By the year 2000, it is projected that AIDS will be the leading cause of death in Thailand, with 2–4 million HIV-infected individuals and 650 000 AIDS cases . Several studies have shown that HIV infection increases the annual risk of developing clinical tuberculosis 6–26-fold [2–3]. Since infection with Mycobacterium tuberculosis is prevalent in Thailand, the HIV/AIDS epidemic will adversely affect tuberculosis control [4–5].
Isoniazid preventive therapy for tuberculosis has been shown to be effective in reducing the risk of developing clinical tuberculosis among tuberculin-positive asymptomatic HIV-infected individuals . This fact prompted health professionals to consider providing tuberculosis preventive therapy to asymptomatic HIV-positive persons.
It is widely recognized, however, that poor adherence or non-adherence to therapy is the major obstacle to tuberculosis control [7–10]. This may also be true for administering tuberculosis preventive therapy to asymptomatic HIV-infected persons [11–13]. Therefore, tuberculosis preventive therapy should only be considered in areas that have a good functioning tuberculosis programme . Few data are available, however, on the adherence of HIV-infected persons to the preventive therapy. This study was conducted to determine the level of and reasons associated with adherence to the preventive therapy among asymptomatic HIV-infected persons in a northern Thai province.
Subjects and methods
Until the HIV/AIDS era, Thailand enjoyed a steady decline in tuberculosis prevalence . Chiang Rai, where the study was conducted, is the northernmost province of Thailand and is one of the epicentres of the HIV/AIDS epidemic. Tuberculosis is re-emerging there as well. Data from the provincial hospital showed a marked increased in HIV prevalence rate among new tuberculosis patients from 1.5% in 1990 to 45.5% in 1994. The number of new tuberculosis cases also increased during the same time period.
Since November 1993, in response to the dual epidemic, the provincial hospital has commenced the tuberculosis preventive therapy programme for asymptomatic HIV-infected individuals. A special clinic at the hospital, staffed by five nurses and two clerks, provided this service every workday from 8:00 a.m.–4:00 p.m. One nurse and one clerk worked each day, under supervision of a clinician. Participants received isoniazid, from the nurse, free of charge.
The programme sequentially recruited participants from four sources: (i) blood donors who had indicated a desire to know their HIV test result; (ii) persons attending the hospital's anonymous counselling and testing clinic; (iii) patients attending the hospital's outpatient clinics for other illnesses but found to be HIV-infected, and (iv) female commercial sex workers enrolled in a cohort study being conducted in Chiang Rai.
Potential participants received an explanation about the programme, were invited to join the study and then asked to provide written consent. They were then screened with a questionnaire on physical illnesses, physically examined, and tested for tuberculin skin reactivity using the Mantoux procedure. Skin test reading was done 2–5 days after testing. Chest radiograph was performed in every case. Sputum examination for acid-fast bacilli (AFB) was performed on participants with an abnormal chest radiogram.
Eligible participants were provided with daily 300 mg dose of isoniazid, and one tablet of vitamin B complex for 9 months. The participants were required to collect their supply of drugs on a monthly basis. At each visit, participants were instructed to bring leftover medicine with them so that programme staff could count the pills. This protocol was approved by the Research Ethics Committee of the Ministry of Public Health of Thailand.
Until August 1994, 463 persons (219 blood donors, 180 outpatients, 37 female commercial sex workers, and 27 anonymous clinic clients) were referred for preventive therapy. Among these four groups, 184 (84.0%), 166 (92.2%), 35 (94.6%) and 27 (100%), were enrolled in the programme, respectively. Although the enrolment rate was slightly higher in females (130 out of 139 or 93.5%) than in males (282 out of 324 or 87.0%, P = 0.04), enrolment population did not differ in terms of age, marital status, educational level, occupation, location of residence and history of physical symptoms (data not shown). The majority (39 out of 51 or 76.5%) of those who did not enrol did not come for a skin test reading. Twelve people did not enrol for various reasons: five already had AIDS, five were unwilling to take any medicine, one planned to move out of the province and one person was referred elsewhere for treatment of latent syphilis.
The study was composed of three parts:
The completion rate and level of adherence for the 412 enrolled individuals were determined. Completion rate was defined as the proportion of the participants who successfully completed the 9-month preventive therapy without defaulting. Default criteria was failure to take medication for more than 60 consecutive days during the 9-month period. Level of adherence was measured by pill count. The level of adherence was good if more than 80% (216 pills) were taken during the 9-month period .
Data were double-entered into a dBASE data file and analysed with Epi-Info version 5.01b (Centers for Disease Control and Prevention, Atlanta, Georgia, USA) and EGRET version 0.26.61 (Statistics and Epidemiology Research Corporation, Seattle, Washington, USA). Both exploratory data analysis and logistic analysis were carried out. Doses of isoniazid pills taken were used as the outcome variable and were dichotomized using 216 pills as the cut-off in the logistic equation. Odds ratios (OR) and associated 95% confidence intervals (95% CI) were used as measures of strength of association between the outcome variable and exposures.
