Remember me
ORIGINAL ARTICLE
Year : 2023 | Volume
: 22
| Issue : 4 | Page : 470-480
SD-Bioline malaria rapid diagnostic test performance and time to become negative after treatment of malaria infection in Southwest Nigerian Children
Adebola Emmanuel Orimadegun1, Hannah O Dada-Adegbola2, Obaro S Michael3, Akinlolu Adedayo Adepoju4, Roland Ibenipere Funwei5, Fiyinfoluwa Ibukun Olusola3, IkeOluwapo O Ajayi6, Oluwatoyin O Ogunkunle4, Olusegun George Ademowo7, Ayodele Samuel Jegede8, Ebenezer Baba9, Prudence Hamade10, Jayne Webster11, Daniel Chandramohan11, Catherine Olufunke Falade7
1 Institute of Child Health, College of Medicine, University of Ibadan, Ibadan, Nigeria
2 Departments of Medical Microbiology, College of Medicine, University of Ibadan, Ibadan, Nigeria
3 Department of Pharmacology and Therapeutics, College of Medicine, University of Ibadan, Ibadan, Nigeria
4 Department of Paediatrics College of Medicine, University of Ibadan, Ibadan, Nigeria
5 Department of Pharmacology, Babcock University, Ogun State, Nigeria
6 epartment of Epidemiology and Biostatistics, College of Medicine, University of Ibadan, Ibadan, Nigeria
7 Department of Pharmacology and Therapeutics, College of Medicine, University of Ibadan; Institute for Advanced Medical Research and Training, College of Medicine, Ibadan, Nigeria
8 Department of Sociology, Faculty of the Social Sciences, University of Ibadan, Ibadan, Nigeria
9 Malaria Consortium Regional Office for Africa, Kampala, Uganda
10 Malaria Consortium, London, United Kingdom
11 London School of Tropical Medicine and Hygiene, London, United Kingdom
Correspondence Address:
Catherine Olufunke Falade
Department of Pharmacology and Therapeutics, College of Medicine, University of Ibadan, Ibadan
Nigeria
Source of Support: None, Conflict of Interest: None
CheckDOI: 10.4103/aam.aam_220_21
Abstract
Context and Aim: Given the challenges of microscopy, we compared its performance with SD-Bioline malaria rapid diagnostic test (MRDT) and polymerase chain reaction (PCR) and evaluated the time it took for positive results to become negative after treatment of children with acute uncomplicated malaria. Subjects and Methods: We present the report of 485 participants with complete MRDT, microscopy, and PCR data out of 511 febrile children aged 3–59 months who participated in a cohort study over a 12-month period in rural and urban areas of Ibadan, Nigeria. MRDT-positive children received antimalaria and tested at every visit over 28 days. Speciation was also carried out by PCR. Results: With microscopy as the gold standard, SD-Bioline™ had 95.2% sensitivity, 66.4% specificity, 67.5% positive predictive value (PPV), and 94.9 negative predictive value (NPV), while with PCR the findings were 84.3% sensitivity, 66.5% specificity, 72.7% PPV, and 80.1% NPV. PCR speciation of malaria parasites revealed 91.6% Plasmodium falciparum, 18.9% Plasmodium malariae, and 4.4% Plasmodium ovale. Among the 47 children with P. malariae infections, 66.0% were coinfected with P. falciparum, while 54.6% cases of P. ovale occurred as coinfections with P. falciparum. The median time to a negative MRDT was 23.2 days, while the median time to a negative malaria microscopy was 3.8 days. The two survival curves were significantly different. Conclusions: The SD-BiolineTM MRDT performed well, with remarkable persistence of rapid test-positive for an average of 23 days post treatment. The prevalence of P. malaria is somewhat greater than expected.
