Malaria diagnosis using microscopy is currently the gold standard. However, malaria rapid diagnostic tests (mRDTs) were developed to simplify the diagnosis in regions without access to functional microscopy.
The objective of this study was to compare the diagnostic accuracy of mRDT CareStatTM with microscopy.
This study was conducted in the paediatric primary care clinic of the Federal Medical Centre, Asaba, Nigeria.
A cross-sectional study for diagnostic accuracy was conducted from May 2016 to October 2016. Ninety-eight participants were involved to obtain a precision of 5%, sensitivity of mRDT CareStatTM of 95% from published work and 95% level of confidence after adjusting for 20% non-response rate or missing data. Consecutive participants were tested using both microscopy and mRDT. The results were analysed using EPI Info Version 7.
A total of 98 children aged 3–59 months were enrolled. Malaria prevalence was found to be 53% (95% confidence interval [CI] = 46% – 60%), whilst sensitivity and specificity were 29% (95% CI = 20% – 38%) and 89% (95% CI = 83% – 95%), respectively. The positive and negative predictive values were 75% (95% CI = 66.4% – 83.6%) and 53% (95% CI = 46% – 60%), respectively.
Agreement between malaria parasitaemia using microscopy and mRDT positivity increased with increase in the parasite density. The mRDT might be negative when malaria parasite density using microscopy is low.
Malaria is one of the leading causes of morbidity and mortality worldwide, especially in under 5-year-old children and pregnant women in sub-Saharan Africa, where over 80% of cases and at least 90% of malaria deaths occur.
The objectives of this study were to determine the sensitivity and specificity of mRDT and factors that affect these parameters using microscopy as the gold standard, to calculate the positive and negative predictive values of mRDT in the study population and to determine the association between parasite density and mRDT.
We hypothesized that mRDT has sensitivity and specificity similar to microscopy.
The study was conducted in the Children Outpatient Clinic (CHOP) of the Federal Medical Centre, Asaba. The clinic has three consulting rooms that are run by family physicians (Residents and Consultants) in the Family Medicine department. An average of 100 malaria cases are seen monthly amongst under five children presenting to the clinic. Asaba is the capital city of Delta State and shares boundary with Anambra State on the eastern coast of the Niger River.
Although some work has been carried out on this topic, there is paucity of published work on the subject in the South – South geopolitical zone of Nigeria, and the adoption of the mRDT is yet to be popular. This study is therefore aimed at bridging this gap and also to find out the parasite density at which the mRDT will become positive with the aim of making recommendations for its adoption during diagnosis of malaria, especially where the microscopic diagnosis is not feasible.
The study was a cross-sectional study, comparing diagnostic accuracy of malaria using mRDT CareStatTM (a histidine-rich protein-2
Participants were consecutively recruited into the study until the desired sample size was achieved. Finger prick blood sample was obtained, and one thin and one thick blood smears were prepared, stained with 10% Giemsa and read for the presence of malaria parasite by a microscopist. The microscopist was a qualified laboratory scientist with the Federal Medical Centre, Asaba. He has been doing malaria microscopy for 10 years and has been certified as a malaria microscopist. Ten out of the 98 slides were also randomly selected and sent to the parasitology unit of the Medical Microbiology Laboratory of the University College Hospital Ibadan. A drop of blood (about 5 µL) was also taken from the thumb by a dropper that came with the rapid diagnostic test kit. The malaria rapid tests were conducted by the second author, strictly following the instructions on the leaflet in the CareStartTM packs. The sample was introduced into the kit chamber and two drops of the buffer solution were introduced and left for 5–10 min. The results were read and recorded as positive or negative for malaria parasite. The microscopist was independent and hence not aware of the mRDT results obtained. This was performed to remove bias in his interpretation of the blood films sent to the laboratory. The results of the microscopy and the mRDT tests were compared using EPI InfoTM 7 (7.1.5) and the data were summarised using proportions, frequency and percentages. The sensitivity, specificity and predictive values of mRDT were compared with that of the microscopy. Regression analysis was used to determine the relationship between malaria parasite density and the positivity of mRDT.
The CHOP of the Federal Medical Centre, Asaba, Delta State, Nigeria, is being run by the Family Medicine Department, and it provides primary care delivery for children attending the hospital. The hospital is situated along Nnebisi road in the west-end area of the town close to Saint Patrick College, Asaba.
