Quakyi et al. Malar J (2018) 17:468 https://doi.org/10.1186/s12936-018-2613-x Malaria Journal RESEARCH Open Access Diagnostic capacity, and predictive values of rapid diagnostic tests for accurate diagnosis of Plasmodium falciparum in febrile children in Asante-Akim, Ghana Isabella A. Quakyi1, George O. Adjei2, David J. Sullivan Jr.3, Amos Laar4, Judith K. Stephens1, Richmond Owusu5, Peter Winch6, Kwame S. Sakyi7, Nathaniel Coleman1, Francis D. Krampa8, Edward Essuman1, Vivian N. A. Aubyn9, Isaac A. Boateng10, Bernard B. Borteih1, Linda Vanotoo11, Juliet Tuakli12, Ebenezer Addison13, Constance Bart‑Plange9, Felix Sorvor1 and Andrew A. Adjei14* Abstract Background: This study seeks to compare the performance of HRP2 (First Response) and pLDH/HRP2 (Combo) RDTs for falciparum malaria against microscopy and PCR in acutely ill febrile children at presentation and follow‑up. Methods: This is an interventional study that recruited children < 5 years who reported to health facilities with a history of fever within the past 72 h or a documented axillary temperature of 37.5 °C. Using a longitudinal approach, recruitment and follow‑up of participants was done between January and May 2012. Based on results of HRP2‑RDT screening, the children were grouped into one of the following three categories: (1) tested positive for malaria using RDT and received anti‑malarial treatment (group 1, n = 85); (2) tested negative for malaria using RDT and were given anti‑malarial treatment by the admitting physician (group 2, n = 74); or, (3) tested negative for malaria using RDT and did not receive any anti‑malarial treatment (group 3, n = 101). Independent microscopy, PCR and Combo‑RDT tests were done for each sample on day 0 and all follow‑up days. Results: Mean age of the study participants was 22 months and females accounted for nearly 50%. At the time of diagnosis, the mean body temperature was 37.9 °C (range 35–40.1 °C). Microscopic parasite density ranged between 300 and 99,500 parasites/µL. With microscopy as gold standard, the sensitivity of HRP2 and Combo‑RDTs were 95.1 and 96.3%, respectively. The sensitivities, specificities and predictive values for RDTs were relatively higher in micros‑ copy‑defined malaria cases than in PCR positive‑defined cases. On day 0, participants who initially tested negative for HRP2 were positive by microscopy (n = 2), Combo (n = 1) and PCR (n = 17). On days 1 and 2, five of the children in this group (initially HRP2‑negative) tested positive by PCR alone. On day 28, four patients who were originally HRP2‑ negative tested positive for microscopy (n = 2), Combo (n = 2) and PCR (n = 4). Conclusion: The HRP2/pLDH RDTs showed comparable diagnostic accuracy in children presenting with an acute febrile illness to health facilities in a hard‑to‑reach rural area in Ghana. Nevertheless, discordant results recorded on day 0 and follow‑up visits using the recommended RDTs means improved malaria diagnostic capability in malaria‑ endemic regions is necessary. Keywords: Febrile, Children < 5 years, Rapid diagnostic test (RDT), Malaria, HRP2, Combo *Correspondence: aaadjei@ug.edu.gh 14 Worldwide Universities Network, University of Ghana, P.O. Box LG 13, Legon, Accra, Ghana Full list of author information is available at the end of the article © The Author(s) 2018. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creat ivecom mons. org/licen ses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creat iveco mmons .org/ publi cdoma in/zero/1.0/) applies to the data made available in this article, unless otherwise stated. Quakyi et al. Malar J (2018) 17:468 Page 2 of 9 Background similar to PfHRP2 and detected by HRP2-based RDTs Malaria remains one of the most common causes of [15]; however, PfHRP3 is not detected by all RDTs. febrile illness among people living in tropical and sub- Since their adoption in Ghana, RDTs have contributed tropical regions. Globally, an estimated 214 million cases to reduction in presumptive treatment [17] but they are were reported which resulted in about 438,000 malaria not readily available in sufficient quantity even for sec- mortalities [1]. For sub-Saharan Africa (SSA) the malaria ondary and referral facilities. burden has continuously been overwhelming, with the Recent studies in Peru reported field isolates that lack incidence of new malaria cases in SSA accounting for one or both antigens (PfHRP2 and PfHRP3) and that 89% of new malaria cases and 91% of malaria deaths in poses a significant problem for diagnosis. Similar find- 2015 [1]. While deaths due to malaria have dropped ings have also been reported in African countries, includ- from near 2 million to fewer than 500,000 each year with ing Ghana [18, 19]. This may necessitate the use of HRP2 access to timely diagnosis and effective artemisinin com- in combination with other antigens that are more con- bination therapy, malaria cases have had a less dramatic served within the parasite (e.g., pLDH or aldolase) [20], drop [2]. The goal of a timely malaria diagnosis in Africa to improve diagnostic accuracy. is to quickly distinguish life-threatening falciparum The aim of the study was to evaluate the implications malaria from other causes of illness. of a negative malaria test outcome in relation to clinical Microscopy remains the gold standard for malaria diag- diagnosis, and to demonstrate the implications of car- nosis because it is inexpensive, has high sensitivity