High Burden of Febrile Sub-microscopic Plasmodium Mixed Infections in Central India: A Cross-Sectional Study

Abstract

Introduction

Malaria continues as a public health threat through symptomatic/febrile cases, asymptomatic and low-density infections of Plasmodium falciparum, P. vivax, and their mixed infections. Mixed infections have not been studied much regarding their burden, clinical manifestations, and implications, and therefore, this study was conducted.

Methods

Febrile patients were recruited from four patient-care settings from June to November 2020 through the collection of dried blood spots (DBS) and their paired microscopy and/or rapid diagnostic test (RDT) data. Polymerase chain reaction (PCR)-based molecular diagnosis of both parasite species was performed from genomic DNA isolated from the DBS. Clinico-demographic details were recorded from patients from one of the sites, wherein patients with mixed infections were telephonically followed for subsequent clinical development.

Results

Out of the 1030 samples collected and analyzed, 27% (280) were infected with P. falciparum and/or P. vivax: 188 (18%) mono-P. falciparum, 6 (0.5%) mono-P. vivax and 86 (8%) mixed. None of the infections were detected by microscopy and/or RDT, meaning that all 27% were febrile sub-microscopic infections with 8% burden of mixed infections. The quality of microscopic slides was found to be unsatisfactory when a sub-sample of slides was cross-examined by level 1-competent microscopists. None of the nine mixed-infection patients from Gandhi Medical College and Hospital (GMCH) reported recurrences or any clinical development during the 12-month follow-up. No clinically/statistically significant difference was observed between mono- and mixed infections.

Conclusions

A high 27% febrile sub-microscopic Plasmodium infections with 8% mixed infections represent a significant challenge for malaria elimination, considering the quality of microscopy and the fact that Madhya Pradesh is classified under category 1 in the National Strategic Plan for malaria elimination 2023–2027.

Supplementary Information

The online version contains supplementary material available at 10.1007/s40121-025-01297-x.

Key Summary Points

Why carry out this study?

Malaria due to mixed infections of Plasmodium falciparum and P. vivax might have a different burden and clinical spectrum.

However, such infections have been understudied and underreported.

Therefore, this study was conducted to estimate the burden of malaria due to P. falciparum and P. vivax mixed infections and to study their clinical profile and recurrence potential.

What was learned from the study?

A high 27% febrile sub-microscopic Plasmodium infections with 8% P. falciparum and P. vivax mixed infections represent a significant challenge for malaria elimination.

This burden of febrile Plasmodium infections was not detected by the routine malaria surveillance, even in a tertiary care setting in central India.

This research raises an important clinical and public health issue, as such infections pose an undetected threat to malaria elimination efforts.

The technicians responsible for microscopy-based peripheral blood smear examination require rigorous, continuous training to ensure high diagnostic accuracy and quality.

Introduction

In 2023, malaria affected an estimated 263 million people worldwide, an increase of 11 million estimated cases over 2022 [1]. India is a major contributor (~ 2 million cases) to the global malaria burden after certain African countries and Pakistan. Although India aims for malaria elimination by 2030, there was a 29% increase in the number of reported cases since 2022 and it remains the leading contributor of malaria in the World Health Organization’s (WHOs) South-East Asia Region [1].

Although nine species of Plasmodium are known to cause human malaria, Pf and Pv have been the dominant species in India. Despite a nearly equal contribution of both species in India, Pf and Pv have distinct geographical niches. Whereas Pf is dominant in Odisha, Chhattisgarh, and eastern part of Madhya Pradesh (MP), Pv has its dominance in West Bengal, Gujarat and certain northern/southern parts of India.

In addition to their mono-infections (single species in a host), Pf and Pv are also found as mixed infections (> 1 species in a host). Such Pf-Pv mixed infections may stymie malaria elimination as they can impede diagnosis and thus, treatment. It is important to note that different species exhibit distinct characteristics, give rise to distinct clinical features, and require different treatments. Accurate diagnosis and effective treatment are two critical components of malaria control and elimination, which are hampered by mixed infections. For example, subsequent detection of Pv after Pf treatment could be due to low-density Pv infection that went undetected by light microscopy (MS) during initial diagnosis [2, 3]. For mixed Plasmodium species infections, MS has a very low sensitivity (~ 50%) and the molecular-based tools are the best to detect such infections [4–6].

