Annals of Tropical Paediatrics (2003) 23, 47–50 Low level of compliance with tuberculosis treatment in children: monitoring by urine tests AYS ¸ E PALANDUZ, DERYA GU ¨ LTEKIN, ELA ERDEM & NIMET KAYAALP Department of Paediatrics, S ¸ is ¸li Etfal Hospital, I stanbul, Turkey (Accepted November 2002) Summary Patient compliance should be ensured in an eVective tuberculosis control programme. We measured patient compliance by detecting antituberculous drugs in the urine of 237 outpatients receiving one to three antituberculous drugs. Positive controls were 20 hospitalised patients, supervised to receive isoniazid (INH), rifampicin (RIF) and pyrazinamide (PZA), and negative controls were not on any drugs. Among the 237 study patients, only 67% were found to be taking the appropriate treatment and 8% had taken none. We conclude that a remarkable number of patients (33%) were non-compliant with treatment. The detection of antituberculous drugs in the urine is a quick, simple and inexpensive means of measuring adherence to treatment. Unless directly observed therapy (DOT) is adopted, we recommend routine urine testing for antituberculous drugs to identify defaulting patients. Introduction were not. 2 The use of pill counts is another method but patients may deceive by altering The problem of non-adherence to therapy the number of pills and the presence of an has been recognised since Hippocrates. 1 In appropriate number of pills does not mean the case of tuberculosis, non-adherence is that the missing pills have been ingested by potentially more serious than inattention to the patient. 3 Micro-electronic pill dispensers therapy for any other disease because not are used to measure drug compliance but only does it result in substantial morbidity they also do not prove that the drug has and mortality in individual patients but it is been taken. 4,5 Actual ingestion of medication also potentially a public hazard as it contri- is best measured by testing serum and urine butes to the development of drug resistance. for presence of the drug. 6 Serum tests are Directly observed therapy is the best way invasive, expensive and not easily applicable. to ensure compliance. If this method can not We tested compliance by performing be adopted, adherence to treatment should simple urine tests to demonstrate the pre- be measured periodically throughout the sence of isoniazid (INH), rifampicin (RIF) course of treatment. There are several ways and pyrazinamide (PZA) in the urine of of measuring adherence to antituberculous tuberculous patients. therapy. One is to question the patients but this is not reliable as it has been reported that, when questioned, patients frequently Materials and Methods expressed themselves adherent when they The study was conducted in S ¸ is ¸li Etfal Research and Training Hospital, I stanbul. Reprint requests to: Dr Ays ¸e Palanduz, Hasan Halife A total of 277 subjects treated either for Mahallesi, Sehit Kubilay Sokak, 14/1, Fatih, 34080, I stanbul, Turkey. E-mail: apalanduz@yahoo.com pulmonary tuberculosis or latent infection © 2003 The Liverpool School of Tropical Medicine DOI: 10.1179/000349803125002869