Ethics in Global Health Practice

spiderman2017

ARMSTRONG-2010-Developing_World_Bioethics.pdf

Home >Applied Sciences homework help >Ethics in Global Health Practice

FAIRNESS AND EQUITY IN THE PROVISION OF ANTI-RETROVIRAL THERAPY: SOME REFLECTIONS FROM LESOTHOdewb_267 129..140

RUSSELL ARMSTRONG

Keywords Africa, antiretroviral therapy, bioethics, distributive justice, health priorities, HIV/AIDS

ABSTRACT The number of people in immediate need of anti-retroviral treatment (ART) in the southern African region continues to significantly exceed the capacity of health systems there to provide it. Approaches to this complex rationing dilemma have evolved in different directions. The ethical concepts of fair- ness and equity have been suggested as a basis to guide the development of approaches to select patients for ART. This article reports the results of a case study on patient selection at a rural ART clinic in Lesotho. The purpose of the study was to examine whether or not such concepts had relevance or operative value for a treatment team providing ART in rural Lesotho. The study found that while concepts of fairness and equity were relevant to the work of the treatment team, patient selection practices did not necessarily reflect what these concepts entail. The idea of fairness as a structured, formalized selection process did not figure in the approach to ART provision at the site. A less formal, ‘first-come-first-served’ approach was adopted. While there was knowledge among some team members that social, economic or geographic conditions inhibit individuals and groups from gaining access to ART and that this was inequitable, it was felt that there was little they could do to try to mediate the impact of these condi- tions. The study’s findings pose importance questions about the approach to ART programming in resource constrained settings. The findings also question the relevance of trying to achieve fairness and equity when the gap between need for care and capacity to provide it remains so large.

INTRODUCTION

The HIV epidemic in the southern African region is a societal issue of enormous magnitude. This region con- tains the four countries with the highest HIV prevalence globally: Lesotho: 23.2%, Botswana: 23.1%, Swaziland: 26.1% and South Africa: 18.1%.1 In 2007, there were approximately 5.7 million HIV-positive individuals in South Africa alone.2 Despite a massive effort to scale-up public sector access to anti-retroviral treatment (ART) in the region, the need for treatment continues to exceed the

capacity of these countries to provide it.3 In South Africa, for example, at the end of 2007, of the 1.3 million indi- viduals in immediate need of ART, only 28% were receiv- ing it.4 For the same period in Lesotho the proportion was only 26%.5 Although Lesotho has made a pledge towards achieving a target of 80% coverage for those in need of ART by 2010, progress is limited by difficult

1 Joint United Nations Programme on HIV/AIDS (UNAIDS). 2008. 2008 Report on the Global AIDS Epidemic. Geneva: UNAIDS: 215. Available at: http://data.unaids.org/pub/GlobalReport/2008/JC1510_ 2008GlobalReport_en.zip [Accessed 2 Oct 2008]. 2 Ibid.

3 World Health Organization (WHO). 2008. Towards Universal Access: Scaling Up Priority HIV/AIDS Interventions in the Health Sector: Progress Report 2008. Geneva: WHO: 7. Available at: http://www. who.int/entity/hiv/pub/towards_universal_access_report_2008.pdf [Accessed 2 Oct 2008]: 7. 4 UNAIDS, op. cit. note 1, p. 271. 5 Ibid: 269; Government of Lesotho (GOL). 2007. Annual Joint Review Report 2007/2008 FY. Maseru, Lesotho: Ministry of Health and Social Welfare: 53.

Address for correspondence: Box 0213, Maseru West 105, Lesotho, Southern Africa. Email: [email protected]

Developing World Bioethics ISSN 1471-8731 (print); 1471-8847 (online) doi:10.1111/j.1471-8847.2009.00267.x Volume 10 Number 3 2010 pp 129–140

bioethics developing world

challenges which include inadequate primary health care systems and a lack of skilled personnel to deliver and monitor treatment programs.6 Such challenges take time to resolve and as efforts are made to do so, certain indi- viduals will receive treatment and others will not. Those individuals considered in need of treatment are already in the later stages of HIV disease and most of them have only a limited time before their condition becomes unre- sponsive to ART.7 An estimated 18,000 individuals died of HIV-related causes in 2007 in Lesotho (approximately 1% of the population); there were an estimated 350,000 HIV-related deaths in South Africa during this same period.8

While it was known at the beginning of the regional effort to provide ART that rationing processes would be required, at least in the scale-up phase, effective ways of selecting those individuals who should go first and those who should continue to wait have not developed as was anticipated originally. Despite the creation of model eli- gibility criteria and patient selection processes by teams of experts associated with the World Health Organiza- tion (WHO), the advice of prominent legal and ethical scholars and consensus statements by globally represen- tative stakeholder groups, approaches to this complex rationing dilemma have evolved in different directions.9

This article discusses some of the ethical problems that have emerged using a case study of a health care team working at a rural hospital in Lesotho. The team’s expe- rience challenges the practical value of the advice given by the WHO and others. The study contributes to a growing body of similar experiences, and suggests the need for much deeper ethical reflection and ongoing ques- tioning of the dominant approach to ART provision in Lesotho and elsewhere.

BACKGROUND

In 2004, the WHO proposed some clinical and social criteria to assist in selecting patients for ART in settings where the need was greater than the capacity.10 In

general, the criteria targeted individuals who were already experiencing advanced symptoms of HIV- disease. The criteria also targeted those who were most likely to be able to take ART regularly and correctly over their lifetime. Overall, the WHO suggested the following aim for the design and implementation of ART pro- grams, including patient selection processes:

Remember: ARV therapy for the individual is rarely an emergency! The public health emergency is to get large numbers of the right patients on treatment with good adherence and good overall HIV chronic care.11

Patient selection was designed to put the ‘right’ patients on ART through a process of clinical and non-clinical screening and eligibility determination. In settings like South Africa or Lesotho, however, this approach has proved insufficient as an effective rationing tool.12 The number of individuals who meet the suggested eligibility criteria continues to exceed spaces available for treat- ment in many settings. There is a step between those who are eligible and those who actually gain access to treatment that is a much more direct process of ration- ing that was not part of the WHO model. Very few pro- grams, if any, have attempted this step. R. Stewart et al., noting such challenges in South Africa, lamented the absence of more explicit patient selection strategies and the resulting form of indirect or implicit rationing that was occurring there:

Currently this rationing is indirect and implicitly favours those who are informed, can afford it, are in proximity to facilities, and/or have time to wait in queues. In some instances, clinicians are forced to make choices as to who can or cannot receive treat- ment, decisions they are ill-equipped to handle. Implicit rationing is likely to increase the inequity in provision of, and access to services, while undercutting the potential societal benefits of the programme.13

In the absence of explicit guidelines, the rationing dilemma falls at the feet of the health care providers in treatment programs who may (or may not) develop their own ad-hoc approaches. These may not be sufficient to achieve an effective distribution of the benefits of ART6 GOL, op. cit. note 5, p. 55.

