INT J TUBERC LUNG DIS 15(12):1637–1642 © 2011 The Union http://dx.doi.org/10.5588/ijtld.10.0558

Adherence to isoniazid preventive therapy in children exposed to : a prospective study from Guinea-Bissau

V. F. Gomes,* C. Wejse,*† I. Oliveira,* A. Andersen,* F. J. Vieira,‡ L. J. Carlos,§ C. S. Vieira,‡ P. Aaby,* P. Gustafson¶ * Bandim Health Project, Indepth Network, Bissau, Guinea-Bissau; † Department of Infectious Diseases, Aarhus University Hospital, Aarhus, Denmark; ‡ National Tuberculosis Referral Hospital (Hospital de Pneumologia Raoul Follereau), Bissau, § Hospital Nacional Simao Mendes, Bissau, Guinea-Bissau; ¶ Infectious Diseases Research Group, Department of Clinical Sciences, Lund University, Malmö, Sweden

SUMMARY

OBJECTIVE: To assess adherence to isoniazid preven- fied, 1895 (72%) were evaluated for eligibility for IPT, and tive therapy (IPT) in children exposed to adult pulmo- 820 were enrolled in the study: 609 were aged ⩽5 years nary tuberculosis (TB) at home. and 211 aged 5–15 years. A total of 79% of the pre- METHODS: Children were enrolled on IPT if they were scribed doses were taken, with 65% of the children tak- aged ⩽5 years or 5–15 years and presented a tuberculin ing >80% of their doses. In all, 51% completed more skin test induration of ⩾10 mm. Children were included than 6 consecutive months of IPT. from the demographic surveillance system of the Bandim CONCLUSION: Overall adherence to IPT was better Health Project in Bissau, Guinea-Bissau. The main out- than previously reported from TB-endemic areas, with come measures were adherence, completion rates and 76% of the children completing at least 6 months of side effects during 9 months of IPT. The main outcome treatment, with more than 80% adherence. was 6 consecutive months of at least 80% adherence. KEY WORDS: latent TB; isoniazid preventive therapy; RESULTS: A total of 2631 children were identified as TB exposure; adherence; Guinea-Bissau contacts of adult TB cases. Among the children identi-

ISONIAZID (INH) was introduced as tuberculosis apy (IPT) trials in the United States showed an opti- (TB) chemoprophylaxis nearly 50 years ago.1 Clini- mal duration of treatment of around 9 months.8 The cal trials have shown the ability of INH to prevent American Thoracic Society, the International Union the progression of latent TB (LTBI) to active Against Tuberculosis and Lung Disease (The Union) TB.2 In children with a positive tuberculin skin test and the World Health Organization have recom- (TST), INH reduced TB morbidity by 94% during mended 6–9 months INH prophylaxis in children the medication year, and by 70% over the subse- aged <5 years exposed to adults with smear-positive quent 9-year period.3,4 In the absence of re-infection, TB.2,7,9 The International Standards for TB care have which occurs with far greater frequency in TB-endemic made the same recommendation.10 TST and chest areas, protection appears to be long-lasting, present radiograph (CXR) are no longer regarded as pre- for up to 20 years after initiation of treatment.5 requisite screening tests in settings where these tests The optimal duration of INH treatment for LTBI are not readily available.11 has been the subject of extensive debate in the past Studies have shown that 60–80% of children with 20 years.6 Clinical trials conducted by the Inter- prolonged contact with a sputum smear-positive case national Union Against Tuberculosis Committee on become infected.12,13 When the source case is smear- Prophylaxis using 3, 6 and 12-month regimens showed negative, 30–40% of children become infected.14,15 the protective effi cacy of respectively 32%, 69% and Without IPT, 40–50% of infants and 15% of older 93% among individuals adhering to treatment, la- children with LTBI will develop active TB within belled ‘completer-compliers’.7 INH preventive ther- 2 years.16 One study from South Africa diagnosed ac- tive TB in 16% of children aged <5 years of age ex- 17 This article is dedicated to the memory of E Dju, fi eldwork super- posed to an adult TB case at home. visor, who was killed in a tragic car accident while conducting the Public health priority for childhood TB has been study. low,18 and treatment of LTBI has been largely ignored