Participants who had missed a scheduled appointment for more than a month were identified. If they subsequently returned to the programme, they were interviewed on the date of return to determine the reasons for non-adherence. Interviews were held in a private counselling room by a social scientist, one of the authors (J.N.). With verbal informed consent from the participants, notes of the interview were recorded by the interviewer. At the end of the interview session, the recording was verified by the participants. Content analysis was later done by the interviewer.
Small focus group discussion among those who had successfully completed the therapy was held to determine motivation and methods used in order to adhere to the programme. We grouped participants achieving a 95–100% adherence rate by homogeneous attributes based on sex, marital status, education and occupation. Five group discussions were organized: one group of single males; two groups of married males; one group of married females; and one group of female widows. For each group, about six to 12 participants were invited but only four to eight could join the discussion. The others did not participate because they were too busy with their work.
Each focus group discussion was moderated by the same social scientist (J.N.). Discussions were held in a private room inside the hospital without the presence of health workers. With permission from all of the participants, the discussions were recorded. Discussion contents were later transcribed and analysis was carried out by the moderator of the discussions.
Part I: Level of adherence and factors associated with adherence
Of the 412 participants enrolled in the programme, 286 (69.4%) completed the 9-month preventive therapy, 109 (26.5%) defaulted, and 17 (4.1%) developed clinical AIDS (including tuberculosis) or died before either completing the programme or defaulting. Five tuberculosis cases, four cases of pulmonary tuberculosis and one of tuberculosis meningitis were reported. Of the 109 defaulters, 43 (39.4%) returned to restart the therapy and 20 subsequently succeeded. Completion rate including these participants rose to 74.3% (306/412). Figure 1 shows the follow-up status of the participants at different time points. It can be observed that default is most frequent in the early months after beginning treatment.
Table 1 shows dosage of isoniazid taken during 9 months after the initiation of treatment. Based on the 216 pill cut-off, adherence rate was 67.5% (278 out of 412 cases). Fourteen participants completed the 9-month therapy but took fewer than 80% of the pills prescribed. Two defaulters and four cases who had AIDS or died were considered to be adherent because they were lost to follow-up or became ill or died after taking more than 216 pills.
Table 2 compares demographic characteristics of adherents and non-adherents. Age, educational level, and distance from home to hospital were not associated with adherence. Females were more likely to adhere to the therapy than males, especially when commercial sex workers were excluded from the analysis. Married people, survival spouses and self-employed people were more likely to be adherent. Outpatients showed higher adherence. Female commercial sex workers were less adherent than blood donors. Participants who had a history of physical symptoms showed less adherence than those without past symptoms.
Table 3 shows adjusted OR of factors predicting adherence based on a logistic regression model. After adjustment for sex and recruitment source, those who had a history of physical symptoms were twice as likely to be non-adherent to preventive therapy than those without symptoms.
Part II: Reasons for non-adherence
Of the total 412 participants, 109 defaulted. Most of the participants wanted to keep their HIV status confidential; they did not allow nurses to visit their home. Because of this, defaulters were tracked only by mail. Fifty defaulters returned after receiving reminder letters, and were interviewed for their reasons. Among these, 43 restarted the therapy while seven declined. A different group of 22 participants, who missed appointments for more than 1 month but had not yet been classified as defaulting, were also interviewed for their reasons. The interview results of these 72 participants are summarized in Table 4.
Out-migration for job search in other provinces is the predominant reported reason for non-adherence. Other significant reasons include denial of HIV-infection status, perceived side effects of isoniazid, and misunderstanding about duration of the preventive therapy. Of the seven defaulters who permanently withdrew from the programme, four denied their HIV status, and three perceived side effects of isoniazid.
Part III: Reasons for adherence
A total of 28 participants who demonstrated good adherence were included in one of the five focus group discussions. Despite having completed the 9-month programme, about a quarter of the participants still did not know about the effect of isoniazid in preventing clinical tuberculosis. Some of them thought that isoniazid reduced blood HIV concentration or that it prevented other AIDS-related complications. Many of them said that they took the medication because it was recommended by doctors or nurses. Some said that they were concerned about their children and family, and these concerns motivated them to prolong their life including the taking of isoniazid. Most of these participants accepted their HIV status well, and they were satisfied with the service and with the providers. Some said they were encouraged by the providers to complete the medicine and to fight the HIV infection.
Without staff instruction, several participants developed their own medicine reminder systems, including storing medication in a visible location, marking a calendar after taking medication, and putting a reminder on a cosmetic desk.
