External quality assessment of malaria microscopy in the Democratic Republic of the Congo

Abstract

Background: External quality assessments (EQA) are an alternative to cross-checking of blood slides in the quality control of malaria microscopy. This study reports the findings of an EQA of malaria microscopy in the Democratic Republic of the Congo (DRC).

Methods: After validation, an EQA slide panel and a questionnaire were delivered to diagnostic laboratories in four provinces of DRC. The panel included three samples for diagnosis (sample 1: Plasmodium falciparum, 177,000/μl, sample 2: P. falciparum, 2,500/μl, sample 3: no parasites seen), one didactic sample (Howell-Jolly bodies) and one sample for assessing the quality of staining. Participating laboratories were addressed and selected through the network of the National Tuberculosis Control Programme. Participants were asked to return the responses together with a stained thin and thick blood film for evaluation of Giemsa stain quality.

Results: Among 174 participants (response rate 95.1%), 26.2% scored samples 1, 2 and 3 correctly and 34.3%, 21.5% and 5.8% of participants reported major errors in one, two or three samples respectively. Major errors included reporting "no malaria" or "non-falciparum malaria" for Plasmodium falciparum-positive samples 1 and 2 (16.1% and 34.9% of participants respectively) and "P. falciparum" for Plasmodium negative sample 3 (24.0%). Howell-Jolly bodies (didactic sample) were not recognized by any of the participants but reported as "P. falciparum" by 16.7% of participants. With parasite density expressed according to the "plus system", 16.1% and 21.5% of participants scored one "+" different from the reference score for samples 1 and 2 respectively and 9.7% and 2.9% participants scored more than two "+" different. When expressed as counts of asexual parasites/μl, more than two-thirds of results were outside the mean ± 2SD reference values. The quality of the Giemsa stain was poor, with less than 20% slides complying with all criteria assessed. Only one quarter of participants purchase Giemsa stain from suppliers of documented reliability and half of participants use a buffered staining solution. One third of participants had participated in a formal training about malaria diagnosis, half of them earlier than 2007.

Conclusion: The present EQA revealed a poor quality of malaria microscopy in DRC.

Figure 1

Distribution of parasite densities expressed per μl of blood for sample 1 and 44 participants. Mean and SD values refer to the expected result as counted by expert microscopy.

Figure 2

Distribution of parasite densities expressed per μl of blood for sample 2 and 41 participants. Mean and SD values refer to the expected result as counted by expert microscopy.

Figure 3

Distribution of slide positive rates among the participants (168 valid answers, data represent numbers).

Figure 4

Numbers of samples processed monthly by the participants associated with the percentage of correct answers for the three diagnostic samples.