Obtaining valid laboratory data in clinical trials conducted in resource diverse settings: lessons learned from a microbicide phase III clinical trial.

BACKGROUND: Over the last decade several phase III microbicides trials have been conducted in developing countries. However, laboratories in resource constrained settings do not always have the experience, infrastructure, and the capacity to deliver laboratory data meeting the high standards of clinical trials. This paper describes the design and outcomes of a laboratory quality assurance program which was implemented during a phase III clinical trial evaluating the efficacy of the candidate microbicide Cellulose Sulfate 6% (CS) [1]. METHODOLOGY: In order to assess the effectiveness of CS for HIV and STI prevention, a phase III clinical trial was conducted in 5 sites: 3 in Africa and 2 in India. The trial sponsor identified an International Central Reference Laboratory (ICRL), responsible for the design and management of a quality assurance program, which would guarantee the reliability of laboratory data. The ICRL provided advice on the tests, assessed local laboratories, organized trainings, conducted supervision visits, performed re-tests, and prepared control panels. Local laboratories were provided with control panels for HIV rapid tests and Chlamydia trachomatis/Neisseria gonorrhoeae (CT/NG) amplification technique. Aliquots from respective control panels were tested by local laboratories and were compared with results obtained at the ICRL. RESULTS: Overall, good results were observed. However, discordances between the ICRL and site laboratories were identified for HIV and CT/NG results. One particular site experienced difficulties with HIV rapid testing shortly after study initiation. At all sites, DNA contamination was identified as a cause of invalid CT/NG results. Both problems were timely detected and solved. Through immediate feedback, guidance and repeated training of laboratory staff, additional inaccuracies were prevented. CONCLUSIONS: Quality control guidelines when applied in field laboratories ensured the reliability and validity of final study data. It is essential that sponsors provide adequate resources for implementation of such comprehensive technical assessment and monitoring systems. TRIAL REGISTRATION: ClinicalTrials.gov NCT00153777 and Current Controlled Trials ISRCTN95638385.

Prevalence of sexually transmitted infections among adolescents in Kampala, Uganda, and theoretical models for improving syndromic management.

PURPOSE: To estimate the prevalence of treatable sexually transmitted infections (STI) in adolescents visiting a youth health clinic. To evaluate the algorithm for management of the abnormal vaginal discharge (AVD) syndrome recommended in Uganda's national guidelines and compare it with other theoretical flowchart models. METHODS: Sexually experienced adolescents who were visiting an urban youth health clinic in Kampala, Uganda were examined and interviewed (with their consent) about their socio-demographic background, sexual risk factors, and genital symptoms. Samples taken for Chlamydia trachomatis (CT), Neisseria gonorrhoea (NG), and Trichomonas vaginalis (TV) were analyzed by polymerase chain reaction (PCR). Rapid plasma reagin (RPR) was used with confirming treponema pallidum hem agglutination (TPHA) for syphilis diagnosis. One hundred ninety-nine females and 107 males were examined. Performance of the national algorithm was compared with different theoretical algorithms. RESULTS: Prevalence of CT, NG, TV and syphilis was 4.5%, 9.0%, 8.0%, and 4.0%, respectively, for girls and 4.7%, 5.7%, 0%, and 2.8%, respectively, for boys. We found that 20.6% of the females and 13.2% of the males had at least one STI. The national AVD flow chart had a sensitivity of 61%, a specificity of 38.5% and a positive predictive value (PPV) of 11.6%. All the models had PPV of less than 20% and sensitivity less than 85%. The best performing algorithm using risk and protective factors, rather than symptoms, implicated a sensitivity/specificity and PPV of 82.6%/47% and 17.3%, respectively (p = .012). CONCLUSIONS: An algorithm for management of STI using behavioral and demographic factors in this population demonstrated enhanced sensitivity, specificity, and PPV.