Despite having importance, Pf–Pv mixed infections have neither been reported exclusively by the National Center for Vector Borne Diseases Control (NCVBDC) India, the national malaria control program, nor studied comprehensively in India. The NCVBDC started reporting mixed infections from 2019 and mixed infections involving Pf were considered/reported as Pf mono-infections.

The burden of mixed infections could be appreciated from individual research studies, which are mostly cross-sectional and regional [7]. The prevalence of Plasmodium mixed infections by molecular-based tools ranges between 2% and 45% in India, which is a significant burden [5, 6, 8–11].

MP has infections from both species [12] and this makes MP a good geographical site to study their burden in India. This study therefore targeted MP to estimate PCR-based prevalence of Pf–Pv mixed infections in febrile patients and to study their clinical profile/complications. This study also aimed to determine whether the identified Plasmodium infections meet the detection threshold of light MS and/or RDTs thus identifying the issues related to the quality of MS.

Methods

Ethical Approval

Approvals from the institutional human ethics committees of Indian Council of Medical Research-National Institute of Malaria Research, New Delhi (ICMR-NIMR; PHB/NIMR/EC/2020/62) and Gandhi Medical College and Hospital, Bhopal, MP (GMCH; 854/MC/IEC/2020) were obtained to conduct the study. Voluntary informed written consent was obtained from all the participants before enrolling them in the study. This study was conducted in accordance with institutional and national ethical standards and the 2013 revision of the Declaration of Helsinki.

Data Collection and Initial Screening

The primary data collection site was the medicine outpatient department (mOPD) of the GMCH, Bhopal, MP, India (a tertiary care hospital). The total number of fever/malaria cases from 2017–2019 reported from GMCH mOPD was analyzed to obtain the month-wise trend of fever/malaria. Based on this trend, a 12-month study was conducted that included cross-sectional (6-month hospital-based study during malaria transmission season from June to November 2020) and longitudinal (12-month telephonic follow-up) components.

With single finger-prick blood from all consenting febrile patients attending the GMCH mOPD, a thick/thin peripheral blood smear (PBS) was made, a rapid diagnostic test (RDT) was performed using bivalent Pf/Pv RDT and three DBS were made on sterile grade 1 qualitative Whatman^® filter paper (GE Healthcare, UK), for the point-of-care (immediate) and molecular screening of Plasmodium infection. Patients were advised to seek a clinical opinion irrespective of the RDT result. PBS were made and examined by the District Malaria Office (DMO) staff as a part of their routine malaria surveillance, whereas RDT and molecular diagnosis were performed by the study investigators.

Patients’ socio-demographic characteristics were recorded at the time of sample collection along with consultants’ prescriptions. The outcomes of the prescribed lab investigations of the recruited patients were recorded retrospectively from the pathology department. All patients from GMCH with mixed PfPv infection, irrespective of the mode of diagnosis, were telephonically followed (monthly) to identify self-reported complications and recurrences, if any, until 12 months from the date of diagnosis. During follow-up, patients were asked about fever and any other complaints between the follow-up day and the most recent prior contact.

In addition, anonymized PBS and DBS from febrile patients were contributed to the analyses by Jai Prakash District Hospital (JP), Jawaharlal Nehru Hospital (JNH), and DMO, Bhopal, MP, India, after obtaining permission. Ten percent of all PBS were re-examined for the quality of MS by the WHO-certified L1 microscopists of ICMR-NIMR, New Delhi, India.

Molecular Diagnosis

Molecular detection of Pf/Pv was done through PCRs with half of the samples tested by nested PCR (nPCR) and the other half through quantitative real-time PCR (qPCR) to determine the parasite load. Genomic DNA extraction from DBS was done using QIAamp^® DNA mini kit (Qiagen, DE) as per the manufacturer’s instructions. Nested PCR was carried out using primers targeting the Plasmodium 18S rRNA gene [13, 14]. To screen Pf and Pv, separate reactions were carried out for each sample using species-specific primers. qPCR was done using a hydrolysis probe-based method with previously published primers/probes based on the 18S rRNA gene for Pf [15], Plasmodium genus, and Pv [16]. To determine the parasite load of the test samples for classifying them as sub-microscopic, internal controls were prepared by limiting dilution and a standard curve was generated to identify the respective cycle threshold (ct) for each dilution. Briefly, DNA from Pf/Pv-infected patients with known parasite density was serially diluted with nuclease-free water to make internal controls ranging from 10 to 0.1 parasites/μl (ppm) of whole blood. Test samples were interpreted as negative if they had no ct values. Ct value corresponding to 10 ppm was considered as the threshold for classifying sub-microscopic infections, considering the lower bound of the limit of detection (LoD) for microscopy (10–100 parasites/μl) [17, 17]. For DNA control, human RNaseP gene was amplified for each sample using primers and probes as described elsewhere [16]. PCR conditions and sequences of primers and probes are mentioned in Tables S1 and S2 in Supplementary Material.

