Abstract
Background
In India, Plasmodium falciparum and Plasmodium vivax remain in circulation. Accurate detection of the parasite species remains crucial for prompt initiation of treatment and reducing onward transmission.
Methods
A cross-sectional study across 12 sites of varying malaria endemicities was conducted from September 2023 to April 2024. Febrile participants were tested for malaria using rapid diagnostic tests (RDTs) and microscopy. Malaria positivity proportions along with 95% confidence intervals (95% CI) were presented separately by parasite species. The diagnostic performance of the RDT was compared against microscopy.
Results
A total of 10,290 febrile participants were tested by both RDT and microscopy: 1,516 (14.7%, 95% CI 7.7–21.8%) malaria cases were identified by RDT and 1,436 (14.0%, 95% CI 6.9–21.1%) by microscopy. Of the 1,516 RDT positives, 1,105 (72.9%) had P. falciparum mono-infection, 290 (19.1%) had P. vivax mono-infection, and 121 (8.0%) had P. falciparum and P. vivax mixed infections. The sensitivity and specificity of RDT were 95.0% [95% CI 94–96%] and 99% [95% CI 98–99%], respectively, for detecting P. falciparum mono-infection, 83% [95% CI 78–87%] and 100% [95% CI 99–100%] for detecting P. vivax mono-infection, and 88% [95% CI 80–93%] and 100% for detecting a mixed infection of P. falciparum and P. vivax. Overall, 43 (0.4%) participants who were RDT negative were found to have malaria on subsequent microscopic examination.
Conclusion
Approximately 15% of the febrile participants tested were identified as malaria positive by RDT, of which nearly one-fifth were P. vivax mono-infections and 8% harboured P. falciparum and P. vivax mixed infections. Low sensitivity of the RDTs for identifying P. vivax underscores an urgent need for developing reliable diagnostics.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12936-025-05556-7.
Keywords: Plasmodium falciparum, Plasmodium vivax, India, Malaria elimination, Survey, Transmission
Background
Early and accurate disease diagnosis is one of the cornerstones of global malaria control and elimination. Microscopy of peripheral blood smears remains the ‘gold standard’ for malaria diagnosis [1, 2], but rapid diagnostic tests (RDTs) can be used where microscopy is not feasible. The operational suitability makes RDTs a feasible alternative for deployment in remote areas with limited access to healthcare and among hard-to-reach population [3, 4]. Therefore, RDTs remain a crucial component in the management and control of malaria globally. In 2022, approximately 345 million RDTs were distributed worldwide through the national malaria programmes, of which 12.1 million were distributed in the WHO South-East Asia region [5].
The WHO South-East Asia region has reported a consistent decline in malaria cases over the past 15 years, with confirmed cases reducing from 2.6 million in 2010 to 809,000 in 2022. Of the latter, India has accounted for approximately two-thirds, with both Plasmodium falciparum and Plasmodium vivax contributing to the disease burden [6]. A majority of the cases of P. vivax and P. falciparum in India have emerged from a few states: Odisha, Jharkhand, Chhattisgarh, West Bengal, Tripura, Mizoram, Maharashtra, and Uttar Pradesh [6]. Across these states, RDTs have remained the mainstay of malaria control efforts. However, several diagnostic-related challenges have emerged in some of these states in recent times, such as the failure to detect low-density parasitaemia and limited accuracy to detect mixed infections [7–9].
Some of the most widely used RDTs target histidine-rich proteins (PfHRP2) encoded by the pfhrp2 gene for detection of P. falciparum malaria. Detection of P. vivax malaria utilizes Plasmodium lactate dehydrogenase (pLDH) enzymes, which is present in metabolically viable P. vivax parasites. Partial or a complete deletion of the pfhrp2/3 genes results in minimal or no PfHRP2 production, and allows parasites to evade detection of P. falciparum malaria by RDTs leading to false-negative results [10]. Parasites harbouring partial or complete deletions of the pfhrp2/3 genes have also been reported in several studies from India [11, 12]. The emergence of such “diagnostic-resistant” P. falciparum malaria has thus jeopardised malaria control efforts in India and the broader region. Therefore, timely and efficient monitoring of pfhrp2/3 gene deletion status is crucial to inform the optimal diagnostic strategies. This is particularly important as the triad of resistance against antimalarials (in particular artemisinin) [13, 14], diagnostics, and insecticides threatens to seriously undermine the sustainability of the progress made over the past two decades. To evaluate the operational accuracy (diagnostic accuracy) of the current RDTs in field deployment and assess the current malaria status, a survey was undertaken across 12 different sites in India, representing varying levels of transmission endemicity, from September 2023 to April 2024.
This study provides an estimate of malaria positivity among screened febrile participants and evaluates the diagnostic performance of rapid diagnostic tests (RDTs) against microscopy. The findings will assist the national programme in identification of high-burden areas, aid policymakers in tracking progress towards malaria elimination goals, guide resource allocation, and ultimately contribute towards optimizing strategies and resources in the combat against malaria.
Methods
Study design
A cross-sectional study was conducted across 12 different sites in nine states of India of varying malaria endemicities between September 2023 and April 2024 (Fig. 1).
Fig. 1.