Correspondence to: Victor Francisco Gomes, Projecto de Saúde de Bandim, Apartado 861, 1004 Bissau Codex, Guinée- Bissau. Tel: (+245) 320 14 89/320 44 60. Fax: (+245) 320 16 72. e-mail: [email protected] Article submitted 27 August 2010. Final version accepted 14 June 2011. [A version in French of this article is available from the Editorial Offi ce in Paris and from the Union website www.theunion.org] 1638 The International Journal of Tuberculosis and Lung Disease outside of high-income countries.6 In low-income the initiation of INH prophylaxis, the children were countries IPT is still not in general use, partly due to evaluated for active TB. For children aged ⩽5 years, concerns about the risk of creating drug resistance, a ‘weight by age chart’ and the Keith Edwards score25 risk of toxicity and low protection due to poor ad- were applied. Human immunodefi ciency virus (HIV) herence,3,19–21 but also due to lack of structural and testing was not routinely performed; however, if the operational requirements for implementing a pro- investigation suggested active TB, children were sub- phylaxis policy. The use of INH has recently been mitted to a careful assessment of all evidence from emphasised as an important tool for childhood TB history, clinical examination and relevant investiga- control in both low- and high-income countries.22 tions, including HIV and CXR. We set out to evaluate operational aspects of im- Broad spectrum antibiotics were administered for plementing self-administered 9-month INH prophy- 10–15 days. Children who failed to improve clini- laxis to children living in the same house as an adult cally and radiologically after 2 weeks and without with pulmonary TB. We assessed adherence and side other explanatory disease were given a full TB treat- effects over 9 months. ment regimen based on the national protocol. Chil- dren who developed signs and symptoms suggestive MATERIAL AND METHODS of active TB while on IPT were evaluated and treated in a similar way. No active screening was performed Setting for other members of the household; however, they The present study was conducted from September were referred for further investigation if they pre- 2005 to October 2007 in six areas in Bissau, Guinea- sented with symptoms. Bissau, that are covered by the Bandim Health Proj- ect (BHP), a Health and Demographic Surveillance INH administration, adherence and completion Site (HDSS) that is part of the International Network INH tablets were administered at 5 mg/kg daily to- for the Demographic Evaluation of Populations and gether with pyridoxine (vitamin B6) tablets. Vitamin Their Health (INDEPTH) in low-income countries. B6 dosage was 25 mg for children receiving <100 mg The population of around 102 000 is followed through INH and 50 mg for children receiving ⩾100 mg regular censuses and registered with information on INH.26 Medicines were provided every 2 weeks by a sex, ethnic background, and dates of birth, and fi eld assistant. Pill counts were performed at each migration. Since 1996, a TB surveillance system im- visit and evaluated on a monthly basis. plemented in collaboration with the national TB re- ferral hospital has identifi ed adult TB cases using Follow-up and side effects passive and active case fi nding.23 The incidence of Study children were visited after 1, 4, 7 and 9 months adult intrathoracic TB in the area is high, at 471 per of INH treatment. Follow-up visits at 1 and 7 months 100 000 person-years.23 were performed by the research clinician at the local health centre, and at 4 and 9 months by a fi eld assis- Participant and patient recruitment tant at the child’s home. Evaluation at follow-up vis- During the study period, culture for Mycobacterium its included questions about side effects and a physi- tuberculosis and drug susceptibility testing were not cal assessment of signs and symptoms of hepatotoxicity, available. TB diagnosis relied solely on sputum smear e.g., jaundice and vomiting. Weight was registered microscopy. Once an adult from the study area was and entered on a weight-by-age chart to detect failure identifi ed with pulmonary TB, the project assistant to thrive or malnutrition during IPT. went to the patient’s house and updated the census.24 Routine laboratory monitoring of transaminase TST was performed using the Mantoux technique levels during treatment was not performed unless a among all children aged <15 years. The standard child presented signs or symptoms suggestive of 2 tuberculin units of purifi ed protein derivate RT23 hepatotoxicity. with polysorbate 80 (Tween 80, Statens Serum Insti- tut, Copenhagen, Denmark) was administered intra- Statistical analysis dermally to the volar surface of the left forearm by a Standardised questionnaires were used to collect in- trained nurse. The size of the reaction to the tubercu- formation and data were entered using dBase V soft- lin was measured by a trained reader at a second visit ware. Statistical analyses were conducted using STATA 48–72 h after the inoculation. version 10 (Stata Corp, College Station, TX, USA). A Children aged <15 years living in the house when signifi cance level of 5% was used in the analyses. We the adult started treatment were eligible for inclu- present total adherence as well as the total number of sion; children aged <5 years were eligible for IPT, re- months and the number of consecutive months com- gardless of TST status, while children aged 5–15 pleted. Two defi nitions of a completed month of years were eligible if the mean of the longitudinal and treatment were applied: 80% and 60% adherence. In transversal indurations of the TST reaction was ⩾10 a month with loss to follow-up, adherence was de- mm. Children were enrolled in the study and placed fi ned as tablet intake divided by the number of days on IPT if a parent or guardian gave consent. Prior to until the patient stopped attending. Adherence to IPT in Guinea-Bissau 1639

