This study explored the validity of self-reported sulfadoxine-pyrimethamine IPTp by testing for presence of SDX in maternal blood at delivery using HPLC. Two main findings of this study are that self-report on sulphadoxine-pyrimethamine IPTp use is unreliable not only for knowing whether the pregnant patient took the SP or not but also for finding out when the patient took the drug. Several patients who reported not having taken SP were found to have the drug derivatives in their blood. Further, some patients who reported having taken the drug before 9 weeks preceding baby delivery (when SDX would be too low to be detected in blood by HPLC) were also found to have the drug in the blood. On the other hand, some patients claiming to have taken SP within 9 weeks before delivery (when blood SDX would be detected by HPLC) actually did not have detectable SDX blood levels.
Interestingly, participants who self-reported IPTp use during their present pregnancy were more likely to have SDX in their circulation at delivery, although the level of agreement was only slight as assessed by kappa statistics. Although only 29.2% of participants who reported IPTp use actually had SDX in their blood at the time of delivery, 25% of participants who reported not taking IPTp had SDX in blood at delivery. This finding questions the validity of self-reported data in estimating the IPTp coverage. The findings of this study concur with a previous study in Uganda which found low validity of caretakers’ report on use of anti-malarials and antibiotics [
11]. Okura
et al. found that the young and more educated were more likely to report correctly on the prevalence of diseases such as hypertension, diabetes and history of myocardial infarction [
19]. Another study found a good agreement of self report with diagnosis of diabetes and hypertension [
13]. In another study in Kenya, which looked at anti-malarial drugs before initiating treatment in participants who reported no use of drug in 28 days prior to enrolment, it was found that the proportion of participants with residual anti-malarials was high and self-report on drug intake was unreliable [
12]. Yet another study found that self reported compliance in use of antibiotics among sexually transmitted disease patients was also unreliable [
20]. Thus, several previous studies concur with the findings of the present study where self-report data on IPTp use is only in slight agreement (kappa = 0.03) with results of HPLC detection of the drug ingredients in blood. In view of the demonstrable weakness of self reports, a previous study has suggested increasing the validity of self-reported data through focus group discussions, using language with which the respondents are very familiar, sequencing the questions from the least to the most threatening, using open-ended and direct questions [
21]. In a review on validity of self reported data, Brener
et al.[
22] noted that validity can be improved when the patients understand the questions and are able to recall, their answers are anonymous and there is no fear of reprisals [
22]. Using focus group discussions may reduce the fear of reprisals and increase anonymity, which improve validity of self-report. In contrast to household surveys, the present study was undertaken in a hospital setting, which may have led to selective recall bias for fear of possible repercussions. A previous review indicated that the validity of self-reported data may be affected by cognitive issues including clarity of the questions, memory needed to answer the questions and influence of the survey settings [
22].
IPTp use and other characteristics
Participants who reported IPT use during pregnancy tended to be younger in age, more educated, and reported having received iron supplementation during pregnancy (Table
3). In other words, the more educated participants were more likely to have reported IPTp intake than the less educated. There was a high ITN coverage (89.2%) in the participants, an encouraging finding which is important for prevention of malaria in pregnancy. This high bed net use could be due to the relatively high socio-economic status of the participants as the study population was largely urban and had access to contemporary distribution of ITNs to pregnant mothers free of charge. A recent study in Tanzania found that timely uptake of IPTp depends more on practices of health workers at the health units than individual characteristics of pregnant women and that early ANC attendance did not influence IPTp use [
23]. Although 98.5% of the participants reported having attended the antenatal clinic at least once during that pregnancy, only 58.8% reported IPTp use during the current pregnancy.
This lower than expected use of IPTp could reflect suboptimal care at the antenatal clinics, lack of drugs in the health units and inadequate sensitization of the health workers which have been found to affect uptake [
24]. Significantly, in areas with high transmission, pregnancy malaria still causes considerable morbidity and mortality in spite of high bed net use [
25]. This emphasizes the importance of improving IPTp coverage to reduce the incidence and effects of pregnancy malaria efficiently.
Of the participants who reported non-use of IPT during pregnancy, 25% (n = 84) actually had the drug in circulation at the time of delivery. This finding suggests false reporting by the participants, which could be due to recall bias or the possibility that participants were not informed about the drugs given during the pregnancy, since the participants who reported using any sulpha drug during their pregnancy were excluded from the study.