Statistical Analysis

Descriptive analysis was done by calculating ratios and proportions (n/N; %) for qualitative data and mean and standard deviation for quantitative data. Mean values were statistically compared using an independent-samples t test followed by estimation of P value for statistical significance, where a P value < 0.05 was considered significant.

Results

Low Prevalence of Malaria During 2017–2019

Record-based analysis (2017–2019) from GMCH mOPD showed that ~ 4100 febrile patients were screened for malaria yearly by MS, with the biggest surge between June and November across the period (Fig. 1). The prevalence of malaria ranged from 0.1% (6/4398; 2019) to 0.5% (25/4482; 2017). Pf and Pv were the only species found by MS with a predominance of Pv (Fig. 1) as reflected by the Pv-to-Pf ratio in 2017 (25:2; 11.5), 2018 (7:1; 7), and 2019 (5:1; 5) with no mixed infections. The majority of malaria cases (77%; 30/39) were reported between June and November.

Fig. 1.

Malaria trend of patients that attended medicine OPD, GMCH, Bhopal, MP during 2017–2019. The figure displays the number of Pf and Pv-infected patients screened for malaria by light microscopy during 2017–2019. Here, the number of fever patients (blue) is plotted on the primary axis while the number of patients infected with Pf (red) and Pv (green) is depicted on the secondary axis. OPD outpatient department, GMCH Gandhi Medical College and Hospital, MP Madhya Pradesh, Pf Plasmodium falciparum, Pv Plasmodium vivax

High Prevalence of P. falciparum and P. vivax Mixed Infections (2020)

Seventy-four fever patients (with PBS, RDT, and DBS) were recruited from GMCH mOPD between June and November 2020 and 956 PBS and DBS were obtained from additional sites (JNH, JP, and DMO) as shown in Table 1.

Table 1.

Number of patients screened for malaria during June to November 2020 from four different sites

	GMCH	JNH	JP	DMO
June 2020	4	26	23	15
July 2020	40	197	76	49
August 2020	10	43	95	44
September 2020	13	42	129	48
October 2020	7	5	0	46
November 2020	0	0	50	68
Total (1030)	74	313	373	270

The table displays the number of patients screened for malaria and number of blood samples (dried blood spots) collected in each month of the data collection tenure

GMCH Gandhi Medical College and Hospital, JNH Jawaharlal Nehru Hospital, JP Jai Prakash District Hospital, DMO District Malaria Office

Only two samples (02/1030; 0.19%) were positive for Plasmodium spp. (one each mono-Pf and mono-Pv) by MS (both from DMO). These two samples were re-tested for Plasmodium genus by qPCR, and were found to be negative. All samples from GMCH were RDT-negative.

PCR found 280 (280/1030; 27.18%) samples positive for Plasmodium spp. with 188 as mono-Pf (188/1030; 18.25%), six (0.58%) as mono-Pv and 86 (8.34%) as Pf/Pv mixed infections (Tables 2 and 3).

Table 2.

PCR outcome of the blood samples collected from four different sites

Data collection sites	No. of samples screened by PCR	Mono-Pf n (%)	Mono-Pv n (%)	Mixed PfPv n (%)	Total malaria-positive n (%)
GMCH	74	14 (19)	6 (8)	9 (12)	29 (39)
JNH	313	75 (24)	0	33 (11)	108 (35)
JP	373	37 (10)	0	40 (11)	77 (21)
DMO	270	62 (23)	0	4 (1)	66 (24)
Total	1030	188 (18)	6 (0.5)	86 (8)	280 (27)

The table displays the number of samples screened either by nested or qPCR and number of samples (along with prevalence %) detected positive for mono-Pf, mono-Pv and Pf-Pv mixed infections.