The study population were children who were under 5 years of age attending the Children Outpatient Clinic of the Federal Medical Centre, Asaba. The inclusion criteria were as follows: children whose parents or guardian provided informed consent, children aged between 3 and 59 months, children with axillary temperature ≥ 37.5 °C at presentation or a history of fever within the previous 48 h, children presenting with symptoms and signs comparable with the clinical picture of malaria. The exclusion criteria were as follows: refusal of the parents or guardians to provide informed consent, children with signs of severe illness or unconscious at presentation and those who were enrolled in other clinical studies.
The data were collected using an interviewer-administered questionnaire in English and a translation into pidgin English for parents who did not understand English.
The data collected were analysed using the EPI Info Version 7 statistical package. Parasite density was assessed with the thick film, whilst parasite speciation was assessed with the thin film. The slide was considered negative when no parasite was seen or detected after screening 200 high power fields (see
Diagnostic accuracy of the malaria rapid diagnostic test compared with microscopy.
mRDT Result (Index Test Result) | Microscopy Result (Reference Standard) | Total | |
---|---|---|---|
Positive | Negative | ||
Positive | (TP) | (FP) | (TP + FP) |
Negative | (FN) | (TN) | (FN + TN) |
FN, false negative; FP, false positive; mRDT, malaria rapid diagnostic test; TN, true negative; TP, true positive.
Sensitivity = TP / (TP + FN) X 100.
Specificity = TN / (FP + TN) X 100.
Positive predictive value = TP / (TP + FP) X 100.
Negative predictive value = TN / (FN + TN).
Ethical clearance to conduct the study was obtained from the Federal Medical Centre, Asaba, Nigeria (Ethical Clearance Number: FMC/ASB/T/A81/66) on 16 December 2015. The study was conducted in compliance with International Conference on Harmonisation – Good Clinical Practice (ICH GCP) and the Declaration of Helsinki.
A total of 98 children aged between 3 and 59 months were recruited and enrolled in this study. There were 53 males and 45 females, giving a male–female ratio of 1:0.85. Most of the participants were in the age group of 3–24 months, accounting for 50 members (51.02%) of the total participants. The mean age ± standard deviation (s.d.) of the participants was 26.2 months ± 15.7 (range = 3–59 months) as shown in
Age category of participants in months.
Age range(months) | Proportion | Percentage | Gender | Mean age | ±s.d. | |
---|---|---|---|---|---|---|
Male | Female | |||||
3–12 | 25 | 25.5 | - | - | - | - |
13–24 | 25 | 25.5 | - | - | - | - |
25–36 | 20 | 20.4 | 54 | 46 | 26.2 | 15.7 |
37–48 | 19 | 19.4 | - | - | - | - |
49–59 | 9 | 9.2 | - | - | - | - |
s.d., standard deviation.
The most common clinical symptom was fever. Eighty-eight of the participants (89.8% [95% CI = 82.03% – 95.0%]) presented with fever. About 74.5% of the participants had raised temperature of ≥ 37.5 °C, with a mean ± s.d. temperature of 38.0 °C ± 0.9 °C (range = 36.0 °C – 39.9 °C) as shown in
Clinical features of the participants.
Characteristics | Frequency |
Percentage | 95% CI |
---|---|---|---|
Fever | |||
At presentation | 88 | 89.8 | 82.23–94.37 |
Before presentation | 83 | 84.7 | 76.27–90.50 |
Refusal of feeds | |||
At presentation | 65 | 66.3 | 56.52–74.91 |
Before presentation | 51 | 52.0 | 42.26–61.67 |
Irritability | |||
At presentation | 26 | 26.5 | 18.80–36.04 |
Before presentation | 16 | 16.3 | 10.31–24.89 |
Vomiting | |||
At presentation | 32 | 32.7 | 24.17–42.44 |
Before presentation | 29 | 29.6 | 21.46–39.26 |
Raised temperature > 37.5 °C | |||
At presentation | 73 | 74.5 | 65.05–82.08 |
Mean temperature ±s.d. | 38.0°C ± 0.9°C | - | 36.0–39.9 |
s.d., standard deviation; CI, confidence interval.