and egiver adherence or otherwise to a negative RDT test in allows Plasmodium species identification and quantifica- a rural setting in an endemic area. To this end, the diag- tion of parasite density [3, 4], however, in rural settings nostic utility of the HRP2 (First Response), pLDH/HRP2 it is often unavailable due to the lack of facilities, exper- (Combo), microscopy, and PCR were compared in the tise and constant power supply. The advent of rapid diag- following three groups of acutely ill febrile children at nostic tests (RDTs) for malaria diagnosis is therefore an presentation (day 0) and follow-up: (i) RDT-positive chil- important development to enhance early diagnosis. The dren who received anti-malarials; (ii) RDT-negative chil- implementation of RDTs has contributed to the timely dren who received anti-malarials; and, (iii) RDT-negative diagnosis and management of malaria in some endemic children who did not receive anti-malarials. countries. There was more than a two-third reduction in anti-malarial drug dispensing for children under-five Methods years upon use of RDTs in some African countries [5, 6]. Study sites Although RDTs have simplified the diagnosis of This study was conducted in two health facilities (Kon- malaria, WHO and other agencies advocate that coun- ongo-Odumase Government Hospital and Juansa Health tries test the sensitivity and specificity of malaria RDTs Centre) in Asante Akim North District of the Ashanti before giving approval [7–9]. In addition, quality control Region of Ghana. The region is one of ten in the coun- systems that assure the quality of each batch of RDTs try and is located in the Forest Zone where there are should be implemented through systematic surveillance two distinct seasons: a wet season (April to October) and monitoring. Where RDTs are available, there is con- when malaria transmission is highest and a dry season siderable evidence that clinicians treat febrile presen- (December to March). The district occupies an area of tations with anti-malarial drugs, even when the result 1462 sq km. The main economic activities of the district of the RDT is negative for the presence of the parasite are subsistence farming, animal husbandry and petty antigen. It is reported that about half of all negative RDT trading. patients were prescribed anti-malarial drugs [10–14]. RDTs for malaria are based on the detection of one of Konongo‑Odumase Government Hospital three antigens, histidine-rich protein-2 (HRP2), lactate The hospital provides both general, specialized and refer- dehydrogenase (LDH) and aldolase, which distinguish ral services to residents in Konongo-Odumase township, the differences in the sensitivity and specificity seen in surrounding communities and other residents of Ashanti RDT test kits [8]. The majority of commercially available Region. This hospital serves a population of approxi- malaria RDTs target PfHRP2 [8, 15]. Performance testing mately 100,000 with a 50-bed capacity and staff strength of RDTs revealed PfHRP2 to be a more sensitive antigen of over 250 healthcare providers. for detecting Plasmodium falciparum infections than other antigens, such as Plasmodium lactate dehydroge- Juansa Health Centre nase (pLDH) [16]. Plasmodium falciparum also produces The Juansa Health Centre (JHC), located between Kon- histidine-rich protein 3 (PfHRP3), an antigen highly ongo-Odumase and Agogo has a 12-bed capacity, is Quakyi et al. Malar J (2018) 17:468 Page 3 of 9 headed by a physician assistant and provides services to Sampling procedure over 15,000 people. After obtaining informed consent, the participants were examined by the physician-in-charge, and a study Study population questionnaire was administered to the parent/guard- The study included all children under 5  years who ian. This was followed by RDT screening. Based on the reported to the health facilities with a history of fever results of the RDT and decision of the admitting physi- within the previous 72 h or a documented axillary tem- cian, the children were grouped into one of the follow- perature of 37.5 °C. ing three categories. Study design • Children < 5  years who presented with fever and The study, conducted between the months of Janu- tested positive for malaria using RDT and received ary and May 2012, employed longitudinal methods that anti-malarial treatment (Group 1). included interventional and quantitative approaches. The • Children < 5  years who presented with fever and sampling strategy and procedures are detailed in Fig.  1. tested negative for malaria using RDT and were given A total of 260 participants were enrolled from the two anti-malarial treatment by the admitting physician facilities. Informed consent was obtained from parents/ (Group 2). guardians of the children after detailed explanation of the • Children < 5  years who presented with fever and purpose and procedures of the study. Parents/guardians tested negative for malaria using RDT and did not were assisted to complete an interviewer-based, semi- receive anti-malarial treatment (Group 3). structured questionnaire at the appropriate literacy level. All children <5 with fever within (72hrs) and documented temperature of 37.5oC Administer T5-1 Screen with RDTs (HRP2+Combo) Day 0—Lab Instructions Temperature at axillary Selection is by HRP2 1. Take venous blood sample of 0.5ml 2. Fill 3 capillary tubes a. One tube for electrophoresis (seal and store at 4oC) b. Spin 