GMCH Gandhi Medical College and Hospital, JNH Jawaharlal Nehru Hospital, JP Jai Prakash District Hospital, DMO District Malaria Office, Pf Plasmodium falciparum, Pv Plasmodium vivax, PCR polymerase chain reaction

Table 3.

Distribution of blood samples tested either by nested or qPCR

	qPCR				Nested PCR
	No. of samples screened	Mono-Pf	Mono-Pv	Mixed PfPv	No. of samples screened	Mono-Pf	Mono-Pv	Mixed PfPv
GMCH	0	0	0	0	74	14	6	9
JNH	266	40	0	33	47	35	0	0
JP	186	20	0	24	187	17	0	16
DMO	59	20	0	1	211	42	0	3
Total	511	80	0	58	519	108	6	28

The table displays the number of blood samples tested positive for mono-Pf, mono-Pv and Pf-Pv mixed infections either by qPCR or nested PCR from all four data collection sites.

GMCH Gandhi Medical College and Hospital, JNH Jawaharlal Nehru Hospital, JP Jai Prakash District Hospital, DMO District Malaria Office, Pf Plasmodium falciparum, Pv Plasmodium vivax, qPCR quantitative polymerase chain reaction

The Pf/Pv ratio reported by MS (GMCH; 2017–2019) was reversed by PCR (2020) from all four sites (Fig. 2A). During 2017–2019, Pv accounted for > 80% of all cases, with no Pf/Pv mixed-infection. Pv contribution to total cases from all four sites was reduced to 2% in 2020. The primary contributors to the total malaria cases at GMCH were mono-Pf infections (48%; 14/29) followed by Pf/Pv mixed infections (31%; 9/29). Furthermore, the reported overall prevalence of malaria (2017–2019) was ≤ 0.5% (by MS), far lower than that reported in 2020 by PCR (27%; Fig. 2B).

Fig. 2.

Proportion (A) and prevalence (B) of mono-Pf, mono-Pv and Pf–Pv mixed infections in blood samples examined between 2017 and 2020. The data from 2017 to 2019 represent light microscopy outcome (of samples screened at GMCH), whereas the data from 2020 marked with a ‘*’ is a PCR outcome of the screened blood samples (at all four data collection sites). The x-axis of A represents the proportion of positive cases, whereas that of B represents the prevalence of positive cases (as %). The year of data collection has been displayed along with the total number of positive cases and the total number of blood samples screened for malaria on the y-axis of A and B, respectively. Pf Plasmodium falciparum, Pv Plasmodium vivax, GMCH Gandhi Medical College and Hospital, PCR polymerase chain reaction

Molecular-based assays (qPCR and nPCR) yielded equal sensitivity for Plasmodium spp. prevalence (27%): qPCR (138/511) and nPCR (142/519) (Table 3). However, the species-specific prevalence for Pf/Pv mixed infections was high in qPCR (11%; 58/511) compared to nPCR (5%; 28/519). In contrast, nPCR revealed a higher prevalence for mono-Pf (21%; 108/519) than reported by qPCR (16%; 80/511). Furthermore, nPCR yielded a very low prevalence 1.1% (6/519) of Pv, whereas qPCR detected none (Table 3).

Based on the ct values of the internal controls, it appears that all the infections were sub-microscopic, since their ct values corresponded to parasite density of < 10 parasites/µl.

Clinical and Epidemiological Profile of the Recruited Patients (n = 74) from GMCH

The mean age of Plasmodium-infected patients was 40 (n = 29), whereas it was 33 (n = 45) in uninfected patients (P = 0.02). Further, the difference in mean fever duration and mean hemoglobin level (not advised for all patients) between these two groups was not statistically significant. Only 21% (6/29) of febrile Plasmodium-infected patients were recommended a malaria diagnosis. None of the patients were provided antimalarials based on suspicion. However, the clinicians prescribed antibiotics to 28% (8/29) of Plasmodium-infected patients without any known cause of fever (Table 4). The clinicians did not suspect malaria in 79% (23/29) of fever patients (positive with either mono-Pf and Pv or Pf/Pv mixed-infection) who went untreated and did not return back to the OPD during the data collection period. Somewhat similar findings were observed for mono-infections (Pf or Pv) and Pf/Pv mixed infections. The difference in mean age and duration of fever between patients with mono-infections and patients with Pf/Pv mixed infections was likewise, not significant. Apart from the parameters listed in Table 4, other existing complaints (loss of appetite, chest discomfort, anemia, vomiting, etc.) and comorbidities (diabetes, hypothyroid, hypertension, etc.) were also investigated, but no noteworthy findings were observed. Complete blood count, urine examination, chest X-ray, renal, and liver function tests, and other advised investigations were also within normal limits.

