Malaria risk, treatment-seeking, and prevention in Northeastern India: clinic-based surveillance of persistent transmission Authors: Phibansuk Lyngdoh, Rajiv Sarkar, Mark L Wilson, Innang Lyngdoh, Anne Kessler, Jane M Carlton, Sandra Albert Published: 2025 ABSTRACT AbstractBackgroundMalaria has historically been a public health concern in Meghalaya with year-round transmission peaking during the monsoon and post-monsoon seasons. This study investigated malaria prevalence among suspected cases presenting at health centres across three districts of Meghalaya, in the northeastern region (NER) of India.MethodsA multi-site, facility-based, cross-sectional study was conducted in five primary health centres (PHC) and one district hospital from three geographically distinct regions of Meghalaya: West Khasi Hills [KH] district, West Jaintia Hills [JH] district, and South Garo Hills [GH] district. Individuals presenting at the selected healthcare facilities with malaria-like symptoms between 2018 and 2021 were enrolled. Malaria cases were detected by light microscopy, RDT, and PCR. Malaria prevalence was estimated as the number of Plasmodium infections divided by the total number of participants providing a blood sample. Logistic regression analysis was conducted to assess the risk factors of malaria.ResultsA total of 1,031 participants provided a blood sample for testing. Forty-five (4.4%) participants had a Plasmodium infection, most (60%) of which were Plasmodium falciparum. Malaria prevalence ranged from 1.8% in KH to 12.3% in GH. Diagnosis of malaria within the past 12 months (OR = 9.2; P < 0.001) and performing outdoor activities before bedtime (OR = 2.0; P = 0.023) or early morning (OR = 3.1; P = 0.010) were significantly associated with Plasmodium infection.ConclusionThe results highlight heterogeneities in disease prevalence within an endemic area, and emphasizes the significance of local risk- factors, such as outdoor activities during peak mosquito hours, in sustaining malaria transmission. Targeted interventions, community education, and better access to prevention and treatment can play an important role in reducing the disease burden.Keywords: Northeast India, Malaria, Prevalence, Risk factors, Plasmodium infection Background Historically, malaria has been a public health concern in Meghalaya, a state in the northeastern region (NER) of India. Significant progress in controlling transmission has been achieved over the past decade, primarily due to targeted interventions, enhanced disease surveillance, and active community participation in use of preventive measures [1]. The introduction of artemisinin-based antimalarial drugs for the treatment of malaria and widespread distribution of long-lasting insecticidal nets (LLINs) may have played a role in lowering malaria incidence, although other factors too may have played a role [1, 2]. For example, indoor residual spraying (IRS) has served as a complementary control method, though its coverage remains limited in certain areas [1, 2]. Previous studies have shown a diverse range of Anopheles species with potential roles in malaria transmission inhabiting the study region [1]. Anopheles jeyporiensis was the most abundant species in both West Khasi Hills and West Jaintia Hills, while Anopheles maculatus was more prevalent in West Khasi Hills. Both species have previously been reported to carry Plasmodium falciparum in the region [1], highlighting their potential importance as local vectors. Notably, the primary forest-associated vectors, Anopheles baimaii and Anopheles minimus, were absent in those samples [1], consistent with broader regional studies indicating their decline in northeast India [3, 4].This shift suggests that species historically considered secondary vectors may now be contributing more significantly to malaria transmission, although their exact roles remain to be fully characterized [5]. Plasmodium transmission in Meghalaya occurs year-round, with peak malaria incidence during the monsoon season (May–July) [6]. Travel history is a major risk factor, and individuals with prior malaria episodes also face a higher risk [7]. Earlier community-based studies from Meghalaya have provided insights into malaria epidemiology, emphasizing the role of behavioural and environmental factors such as vector exposure and human mobility in malaria epidemiology [1, 7]. Despite these community-level findings, research on the clinical epidemiology of malaria in Meghalaya remains limited, with little attention given to patient-specific factors influencing malaria prevalence. Understanding these factors is crucial for refining malaria control strategies, identifying high-risk groups, and tracking disease trends. By analysing suspected malaria cases at the health centre level, valuable insights into disease burden and key risk factors can be obtained, helping to strengthen prevention, control, and the more recent move toward elimination [1]. This facility-based surveillance