This 9-month programme demonstrated a slightly higher adherence rate than the 62% reported from a 6-month isoniazid regimen study in Uganda . Although this suggests that good adherence is possible, careful attention must be paid to improving treatment adherence. There is little value in increasing the number of persons beginning preventive therapy, if they will not complete it . At least three major issues must be considered to increase adherence: participant selection, enrolment and follow-up.
Our study implies that people should be selected from appropriate groups for preventive therapy. Females were more likely to adhere than males, as shown in other studies [8–9]. Married women and widows were more motivated to adhere to treatment regimens because they wanted to prolong their ability to take care of their children. Female commercial sex workers were likely to be non-adherent mostly because they tend to move to work in other provinces [K. Limpakarnjanarat, personal communication, 1995].
Before enrolment, participants should be evaluated for potential job migration, acceptance of their HIV status, and understanding of treatment duration and possible side effects of isoniazid. Other studies in Thailand [18–20] show that job migration is an important reason for non-adherence. As Ferebee  pointed out, for the poor, problems of daily existence are often so acute and immediate that little time or energy is left to be concerned about preventive treatment for a disease they only might have in the future. To minimize nonadherence due to migration, at the time of enrolment, migration potential should be assessed, adequate contact information about the programme should be provided to the participants, and arrangement should be made so that those who migrate can collect their supply of drugs in advance.
The popular image of AIDS is influenced by the media, which sometimes creates fear and anxiety that results in stigmatization of infected persons [22–23]. Therefore, it is possible that those who did not return for skin test reading and the defaulters denied their HIV status. Some follow-up losses may also be due to death, such as suicide. It was found that the suicide rate among persons with AIDS is 21–36-fold higher than that of the general population . This highlights the importance of counselling prior to enrolment. In addition, since the treatment is preventive, not curative for acute diseases, and a significant proportion of the defaulting occurred soon after initiation of the treatment, more time should be allowed to prepare the participants.
Some studies have suggested that home visits improved adherence [11,25]. Therefore, if possible, home visits should be used to follow-up the defaulters. The major limitation on drawing conclusions from this study and the programme is that the reasons for default in 59 cases are unknown. Visits to these participants would be invaluable for proper study closure.
Participants who developed an acute illness might take only drugs that treated the illness and might ignore their treatment with isoniazid. In addition, prescribing increased numbers of drugs is associated with nonadherence [26–28]. A good relationship between healthcare providers and participants, however, can enhance adherence, as shown in this and other studies [29–34]. Participants who believe in healthcare providers tend to comply more than those who do not. As Donovan and Blake  indicated, however, patients are not ‘blank sheets’. They have many beliefs and disease information which may not coincide with those of medical profession, and they carry out their own cost-benefit analyses for each treatment they are offered. Hence, providers should recognize and understand these individual differences, and treatment should be offered based on the individual's needs and in context with their concerns.
Although this study has demonstrated that a programme of tuberculosis preventive therapy for HIV-seropositive persons can be implemented in a developing country, further adjustments of the programme based on these findings are necessary to improve patient adherence. The prerequisites for considering preventive therapy as a public health measure in developing countries should include: close collaboration between the National AIDS and the Tuberculosis Control Programme, achievement of the World Health Organization's target for tuberculosis control , and administration of preventive therapy in a supervised fashion .
We are grateful to R. Srismith, the Director of the Chiang Rai Hospital, and staff nurses of Department of Preventive and Social Medicine, Chiang Rai Hospital. We thank the Research Institute of Tuberculosis, Japan, and the Japanese Foundation for AIDS Prevention for their support. We thank K. De Cock, R. Chaisson, V. Diwan, C. Sitthi-amorn, A. Winkvist, T. Mastro, C. Murray, and other experts for their review and comments on the preliminary manuscript.
1. AIDS Division, Ministry of Public Health Thailand
: HIV/AIDS situation in Thailand
. HIV/AIDS Update
2. Selwyn PA, Hartel D, Lewis VA, et al.
: A prospective study of the risk of tuberculosis among intravenous drug users with human immunodeficiency virus infection
. N Engl J Med
3. Braun MM, Badi N, Ryder RW, et al.
: A retrospective cohort study of the risk of tuberculosis among women of childbearing age with HIV infection in Zaire
. Am Rev Respir Dis
4. Spinacci S: Guidelines for tuberculosis treatment in adult and children in national tuberculosis programmes
. Thai J Tuberc Chest Dis
5. Xavier RJ: AIDS and tuberculosis: a dangerous synergy
. In AIDS in the World
. Edited by Mann JM, Tarantola DJM, Netter TW. Boston: Harvard University Press; 1992:148–163.