Table 4.

Clinical profile of the recruited patients (n = 74) from medicine OPD, GMCH

	Malaria-negative (n = 45)	Malaria-positive (n = 29)	Pf (n = 14)	Pv (n = 6)	Mixed PfPv (n = 9)
Mean age (years)	32.79	40.17	41.45		37.33
Standard deviation	13.83	13.13	14.00		10.99
Two-tailed P value	0.02**		0.44
Mean duration of fever	9	4	3		4
Standard deviation	20.07	6.57	7		6
Two-tailed P value	0.15		0.71
Mean hemoglobin*	12.4 (n = 2)	12.36 (n = 5)	12.5 (n = 3)	11.70* (n = 1)	12.50* (n = 1)
Standard deviation	3.11	0.83	–	–	–
Two-tailed P value	0.97		–	–	–
Prescribed antibiotics (%)	22.22 (10/45)	27.59 (8/29)	14.29 (2/14)	50.00 (3/6)	33.33 (3/9)
COVID-19 tested (%)	35.56 (16/45)	37.93 (11/29)	35.71 (5/14)	33.33 (2/6)	44.44 (4/9)
Malaria suspected by the clinicians (%)	8.89 (4/45)	20.69 (6/29)	21.43 (3/14)	16.67 (1/6)	22.22 (2/9)
Male/female ratio (M/F)	0.88 (21/24)	0.61 (11/18)	0.56 (5/9)	0.50 (2/4)	0.80 (4/5)

The table displays different clinical parameters of malaria-negative (n = 45) and malaria-positive patients (n = 29) with mono-Pf, mono-Pv and Pf-Pv mixed. Details for all the above parameters were available for all the recruited patients (n = 74), with the exception of hemoglobin, which was not advised to all patients; thus, the ‘n’ given below the ‘mean hemoglobin’ reflects the total number of patients whose hemoglobin was tested. Further, for mono-Pv and Pf–Pv mixed the hemoglobin reading was available for only one patient, hence, it has been mentioned with ‘*’ because it does not represent the mean. The significant P value has been marked with ‘**’ wherever applicable

Pf Plasmodium falciparum, Pv Plasmodium vivax, OPD outpatient department, GMCH Gandhi Medical College and Hospital

All Pf/Pv mixed-infection cases (n = 9) detected from GMCH were telephonically followed (monthly) till 12-months from the date of diagnosis. All patients were reachable throughout the follow-up period but no subsequent incidences of fever or other complaints were reported.

Discussion

India has committed to eliminating malaria by 2030 and has developed the National Strategic Plan (NSP) for the task [18]. MP is the second largest Indian state by area and falls under category 1 (elimination phase) in the NSP with all its constituting districts having annual parasite incidence (API) < 1 per 1000 population. MP has shown drastic improvement in the progress towards malaria elimination as evidenced by its shift from category 3 in 2015 (API > 1 per 1000 population) to category 1 in 2022 [18]. However, research studies involving microscopy and PCRs continued to show higher than reported burden in multiple states across India, including MP [10, 19, 20] with malaria prevalence as high as 32% [20].

Moreover, MP has both Pf and Pv infections, but their mixed infections were not reported by NCVBDC until 2019. Considering the biological, clinical, and epidemiological challenges associated with Pf/Pv mixed infections, it becomes important to study their burden by molecular-based tests. This study was therefore conducted in central MP to determine the clinical epidemiology of malaria due to Pf and Pv and their mixed infections using PCRs and to ascertain whether these infections are microscopic or sub-microscopic.

The study reported a high burden of malaria (27%) by PCR in 2020 from four different sites, where the burden reported by MS was only 0.1% (02/1030). Interestingly, qPCR labeled these two samples as negative.