RESULTS Index cases and included children From September 2005 to October 2007, 2631 children were identifi ed from the census fi les as household con- tacts of the identifi ed index cases diagnosed with pul- monary TB (Figure 1). Of the identifi ed children, 729 (28%) were absent or travelling and seven children ex- posed to TB died before the enrolment visit. In all, 1895 children were evaluated for eligibility for IPT. Of these, 736 (39%) were aged ⩽5 years; 609 (83%) were enrolled in the IPT. Of the 1159 children aged >5 years, 253 (22%) were TST-positive, of whom 211 (83%) were enrolled on IPT. A total of 820 children Figure 1 Flow chart of inclusion. TST = tuberculin skin test. were included on IPT. The elapsed time, on average, be- tween adult diagnosis and TST placement was 57 days. Treatment termination Risk factors for non-completion were assessed us- No child was diagnosed with active TB using the ing <6 consecutive months of treatment as the out- Keith Edward score at the time of inclusion. Active come. We used risk ratios using Cox regression with TB was diagnosed in two children after respectively 3 fi xed risk time and robust standard errors27 to evalu- and 5 weeks of IPT. Both diagnoses were based on ate risk factors. The following risk factors were eval- clinical and CXR fi ndings. Both children were tested uated both in univariate analyses and jointly in a for HIV. One of these children was HIV-1-positive multivariate analysis: sex, age group, ethnicity, reli- and enrolled in an antiretroviral (ARV) programme. gion, residential area, care givers’ ability to read, edu- Two children died during IPT: for one, a 6-month old cational level of the parents, proximity to the index boy, hospital records gave the cause of death as se- TB case, adult crowding (number of persons ⩾15 years vere and anaemia; the mother of the second living in the same household) and child crowding child, a 2-year-old girl who died 2 months after the (number of persons <15 years living in the same initiation of IPT, reported that she had had diarrhoea, household). The following factors, obtained from the cough and fever prior to death. general BHP registration system, were used to evalu- ate the impact of socio-economic status on treatment Adherence and completion adherence: straw roof, electricity and indoor toilet. A In all, the children took 79% of the prescribed tab- backward selection procedure, sequentially eliminat- lets. Individually, 83% of the children took more than ing insignifi cant factors at a 5% signifi cance level, was 60% of the prescribed doses, while 65% took more used to assess the effect of the variables evaluated in than 80%. As the urban population in Guinea Bissau is the multivariate analysis. very mobile, 21% of the children were lost to follow- up. To assess the continuity of drug consumption, we Ethical approval looked at the number of months of completed treat- Parents or care givers were informed in written Por- ment. Table 1 shows the total number of months com- tuguese and verbally in the common language, Creole, pleted and the longest period of continuously com- before the child was enrolled in the study. The study pleted months. Figure 2 illustrates these graphically. protocol was approved by the Guinea-Bissau Ethics Generally 6–9 months of IPT is recommended. Con- Committee. sidering all children, censoring those who migrated,

Table 1 Number of children completing at least 1, 2, 3, 4, 5, 6, 7, 8 or 9 months of treatment

