Potential overestimation of parasite loss in thick smears
Georges Snounou, CNRS France and NUS Singapore
21 October 2008
The loss of parasite material during staining had been recognized early on by malariologists when it was noted that parasites transferred from smears of malaria-infected persons to those of uninfected individuals when staining was carried out in the same container (Brooke and Donaldson, 1948, Public Health Reports, 63:991).
The molecular analysis presented in the article by Bejon et al. led them to conclude that thick smears counts underestimate the true parasite levels by ten-fold (the log difference from Fig. 1). This implies that 9 out of 10 parasites are lost when the thick smear is stained (the authors discount the probability of parasites being obscured in the thick smear). The calculations were based on a the assumption that 100 high powered fields correspond to 1 microlitre of blood. In the reference quoted for this value ([7] Greenwood and Armstrong, 1991, Transactions of the Royal Society of Tropical Medicine and Hygiene, 85:186), the equivalent figure was 500 high powered fields.
We suggest the value of 500 is the correct one, thus the actual loss in the sensitivity of thick smear microscopy for parasite load evaluation would be two-fold rather than ten-fold. We feel that the loss observed is mainly due to difficulties in identifying parasites when these are present in very low numbers in a thick smear, rather than because of a massive loss of parasite material during Giemsa staining. This could be easily confirmed by comparing DNA content in blood at high parasite levels (> 10 000 parasites per ml of blood) with thick smear counts.
In conclusion, we agree that thick smears at low parasite densities underestimate true parasite loads but not to the extent concluded in the article. On the other hand we agree that the more sensitive molecular techniques of parasite detection should be applied when measuring intervention outcomes, such as those of vaccination studies.
Georges Snounou (CNRS, Paris, France; National University of Singapore, Singapore) and Mehul Dhorda (Epicentre, Mbarara, Uganda)
Potential overestimation of parasite loss in thick smears
21 October 2008
The loss of parasite material during staining had been recognized early on by malariologists when it was noted that parasites transferred from smears of malaria-infected persons to those of uninfected individuals when staining was carried out in the same container (Brooke and Donaldson, 1948, Public Health Reports, 63:991).
The molecular analysis presented in the article by Bejon et al. led them to conclude that thick smears counts underestimate the true parasite levels by ten-fold (the log difference from Fig. 1). This implies that 9 out of 10 parasites are lost when the thick smear is stained (the authors discount the probability of parasites being obscured in the thick smear). The calculations were based on a the assumption that 100 high powered fields correspond to 1 microlitre of blood. In the reference quoted for this value ([7] Greenwood and Armstrong, 1991, Transactions of the Royal Society of Tropical Medicine and Hygiene, 85:186), the equivalent figure was 500 high powered fields.
We suggest the value of 500 is the correct one, thus the actual loss in the sensitivity of thick smear microscopy for parasite load evaluation would be two-fold rather than ten-fold. We feel that the loss observed is mainly due to difficulties in identifying parasites when these are present in very low numbers in a thick smear, rather than because of a massive loss of parasite material during Giemsa staining. This could be easily confirmed by comparing DNA content in blood at high parasite levels (> 10 000 parasites per ml of blood) with thick smear counts.
In conclusion, we agree that thick smears at low parasite densities underestimate true parasite loads but not to the extent concluded in the article. On the other hand we agree that the more sensitive molecular techniques of parasite detection should be applied when measuring intervention outcomes, such as those of vaccination studies.
Georges Snounou (CNRS, Paris, France; National University of Singapore, Singapore) and Mehul Dhorda (Epicentre, Mbarara, Uganda)
Competing interests
No conflict of interest