7 WHO, op. cit. note 3, p. 17. 8 UNAIDS, op. cit. note 1, p. 217. 9 WHO. 2004. Participant Manual for the WHO Basic ART Clinical

Training Course. Geneva; WHO; R. Macklin. 2004. Ethics and Equity in Access to HIV Treatment – 3 by 5 Initiative. Geneva: WHO. Available at: http://www.who.int/ethics/en/background-macklin.pdf [Accessed 4 Nov 2006]; N. Daniels. 2004. How to Achieve Fair Distribution of ARTs in 3 by 5: Fair Process and Legitimacy in Patient Selection. Geneva: WHO. Available at: http://www.who.int/ethics/en/background-daniels. pdf [Accessed 4 Nov 2006]; WHO. 2004. Guidance on Ethics and Equitable Access to HIV Treatment and Care. Geneva: WHO. Available at: http://www.who.int/ethics/Guidance%20on%20Ethics%20and%20 HIV.pdf [Accessed 4 Nov 2006]. 10 WHO, op. cit. note 3.

11 WHO. 2006. Chronic HIV Care with ARV Therapy and Prevention; Draft: June, 2006. Geneva: WHO: H26. Available at: www.who.int/ 3by5/publications/documents/chronCareModGenDraftRev1.pdf [Accessed 1 Dec 2007]. 12 S. Rosen et al. Hard Choices: Rationing Antiretroviral Therapy for HIV/AIDS in Africa. Lancet 2005; 365: 354. 13 R. Stewart, A. Padarath & C. Milford. Emerging Threats to Equi- table Implementation of ART in South Africa. Acta Academic Supple- mentum 2006; 1: 286–308. See also S. Bennett & C. Chanfreau. Approaches to Rationing Antiretroviral Treatment: Ethical and Equity Implications. Bull World Health Organ 2005; 83(7): 541–547. These authors found similar results in their review of patient selection strate- gies in Mexico, Senegal, Thailand and Uganda.

130 Russell Armstrong

and could even compromise the ability of treatment pro- grams generally to mitigate the larger, more devastating social and economic impacts of HIV/AIDS in settings like southern Africa.

What form should selection strategies take? With spe- cific reference to providing ART, R. Macklin and N. Daniels provided some guidance at the beginning of the global scaling-up process.14 D. McCoy and B. Kalanda et al., through Equinet, also contributed to the debate.15

The concepts of equity and fairness linked the different perspectives. Macklin, addressing equity, advised that there was a need to choose between different principles of equity and to develop corresponding approaches to patient selection.16 She noted that, although a ‘first-come- first-served’ approach might be chosen, where treatment is offered to eligible patients sequentially, based on their order of arrival at a treatment site, it would likely not achieve equity. Consider the case of a ship that founders, with only limited space in its life boats. Some additional selection criteria are usually applied with a view to trying to preserve some essential aspects of the group or com- munity that is at risk (‘women and children first’, for example). In addition, it is considered unfair and inequi- table that those who get seats in boats first are those who are closest, who can run faster or who have knowledge of where the boats are. It is for similar reasons that other approaches to equitable selection for ART should be considered, according to Macklin. She stepped back, however, from recommending which approach should be followed. She noted only that the decision would be complex and difficult and that the ethical debate on rationing problems offered no clear guidance on which principles or values should predominate.17

Macklin concluded her advice with an emphasis on the need for formal, inclusive, deliberative processes at all levels of ART delivery to decide which of the available approaches to equity should guide patient selection.18 In this respect, she relied on Norman Daniels’ approach to rationing problems. In earlier work, Daniels and his

colleague James Sabin, proposed a reliance on principles of procedural justice when faced with the dilemma of not knowing which values or other criteria should guide rationing decisions.19 In his later comments on the diffi- cult challenge of patient selection for ART, Daniels noted that there could be no substantively fair outcome where fairness meant that all who needed a benefit would receive it.20 Some would get treatment and others would not. All that could be aspired to in this setting would be a fair process of choosing between those in need. This led him to propose his ‘accountability for reasonableness’ model as a tool to assist in the development of patient selection processes and rationing strategies. According to the model, provided the selection process was procedur- ally fair, and was viewed as such by all involved, even those who did not benefit from the outcome would still accept the selection decision. As Daniels had earlier stated:

In the absence of broadly accepted consensus on prin- ciples of fair distribution, the problem of fair allocation becomes one of procedural justice. The basic idea behind this appeal to procedural justice is quite famil- iar. When we lack consensus on principles that tell us what is fair, or even when we have general principles but are burdened by reasonable disagreements about how they apply, we may nevertheless find a process or procedure that most can accept as fair to those who are affected by such decisions. That fair process then deter- mines for us what counts as a fair outcome.21

This approach was created in the context of rationing dilemmas in developed countries where the issues at stake were predominately access to advanced technological interventions or those still in the experimental stage. When Daniels recommended this approach in the context of rationing access to ART, it had not, as yet, been tested in a settings like Lesotho where the rationing dilemma affects a substantial portion of the population; where the impact of HIV-related morbidity and mortality on the social and economic viability of the country is significant in the absence of treatment programs; and where such an evolved conception of distributive justice in terms of health care resources has not necessarily taken hold within the national approach to providing ART.22

Researchers at Equinet were more direct in their response to the challenge of patient selection for ART. They argued for a particular type of equity as the only