Abstract in French Résumé
Contexte et objectif: Compte tenu des défis de la microscopie, nous avons comparé le test de diagnostic rapide du paludisme SD-Bioline (MRDT) avec la réaction en chaîne par polymérase (PCR) et évalué le temps qu'il a fallu pour que des résultats positifs deviennent négatifs après le traitement d'enfants atteints de paludisme aigu non compliqué. Sujets et méthodes: Nous présentons le rapport de 485 participants avec des données complètes de MRDT, de microscopie et de PCR sur 511 enfants fébriles âgés de 3 à 59 mois qui ont participé à une étude de cohorte sur une période de 12 mois dans les zones rurales et urbaines d'Ibadan, Nigeria. Les enfants positifs au MRDT ont reçu un antipaludique et ont été testés à chaque visite pendant 28 jours. La spéciation a également été réalisée par PCR. Résultats: Avec la microscopie comme référence, SD-Bioline TM avait une sensibilité de 95,2 %, une spécificité de 66,4 %, une valeur prédictive positive (VPP) de 67,5 % et une valeur prédictive négative (VPN) de 94,9 %, tandis qu'avec la PCR, les résultats étaient de 84,3 % de sensibilité, 66,5 % de spécificité, 72,7 % de VPP et 80,1 % de VPN. La spéciation par PCR des parasites du paludisme a révélé 91,6 % de Plasmodium falciparum, 18,9 % de Plasmodium malariae et 4,4 % de Plasmodium ovale. Parmi les 47 enfants atteints d'infections à P. malariae, 66,0 % étaient co-infectés par P. falciparum, tandis que 54,6 % des cas de P. ovale se sont produits sous forme de co-infections par P. falciparum. Le délai médian jusqu'à un MRDT négatif était de 23,2 jours, tandis que le délai médian jusqu'à une microscopie négative du paludisme était de 3,8 jours. Les deux courbes de survie étaient significativement différentes. Conclusions: Le SD-BiolineTM MRDT a donné de bons résultats, avec une infection à P. malariae un peu plus élevée que attendu dans la population et persistance remarquable des résultats positifs aux tests de diagnostic rapide pendant une moyenne de plus de 23.
Mots-clés: Paludisme, microscopie, Nigéria, réaction en chaîne par polymérase, test de diagnostic rapide, spéciationjours après le traitement
Keywords: Malaria, microscopy, Nigeria, polymerase chain reaction, rapid diagnostic test, speciation
Introduction 
Malaria remains a major cause of morbidity and mortality in Sub-Saharan Africa, where 90% of the global malaria burden occurs. Over 90% of deaths occurred in children under five years of age. Nigeria bears a disproportionately heavy burden of cases and deaths.[1] Effective management of the disease depends on accurate diagnosis followed by prompt and effective treatment. Presumptive diagnosis of malaria based on clinical features alone, even in experienced hands, leads to overdiagnosis of malaria.[2],[3],[4] The problems associated with misdiagnosis include delay or failure to diagnose potentially serious clinical conditions such as pneumonia and unnecessary exposure to antimalarial drugs with the possible consequence of adverse drug effects or selection of drug-resistant strains of Plasmodium.[2],[3],[4] The World Health Organization in 2010 recommended parasite-based confirmation of clinically suspected malaria by microscopy or malaria rapid diagnostic test (MRDT) before administration of artemisinin-based therapy.
Even though microscopy remains the gold standard for diagnosing malaria, there are difficulties that limit its use. Among the obstacles are inconsistent electricity to power the few accessible microscopes in malaria-endemic nations, a scarcity of competent microscopists, and the parasite detection limit by microscopy. Many studies in malaria-endemic countries have shown that quality-assured MRDTs are effective alternatives to microscopy in the diagnosis of malaria.[5],[6],[7] The use of MRDTs in children with febrile diseases has also been shown to reduce antimalarial medication overprescription.[8] The introduction of MRDTs into the community in Zanzibar resulted in a decrease in malaria prevalence and improved management of febrile infections, many of which tested negative for malaria.[9] MRDTs were also used to recommend the proper use of antibiotics in patients who tested negative for malaria in that research. In a study conducted in Uganda, histidine rich protein 2 (HRP2)-based MRDTs were found to have higher sensitivity than expert microscopy at a rural center.[10] Polymerase chain reaction (PCR) diagnostic method is reported to be the most sensitive and specific for detecting malaria parasite at densities as low as five parasites/μL of blood.[11] However, PCR requires expensive equipment, reagents, trained personnel, and stable electricity supply. These are facilities that are neither affordable nor available in most malaria endemic areas.
Nigeria revised its malaria treatment guidelines in 2011 to comply with the WHO requirement that malaria be diagnosed parasite-based wherever possible.[12] Because Plasmodium falciparum causes more than 95% of malaria infections in Nigeria, HRP-2-based MRDTs were determined to be the best fit for the country.[12] In addition, HRP-2-based (protein) MRDTs are less expensive and more resistant to high tropical temperatures and humidity than enzyme-based pLDH and aldolase-based MRDTs.[13] MRDTs are simple to use and quick to execute, but they have certain limitations. Antigen–antibody responses drive MRDTs. False-positive results are caused by antibody persistence and cross reactivity, whereas deletion or mutation of the HRP-2 gene causes false-negative results.[14] The HRP-2 protein is unique to P. falciparum; hence, MRDTs may miss non-falciparum infections. These tests remain positive for months after a malaria infection has been cured.[15] These tests fail to distinguish between recently cured and new/current infections.