Both microscopy and mRDT were conducted on every participant. Fifty-two children (53.06% [95% CI = 40.72% – 61.26%]) out of 98 were found positive for the microscopy test, whilst 20 (20.41% [95% CI = 12.93% – 29.74%]) were positive with the mRDT test as shown in
Flow chart showing the result of microscopy test (the gold standard).
Comparison of the microscopy with the malaria rapid diagnostic test.
Of the 20 participants who tested positive with the rapid diagnostic test, 19 gave a history of fever at the time of presentation. However, there was no significant relationship between fever and positivity of the rapid diagnostic test as shown in
Relationship between fever and malaria rapid diagnostic test positivity.
Fever at presentation | Positive mRDT | Total | Chi square | ||||
---|---|---|---|---|---|---|---|
Yes | No | ||||||
% | % | ||||||
Yes | 19 | 21.6 | 69 | 78.4 | 88 | 0.741 | 0.64 |
No | 1 | 10% | 9 | 90% | 10 | - | - |
mRDT, malaria rapid diagnostic test.
There was no significant relationship between the temperature at presentation and the malaria parasite count as shown in
Relationship between parasite count and temperature.
Temperature | Results of microscopy | Chi-square | |||||||
---|---|---|---|---|---|---|---|---|---|
+ | ++ | +++ | ++++ | Nil | Total | ||||
High temperature (≥ 37.5°C) | 28 | 7 | 5 | 0.0 | 33 | 73 | 3.94 | 0.21 | |
% | 38.4 | 9.6 | 6.8 | 0.0 | 45.2 | 100 | - | - | |
Normal temperature (< 37.5°C) | 7 | 4 | 0 | 1 | 13 | 25 | - | - | |
% | 28.0 | 16.0 | 0.0 | 4.0 | 52.0 | 100 | - | - | |
The linear regression plot of the relationship between the natural logarithm of the parasite density and the parasite count of the participants with malaria parasitaemia is shown in
Linear regression of the parasite count against natural logarithm of parasite density of participants with malaria parasitaemia.
The percentage agreement of positive results of mRDT and parasite count using microscopy of +, ++, +++ and ++++ was 14.3%, 45.5%, 80% and 100%, respectively. The percentage agreement of negative results was 10.9%. The proportion is shown in
Comparing malaria rapid diagnostic test positivity with parasite count.
Result mRDT | Results of microscopy | Total | |||||||||
---|---|---|---|---|---|---|---|---|---|---|---|
+ | ++ | +++ | ++++ | Nil | |||||||
% | % | % | % | % | |||||||
Negative | 30 | 85.3 | 6 | 54.6 | 1 | 20.0 | 0 | - | 41 | 89.4 | 78 |
Positive | 5 | 14.3 | 5 | 45.5 | 4 | 80.0 | 1 | 100 | 5 | 10.9 | 20 |
mRDT, malaria rapid diagnostic test.
The percentage agreement of positive results of mRDT and parasite count using microscopy is calculated as follows:
The sensitivity and specificity of the mRDT were 29% and 89%, respectively, whilst the positive and negative predictive values were 75% and 54%, respectively. The prevalence of malaria using microscopy in this study population was 53%. The false positive and false negative rates were 10.9% and 71.2%, respectively, as shown in
Diagnostic accuracy of malaria rapid diagnostic test using microscopy as the gold standard.
mRDT Result (Index Test Result) | Microscopy result (Reference Standard) | Total | |
---|---|---|---|
Positive | Negative | ||
Positive | 15 | 5 | 20 |
(TP) | (FP) | (TP + FP) | |
Negative | 37 | 41 | 78 |
(FN) | (TN) | (FN + TN) | |
FN, false negative; FP, false positive; mRDT, malaria rapid diagnostic test; TN, true negative; TP, true positive.
This study aimed at comparing the diagnostic accuracy of mRDT with microscopy amongst under five children so as to deploy mRDT for prompt diagnosis and treatment of malaria amongst children presenting to our hospital.
The prevalence of malaria in this study using microscopy as a reference diagnostic test was 53% (95% CI = 46% – 60%). This value was lower than that recorded by Samadoulougou et al.