Study sites and overall malaria endemicity
Selection of study sites
The study sites were selected in two stages. In the first stage, seven sites were chosen: Dantewada (Chhattisgarh), West Singhbhum (Jharkhand), Lunglei & Lawngtlai (Mizoram), Balaghat (Madhya Pradesh), Kothagudem (Telangana), and South Tripura (Tripura). The selection was based on pragmatic considerations such as (i) ability to access the sites as some of the endemic regions are in remote forest areas with limited transportation network, (ii) capacity to collect and process study samples, and (iii) the capacity of the research team to diagnose and treat participants presenting with malaria. In the second stage, an additional five sites were selected based on outputs of a geostatistical modelling (manuscript in preparation) and these included: Bokaro (Jharkhand), North Goa (Goa), Uttara Kannada and Mangalore (Karnataka), and Lakhimpur (Assam).
Description of study sites
Balaghat district in Madhya Pradesh has dense forests, hilly terrain, perennial streams, and experiences perennial malaria transmission. Villages in the district become inaccessible during the monsoon (July–September), and both P. falciparum and P. vivax are endemic. Dantewada in Chhattisgarh, along with six districts in the Bastar division has a year-round transmission, primarily of P. falciparum; the division has dense forest cover, and remote tribal settlements. In Jharkhand, West Singhbhum is a high-transmission zone, with intense seasonal peaks during the monsoon, whereas Bokaro has a low to moderate seasonal transmission. Kothagudem in Telangana has forested terrain and faces seasonal to perennial transmission. South Tripura has perennial transmission with peaks during monsoon, and has a large, susceptible tribal population. Lunglei and Lawngtlai in Mizoram also report perennial transmission, exacerbated by cross-border movement. Lakhimpur district in Assam experiences low seasonal transmission. In Karnataka, both Mangalore and Uttar Kannada report low transmission. Goa, although having achieved zero indigenous cases in 2023, faces the risk of imported transmission, particularly among migrant workers, necessitating continued surveillance. The status of malaria endemicity at the selected study sites is presented in Fig. 1.
Screening of the study participants and recruitment
Study participants were recruited from various community health centres (CHCs) within the district identified for surveillance. The selected CHCs had sufficient facilities and resources for case management. Those who presented with signs and symptoms of malaria (including but not limited to headache, body aches, fever, chills, and weakness), aged older than 1 year, and providing informed consent/assent were screened for malaria status using an RDT and microscopy.
Diagnostics and sample collection
The bivalent SD Bioline Malaria Ag P.f./P.v. rapid diagnostic test (RDT), which detects Plasmodium falciparum-specific histidine-rich protein 2 (HRP2) and Plasmodium vivax-specific parasite lactate dehydrogenase (Pv-pLDH) antigens, was used alongside blood smear examination to screen febrile patients in field settings by trained staff. The slides prepared in the field were validated by two WHO certified level-1 microscopists at Indian Council of Medical Research-National Institute of Malaria Research (ICMR-NIMR), New Delhi. In case of discrepancy, a third independent microscopist served as a tie-breaker. For each sample, thick and thin smears were prepared and stained with 10% Giemsa stain. The smears were allowed to air-dry and were protected from dust, labelled with a unique ID. Basic information collected at the time of screening included the age and sex of the participant along with the date of testing and their geographic location. This study was a cross-sectional survey and there were no participant follow-up visits. Name, address, and other personal identifiers were not recorded in the study database. Fully anonymised data were archived indefinitely at the ICMR-NIMR Delhi, as per the institutional guidelines.
Quality assurance
All study personnel involved in the implementation first received on-site training on the principles of good clinical practice (GCP), informed consent/assent procedures, specimen collection, labelling, storage, and shipment before recruitment commenced. All the logistics were shipped to the study site in batches. Feedback on sample quality and suggestions for corrective action were sent to the site personnel.
Sample size approximation
The sample size at each participating site was determined by logistical limitations rather than statistical considerations. The primary limitation was the number of malaria cases, which is declining at many sites with ongoing malaria control efforts. This study aimed to detect approximately 203 P. falciparum positive patients per site. The precision of the estimated proportion of 5% (95% confidence interval, 95% CI) will be ± 3% with this sample size. Further assuming no results can be obtained from ~ 8–10% of the specimens, a target sample size of 220 P. falciparum positive cases was desired at each survey site.
Statistical analysis
Malaria positivity proportion was estimated separately for each study sites and presented with 95% CI computed using Wilson’s method. An overall malaria positivity proportion was estimated by combining data across the 12 sites; 95% CI was estimated by adjusting for within-site correlation using Fleiss’s method [15]. The results were presented separately by the diagnostic methodology used (RDT or microscopy), and separately by the parasite species (P. falciparum, P. vivax, mixed infection, or other species). The accuracy of the RDT was compared against the results from microscopy, and the following performance measures computed using epiR library in R software were reported: sensitivity, specificity, and negative and positive predictive values (NPVs and PPVs) [16]. Due to the occurrence of triple zeros across multiple sites, the performance measures were summarised by pooling data across the studies without undertaking meta-analysis and the results were presented separately for each of the parasite species. All analyses were conducted in R (R Studio) [17].
Informed consent
All prospective participants were informed about the study in both oral and written forms. The written consent documents embodied the elements of informed consent described in the current edition of the Declaration of Helsinki and adhered to the ICH guidelines. Prospective participants were informed that their decision regarding participation would have no impact on the care they received and those not consenting to participate received standard care per local protocol.