1 month 2 months 3 months 4 months 5 months 6 months 7 months 8 months 9 months (n = 820)* (n = 806)* (n = 791)* (n = 783)* (n = 761)* (n = 745)* (n = 732)* (n = 705)* (n = 683)* n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) n (%) Total months of completed treatment Cut-off 80%† 819 (100) 785 (97) 754 (95) 705 (90) 642 (84) 568 (76) 482 (66) 381 (54) 217 (32) Cut-off 60%† 820 (100) 802 (100) 783 (99) 750 (96) 712 (94) 659 (88) 602 (82) 515 (73) 352 (52) Continuous months of completed treatment Cut-off 80%† 819 (100) 764 (95) 672 (85) 580 (74) 469 (61) 378 (51) 299 (41) 244 (35) 217 (32) Cut-off 60%† 820 (100) 790 (98) 742 (94) 687 (88) 577 (76) 505 (68) 444 (61) 373 (53) 352 (52)

* Number of children still at the house. † The adherence cut-off used to define a completed month. 1640 The International Journal of Tuberculosis and Lung Disease

IPT.28,29 Regarding the number of completed months with more than 80% adherence, 76% of the children completed at least 6 months of treatment in total. Other studies have shown that the rate of completion of a 6-month course of self-administered INH mono- therapy ranges from 6% to 60%, with rates of 20– 30% in most series.22,30,31 Our fi ndings are similar to those of a recent study from Greece, which reported a compliance of 65%.32 The high adherence in our study may partially be explained by the provision of medicines at home by a fi eld assistant who showed a high level of commitment to the work. Figure 2 Frequency of children completing at least 1, 2, 3, 4, The excellent safety and tolerability of IPT re- 5, 6, 7, 8 or 9 months of treatment with 80% adherence. Both ported in a study of HIV-positive children suggest total and consecutive number of months is illustrated. that, in general, INH prophylaxis is safe in children.33 A study from South Africa reported no side effects we observed that 76% completed at least 6 months among 180 children enrolled in INH treatment.17 of treatment, with more than 80% adherence. Another study reported 0% liver injury in children aged 0–14 years,21 while a study of the pharmaco- Reasons for non-adherence kinetics of INH in children showed that young chil- Overall the children failed to take 21% (46 260/ dren eliminate INH faster than older children, and 223 860) of the prescribed doses. Migration ac- children as a group faster than adults.34 Indeed, sev- counted for 41% (19 085/46 260) of the missed doses, eral studies have shown that INH toxicity is strongly while travelling accounted for 17% (8072/46 260). related to the age of the patients, the greatest risk of Forgetting to take the tablets, as stated by the care developing INH-related hepatoxicity being found giver, contributed to 40% (18 724/46 260) of the among individuals aged >35 years.35–37 The present missed doses. Tablets not being distributed by study study confi rmed limited toxicity in children. assistants and other reasons for not taking the tablets The urban population in Guinea-Bissau is very mo- accounted for respectively 0.2% (87/46 260) and bile. In the present study, 21% of the children moved 0.6% (292/46 260). out of the study area during the 9 months of follow- Risk factors for non-completions up. Migration and travelling together accounted for 59% of the missed doses of INH. Adherence may be Table 2 presents risk factors for not completing at infl uenced by the comprehension of the disease and least 6 consecutive months of IPT. Children aged the need for treatment; adherence may be improved by >5 years were signifi cantly more likely to be fully specifi c interventions such as better communication.20 adherent than those aged ⩽5 years (P = 0.003). The We also found that children aged ⩽5 years had worse socio-economic indicators were not associated with adherence than older children. This may be due to long completion. breast-feeding practices (2–3 years), causing mothers Side effects to take small children with them when travelling. Four children reported transient itching after 1 month Limitations of IPT and in one child a mild enlargement of the liver was observed without clinical signs of hepatitis; Adherence assessment in the study was based exclu- no cases of jaundice or dermatitis were reported. sively on pill counts. This method does not provide Liver transaminases were measured in one child with information on whether the tablets were actually vomiting; liver transaminases were normal and fur- taken, and adherence can therefore be overestimated. ther clinical evaluation revealed no signs of side ef- Assessment of adherence using a combination of ob- fects related to INH treatment. jective and subjective tests has been recommended.38 Implications and conclusion DISCUSSION Adverse events to IPT were few, and overall adher- The main fi ndings of this study were a low frequency ence was high. Migration and travelling accounted of adverse reactions, but also rather low rates of con- for most of the observed non-adherence. The results tinuous treatment completion, mainly due to migra- observed indicate that unsupervised IPT is feasible in tion and travelling. Guinea-Bissau and that IPT is safe in children. The The overall rate of adherence was high. In total, provision of IPT for exposed children can be incor- 79% of the prescribed tablets were taken. Individu- porated into the provision of TB treatment for adult ally, 65% of the children completed treatment, with a TB index cases through the registers under routine total adherence of more than 80%, which matches programme conditions. A separate register could be completion rates observed in previous studies of created for children on IPT and linked using a number Adherence to IPT in Guinea-Bissau 1641