2 tubes for PCV Day 0 Day 0 Day 0 i. Calculate PCV/Hb Group 1 (N=50/50 Konongo/Juansa) Group 2 (N=25/25 Konongo/Juansa) Group 3 (N=50/50 Konongo/ Juansa)ii. Cut tubes, seal and store plasma and pellets • +ve test for HRP2 RDT • -ve test for HRP2 RDT • -ve test for HRP2 RDT 3. Perform thick and thin smear for • received treatment • received treatment • agrees to admission and admitted for observation microscopy • From Konongo or Juansa • from Konongo or Juansa • treatment or no treatment 4. Fill 2 circles on Whatman filter • sent home • sent home • from Konongo or Juansa paper for PCR 5. Collect 20µl for HRP2 and Combo RDT Day 1 Day 1—Lab Instructions • Take axillary temperature, give paracetamol or tylenol, 1. Fill 2 circles on Whatman filter and monitor child paper for PCR • Perform lab instructions for Day 1 2. Collect 20µl for HRP2 and • If microscopy is +ve, give ACT and document Combo RDTs by finger prick 3. Perform thick and thin smear for microscopy Day 2 • Take axillary temperature and monitor child. • Perform lab instructions for Day 2 Day 2—Lab Instructions 1. Finger prick blood (20µl) for • If microscopy is +ve, give ACT and document HRP2 and Combo RDTs, and 2. Fill 2 circles on Whatman filter paper for PCR 3. Perform thick and thin smear for microscopy Day2: If fever persists 4. Perform urine analysis, total • Perform urine analysis, total blood Day 2: if fever resolves blood count if fever persists. count, and necessary tests. • Send child home • Treat child per diagnosis • Send child home when fever resolves Day 4—Lab Instructions 1. Finger prick blood (20µl) for Day 4: Sample at home Day 4: Sample at home HRP2 and Combo RDTs, and • Fill follow up questionnaire • Fill follow up questionnaire Day 4: Sample at home 2. Fill 2 circles on Whatman filter • Request for drug package to confirm • Request for drug package to confirm • Fill follow up questionnairepaper for PCR 3. Perform thick and thin smear for completion of full course completion of full course • Ask if any additional drug was taken microscopy • Ask if any additional drug was taken • Ask if any additional drug was taken • Perform lab instructions for day 4 • Perform lab instructions for day 4 • Perform lab instructions for day 4 Day 28—Lab Instructions 1. Finger prick blood (20µl) for HRP2 and Combo RDTs, and Day 28: Sample at home Day 28: Sample at home Day 28: Sample at home 2. Fill 2 circles on Whatman filter • Fill follow up questionnaire • Fill follow up questionnaire • Fill follow up questionnaire paper for PCR • Perform lab instructions for day 28 • Perform lab instructions for day 28 • Perform lab instructions for day 28 3. Perform thick and thin smear for microscopy Fig. 1 Selection of target patient and associated laboratory instructions: a flow chart Quakyi et al. Malar J (2018) 17:468 Page 4 of 9 Data, sampling and laboratory analysis (BioRad, Hercules, CA, USA). The thermocycler machine Temperature, weight and other demographic character- detects 4 probes, therefore P. falciparum, Plasmodium istics of the children were obtained. Finger and venous vivax and Plasmodium malariae were chosen along with blood specimens were collected. All sample collection the human actin gene control. The primers and probes procedures were done under aseptic conditions. In all, were as follows. 0.5 µL of venous blood and two dried blood spots (DBS) Plasmodium falciparum 18S rRNA- Forward: 5′-CCA were deposited onto Whatman 903 protein saver cards CAT CTAA GG AAG GCAG CAG Reverse: 5′-CCT CCA with about 50 µL of blood for each circle. The DBS were ATTG TT ACT CTG GGA AGG Probe-5′CCC ACC ATT stored at 20 °C. CCA ATT ACAA -Cy5. Plasmodium vivax AMA1 Forward 5′-ACGC CA AGT Working principle of First Response® and Combo RDT TCGG AT TATG G Reverse: 5′-CCGT CA TTTC TT CTT The performance of the First Response® Malaria Ag CATA CT GAG Probe-5′TTGA TCT GA GGC ACT CGC HRP2-HRP2 alone (Premier Medical Corp. Ltd., India) TCCG-TET. and SD Bioline Malaria Ag Pf/Pan- HRPII and panLDH Plasmodium malariae plasmepsin Forward: 5′-CCAA CA (Standard Diagnostic Inc. Suwon City, South Korea, Cat- ATAC ATA CAC AT TAG AACC Reverse: 5′-GTA GGAT AT alogue No: 05FK60) were evaluated following manufac- AAA GCAT ACA CA AAG TG Probe-5′ATC TAGT AA TGG turers’ instructions. Briefly, 20  µL of blood from finger CTCC-TX Red Human beta actin For 5′-GTGC TCA GG stick was used for the RDTs and colour changes observed GCTT CT TGT CC Rev 5′-CCAT GT CGTC CCA GT TGG T after 15 min. For each RDT, cassettes were first labelled Probe-5′ACC CAT GCC CAC CAT CAC GCCC-FAM. with the sample number, then 10 µL of whole blood was The human actin gene was used as an extraction con- added to the sample well and the assay buffer completely trol and PCR was performed in duplicate from the single emptied into the buffer well. The RDT reaction was con- extraction of each sample. sidered as positive when two colour bands were seen at A cycle count of 34 was used for the cut off to separate the control (C) and test (T) labels. The reaction was con- positive and negative PCR samples. The efficiencies for sidered negative when only one band was seen at the con- the amplifications were 150% for P. falciparum, 101% for trol (C) label. The reaction was considered invalid when P. malariae and 70% for P. vivax. no bands were seen at both control and test labels and or when a band was seen at the test label but not at the con- trol label. All invalid reactions