Ethics approval
Ethics approval for the study was granted by the Institutional Ethics Committee of NIMR (NIMR/IEC-M/2023/970/V2-OCT/14).
Results
During the study period (September 2023 to April 2024), 10,290 febrile participants who consented to inclusion in the study were tested by RDT and subsequently by microscopy across 12 different study locations (Fig. 1, Table 1). Of these, 4,237 (41.2%) participants were from four highly endemic sites (Lawngtlai, Dantewada, Lunglei, and West Singhbhum), 3,477 (33.8%) were from three moderately endemic sites (Kothagudem, South Tripura, and Balaghat), and 2,576 (25.0%) were from 5 sites with low malaria endemicity (North Goa, Lakhimpur, Mangalore, Uttara Kannada, and Bokaro). Overall, 5,349 (52%) participants were female and 4,941 (48%) male. A total of 1,022 (9.9%) participants were < 5 years old, 2,774 (27.0%) were between 5 and < 15 years old, and the remaining 6,494 (63.1%) were aged 15 years or older (See supplemental file S1).
Table 1.
Malaria diagnosis using rapid diagnostic test and microscopy
| Study site | Endemicity | Sample period | Method | Tested | Negative | Pf monoinfection | Pv monoinfection | Pm monoinfection | Pf + Pv | Pf + Pm |
|---|---|---|---|---|---|---|---|---|---|---|
| Bokaro (Jharkhand) | Low | 11/2023–12/2023 | RDT | 241 | 241 (100%) | 0 (0.0%) | 0 (0.0%) | – | 0 (0.0%) | – |
| Microscopy | 215 | 215 (100%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | |||
| Lakhimpur (Assam) | Low | 10/2023–02/2024 | RDT | 644 | 641 (99.5%) | 1 (0.2%) | 2 (0.3%) | – | 0 (0.0%) | – |
| Microscopy | 644 | 641 (99.5%) | 1 (0.2%) | 2 (0.3%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | |||
| Mangalore (Karnataka) | Low | 11/2023–04/2024 | RDT | 642 | 620 (96.6%) | 4 (0.6%) | 18 (2.8%) | – | 0 (0.0%) | – |
| Microscopy | 642 | 620 (96.6%) | 4 (0.6%) | 18 (2.8%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | |||
| North Goa (Goa) | Low | 01/2024–03/2024 | RDT | 647 | 636 (98.3%) | 11 (1.7%) | 0 (0.0%) | – | 0 (0.0%) | – |
| Microscopy | 647 | 636 (98.3%) | 11 (1.7%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | |||
| Uttar Kannada (Karnataka) | Low | 01/2024–02/2024 | RDT | 402 | 401 (99.8%) | 1 (0.2%) | 0 (0.0%) | – | 0 (0.0%) | – |
| Microscopy | 402 | 401 (99.8%) | 1 (0.2%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | |||
| South Tripura (Tripura) | Moderate | 11/2023–03/2024 | RDT | 906 | 869 (95.9%) | 29 (3.2%) | 7 (0.8%) | – | 1 (0.1%) | – |
| Microscopy | 904 | 876 (96.9%) | 22 (2.4%) | 6 (0.7%) | 0 (0.0%) | 0 (0.0%) | 0 (0.0%) | |||
| Balaghat (Madhya Pradesh) | Moderate | 09/2023–10/2023 | RDT | 890 | 618 (69.4%) | 227 (25.5%) | 17 (1.9%) | – | 28 (3.1%) | – |
| Microscopy | 888 | 632 (71.2%) | 222 (25%) | 10 (1.1%) | 0 (0.0%) | 24 (2.7%) | 0 (0.0%) | |||
| Kothagudem (Telangana) | Moderate | 10/2023–11/2023 | RDT | 1,681 | 1,549 (92.1%) | 121 (7.2%) | 6 (0.4%) | – | 5 (0.3%) | – |
| Microscopy | 1,681 | 1,571 (93.5%) | 104 (6.2%) | 5 (0.3%) | 0 (0.0%) | 1 (0.1%) | 0 (0.0%) | |||
| West Singhbhum (Jharkhand) | High | 11/2023–02/2024 | RDT | 614 | 310 (50.5%) | 259 (42.2%) | 30 (4.9%) | – | 15 (2.4%) | – |
| Microscopy | 614 | 317 (51.6%) | 248 (40.4%) | 34 (5.5%) | 0 (0.0%) | 15 (2.4%) | 0 (0.0%) | |||
| Dantewada (Chhattisgarh) | High | 09/2023–10/2023 | RDT | 1,059 | 834 (78.8%) | 182 (17.2%) | 23 (2.2%) | – | 20 (1.9%) | – |
| Microscopy | 1,058 | 865 (81.8%) | 163 (15.4%) | 16 (1.5%) | 0 (0.0%) | 14 (1.3%) | 0 (0.0%) | |||
| Lawngtlai (Mizoram) | High | 10/2023–03/2024 | RDT | 1,671 | 1,286 (77%) | 191 (11.4%) | 146 (8.7%) | – | 48 (2.9%) | – |
| Microscopy | 1,668 | 1,271 (76.2%) | 176 (10.6%) | 174 (10.4%) | 2 (0.1%) | 45 (2.7%) | 0 (0.0%) | |||
| Lunglei (Mizoram) | High | 11/2023–03/2024 | RDT | 893 | 769 (86.1%) | 79 (8.8%) | 41 (4.6%) | – | 4 (0.4%) | – |
| Microscopy | 891 | 773 (86.8%) | 73 (8.2%) | 39 (4.4%) | 0 (0.0%) | 5 (0.6%) | 1 (0.1%) | |||
| Overall | Overall | 09/2023–04/2024 | RDT | 10,290 | 8,774 (85.3%) | 1,105 (10.7%) | 290 (2.8%) | – | 121 (1.2%) | – |