Table 2 Risk factors for not completing at least 6 consecutive months of treatment

Non-completion n/N (%) Risk ratio (95%CI) Risk ratio (95%CI)* Risk ratio (95%CI)† P value† Sex Male 213/393 (54) Reference 1 0.87 Female 229/427 (54) 0.99 (0.87–1.12) 1.02 (0.90–1.16) Age, years 0–4 344/601 (57) Reference 1 1 0.003 5–14 98/219 (45) 0.78 (0.66–0.92) 0.79 (0.67–0.94) 0.79 (0.67–0.93) 0.005 Ethnicity Pepel 141/241 (59) Reference 1 0.32 Balanta 46/87 (53) 0.90 (0.72–1.13) 0.93 (0.74–1.17) Manj/Manc 114/216 (53) 0.90 (0.76–1.06) 0.93 (0.78–1.11) Mand/Fula 64/136 (47) 0.80 (0.65–0.99) 0.73 (0.52–1.02) Other 69/121 (57) 0.97 (0.81–1.18) 0.93 (0.73–1.18) Data missing 8/19 (42) 0.72 (0.42–1.23) 0.81 (0.46–1.45) Religion Animist 176/303 (58) Reference 1 0.40 Catholic 133/259 (51) 0.88 (0.76–1.03) 0.88 (0.75–1.04) Muslim 107/210 (51) 0.88 (0.74–1.03) 1.07 (0.81–1.41) Other 25/45 (56) 0.96 (0.72–1.26) 0.89 (0.68–1.18) Data missing 1/3 (33) 0.57 (0.12–2.85) 0.54 (0.13–2.23) Area Bandim 1 169/283 (60) Reference 1 1 0.05 Bandim 2 53/114 (46) 0.78 (0.63–0.97) 0.78 (0.62–0.97) 0.81 (0.65–1.00) Belem 48/95 (51) 0.85 (0.68–1.06) 0.82 (0.65–1.04) 0.82 (0.66–1.02) Mindara 41/65 (63) 1.06 (0.86–1.30) 1.05 (0.84–1.32) 1.07 (0.86–1.32) Cuntum 1 76/148 (51) 0.86 (0.72–1.03) 0.87 (0.72–1.06) 0.86 (0.72–1.03) Cuntum 2 55/115 (48) 0.80 (0.65–0.99) 0.81 (0.64–1.02) 0.79 (0.64–0.98) 0.046 Literate Yes 289/532 (54) Reference 1 0.81 No 152/285 (53) 0.98 (0.86–1.12) 0.82 (0.66–1.02) Data missing 1/3 (33) 0.61 (0.12–3.05) 0.60 (0.11–3.35) School education Yes 242/457 (53) Reference 1 0.77 No 182/332 (55) 1.04 (0.91–1.18) 1.23 (0.99–1.52) Data missing 18/31 (58) 1.10 (0.80–1.50) 1.29 (0.91–1.82) Proximity Same bed 42/82 (51) Reference 1 1 0.06 Different bed 52/110 (47) 0.92 (0.69–1.23) 0.92 (0.69–1.22) 0.90 (0.67–1.19) Different room 330/603 (55) 1.07 (0.85–1.34) 1.00 (0.80–1.24) 1.01 (0.81–1.26) Separate house 18/25 (72) 1.41 (1.02–1.94) 1.42 (1.02–1.98) 1.43 (1.03–1.98) 0.039 Roof Straw 26/57 (46) Reference 1 0.37 Zinc 385/701 (55) 1.20 (0.90–1.61) 1.19 (0.90–1.59) Data missing 31/62 (50) 1.10 (0.75–1.60) 1.60 (0.84–3.05) Electricity Yes 118/215 (55) Reference 1 0.40 No 286/523 (55) 1.00 (0.86–1.15) 0.96 (0.80–1.14) Data missing 38/82 (46) 0.84 (0.65–1.10) 0.85 (0.52–1.39) Toilet Outdoor 353/643 (55) Reference 1 0.35 Indoor 54/100 (54) 0.98 (0.81–1.19) 0.98 (0.79–1.22) Data missing 35/77 (45) 0.83 (0.64–1.07) 0.78 (0.49–1.25) Crowding (>2 adults) No 106/196 (54) Reference 1 0.98 Yes 322/597 (54) 1.00 (0.86–1.16) 1.04 (0.89–1.22) Data missing 14/27 (52) 0.96 (0.65–1.41) 0.75 (0.29–1.95) Crowding (>2 children) No 105/179 (59) Reference 1 0.32 Yes 322/613 (53) 0.90 (0.78–1.03) 0.85 (0.72–0.99) Data missing 15/28 (54) 0.91 (0.63–1.32) 1.14 (0.46–2.81)

