Summary of studies evaluating the malaria diagnostic accuracy of Cell-Dyn series analysers using the side-scatter/depolarized side-scatter plot abnormal depolarizing events criterion – Including observations for each study and blinding. First author, year Number of participants and Standard reference Sensitivity Specificity ¶ Blinding Observations and country diagnoses test Index test criterion % % Total: 224 directed samples from 175 Mendelow, 1999, Microscopy*, PfHRP2 CD** 3500 Race differences, probably in patients, P. falciparum: 93, Species + 72 96 South Africa [18] (Para-sight F) relation to immunity. not specified: 2 ≥1 depolarizing events‡ Hänscheid, 2001, Total: 174, P. falciparum: 48, P. CD 3500 † Microscopy + 95 88 Portugal [39] vivax: 6, P. ovale: 1, P. malariae: 2 ≥2 depolarizing events Depolarizing purple events above Total: 113, P. falciparum: 46, P. CD 3500 a line that starts horizontally Wever, 2002, The vivax: 5, P. ovale: 4, no Microscopy, QBC Either ≥1 depolarizing + 62 96 through granularity signal channel Netherlands† [36] differentiation for P. vivax or P. events or 25 and then continues at 22.5° ovale: 3 pseudoreticulocytosis angle with the x axis.

Grobusch, 2003, Total: 403, P. falciparum: 87, P. Semi-immune sensitivity: 73.7% Microscopy, PCR CD 3000 + 48.6 96.2 Germany† [27] vivax: 13, P. ovale: 5, P. malariae: 2 ≥1 depolarizing events No- immune sensitivity: 28.6% Depolarizing purple events above Total: 831, P. falciparum: 334, P. Microscopy, HRP2 Scott , 2003, South CD 4000 channel 25 and within the side- vivax: 7, P. ovale: 1, P. malariae: 2, (Pf/Pv ICT), P-LDH + 80.2 87.3 Africa [35] scatter/depolarized side-scatter mixed or unspecified: 6 (OptiMAL), PCR ≥1 depolarizing events boundary shown in Figure 1 Suh, 2003, South Considered black or green coded Total: 168, P. vivax: 68 Microscopy CD 4000 - 91.2 100 Korea [32] ≥1 depolarizing events events Total: 799 (directed: suspected of QBC, PfHRP2 Directed Directed Depolarizing events above Dromigny, 2005, malaria 123, non-suspected random (MAKROmed), channel 50 and within the side- CD 3200 - 92.9 93.8 Senegal [30] samples 676) P. falciparum: 68, Microscopy, nested ≥1 depolarizing events Random Random scatter/depolarized side-scatter treated or subclinical: 83 PCR 90.2 96.7 boundary shown in Figure 1 Padial, 2005, Total: 411, P. falciparum: 35 , P. no exact definition of CD positive Equatorial Guinea Microscopy, PCR - 72 98 ovale: 3, mixed: 1 CD 4000 criterion given [41] Josephine, 2005, Total: 889, P. vivax: 12, P. malariae: Microscopy (peripheral no exact definition of CD positive - 100 100 Malaysia [40] 3, P. falciparum: 1 blood smear) CD 4000 criterion given de Langen, 2006, Microscopy , PfHRP2 CD 3700 no exact definition of CD positive Total: 208, P. falciparum: 90 - 93 97 Namibia [42] (Immuno-Mal) ≥1 depolarizing events criterion given CD 3000*** Hänscheid, 2008, Children, total: 368, P. falciparum: Specialized strategy for each type Microscopy ≥1 depolarizing purple - a) 96% a) 96% Gabon [34] 152 events of event*** Green-coded events b) 85% b) 96% Hänscheid, 2009, Pregnant patients, total 685, P. CD 3000 Microscopy - 86.8 78.5 Gabon [43] falciparum: 86 ≥1 depolarizing events Rathod, 2009, India Total: 523, P. falciparum :73, P. CD 3200 P. vivax :sensitivity: 63.0%, Microscopy - 62.2 25.3 [33] vivax: 62 ≥1 depolarizing events specificity:61.3% ¶Index diagnostic test: abnormal depolarizing events in the side-scatter/depolarized side-scatter plot. ‡All studies use the instrument’s diagonal separation line for eosinophils and neutrophils in the side-scatter/depolarized side-scatter plot , unless otherwise specified. *Microscopy for all studies corresponds to thick film evaluation except where otherwise specified. **CD: Cell-Dyn. IFAT: indirect fluorescence antibody test. QBC: Quantitative buffy coat. PCR: Polymerase chain reaction. ICT: immunochromatography. †Imported malaria. ***For purple events, these were considered positive if present above a line traced at 5 pixels from the x axis. For green events, a special gate was created to identify haemozoin-laden granulocytes, with the intention to exclude eosinophils. In accordance with studies using flow cytometric cell sorting [27], the largest possible gate to the left and above the usual location of the eosinophil population was created which did not contain any eosinophils [34]. For this, CBC analyses from children without malaria or pseudoreticulocytosis were used [34].