| Microscopy | 10,254 | 8,818 (86.0%) | 1,025 (10.0%) | 304 (3.0%) | 2 (0.0%) | 104 (1.0%) | 1 (0.0%) |
Percentage in parenthesis represents row percentages. All percentage rounded to 1 decimal place. RDT rapid diagnostic test. The bold test represents overall data pooled from all the study sites. Pf = P. falciparum. Pv = P. Vivax. Pm = P.malariae
Overall malaria positivity using RDT and microscopy
Of the 10,290 participants, 1,516 (14.7%, 95% confidence interval (CI) 7.7–21.8%) tested positive for malaria (any species; either mono-infection or a mixed infection) by RDT (Fig. 2). Paired microscopic slide results were not available in 36 (0.3%) participants; reasons included damage during transportation, staining issues, and poor-quality smear. Of the 36 with missing microscopic results, 34 was RDT negative (any malaria) and 2 were RDT positive for P. falciparum. Among 10,254 participants in whom microscopy slides were available, 1,436 (14.0%, 95% CI: 6.9%–21.1%) tested positive for malaria (any species).
Fig. 2.
Malaria prevalence using microscopy and RDT for each of the study sites. 95% CI estimated using Wilson’s method. Current endemicity status for each site shown in parenthesis
Malaria positivity estimates across the study sites are presented in Table 1 and Fig. 2. In the high endemicity sites, the proportion of participants who tested positive for malaria by RDT and microscopy were 21.2% (225/1,059) and 18.2% (193/1,058) respectively in Dantewada, 23.0% (385/1,671) and 23.8% (397/1,668) in Lawngtlai, 13.9% (124/893) and 13.2% (118/891) in Lunglei, and 49.5% (304/614) and 48.4% (297/614) in West Singhbhum. The corresponding estimates in moderately endemic sites were 30.6% (272/890) and 28.8% (256/888) in Balaghat, 7.9% (131/1,681) and 6.5% (110/1,681) in Kothagudem, and 4.1% (37/906) and 3.1% (28/904) in South Tripura. The corresponding estimates in areas of low endemicities were: 0% (0/241) and 0% (0/215) in Bokaro, 0.5% (3/644) and 0.5% (3/644) in Lakhimpur, 3.4% (22/642) and 3.4% (22/642) in Mangalore, 1.7% (11/647) and 1.7% (11/647) in North Goa, and 0.2% (1/402) and 0.2% (1/402) in Uttara Kannada. Further details are presented in supplemental file S1.
Causative parasite species
Of the 1,516 RDT positives, 1,105 (72.9%) were P. falciparum mono-infection, 290 (19.1%) were P. vivax mono-infection, and the remaining 121 (8.0%) had P. falciparum and P. vivax mixed infections. Of the 1,436 who tested positive using microscopy, 1,025 (71.4%) had P. falciparum mono-infection, 304 (21.2%) had P. vivax mono-infection, 2 (0.1%) were identified as P. malariae mono-infection, 104 (7.2%) presented with a mixed P. falciparum and P. vivax, and 1 (0.1%) patient had a mixed P. falciparum and P. malariae. Species specific breakdown of the malaria status is presented in Table 1.
Performance of RDT for detecting P. falciparum mono-infection
The diagnostic accuracy of the RDTs in detecting malaria by causative parasite species is presented in Tables 2, 3, and 4. Among 10,254 participants for whom both the RDT and microscopy results were available, 8,987 (87.3%) tested negative by both methods i.e., microscopy and RDT, and 1,087 (10.6%) tested positive by both methods. Further 43 participants (0.4%) were RDT negative but were subsequently found to be microscopy positive, while the remaining 137 (1.3%) were RDT-positive but subsequently tested negative by microscopy. Overall, this resulted in sensitivity and specificity estimates for the RDT (compared to microscopy) of 95.0% [95% CI 93.0–97.0%] and 98.0% [95% CI 97.0–98.0%] respectively in high endemicity areas, 95.0% [95% CI 92.0–97.0%] and 99.0% [95% CI 98.0–99.0%] in the areas of moderate endemicities, and 100% sensitivity and specificity were observed in the areas of low endemicities (Table 3). Pooled across the study sites, the overall sensitivity and specificity estimates of the RDT was 95.0% [95% CI 94.0–96.0%] and 99.0% [95% CI 98.0–99.0%] respectively (Table 2). See Tables 2 and 3 for the predictive values and further accuracy measures.
Table 2.