investigated malaria prevalence among suspected cases presenting at government healthcare facilities in three districts of Meghalaya: West Jaintia Hills (JH), West Khasi Hills (KH), and South Garo Hills (GH). It also assessed the epidemiological patterns, risk factors, and health-seeking behaviour of patients with malaria-like symptoms with the goal of helping policymakers and healthcare providers take data-driven decisions to strengthen India’s malaria elimination efforts. Methods Study design, area, and period Meghalaya is a mountainous, heavily forested state in Northeast India, bordered by the state of Assam in the north and the country of Bangladesh to the south [8]. The population of Meghalaya is estimated to be over 3.5 million, representing predominantly tribal people mostly living in rural, forested settings [9]. The economy is largely agrarian involving mixed farming. The climate is among the wettest in the world, especially during the rainy season (e.g., May–September). With temperatures ranging from 23 to 28 °C and high relative humidity (> 70%), the conditions are ideal for mosquito reproduction and perpetual transmission. The area’s major and minor transmission seasons take place between May–July and between September–November, respectively. While both P. falciparum and Plasmodium vivax are present across the state, P. falciparum is predominant in most regions [2]. A multi-site, facility-based, cross-sectional study was conducted over a three-year period from September 2018 through November 2021 in three districts of Meghalaya. The facilities included: Nonglang Primary Health Centre (PHC) in West Khasi Hills district (KH), Barato PHC and Nartiang PHC in West Jaintia Hills (JH), and Rongara PHC, Siju PHC and Baghmara Civil Hospital in South Garo Hills (GH) (Fig. 1). Selection of the six health centres was based on state control programme Plasmodium infection records with the goal of capturing both high and low prevalence settings in the three districts. Map of Meghalaya, highlighting the health facilities included in the study Based on an anticipated 18.2% prevalence of malaria as reported in an earlier cross-sectional study [10], a relative precision of 20%, a non-response rate of 20% and a design effect of 2, the sample size was estimated at 1036, with a conservative 95% confidence interval. Ethical approval Approval to undertake the study was obtained from the Institutional Review Board at New York University, New York, NY, USA and the University Research Ethics Committee of Martin Luther Christian University, Shillong, Meghalaya, India. Written informed consent was obtained from all adult participants (≥ 18 years of age) and parental consent from participants aged < 18 years. Additionally, assent was obtained from participants aged 7–17 years. Data and sample collection and analysis Socio-demographic and behavioural information of study participants was collected through a structured interview consisting mainly of close-ended questions. The questionnaire was developed by the study team, initially in English and subsequently translated into Khasi, Pnar, and Garo (the local languages). It was pre-tested to ensure accuracy of translation, language clarity, and comprehension before being administered in the field. The questionnaire was administered to individuals aged 1–69 years who presented to the health centres with at least two malaria-like symptoms, which included fever, chills, sweating, headache, muscle aches, nausea, vomiting, or diarrhoea. Individual questionnaires were administered at the clinic to consenting participants or their parent/guardian (in case of children) in their native language (e.g., Khasi, Pnar, Garo), to gather information on age, gender, education, occupation, knowledge of malaria, malaria episodes during previous year, fever episodes during previous 48 h, travel history (past two weeks), and malaria prevention methods utilised. Questions on malaria-related risk behaviours and practices were also asked. Blood sample collection and Plasmodium detection Up to 1 ml of blood was drawn via finger prick, and P. falciparum and/or P. vivax infections were detected at the point of care using both a bivalent rapid diagnostic test (RDT; FalciVax) and an ultra-sensitive RDT (Abbott Alere). Light microscopy evaluation of Giemsa-stained blood smears and Polymerase Chain Reaction (PCR) amplification of concentrated DNA extracted from microvette RBC pellets were also performed on all samples. Species-specific Plasmodium infections were detected using a single-step PCR targeting Pfr364 (for P. falciparum) and Pvr47 (for P. vivax) [11]. Statistical analysis Data were collected electronically and stored in a secure, firewall-protected REDCap (Research Electronic Data Capture) database [12]. Data analysis was conducted using Stata version 17 for Windows (StataCorp, College Station, TX, USA). A complete-case analysis approach was used, whereby participants with missing data on key variables