14 Macklin, op. cit. note 9; Daniels, op. cit. note 9. 15 D. McCoy. 2003. Discussion Paper 10: Health Sector Responses to HIV/AIDS Treatment in South Africa: Addressing Equity. Harare, Zimbabwe: Equinet. Available at: www.equinetafrica.org/bibl/docs/ DIS10aids.pdf [Accessed 11 Nov 2006]; B. Kalanda, I. Makwiza & J. Kemp. 2004. Proposed Framework for Monitoring Equity in Access and Health Systems for Antiretroviral Therapy (ART) Programs in Southern Africa. Lilongwe, Malawi: Equinet. Available at: www.equinetafrica. org/bibl/docs/DIS24aids.pdf [Accessed 11 Nov 2006]. Equinet is the Regional Network on Equity in Health in Southern Africa, a network of professionals, civil society members, policy makers, state officials and others within the region who act as an equity catalyst to promote and realize shared values of equity and social justice in health. See http:// www.equinetafrica.org 16 Maklin, op. cit. note 9, pp. 4–6. She proposed three such principles: a) utilitarian; b) egalitarian; or c) maximin. 17 Ibid: 3. 18 Ibid: 12.

19 N. Daniels & J.E. Sabin. 2002. Setting Limits Fairly: Can We Learn to Share Medical Resources? New York, NY: Oxford UP. 20 Daniels, op. cit., note 9, pp. ii–iv. 21 Daniels, op. cit. note 19, p. 4. 22 Daniels and Sabin have themselves acknowledged the potential limited applicability of the accountability for reasonableness model in developing world settings. See Daniels, op. cit. note 19, pp. 164–167.

Results from a Case Study of a Rural ART Program 131

effective outcome of patient selection processes. For- mally, for them, equity was the absence of discrimination in an ART program on the basis of arbitrary or non- relevant criteria. Substantively, it was the alleviation of unfair or avoidable health disparities in the distribution of ART. As they noted: ‘Equity implies an approach that gives more to those who have little, and thus less to those who have much.’23 Practically this meant using patient selection to break through the barriers of health inequi- ties by, in effect, prioritizing the provision of ART towards those least able to access health care services or to benefit from health care resources. It also meant lim- iting or refusing access to those with comparative social or economic advantages who were able to obtain ART from alternative sources. McCoy warned that unless pro- active attempts were made to ensure equity, the distribu- tion of ART was likely to follow the path of existing inequities and not to benefit those in greatest need.24 R. Loewenson and McCoy went further to suggest that, in the extreme, corruption and abuse within treatment programs could also occur.25 For the Equinet group, the provision of ART on a large scale was a major opportu- nity to strengthen weak and dysfunctional health systems. It was also a chance to try to address and resolve entrenched inequities in many southern African settings in terms of which individuals and groups, based on social and economic condition, had ready access to health care services and which had little or none.

When taken together, all of the commentators sug- gested that (procedural) fairness and (substantive) equity should guide the development of rationing strategies. They further suggested that the following characteristics of the patient selection process (over and above issues of eligibility), where they were apparent, had the greatest potential to achieve ethically acceptable outcomes:

� The existence of formal, explicit eligibility criteria developed through a structured, inclusive and trans- parent process of deliberation; and,

� The existence of a structured, inclusive, transparent and publicly accessible process for applying the crite- ria and for deliberating on exceptions.

How have health care teams that work at the front-line of ART program delivery fared in terms of this advice? Have the proposed relationships between patient selec- tion processes, fairness and equity been absorbed and implemented in these settings? The limited evidence that is emerging in the literature regarding this suggests that it has not. Take for example the early experience in Malawi

were, according to A.S. Muula, although 150,000 indi- viduals were originally estimated to need antiretroviral therapy in 2004, only between 25,000 and 50,000 were to receive it over a five-year roll-out period. While the National AIDS Commission did organize a broad, delib- erative process to discuss eligibility criteria and priority groups, ultimately a first-come-first-served approach was chosen mainly because no consensus could be reached on which groups should be treated first.26

A detailed account of patient selection challenges was provided in a case study of a Médecins-sans- frontières-sponsored treatment project in Khayelitsha, South Africa.27 The treatment team adopted a first-come- first-served approach using clinical and social criteria to evaluate patients’ eligibility for ART, similar to the WHO model. Committees were used to assess individual patients for treatment. However, few if any patients were deemed not eligible. They were simply given more time and assistance to meet the eligibility criteria: ‘We never definitively give a “no” . . . We just say, “not ready.” ’28

There was extreme reluctance on the part of the team to enter into more direct forms of selection and more explicit approaches to rationing.29 A study of the public sector ART program in Free State, South Africa shed further light on patient selection challenges.30 These authors highlighted the many indirect mechanisms of selection that result from program design. In the absence of explicit guidelines or directions, implicit rationing approaches evolve where otherwise eligible individuals simply drop out of ART programs because they are too cumbersome, in terms of the number of steps involved (and the number or repeat clinic visits required), too dif- ficult to access geographically, or too burdensome with respect to some of the eligibility criteria that are imposed. What is evident from these examples is that serious, ongoing challenges continue to arise in the provision of ART in these settings. As the case study presented below

23 McCoy, op. cit. note 15, p. 11. 24 Ibid. 25 R. Loewenson & D. McCoy. Access to Antiretroviral Treatment in Africa: New Resources and Sustainable Health Systems are Needed. Br Med J 2004; 328: 242.

26 A. S. Muula. Ethical and Programmatic Challenges in Antiretroviral Scaling-up in Malawi: Challenges in Meeting the World Health Orga- nization’s ‘Treating 3 Million by 2005’ Initiative Goals. Croat Med J 2004; 45(4): 415–421; For a similar example, see also WHO. 2006. Equity and Fair Process in Scaling Up Antiretroviral Treatment: Poten- tials and Challenges in the United Republic of Tanzania. Geneva: WHO. Available at: www.who.int/entity/ethics/Equity%20and%20Fair %20Practice%20in%20ART_Tanzania.pdf [Accessed 4 Nov 2006]. 27 R.C. Fox & E. Goemaere. They Call It ‘Patient Selection’ in Khayelitsha: The Experience of Médecins Sans Frontières-South Africa in Enrolling Patients to Receive Antiretroviral Treatment for HIV/ AIDS. Camb Q Healthc Ethics 2006; 15: 302–312. 28 Ibid: 306. 29 S. Benatar. Facing Ethical Challenges in Rolling Out Antiretroviral Treatment in Resource-Poor Countries: Comment on ‘They Call It “Patient Selection” in Khayelitsha.’ Camb Q Healthc Ethics 2006; 15: 322–330. 30 N. Jacobs, H. Schneider & H.C.J. van Rensburg. Rationing Access to Public-Sector Antiretroviral Treatment During Scale-Up in South Africa: Implications for Equity. Afr J AIDS Res 2008; 7(1): 19–27.