The 2011 Nigeria National Malaria treatment guidelines also stipulated that ACT treatment be restricted to MRDT-positive cases only as recommended by the WHO. This underscores the need to periodically evaluate the performance of commonly used HRP-2-based tests that also score high on the WHO/FIND list.[13],[16] During this study, we tested the SD-Bioline, an HRP-2-based MRDT, against microscopy and PCR in febrile children with clinical features consistent with malaria. We also evaluated the time it took for positive MRDT results to convert to negative results after treatment of children for acute uncomplicated malaria by testing with MRDT at every visit. In addition, the speciation of infecting Plasmodium was carried out by PCR.
Subjects and Methods Study setting and design
The study was conducted in two centers in southwestern Nigeria; an urban center located at a secondary healthcare facility and a rural center located at a primary health care facility. The rural site of the study was located at a primary healthcare center in Idi-Ayunre in Oluyole Local Government Area about ten kilometers away from the city of Ibadan while the urban site was located in St Mary's Catholic Hospital, Eleta, Ibadan. St. Mary's Hospital is a secondary level hospital located within the urban slum of Ibadan. Ibadan is a heavily populated city in Southwest Nigeria where malaria transmission is intense. The study reported here is part of a larger study which evaluated the consequences of restricting ACT treatment to only MRDT-positive under-five children.[17] Enrollment into the study was carried out between middle November 2013 and middle of November 2014.
Sample size calculation
Assuming the expected sensitivity and specificity was 95% and 87%, respectively, adapted from the report by de Oliveira et al.,[6] expected prevalence of MRDT positivity of 60%, desired precision of 0.05 at 95% level of confidence, the estimated optimum sample size for the study was 436. Using the sample size formula for single proportion, this sample size (436) would achieve a precision of 0.025 and 0.05 for sensitivity and specificity, respectively.
Study participants
Children aged 3 to 59 months living within 15 kilometers of the two study centers were enrolled into the study if they were febrile (axillary temperature >37.4°C) or have a history of febrile illness within 48 h of presentation, absence of any danger sign, and provision of a signed informed consent by the parent or guardian. Children that had features of severe malaria or other severe illness were not enrolled but were attended to clinically and referred to a tertiary health center for further care as considered necessary.
Clinical and Laboratory procedures
Enrollees received detailed clinical evaluation: biodata, history of the presenting illness, and clinical examination which were documented in case record forms specifically designed for the study. The children had their weights and heights measured. Capillary blood samples were taken from a finger prick for MRDT, preparation of thick blood film for microscopy, and blood spot on filter paper for PCR.
Malaria rapid diagnosis procedure
MRDT was done using SD-Bioline™ (an HRP-2-Based rapid diagnostic test [RDT] kit) manufactured by Standard Diagnostics, Inc., South Korea Batch No-Lot 082365. The date of manufacture was July 5, 2013, and the expiratory date was July 4, 2015. The test kits were stored in an air-conditioned room between 4°C and 30°C according to the manufacturer's instructions and taken in batches of 25 to the field study sites as needed. Trained clinical research staff conducted the MRDT procedure. At the point of enrollment, the test device was carefully removed from its pouch and the enrollee's details: study identification number, initials, date, and study site were written at the back of the cassette with a permanent marker following which the test device was placed on a flat horizontal surface away from direct sunlight and insects. The pulp of the left middle finger of the enrollee was cleaned with cotton wool soaked in 70% alcohol and allowed to dry. The cleaned pulp of the left middle finger was pricked with the lancet supplied with the MRDT cassette and squeezed gently to obtain drops of blood. Five to 10 μl of blood was collect from the prick site using the loop provided by the manufacturer and dropped in the designated sample well labeled A following which 4 drops of buffer was dropped into the round hole marked B holding the plastic dropper vertically. The result was read at the end of 15 min according to the manufacturer's instructions.
Interpretation of MRDT test results were done at the clinical study sites independent of microscopy results. Each test was read by two trained research staff and the supervising research medical officer re-read the results if there was a discrepancy. The test was considered positive when both the test and control bands were visible in their respective windows, negative when only the control band was visible and invalid when the control band was absent irrespective of whether the test band was visible or not.
Malaria microscopy
Blood sample for thick smears and blood spots on filter paper were also collected from the same finger-prick as that used for rapid diagnostic testing. Duplicate thick blood smears were prepared on separate slides bearing the date, participant identification numbers and initials. Smears were stained with fresh 10% Giemsa stain at pH 7.2 and examined under the high-power objective lens of a light microscope (×1000 magnification) for identification and quantification of asexual stages of malaria parasites. Microscopy was done independent of the rapid diagnostic test results. A blood smear was considered negative only after the inspection of at least 200 high power fields of the thick smear. Parasite density was calculated using an assumed white blood cell count of 8000/cm3. All slides were read by two expert microscopists at the University College Hospital laboratory of the project.