Because GMCH was designated as a COVID-dedicated hospital in 2020, patients with general febrile illnesses were directed to district hospitals instead. Moreover, the government-imposed lockdown further restricted patient movement and access to healthcare facilities. In addition, all fever patients who were supposed to report to the mOPD were diverted to a specially created ‘fever OPD’ in 2020, where all unauthorized entrances were restricted. As a result, unexpectedly, the flow of fever patients was adversely affected and only 74 fever patients were recruited from GMCH mOPD between June and November 2020 (as against the expected number of 2642—the average number of fever cases from June–November (2017–2019).

The burden of malaria by PCR in the primary data collection site (GMCH) was 29%, with 12% alone contributed by Pf/Pv mixed infections. In contrast, the burden reported by MS was zero in 2020. Record-based analysis of malaria burden in last 3 years (2017–2019), revealed that the burden was between 0.1% and 0.5% (with no mixed infections) by MS. The discordance between the MS and PCR continues with the contribution and proportion of Plasmodium species to the overall burden. Based on the record-based data (2017–2019), it appears that there might be cases of Pf/Pv mixed infections that were misdiagnosed as mono-Pv infections by MS as the diagnostic accuracy of MS is compromised in Pf–Pv mixed infections [5].

Molecular diagnosis of Plasmodium species by conventional nested (nPCR) and quantitative probe-based (qPCR) shows that nPCR detected mixed PfPv infections in 28 of 519 samples (5%), whereas qPCR identified 58 of 511 samples as positive (11%), suggesting higher sensitivity of qPCR in detecting mixed infections that were reported as mono-infections by nPCR. Two different PCR techniques were used for molecular diagnosis of Plasmodium infections on different set of samples and we believe that a direct comparison between nPCR and qPCR results may be challenging due to their fundamentally different primer binding sites and detection principles even using the same diagnostic target, 18S rRNA in this case.

This was verified by performing qPCR on a subset of the samples previously classified as mono-infections by nPCR. Among 13 mono-Pf infections detected by nPCR, qPCR confirmed nine as mono-Pf and identified four as mixed PfPv infections. Similarly, out of six mono-Pv infections by nPCR, qPCR confirmed three as mono-Pv, detected one as mixed PfPv, and reported two as negative.

This additional analysis clarifies that the higher prevalence of mixed infections observed by qPCR most likely arises from its improved detection sensitivity. It is noteworthy that, based on Cq values, all analyzed samples had parasite densities below 10 parasites/µl, indicating sub-microscopic infections. Such samples fall within a “grey zone” of detection, where PCR-based methods may exhibit stochastic variability due to limited template availability, explaining the occasional non-reproducibility and the two qPCR-negative samples. Similar results were obtained by authors in other ongoing epidemiological studies in different regions of India (Mizoram, West Bengal, Assam), where both methods were applied in parallel (manuscripts in preparation).

In addition, several other technical parameters, such as initial blood volume used for DNA extraction, elution volume, input template amount, reaction mix composition, primer/probe design and efficiency, and instrument sensitivity may collectively contribute to minor differences in detection outcomes, particularly in low-parasitemia samples.

Because the discordance between nPCR and qPCR results on the same set of mono-Pv infected samples (n = 6) was reported in two samples, the prevalence of mono-Pv infections by nPCR might be false in about one-third of cases. However, due to a very small sample size of 6 samples for such analysis, and very low overall reported burden of mono-Pv infections in the study (6/1030), this discordance does not appear to affect the overall interpretation of the findings.

In order to verify the results of MS for the current study, 10% of the smears were analyzed by expert microscopists at the WHO-approved laboratory of the ICMR-NIMR but the quality of the smears was poor for them to draw any conclusions. This might explain the two samples found positive by MS turned out to be negative by qPCR. It is important here to recall that MS is a skill-based method and it requires quality assurance (with continuous training) starting from preparation of smear to slide analysis [21–24].

MS is considered to be highly reliable since it can not only provide specific diagnoses (species identification) but can also quantify the parasite load. With an experienced microscopist, it can have a LoD of up to 10 parasites/µl of blood but under typical field conditions, the LoD is ~ 100 parasites/µl [17, 25]. Therefore, a reliable microscopy diagnosis necessitates a skilled microscopist.

Since all detected Plasmodium infections in febrile patients had negative MS/RDT results, all remained untreated and might have served as reservoirs and perhaps contributed to malaria transmission. However, for the current study qPCR results revealed that the parasite load of the tested samples was below the LoD of MS, despite the fact that all patients had fever (febrile sub-microscopic infections or fSMIs).