132 Russell Armstrong

demonstrates, the challenge is a reflexive one. On the one hand, there is an extreme reluctance to engage in direct rationing and to prefer more implicit and somewhat ques- tionable responses to the provision of ART from an ethical perspective. On the other hand, dominant norms of justice, in particular distributive justice, appear to have little valence for health care providers who must make patient selection decisions within specific, real-time contexts.

STUDY PURPOSE AND DESIGN

The purpose of this study was to examine a patient selec- tion process from the perspective of equity and fairness using the experience of a health care team working at an ART site in rural Lesotho. The study followed an explor- atory case study design.31 It occurred at St. Charles Mission Hospital in Seboche, Lesotho. The hospital pro- vides ART through its outpatient department as part of a collaborative project with the Government of Lesotho and a Swiss-based international development organiza- tion called SolidarMed. The site was chosen because it is representative of the setting where most local ART pro- grams operate in Lesotho (i.e. outside of urban settings, and where the population is impoverished and geographi- cally disbursed across extremely difficult terrain). The period of analysis was from May 2005 to December 2007.

There were two components to the method structure: a qualitative component to document and assess the patient selection process and; a quantitative component to measure substantive equity. Data for the study was collected through semi-structured interviews of key in- formants, document reviews, direct observation and the abstraction of aggregate, anonymized data on patient characteristics from monthly progress reports and treat- ment registers. The qualitative data analysis involved iso- lating and coding data segments, arranging them into thematic groups and then using an iterative process of combining and distilling the segments to come up with concise, robust explanations of what was observed.32 The quantitative analysis examined variations in key charac- teristics (age and sex) of the population receiving treat- ment against those in need of treatment, including a comparison of these groups against the general popula- tion living around St. Charles. Chi-squared was used to test for the independence of any observed variations.33

The study was approved by the Health Research Ethics Committee of the Faculty of Health Sciences, University of the Witwatersrand; the Research Ethics Committee of the Ministry of Health and Social Welfare, Government of Lesotho; and the Medical Superintendent of St. Charles Mission Hospital on behalf of the Board and the administration of the facility.

FINDINGS

Patient selection at St. Charles Hospital

To situate the discussion of fairness and equity in relation to patient selection at St. Charles Hospital, it is first necessary to briefly describe the setting in which the hos- pital provided its services and the patient selection process itself. At the time of the study, the hospital oper- ated in a rural environment where poverty was dominant. Most individuals were unemployed and occupied in sub- sistence level agriculture. Remittances from mine workers employed in neighbouring South Africa were a primary source of income for many families. ART was provided by a healthcare team that included four expatriate doctors (three Swiss and one Congolese), one nurse clini- cian, two nurses, four HIV counsellors, two pharmacy technicians and one expert patient (an HIV+ individual on ART who counselled and educated others). A Swiss individual from SolidarMed provided administrative support to the ART program. Individuals seeking ART travelled to the hospital to receive it across a terrain that was mountainous and poorly serviced by the public trans- port system (taxis). Some communities were only acces- sible by foot or on horseback. According to the health care team, money for transport was the major barrier preventing individuals from accessing the ART program at St. Charles.

There was no official estimate of HIV prevalence for the hospital’s catchment area and hence, an accurate esti- mate of the need for ART could not be calculated. Need for ART could be broadly estimated by applying the district level HIV prevalence estimates (adult male 18.7%; adult female 24.3%; children 6% or less) to the catchment population (38,182) giving an estimated 5,873 HIV+ adults and children.34 According to program data, there were 692 HIV+ individuals enrolled in the ART program,

31 R.K. Yin. 1994. Case Study Research: Design and Methods. Second Edition. Thousand Oaks, CA: Sage Publications: 38–44. 32 M.B. Miles & A.M. Huberman. 1994. Qualitative Data Analysis: An Expanded Sourcebook. Second Edition. Thousand Oaks, CA: Sage Pub- lications: 85. 33 This aspect of the method followed the suggestions of Kalanda, op. cit. note 15.

34 GOL. 2005. Lesotho Demographic and Health Survey 2004. Maseru, Lesotho: Ministry of Health and Social Welfare & Bureau of Statistics; United National Development Program (UNDP). 2007. Human Devel- opment Report 2007/2008. Geneva: UNDP. Available at: http:// hdr.undp.org/en/media/hdr_20072008_en_complete.pdf [Accessed 20 Jan 2008]; GOL. 2007. Population Census 2006. Maseru, Lesotho: Bureau of Statistics. Available at: www.bos.gov.ls/Census2006b.htm [Accessed 14 Dec 2007]. The population estimate for the St. Charles catchment area combines the constituencies of Mechachane and Hololo.

Results from a Case Study of a Rural ART Program 133

representing 24% of the number of individuals who had sought HIV testing at the hospital during the study period. The WHO estimates that between 15% and 19% of the HIV+ population is in immediate need of ART.35 This gives a range of between 873 and 1105 individuals in need of ART in the St. Charles catchment area. At the end of the study period, 462 individuals had been enrolled on ART representing approximately 67% of the HIV+ group. Operational reports for the provision of ART showed that the patient group was predominantly adult females (65%). Adult men and children of both sexes made up a much smaller proportion of the patient popu- lation (25% and 10% respectively). This differed mark- edly from national census estimates of sex distribution (49.5% male) and international estimates of age distribu- tion for the Lesotho population (40.4% of the total popu- lation is under the age of 15).36

Figure 1 shows the main stages an individual passes through in gaining access to ART at St. Charles Hospital.

Individuals entered the ART program by way of a clinical finding of HIV infection.

