AIDS and Behavior

, Volume 14, Issue 4, pp 794–798

Adherence to Antiretroviral Therapy in Patients Participating in a Graduated Cost Recovery Program at an HIV Care Center in South India


  • Ashita S. Batavia
    • Weill/Cornell Medical College
  • Kavitha Balaji
    • YRG CARE
  • Elizabeth Houle
    • Department of MedicineJohns Hopkins Hospital
  • Sangeetha Parisaboina
    • YRG CARE
  • A. K. Ganesh
    • YRG CARE
    • Department of MedicineBrown Medical School/Miriam Hospital
  • Suniti Solomon
    • YRG CARE
Original Paper

DOI: 10.1007/s10461-009-9663-6

Cite this article as:
Batavia, A.S., Balaji, K., Houle, E. et al. AIDS Behav (2010) 14: 794. doi:10.1007/s10461-009-9663-6


In resource-constrained settings, the most frequently cited barrier to optimal antiretroviral therapy (ART) adherence among HIV-infected patients has been the cost of medications. In recent years many subsidized medication programs have been developed to improve ART affordability. A Graduated Cost Recovery program at the largest care center in South India has enrolled 839 eligible patients into four tiers based on an evaluation of their financial information and willingness to pay, of these patients 635 consented to participate in this study. Patients in Tier 1 receive first-line ART at no cost, whereas patients in Tiers 2, 3, and 4 pay 50, 75, and 100%, respectively of the cost of first-line medications based on an assessment of their means. Adherence rates of 95% or greater on 3-day recall were achieved by 84.6% of Tier 1 (n = 156), 71.6% of Tier 2 (n = 141), 72.3% of Tier 3 (n = 242), and 79.2% of Tier 4 (n = 96). These findings suggest patients are highly motivated and that the provision of no-cost ART can promote higher rates of optimal adherence.


Antiretroviral therapyIndiaHIVCostsAdherence


Antiretroviral therapy (ART) has been shown to dramatically improve the quality of life and health outcomes for HIV-infected patients [1, 2]. When used consistently and appropriately, ART can slow disease progression, decrease the incidence of opportunistic infections, and reduce mortality [26]. However, unlike medications for other chronic diseases, very high levels of ART adherence are necessary for the long-terms benefits of treatment to be realized [7]. A prior study found that adherence rates of 95% or greater are necessary to achieve virologic success [8]. Although it has been established that ART adherence is vital to treatment success, adherence rates have been poor in several studies from diverse clinical settings [810]. There are multiple explanations for poor adherence, but in the Developing World the most frequently cited barrier to optimal adherence has been the cost of medications [1115].

This study was conducted in India, a country that is home to almost 2.5 million HIV-infected individuals, a number that only ranks behind South Africa and Nigeria [16]. Generic first-line ART was introduced to the Indian market in 1994 at US$658 per month [17]. Since then, costs have decreased substantially and, at the time of this writing, our study site, the YRG Center for AIDS Research and Education (YRGCARE) was able to provide first-line ART at 1250–1750 INR (US$25–35) per month. At the current price, the annual costs of a HAART regimen would represent 11–16% of India’s average per capita income [17]. Recent studies in India continue to show that most common barrier to optimal ART adherence is cost despite recent decreases in costs [18, 19]. Indian patients report various strategies for coping with high medication costs. Besides taking self-imposed drug-holidays where ART is abruptly stopped and restarted based on the patient’s ability to afford medications, patients also reported borrowing money, sacrificing other necessary expenses, selling property and valuables, and accumulating debt to pay for their medications [20]. In many cases such drastic actions do not guarantee continuous ART access and the risks of frequent treatment interruptions include an increase in HIV-replication, and the development of viral mutations and drug-resistance that is challenging, not to mention more costly to manage [16, 2123].

There are many different funding sources working to provide low-cost ART in India, among them is the Global Fund to fight AIDS, Tuberculosis and Malaria (GFATM). At YRGCARE the GFATM supports a Graduated Cost Recovery (GCR) Program that provides financial assistance for ART based on a patient’s socioeconomic information. Patients in Tier 1 of this program receive their medication free of charge, whereas patients in Tiers 2, 3, and 4 are expected to contribute 50, 75, and 100%, respectively to their ART costs. Each quarter, the GCR program also supports an HIV/AIDS laboratory monitoring-tests package, physician consultation, and counseling for patients and their families [24]. This study explores the role of financial assistance for ART on patients’ self-reported adherence.


Study Setting

The YRG Centre for AIDS Research and Education (YRGCARE), a tertiary care HIV clinic based in Chennai, India that provides comprehensive care for over 10,000 HIV-infected individuals with more than 4,000 patients currently receiving antiretroviral therapy. YRGCARE patients are advised to initiate highly active antiretroviral therapy (HAART) at CD4 counts of less than 200 cells/μl, or with CD4 counts between 200 and 350 cells/μl and an AIDS-defining illness. Patients are scheduled for medical evaluation every 3 months or as clinically directed.


All 839 adult patients enrolled in the GFATM-funded GCR program at YRGCARE were eligible for this study. Eligible patients were recruited between February and May 2007 during their regularly scheduled quarterly appointments at YRGCARE. In total, the purpose of the study was explained by trained study counselors to 636 eligible patients of whom 635 provided informed consent and were able to participate in this study.