Treatment and follow-up of participants
Children who tested positive to MRDT were treated with oral artesunate–amodiaquine (ASAQ™; Sanofi Aventis) supervised according to the national treatment guidelines.[18] Enrollees with septic foci were treated as appropriate with antimicrobial agents while others received antipyretic analgesic and/or vitamin supplements. All enrollees were followed up daily until day 3 and subsequently on days 7, 14, 21, and 28. Parents and guardians of enrolled children were encouraged to report in the clinic whenever the children were unwell, or they had concerns about the health of their children/wards. Enrollees received thorough history taking, physical examination and finger prick for thick blood smear at every follow-up visit. Repeat MRDT testing was also done at Days 3, 7, 14, 21, and 28.
Polymerase chain reaction assay
Finger-prick blood was blotted on Whatman 3MM filter paper, dried in dust free area, wrapped in plastic zip-lock sample bags with silica gel to prevent DNA contamination until analysis.
DNA isolation
DNA was extracted from filter paper using QIAamp™ DNA Mini kit blood and tissue (QIAGEN Germany) according to the manufacturers' instructions. Briefly, three punched out circles from the dried blood spots were placed into a 1.5 ml microcentrifuge tube and 180 μl ATL (tissue lysing buffer) was added and incubated at 85°C for 10 min. Twenty microliters of proteinase K was added to the mix and incubated for 1hr at 56°C. 200 μl of ATL buffers (tissue lysing buffer) was added to the sample and was incubated for another 10 min at 70°C. Two hundred microliter of 96%–100% ethanol was added to the sample, vortexed thoroughly and then transferred to the QIAamp Mini spin column. The sample was centrifuged at 8000rpm for 1 min and the flow through (filtrate) discarded following which 500 μl of wash buffers (AW1 and AW2) were added to the mini spin column and then centrifuged for 1 min at 8000rpm for AW1 and 14000 rpm for AW2 for 3 min. Finally, 150 μl of Buffer AE (Elution buffer) was added to the spin column, incubated for 1 min at 25°C and centrifuged at 8000rpm for 1 min to obtain the final yield of the DNA. The extracted DNA was stored at –20°C until use.
Specie differentiation
Identification of malaria-infected samples was carried out by nested PCR. The target gene was the small subunit rRNA. A primary reaction of the 18srRNA Plasmodium genus was amplified using specific primers for forward (rPLU5) and reverse (rPLU6) as previously described.[19],[20] The first amplification was done with an initial denaturation at 95°C for 5 min, this was followed by 25-cycles at 94°C for 1 min, annealing at 58°C for 2 min, and extension at 72°C for 2 min and final extension at 72°C for 5 min. A mixture of 10x PCR buffer 2 μl, MgCl2 0.8 μl, species-specific primers (forward and reverse) 0.5 μl, dNTP 0.8 μl, Taq 0.2 μl (New England Biolabs Inc. USA), 5 μl of DNA template and a nuclease free PCR water in a final volume of 20 μl for both reactions. Two microliters of the primary PCR product was used as the template with species-specific primer pairs designed to amplify the 18srRNA gene of P. falciparum, P. vivax, Plasmodium Malariae, and Plasmodium ovale performed in the Nested PCR for 30 cycles following a standard protocol.[19],[20] PCR products were mixed in a loading mixture containing 1 μl SYBR green, 2.5 μl loading buffer and 1.5 μl nuclease free PCR water (5 μl volume). Gel electrophoresis (100 volts) was done on 1.5% agarose gel alongside with a 100 bp DNA ladder (New England Biolabs Inc., USA) for DNA size standards to separate PCR products to allow sizing of species bands. Upon completion of the gel electrophoresis, gels were placed in a gel imaging cabinet and digitally photographed under ultraviolet light. Gel images were recorded on PCR viewer, printed and corresponding sample lanes were scored visually for the presence of P. falciparum, P. vivax, P. malaria, and P. ovale. In each of the reactions, positive and negative controls were included as part of quality control measures.
Data analysis
Data were entered into a study database and analyzed using SPSS Version 20 software (IBM-SPSS Inc., IL, USA). Parasite densities obtained by microscopy were presented as geometric means. Univariate analysis used Chi-square and Fisher's exact test for binary data. The sensitivity, specificity, and positive and negative predictive values (NPVs) of the SD-Bioline MRDT were compared with microscopy and conventional PCR as reference standards. MRDT results were compared with Microscopy or PCR as gold standards, while microscopy was compared with PCR as gold standard during data analysis. MRDT results were considered true positives (TP) or true negatives (TN) if they were concordant with microscopy or PCR results and false negative or false positive if otherwise. The results of microscopy were also compared with PCR results in the same manner. The agreement between diagnostic methods was assessed by calculating the kappa statistics (k) and corresponding standard error of mean;[21]k-value of 0.4–0.6 was considered moderate agreement, while 0.61–0.8 was considered substantial agreement and 0.81–0.99 almost perfect agreement. Statistical significance was set at P ≤ 0.05.