Fever in such fSMIs may/may not be due to Plasmodium parasites. The plausibility of Pf/Pv causing fever are extremely low as the pyrogenic threshold of Pf has been reported to be 1900–10,700 parasites/µl and that of Pv is even lower [26]. However, variation on this threshold has also been reported depending upon the age, endemicity, and seasonality [27, 28] but the reported variation has always been above the threshold of MS.

Further, the parasite density estimated from the PBS might not reflect the entire parasite biomass present in the body [29–31]. Noteworthy here is that there might be other human-infecting non-Pf/Pv species in the patients causing fever that were not observed by MS (due to low-quality PBS) and not screened by RDT/PCR. In addition, since the data collection was at the peak of COVID-19, mild COVID-19 infections that are missed by diagnostic methods used at the site could also lead to fever in studied patients.

Despite the argument about the cause of fever in such patients with demonstrated Pf/Pv infections, these infections are a potential threat to malaria transmission and the guidelines recommend treating such infections irrespective of symptoms [32]. The patients from the current study could not be treated for malaria because there was no point-of-care (POC) PCR-based diagnosis available at the recruitment sites. It is highly recommended in elimination settings that such POC diagnostics should be developed and deployed, particularly at tertiary care settings. Such modalities, like Truenat^®, are available but they have some technical concerns that need to be resolved [33].

The overall prevalence of malaria in MP was between 0.03% and 0.07% from 2020–2023 based on the total PBS examined and RDTs done (NCVBDC; Table S3 in Supplementary Material). When examining the species contribution, Pf varied in its contribution, accounting for 58% (3914/6760), 51% (1635/3181), 62% (2376/3826) and 31% (1169/3794) of all malaria cases reported from MP in 2020, 2021, 2022, and 2023, respectively (Table S3 in Supplementary Material). In the year 2023, this figure got reversed and Pv (68%; 2568/3794) dominated Pf. The mixed infections of Pf and Pv in these years were between 1% and 2% (Table S3 in Supplementary Material).

The contributory figure of NCVBDC also contrasts with the data derived from the GMCH alone for the years 2017–2019. There are very few studies that come up when we look at data from individual studies for MP. According to a study conducted in Balaghat district (MP) in 2013–2014, 32% of febrile patients had Plasmodium infection, with 88% Pf, 11% Pv, and 1% Pf/Pv mixed-infection [20]. In the same study, 29% of non-febrile patients also had Plasmodium infection, with similar contributions from Pf (88%), Pv (10%), and Pf/Pv mixed (2%). The species contribution shown in this study also differs from the GMCH MS data (2017–2019), and the prevalence of malaria was much higher than reported by NCVBDC. Another study conducted from July–December, 2014 in 8 Indian states (including MP) found 16% Pf/Pv mixed infections, which were initially diagnosed as mono-Pf by MS [19]. Amongst them, 20% (47/239) were contributed by MP alone. From these findings, it is suggested that the burden reported by NCVBDC by MS is only the tip of the iceberg in terms of pathogen reservoir.

The study also aimed to uncover the clinical consequences of Plasmodium mono- and mixed infections, if any. However, except age group, all findings were statistically insignificant. Though information on the age group of individuals more prone to malaria is scarce, studies have found that the malaria burden remains higher in younger age groups [34–36].

A few studies evaluated clinical consequences in Plasmodium mono and mixed infections and found conflicting outcomes. A retrospective analysis done in an Indian hospital, revealed patients with Pf/Pv mixed infections had less severe complications than mono-Pf infections [37]. Another multicenter analysis found that Pf/Pv mixed infections were more likely to cause severe anemia and multiple organ failure than mono-Pf infections [38]. These findings were consistent with three more studies conducted on an Indian population, which found that individuals with mixed infections are more likely to acquire severe malaria than patients with mono-infections [39–41]. The majority of the research, however, were conducted retrospectively on previously existing data and elaborative research should be undertaken on a large number of patients (from different settings) with direct observation of clinical symptoms.

In the current study follow-up, no subsequent complaints and fever were reported. Though the follow-up period was at the peak of COVID-19 pandemic when people were scared and did not want to step out due to mild complaints. Therefore, when contacted by phone, the majority of them stated, “I have no further complaints or fever”. Since none of the Pf/Pv mixed-infection patients reported of fever during follow-up, rescreening for malaria was not carried out, and as a result, no recurrence was reported. The follow-up could have been improved if direct visits to patients were made instead of telephonic calls, but because of the pandemic peak, it was not possible.

