INT J TUBERC LUNG DIS 8(3):318–322 © 2004 IUATLD
Defaulting from anti-tuberculosis treatment in a teaching hospital in Rio de Janeiro, Brazil C. L. G. Salles,* M. B. Conde,* C. Hofer,* A. J. L. A. Cunha,† A. L. Calçada,* D. F. Menezes,* L. Sá,* A. L. Kritski* * Unidade de Pesquisa em Tuberculose-Serviço de Pneumologia-Departamento de Clínica Médica, Hospital Universitário Clementino Fraga Filho da Universidade Federal do Rio de Janeiro, Rio de Janeiro, † Departamento de Pediatria, Instituto de Pediatria e Puericultura Martagão Gesteira da Universidade Federal do Rio de Janeiro, Rio de Janeiro, Brazil SUMMARY S E T T I N G : Few studies have investigated factors associated with defaulting from anti-tuberculosis (TB) therapy in hospital settings. O B J E C T I V E : To identify the factors associated with defaulting from treatment among TB in-patients in Rio de Janeiro city, Brazil. D E S I G N : Case-control study. M E T H O D S : All study participants initiated anti-tuberculosis treatment in a teaching hospital. A defaulting case was defined as a person who did not return for antituberculosis medications after 60 days. Cases and controls were interviewed by a trained health care worker using a standardized form. R E S U L T S : From 1 January to 31 December 1997, 228 TB cases were registered. After a review of the medical records, 39 were excluded. Household visits were performed in 189 patients; 46 subjects were identified as cases
and 117 as controls. Defaulting from anti-tuberculosis treatment was observed in 66 cases (28.9%) before and in 46 (20.2%) after a home visit. After multivariate analysis, the strongest predictors of defaulting from treatment were: 1) returning card not provided (OR 0.099; 95%CI 0.008–1.2; P 0.07), 2) not feeling comfortable with a doctor (OR 0.16; 95%CI 0.33–0.015; P 0.001), and 3) blood pressure not measured (OR 0.072; 95%CI 0.036–0,79; P 0.024). C O N C L U S I O N S : In this hospital, the factors associated with defaulting from anti-tuberculosis treatment highlight the necessity for a structured TB Control Program. It is expected that the implementation of such a program, pursuing specific approaches, should enhance completion of anti-tuberculosis treatment and cure. K E Y W O R D S : tuberculosis; treatment; defaulting; HIV
TUBERCULOSIS (TB) remains a serious public health problem worldwide despite being a curable disease. The World Health Organization (WHO) estimates more than 8 million new cases and 1.87 million deaths each year due to TB.1 Brazil has almost 40 million individuals infected by Mycobacterium tuberculosis, and 90 000 new cases are notified every year.2 In Rio de Janeiro city 10 000 TB cases were notified in 1997, 33% of whom were managed in 25 hospitals that had no structured tuberculosis control activities.3 One of the most serious problems for TB control is default from treatment. In Brazil, the overall default rate ranges from 12% to 15%, while in urban areas it is from 28% to 30%.2,4 Several studies have investigated the risk factors associated with defaulting from anti-tuberculosis therapy.5–10 However, most of these
studies refer to out-patients managed in community health centers (OCHC). The purpose of the present study was to evaluate patients undergoing anti-tuberculosis therapy and to determine the rate and predictors of default in the out-patient clinic of a teaching hospital in Rio de Janeiro, Brazil.
METHODS A case-control study was conducted from 1 January to 31 December 1997 among patients with TB who initiated anti-tuberculosis treatment at the Hospital Universitario Clementino Fraga Filho (HUCFF) of Universidade Federal do Rio de Janeiro (UFRJ), a teaching hospital and reference center for acquired
Correspondence to: Cristiane L G Salles, MD, PhD, Unidade de Pesquisa em Tuberculose-Serviço de Pneumologia-Departamento de Clínica Médica, Hospital Universitário Clementino Fraga Filho da Universidade Federal do Rio de Janeiro, Rua Conselheiro Zenha 57, apt 303, 20550-090 Rio de Janeiro, Brazil. Tel: (21) 2550 6903. Fax: (21) 2550 6903. e-mail:
[email protected] or
[email protected] Article submitted 22 November 2001. Final version accepted 21 August 2003.
