the respondent would express as their maximum WTP in terms of price

3) The relevant range of price/demand data for this study would not be available because

no other provider in the area offers cataract surgery at prices as low as the ones of interest here

Externalities that benefit households cannot be separated from the benefits that accrue to the patient using ex post data and therefore cannot be evaluated

4.2 WTP and'Demand'

A reasonable definition of willingness to pay for a medical care service is the maximum

the service considers themselves to be just as well off with fewer resources as they would have been forgoing the service and retaining their resources It is the accounting for 'fewer resources' that makes this concept analogous to demand WTP surveys in health care are not intended to elicit a valuation of the health outcome directly In other words, they do not ask the question "How much do you value normal vision in cash?" Clearly this would not give useful data because most people would argue that their vision is priceless, as they would not give it up for any amount of money However, in the reverse situation, resource limitations do not actually allow us to consider paying infinite amounts of money to obtain a treatment for the deterioration of visual acuity This distinction seems simplistic but becomes very

important when WTP questions are designed Failure to correctly convey the need for the respondent to keep their budget and other consumption needs in mind will result in serious hypothetical bias and potentially useless data

An appropriately designed and administered survey will give us individuals' a priori maximum WTP which can then be aggregated to represent the 'demand' for the service The approximation is crude because demand is a measure that provides information about an individual's marginal demand for a good given a change in price WTP data do not provide information on marginal demand at the individual level because the survey obtains the

maximum WTP for a single unit of a good from each respondent and so will only inform as to whether or not the respondent would purchase one unit of the good at the inquired price level

We would have no information about the next unit of consumption and we cannot assume that they would pay the same price for the next unit, even if the cataract is bilateral, due to the

expectation of diminishing marginal return on each unit consumed (If people only get

unilateral cataract surgery then the marginal WTP curve and this demand curve would be the same.) However, in this case, the marginal demand is all the information we need at an

individual level to determine the level of access since, in this environment, cataract removal is usually only sought/provided for one eye even if the cataract is bilateral, usually the one with lower vision However, the aggregation of responses at each price across individuals from a representative sample does inform us as to whether or not if we produce another unit of a good (the marginal good), it would find a buyer at each price level The information that we require is how many surgeries with amenities we can 'sell' at each price level, which is captured in the determination of demand for the marginal unit That is, we can use the data to construct a market demand curve for cataract surgery To make the model practical as a business plan we may need to determine which prices/amenities will maximize revenues in the future but this is beyond the scope of this analysis For this study we will assume that amenities are produced at some minimal cost and we will not attempt to account for

preference amongst amenities by potential consumers

4.3 WTP and Social Welfare

Ideally we would like to determine the impact of improving equity of access to cataract surgery on social welfare Many studies have shown that it is not just the absolute level of utility (from our consumption of goods and services) that determines our welfare as

individuals but also our level of utility relative to the utility of those around us (Kawachi et al., 2002) Given our definition of WTP, it should be theoretically possible to write down a social welfare function that is a function of individual utilities from cataract surgery, income and relative utility status, to see whether some reduction of inequity benefits society In fact this is very difficult to do First, the aggregation of individuals' utilities requires some

assumptions about the marginal utility that each individual gets from the consumption of the good, in this case from one cataract surgery Those assumptions would be based on which

view o f justice' one has A utilitarian view would imply that each individual's unit of utility

is equal and so reducing inequality results in no increase in total social welfare This makes sense for cataract surgery, as improving vision for X is clearly as good as improving vision for

Y once we have ascertained that both are in need Clearly, this is not the view of this study since HKI/GEH values increasing social welfare, by increasing the CSR, and also values increasing the uptake of surgery amongst the poorest, by reducing inequity That is, we would

in some sense 'prefer' to give the surgery to X if he is poorer The focus on the poor

describes a view that is similar to a Rawlsian view of total social welfare that depends only on the utility of the poorest individual and so an increase in this individual's utility is an increase

in social welfare However since our primary goal is increasing the CSR the Rawlsian view seems extreme Therefore we could attempt some 'in-between' view where individual

utilities are weighted before determining the net benefit of a decrease in inequity

Unfortunately weighting does not actually solve this issue because the question of what

weights to apply is debatable and can severely limit the acceptability of results

