Maximizing microscopy as a diagnostic tool in peripheral health centres of BU endemic areas in Ghana.

Buruli ulcer (BU) disease, a skin condition caused by Mycobacterium ulcerans (M. ulcerans) is endemic in remote rural areas. Disease diagnosis on clinical basis alone can be misleading, requiring definitive diagnosis based on laboratory tests. Resource constraints in BU endemic areas make microscopy for the detection of acid fast bacilli (AFB) an important and useful method. It is rapid, user-friendly, convenient and cheap. Despite its usefulness, its performance is relatively low. This study investigated modifications of the current method aimed at improving its performance. Forty (IS) 2404 polymerase chain reactions (PCR) positive BU samples were processed by eight physical (centrifugation and overnight sedimentation) and chemical (phenol ammonium sulphate and sodium hypochlorite) modifications of the current direct method. Assessments were based on standard AFB evaluation coupled with in house criteria; positivity (P), clarity and contrast (C) release of bacilli from specimen (R). Overall AFB positivity rate was 64% (409/640). Each protocol had 80 smears. The percentage positivity (P) for the conventional method was 58% (46/80) smears. The highest positivity rate of 57/80 (%) was by protocol 7 (5% phenol in 4% ammonium sulphate (PhAS) and concentrated by overnight gravitational sedimentation). The least positivity rate at 35% (28/80) was by protocol 1 (smears from direct application of swab tips). The differences in performance between the two chemical tested; 5% phenol in 4% ammonium sulphate (PhAS) and 3.5% NaHOCl was significant (p<0.05). The differences between the two physical methods were however not significant (p>0.05). This study concluded that BU samples treated with a solution of 5% phenol in 4% ammonium sulphate and concentrated by either centrifugation or overnight sedimentation is useful for maximizing AFB detection by bright field microscopy. This can be useful in rural health facilities with resource constraints.