Following this, an eligibility assessment was per- formed. The eligibility criteria originated from different sources, including guidelines published by WHO in 2004 and revised in 2006; national ART treatment guidelines for Lesotho; practices based on the WHO’s Integrated Management of Adult and Adolescent Illness (IMAI) program and; practices based on what the team at St. Charles understood the national approach to be even when the source of such understanding could not be fully substantiated.37 At the time of the study, there was not one, overall document either at St. Charles or nationally that set out all of the eligibility criteria in use.

Following the WHO model, there were two types of criteria to determine eligibility for ART. Clinically, an individual was required to be in the advanced, symptom- atic stages of HIV-disease, including a CD4+ lymphocyte count of �200 cells/ml2 and the presence of one or more

AIDS-defining conditions. Once an individual met the clinical eligibility criteria, he or she was required to at tend a minimum of three adherence preparation sessions where the importance of adherence to the ART regimen was explained in detail. Patients were further required to disclose their status to another individual (usually a family member or close friend) who also attended the sessions. During the last adherence session, the patient’s readiness, or his or her knowledge of ART and the impli- cations of starting treatment, were assessed by either the ART Nurse or the HIV/AIDS Coordinator using an oral test. If the patient was not successful, more adherence preparation sessions were scheduled. After the readiness test, a patient was further required to give active consent to start treatment (indicating willingness) in the presence of the doctor who then prescribed ART. If, during the consultation, the doctor determined that the patient was not ready or not willing, he or she was referred back to the ART Nurse for more adherence preparation sessions. There appeared to be no official limit to the number of times a patient could be referred back to adherence coun- selling. The treatment team took a very enabling approach, offering as much assistance as was necessary. However, one informant noted that after several sessions, patients themselves would simply not return. The team could not provide an estimate of the number of patients who did not return for additional adherence sessions after failing to pass the initial readiness test.

Some selection criteria had exceptions, others did not. Presumably, the purpose of exceptions was to recognize special circumstances and enable access to treatment in those cases where individuals might otherwise have been ineligible for ART. In the eligibility process, a clinical finding of HIV infection was required, usually in the form of a positive result from an HIV test administered at the hospital. However, if a patient was unable to give consent to be tested or to participate in pre-test counselling because of impaired capacity, testing did not take place and an individual could not proceed through the patient selection process (this affected individuals with disabili- ties or the gravely ill; there was controversy among the team members on this issue). Pregnant women could be started on ART after only two adherence counselling sessions when other patients were required to attend three. On the requirement for disclosure and for a treat- ment supporter, discretionary room was limited. Some individuals were able to begin ART without having dis- closed their status or identifying a treatment supporter. It was not clear from the key informants on what basis such an exception was granted other than the persuasiveness of the particular patient to argue that either disclosing HIV status or identifying a treatment supporter was impossible to do. What was clear was that exceptions to eligibility criteria were limited. None appeared to have emerged from a formal, deliberative process. Most had

35 WHO. 2005. Summary Country Profile for HIV/AIDS Treatment Scale-Up: Lesotho. Geneva: WHO. Available at: http://www.who.int/ hiv/HIVCP_LSO.pdf [Accessed 1 Dec 2007]. This estimate used a CD4+ lymphocyte count of �200 cells/ml2 as one parameter for calculating need for ART. During the study period, the WHO was changing this parameter to �350 cells/ml2, a change that would lead to a much larger number of individuals in immediate need of ART. 36 Bureau of Statistics, op. cit. note 34; UNDP, op. cit. note 34. 37 WHO. 2006. Antiretroviral Therapy for HIV Infection in Adults and Adolescents: Recommendations for a Public Health Approach. Geneva: WHO: 70–71. Available at: http://www.who.int/hiv/pub/guidelines/ artadultguidelines.pdf [Accessed 1 Dec 2007]; GOL. 2004. National Anti-Retroviral Treatment Guidelines. Maseru, Lesotho: Ministry of Health and Social Welfare; WHO. 2004. Participant Manual for the WHO Basic ART Clinical Training Course. October 2005 Update. Geneva. WHO; WHO. 2006. Chronic HIV Care with ARV Therapy and Prevention:Draft: June, 2006. Geneva: WHO. Available at: www. who.int/3by5/publications/documents/chronCareModGenDraftRev1. pdf [Accessed 1 Dec 2007].

134 Russell Armstrong

been adopted by the ART team as a result of either training curricula or specific statements in national or international guidelines.

Fairness

In addition to an objective assessment of the patient selection process from a procedural fairness perspective, fairness was also explored in the study using a simple open-ended question: ‘Do you think the [ART patient selection] process is fair?’ Informants viewed fairness as both an absence of (moral) judgement and the consis- tent application of the national guidelines as they were understood by the treatment team. According to one informant:

I didn’t judge that person. I treated that person like every other client. I didn’t do any favour. I didn’t impinge upon his or her right. I just did what was agreed upon. . . . We are not only making our deci- sions. We are following the guidelines.

In the words of another: ‘I try to treat everyone the same. I just sit in my office. I just see the people I see.’ Fairness also meant that no patients were refused: ‘We don’t refuse anybody . . . Anyone who needs help from us, they get it. We don’t pick and choose.’ When a patient did not proceed through the steps towards gaining access to ART, it was not because they were not selected for treat- ment. It was because the eligibility criteria were not met, a situation that was the responsibility of the patient. To

Individual comes to St. Charles seeking HIV test.

Individual comes to St. Charles seeking medical care and is referred by nurse or doctor for HIV test.

Individual comes to St. Charles with HIV-positive result.

HIV testing. If HIV-positive, sent to laboratory for CD4.

Clinical assessment. Application of clinical eligibility criteria. If eligible, referred for adherence counselling

Adherence counselling. Three sessions (two if pregnant).

Readiness Assessment (same time as last adherence counselling session.).

Return to doctor for first prescription and willingness assessment.

Collection of ARVs from dispensary. Further ‘readiness’ assessment.

CD4 test. Other laboratory diagnostics if required.

If not ‘ready’, repeat adherence counselling.

If not ‘willing’, refer back to adherence counselling.

If not ‘ready’, issue medications but refer back to adherence counselling.

