Conclusion
Our results show high agreement between LFA, LA and India ink in different samples and a perfect agreement between LFA in different samples. The high agreement shows that LFA is a reliable POC diagnostic test. The results on individual tests show that there was almost perfect agreement between LFA and LA on CSF and serum. The test demonstrated high level of sensitivity and specificity of LFA compared to LA on sera and CSF. These findings are consistent with similar studies conducted in South Africa and USA that show high sensitivity using CSF and serum (12,18). Comparable results were reported in a study on multisite validation of cryptococcal antigen lateral flow assay in Uganda and South Africa (7). The strong agreement between the LFA and LA tests is an indicator that LFA test on whole blood, CSF and serum is as good as LA test on sera and CSF.
The findings from comparison of LFA to India ink microscopy using CSF demonstrated high sensitivity, specificity, and predictive values. This is in contrast to the findings from the expert opinion and other studies that documented lower sensitivity and NPV for CSF microscopy against LFA (7,9). The India ink microscopy requires laboratory infrastructure, dependent on fungal concentration and is highly operator dependent rather than the test performance.
On comparison of LFA to culture using CSF, there was high sensitivity, low specificity, and moderate agreement with a weak kappa value. The findings were consistent with other studies that documented high sensitivity and low specificity (14,19). The findings on high sensitivity, low specificity and a weak kappa value were similarly demonstrated when CSF culture was compared to LA and India ink using CSF. However, other studies documented low sensitivity in CSF culture when compared against other diagnostic tests (7–9).
LFA on capillary blood was compared with LFA on serum and CSF. The LFA results on capillary blood had an ideal concordance with LFA serum and CSF results. LFA had a very high positive and negative predictive values both on serum and CSF, a characteristic that makes it good for an accurate diagnosis of cryptococcal meningitis. The high sensitivity and specificity of the test and its ability to be easily performed at the bedside and giving accurate results rapidly allows for prompt and timely initiation of treatment (9). The findings were comparable with a similar study on evaluation of LFA using serum, CSF and capillary blood in Uganda (15,18,20).
Recommendation
The evidence of the perfect agreement between LFA on capillary blood, serum and CSF, high sensitivity and specificity, ease of performance, along with rapid results may indicate LFA using capillary blood POC test as the method of choice for CM diagnosis. Although LFA meets World Health Organization assured criteria for POC diagnostic tests in resource-limited settings, we recommend use of LFA test as a POC test in resource limited settings for the diagnosis of CM.
Limitation of the study
The limitations of this study include participants not yielding CSF sample due to dry taps, thus reducing the CSF samples that were analysed. The difference on CSF samples analysed had no major implications to the overall evaluation since there were more than one sample type used in the evaluation. Capillary blood could only be used on LFA test thus, there was no uniform use of the sample type across other testing procedures.