Defaulting from anti-tuberculosis treatment
immune-deficiency syndrome (AIDS) and TB in Rio de Janeiro city, Brazil. Patients were identified from the registration card at the hospital pharmacy, where anti-tuberculosis drugs were issued to patients. A patient who initiated anti-tuberculosis treatment but failed to return to the health center for 60 days or more was defined as a default case and included in the study. The data on default cases were compared with the data on TB patients who finished anti-tuberculosis treatment within the recommended period. The standard treatment regimen consists of self-administered daily isoniazid (H), rifampin (R), and pyrazinamide (Z) for 2 months, followed by daily isoniazid and rifampin for 4 months (2HRZ/4HR). The study was approved by the Committee for the Protection of Human Subjects (CPHS) of the Medical School of the Federal University of Rio de Janeiro. Informed written consent was obtained from all participants prior to the study. After reviewing all those cases registered for antituberculosis therapy, a standardized form was filled out by trained health care workers (HCWs) based on the patient’s medical records. Household visits were made to non-adherent and cured cases by a trained nurse. Cases who moved, transferred to another health care center, died, had a change in diagnosis or did not provide written informed consent to take part in the study were excluded. Patients underwent a standardized interview to collect information on the following variables: demographic data, education status, working conditions, family environment, alcoholism, smoking habits, use of illegal drugs, human immunodeficiency virus (HIV) infection, family monthly income, use of concomitant medication, alterations in behavior, comorbidities, doctor-patient relationship and activities during the medical examination (blood pressure gauge, physical examinations, weight, appointment card to return to hospital, waiting time at appointments, absenteeism from consultation and awareness of the disease). A case of AIDS was considered in the presence of OPAS/Caracas or Centers for Disease Control criteria, according to the case.11,12 Economic status was evaluated by the total household income. To assess further alcohol consumption practices, a previously tested and validated series of four simple questions about specific habits or symptoms associated with heavy drinking, described as the CAGE system, was used.13–15 The health service was evaluated by whether or not the patient was weighed and whether blood was taken during the physical examination. The patient relationship with HCWs (nurse and physician) was evaluated by the level of satisfaction with the health assistance and personal attention provided during consultations, as well as the time spent waiting during an appointment. Patients were also evaluated in their awareness and knowledge about their willingness to be treated.
319
Statistical methods The Student’s t-test was used to assess normally distributed continuous variables, and the Wilcoxon rank sum test was used to assess non-parametric data. Cases and matched control patients were evaluated for differences in their characteristics and responses using the Mantel-Haenszel 2 test for proportions. Univariate matched odds ratios (OR) with 95% confidence intervals (CI) were calculated to test associations with defaulting from treatment. Conditional logistic regression was performed to determine which variables were independently associated with defaulting from treatment. A forward stepwise procedure was used to enter variables that were significant on univariate analysis (P 0.20) into a multivariate model. Data were entered and analyzed using the statistical software packages Epi Info 6.0 (CDC, Atlanta, GA, USA) and Stata® (version 5.0, Stata Corporation, College Station, TX, USA).
RESULTS From 1 January to 31 December 1997, 228 TB cases were registered. After a review of the medical records, 189 patients were identified to receive household visits. The flow chart and the distribution of cases is shown in the Figure. Thirty-nine records were excluded because cases had moved (29 cases), transferred to a different health unit (two cases), had a change in diagnosis (two cases) or died (six cases). After the household visit, 123 (65%; 123/189) were classified as having received their anti-tuberculosis treatment according to record review, 117 (95%; 117/123) were confirmed as finishing treatment, one (0.8%; 1/123) was transferred to another health unit and five (4%; 5/123) had defaulted from treatment. Among the 66 cases classified as defaulting from treatment by medical record review, 41 (62.1%; 41/66) continued to default after the house-
Figure
Flow chart of cases. TB tuberculosis.
