Hello and thanks for your question.

To start, please check that you are using the correct datafile. The data on malaria parasitemia tests for individual children are available in the person recode file (PR) which contains data for every individual in a household. For the Ghana 2014 DHS this is the GHPR70FL file.

The standard variables to check first for data on malaria parasitemia test results are hml33 (result of malaria measurement), hml32 (final results of malaria from blood smear) and hml35 (result of malaria rapid test). It is always possible that non-standard, country-specific variables are available in the dataset. In this case, a search for "mal" turns up several country specific variables related to malaria testing (sh212c, sh212f7, sh212i, sh212q).

I'm not sure where you are seeing that only 33 blood tests were done. When I tabulate hml33 I see that 3,192 of 3,344 children had malaria measurements. Tabulating hml35 shows a distribution of 58% negative and 42% positive RDT test results from 3,191 children. Tabulating sh212c gives us the results of the species classification from the RDT results (59% Pf, 3% pan, 37% Pf and pan out of 1,346).

We have noticed that the standard recode variable for microscopy data, hml32, is unpopulated with data in this dataset and we are exploring the reasons for this. Microscopy testing was done in the survey and the results are available in the survey's final report. Once we have identified the problem and corrected the recode file we will make it available to users and you should be automatically notified that a new version of the data are available for download.