Accuracy of rapid diagnostic test for detecting malaria infection, by parasite species
| Parasite species | Microscopy | RDT | Sensitivity [95% CI] | Specificity [95% CI] | PPV [95% CI] | NPV [95% CI] |
|---|---|---|---|---|---|---|
| (positive/tested) | (positive/tested) | |||||
| Pf mono-infection | 1,025/10,254 | 1,103/10,254 | 0.95 [0.94–0.96] | 0.99 [0.98–0.99] | 0.88 [0.86–0.90] | 0.99 [0.99–1.00] |
| Pv mono-infection | 304/10,254 | 290/10,254 | 0.83 [0.78–0.87] | 1.00 [0.99–1.00] | 0.87 [0.82–0.90] | 0.99 [0.99–1.00] |
| Pf + Pv mixed infection | 104/10,254 | 121/10,254 | 0.88 [0.80–0.93] | 1.00 | 0.75 [0.67–0.83] | 1.00 |
| Any malaria | 1,436/10,254 | 1,514/10,254 | 0.96 [0.95–0.97] | 0.98 [0.98–0.99] | 0.91 [0.89–0.92] | 0.99 [0.99–0.99] |
RDT rapid diagnostic test, NPV negative predictive value, PPV positive predictive value; 95% confidence interval (95% CI) are based on exact method
Table 3.
Accuracy of the rapid diagnostic test for detecting Pf mono-infection, by malaria endemicity
| Endemicity | Microscopy | RDT | Sensitivity [95% CI] | Specificity [95% CI] | PPV [95% CI] | NPV [95% CI] |
|---|---|---|---|---|---|---|
| (positive/tested) | (positive/tested) | |||||
| High (n = 4,231) | 660/4,231 | 710/4,231 | 0.95 [0.93–0.97] | 0.98 [0.97–0.98] | 0.88 [0.86–0.91] | 0.99 [0.99–0.99] |
| Moderate (n = 3,473) | 348/3,473 | 376/3,473 | 0.95 [0.92–0.97] | 0.99 [0.98–0.99] | 0.88 [0.84–0.91] | 0.99 [0.99–1.00] |
| Low (n = 2,550) | 17/2,550 | 17/2,550 | 1.00 [0.80–1.00] | 1.00 | 1.00[0.80–1.00] | 1.00 |
RDT rapid diagnostic test, NPV negative predictive value, PPV positive predictive value; 95% confidence interval (95% CI) are based on exact method
Table 4.
Accuracy of the RDT for detecting Pv mono-infection, by malaria endemicity
| Endemicity | Microscopy (positive/tested) | RDT (positive/tested) | Sensitivity [95% CI] | Specificity[95% CI] | PPV [95% CI] | NPV [95% CI] |
|---|---|---|---|---|---|---|
| High (n = 4,231) | 263/4231 | 240/4231 | 0.81 [0.76–0.86] | 0.99 [0.99–1.00] | 0.89 [0.85–0.93] | 0.99 [0.98–0.99] |
| Moderate (n = 3,473) | 21/3473 | 30/3473 | 0.81 [0.58–0.95] | 1.00 [0.99–1.00] | 0.57 [0.37–0.75] | 1.00 |
| Low (n = 2,550) | 20/2550 | 20/2550 | 1.00 [0.83–1.00] | 1.00 | 1.00 [0.83–1.00] | 1.00 |
RDT rapid diagnostic test, NPV negative predictive value, PPV positive predictive value; 95% confidence interval (95% CI) are based on exact method
Performance of RDT for detecting P. vivax mono-infection
The sensitivity and specificity of the RDT were respectively 81.0% [95% CI 76.0–86.0%] and 99.0% [95% CI 99.0–100%] in the areas of high endemicity, 81.0% [95% CI 58.0–95.0%] and 100% [95% CI 99.0–100%] in the areas of moderate endemicities, and 100% sensitivity and specificity were observed in the areas of low endemicities (Table 4). Pooled across the 12 study sites, the overall sensitivity and specificity of the RDT for detecting a P. vivax mono-infection were 83.0% [95% CI 78.0–87.0%] and 100% [95% CI 99.0–100%] (Table 2). See Tables 2 and 4 for the predictive values and further accuracy measures.
Performance of RDTs for detecting mixed P. falciparum and P. vivax infection
The overall sensitivity and specificity of the RDT for detecting a mixed P. vivax and P. falciparum infection were 88% [95% CI: 80%–93%] and 100% respectively (see Table 2). See Table 2 for the predictive values.
Relationship between age, sex and malaria status
In a multivariable logistic regression containing age, sex, and transmission endemicity, the following variables were associated with increased odds of test positivity by RDTs: every 5 yearly increase in age was associated with 12% lower odds of testing positive (adjusted odds ratio (AOR): 0.88, 95% CI 0.86–0.90), being male (AOR: 1.24, 95% CI 1.10–1.39) (compared to females), and residing in the areas of high malaria endemicity (AOR: 17.33, 95% CI 12.55–24.68) or moderate endemicity (AOR: 7.81, 95% CI 5.62–11.21) (compared to areas of low transmission) (Fig. 3 and see supplemental Table 1 and supplemental Table 2).
Fig. 3.