320
The International Journal of Tuberculosis and Lung Disease
Table 1 Social and demographic data among 163 patients evaluated Variables Sex Male Female Race* White Non-white HIV positive† Yes No Alcoholism Yes No Monthly earnings US$250 Yes No Unemployed‡ Yes No
Default Cure (n 46) (n 117)
OR (95%CI)
P
31 15
68 49
1.99 (0.68–3.28)
0.27
24 22
62 54
1.05 (0.5–2.2)
0.88
16 10
32 41
2.05 (0.75–5.67)
0.18
5 17
16 48
0.88 (0.24–3.15)
0.83
19 3
45 21
2.9 (1.09–8.47)
0.03
27 11
58 39
1.65 (0.68–4.06)
0.22
* 1 case without information. † 64 cases without information. ‡ 28 cases without information. OR odds ratio; CI confidence interval; HIV human immunodeficiency virus.
hold visit, while 15 (22.7%; 15/66) died, nine (13.6%; 9/66) were transferred and one (1.5%; 1/66) was still on treatment. The social and demographic data of 163 TB patients are presented in Table 1. Table 2 shows the health care workers’ activities related to anti-tuberculosis therapy. Univariate analysis showed that patients who completed treatment differed from those patients who Table 2 Health care worker (nurse and physician) activities during tuberculosis treatment Variables Blood pressure evaluation* Yes No Lack of weighing* Yes No Lack of abdominal examination* Yes No Returning card given† Yes No Long wait at an appointment‡ Yes No
Default Cure (n 46) (n 117)
OR (95%CI)
P
28 18
100 16
4.02 (1.69–9.59)
0.099
26 20
40 76
2.47 (1.16–5.27)
0.01
34 12
15 101
2.38 (0.93–6.04)
0.04
3 19
1 65
7 15
3 61
* 1 case without information. † 75 cases without information. ‡ 77 cases without information. OR odds ratio; CI confidence interval.
10.26 (0.86–272.3) 0.04
9.49 (1.83–61.41) 0.0006
defaulted from treatment in their monthly earnings (US$250, P 0.02, Table 1). Patients who reported they were not weighed (P 0.01), did not have their blood pressure taken (P 0.0003) or have an abdominal examination (P 0.04) were significantly more likely to have defaulted from treatment (Table 2). Not having an appointment card for returning to the hospital (P 0.04) and reporting a long wait at an appointment (P 0.002) were mainly mentioned by subjects who did not complete treatment. No difference was detected between the two groups as regards alcoholism, drug use or HIV seropositivity. After multivariate analysis, lack of an appointment card (OR 0.099; 95%CI 0.008–1.2; P 0.07), lack of good relation with the doctor (OR 0.16; 95%CI 0.33–0.015; P 0.001) and lack of blood pressure measurement (OR 0.072; 95%CI 0.036–0.79; P 0.024) were significantly associated with defaulting from treatment.
DISCUSSION Failure to complete the treatment regimen has been historically cited as one of the most challenging problems in tuberculosis treatment.16 Factors usually associated with defaulting from treatment are related to the complexity, duration, adverse effects, cost and access to health services.17 Age, sex, race and education level are not usually identified as predictors of failing anti-tuberculosis treatment.9 The rate of default from treatment in our study was 20.2%. The rate of patients with active TB who abandoned treatment ranges from 8% to 25%.5 In Brazil, the treatment default rate among patients managed at out-patient health units ranged from 15% to 28% in different settings.2,4,18 There are no data regarding this rate among patients cared for in general hospitals. In the present study, the factors associated with defaulting from treatment were mainly related to health service activities such as the absence of physical examination or long waiting time for a medical consultation. These findings are quite different from those mentioned by other reports, which described alcoholism, homelessness, use of illegal drugs and HIV infection as the most important risk factors associated with treatment default.5,11,12 In Rio de Janeiro City, in the same period, another study was performed among patients managed in seven out-patient health care units, in which alcoholism and illicit drug abuse were identified as the most significant factors associated with treatment default.19 TB patients who seek care, diagnosis and treatment in a teaching hospital may differ from those who seek care and are managed in community health centers; hence, our results and conclusions may not be applicable to all TB patients in Rio de Janeiro City. However, this study suggests that household visits to con-
Defaulting from anti-tuberculosis treatment
firm anti-tuberculosis treatment results can reduce the rate of default: the initial default rate of 28.9% fell to 20.2% after household visits. Future research on TB patients treated in general hospitals in Rio de Janeiro City should focus on developing strategies that will enhance completion of anti-tuberculosis treatment and improve cure among those patients with different co-morbidities. Acknowledgements This work was supported by an NIH Fogarty International Training Grant (5 D43TW00018), Cornell University, NY, USA. The authors thank Dr Warren D Johnson, Jr, for his generous support.