The time for a previously positive MRDT to become negative was also noted as was the time of any conversion from negative MRDT result to positive. Patients who presented with recurrent malaria parasitemia on microscopy before MRDT had become negative were treated with 6-dose artemether–lumefantrine and excluded from the analysis. MRDT results of those children who were lost to follow-up were considered as at their last visit. The probability of a test becoming negative over time was calculated with Kaplan–Meier survival function. We compared the survival curves for malaria microscopy and MRDT using the Log-rank (Mantel-Cox) test.
Ethical considerations
Ethical approvals for the study protocol were obtained from University of Ibadan/University College Hospital Ethical Review Committee, the Oyo State Ministry of Health Ethics Committee and The London School of Tropical Medicine and Hygiene IRB. Participation in the study was voluntary and based on written or witnessed verbal informed consent of parent or caregiver of children who fulfilled the inclusion criteria.
Results Demographic profiles of participants:
One thousand one hundred and two children were screened between November 18, 2013, and November 21, 2014. Five hundred and eleven children who satisfied the inclusion criteria were enrolled into the main study. The results of 485 enrollees with complete MRDT, microscopy, and PCR data points are presented in this paper. Of these 485 children, 340 (70.1%) were enrolled at the rural site while 145 (29.9%) were enrolled at the urban site. The mean age of the enrollees was 26.6 months ± 15.7, while the mean temperature was 37.7°C ± 1.19 (range 35.6°C–40.7°C). Further demographic and clinical details of enrollees are shown on [Table 1]. Parents/care givers of 91 (18.8%) enrollees admitted having given antimalarial drugs to their children/wards during the index febrile illness before presentation at the clinic. Antimalarial drugs administered include mefloquine (1), artemisinin monotherapy (2), amodiaquine (3), chloroquine (31), artesunate–amodiaquine (1), and artemether–lumefantrine (49). A significantly higher proportion of children in the urban study site than rural study site had received antimalaria drugs ([24.8%; 36/145] versus [16.2%; 55/340] ρ = 0.031) before presentation. However, there was no significant difference in the proportion that received ACT ([11.0%; 16/145] versus [10%; 34/340] ρ = 0.872).
Table 1: Clinical and demographic characteristics of enrollees at enrollmentMalaria parasite detection by microscopy
Malaria parasite was detected by microscopy in 42.3% (205/485) of the enrollees. This consisted of 50.3% (171/340) of the enrollees from the rural study site and 23.4% (34/145) of those from the urban site. The geometric mean parasite density was also higher among rural children compared to those enrolled at the urban study site [Table 1]. It is noteworthy that 13.2% (27/205) of parasitemic enrollees had parasite density <500/μL, while 13.7% (28/205) had parasite density over 100,000/μL with the majority in between. There was a nonsignificant difference in the geometric mean parasite density during the rainy season months and dry season months with 9,396/μL and 6,842/μL, respectively (ρ = 0.318). The prevalence of malaria parasite by microscopy was significantly higher (51.4%; 75/146) among children aged 36 to 59 months compared with those aged 3 to 35 months (38.3%; 130/339) (odds ratio [OR] = 0.59; 95% confidence interval [CI]: 0.398–0.871).
Malaria parasite detection by SD-Bioline malaria rapid diagnostic test
Overall, 289/485 (59.6%) enrollees tested positive to MRDT. About two-thirds (67.6% [230/340]) of the enrollees from the rural study site tested positive to MRDT, while 40.7% (59/145) had similar result in the urban study site (ρ < 0.0001). There were 10 (2.1% of 385) cases of false-negative results and 94 false-positive MRDT readings when microscopy was used as the gold standard giving a sensitivity of 95.1% and a NPV of 94.9%. The parasite density of enrollees who had false-negative result ranged between 76/μL and 5,538/μL with two of the ten recording parasite densities of 2476/μL and 5538/μL. The geometric mean parasite density for false-negative cases was 480/μL, while that of true-positive cases was 9212/μL. More details on the performance of SD-Bioline™ against microscopy are shown on [Table 2]. The performance of SD-Bioline MRDT at parasite densities >200/μL is similar to its overall performance [Table 2]. The κ-statistics were 0.583 and 0.577 for the total study population and at parasite density >200/μL. SD-Bioline detected PfHRP-2 antigenemia in 80% (12/15) of enrollees with parasite density ≤200/μL of asexual parasites and in 81.5% (22/27) of those with parasite density <500/μL.
