The malaria prevalence of 23% using microscopy in our study was within the range of 9.7% in Ketu South to 31% in Adaklu districts recorded in the study region in 2020. But this rate was significantly higher by 8.1% point above the rate of 15% recorded in the Ho Municipality in the study township in 2020(31). The variation could be due to an anecdotal increase in clinical cases of malaria among cohorts of military personnel returning from international peacekeeping in malaria endemic regions. Inversely, the relatively low rate of 15.1% recorded in the DHIMS database for Ho Municipality could be due to limitations associated with mRDTs data in the Municipality(32), or just a natural reflection of reductive trends in incidence of malaria observed in Ghana(7).
In addition, the malaria prevalence of 23% using microscopy was relatively higher than the mean rate of 12.3% using candidate mRDTs in this study. This phenomenon of recording higher rates of malaria infection by microscopy relative to lower rates using CareStart® mRDTs was reported in a number of comparable studies, including Ghana(33) and other African counties(34,35). The mean rate of 49.5% sensitivity recorded for candidate mRDTs in this study was significantly below the manufacturers’ mean rates of 96.2% and WHO’s threshold of ≥ 95%(36). However, the mean rate of 99% specificity recorded was comparable to the manufacturers’ mean specificity of 98.8% and the WHO’s threshold of ≥ 90%( 25).
This phenomenon of lower sensitivity relative to higher specificity rates for mRDTs was also reported in previous multicentre studies in same region (32,37) and other regions in Ghana (38,39), as well as studies in other African countries(35,40,41). Similar to pfhrp2-based CareStart, relatively lower mean rates of sensitivity (72.7%) and specificity (62.7%) rates for using pfhrp2-based Paracheck were also previously reported in Ghana(33,42),and West-African countries including Monrovia(43), Mali(44) and Nigeria(45), as well as other African countries, including Zimbabwe(46), Uganda(47), Malawi(48), Ethiopia(49), and Tanzania(50). The most cited reasons for this phenomenon included, but were not limited to, variations in geo-demographics of the clinical population (33–35), and impact of experience microscopy (38).
Moreover, the disconcordance between candidate mRDTs and microscopy is mathematically associated with a poor sensitivity of tested mRDTs on the one hand, and low sensitivity rates associated with false negativity of tested mRDTs on another hand (51). The most cited reasons for false negativity associated with mRDTs are technical, biological, social, infrastructural, regulatory, and economic barriers (51). As Miller and Sikes (2015) outlined, the “technical barriers” include “susceptibility to heat and humidity”, “inadequate characterization of the affinity agent and antigen”, “prozone effect”, “subjective interpretation of test lines”, “time-dependence of signal development”, “insensitivity to low disease loads” and “unsuitability for pathogen quantification” (51).
Regarding the factors of “heat and humidity”, the storage conditions were traceable in the study laboratory and the metrics were within the normal ranges. However, records of quality control on maintenance of optimum storage condition during transportation from the national through regional storage facilities to the study facility were not traceable for verification. Although there was no gross observation of inaccurate low intensity test lines due to either overabundant antigens or overabundant antibodies, the association between our false negative samples and prozone phenomenon was experimentally not evaluated. However, due to experience of the Medical laboratory Scientists handling the mRDTs, as well as documentary evidence on the internal quality control mechanism of the sample collection, processing and reading of the mRDT, we could rule out the association between our false negative samples and the factor of the ‘subjective interpretation of test-lines’, on one hand and “time-dependence of signal development’, on another hand.
In addition, biologic factors include lack of biomarkers for a specific pathogen, limited ability of the biomarker to track heterogenic genes emitted by drug resistance or mutation and cross-reactivity in conditions such rheumatoid and Schistosomiasis(51). Of the 9 characteristics of pathogen identification biomarkers, the absence of diagnostic biomarker for a pathogen and the limited ability of the biomarker to track all stages of the pathogen’s life cycle, were strongly implicated in our study. Although we were unable to experimentally rule out the association between negative mRDT samples and genetic heterogeneity of HRP2 vs HRP3 genes as implicated in a number of studies across the globe (52), the travel and exposure history associated with the study population, strongly suggest the circulation of non-falciparum species of plasmodium among the clinical cases. Although the study site is a Military Hospital open to public, the attendees were mostly military personnel who have had extensive history of travelling on international peacekeeping missions in malaria endemic regions and their families. Therefore, they may have been exposed to non-falciparum species of Plasmodium (53), which escaped detection using the candidate mRDTs, which are pfhrp2-based and specific for only P.falciparum. Indeed, using microscopy, non-falciparum species of Plasmodium (P) including P. malariae, P. ovale, and P. vivax were detected in 84% of samples that were negative for malaria using candidate mRDTs.
Furthermore, although we could not experimentally verify the association between our negative samples and the factor of low parasite density in this study, the remaining 16% of samples, which were negative using the candidate mRDTs, could be due to a lower parasite antigen titre in the blood at the time of sampling. Low parasite density was known to be associated with early infection and/or self-medication, which is fast becoming a common practice in Africa (54, 55). However, even though the candidate mRDTs showed a higher specificity to detect true negative cases, poor sensitivity metrics render them unreliable in the detection of malaria cases. The utilization of ecologically incentive mRDTs is detrimental to managing malaria among at-risk populations including children under five years of age, pregnant women and our military personnel exposed to non-falciparum species of Plasmodium. Consequently, this will compromise global health strategies to eradicate epidemic malaria by 2030.
4.1 Limitations:
We were unable to overcome the inherent limitations associated with the pilot study. These included lack of funding to extend the study period to cover both raining and non-raining transmission seasons of malaria, and the use of limited sampling (Randomization vs consecutive) and analysis methods (Microscopy vs PCR). So, predictions, assumptions and interpretation of our result should, therefore, be done within the context of our study design and settings.
4.2 Conclusions:
Both mRDTs showed lower discriminatory accuracy compared to microscopy in this study. The observed variation between the metrics of ‘external validity’ and ‘ecologic validity’ has Research and policy implications.
4.3 Recommendation:
Due to the proliferation of non-falciparum species of plasmodium in different ecologies, Ghana should as matter of urgency accelerate and strengthen the implementation of the WHO’s recommendation to switch from monobiomaker based mRDTs to multiple- biomarker based mRDTs for confirmatory malariology. Additionally, it is essential to maintain a regular workshop for medical store managers at the regional and district levels to review the quality control mechanism on transport, storage and handling of RDT kits. Furthermore, it is crucial to conduct further research to consolidate our understanding of ecological risk factors, which could affect the validity and reliability of mRDT in primary health care centres in Ghana. The National Malaria Elimination Programme (NMEP) should refine strategy to contain our ‘military population’ as an emerging reservoir for the non-falciparum plasmodium parasite in Ghana. Additionally, the NMEP and health training institutions must intensify the diagnostic capacities of health workers, using diversified mRDTs.