Menu
Menu
Menu
Menu
Background: Tuberculosis diagnosis in pregnancy is complex because tuberculosis symptoms are often masked by physiological symptoms of pregnancy. Untreated tuberculosis in pregnant and postpartum women may lead to maternal morbidity and low birth weight. Tuberculosis in HIV-positive pregnant women increases the risk of maternal and infant mortality. Objective: This study aimed to determine tuberculosis prevalence stratified by HIV status and identify screening algorithms that maximise detection of active tuberculosis among pregnant and postpartum women in Eswatini. Methods: Women were enrolled at antenatal and postnatal clinics in Eswatini for tuberculosis screening and diagnostic investigations from 01 April to 30 November 2015 in a cross-sectional study. Sputum samples were collected from all participants for tuberculosis diagnostic tests (smear microscopy, GeneXpert, MGIT culture). Blood and urine samples were collected from HIV-positive women for cluster-of-differentiation-4 cell count, interferon gamma release assay and tuberculosis lateral flow urine lipoarabinomannan tests. Results: We enrolled 990 women; 52% were pregnant and 47% were HIV-positive. The prevalence of tuberculosis among HIV-positive pregnant women was 5% (95% confidence interval [CI]: 2-7) and among postpartum women it was 1% (95%CI: -1-3). Tuberculosis prevalence was 2% (95%CI: 0-3) in HIV-negative pregnant women and 1% (95%CI: -1-2) in HIV-negative postpartum women. The national tuberculosis symptom screening tool failed to identify women who tested tuberculosis-culture positive. Conclusion: Routine tuberculosis symptom screening alone is insufficient to rule out tuberculosis in pregnant and postpartum women. Only sputum culture maximised the detection of tuberculosis, indicating a need to balance access and cost in developing countries. Copyright:
Ethical approval was obtained from the Eswatini National Health Research Board (formerly Scientific and Ethics Committee [Approval reference: MH/599C]), CDC Institutional Review Board (IRB) (Reference: CGH-HSR Tracking #: 2015-196), and University Research Co. LLC IRB (02 March 2015). Participants signed informed consent written in their preferred language (SiSwati or English). Participants received no incentives but were reimbursed transport costs for additional visits to read TST. Diagnostics tests and treatment, if required, were free of out-of-pocket charges. We conducted a cross-sectional study enrolling pregnant and postpartum women, aged 18 years and older, attending antenatal and postnatal care clinics, from 01 April to 30 November 2015 at three public health facilities in three of the four regions of Eswatini. Sociodemographic and clinical data, including past tuberculosis screening results where applicable, were collected. Eswatini is categorised by WHO as a high tuberculosis/HIV burdened country with a co-infection rate of 70% and a tuberculosis incidence rate of 398 per 100 000 population.27 Participants were pregnant and postpartum women who were not on anti-tuberculosis treatment at enrolment or who had not taken anti-tuberculosis medicines, including isoniazid for tuberculosis preventive therapy, within the 2 months preceding enrolment, based on documented evidence from patient clinical records, and who provided informed consent. Four groups of women were enrolled: HIV-positive pregnant, HIV-negative pregnant, HIV-positive postpartum, and HIV-negative postpartum. A sample size of 183 in each group was determined and a full narrative of sample size calculation is fully described in our protocol paper titled ‘Screening in Maternity to Ascertain Tuberculosis Status (SMATS) study’.4 Participants were consecutively enrolled until the sample size was reached. All participants were screened using the WHO-recommended national tuberculosis four-symptom screening (standard NTBSS) tool. Participants held clinic cards which were checked for evidence of tuberculosis symptom screening at their last clinical encounter to ascertain routine tuberculosis screening coverage at their previous visit. Participants were screened as positive using the standard NTBSS tool, if they had a cough lasting at least 2 weeks,16 or a cough lasting less than 2 weeks plus any other symptom of fever or unexplained loss of weight or night sweats, or if any two symptoms were present.16,17 Enhancements of the tuberculosis screening tool were done by adding to the four symptom screening tool any history of contact with a person on tuberculosis treatment or who had been diagnosed with tuberculosis and the presence of tuberculosis symptoms within the household inhabitants. We measured sensitivity, specificity and predictive values (positive and negative) of the WHO-recommended four symptom tuberculosis screening tool among HIV-positive and HIV-negative pregnant and postpartum women compared with sputum culture, the gold standard for M. tuberculosis detection. Even though chest radiographs are not contraindicated in pregnancy, chest radiographs were only carried out among postpartum women (both HIV-positive and HIV-negative) to