Probability of RDT confirmed malaria status by age, sex, and transmission setting. Predictions obtained from a multivariable logistic regression that contained age, transmission setting and sex. The model had age as a linear effect which was the most parsimonious model. For comparison of different model fits, see supplemental file 1
Discussion
Overall, the proportion of participants who were RDT positive was approximately 15% in this study. This estimate is similar to previous studies conducted over multiple sites in India that had reported approximately 18% positivity [18]. Malaria positivity proportion was found to be higher among males who had 1.24-fold increased odds of testing positive compared to females; this finding are in line with a previous review that indicated a generally higher disease burden (DALYs and deaths) among males [19] and also concur with the findings from a longitudinal survey among adult population [20].
Of those who tested positive for malaria, nearly three-quarters had P. falciparum mono-infection and approximately one-fifth tested positive for P. vivax mono-infection and 8% harboured mixed infections. However, there was a marked heterogeneity in the dominant causative species across the study sites. For example, P. vivax accounted for 80% (18/22) of total malaria cases in Mangalore-Karnataka and more than one-third across the two Mizoram sites. For the remaining sites, malaria was largely caused by P. falciparum. In particular, Lawngtlai, Mizoram—which is an area of high malaria endemicity, had substantially higher proportion of P. vivax infection compared to the rest of the sites. This is consistent with the known uneven distribution of P. vivax in India with a regional variation in burden and previously reported fluctuations in the dominance of P. falciparum and P. vivax as the leading causative species for malaria in India. Plasmodium falciparum accounted for 63% of all reported malaria in 2017, and this dropped to 46% in 2019, followed by a resurgence to 64% in 2020, and has since consistently remained above 50% [21]. Overall, identification of substantial proportion of P. vivax infections in this survey may indicate an increasing importance of P. vivax malaria in the country that historically had a large burden of P. falciparum malaria (See Fig. 1 in [22]). These shifts and fluctuations highlight ongoing challenges in malaria control and the need to implement site-specific malaria control strategies tailored to the local needs.
The observed malaria positivity proportion in this study while largely corroborated with the underlying disease endemicity, some results were notably distinct. First, West Singhbhum, a site with known high transmission, had nearly half of the patients testing positive which suggests a substantial malaria burden in this location beyond what was previously observed. The high positivity proportion observed could be due to the timing of the survey which coincided with the peak of the malaria season (November to February), right after the monsoon leading to a surge in the cases. Second, Balaghat, currently considered a moderately endemic site, had a malaria positivity of approximately 30% among screened participants. This contrasts with previous findings at the site, which reported monthly slide positivity rates of < 5% between 2008 and 2020 [23]. If true, this finding likely suggests the need for re-consideration of endemicity status of the Balaghat. A potential explanation for this discordance could be the inclusion of higher proportion of the children < 10 years enrolled at this site relative to other study sites (See supplemental Figs. 1–4). However, it remains unclear whether the 30% positivity in this site reflects a true increase in transmission over the years or remains simply an artefact of the sampling strategy; this warrant further epidemiological investigation.
Overall, there were 43 (0.4%) participants in whom RDT was negative but were identified as malaria positive subsequently by microscopic examination. Of these 43, some reported hrp2 gene deletion. Such finding warrants further caution to be exercised and emphasizes the need for continuously monitoring the diagnostic performance of RDTs in India.
In this study, a higher sensitivity of RDTs was observed in the areas of low transmission for both P. falciparum and P. vivax, and the sensitivity progressively declined with increasing endemicity status. The performance of the RDTs was also generally better for P. falciparum than for P. vivax. In addition, the precision of the estimated performance measures was also much higher for P. falciparum than for P. vivax. The poor sensitivity of RDTs in identifying P. vivax and mixed infections is a particular note of concern, as the underlying drug for treatment of P. falciparum and P. vivax malaria remains different. In particular, the sensitivity of the RDT for P. vivax observed in this study is much lower than the 95.5% reported earlier [24]. It is known that SD Bioline RDT have a sensitivity below 90% for P. vivax at parasite density < 100/µL. In this study, explicit quantification of the parasitaemia levels was not undertaken and thus, low parasite density among the participants could not be ruled out as a potential explanation for the lower sensitivity observed. Taken together, these results suggest that a better and more accurate diagnostic test is warranted for P. vivax in the context of India. However, the specificity of RDT was high regardless of the parasite species along with the negative predictive values close to 100% suggesting a high accuracy in ruling out the disease.
There are some key limitations of this study. This study aimed to detect approximately 220 P. falciparum cases per site and this was not achieved across majority of the sites, especially in the areas of low and moderate transmission. The selection of the study sites was also not random and was determined based on operational feasibility in the first stage and guided by mathematical modelling in the second stage. In addition, limited participant level covariates were collected and therefore evaluation of predictors of test positivity was limited to age, sex, and transmission setting. Influence of predictors such as history of malaria in the recent past, travel history and other key factors such as housing conditions and bed-net usage could not be assessed. Information regarding the number of participants who were invited for screening and reasons for exclusion were also not explicitly recorded. Further investigation of the reasons for exclusion (such as declining consent or non-eligible age range) could have provided further insights. Data on hrp2 gene deletion is not explicitly reported here and this will be presented as a subsequent report (manuscript in preparation). Finally, while hrp2 deletion explained some of the discordant results between RDT and microscopy, few additional factors that could have led to the discordance were not investigated. In particular, hrp2 antigen is known to persist in patients for several weeks even after the resolution of infection which could have explained some of the discordance between RDT and microscopy [20]. However, as stated earlier, additional information regarding participant’s recent malaria history wasn’t collected and this remains a limitation. Additional information regarding parasite density could have been potentially useful as previous research indicates a decreasing accuracy of RDT at low parasite density (See Fig. 4 in [25]), and this can be considered as part of future research, particularly for P. vivax.