References 1 Dye C, Scheele S, Dolin P, Pathania V, Raviglione M. Global burden of tuberculosis. Estimated incidence, prevalence, and mortality by country. JAMA 1999; 282: 677–686. 2 Ruffino Netto A. Controle da tuberculose no Brasil. Atividades implementadas em 1999. Bol Pneum San 1999; 7: 58–66. 3 [SMS-RJ] Secretaria Municipal de Saúde do Rio de Janeiro. Saúde em Foco. Informe especial sobre os números produzidos pela Secretaria Municipal de Saúde (Ano VII). Rio de Janeiro, Brazil: Secretaria Municipal de Sandé, 1998; 19: 7. 4 Campos H S, Melo F A F. Efetividade do esquema 3 (3SZEEt/ 9Eet) no retratamento da tuberculose na rotina das Unidades de Saúde. Bol Pneum San 2000; 8: 7–14. 5 Addington W W. Patient compliance: the most serious remaining problem in the control of tuberculosis in the United States. Chest 1979; 76: 741–743. 6 Andersen S, Banerji D. A sociological inquiry into an urban tuberculosis control programme in India. Bull World Health Organ 1963; 29: 685–700. 7 Beltran O R P, Mosca C A, Eisele C, Carosso A. El abandono del tratamiento en tuberculosis. Realidad actual e perspectivas
8
9
10 11
12
13
14
15
16
17 18
19
321
futuras. Rev Argentina de Tuberculosis, Enfermedades Pulmonares y Salud Publica 1983; XLVI: 11–19. Fox W. Compliance of patients and physicians: experiences and lessons from tuberculosis. II. British Med J 1983; 287: 101–105. Sumartorjo E. When tuberculosis treatment fails: a social behavioral account of patient adherence. Am Rev Respir Dis 1993; 147: 1311–1320. Sbarbaro J A. Public health aspects of tuberculosis: supervision of therapy. Clin Chest Med 1985; 1: 253–263. Ferrer X, Kirschbaun A, Toro J. Compliance with tuberculosis treatment in adults in Santiago, Chile. Boletin de la Oficina Sanitaria Panamericana 1991; 11: 423–431. Crespo R, Valdes L, Bobe I, Cayla J. Difficulties in carrying out the treatment as seen by the tuberculous patient. Atención Primaria 1992; 9: 260–262. Centers for Disease Control and Prevention. Revised classification system for HIV infection and expanded surveillance case definition for AIDS among adolescents and adults. Morbid Mortal Weekly Rep 1992: 41(RR-17): 1–18. Working Group on AIDS Case Definition. Pan American Health Organization. Caracas, Venezuela: PAHO, February 20–22, 1989. Mayfield M, McCleod G, Hall P. The CAGE questionnaire: validation of a new alcoholism screening instrument. Am J Psychiatry 1974; 131: 1121–1123. Bush B, Shaw S, Cleary P, Delbanco T, Aronson M. Screening for alcohol abuse using the CAGE questionnaire. Am J Med 1987: 82: 231–235. Ewing J. Detecting alcoholism: the CAGE questionnaire. JAMA 1984; 252: 1905–1907. Diniz L S, Gerhardt G, Miranda J Á, Manceau J. Efetividade do tratamento da tuberculose em oito municípios de capitais brasileiras. Bol Pneum San 1995; 3: 6–18. DeRiemer K, Brito Z, Lauria L, Kristki A L, Reingold. A. The association between illicit drug use and treatment defeaulting among TB patients in Rio de Janeiro, Brazil. Int J Tuberc Lung Dis 1998; 2 (Suppl 2): S349.
