Developing strategies for TB screening among HIV-infected and HIV-uninfected pregnant and postpartum women in Swaziland Marianne Calnan Deputy Director www.urc-chs.com
Background Recent research suggests that pregnant women and new mothers are at an increased risk of TB. The absolute risk of TB in pregnancy is related to whether the woman /mother has been exposed to TB and become infected with it. Women who have lived in parts of the world where TB is endemic, those who have been in close contact with a case of active disease; or who are immunosuppressed are at particular risk. However, data suggests that current TB screening methods are missing a significant number of pregnant women who may be infected with Tuberculosis, partly due to similarity of symptoms between pregnancy and TB.
Situation in Swaziland HIV prevalence rate in gen pop n (18 49 age group): 31% 1 HIV prevalence among ANC clients: 42% 2 (ANC HIV surveillance 2010), more recent anecdotal data (49%) ANC coverage: 85% 3, PMTCT coverage:90% 3 EID positivity rate: about 2.5% TB incident cases: 773 per 100,000 HIV co-infection rate in TB patients: 76% Proportion of TB patients who are children: 11% Proportion of TB patients who are women: 48.5% Rates of TB among pregnant women: program data 0.4% 1. Swaziland Ministry of Health. (2012). Swaziland HIV Incidence Measurement Survey ( SHIMS ). Mbabane. 2. Eaton, J. W., Rehle, T. M., Jooste, S., Nkambule, R., Kim, A. A., Mahy, M., & Hallett, T. B. (2014). Recent HIV prevalence trends among pregnant women and all women in sub-saharan Africa: implications for HIV estimates. AIDS (London, England), 28 Suppl 4(4), S507-14. 3. Countdown to 2015. (2016). A Decade of Tracking Progress for Maternal, Newborn and Child Survival. The 2015 Report Swaziland.
TB Screening in ANC and PNC Unit: Ministry of Health TB Screening Tool Serial Number: Patient Data Date: / / / Question 1 Adult Current Cough? (a) Question 2 Fever for 2 weeks or more? Question 3 Night Sweats for 2 weeks or more? Question 4 History of contact with a TB case? Question 5 Recent or ongoing unintentional weight loss? Pre/ART/ART No. Age: Sex: M F Weight (in Kgs): Children < 15 Current Cough? (a) Fever for 2 weeks or more? Poor weight gain/failure to thrive History of contact with a TB case? a) If Yes to Q1 probe for duration of cough and if duration is more than 2 weeks consider that as a presumptive TB case. b) If Yes to Q1 but duration is less than 2 weeks it must be accompanied by Yes in any of the other questions i.e. Q2-Q4. c) If No to Q1, consider as presumptive TB if at least two (2) of the other symptoms i.e. Fever & Night sweats etc. exists. d) For Children, History of Contact with TB case alone should be considered presumptive TB case. Night sweats? ACTION TAKEN SCREEN RESULT OSITIVE (Presumptive T NEGATIVE Note: If a case satisfies the instructions in (a) to (d) then select POSITIVE; but if it does not, then select NEGATIVE putum collected for Lab Investigation Y No erred to Doctor for Clinical Assessment No
Study Aim To identify TB screening algorithm for all pregnant and lactating women, including HIV infected, that maximizes the number of true TB cases detected in a timely, affordable manner. Objectives To measure the sensitivity, specificity, positive and negative predictive values of the WHO-recommended four-symptom screening tool among HIV infected and non HIV infected pregnant and lactating women in Swaziland, compared with the gold standard of MTB sputum culture To measure the sensitivity, specificity, positive and negative predictive values of proposed enhanced TB screening algorithms among HIV infected and non HIV infected pregnant and lactating women in Swaziland, compared with the gold standard of MTB sputum culture To measure the prevalence of bacteriologically confirmed TB among the study sample of HIV infected and non HIV infected pregnant and lactating women
Study methods Cross-sectional analytic study Conducted in 3 public health units over 7 months Non-random, consecutive sampling of HIV positive and HIV negative pregnant and lactating women Sample size was 990 HIV positive pregnant women: 254, 25.7% HIV negative pregnant women: 262, 26.5% HIV positive lactating women: 216, 21.8% HIV negative lactating women: 258, 26%
