Inter-observer agreement according to malaria parasite density

Background: Recent developments in diagnostic techniques for malaria, particularly DNA probes and sero-immunology, have raised questions as to how these techniques might be used to facilitate malaria diagnosis at the most peripheral levels of the primary health care system. At present, malaria diagnosis is based on the standard microscopic examination of blood films in most field epidemiologic studies and is likely to remain so in the immediate future in Africa. The objective of this study was to assess inter-observer agreement for the examination of Giemsa-stained slides for Plasmodium falciparum parasites. Methods: Children aged 0 to 10 years were enrolled yearly in Bancoumana village (West Africa), mainly during the transmission season (June to October). The blood smears obtained from the persistently negative children in June 1996, August 1996, October 1996 and March 1997 were systematically re-examined. A stratified random sample (10%) proportional to the following parasite density classes 1-100, 101-5000, and 5001 and over was taken from the slides collected. The kappa statistics and the intra-class correlation were used as measures of agreement the first and the second slide examinations. Results: The weighted kappa statistic, widely used as a chance-corrected measure for nominal agreement, showed excellent inter-observer agreement (K-w= 0.7926; 95% CI [0.7588, 0.8263]; p= 0.01). The intra-class correlation co-efficient had the same value of 0.7926 confirming the appropriateness of the weighted kappa statistic. Interobserver agreement for slides read as negative by one observer, or as containing more than 100 parasites per mu l, was excellent: 97% (493/506) and 92% (145/158), respectively. In contrast, the inter-observer agreement for slides read by one observer as containing 1-100 parasites/mu l was poor, 36% (96/268). Conclusions: In field conditions in Mali, there was a high reproducibility for slides reported as negative or as having more than 100 parasites per mu l. However, smears with readings of 1-100 parasites per mu l were less reproducible and should be re-examined carefully.