RÉSUMÉ C A D R E : Peu d’études ont été consacrées aux facteurs associés à la défaillance à l’égard du traitement antituberculeux dans un contexte hospitalier. O B J E C T I F : Identifier les facteurs associés à la défaillance à l’égard du traitement parmi les patients tuberculeux (TB) hospitalisés à Rio de Janeiro-ville, Brésil. S C H É M A : Etude cas contrôle. M É T H O D E S : Tous les participants à l’étude ont commencé un traitement anti-tuberculose dans un hôpital d’enseignement. On a défini comme cas de défaillance une personne qui n’est pas revenue rechercher ses médicaments antituberculeux après 60 jours. Les cas et les contrôles ont été interviewés par un travailleur de soins de santé entraîné en utilisant un formulaire standardisé. R E S U L T A T S : Entre le 1er janvier et le 31 décembre 1997, 228 cas de TB ont été enregistrés. Après révision des dossiers médicaux, 39 en ont été exclus. Les visites au domicile ont été assurées chez 189 patients ; 46 sujets ont été identifiés comme cas et 117 comme contrôles. On a
observé une défaillance à l’égard du traitement antituberculose dans 66 cas (28,9%) avant les visites domiciliaires et dans 46 cas (20,2%) ultérieurement. Après une analyse multivariée, les prédicteurs les plus puissants de la défaillance à l’égard du traitement anti-TB ont été : 1) l’absence de fourniture d’une fiche pour retour (OR 0,099 ; IC95% 0,008–1,2 ; P 0,07), 2) un sentiment de manque de confiance à l’égard du médecin (OR 0,16 ; IC95% 0,33–0,015 ; P 0,001) et 3) le fait que la pression sanguine n’ait pas été mesurée (OR 0,072 ; IC95% 0,036–0,79 ; P 0,024). C O N C L U S I O N S : Dans cet hôpital, les facteurs associés à la défaillance à l’égard du traitement anti-tuberculose font ressortir la nécessité d’un programme structuré de lutte antituberculeuse. On s’attend à ce que la mise en œuvre d’un tel programme poursuivant des approches spécifiques améliore l’achèvement du traitement antituberculose et les guérisons.
322
The International Journal of Tuberculosis and Lung Disease
RESUMEN M A R C O D E R E F E R E N C I A : Pocos estudios han investigado los factores asociados con el abandono del tratamiento antituberculoso en el contexto hospitalario. O B J E T I V O : Identificar los factores asociados con el abandono del tratamiento en pacientes TB, en la ciudad de Río de Janeiro, Brasil. D I S E Ñ O : Estudio casos-control. M É T O D O : Todos los participantes en el estudio iniciaron un tratamiento antituberculoso en un hospital docente. Un caso de abandono fue definido como una persona que no vuelve por sus medicamentos para la TB después de 60 días. Los casos y los controles fueron entrevistados por un trabajador de la salud entrenado, utilizando un formulario estandarizado. R E S U L T A D O S : Del 1 de enero al 31 de diciembre de 1997, se registraron 228 casos de TB. Después de la revisión de las fichas médicas, 39 casos fueron excluidos. Se realizaron visitas a domicilio para 189 pacientes y se identificaron 46 sujetos como casos y 117 como con-
troles. Se observó un abandono del tratamiento antituberculoso en 66 casos (28,9%) antes de las visitas a domicilio y en 46 (20,2) después de ellas. Después del análisis multivariado, los factores de predicción más potentes para el abandono fueron : 1) el hecho de no haber entregado al paciente una tarjeta de retorno (OR 0,099 ; IC95% 0,008–1,2 ; P 0,07), 2) sentimiento de desconfianza con respecto al médico (OR 0,16 ; IC95% 0,33–0,015 ; P 0,001) y 3) el hecho que la presión arterial no fue tomada (OR 0,072 ; IC95% 0,036–0,79 ; P 0,024). C O N C L U S I Ó N : En este hospital, los factores asociados con el abandono del tratamiento antituberculoso ponen en evidencia la necesidad de un programa estructurado de control de la TB. Se espera que la implementación de tal programa, que persiga objetivos específicos, va a lograr un mejor cumplimiento del tratamiento y mejores tasas de curación.