were repeated to deter- Case definition mine results as either positive or negative. True positives (TP) for RDT were defined as PCR posi- tive and/or microscopy positive. False positives (FP) Microscopy were cases in which PCR and microscopy negatives were Thick and thin blood films were prepared on slides positive for RDT. True negatives (TN) were negative by and stained with 10% Giemsa and examined using oil all three methods. False negatives (FN) were those cases immersion magnification with a light microscope. Two that were negative by RDT but positive for PCR and/or independent microscopists examined slides for asexual microscopy. parasite stages. Parasite density was quantified in thick films by counting asexual P. falciparum parasites against 200 leukocytes and multiplied by 40, assuming a standard Counselling and follow‑up of patients with initial negative leukocyte count of 8000 leukocytes/µL. and positive results Follow-ups were done by registered nurses and was coor- dinated by research assistants, the biomedical scientist PCR About 100 µL of blood previously blotted on two circles at Konongo, and the Municipal Director of Health Ser- of Whatman 903 Protein Saver cards filter paper were vices. Children who tested positive by RDT and received dried and stored at room temperature (20 °C). Five 3-mm anti-malarials (Group 1) were followed up as outpatients diameter punches were processed with a commercial on day 4 and day 28 in their homes. Children who tested 96-well kit (Promega, Fitchburg, WI, USA) to extract negative and received anti-malarials (Group 2) were also the DNA from approximately 25  µL of the dried blood followed up as outpatients on day 4 and day 28 at their into an eluted volume of 200 µL of water. Multiplex PCR respective homes. Children who tested negative and did from 10 µL (1/20 of 25–1.25 µL of blood equivalents) of not receive anti-malarials (Group 3) were placed under the extracted DNA volume was performed in real time observation overnight. The plan for observing and fol- (qPCR) on a CFX 384 Detection System Thermocycler lowing up of participants is detailed in Fig. 1. Quakyi et al. Malar J (2018) 17:468 Page 5 of 9 Statistical analysis Hospital and Juansa Health Centre, both in the Asante- Data were entered into spreadsheets using Microsoft Akim North District. The mean age was 22 months and Excel and analysed with the Statistical Package for Social females accounted for nearly 50% (49.8%) of the study Sciences (SPSS) version 17 (SPSS Inc., Chicago, IL, participants. At the time of diagnosis, the mean body USA). Simple descriptive statistics were used to analyse temperature was 37.9 °C (range 35–40.1 °C). the demographic data. The malaria parasite density was log transformed before analysis. Significant levels were Comparison of microscopy, qPCR, HRP2, Combo RDTs measured at 95% confidence intervals and values were Tables  1 and 2 show the results of all four diagnos- considered significant at P < 0.05. Sensitivity and specifi- tic methods deployed in the study: HRP2-RDT 32% cities of the tests were calculated from the TP, TN, FP, (83/260), Combo-RDT 31% (81/260), microscopy 31% and FN test results using the formulae below. (81/260), and qPCR 38% (98/259). Microscopic para- Sensitivity = TP/(TP + FN), Specificity = TN/ site density ranged between 300 and 99,500  parasites/ (TN + FP), Positive predictive value = TP/(TP + FP), µL (Table  1). Thin blood film showed P. falciparum in Negative predictive value = TN/(TN + FN). The values all blood specimens except three individuals who were obtained were expressed as percentages by multiplying positive for P. malariae (two of which were mixed with by 100. P. falciparum). None was positive for P. vivax by qPCR; P. falciparum schizonts were observed in one sample. No Results gametocytes were detected at the microscopic level. Characteristics of study participants There were ten negative samples for qPCR, which were A total of 260 children < 5 years reporting with fever were positive for RDTs and microscopy. With microscopy as recruited from the Konongo-Odumase Government gold standard, the sensitivity of HRP2 and Combo-RDTs Table 1 Comparison of rapid diagnostic test (RDTs) HRP2 (First Response®) and pLDH/HRP2 (Combo®) with microscopy Tests results GS: Microscopya Prevalence Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) (n = 260) Positive Negative HRP2 Positive 77 6 31.3 (25.7–37.3) 95.1 (87.8–98.6) 96.6 (92.8–98.8) 92.8 (85.4–96.6) 97.7 (94.3–99.1) Negative 4 173 Combo Positive 78 3 31.2 (25.9–37.2) 96.3 (89.6–99.2) 98.3 (95.2–99.7) 96.3 (89.4–98.8) 98.3 (95.1–99.4) Negative 3 176 GS gold standard, PPV positive predictive value, NPV negative predictive value, HRP2 histidine-rich protein 2, pLDH lactate dehydrogenase a Mean parasite density 29,721.6/µL (300–99,500/µL), Trophozoites 50, Schizont 1 Table 2 Comparison of  rapid diagnostic test (RDTs) HRP2 (First R esponse®) and  pLDH/HRP2 ( Combo®) and  microscopy with PCR Tests results GS: PCR (n = 246) Prevalence Sensitivity (95% CI) Specificity (95% CI) PPV (95% CI) NPV (95% CI) Positive Negative HRP2 Positive 72 10 39.8 (33.7–46.3) 73.5 (63.6–81.9) 93.2 (87.9–96.7) 87.8 (79.6–93.0) 84.2 (79.2–88.1) Negative 26 138 Combo Positive 71 9 39.8 (33.7–46.3) 72.5 (62.5–81.0) 93.9 (88.8–97.2) 88.6 (80.5–93.8) 83.7 (78.8–87.7) Negative 27 139 Microscopy Positive 70 9 39.8 (33.7–46.3) 71.4 (61.4–80.1) 93.9 (88.8–97.2) 88.6 (80.3–93.7) 83.2 (78.4–87.2) Negative 28 139 GS gold standard, PCR polymerase chain reaction, PPV positive predictive value, NPV negative predictive value. HRP2 histidine-rich protein 2, pLDH lactate dehydrogenase, CI confidence interval Quakyi et al. Malar J (2018) 17:468 Page 6 of 9 was 95.1 and 96.3%, respectively. The sensitivities, spe- children in this group who tested positive on days 4 and cificities and predictive values for RDTs were relatively 28 were referred for further management. higher in microscopy-defined malaria cases than in qPCR positive-defined cases. HRP2‑negative children treated with anti‑malarials In this group (n = 68) (children < 5  years who presented Microscopy and Combo results of HRP2‑negative febrile with fever and tested negative using RDT and received children during 28‑day follow‑up anti-malarial treatment), one child tested positive by HRP2‑negative children not treated with anti‑malarials microscopy, and nine children tested positive by PCR on All febrile children who were initially HRP2-negative day 0. (n = 95) and did not receive anti-malarials were fol- lowed up. Day-0 results of initially HRP2-negative chil- HRP2‑positive children treated with anti‑malarials dren found later to be positive were microscopy (n = 2), A total of five children in this group tested positive on Combo (n = 1) and PCR (n = 17) (Table 3). On days 1 and day 4 for HRP2, microscopy and Combo tests and 10 by 2, five of the children in this group tested positive by PCR PCR. However, on day 28, two children were positive by alone. On day 4, children who were originally HRP2-neg- microscopy and eight by PCR. ative tested positive for microscopy (n = 1) and Combo (n = 1) (Table  3). On day 28, four patients who were Discussion originally HRP2-negative tested positive for microscopy Malaria remains a major public health problem in many (n = 2), Combo (n = 2) and PCR (n = 4) (Table 3). A child countries. In the quest to effectively manage cases, early in Group 3 was positive for all three malaria tests on day diagnosis and prompt treatment with efficacious anti- 4, whereas three were positive only for PCR. On day-28 malarials is advocated [21]. The WHO recommends all follow-up, two children were positive for all three tests, patients receive parasitological confirmation by micros- whereas four children were positive for PCR (Table 3). It copy or RDTs before malaria treatment begins [7]. How- is noteworthy that all children in this group who initially ever, although RDTs are a good alternative to microscopy tested negative by HRP2 and later tested positive with in resource-poor settings, RDTs cannot quantify the microscopy at follow-up, were treated with an appropri- parasite load and are ineffective for diagnosing recently ate anti-malarial and dropped out of the group. More so, treated individuals. The results indicate that both HRP2 and Combo RDTs recorded high sensitivity when microscopy was used as Table 3 Follow-up results for  all tests (microscopy, PCR gold standard. These sensitivity rates are comparable and Combo) after initial testing with HRP2 with reports from previous studies [22], but higher than Follow‑up Test Group 1 Group 2 Group 3 reported by Sani et  al. [23] in Nigeria. It is noteworthy that both sensitivity and specificity values for HRP2 and Day 0 Microscopy 67 1 2 Combo RDTs in this study meet the minimal standard of HRP2 72 0 0 95% for P. falciparum [9]. Most commercially available Combo 70 0 1 RDTs detect PfHRP2 alone or a combination of PfHRP2 PCR 72 9 17 and pLDH. The choice of PfHRP2 is influenced among Day 1 Microscopy – – 0 others by its specificity to the predominant cause of HRP2 – – 0 malaria, P. falciparum. In endemic areas, it is also char- Combo – – 0 acteristic for HRP2 antigen to be produced at the asexual PCR – – 5 and early gametocyte stages of P. falciparum life cycle Day 2 Microscopy – – 0 [24], and its persistence possibly explains the false posi- HRP2 – – 0 tives recorded [25, 26]. In addition, HRP2 antigens are Combo – – 0 produced by the schizonts at an early stage of the para- PCR – – 5 site, even before the parasites are initially released into Day 4 Microscopy 5 1 1 peripheral circulation [22], while pLDH is more con- HRP2 5 1 1 served and is cleared after a relatively shorter period. Combo 5 1 1 Indeed, it has recently been shown that a large propor- PCR 10 0 3 tion of children (up to 25%) treated for malaria based on Day 28 Microscopy 2 3 2 positive HRP2-RDT results were children who were not HRP2 0 0 2 infected with malaria, if microscopy is taken as the gold Combo 0 0 2 standard [27]. Aside from HRP2 persistence, other possi- PCR 8 2 4 ble reasons for false-positive results include non-specific Quakyi et al. Malar J (2018) 17:468 Page 7 of 9 bindings or inference with other immunological or infec- in the result and knowledge of alternative diagnosis [14, tious factors [28–31]. 