Abstract:
Background: Screening and treatment for latent TB infection is necessary in the clinical
settings where people with HIV infection receive their care. Tanzania’s National TB and
Leprosy Program (NTLP) has started to provide Isoniazid Preventive Therapy (IPT) to HIV
infected patients with latent TB infection in 14 pilot sites. A screening tool which includes
fever = 2 weeks, cough = 2 weeks, hemoptysis, noticeable weight loss for new patients or 3 kg
weight loss in a month and excessive sweating at night for = 2 weeks is used to rule out active
TB infection. Patients are considered to possibly have active TB when they present with any
of the 5 symptoms in the tool and are further subjected to other TB investigations.
Objective: To determine the usefulness of the NTLP screening tool in identifying patients
eligible for IPT and the role of Tuberculin skin test (TST) among patients attending Muhimbili
national hospital (MNH) HIV clinic.
Study design and setting: Descriptive cross sectional study among HIV infected patients.
Methodology: Socio demographic data was obtained using structured questionnaires. Patients
underwent physical examination, chest x-ray (CXR), TST, induced sputum for acid fast bacilli
(AFB) microscopy and mycobacterial culture, CD4 count and complete blood count.
Results: A total of 373 patients were enrolled, 72.1% being fema les. Active TB was found in
4.1% (using culture) and 9.2% (NTLP TB definition) of the participants. The sensitivity and
specificity of the NTLP TB screening tool was 71.4% and 75.9% respectively, with PPV and
NPV of 11.4% and 98.4% respectively. A CXR identified 3 of the 4 participants with culture
confirmed MTB that were missed by the screening tool. Cough = 2 weeks and ARV use were
independent predictors of sputum culture defined TB. A positive TST was found among 24%
of the participants and no relationship was observed between TST reactivity and TB.
vii
Conclusion and recommendation:
The prevalence of PTB among patients attending MNH HIV clinic is high (4.1% and 9.2%
using culture and NTLP PTB definition respectively). The screening tool showed a good
sensitivity and specificity for TB with a high negative predictive value, making it a good
screening tool in ruling out active TB. Whenever possible a CXR should be done as this may
improve the sensitivity of the tool. No relationship was observed between TST and TB