Table 2: SD-Bioline malaria rapid diagnostic test performance compared with microscopy and polymerase chain reactionDiagnostic performance using PCR as gold standard
PCR detected malaria in 51.3% (249/485) of the enrollees. SD-Bioline recorded 39 false-negative results when compared with PCR while microscopy recorded 16 false-positive results. When the performance of SD-Bioline and microscopy was compared to PCR as gold standard, microscopy had the better performance with a specificity of 93.2%, a positive predictive value (PPV) of 92.2%, and κ-statistic of 0.688 [Table 2] and [Table 3]. The prevalence of malaria was least in children <12 months of age and increased steadily with age by all three diagnostic tests [Figure 1]. Furthermore, [Table 4] shows the levels of agreement (k-statistics) between paired comparisons of the mRDT with microscopy and PCR methods, as well as microscopy with PCR methods.
Figure 1: Distribution of plasmodium prevalence by age group and diagnostic method
Table 3: Performance of microscopy against polymerase chain reaction as gold standard among under-5 febrile children in Southwest Nigeria
Table 4: Performance of three diagnostic methods (SC-Bioline malaria rapid diagnostic test, microscopy and polymerase chain reaction) among febrile under-5 children in Southwest NigeriaThe performance of SD-Bioline™ was significantly better during the rainy season than during the dry season [Table 5]. Prior use of ACT did not seem to affect the performance of MRDT. The proportion of enrollees who had a false-positive MRDT results among those who admitted to a history of ACT use prior to presentation was 30.0% (14/36) compared with 34% (83/244) among who claimed not to have taken ACT (ρ = 0.738).
Table 5: Performance of SD-Bioline™ malaria rapid diagnostic test using microscopy as gold standard during rainy and dry seasons among febrile under 5 children in Southwest NigeriaPlasmodium species detected
P. falciparum was the most common specie detected by molecular biology technique (228/249; (91.6%), followed by P. malariae (47; 18.9%) and P. ovale (11 (4.4%). No case of P. vivax was detected. Most cases of P. malariae and P. ovale occurred as coinfection with P. falciparum [Table 1]. One enrollee had coinfection with P. falciparum, P. malariae, and P. ovale at the same time.
[Table 6] shows the percentage of plasmodium species detected by mRDT versus the total number of plasmodium species identified by PCR.
Time to conversion of malaria rapid diagnostic test positivity to negative
There were 289 MRDT-positive cases on day 0 of which only 195 (67.5%) had patent parasitemia giving a false-positive rate of 32.5%. The proportion of enrollees who tested positive to MRDT gradually decreased over time with 81.5% still MRDT positive by day 28 [Figure 2]. When the proportion of enrollees in whom parasitemia recurred or appeared for the first time are removed, these leaves 66% persistent MRDT positivity which represents the proportion of false-positive cases by day 28. The MRDT result did not convert to negative before parasite recurrence in any of the enrollees who tested positive at day 0 and had parasite recurrence.
Figure 2: Proportion of microscopy +ve and MRDT +ve enrollees on various study visit daysThe mean length of time that positive MRDT result remained positive was significantly longer among enrollees who had patent malaria parasitemia at day 0 than those who did not have patent malaria parasitemia (26.33 ± 4.9 days versus 23.0 ± 8.8 days: ρ ≤ 0.0001). Positive MRDT results remained consistently positive for significantly longer as parasite densities increased starting with 2,000/μL or more and 20,000/μL or more (ρ = 0.008 and ρ = 0.001, respectively). [Figure 2] compares the proportion of microscopy positive and MRDT positive enrollees on different study visit days, while [Figure 3] shows the time to negative mRDT results for both as a survival curve. The average time to obtain negative mRDT results was 23.3 days (95% CI = 21.6, 24.8), while the time to receive negative malaria microscopy results was 3.8 days (95% CI = 3.5, 4.1). The two survival curves were significantly different as shown in [Figure 3] (Chi-square = 528.3, df = 1; P < 0.0001).