Consequently, it is usually recommended that cost-benefit analyses with WTP data be

presented without weighting even though this actually implicitly implies equal weighting and not no weighting While this analysis is concerned with equity and social welfare it stops short of estimating how much better off this society will be due to the reduced inequality and

firstly, it can be used to asses both the impact of those WTP is greater than 0 and those whose WTP is less equal to zero (or less than zero if willingness to accept is measured.) Secondly, it captures a valuation that accounts for individual budget constraints and therefore the rest of the social welfare function in aggregate In practice, the determination of who consumes and therefore gets utility from observing an ex post demand curve, does not allow use to observe non-consumers Since price determines who is in the demand curve, anyone whose WTP is below the prices relevant to the demand curve is presumed to have zero demand In health care markets however, policy makers are constantly 'deconstructing' the market in order to incorporate equity considerations WTP valuations affords policy makers the opportunity to observe the hypothetical market that could exist given a different set of prices than the ones observed in real markets, in order to determine what sort of policies might allow these

individuals, previously with zero demand, to become part of the market Consequently WTP gives us the ability to directly observe the way that prices may impact equity when they limit access to health care

4.4 WTP and Externalities

A positive externality of a good is that benefit that derives from the consumption of a good

to anyone other than the consumer and the seller We anticipate that household members stand to gain something from the improvement in visual acuity that accrues solely to the patient who gets cataract surgery WTP as assessed by some other member of the household could give us a measure of this valuation The question is whether to count the household valuation of WTP as the externality in its entirety or whether to count it as an externality only where there is some positive difference between the households' WTP and the patient's WTP Economic theory suggests that externalities are 'additional' to the demand that occurs at the

individual level and that markets at equilibrium produce less than the socially optimum

amount A WTP assessment from both the patient and a household representative might allow

us to test whether there are 'real' positive externalities that accrue from patients obtaining cataract surgery and how large these may be An added dimension to this analysis would be

to distinguish between whether this externality is not in fact, altruism, since the population being examined here is elderly, poor and without insurance and so may be paying for surgery from alternative funding sources, including their households if they reside with their children

An externality is a benefit that accrues to a third party who does not consume it An analysis conducted from a societal point of view that surveyed members of a community who do not stand to benefit directly from other community members' cataract surgeries for example, would be an appropriate way to assess an externality If the finding was that no-one placed any value on cataract surgery then we might conclude that there is no externality to providing cataract surgery to those who need it If on the other hand it was found that these community members did place a value on providing surgery even if they did not directly benefit from it and if they didn't have to contribute to the cost of providing the surgery, then we might conclude that there are indeed externalities to providing cataract surgery

However, this analysis is focused on the financial access to cataract surgery and the

respondents being surveyed are of only two types They are either potential surgery patients who would directly benefit from surgery or they are members of the households of potential patients who could be said to benefit indirectly from having a previously visually impaired household member cured of the impairment We may find that household members do not value the cataract surgery and conclude that they do not benefit from it Alternatively we Altruism would be the case where the third party pays for the good but receives none of the benefit

4.5 WTP and Altruism

Pure altruism is described as behavior that promotes the welfare of others without

conscious regard for one's own self-interests while impure altruism, 'reflects the value of doing good' WTP has been used to assess altruism in term of WTP for insecticide treated nets in Nigeria (Onwujekwe & Uzochukwu, 2004) It was found that WTP estimates had low levels of positive predictive validity in that environment with a Phi correlation of only 49 between stated and actual altruistic WTP However previous studies of WTP and actual behavior for the same intervention in the same region have found higher levels of correlation between stated and actual levels of WTP Specifically, Onwujekwe et al found that 76% of those who where hypothetically willing to pay actually purchased bednets and that the WTP technique correctly predicted the choice of 80% of the respondents in that study (Onwujekwe

et al., 2001) Therefore one might conclude that altruism as measured by stated WTP has a tendency to overestimate altruistic WTP or is simply not an appropriate measure of altruism because the tendency towards hypothetical bias cannot be overcome However, for this study

we believe that it is reasonable to propose that this bias may be less of an issue where the benefit is directed at a household member

In the final analysis, the conclusion of whether or not a deviation in valuation exists between households and patients will depend upon the data that we obtain for household characteristics and the caregiving needs of the patient If the household benefits directly from the cataract surgery because a paid caregiver is no longer required or the patient can return to some income generating work then we cannot conclude that a positive difference in valuation

is altruism but would determine it to be an externality Alternatively we may have to further examine the reasons for a negative difference in valuation i.e when a patient gives a higher