A comparision of the yield of three tuberculosis screening modalities among people living with HIV: a retrospective quasi-experiemental study

BACKGROUND: The Intensified Case Finding (ICF) tool was approved for TB screening in 2011; however there is still paucity of robust data comparing yields of the different ICF screening modalities. We compared yields of three different screening modalities for TB among Patients Living with HIV (PLHIV) in Uganda in order to inform National TB Programs on the most effective TB screening method. METHODS: This was a retrospective quasi-experimental study conducted at an Out-Patient HIV/AIDS clinic in Uganda. We set out to determine yields of three different TB screening modalities at three time periods: 2006/07 where Passive Case Finding (PCF) was used. Here, no screening questions were administered; the clinician depended on the patient’s self report. In 2008/09 embedded Intensified Case Finding Tool (e-ICF) was used; here a data capture field was added to the patient clinical encounter forms to compel clinicians to screen for TB symptoms. In 2010/11 Independent Intensified Case Finding Tool (i-ICF) was used; here a screening data collection form, was used, it had the same screening questions as e-ICF. Routine clinical data, including TB status, were collected and entered into an electronic clinical care database. Analysis was done in STATA and the main outcome estimated was the proportional yield of TB cases for each screening modality. RESULTS: The overall yield of TB cases was 11.18 % over the entire period of the study (2006 – 2011). The intervention–specific yields were 1.86 % for PCF, 14.95 % for e-ICF and 12.47 % for i-ICF. Use of either e–ICF (OR: 9.2, 95 % CI: 4.81-17.73) or i– ICF (OR: 7.7, 95 % CI: 4.02-14.78) significantly detected more TB cases compared to PCF (P <0.001). While the yields of the Active Case Finding modalities (e-ICF & i-ICF) were not significantly different (OR: 0.98, 95 % CI 0.76-1.27, P = 0.89). CONCLUSION: The active screening modalities (e-ICF & i-ICF) had a comparable TB yield and were eight to nine times more efficient in identifying TB cases when compared to the PCF. Cost effectiveness studies would be informative.