* Multivariate analysis including all variables. † Multivariate analysis including age, area and proximity, which were significant in a backward selection procedure using a 5% significance level. 1642 The International Journal of Tuberculosis and Lung Disease to the adult TB index case. The National TB Pro- Beyers N. Adherence to isoniazid preventive chemotherapy: a gramme should ensure DOTS delivery using the com- prospective community based study. Arch Dis Child 2006; 91: 762–765. munity approach. Systematic engagement of other 18 Donald P R. Childhood tuberculosis: the hidden epidemic. Int care providers is needed. These fi ndings are relevant J Tuberc Lung Dis 2004; 8: 627–629. to the development of an effective IPT programme in 19 Nelson L J, Wells C D. Global epidemiology of childhood tu- a setting such as Guinea-Bissau. berculosis. Int J Tuberc Lung Dis 2004; 8: 636–647. 20 Starr M, Sawyer S, Carlin J, Powell C, Newman R, Johnson P. Acknowledgements A novel approach to monitoring adherence to preventive ther- apy for tuberculosis in adolescence. J Paediatr Child Health The authors thank the study population for their participation and 1999; 35: 350–354. good reception. They also thank the staff of the Bandim Health 21 LoBue P A, Moser K S. Use of isoniazid for latent tuberculosis Project for their contribution to data collection and management. infection in a public health clinic. Am J Respir Crit Care Med This study was supported by the Swedish International Develop- 2003; 168: 443–447. ment Cooperation Agency/Department for Research Cooperation 22 van Zyl S, Marais B J, Hesseling A C, Gie R P, Beyers N, Schaaf and the Danish International Development Agency. H S. Adherence to anti-tuberculosis chemoprophylaxis and treatment in children. Int J Tuberc Lung Dis 2006; 10: 13–18. 23 Gustafson P, Gomes V F, Vieira C S, et al. Tuberculosis in References Bissau: incidence and risk factors in an urban community in 1 American Thoracic Society. Chemoprophylaxis for the preven- sub-Saharan Africa. Int J Epidemiol 2004; 33: 163–172. tion of tuberculosis. Am Rev Respir Dis 1965; 96: 558–560. 24 Wejse C, Gomes V F, Rabna P, et al. Vitamin D as supplemen- 2 Targeted tuberculin testing and treatment of latent tuberculo- tary treatment for tuberculosis: a double-blind, randomized, sis infection. This offi cial statement of the American Thoracic placebo-controlled trial. Am J Respir Crit Care Med 2009; Society was adopted by the ATS Board of Directors, July 1999. 179: 843–850. This is a Joint Statement of the American Thoracic Society 25 Narayan S, Mahadevan S, Serane V T. Keith Edwards Score (ATS) and the Centers for Disease Control and Prevention for diagnosis of tuberculosis. Indian J Pediatr 2003; 70: 467– (CDC). This statement was endorsed by the Council of the In- 469. fectious Diseases Society of America (IDSA), September 1999. 26 Médecins Sans Frontières. Médicaments essentiels: guide pratique Am J Respir Crit Care Med 2000; 161 (4 Pt 2): S221–S247. d´utilisation. 2nd ed. Paris, France: MSF, 1993: p 78. [French] 3 Ferebee S H. Controlled chemoprophylaxis trials in tuberculo- 27 Barros A J, Hirakata V N. Alternatives for logistic regression sis. A general review. Bibl Tuberc 1970; 26: 28–106. in cross-sectional studies: an empirical comparison of models 4 Jenkins D, Davidson F F. Isoniazid chemoprophylaxis of tuber- that directly estimate the prevalence ratio. BMC Med Res culosis. Calif Med 1972; 116: 1–5. Methodol 2003; 3: 21. 5 Comstock G W, Baum C, Snider D E Jr. Isoniazid prophylaxis 28 Page K R, Sifakis F, Montes de O R, et al. Improved adherence among Alaskan Eskimos: a fi nal report of the Bethel Isoniazid and less toxicity with rifampin vs isoniazid for treatment of Studies. Am Rev Respir Dis 1979; 119: 827–830. latent tuberculosis: a retrospective study. Arch Intern Med 6 Chaisson R E. New developments in the treatment of latent tu- 2006; 166: 1863–1870. berculosis. Int J Tuberc Lung Dis 2000; 4 (Suppl 2): S176–S181. 