44], both of which could be enhanced by improving diag- Moreover, the sensitivity of the HRP2 (First R esponse® nostic capacity for other common febrile illnesses and by Malaria Kit) recorded in this study contrasts with that developing evidence-based guidelines for treatment of reported by Ndamukong-Nyanga et  al. (95 vs 48.5%) in symptomatic RDT-negative patients [42]. Cameroon. For the Combo RDT, Xiaodong et  al. [32] The sensitivity of HRP2 and pLDH as diagnostic mark- reported < 90% sensitivity, which is comparable to sensi- ers in P. falciparum has shown a sensitivity of about 95.2 tivity found in the present study. The positive predictive and 98.5%, respectively, from previous works [45, 46] and value (PPV) and negative predictive values (NPV) for the this is similar to the results reported herein (Tables 1, 2 HRP2 and Combo were comparable and both > 92% are and 3). higher than those reported by others elsewhere for HRP2 A major limitation to this study is that the authors were (a PPV of 62.3% and NPV of 75% [33] and for Combo unable to perform any genotyping or sequencing on sam- RDT, PPV of 38.3% and NPV of 14.3% [34]. ples collected on various follow-up visits. It is therefore When PCR was used as reference, HRP2 and Combo not possible to draw definitive conclusions as to whether RDTs recorded lower sensitivity, specificity, PPV, and seropositivity for malaria parasites, antigens/DNA was NPV (Table  3). The higher accuracy by the more sensi- due to a persistent infection or to new infections. tive PCR may be indicative of false-negative RDT results as is often seen in patients with low parasitaemia [35, 36]. However, false-negative RDT results have also been Conclusion reported at high densities, due to the prozone phenom- HRP2- and pLDH-based RDTs showed comparable diag- enon in HRP2-based RDTs [37], suggesting the continu- nostic accuracy in children presenting with an acute ous need for alternative diagnostic markers for effective febrile illness to health facilities in a hard-to-reach screening that are more predictive in field application and rural area in Ghana. However, the presence of discord- suitable for point-of-care application, where resources ant results between the recommended diagnostic tests and expertise to perform advanced laboratory diagnos- on presentation and during follow-up suggest the need tics are unavailable. for improving diagnostic capability for febrile illness in Some recent studies have reported an increase in false- malaria-endemic areas. positive results of HRP2-based RDTs due to mutations in the antigen [19, 38, 39]. Some parasite strains from Africa Authors’ contributionsDS, AAA, PW, IQ, GOA, AL, conceptualized and designed the study. IQ, PW, and South America have been reported to lack the HPR2 GOA, AL and JS implemented study with training of staff. DS, RO, AL, GOA, JS antigens [19, 38–41]. and contributed to analyses of the data. AAA, AL, JS, GOA, IQ, NC, RO, EE, FK In determining which markers are best diagnostic pref- contributed to writing of the manuscript. NC, RO, EE, KS, JS and FS coordi‑nated the field work at the districts and collected data. All authors read and erences for malaria in RDTs, some studies that compare approved the final manuscript. RDT versus microscopy tend to use PCR as a confirma- tory test. The sensitivity, specificity and predictive values Author details1 Department of Biological, Environmental and Occupational Health Sciences, of the Combo RDT were higher than the HRP2 with PCR School of Public Health, College of Health Sciences, University of Ghana, as the gold standard. In view of the fact that HRP2-based Legon, Accra, Ghana. 2 Centre for Tropical Clinical Pharmacology and Thera‑ RDTs are more sensitive than LDH-based RDTs at low peutics, School of Medicine and Dentistry, University of Ghana, Accra, Ghana. 3 Department of Molecular Microbiology and Immunology, Johns Hopkins parasite densities, the findings are in agreement with the Bloomberg School of Public Health, 615 N. Wolfe St, Baltimore, MD 21205, USA. general conclusions that a positive LDH RDT suggests 4 Department of Population, Family, and Reproductive Health, School of Public 5 a parasite density above HRP2 detection threshold [16], Health, University of Ghana, Legon, Accra, Ghana. Department of Health Policy, Planning and Management, School of Public Health, University while microscopy-positive LDH-negative samples may of Ghana, Legon, Accra, Ghana. 6 Department of International Health, Social reflect low density infections. and Behavioural Interventions Program, Johns Hopkins Bloomberg School The use of RDTs can reduce overprescribing of anti- of Public Health, 615 N. Wolfe St, Baltimore, MD 21205, USA. 7 Department of Public and Environmental Wellness, Oakland University, 3101 Human Health malarial drugs, and studies have shown that health Building, 433 Meadow Brook Rd, Rochester, MI 48309‑4452, USA. 8 Depart‑ workers prescribe anti-malarials to patients with nega- ment of Biochemistry, Cell and Molecular Biology, University of Ghana, Legon, tive RDT results [42]. In endemic areas, the presence of Accra, Ghana. 9 National Malaria Control Programme, Ministry of Health, Accra, Ghana. 10 Asante‑Akim Central Municipal Health Directorate, Ghana Health malaria parasites in blood may not necessarily reflect a Services, Konongo, Ghana. 11 Regional Health Directorate, Ghana Health Ser‑ clinical malaria episode [43], while non-compliance to vices, Accra, Ghana. 12 Child and Associates, Accra, Ghana. 