Evaluation of a clinical index among adult contacts of children with tuberculosis in rural Haiti.

OBJECTIVE: To analyze the prevalence of symptoms among adults with and without tuberculosis who were household contacts of children <6 years old with tuberculosis in rural Haiti. DESIGN: An analytic group comparison study to determine whether a clinical index questionnaire (previous tuberculosis diagnosis, previous tuberculosis treatment, cough >1 month, coughing up of blood in the past year, fever >1 month and weight loss in the past year), administered to adults accompanying the child to the hospital can identify which adult household contacts have pulmonary tuberculosis. Household contacts then completed the clinical index and underwent chest X-ray and sputum smear examination. RESULTS: Thirty-two children had 109 adult household contacts. Of these, 56 completed evaluation and nine had pulmonary tuberculosis. Proxy report (OR 36.6, 95%CI >2.3, P = 0.004) was better than self report (OR 2.9, 95%CI >0.36, P = 0.44) in predicting clinical tuberculosis. Investigating adults with a positive clinical index by proxy would have reduced from 26 to 11 those investigated, and no cases of pulmonary tuberculosis would have been missed. CONCLUSION: In rural Haiti, clinical index by proxy by the adult accompanying the child to the hospital can identify which adult household contacts should be investigated for pulmonary tuberculosis.