29 Rennie T W, Bothamley G H, Engova D, Bates I P. Patient 7 International Union Against Tuberculosis Committee on Proph- choice promotes adherence in preventive treatment for latent ylaxis. Effi cacy of various durations of isoniazid preventive tuberculosis. Eur Respir J 2007; 30: 728–735. therapy for tuberculosis: fi ve years of follow-up in the IUATLD 30 Levesque J F, Dongier P, Brassard P, Allard R. Acceptance of trial. Bull World Health Organ 1982; 60: 555–564. screening and completion of treatment for latent tuberculosis 8 Rieder H L. Interventions for tuberculosis control and elimina- infection among refugee claimants in Canada. Int J Tuberc tion. Paris, France: International Union Against Tuberculosis Lung Dis 2004; 8: 711–717. and Lung Disease, 2002: pp 127–146. 31 Coly A, Morisky D. Predicting completion of treatment among 9 Enarson D, Rieder H L, Arnadottir T, Trébucq A. Management foreign-born adolescents treated for latent tuberculosis infec- of tuberculosis. A guideline for low-income countries. 5th ed. tion in Los Angeles. Int J Tuberc Lung Dis 2004; 8: 703–710. Paris, France: International Union Against Tuberculosis and 32 Spyridis N P, Spyridis P G, Gelesme A, et al. The effectiveness Lung Disease, 2000: pp 1–89. of a 9-month regimen of isoniazid alone versus 3- and 4-month 10 Hopewell P C, Pai M, Maher D, Uplekar M, Raviglione M C. regimens of isoniazid plus rifampin for treatment of latent tu- International standards for tuberculosis care. Lancet Infect Dis berculosis infection in children: results of an 11-year random- 2006; 6: 710–725. ized study. Clin Infect Dis 2007; 45: 715–722. 11 Marais B J, Pai M. New approaches and emerging technologies 33 Zar H J, Cotton M F, Strauss S, et al. Effect of isoniazid prophy- in the diagnosis of childhood tuberculosis. Paediatr Respir Rev laxis on mortality and incidence of tuberculosis in children 2007; 8: 124–133. with HIV: randomised controlled trial. BMJ 2007; 334: 136. 12 Singh M, Mynak M L, Kumar L, Mathew J L, Jindal S K. Prev- 34 Schaaf H S, Parkin D P, Seifart H I, et al. Isoniazid pharma- alence and risk factors for transmission of infection among cokinetics in children treated for respiratory tuberculosis. Arch children in household contact with adults having pulmonary Dis Child 2005; 90: 614–618. tuberculosis. Arch Dis Child 2005; 90: 624–628. 35 Martinez S A, Calpe Calpe J L, Llavador R G, Ena M J, Calpe 13 Marais B J, Gie R P, Schaaf H S, et al. The natural history of A A. Primary prevention and treatment of latent tuberculosis childhood intra-thoracic tuberculosis: a critical review of liter- infection with isoniazid: effi cacy of a control program, 1997– ature from the pre-chemotherapy era. Int J Tuberc Lung Dis 2002. Arch Bronconeumol 2005; 41: 27–33. 2004; 8: 392–402. 36 Garibaldi R A, Drusin R E, Ferebee S H, Gregg M B. Isoniazid- 14 Gedde-Dahl T. Tuberculous infection in the light of tuberculin associated hepatitis. Report of an outbreak. Am Rev Respir matriculation. Am J Hygiene 1952; 56: 139–214. Dis 1972; 106: 357–365. 15 Zeidberg L, Gass R, Dillon A, Hutchesson R. The Williamson 37 Kopanoff D E, Snider D E Jr, Caras G J. Isoniazid-related hepa- County Tuberculosis Study. A twenty-four-year epidemiologic titis: a US Public Health Service cooperative surveillance study. study. Am Rev Respir Dis 1963; 87: 1–88. Am Rev Respir Dis 1978; 117: 991–1001. 16 Shingadia D, Novelli V. Diagnosis and treatment of tuberculo- 38 Stephenson B J, Rowe B H, Haynes R B, Macharia W M, Leon sis in children. Lancet Infect Dis 2003; 3: 624–632. G. The rational clinical examination. Is this patient taking the 17 Marais B J, van Zyl S, Schaaf H S, van Aardt M, Gie R P, treatment as prescribed? JAMA 1993; 269: 2779–2781. Adherence to IPT in Guinea-Bissau i