eliminate risk of radiation exposure to the foetus. Two samples of sputum (for Xpert® MTB/RIF, smear microscopy and culture using BACTEC TM MGIT 960) were collected from all participants.4 Sputum samples were collected through the production of spontaneous self-expectorated phlegm (preferred method), sputum induction through nebulising with hypertonic saline or, if both the above failed, the participant received a sputum container to take home and attempt to produce an early morning sputum sample and bring it back to the health facility. Testing of sputum samples was done at the National Tuberculosis Reference Laboratory in Mbabane. Tuberculosis culture was the gold standard test and in situations where sputum samples were insufficient for the three tests, culture was prioritised ahead of Xpert® MTB/RIF and smear microscopy. A urine sample for the LF-LAM test and two 4 mL blood samples for IGRA testing and cluster-of-differentiation-4 (CD4) cell count testing were collected. Interferon-gamma release assays and LF-LAM were only done for HIV-positive women due to limited evidence on IGRA use28 and existing WHO recommendations on LF-LAM use in HIV-positive individuals.29 TST was also done and the induration was read after 48 hours – 72 hours. IGRA and TST procedures were explained to all participants and only HIV-positive participants were then asked which test they preferred between IGRA and TST. Sample collection and storage followed national standard operating procedures for urine and blood collection and manufacturer’s instructions for the DetermineTM Tuberculosis LF-LAM test (Abbott Laboratories, Lake Bluff, Illinois, United States) and IGRA testing.4 All sputum specimens were analysed by means of: (1) Xpert® MTB/RIF assay (Cepheid, Sunnyvale, California, United States), (2) concentrated Ziehl-Neelsen microscopy, (3) liquid–medium culture method (BACTECTM MGIT 960TM TB Diagnostic System; Becton, Dickinson & Company [BD], Crystal Lake, New Jersey, United States), and (4) MGIT 960TM DST (BD, Crystal Lake, New Jersey, United States). Positive cultures were identified as M. tuberculosis using the tuberculosis Ag MPT64 Rapid® assay (Standard Diagnostics, Inc., Yongin, South Korea).30 Interferon-gamma release assays and CD4 cell count testing were done at Lancet Laboratory in Johannesburg, South Africa. IGRA blood specimens were collected directly into IGRA tubes used for QuantiFERON-TB Gold in-Tube assay (Cellestis Ltd, Carnegie, Victoria, Australia).31,32 CD4 cell count tests were conducted by BD FACSCalibur™ flow cytometry (BD Biosciences, San Jose, California, United States) using venous blood collected in sterile four millilitre BD Vacutainer EDTA tubes by trained study nurses. The data collection tools were matched to the tools used for routine data collection at health facilities. Demographic fields in the data collection forms were adapted from client cards. Data fields for tuberculosis symptom screening were adapted from the national tuberculosis screening tool. Patient information was anonymised. Data were entered in Epi Info™ (Centers for Disease Control and Prevention, Atlanta, Georgia, United States) and Research Electronic Data Capture (REDCap; Vanderbilt University, Nashville, Tennessee, United States), and data extraction tools were cross-checked to validate conflicting fields. Laboratory results were compared with the electronic study results file generated from the laboratory, and participant identity numbers were used to relate the data. We evaluated both option one and option two of the WHO four symptom screening algorithms. TST numeric readings were recoded as positive, if the length of the induration was ≥ 5 mm for HIV-positive participants or ≥ 10 mm, if the participant was HIV-negative.32,33 All other lengths, including 0 mm, were recoded as negative according to existing literature and CDC guidance on interpretation of TST results.32,33 IGRA was done at a private laboratory according to manufacturer’s recommendations and the differences in readings between QuantiFERON-tuberculosis Gold in-Tube tuberculosis antigen, tuberculosis nil and tuberculosis mitogen were used to interpret positive, negative and indeterminate results, respectively (Figure 1).34 Interpretation of interferon-gamma release assays results. Frequencies and proportions were used to describe participant characteristics and related clinical data. Diagnostic parameters of sensitivity, specificity and positive and negative predictive values analyses for tuberculosis symptoms, and tuberculosis diagnostic tests were calculated in STATA version 13 (© 1985–2013 StataCorp LLC, College Station, Texas, United States). Using logistic regression, associations between culture-positive tuberculosis and HIV status and pregnancy or postpartum status variables were determined. Other sociodemographic and clinical factors were considered for inclusion in the multivariate model if the p-value was ≤ 0.1 on bivariate analysis. Factors that perfectly predicted the outcome were excluded. Estimates were reported with 95% confidence intervals (CIs) and corresponding p-values.
N/A