13 Kpone Katamanso 14 RDT-negative results by prescribing anti-malarial drugs District Health Directorate, Tema, Ghana. Worldwide Universities Network, University of Ghana, P.O. Box LG 13, Legon, Accra, Ghana. may neglect an underlying infection. Factors associated with compliance to negative RDT results include trust Quakyi et al. Malar J (2018) 17:468 Page 8 of 9 Acknowledgements detection in combination malaria rapid diagnostic tests and implications We are grateful to all the field workers and staff of the District Health Manage‑ for clinical management. Malar J. 2015;14:115. ment for their support in coordination of the study. This Research was sup‑ 17. Baiden F, Bruce J, Webster J, Tivura M, Delmini R, Amengo‑Etego S, ported by CHAI budget. The content is solely the responsibility of the authors et al. Effect of test‑based versus presumptive treatment of malaria in and does not necessarily represent the official views of the funders. under‑five children in rural Ghana—a cluster‑randomised trial. PLoS ONE. 2016;11:e0152960. Competing interests 18. Cheng Q, Gatton ML, Barnwell J, Chiodini P, McCarthy J, Bell D, et al. The authors declare that they have no competing interests. Plasmodium falciparum parasites lacking histidine‑rich protein 2 and 3: a review and recommendations for accurate reporting. Malar J. Publisher’s Note 2014;13:283. 19. Koita OA, Doumbo OK, Ouattara A, Tall LK, Konaré A, Diakité M, et al. Springer Nature remains neutral with regard to jurisdictional claims in pub‑ False‑negative rapid diagnostic tests for malaria and deletion of the lished maps and institutional affiliations. histidine‑rich repeat region of the hrp2 gene. Am J Trop Med Hyg. 2012;86:194–8. Received: 4 March 2018 Accepted: 5 December 2018 20. Jain P, Chakma B, Patra S, Goswami P. Potential biomarkers and their applications for rapid and reliable detection of malaria. Biomed Res Int. 2014;2014:852645. 21. WHO. Expert committee on malaria. Twentieth report. Geneva: World Health Organization; 2000. References 22. Ameyaw B, Mutocheluh M, Owusu M, Khalifah A, Duneeh R, Narkwa P. 1. WHO. World malaria report 2016. Geneva: World Health Organization; Performance of First Response® and CareStart™ malaria rapid diagnostic 2016. tests for the detection of Plasmodium falciparum in a tertiary hospital in 2. Murray CJL, Rosenfeld LC, Lim SS, Andrews KG, Foreman KJ, Haring D, Ghana. Br J Med Med Res. 2017;19:1–7. et al. Global malaria mortality between 1980 and 2010: a systematic 23. Sani UM, Jiya NM, Ahmed H. Evaluation of a malaria rapid diagnos‑ analysis. Lancet. 2012;379:413–31. tic test among febrile children in Sokoto, Nigeria. Int J Med Med Sci. 3. Kyabayinze DJ, Tibenderana JK, Odong GW, Rwakimari JB, Counihan H. 2013;3:433–40. Operational accuracy and comparative persistent antigenicity of HRP2 24. Rock EP, Marsh K, Saul AJ, Wellems TE, Taylor DW, Maloy WL, et al. Com‑ rapid diagnostic tests for Plasmodium falciparum malaria in a hyperen‑ parative analysis of the Plasmodium falciparum histidine‑rich proteins demic region of Uganda. Malar J. 2008;7:221. HRP‑I, HRP‑II and HRP‑III in malaria parasites of diverse origin. Parasitology. 4. Chanie M, Erko B, Animut A, Legesse M. Performance of CareStart™ 1987;95:209–27. Malaria Pf/Pv Combo test for the diagnosis of Plasmodium falciparum 25. Moody A. Rapid diagnostic tests for malaria parasites. Clin Microbiol Rev. and Plasmodium vivax infections in the Afar Region, North East Ethiopia. 2002;15:66–78. Ethiop J Health Dev. 2011;25:206–11. 26. Humar A, Ohrt C, Harrington MA, Pillai D, Kain KC. Parasight F test 5. Thiam S, Thior M, Faye B, Ndiop M, Diouf ML, Diouf MB, et al. Major compared with the polymerase chain reaction and microscopy for the reduction in anti‑malarial drug consumption in Senegal after nation‑wide diagnosis of Plasmodium falciparum malaria in travelers. Am J Trop Med introduction of malaria rapid diagnostic tests. PLoS ONE. 2011;6:e18419. Hyg. 1997;56:44–8. 6. Ishengoma DS, Francis F, Mmbando BP, Lusingu JPA, Magistrado P, Alifran‑ 27. Kiemde F, Bonko MDA, Tahita MC, Lompo P, Rouamba T, Tinto H, et al. gis M, et al. Accuracy of malaria rapid diagnostic tests in community stud‑ Accuracy of a Plasmodium falciparum specific histidine‑rich protein ies and their impact on treatment of malaria in an area with declining 2 rapid diagnostic test in the context of the presence of non‑malaria malaria burden in north‑eastern Tanzania. Malar J. 2011;10:176. fevers, prior anti‑malarial use and seasonal malaria transmission. Malar J. 7. WHO. New perspectives: malaria diagnosis: report of a joint WHO/USAID 2017;16:294. informal consultation. Geneva: World Health Organization; 2000. p. 57. 28. Iqbal J, Sher A, Rab A. Plasmodium falciparum histidine‑rich protein 8. Rosenthal PJ. How do we best diagnose malaria in Africa? Am J Trop Med 2‑based immunocapture diagnostic assay for malaria: cross‑reactivity Hyg. 2012;86:192–3. with rheumatoid factors. J Clin Microbiol. 2000;38:1184–6. 9. Murray CK, Gasser RA, Magill AJ, Miller RS. Update on rapid diagnostic 29. Maltha J, Gillet P, Cnops L, Bottieau E, Van Esbroeck M, Bruggeman C, et al. testing for malaria. Clin Microbiol Rev. 2008;21:97–110. Evaluation of the rapid diagnostic test SDFK40 (Pf‑pLDH/pan‑pLDH) for 10. Ansah EK, Narh‑Bana S, Epokor M, Akanpigbiam S, Quartey AA, Gyapong the diagnosis of malaria in a non‑endemic setting. Malar J. 2011;10:7. J, et al. Rapid testing for malaria in settings where microscopy is available 30. Leshem E, Keller N, Guthman D, Grossman T, Solomon M, Marva E, et al. and peripheral clinics where only presumptive treatment is available: a False‑positive Plasmodium falciparum histidine‑rich protein 2 immuno‑ randomised controlled trial in Ghana. BMJ. 2010;340:635. capture assay results for acute schistosomiasis caused by Schistosoma 11. Reyburn H, Mbakilwa H, Mwangi R, Mwerinde O, Olomi R, Drakeley C, mekongi. J Clin Microbiol. 