Enrollment All pregnant and lactating women accessing antenatal care (ANC) or post natal care at three maternal child health (MCH) clinics in Swaziland. Inclusion Criteria: Pregnant woman (any stage of pregnancy) or lactating woman (up to 6 months post- delivery) Was aged 18 or above Was willing and able to provide free informed consent Was not currently diagnosed as having active TB or TB treatment Had not been on TB treatment in the past 2 months Had not enrolled in the study less than three months prior Exclusion Criteria: Was not a pregnant woman or lactating woman (up to 6 months lactating) Was aged below 18 Was unwilling or unable to provide free informed consent Was currently diagnosed as having active TB or TB treatment Had been on TB treatment in the past 2 months. Had been enrolled in the study less than three months prior
Study Procedures & Data Collection All study participants had demographic and social history collected. Regardless of symptoms, all participants underwent TB evaluation that included: TB screening questionnaire TST IGRA Urine TB Lam Sputum collection for GeneXpert MTB Rif MTB culture AFB Microscopy Chest Xray If HIV positive, CD4 count
Demographics Variable Number (n, %) Age 26 years (median) Region Hhohho Manzini Shiselweni 355, 35.9% 343, 34.6% 292, 29.5% Marital status Single Married Cohabiting Widowed 577, 58.3% 315, 31.8% 97, 9.8% 1, 0.1% Employment Unemployed Employed 690, 60.6% 300, 30.3% Patient type Pregnant Lactating 516, 52.1% 474, 47.9% Pregnancy trimester First Second Third 101, 10.2% 219, 22.1% 196, 19.8% HIV status Negative Positive 520, 52.5% 470, 47.5%
Results Participant Enrollment HIV infected pregnant women: 254 HIV uninfected pregnant women: 262 HIV infected lactating women: 216 HIV uninfected lactating women: 258 Prevalence of TB in study population Overall TB prevalence: 3% Pregnant women: 3.8% (2.2% - 6.2%) Lactating women: 1.9% (.7% - 4.1%) Among HIV infected: 3.4% HIV uninfected: 2.5%
Sensitivity and Specificity of the Screening tool Four symptom screening tool Sensitivity: 14.3% (3.05% - 36.1%) Specificity: 82.2% (79.1% - 84.2%) PPV: 2.4% NPV: 96.9% All the symptoms with the exclusion of cough had a sensitivity of 0%. Cough had a sensitivity of 14.3% (3.05% - 36.3%) No significant difference whether HIV infected or not, pregnant or lactating.
Sensitivity of diagnostic tools in addition to the TB screening tool using TB culture as the Gold Standard Test Sensitivity Specificity PPV NPV Xpert MTB 0% 98.39% 2.1% 97.9% Smear Microscopy 9.52% (1.1% - 30.38%) 99.71% (98.9% - 99.96%) IGRA 30% (6.67% - 65.25%) 65.78% (57.13% - 68.19%) TST 27.78% (9.69% - 53.48%) History of contact 4.76% (0.12% - 23.82%) 67.11% (62.93% - 71.09%) 77.53% (74.2% - 80.62%) TB Lam 10% (0.25% - 44%) 94.01% (90.8% - 96.35%) 50% (6.76% - 93.24%) 2.54% (0.53% 7.25%) 2.78% (0.9% - 6.36%) 0.65% (0.02% - 3.06%) 2.93% (1.35% - 5.49%) 97.2% (95.7% - 98.36%) 96.52% (92.96% 98.59%) 96.49% (94.07% - 98.12%) 96.35% (94.0% - 97.76%) 97.07% (94.51% - 98.65%) Both IGRA and TST were the most valuable additional diagnostic tool among the women since they were more likely to pick up more women with active Tuberculosis disease
Sensitivity of diagnostic tools in addition to the TB screening tool compared to HIV status Diagnostic test (disaggregated by HIV status) # diagnosed against Gold Standard Culture 21 15 Pregnant, 6 Lactating 11 HIV+, 9 HIV- P value Sensitivity Specificity Gold Standard TB Lam 1 (HIV+) <0.001 11.11% ** MTB rif 0 0.286 0% 100%** Smear Microscopy 2 >0.99 8.3% 88.89% IGRA 3 (HIV+) <0.001 30% ** TST 6 (1:1 HIV+:HIV-) 0.32 18.8% 57.14% Among HIV infected women, the IGRA was more sensitive to detecting women who had active TB disease than the other tools.
Conclusion The 4 question screening tool misses out 85% of all TB cases. Additional tests to the screening questionnaire such as TST and IGRA (for HIV+ women) improve the sensitivity to 30% There is need to find more sensitive diagnostic tests for pregnant and lactating women to improve early diagnosis of TB.
Acknowledgements MoH Swaziland PEPFAR Staff of MCNCH clinics of Mbabane, Raleigh Fitkin Memorial and Hlathikulu hospitals Study participants URC