Clinical eligibility assessment

Non-clinical eligibility assessment

Ready? Willing?

START

Figure 1. Patient Selection for ART at St. Charles Hospital

Results from a Case Study of a Rural ART Program 135

the extent that there was patient selection, it was implicit. Where selection occurred, it was viewed by the team as self-selection, meaning that patients self-selected not to continue with the process of meeting eligibility criteria for reasons that were beyond the team’s control.

Aspects of unfairness emerged within the treatment team with respect to patients who were very ill and unable to participate in adherence counselling or even unable to provide consent to be tested for HIV. At the time of the study, there was no clear national guideline to help treat- ment providers in these situations.38 The differences within the treatment team regarding this issue mirrored the foreign and local mix of healthcare staff. One (Swiss) informant, noting the reluctance of hospital staff to make exceptions in difficult circumstances described the situa- tion in this way: ‘You follow national guidelines [in Lesotho] even if they are sometimes not logical.’ For the local members of the treatment team, their perspective was as follows: ‘We have the guidelines. We went for IMAI. We have the guidelines. We don’t discuss it.’ Doctors felt constrained by this attitude as it limited their autonomy to test and initiate ART on their own regard- less of the views of other staff. According to one doctor: ‘I cannot do it [perform an HIV test on a patient who is not competent]. They [the nursing staff] would not start [ART]. I just have to let them die. It is not okay.’

Equity

At St. Charles, individuals may not have been able to receive ART not because they were not eligible but because they could not travel to the hospital to enter the program, for example, or they were unable to pay the nominal user fee for the provision of outpatient services. Equity in this study was used as a measure of what efforts were made to overcome such barriers or exclusions. It was explored in two ways in the study: through the key informant interviews and through an analysis of opera- tional data. In the interviews, the researcher explained equity as all individuals having equal chances to receive treatment suggesting that where there were barriers or challenges, efforts were made to address these inequities. For some informants, the opportunity to access treat- ment was the same for all. What hindered access to treat- ment were individual barriers that could not be resolved by the treatment team: ‘We don’t hinder anyone. The clients hinder themselves.’

Other members of the treatment team had different views. According to one informant, ‘Some people don’t

get treatment because they are too far.’ She suspected that the very sick or the very poor did not return to the clinic once they were seen initially. ‘I can imagine they won’t come back. They are badly dressed or the transla- tor says they won’t come back.’ In the words of another informant: ‘The more you know, the closer you are, you are more likely to come for service. I’m not sure. But it’s not only for ARVs, it is for everything.’ According to still another, there were insurmountable challenges in the way of achieving equity, including geography, infrastructure and lack of staff: ‘We are not marginalizing. We are forced by circumstances.’ One informant felt that indi- viduals should not all have the same chances and that there should be priorities: ‘I don’t think everyone should have the same chance. I am more reluctant to put older people on treatment unless they are care takers.’ For this informant, younger people, children and mothers were priorities. ‘We have to do a triage. . . . I don’t tell them [older patients] I won’t put you on treatment. I just don’t insist.’

Despite variances in the interpretation of equity and its application to the provision of ART, team members did acknowledge that social, economic or geographic condi- tions inhibited individuals and groups from gaining access to ART and that this was inequitable. As a result, the team at St. Charles made efforts to address these barriers. It provided reimbursement for transport costs for those patients with this difficulty. The primary care outreach team conducted regular visits across the region and integrated HIV testing and counselling and patient follow-up within these more general efforts. The hospital was in the process of implementing decentralization of ART, meaning that it was attempting to make it available at local community health centres across the region rather than exclusively through the hospital’s outpatient clinic.

Equity was also assessed in the study by examining program reports and other sources of operational data for significant variations in the characteristics of those individuals receiving treatment against those determined to be in need but not yet on ART. Table 1 shows expected and observed values for the population surrounding St. Charles Hospital and those seeking HIV counselling, testing and initiation on ART. Expected values were only broadly determined from census data, national pre- valence estimates and other published sources.39 No information on the specific characteristics of the catch- ment population of St. Charles and the need for ART was available.

The proportions of those seeking HIV testing in the Seboche were 75% women, 19% men and 5% children, dramatically different than the estimated population dis- tribution of 30% women, 29.6% men and 40.4% children.

38 Revised national ART guidelines were under development at the time of the study. The new guidelines will provide mechanisms for determin- ing HIV status and initiating treatment when patients are clearly suffering from AIDS-related illnesses but are incapable of providing consent. See GOL. 2007. Lesotho National ART Guidelines: Working Draft. Maseru, Lesotho: Ministry of Health and Social Welfare. 39 See note 34.

136 Russell Armstrong

Those seeking testing were overwhelmingly adult women. Of those who presented for testing and were found to be HIV-positive, they were also overwhelmingly women (67%). Of the males in the population who sought testing, a much larger proportion were HIV-positive (36%) than what was estimated nationally (18.7%) in Lesotho. Com- paring the estimated need for ART (at 15% and at 19%)40

with the number of those initiated on ART at St. Charles, one could generally conclude that the program had reached between 41% and 53% of the overall immediate need for treatment. For men the range was 30% to 38%; for children it was 24% to 30%.

Using a chi-squared test for independence of these variances showed significant results [rejection region: c2 � 14.8602 df = 4 p = .005; result: c2 = 3130.882]. The results leant weight to the observation that adult males and children of both sexes were significantly under- represented among those seeking HIV testing and those estimated to be in need of ART in the geographic region serviced by St. Charles Hospital. When the result was shared with key informants, it was noted that men gen- erally do not access health care services until they are very sick and that many men whose families live in the region around Seboche were employed in South Africa. It was also noted for children that: ‘Treatment is too difficult and not sustainable.’

DISCUSSION: THE ETHICS OF PATIENT SELECTION

What light can fairness and equity shed, then, on patient selection at St. Charles? First of all, it should be noted that none of the key informants felt that the ART program was either inundated or overwhelmed by demand for treatment. Rather, it was a case of many individuals in need of treatment not coming to the hos- pital to enter the program. There was no overall strategy to target or prioritize segments of the community around the treatment site. There was no evidence of any discus- sion in terms of which groups in the community should gain immediate access to treatment and which groups should continue to wait. The team believed its role to be to administer the eligibility criteria, to follow ‘guidelines’, and not to engage in any more direct process of rationing or prioritization. Those things that prevented more adults and children from coming forward for treatment were largely beyond what the treatment team could influence.