6. Pape WJ, Jean SS, Ho JL, et al.
: Effect of isoniazid prophylaxis on incidence of active tuberculosis and progression of HIV infection
7. Cuneo WD, Snider DE: Enhancing patient compliance with tuberculosis therapy
. Clin Chest Med
8. Van Der Werf TS, Dade GK, Van Der Mark TW: Patient compliance with tuberculosis treatment in Ghana: factors influencing adherence to therapy in rural service programme
9. Chuah SY: Factors associated with poor patient compliance with antituberculosis therapy in Northwest Perak, Malaysia
10. Sumartojo E: When tuberculosis treatment fails: a social behavioral account of patient adherence
. Am Rev Respir Dis
11. Aisu T, Raviglione MC, van Praag E, et al.
: Preventive chemotherapy for HIV-associated tuberculosis in Uganda: an operational assessment at a voluntary counselling and testing centre
12. Twumasi P: Preventive therapy for tuberculosis [letter]
13. Miller B: Preventive therapy for tuberculosis
14. International Union Against Tuberculosis and Lung Disease, Global Programme on AIDS, World Health Organization Tuberculosis Programme: Tuberculosis preventive therapy in HIV-infected individuals. A joint statement of the International Union Against Tuberculosis and Lung Disease (IUATLD) and the Global Programme on AIDS and the Tuberculosis Programme of the World Health Organization (WHO)
. Tubercle Lung Dis
15. Payananda V, Kladphuang B, Talkitkul N, et al.
: Information in Preparation for an External Review of the National Tuberculosis Programme, Thailand
1995. Bangkok: Tuberculosis Division, Department of Communicable Disease Control, Ministry of Public Health.
16. Tuberculosis Control Programme, World Health Organization: Feasibility of Isoniazid Preventive Therapy of Tuberculosis in HIV-infected Persons in Developing Countries: A Research Protocol. Geneva: WHO; 1994 p14 (mimeograph).
17. Wobeser W, To T, Hoeppner VH: The outcome of chemoprophylaxis on tuberculosis prevention in the Canadian Plains Indians
. Clin Invest Med
18. Kuttawat K: Factors affecting the regularity of pulmonary tuberculosis patient treatment
. Masters Degree Thesis in Medical and Health Social Science, Mahidol University
, 1984, p49.
19. Kamutthamart P: The study on problems of tuberculosis patients who had irregular treatment, and the solution
. Thai J Tuberc Chest Dis
1987, Apr-June: 77–80.
20. Ngamvithayapong J, Sarapirom S, Yanai H: A focused ethnographic study of tuberculosis in Ubonratchathani Province
. Com Dis J
21. Ferebee SH: Controlled chemoprophylaxis trials in tuberculosis. A general review
. Adv Tuberc Res
22. Helman C: Culture, Health and Illness (2nd edn). Stoneham: Butterworth Heinemann; 1990:100.
23. Ngamvithayapong J, Uthaivoravit W, Sawanpanyalert P, Yanai H, Ngamvithayapong A: Participatory development of counselling media for asymptomatic HIV-infected persons
. Thai AIDS J
24. Weitz R: Life with AIDS. London: Rutgers University Press; 1991:73–74.
25. Sanmarti LS, Megias JA, Gomez MNA, et al.
: Evaluation of the efficacy of health education on the compliance with anti-tuberculosis chemoprophylaxis in school children: a randomized clinical trial
. Tubercle Lung Dis
26. Homedes N, Ugalde A: Patient's compliance with medical treatments in the third world. What do we know
? Health Policy Plann
27. Brenner E, Pozsik C: Cases holding
. In Tuberculosis: A Comprehensive International Approach
. Edited by Reichman LB, Hershfield ES. New York: Marcel Dekker; 1993:183–205.
28. Ngamvithayapong J, Akarasewi P, Rawangphan N, Yanai H, Wongkhomthong S: Confronting tuberculosis problems in AIDS era: the perspectives from various stakeholders
. Third International Conference on AIDS in Asia and the Pacific
. Chiang Mai, September 1995 [abstract FP9C2].
29. Dudley DL: Why patients don't take pills
30. Morisky DE, Malotte CK, Choi P, et al.
: A patient education program to improve adherence rate with antituberculosis drug regimen
. Health Educ Quart
31. Barnwell M, Chitkara R, Lamberta F: Tuberculosis prevention project
. J Natl Med Assoc
32. Barnhoorn F, Adriaanse H: In search for factors responsible for noncompliance among tuberculosis patients in Wardha district, India
. Soc Sci Med
33. Grange JM, Festenstein F: The human dimension of tuberculosis control
. Tubercle Lung Dis
34. Pozsik CJ: Compliance with tuberculosis therapy
. Med Clin N Amer
35. Donovan JL, Blake DR: Patient non-compliance: deviance or reasoned decision-making
? Soc Sci Med
36. De Cock KM, Grant A, Porter JDH: Preventive therapy for tuberculosis in HIV-infected persons: international recommendations, research, and practice