MP was among the only three states that shifted directly from category 3 to category 1 between 2015 and 2022. This is a commendable achievement towards regional elimination of malaria provided the robustness of surveillance and quality of microscopy are ensured. Since the study revealed a high burden (27%) of infections with Pf and Pv in febrile patients, a high index of suspicion should be exercised towards malaria surveillance, diagnosis, and transmission despite all the reported infections being sub-microscopic. This is even more critical considering the quality of smears being examined in a sub-sample.

Limitations

Although the study reported a high burden of fSMI in tertiary healthcare settings, the findings need to be interpreted with certain limitations. Since this was a period-based study done in tertiary care settings, a priori sample size calculation was not done. Based on the febrile patients’ average annual footfall in the immediate previous 3 years, we expected that > 2500 patients would be recruited within the study period. However, as mentioned above, the pandemic restricted this number to only 74 and remedial actions to increase the sample size were immediately made by including three other sites. The final sample size thus achieved was 1030 for estimation of prevalence. Post facto sample size estimation revealed a sample size of 900 assuming 10% prevalence of mixed species Plasmodium infection in MP, India (based on published prevalence range of 8–13%), 20% relative error and 80% power, thus circumventing the sample size-based limitations for estimation of prevalence. However, despite best possible efforts, for clinical-based outcomes, the sample size remained limited to only 74 patients and therefore clinical findings need further validation using larger studies in the region. Further, healthcare center-based studies are inherently limited by a lack of external validity (representativeness) and some selection bias as simple random sampling is not possible in such settings. Recurrence-based outcomes are also limited by very small sample size (nine patients) and a lack of in-person follow-up. As discussed above, these limitations were beyond the control of the investigators due to the ongoing peak of the pandemic.

Conclusions

This study reported a high (27%) prevalence of sub-microscopic Plasmodium infections in febrile patients reporting to four health care sites in Bhopal, Madhya Pradesh in 2020. The reported prevalence for Pf was 18%, and Pv was 0.5%. The study also highlights a much higher prevalence of Pf–Pv mixed infections (8%) than that reported by NCVBDC and WHO. Although there was no reported recurrence of any related symptoms in nine Pf–Pv mixed-infection patients during telephonic follow-up for 12 months, post-diagnosis, this needs to be further validated in large-sized studies with sufficient statistical power and internal validity.

Supplementary Information

Below is the link to the electronic supplementary material.

Author Contributions

Dr. Nimita Deora: Literature search, hospital-based survey, data collection, processing, acquisition, analysis and interpretation, drafting the manuscript, figures, and editing of the manuscript critically. Dr. Veena Pande: Supervision, analysis and interpretation, review and editing of the manuscript critically. Dr. Dinesh K. Pal: Supervision, analysis and interpretation, review, and editing of the manuscript critically. Dr. Himanshu Sharma: Supervision, analysis, and interpretation, review, and editing of the manuscript critically. Dr. Simmi Dube: Supervision, analysis and interpretation, review, and editing of the manuscript critically. Dr. Abhinav Sinha: Conceptualization, supervision, analysis and interpretation, review and editing of the manuscript critically.

Funding

The Department of Science and Technology, Government of India, is acknowledged for providing 3-year financial support to ND through the Women Scientist Scheme-A grant (SR/WOS-A/LS-210/2018).  The ICMR-National Institute of Malaria Research, New Delhi, India, is acknowledged for providing intramural funds for research and the applicable rapid service fee for publishing the manuscript.

Data Availability

The data and materials are available from the corresponding author upon reasonable request.

Declarations

Conflict of Interest

The authors Nimita Deora, Veena Pande, Dinesh K. Pal, Himanshu Sharma, Simmi Dube, and Abhinav Sinha have nothing to disclose.

Ethical Approval

Approvals from institutional human ethics committees of ICMR-National Institute of Malaria Research, New Delhi (ICMR-NIMR; PHB/NIMR/EC/2020/62) and Gandhi Medical College and Hospital, Bhopal, MP (GMCH; 854/MC/IEC/2020) were obtained to conduct the study. Voluntary informed written consent was obtained from all the participants before enrolling them in the study. This study was conducted in accordance with institutional and national ethical standards and the 2013 revision of the Declaration of Helsinki.