As for procedural fairness, the treatment team showed only a limited understanding of its role. To them, they applied eligibility criteria that were national and interna- tionally determined. They did this on a first-come-first-

40 WHO, op. cit. note 35.T ab

le 1.

O bs

er ve

d an

d ex

pe ct

ed va

lu es

fo r

up ta

k e

of H

IV te

st in

g an

d A

R T

A B

C D

E F

G H

I J

K L

S ex

/A ge

P op

ul at

io n

% T

ot al

O bs

er ve

d S

ee k

in g

T es

ti ng

E x

pe ct

ed S

ee k

in g

T es

ti ng

E st

im at

ed H

IV +%

O bs

er ve

d H

IV +

E x

pe ct

ed H

IV +

(E *F

)

E st

im at

ed N

ee d

fo r

A R

T (a

t 15

% )

% T

ot al

O bs

er ve

d A

R T

E x

pe ct

ed A

R T

(a t

15 %

)

M al

e 11

,2 64

30 %

55 4

86 3

18 .7

% 19

9 10

4 31

6 36

% 11

9 16

7 F

em al

e 11

,4 92

30 %

21 66

86 3

23 .4

% 46

4 53

7 41

9 48

% 30

1 22

2 C

h il

d (<

15 yr

s) 15

,4 26

40 %

15 8

11 51

6% 29

9 13

8 16

% 42

73 T

o ta

l 38

,1 82

10 0%

28 78

69 2

87 3

10 0%

46 2

Results from a Case Study of a Rural ART Program 137

served basis, treating patients in the sequential order in which they arrived at the hospital, in what was perceived to be an open and non-discriminatory fashion. Where there were evident procedural problems, there was no awareness within the treatment team of the role of struc- tured deliberation in resolving these challenges. As noted earlier, according to the international guidance on fair- ness in patient selection processes, decision criteria must have a clear relationship to an overall program goal. Furthermore, where exceptions are made, they must be objectively justifiable and be applied consistently across the decision-making process.41 Upon analysis, not all cri- teria used at St. Charles were clearly connected to pro- viding ART to the greatest number of patients who would derive the most benefit, the implicit goal of the ART program. The requirement that an individual dis- close his or her HIV status to someone else and identify a treatment supporter was a very significant one, particu- larly within a rural context where stigma against HIV- disease predominates. As Macklin has noted, requiring disclosure infringes on individual privacy and insisting on a treatment supporter limits autonomy.42 Despite state- ments in WHO training materials and in Lesotho’s national guidelines assuring health care providers that such things lead to better adherence, there exists little empirical evidence to support these criteria as improving the efficacy of treatment.43 Given repeated counselling sessions and other tactics that are used with patients in order to persuade them to disclose their status or to locate a treatment supporter, and given the fact that some who do not disclose or cannot locate a treatment supporter are not started on ART, the rationale for the application of these criteria becomes questionable. One must ask what is the value of these efforts if, in the end, their effect on patient adherence is not measurable.

There were larger problems in the patient selection process with respect to exceptions to the eligibility crite- ria, both in terms of how they were determined and how they were applied. Some individuals were able to receive ART without disclosing their HIV status or without engaging a treatment supporter. The determination of this appeared entirely discretionary. The area of greatest concern was the lack of exceptions for the very ill. Here, the treatment team had two challenges. Firstly, in the

absence of a deliberative process to address and resolve differences in point of view, confusion and tension remained within the team and individuals who needed ART were not started. Secondly, national and interna- tional treatment guidelines are silent on when initiating ART is considered futile, particularly in a resource- limited setting where the human resources necessary to provide care like those on the treatment team at St. Charles are precious and limited. Consequently, there was nothing to assist the team in coming to a resolution on when very sick patients could benefit from ART and when they could not.

In relation to the concept of equity, the treatment team showed a better understanding of its role in selection processes and the need to ensure that ART was accessible to the different individuals and groups who required it. There was knowledge that barriers exist in the Seboche region, largely related to poverty and geography. These were addressed with transport subsidies and outreach services, for example. Where knowledge about equity was more limited was in regard to the impact of the eligibility process in relation to poverty, distance and difficult geographic terrain. Completing the eligibility process required repeat visits to the hospital. While there was awareness that some patients stopped coming if the process became too burdensome, there was no knowledge of the full extent to which this acted as a barrier to access- ing treatment.

With respect to other equity challenges, such as the absence of men and children from the treatment program, the team felt somewhat powerless. This pattern is not unique to St. Charles; it reflects what has been observed across the southern African region. It is the result of the complex interplay of social, economic and cultural factors. In the rural areas of Lesotho, men must travel far to seek work while women remain in towns and villages to care for families. Culturally, illness generally, and HIV disease in particular, are stigmatized among males, particularly in rural environments where the tra- ditional attitudes and beliefs about men’s roles still pre- dominate. Lesotho, overall, has very poor child health indicators with very high rates of mortality, even in com- parison to other countries at similar stages of develop- ment.44 It is not surprising, then, to see this reflected in the absence of children amongst those seeking HIV testing and treatment.

CONCLUSION

It is clear that the architecture of criteria and selection processes aimed at achieving fairness and equity have not fully informed the development of the ART

41 Daniels, op. cit. note 9, p. ii. 42 R. Macklin. No Shortage of Dilemmas: Comment on ‘They Call It “Patient Selection” in Khayelitsha.’ Camb Q Healthc Ethics 2006; 15: 316–317. 43 See E.J. Mills et al. Adherence to HAART: A Systematic Review of Developed and Developing Nation Patient-Reported Barriers and Facilitators. PLoS Med 2006; 3(11): 2055. The study found evidence in developed world settings that disclosure within a supportive social network improved adherence. No such evidence was found in develop- ing world settings largely because the issue had never been explored. Similarly, in the WHO ARV treatment guidelines of 2006 (WHO, op. cit. note 22, p.70), a similar statement is made. 44 UNDP, op. cit. note 34.

