**7.2. Enrolment for the study.**

Patients from the outpatient department were systematically screened for symptoms suspected to be malaria and referred to the research clinic. Inclusion criteria were a clinical suspicion of malaria; weight≥ 5kg; resident in Mbarara Municipality available for two weeks follow up period; and signed informed consent from the study subjects or their legal guardians. Exclusion criteria were signs of severe or complicated malaria [15b], signs of severe disease; and women with visible pregnancy or suspicion of pregnancy based on the assessment of the last normal menstrual period.

### **7.3. Sample size**

The required number of patients with positive blood film was calculated using an estimated sensitivity of the RDTs of 90%, an alpha error of 0.05 and a precision of 6%. This number (n-

96) was doubled to permit a stratified analysis by age group (0--4 and ≥5 years. The same parameters were used to calculate the required number of patients with a negative blood film, thus giving a final minimum sample size of 200 blood-film-positive and 200 blood – film –negative patients.

Functions of Dehydrogenases in Health and Disease 173

the end of the study, two test readers and two laboratory technicians involved in preparing the tests completed a questionnaire concerning the ease of use and interpretation of each

Analysis. All data were either recorded directly or transcribed from source data forms to an individually numbered case report form (CRF). Data were double entered and validated using EpiData version 3.1(EpiData Association,Odense, Denmark)and analysed using Stata 9.1 (Stata Corp. college station,TX,USA).The study profile and base-line characteristics were summarized, including comparative tests between age groups (x 2 test ,Ma-Whitey U test). The validity for each test was calculated overall and then stratified by age group, level of parasitemia(parasites/µl 1---99,≥100,≥200,≥500), presence/absence of fever, duration of illness(0—2vs.3 days and above) and a history of taking antimalarials, using comparative tests( x 2 test, Mann-Whitney U test) to compare differences between groups. Kappa statistics were calculated for iter-reader reliability for each test on the day of diagnosis. A test was considered as reliable if k ≥ 0.8. Univariable and multivariate analyses were performed to investigate the association between explanatory factors and the test remaining

Results. Demographical and Parasitological characteristics of study subjects.

had higher parasite densities than those of the other two species.

Between 26 April and 27th July, 2005, 485 patients from the out-patient department were screened. Nine were ineligible (three had severe illness, five were non-residents and one was not in the appropriate age group after completion of recruitment in the under fives. Sixteen patients did not consent to participate in the study; 239 under fives and 221 aged 5 years and above. The mean age was 12 years(SD 13years;Table 2). There were 248 positive blood films with P. falciparum monoinfections(93.6%), P. malariae monoinfections(2.4%), P. falciparum+ P. malariae mixed infections(0.8%) and P. falciparum+ P. vivax mixed infections(0.8%). Of the 212 negative films, nine had gametocytes present. Parasitological characteristics of positive subjects are given in Table 2. Slides positive with P. falciparum

Only Carestart had estimates for all validity parameters greater than 90% (Table 3). Vistapan and Carestart were as sensitive as Paracheck-Pf(P=0.14 and P=0.38 respectively. Parabank was less sensitive than all other tests (P‹0.001 for each comparison). There was no significant difference in specificity between the three pLDH tests, but Parabank had a higher specificity compared with Paracheck.Pf (P= 0.02) for P. falciparum detection. Sensitivity decreased with older age for both Vistapan [97.7% (under fives) vs 85.7%, P‹0.01] and Parabank[95.4%( under fives)vs73.1%, P‹0.001]. Sensitivity increased with axillary temperature≥ 37.5 0C at inclusion for Paracheck.Pf (98.8 vs91.4%, P=0.04), Vistapan(97.6 vs 89.0%, P=0.03) and Parabank(91.8vs81.0%, P=0.04) compared with patients with axillary temperature ‹37.5 0C. Although, the small number of non-falciparum monoinfections does not permit reliable calculations of validity of non-falciparum mutants, all tests detected 100 % (n=6) of the P. malariae monoinfections. Plasmodium vivax was detected in 4/6 infections by Carestart, 2/6

test.

positive at each follow- up visit.

Validity of RDTs.

by Vistapan and 1/6 by Parabank.
