Date Published: November 10, 2009
Publisher: Public Library of Science
Author(s): Andrew Ramsay, Luis E. Cuevas, Catherine J. F. Mundy, Carl-Michael Nathanson, Petros Chirambo, Russell Dacombe, S. Bertel Squire, Felix M. L. Salaniponi, Sera Munthali, Ben Marais. http://doi.org/10.1371/journal.pone.0007760
Abstract: To quantify the likely impact of recent WHO policy recommendations regarding smear microscopy and the introduction of appropriate low-cost fluorescence microscopy on a) case detection and b) laboratory workload.
Partial Text: There are an estimated 9.3 million new cases of tuberculosis (TB) each year, with the majority occurring in low- and middle-income countries (LMICs) . Most LMICs rely almost entirely on direct sputum smear microscopy (DSSM) for routine TB diagnostic services . The 2008 data indicate that 1.4 million of the TB cases were in people living with HIV/AIDS (PLWHA) who are particularly likely to be paucibacillary or smear-negative , .
This retrospective analysis used data from the laboratory register of Bwaila Hospital in Lilongwe District, Central Malawi . The hospital serves urban and rural populations and provides essential medical laboratory services including DSSM, malaria microscopy, basic haematology and clinical chemistry, blood-grouping and cross-matching and stool and urine analysis. At the time of the audit the laboratory was staffed by two professional laboratory workers (one lab technician and one lab assistant), three health surveillance assistants and one cleaner. One health surveillance assistant was responsible for sputum specimen registration, smear preparation and staining (non-microscopy) on five full days per week. One professional laboratory worker was responsible for microscopy on 4.5 days per week, since the week’s first smears prepared for microscopy become available around mid-day on a Monday.
A total of 2449 patients were registered with a median of 86 patients (range 48–178) and 239 (range 140–482) smears examined per week. Of these, 1920 patients were TB suspects and 529 follow ups. In total 6,796 smears, 5,730 (84.3%) from TB suspects and 1,066 (15.7%) from follow-up patients, were examined. Results from thirty smears from TB suspects were not recorded (4 of 128 smear-positive males and 6 of 72 smear-positive females did not submit the third specimen, the remaining 21 missing smears resulted from between 10 and 20 smear-negative TB suspects submitting only 1 or 2 specimens). Eight follow-up patients had 3 smear examinations recorded possibly in error. Five hundred and sixty one (9.7%) of the 5,730 smears from TB suspects and 49 (4.6%) of the 1,066 smears from follow-up patients had ≥1 AFB resulting in 201 (10.5%) and 26 (4.9%) of the TB suspects and follow up patients being smear-positive. A median of 7 (range 3–13) TB suspects and 1 (range 0–3) follow up patient were positive per week and 19 (range 9–38) and 2 (range 0–6) smears from TB suspect and follow up patients were positive each week.
At the time of this audit, two positive smears were required to define a smear-positive case based on microscopy alone in Malawi and the required grading of the smears was not stated in its national guidelines. A person with a single positive smear could be categorized as a smear-positive case if they had radiographic abnormalities consistent with TB, or had a positive TB culture. TB culture was not routinely available for the diagnosis of new cases and many diagnostic centres did not have ready access to chest radiography. The situation is the same in 2009. This audit indicates that adoption of the revised WHO smear-positive case definition could significantly increase case detection in men and women by up to 28% at centres without access to chest radiography or TB culture depending on the grade of smear currently accepted as positive. In contrast to a prospective study conducted in Kenya, this retrospective analysis found that the increase in case detection may be significantly higher in men than in women . An adequate explanation for the difference in findings between these two studies, conducted in similar epidemiological settings, is not immediately obvious.