2011;49:2331–2. et al. Rapid diagnostic tests compared with malaria microscopy for guid‑ 31. Gillet P, Mumba Ngoyi D, Lukuka A, Kande V, Atua B, van Griensven J, et al. ing outpatient treatment of febrile illness in Tanzania: randomised trial. False positivity of non‑targeted infections in malaria rapid diagnostic BMJ. 2007;334:403. tests: the case of human African trypanosomiasis. PLoS Negl Trop Dis. 12. Skarbinski J, Ouma PO, Causer LM, Kariuki SK, Barnwell JW, Alaii JA, 2013;7:e2180. et al. Effect of malaria rapid diagnostic tests on the management of 32. Xiaodong S, Tambo E, Chun W, Zhibin C, Yan D, Jian W, et al. Diagnostic uncomplicated malaria with artemether–lumefantrine in Kenya: a cluster performance of CareStart™ malaria HRP2/pLDH (Pf/pan) combo test randomized trial. Am J Trop Med Hyg. 2009;80:919–26. versus standard microscopy on falciparum and vivax malaria between 13. Hamer DH, Ndhlovu M, Zurovac D, Fox M, Yeboah‑Antwi K, Chanda P, China‑Myanmar endemic borders. Malar J. 2013;12:6. et al. Improved diagnostic testing and malaria treatment practices in 33. Ndamukong‑Nyanga J, Kimbi H, Sumbele I, Emmaculate L, Nweboh M, Zambia. JAMA. 2007;297:2227–31. Nana Y, et al. Assessing the performance characteristics of the “Car‑ 14. Bisoffi Z, Sirima BS, Angheben A, Lodesani C, Gobbi F, Tinto H, et al. Rapid eStartTM Malaria HRP2 pf (CAT NO: G0141, ACCESSBIO)” rapid diagnostic malaria diagnostic tests vs. clinical management of malaria in rural Bur‑ test for asymptomatic malaria in Mutengene, Cameroon. Int J Trop Dis kina Faso: safety and effect on clinical decisions. A randomized trial. Trop Health. 2014;4:1011–23. Med Int Health. 2009;14:491–8. 34. Woyessa A, Deressa W, Ali A, Lindtjørn B. Evaluation of CareStart™ 15. Gatton M, Dunn J, Chaudhry A, Ciketic S, Cunningham J, Cheng Q. Impli‑ malaria Pf/Pv combo test for Plasmodium falciparum and Plasmodium cations of parasites lacking Plasmodium falciparum histidine‑rich protein vivax malaria diagnosis in Butajira area, south‑central Ethiopia. Malar J. 2 on malaria morbidity and control when rapid diagnostic tests are used 2013;12:218. for diagnosis. J Infect Dis. 2017;215:1156–66. 35. Ratsimbasoa A, Randriamanantena A, Raherinjafy R, Rasoarilalao 16. Gatton ML, Rees‑Channer RR, Glenn J, Barnwell JW, Cheng Q, Chiodini N, Ménard D. Which malaria rapid test for Madagascar? Field and PL, et al. Pan‑Plasmodium band sensitivity for Plasmodium falciparum laboratory evaluation of three tests and expert microscopy of samples Quakyi et al. Malar J (2018) 17:468 Page 9 of 9 from suspected malaria patients in Madagascar. Am J Trop Med Hyg. different stages and strains: implications for rapid diagnostic tests. PLoS 2007;76:481–5. ONE. 2011;6:e22593. 36. Harris I, Sharrock WW, Bain LM, Gray KA, Bobogare A, Boaz L, et al. A 41. Amoah LE, Abankwa J, Oppong A. Plasmodium falciparum histidine rich large proportion of asymptomatic Plasmodium infections with low and protein‑2 diversity and the implications for PfHRP‑2 based malaria rapid sub‑microscopic parasite densities in the low transmission setting of diagnostic tests in Ghana. Malar J. 2016;15:101. Temotu Province, Solomon Islands: challenges for malaria diagnostics in 42. Kabaghe AN, Visser BJ, Spijker R, Phiri KS, Grobusch MP, Van Vugt M. an elimination setting. Malar J. 2010;9:254. Health workers’ compliance to rapid diagnostic tests (RDTs) to guide 37. Gillet P, Maltha J, Hermans V, Ravinetto R, Bruggeman C, Jacobs J. Malaria malaria treatment: a systematic review and meta‑analysis. Malar J. rapid diagnostic kits: quality of packaging, design and labelling of boxes 2016;15:163. and components and readability and accuracy of information inserts. 43. Schellenberg JRMA, Smith T, Alonso PL, Hayes RJ. What is clinical malaria? Malar J. 2011;10:39. Finding case definitions for field research in highly endemic areas. Parasi‑ 38. Kumar N, Pande V, Bhatt RM, Shah NK, Mishra N, Srivastava B, et al. tol Today. 1994;10:439–42. Genetic deletion of HRP2 and HRP3 in Indian Plasmodium falciparum 44. Bisoffi Z, Gobbi F, Van den Ende J. Rapid diagnostic tests for malaria. BMJ. population and false negative malaria rapid diagnostic test. Acta Trop. 2014;348:g3846. 2013;125:119–21. 45. Abba K, Deeks JJ, Olliaro P, Naing C‑M, Jackson SM, Takwoingi Y, et al. 39. Wurtz N, Fall B, Bui K, Pascual A, Fall M, Camara C, et al. Pfhrp2 and pfhrp3 Rapid diagnostic tests for diagnosing uncomplicated P falciparum malaria polymorphisms in Plasmodium falciparum isolates from Dakar, Senegal: in endemic countries. Cochrane Database Syst Rev. 2011. https ://doi. impact on rapid malaria diagnostic tests. Malar J. 2013;12:34. org/10.1002/146518 58.CD008 122.pub2. 40. Baker J, Gatton ML, Peters J, Ho MF, McCarthy JS, Cheng Q. Transcrip‑ 46. Wilson ML. Malaria rapid diagnostic tests. Clin Infect Dis. 2012;54:1637–41. tion and expression of Plasmodium falciparum histidine‑rich proteins in Ready to submit your research ? Choose BMC and benefit from: • fast, convenient online submission • thorough peer review by experienced rese archers in your field • rapid publication on acceptance • support for research data, including large and complex data types • gold Open Access which fosters wider collaboration and increased citations • maximum visibility for your research: over 100M website views per year At BMC, research is always in progress. Learn more biomedcentral.com/submissions