Figure 3: Cumulative proportion survival curve for time to malaria test negative results using RDT and Microscopy Discussion In this study, febrile under-five-year-old children in Southwest Nigeria were diagnosed with malaria using SD-Bioline™, an HRP-2-based MRDT, microscopy, and PCR. Malaria prevalence was 59.6%, 42.4%, and 51.4% by MRDT, microscopy, and PCR, respectively. SD-Bioline has 95.2% sensitivity and 94.9% NPV compared with expert microscopy. Our observed prevalence by MRDT, microscopy, and PCR is similar to the report of Sitali et al.[22] working in endemic areas of Zambia. Furthermore, the results of this study, in respect of the performance of MRDT, are similar to earlier reports of HRP-2-based MRDT in the same environment by Falade et al. in 2013[23] and 2016,[24] as well as reports from the Central African Republic,[25] Southwest Uganda,[15] and from India.[26]
In Nigeria, the use of MRDT has become popular and the malaria national treatment guideline stipulates that antimalarial treatment be given to MRDT-positive patients only.[18] The performance of the MRDT with only 2.1% false-negative results is particularly remarkable in detecting febrile children who do not have malaria. This will give the healthcare provider an early indication for a prompt review and investigation for other causes of fever in patients. In addition, many patients will be prevented from receiving unnecessary ACT. Thiam et al.[27] reported a major reduction in antimalarial drug consumption in Senegal, another West African country, following the introduction of the MRDT. The factors responsible for false-negative MRDT results in our study were low parasite count (considering that the geometric mean parasite density of MRDT false-negative cases was significantly lower than that of true-positive cases), possible deletion of HRP-2/HRP-3 genes as reported by Funwei et al.[28] working in Southwest Nigeria, and the presence of non-falciparum malaria in the area.
Infection by non-falciparum species is an important cause of false negatives in HRP-2-based MRDT. This study recorded a significant presence of P. malariae and P. ovale infections, which are non-falciparum malaria infections. It is noteworthy, however, that most of the non-falciparum infections occur as coinfections with P. falciparum. The occurrence of mixed infection of falciparum species is not peculiar to Southwest Nigeria. Sitali et al.[22] working in Zambia also reported a high prevalence of mixed infections by the various species of Plasmodium. May et al.[29] also reported a high prevalence of Plasmodia species in the same study location. The occurrence of non-falciparum infections as coinfections with P. falciparum is advantageous for MRDT detection of malaria infection in such enrollees. The increased proportion of non-falciparum infections detected in this study is in keeping with the report of Yman et al.[30] working in eastern Tanzania. Yman et al.[30] reported a progressive decrease in the prevalence of P. falciparum, as well as a two-to six-fold increase in the occurrence of P. malaria and P. ovale, while the prevalence of P. falciparum decreased following coordinated efforts to control falciparum malaria. In 1999, May et al.[29] have found 99.5% of Plasmodium falciparum infections as mono-infections or in combination with P. malariae or P. ovale in the same study area. Although there are reports of an increasing prevalence of P. vivax in malaria endemic areas,[31] we did not detect any P. vivax among our study samples.
The recorded false positivity rate of 20.4% (94/485) in the present study is high. This is not surprising because of the high malaria transmission in Southwest Nigeria. Hopkins et al.[32] reported high false positivity rates with HRP-2-based MRDT in areas of high malaria transmission in a study conducted in east Africa. Abeku et al.[33] also made similar observations in the Kenyan and Ugandan highlands. Funwei et al.[34] reported extensive diversity in the allelic frequency of Plasmodium falciparum merozoite surface proteins and glutamate-rich proteins in rural and urban settings of southwestern Nigeria at the study site where this study was conducted. This is an indication of continuing high malaria transmission in the location. The continued positive reaction to HRP-2-based MRDT is a result of the slow clearance of the target antigen (HRP-2) from the blood circulation. During this study, all enrollees who tested positive for the MRDT received a full course of artesunate-amodiaquine supervised in line with the National Guidelines for Diagnosis and Treatment of Malaria.[18] This translates to 32.5% of those treated with ASAQ did not have malaria at the time of enrollment, or at least did not have malaria parasitemia at a parasite density that could be symptomatic even after leaving room for the detection threshold of microscopy. This is, however, still a major improvement compared to the period when presumptive treatment for malaria was in practice and the entire 485 enrollees would have received an antimalarial drug. As it stands, 187 children were saved from unnecessary ACT treatment. The unused doses of ACT will be available for other children who truly need the ACTs. On the other hand, it could be argued that false-positive results can delay detection of other potentially serious or treatable causes of fever among such children. However, that was not the case in the clinical study that has already been published.[17]
The high rate of false positivity recorded during this study is consistent with previous reports from regions of high malaria transmission. Persistence of P. falciparum HRP-2 long after clearance of a malaria infection is an important reason for the high false positivity rate of HRP-2-based MRDT in high transmission areas such as in Southwest Nigeria.[35],[36] Another reason that could account for false-positive MRDT results among our study participants is subpatent malaria infection,[37] which is quite common in malaria endemic areas. This can explain the case of 30 study participants in this study in whom microscopy failed to detect malaria parasites, but the presence of HRP-2 and malaria DNA were detected by MRDT and PCR, respectively.