RÉSUMÉ

OBJECTIF : Evaluer l’adhésion au traitement préventif à adultes de TB au total 2631 enfants. Parmi les enfants l’isoniazide (IPT) chez les enfants exposés à leur domicile identifiés, 1895 (72%) ont fait l’objet d’une évaluation à un cas de tuberculose pulmonaire (TB) chez l’adulte. pour l’éligibilité pour l’IPT avec un recrutement total de MÉTHODES : On a recruté les enfants dans l’IPT 820 enfants dans l’étude, 609 âgés de ⩽5 ans et 211 de l orsqu’ils étaient âgés de ⩽5 ans ou de 5 à 15 ans avec un 5 à 15 ans. Au total, 79% des doses prescrites ont été test cutané tuberculinique ⩾10 mm. Les enfants ont été prises et 65% des enfants ont pris plus de 80% des doses inclus à partir du système de surveillance démographique prescrites. Au total, un achèvement de plus de 6 mois du Projet Santé Bandim à Bissau, Guinée-Bissau. Les consécutifs d’IPT a été réalisé chez 51%. mesures principales de résultats ont été l’adhésion, les CONCLUSION : L’adhésion globale à l’IPT s’est avérée taux d’achèvement et les effets collatéraux au cours des meilleure que celle signalée antérieurement dans des 9 mois d’IPT. Le résultat principal a été une adhésion zones endémiques de TB, puisque 76% des enfants ont d’au moins 80% pendant 6 mois consécutifs. achevé au moins 6 mois de traitement avec une adhésion RÉSULTATS : On a identifié comme contacts de cas de >80%.

RESUMEN

OBJETIVO: Evaluar el cumplimiento del tratamiento RESULTADOS: Se detectaron 2631 niños como contactos preventivo con isoniazida (IPT) en los niños que están de casos adultos de TB. Se evaluaron 1895 de estos niños expuestos en su hogar a adultos que padecen tuberculo- (72%) a fin de confirmar si cumplían con los requisitos sis (TB) pulmonar. del IPT y se inscribieron 820 en el estudio: 609 niños MÉTODOS: Se incluyeron en el programa de IPT los ⩽5 años y 211 niños entre los 5 años y los 15 años de n iños ⩽5 años de edad o entre los 5 años y los 15 años edad. Globalmente, se administraron 79% de las dosis que presentaban una reacción cutánea a la tuberculina recetadas y 65% de los niños recibieron más del 80% de ⩾10 mm. Los niños provenían del sistema de vigilancia las dosis previstas. En total, 51% completaron más de demográfica del proyecto sanitario de Bandim a Bissau, 6 meses consecutivos de IPT. en Guinea-Bissau. Los principales variables de valora- CONCLUSION: El cumplimiento global del tratamiento ción fueron el cumplimiento, las tasas de compleción y preventivo con isoniazida fue superior al que se ha co- los efectos adversos del IPT durante 9 meses. El principal municado previamente en las regiones endémicas; el criterio de evaluación fueron 6 meses consecutivos con 76% de los niños completó como minimo 6 meses de un cumplimiento terapéutico como mínimo de 80%. tratamiento, con un cumplimiento superior al 80%.