The species of malaria parasite identified in all study participants was
The result revealed that 52 children (53.1% [95% CI = 40.7% – 61.3%]) out of the 98 were positive for the microscopy test. Thirty seven (71.2%) out of the 52 children who were positive for the microscopy were found to be negative with mRDT (false negative). This gave a high false negative mRDT test when compared with the result of the microscopy in this study. This is very significant in this study, as this may have contributed to the low sensitivity reported. Sensitivity is the proportion of people with disease (malaria) who will have a positive result when tested with the diagnostic test kit (mRDT in this case) in the diagnosis of malaria.
The microscopy further showed that the parasite count of (+) made up 67% of the total population of patients with positive microscopy. Using this ‘plus’ system scale of scoring to calculate the parasite density, it therefore means that about 67% of the participants who had positive results with microscopy had a parasite density of between 10 and 90 parasites/µL of blood.
The sensitivity and specificity of the mRDT in this study were 29% and 89%, respectively. This means that the mRDT kit (CareStatTM) used in this study will be capable of detecting correctly (giving a positive result) only 29 out of 100 children with malaria infection and will give a negative result in 89 out of 100 patients without malaria infection. The very low sensitivity recorded in this study as against the WHO recommendations of about 95% may be because of a high false negative rate of 71.2% (37/52 × 100) of mRDT as compared to the microscopy. The high false negative rate is similar to the findings of Berhane et al. where only 10 out of the 50 microscopically confirmed
The low sensitivity of this study is in agreement with the research conducted by Oyeyemi et al. in Ijebu Ode, western part of Nigeria, who reported a sensitivity and specificity of 42.5% and 87.1%, respectively.
It was also noted from this study that five (10.9%) of 46 children whose microscopy results were negative were positive with the rapid test (false positive). This may be as a result of persistent antigen of the malaria parasite in the blood even after parasite clearance following adequate anti-malaria treatment of the index cases. The persistent antigenaemia may have contributed to the high specificity recorded in this study. This agreed with the work of Batwala et al.
The percentage agreement of positive results of mRDT and parasite count using microscopy was the highest (100%) at parasite count of (++++) and the lowest (14.3%) with parasite count of (+). Many of these (+) using microscopy were missed by the rapid test, thereby giving a low yield in the positivity of the mRDT and, consequently, the sensitivity of the malaria kit at this level of parasite count. This result agreed with the work of Sani et al. in Sokoto, Northern Nigeria, where it was found that the sensitivity of RDT increased consistently from 33% at low parasite density to 93% at high parasite density.
The specificity of this study was comparable with most of other researches. However, the low sensitivity of this study agreed with the work carried out by Kahama-Maro et al. in Dar es Salaam, who found a low sensitivity. The low sensitivity in this study may not be completely explained only by the parasite density of the malaria as documented by researchers like Mawili-Mboumba.
The positive and negative predictive values in this study were 75% and 53%, respectively. This result is slightly different from the findings of Falade et al.,
The false positive and negative rates in this study were 10.9% and 71.2%, respectively. This false negative rate is quite high. Several factors may account for this high rate, which may include low parasite density. According to WHO, false-negative results can be caused by any or a combination of the following:
the procurement and use of poor-quality RDTs
use of the wrong comparator for the diagnostic test, such as poor-quality microscopy for cross-checking negative RDT
poor transport and storage conditions for RDTs, with sustained exposure to high temperature
operator errors during performance and/or interpretation of RDT results (more rarely)
deletion or mutation of
The linear regression plot (
There was only one microscopist who regularly performed malaria microscopy for clinical care of patients in the study location where the volume of work could sometimes be high. There was no cross-checking of a predetermined percentage of the slides by a second microscopist.
The sensitivity and specificity of mRDT compared with microscopy diagnosis of malaria in this study were found to be 29% and 89%, respectively. There was a significant correlation between parasite count and parasite density (
The authors thank Prof. Catherine O. Falade for her assistance and role at various stages of this research.
The authors have declared that no competing interests exist.
O.O. (Lincoln University College, Selangor, Malaysia) was responsible for the experimental and project design, protocol writing, data analysis, manuscript writing and review of the article. B.A.O. (The Ark Medical Centre, Asaba, Nigeria) performed the experiment and participated in the protocol writing and review of the manuscript. A.I.N. (Federal Medical Centre, Asaba, Nigeria) was involved in the writing of the protocol and review of the manuscript.
The research received no specific grant from any funding agency in the public, commercial or not-for-profit sectors.
Data sharing is not applicable to this article as no new data were created or analysed in this study.
The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of any affiliated agency of the authors.