Conclusions
India has made substantial gains in controlling malaria over the past two decades, attributed to roll-outs of vector and parasite control interventions. However, there are still some major challenges as identified in this study. For example, the finding of a high slide positivity proportion in Balaghat, a site currently considered as moderately endemic, suggests the need for further investigation and future studies at this site. Approximately one-fifth of those who tested positive for malaria had microscopically confirmed P. vivax infections and 7.2% of the microscopically confirmed malaria cases harboured P. falciparum and P. vivax mixed infections. Low sensitivity of the RDTs for identifying P. vivax cases, and a further 43 (0.4%) RDT negatives who were subsequently confirmed as microscopy positive suggests need for more robust diagnostics for detection of P. vivax.
Supplementary Information
Supplementary material 1: Figure 1: Age-distribution by study sites. Figure 2: Age-distribution by study sites, with age presented as 10y bins. Figure 3: Plasmodium falciparum malaria slide positivity proportion by age-bins and site using microscopy. Y-axis depicts age-range, and X-axis presents the percentage of the slides identified as positive using microscopy. Figure 4: Age distribution by rapid diagnostic testresults, across areas of high, low and moderate endemicities. Table 1: Comparison of different functional forms of the covariates included in logistic regression model for malaria positivity by a rapid diagnostic test. Table 2: Coefficients from a logistic regression model for malaria positivity by a rapid diagnostic test.
Acknowledgements
The authors extend their sincere gratitude to the study participants and, for infant participants, their guardians, for their involvement in this study. We also wish to thank the medical officers at various CHCs, PHCs, and the ASHA workers and anganwadi staff for their invaluable efforts in sample collection.
Author contributions
Study Conception: PKB, SSP, PJG; Data curation: PKB, SN, SSP; Formal analysis: PD, SSP, SN, PKB; Funding acquisition: PKB, MD, PJG; Investigation: PD, SSP, SN, PKT, WAM, DPS, AM, VHC, SBM, SPS, PKS, KS, RR, RKB, RC, MN, AD, RSK, AG, MR, JAF, NN, MD, AA, PJG, PKB; Methodology: PD, SSP, SN, PKB; Project administration: SSP, PKB; Resources: PKB, PJG; Software: PD, SSP; Supervision: PKB, PJG; Validation: PD, SSP, SN, PKB; Visualization: PD, SSP, SN, PKB; Writing—original draft: PD, SSP, SN, PKB; Writing—review and editing: All authors were involved in reading and critical revision of the initial draft and approved the final manuscript.
Funding
Bill & Melinda Gates Foundation (INV-004713).
Data availability
The data supporting the findings of this study are available from the ICMR-NIMR, New Delhi. The data can be obtained upon reasonable request submitted to the ICMR-NIMR, New Delhi through the corresponding author: saprapbs@yahoo.co.in.
Declarations
Competing interests
The authors declare no competing interests.
Footnotes
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Prabin Dahal, Sauman Singh-Phulgenda, and Shrikant Nema have contributed equally to this work.
Contributor Information
Philippe J. Guérin, Email: philippe.guerin@ndm.ox.ac.uk
Praveen K. Bharti, Email: saprapbs@yahoo.co.in
References
- 1.Das D, Dahal P, Dhorda M, Citarella BW, Kennon K, Stepniewska K, et al. A systematic literature review of microscopy methods reported in malaria clinical trials. Am J Trop Med Hyg. 2021;104:836–41. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Nema S, Ghanghoria P, Bharti PK. Malaria elimination in India: bridging the gap between control and elimination. Indian Pediatr. 2020;57:613–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Balasubramani K, Bharti PK, Amarthaluri C, Chellappan S, Behera SK, Mohanty AK, et al. Spatiotemporal epidemiology of malaria in India from 2007 to 2022. Am J Trop Med Hyg. 2024;111:26–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Singh MP, Rajvanshi H, Bharti PK, Anvikar AR, Lal AA. Time series analysis of malaria cases to assess the impact of various interventions over the last three decades and forecasting malaria in India towards the 2030 elimination goals. Malar J. 2024;23:50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.WHO. World malaria report 2024. Geneva: World Health Organization; 2024. [Google Scholar]
- 6.Malaria. National Center for Vector Borne Diseases Control (NCVBDC). https://nvbdcp.gov.in/index1.php?lang=1&level=1&sublinkid=5784&lid=3689. Accessed 26 Nov 2024.