Data Collection

On providing informed consent, non-identifying demographic information and prescribed ART dosing information was obtained from chart review. Participants were then briefly interviewed by a trained study counselor at YRGCARE in their local language (Tamil, Telegu, or Hindi). During this interview participants were asked, “How many doses have you missed in the past 3-days?” to establish 3-day dose recall. This recall period was selected because participants.

Statistical Analysis

Descriptive statistics were used to describe the characteristics of the study population. The self-reported 3-day dose recall was divided by the total number of prescribed ART doses for the same period, and this value was subtracted from 100 to determine each participant’s adherence rate. Adequate adherence to medications was defined as taking 95% or greater of all prescribed doses in the 72 h prior to the study visit. All statistical analysis was performed with SPSS software (version 13.0; SPSS, Chicago, IL).


The demographic characteristics for participants are presented in Table 1. The only demographic characteristic in Table 1 that is statistically significant between the Tiers is the Percent Achieving ≥95% adherence (P-value < 0.05). Demographic characteristics independently associated with higher rates of adequate adherence included residence in Tamil Nadu State (80.1% in local residents compared to 70.4% who traveled from Andhra Pradesh, P-value = 0.05; data not shown) and education. The frequency of participants achieving adequate adherence increased with each successive level of educational achievement (see Fig. 1).
Table 1

Demographic information and adequate adherence rates


Tier 1 (N = 156)

Tier 2 (N = 141)

Tier 3 (N = 242)

Tier 4 (N = 96)

Percent male





Median age (IQR)

36 (33–41)

39 (33–45)

38 (33–43)

37 (33–44)

Percent married






 Tamil Nadu





 Andhra Pradesh


























 Higher secondary










Median monthly income (IQR)a

US$51.1 ($38.3–$74.7)

US$99.6 ($76.6–$127.7)

US$153.3 ($114.9–$204.3)

US$281.0 ($198.0–$434.2)

Percent achieving ≥95% adherence (P-value < 0.05)





aExchange rate: US$1 = Rs. 39.15
Fig. 1

Education level and adequate adherence rates

Table 1 also shows that among participants that paid for medications (Tiers 2–4), the frequency of persons that achieve adequate adherence increases with their percentage financial contribution to ART. However, the highest percentage of participants meeting the criteria for adherence was seen among those that receive their medications at no cost (Tier 1: 84.6% adherence. P-value < 0.05).


The findings in the current study suggest that the provision of free HAART promotes higher rates of adequate adherence. This is consistent with the findings of other studies where user fees have been associated with frequent treatment interruptions, delays in seeking medical care, and the compromise of family welfare [12, 25, 26]. In India, ART cost accounts for the largest component of HIV-related expenditure for patients [27], and providing free HAART removes a formidable barrier to achieving optimal adherence.

There are claims that patients should contribute to the cost of their medications to give value to treatment and thereby encourage adherence. A study examining treatment-seeking behavior among HIV-infected Indians found 94% of those not currently taking medicines wanted to be on HAART [28]. Studies outlining measures taken by patients to be able to afford treatment also lend credence to the argument that Indian patients are highly motivated to adhere to HAART [20]. The findings of the current study show an association between increasing financial contribution and rates of optimal adherence, for which there may be several valid explanations. GCR tier assignment is based on information at the time of initial interview and is not reconsidered for 1 year, unless there are exceptional circumstances such as job loss, death of breadwinner, and another HIV-positive diagnosis in the family. Patients in Tiers 2 and 3 have fewer assets and may find even the reduced medication costs to be a barrier to adherence if they are exposed to external financial pressures.

Briefly, the association between higher rates of adherence among residents of Tamil Nadu State may be explained by their relative proximity to the ART dispensary at YRGCARE. At the time of the study, discounted HAART for GCR program participants was only available for drugs purchased at YRGCARE. The association between education and adherence has been inconsistently seen in other studies [29]. Exactly how formal education impacts adherence rates is unclear. In India HIV/AIDS education is not always taught in school, this is particularly true for patient educated early in the HIV pandemic. The educational level attained by participants may be a surrogate marker for their ability to access information that reinforces the importance of adequate adherence in a personally meaningful way. Further studies are needed to clarify this hypothesis. No significant association between higher rates of optimal adherence and participant’s gender or marital status was found suggesting that drug costs and proximity to the treatment center were the major factors determining the level of adherence in this South Indian setting.

Limitations of this study include using 3-day recall to assess adherence because of its convenience and low-cost. Self-report is subject to recall bias and a desire to provide what the respondent may consider a socially acceptable answer that may be inaccurate [30]. There is no gold standard for recall period. Shorter periods have been found to be more accurate Wagner and Miller [31], but longer recall periods correlate more closely with biochemical markers and the development of resistance Simoni et al. [32]. A number of participants in this study were on daily dose schedules; therefore the use of 3-day recall was felt to provide more information without compromising accuracy. Also, other factors such as cognitive and psychological function, substance abuse, health beliefs, stigma, and social support which have been reported to affect adherence rates were not able to be analyzed independently in this study [20, 3339]. Nonetheless, these data suggest that the efforts to support the costs of medication and the placement of treatment centers in convenient locations may be most important to ensure optimal utilization of first line HAART in resource-constrained environments.


Sources of Financial Support: The authors would like to thank the Brown/Tufts Fogarty International Clinical Research and Scholars Program, D43 TW000237-14S1 and the National Institute of Health for their supporting this study.

Copyright information

© Springer Science+Business Media, LLC 2010