138 Russell Armstrong

program at St. Charles Hospital. There is a distance between model selection processes and criteria proposed at the international level and the specific practice of patient selection in a rural environment in Lesotho. Is it this distance that is contributing to gap between those on treatment and those in need, or is it something else? Rather than highlighting problems in the program at St. Charles, perhaps the ethical analysis should point to much larger limitations in the overall approach to ART provision in such circumstances. This is not a criticism of the team at St. Charles or its tremendous efforts. Rather, it is a challenge to the global effort to provide ART in resource-limited settings which, as D. McCoy et al. warned, was fraught with pitfalls and would not result in an equitable distribution of ART.45 The pitfalls, however, may not be caused by the lack of more explicit approaches to equity or more formal, deliberative pro- cesses within treatment programs. They may be caused by something more fundamental. As Alvarez has observed, for example, in similar settings of extreme resource scarcity: ‘Some resource scarcities are too severe and way below the amount that can sustain a decent level of health that they undermine any attempt to make allocation ethical. . . .’46

Perhaps, when the gap between need and capacity is large, as it ostensibly is for ART in the Seboche region and across Lesotho, there can be no such thing as an ethical approach. What emerges is an approach that is pragmatic and expedient and that tries to give a limited benefit to as many as possible without engaging intrac- table questions of which individuals to choose and which not. It may be that measuring equity and fairness in such situations is at best fruitless or, worse, irrelevant.

While selection criteria and selection processes help facilitate difficult work in rationing dilemmas by attempt- ing to make the outcome of such choices as fair and equitable as possible, they also serve other ends. With respect to ART, they also balance risk. An ART program improperly implemented or controlled runs the risk of encouraging the development and spread of treatment- resistant strains of HIV. While this has not yet emerged in Lesotho (there was no technology in-country at the time of the study to examine this), there is still the potential that it could. Using a process like the one at St. Charles sets a certain pace for how many people can move through the various stages of preparation and end up on treatment within the limits of a standard working day or a standard working week. The team was not, at the time of the study, inundated with individuals seeking ART. If it wanted to significantly improve the coverage of the

ART program, much bolder steps are required than tink- ering with eligibility criteria or the number of steps in the selection process. To reach significantly more people with ART perhaps only a minimum of steps are required and a new balancing of the risk equation.

What would the outcome be of much wider access and much fewer eligibility steps? Some individuals would die from not taking their treatment properly. Some individu- als would develop drug resistant HIV and potentially transmit it to others. Would these risks be balanced, though, by a significantly greater number of people taking treatment and doing it correctly over the long term? As S. Rosen et al. have noted, early experience with long term retention of patients in ART programs in sub- Saharan Africa is not encouraging, suggesting, perhaps, that the current approach to preparing patients for treat- ment is not having its intended effect.47 As long as this is the case, the premises underlying ART program design should be revisited. This should be done with a view to examining whether the balance between the public health objective of a well-administered treatment program, and an individual’s entitlement to the means to preserve life and health with only the minimal necessary restrictions, is correctly proportioned. Surely, given the consequences of not receiving ART in a timely way and given the great number of those not yet receiving ART, the analysis should lead us to skew this balance in favour of those individuals whose lives would be lost unless some radical changes are made.

Rationing dilemmas generally, and the specific need for patient selection for ART in the southern African region, raise fundamental moral challenges that neither law nor ethics can yet fully address. While the global gap is so large between those in need and those who receive treat- ment, we are firmly entrenched in something that is ineq- uitable and unfair, a distance that it is impossible for a treatment team and, arguably, the national leadership of a impoverished country to even attempt to cross. As long as this is the case, as this exploratory case study has attempted to show, there may be few reasons to go searching for fair and equitable approaches at the program delivery level. Instead, more pragmatic tools are required to enable health care teams to more fully under- stand the ethical dimensions of what they do in ways that enlighten and empower them to develop creative, context-specific approaches to the difficult task they face. There may indeed be ethical problems at this level, and the experience of the treatment team at St. Charles has shown this, but these will not be fully resolved until much larger questions of fairness and equity are addressed regionally and globally.

45 D. McCoy et al. Expanding Access to Antiretroviral Therapy in Sub-Saharan Africa: Avoiding the Pitfalls and Dangers; Capitalizing on the Opportunities. Am J Public Health 2005; 95(1): 20. 46 A.A.A. Alvarez. Threshold Considerations in Fair Allocation of Health Resources: Justice Beyond Scarcity. Bioethics 2007; 21(8): 432.

47 S. Rosen, M.P. Fox & C.J. Gill. Patient Retention in Antiretroviral Therapy Programs in Sub-Saharan Africa: A Systematic Review. PLoS Med 2007; 4(10): 1–11.

Results from a Case Study of a Rural ART Program 139

STUDY LIMITATIONS

A single case study has limitations. Broad generalizations are made on the basis of only a single instance of a phenomenon. For that reason, certain findings may be overstated or overemphasized. The quality of data from program reports and treatment registers could not be independently validated making any conclusions from the chi-squared analysis tentative at best. The study is only a partial look at a much larger question. It could be that other places where ART programs are offered have resolved some of the key challenges and dilemmas brought forward in this analysis. In a single case study design there is no test for general significance of a finding. That will have to await a much larger, more comprehen- sive assessment of patient selection processes.

Acknowledgements

The author would like to thank Dr. Roland Dürig and Maria Thiess from SolidarMed and the members of the HIV treatment team at St. Charles Mission Hospital, in particular ‘Me Mafisa, HIV/AIDS Coor- dinator, ‘Me Mapoloko, ART Nurse, and ‘Me Motšephe, Hospital Administrator. The author would also like to acknowledge the guidance and support of Prof. Ames Dhai and Prof. Donna Knapp van Bogaert of the Steve Biko Centre for Bioethics, University of the Witwatersrand, Johannesburg, South Africa.

Biography

Russell Armstrong is an HIV and AIDS program development consult- ant based in Lesotho, southern Africa. He holds an MScMed in Bioet- hics and Health Law from the Steve Biko Centre for Bioethics, University of the Witwatersrand, Johannesburg, South Africa. He is currently a PhD candidate at the Centre undertaking applied ethics research in distributive justice and HIV and AIDS

140 Russell Armstrong