- 7.Nema S, Verma AK, Bharti PK. Strengthening diagnosis is key to eliminating malaria in India. Lancet Infect Dis. 2019;19:1277–8. [DOI] [PubMed] [Google Scholar]
- 8.Krishna S, Nema S, Sangle R, Ahmad A, Singh A, Kumar D, et al. Low prevalence of Plasmodium falciparum histidine-rich protein 2 and 3 gene deletions in malaria-endemic states of India. PLoS ONE. 2024;19(12):e0315662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Nema S, Kumari M, Verma K, Krishna S, Ali NA, Verma AK, et al. Plasmodium falciparum histidine-rich protein 2/3 gene deletions and repeat motifs in Chhattisgarh, India (2017–2018). Trans R Soc Trop Med Hyg. 2025;119(7):741–7. [DOI] [PubMed] [Google Scholar]
- 10.Poti KE, Sullivan DJ, Dondorp AM, Woodrow CJ. HRP2: transforming malaria diagnosis, but with caveats. Trends Parasitol. 2020;36:112–26. [DOI] [PubMed] [Google Scholar]
- 11.Kumar N, Pande V, Bhatt RM, Shah NK, Mishra N, Srivastava B, et al. Genetic deletion of HRP2 and HRP3 in Indian Plasmodium falciparum population and false negative malaria rapid diagnostic test. Acta Trop. 2013;125:119–21. [DOI] [PubMed] [Google Scholar]
- 12.Verma AK, Bharti PK, Das A. HRP-2 deletion: a hole in the ship of malaria elimination. Lancet Infect Dis. 2018;18(8):826–7. [DOI] [PubMed] [Google Scholar]
- 13.Dondorp AM, Nosten F, Yi P, Das D, Phyo AP, Tarning J, et al. Artemisinin resistance in Plasmodium falciparum malaria. N Engl J Med. 2009;361:455–67. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Ishengoma DS, Gosling R, Martinez-vega R, Beshir KB, Bailey JA, Chimumbwa J, et al. Urgent action is needed to confront artemisinin partial resistance in African malaria parasites. Nat Med. 2024;30:1807–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Statistical Methods for Rates and Proportions. Wiley Series in Probability and Statistics. https://onlinelibrary.wiley.com/doi/book/10.1002/0471445428.
- 16.Stevenson M, Sergeant E, Heuer C, Nunes T, Heuer C, Marshall J, et al. epiR: tools for the analysis of epidemiological data. https://cran.r-project.org/web/packages/epiR/index.html. Accessed 3 Sept 2025.
- 17.RStudio: Integrated Development Environment for R. Posit Software, PBC, Boston, MA. (2025). Posit team. http://www.posit.co. Accessed 3 Sept 2025.
- 18.Krishna S, Mishra S, Tiwari P, Vishwakarma AK, Khandai S, Shrivastava S, et al. Therapeutic efficacy of artemether-lumefantrine for the treatment of uncomplicated Plasmodium falciparum malaria in four malaria endemic states of India. Malar J. 2021;20:229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Kumar A, Valecha N, Jain T, Dash AP. Burden of malaria in India: retrospective and prospective view. Am J Trop Med Hyg. 2007;77:69–78. [PubMed] [Google Scholar]
- 20.Mohan I, Kodali NK, Chellappan S, Karuppusamy B, Behera SK, Natarajan G, et al. Socio-economic and household determinants of malaria in adults aged 45 and above: analysis of longitudinal ageing survey in India, 2017–2018. Malar J. 2021;20:306. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.National Vector Borne Disease Control Programme (NVBDCP). http://www.nvbdcp.gov.in/. Accessed 3 Sept 2025.
- 22.Nema S, Singh A, Krishna S, Poriya R, Dubey S, Ali NA, et al. Unreported mixed Plasmodium species infection may increase vivax malaria in India: a challenge for malaria elimination. Trans R Soc Trop Med Hyg. 2022;116:600–3. [DOI] [PubMed] [Google Scholar]
- 23.Singh MP, Rajvanshi H, Bharti PK, Jayswar H, Singh S, Mehra RK, et al. Evaluation of the model malaria elimination strategy in Mandla district along with its neighbouring districts: a time series analysis from 2008 to 2020. Malar J. 2023;22:45. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Public Report for Bioline Malaria Ag P.f/P.v (PQDx 0125-012-00) | WHO - Prequalification of Medical Products (IVDs, Medicines, Vaccines and Immunization Devices, Vector Control). https://extranet.who.int/prequal/WHOPR/public-report-bioline-malaria-ag-pfpv-pqdx-0125-012-00. Accessed 3 Sept 2025.
- 25.Singh N, Shukla MM, Shukla MK, Mehra RK, Sharma S, Bharti PK, et al. Field and laboratory comparative evaluation of rapid malaria diagnostic tests versus traditional and molecular techniques in India. Malar J. 2010;9:191. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Supplementary material 1: Figure 1: Age-distribution by study sites. Figure 2: Age-distribution by study sites, with age presented as 10y bins. Figure 3: Plasmodium falciparum malaria slide positivity proportion by age-bins and site using microscopy. Y-axis depicts age-range, and X-axis presents the percentage of the slides identified as positive using microscopy. Figure 4: Age distribution by rapid diagnostic testresults, across areas of high, low and moderate endemicities. Table 1: Comparison of different functional forms of the covariates included in logistic regression model for malaria positivity by a rapid diagnostic test. Table 2: Coefficients from a logistic regression model for malaria positivity by a rapid diagnostic test.
Data Availability Statement
The data supporting the findings of this study are available from the ICMR-NIMR, New Delhi. The data can be obtained upon reasonable request submitted to the ICMR-NIMR, New Delhi through the corresponding author: saprapbs@yahoo.co.in.