The performance of SD-Bioline in this study is at variance with another study in Nigeria conducted by Ajumobi et al.[37] which recorded a sensitivity, specificity, PPV and NPV of 100%, 98.5%, 88.6% and 100%. The better performance of SD-Bioline in the study by Ajumobi et al.[38] can be explained by the difference in the level of malaria transmission in the southwest and northcentral regions, where malaria transmission is lower. Ajumobi et al.,[38] reported a malaria prevalence rate of 10.5%, in contrast to a prevalence rate of 43.5% by microscopy in our study. The prevalence of malaria parasitemia in the two different geopolitical zones of Nigeria represents different levels of transmission, with that of the southwest of Nigeria being higher than in the northcentral region.
Three species of Plasmodia namely falciparum, malariae and ovale were identified among the study population enrolled in this evaluation. This is in consonance with the report of Grandesso et al.[15] reporting from Southwest Uganda and May et al.[28] working in south west Nigeria. Non-P. falciparum malaria infections were more common in the rural study site than in the urban site similar to the report by May et al.[29] who had earlier reported the presence of P. malaria and P. ovale in Southwest Nigeria with a higher prevalence in rural than urban areas (26.4% versus 14.8%). We recorded prevalence rates of about 18.8% and 4.4% for P. malariae and P. ovale respectively with no record of P. ovale in the urban center against 5.3% at the rural site. The occurrence of mono specie infections and coinfections by the various species of Plasmodia were common to both our rural and urban study sites.
During this study, we also examined the time taken for positive MRDT test to convert to negative after instituting ACT therapy. The proportion of total study participants who tested positive for MRDT on Day 0 gradually declined over time, with 81.5% remaining MRDT positive on Day 28. When participants with parasite recurrence or the appearance of parasitemia for the first time were excluded, the proportion of study participants with persistent MRDT positivity fell to 66%. This persistent MRDT positivity, which indicates the proportion of false-positive cases by day 28, provides a clear explanation and reflects the situation on day 0 when a considerable number of study participants received MRDT false-positive results. The MRDT result did not convert to negative before parasite recurrence in any of the participants who tested positive on day 0 and had parasite recurrence. SD-BiolineTM and CareStartTM, both HRP-2-based MRDTs, have a median time length of 35-42 days, according to Grandesso et al.[15] Not surprisingly, our findings demonstrate that participants with patent malaria parasitemia at day 0 had a substantially longer mean length of time that positive MRDT result remained positive than those who did not have patent malaria parasitemia. Positive MRDT results remained persistently positive for substantially longer when parasite densities increased from 2,000/μL or higher to 20,000μL or higher. Grandesso et al.[15] also discovered that the higher the initial parasite density, the longer it took for an HRP-2 to go negative. Mayxay et al.[39] proposed an explanation for this by measuring the persistence of detectable levels of HRP-2 after treatment and discovered a clear link between longer persistence and higher levels of initial blood parasite density. Unlike in our study, Dalrymple et al.[40] found a mean of 15 days after treatment to convert to negative in a review of published publications. However, because these findings were compiled from 31 distinct articles published across a decade (1994–2013), drawing conclusions could be difficult.
Conclusions SD-Bioline™, HRP-2-based MRDT showed a good performance as a diagnostic tool among febrile under-5 children in Southwest Nigeria. Our data also show a slight reduction in the normally assumed >95% prevalence of P. falciparum as the causative specie for malaria in Nigeria, while the prevalence of Plasmodium malariae has increased. A remarkable persistence of RDT-positive results for an average of over a 23-day posttreatment was observed in the study population.
Authors contribution
CF and AO conceived and designed the study. CF, AO, OM, OO, RF, FIO, AAA, ASJ, HD-A, IA, AJ, and OGA conducted the study. CF, AO, and DC analyzed and interpreted the data. CF and OA drafted the manuscript. All authors read and approved the final draft of the manuscript.
Acknowledgment
We thank the funders of the project, SuNMaP Nigeria, (Malaria Consortium under the Support for National Malaria Programme funded by UKaid through DFID) and the children, parents, and guardians who participated in the study. We are grateful to our research staff, Mrs. Fatima AbdusSalam, Mrs Bolatito Akinyele, Ms. Grace Egunyomi, and our village health workers. We also wish to appreciate all physicians, nurses, and other medical staff that participated in the study. We thank the Chief Matron of St. Mary's Catholic Hospital, Eleta, for permission to conduct the urban arm of the study. We are also grateful to the Oluyole Local Government Chairman for invaluable support throughout the study.
Financial support and sponsorship
Malaria Consortium under the Support for National Malaria Programme (SuNMaP Nigeria) funded by UKaid through DFID.
Conflicts of interest
There are no conflicts of interest.
References 
Comments (0)