were excluded from the analysis. Summary statistics were presented as mean and standard deviation (SD) or median and interquartile range (IQR) for continuous variables, depending on distribution of data whereas, for categorical variables, they were presented as number and percentage (%). The prevalence of malaria was estimated as the number of Plasmodium infections divided by the total number of participants from whom a sample was obtained. Associations between socio-demographic, environmental, or behavioural risk factors and Plasmodium infections were evaluated using univariate logistic regression analysis, and reported as odds ratio (OR) and 95% confidence intervals (CI). For the analysis combining data from all sites, a multilevel logistic regression analysis with exchangeable correlation matrix was conducted, considering district (JH/KH/GH) as the grouping variable. Results Population sampled and individual characteristics A total of 1031 participants were enrolled across six PHCs surveyed in JH (n = 722), KH (n = 114), and GH (n = 195) from September 2018 through November 2021. Blood samples were obtained and tested from all 1031 participants. The mean (SD) age of participants was 24.1 (15.9) years. Overall, 54.2% of the study participants were female. Of the adult participants (≥ 18 years), 19.8% had no formal education. About half of all participants (51.3%) were engaged in agricultural-related activities. Malaria in the previous 12 months was reported by 8.2% of participants, while 9% reported a history of travel outside the village in the previous 14 days. About one in ten participants (11.6%) reported spending one or more nights in the field during the agricultural season (May- September), with a mean (SD) of 5.8 (4.4) nights. Nearly one-quarter (23.5%) were engaged in ‘outdoor activities’ before 6 am. Nearly three-quarters (70.8%) reported using insecticide treated bed nets (ITNs) while staying in the field. Roughly one-third (34.6%) of all participants were familiar with the typical malarial signs and symptoms, which they reported as fever (94.9%), headache (71.9%), chills (78.3%), and body discomfort (54.3%). District level demographics and occupations The age and gender distributions of participants across the three districts are presented in Table 1, with males comprising between 43.1 and 55.3%. The district-level proportion of adult participants with no formal education ranged from 11.2 to 63.4%. Between 6.9 and 77.5% of the participants were engaged in agricultural-related activities. Among the participants, from 5.0 to 4.5% were housewives. Table 1Socio-demographic and behavioural characteristics of study participants in three districts of Meghalaya state, IndiaCharacteristic or behaviourAll districts(n = 1031)West Khasi hills(n = 114)West Jaintia hills(n = 722)South Garo hills(n = 195)Age (years) 0–4 5–14 15–44 45–59 60 above81 (7.8%)271 (26.4%)537 (52.2%)101 (9.8%)38 (3.7%)(n = 113)5 (4.4%)18 (15.9%)71 (62.8%)12 (10.6%)7 (6.2%)(n = 721)66 (9.1%)198 (27.5%)359 (49.8%)75 (10.4%)23 (3.2%)(n = 194)10 (5.1%)55 (28.3%)107 (55.2%)14 (7.2%)8 (4.1%)Gender Males Females472 (45.8%)558 (54.2%)63 (55.3%)51 (44.7%)325 (45.1%)396 (54.9%)84 (43.1%)111 (56.9%)Education (for participants aged ≥ 18 years) No formal education Primary Middle Secondary/matric (class X) Higher secondary (class XII) Diploma above(n = 641)309 (48.2%)129 (20.1%)70 (10.9%)69 (10.7%)43 (6.7%)21 (3.3%)(n = 80)14 (17.5%)42 (52.5%)13 (16.2%)6 (7.5%)4 (5.0%)1 (1.2%)(n = 445)282 (63.4%)70 (15.7%)29 (6.5%)29 (6.5%)20 4.5%)15 (3.4%)(n = 116)13 (11.2%)17 (14.6%)28 (24.1%)34 (29.3%)19 (16.4%)5 (4.3%)Occupation Agriculture-related Daily wage-labour Housewife Salaried service Student Unemployed Others(n = 641)321 (50.0%)56 (8.7%)128 (20.0%)34 (5.3%)57 (8.9%)27 (4.2%)18 (2.8%)(n = 80)62 (77.5%)6 (7.5%)4 (5.0%)5 (6.3%)1 (1.2%)2 (2.5%)0(n = 445)251 (56.4%)42 (9.4%)84 (18.9%)15 (3.4%)33 (7.4%)6 (1.4%)14 (3.4%)(n = 116)8 (6.9%)8 (6.9%)40 (34.5%)14 (12.1%)23 (19.8%)19 (16.4%)4 (3.5%)Travel history (past 14 days) Yes No(n = 1031)93 (9.0%)938 (91.0%)(n = 114)9 (7.9%)105 (92.1%)(n = 722)76 (10.5%)646 (89.5%)(n = 195)8 (4.1%)187 (95.9%)Stayed in field for ≥ 1 nights Yes No(n = 845)120 (14.2%)725 (85.8%)(n = 96)4 (4.2%)92 (95.8%)(n = 679)112 (16.5%)567 (83.5%)(n = 70)4 (5.7%)66 (94.2%)Take ITN while in field Yes No(n = 120)85 (70.8%)35 (29.2%)(n = 4)3 (75%)1 (25%)(n = 112)79 (70.5%)33 (29.5%)(n = 4)3 (75%)1 (25%)Knowledge of signs/symptoms Yes No(n = 1031)357 (34.6%)674 (65.4%)(n = 114)40 (35.1%)74 (64.9%)(n = 722)181 (25.1%)541 (74.9%)(n = 195)136 (69.7%)59 (30.3%)First choice of treatment Public health sector# Private clinic Pharmacy/Traditional healer/local practitioner Use local herbs/home remedies/medicines at home Nowhere Don’t know(n = 1031)1,002 (97.2%)25 (2.4%)3 (0.3%)1 (0.1%)00(n = 114)104 (91.2%)9 (7.9%)1 (0.9%)000(n = 722)705 (97.6%)14 (1.9%)2 (0.3%)1 (0.1%)00(n = 195)193 (99%)2 (1.0%)0000Second choice of treatment Public health sector Private clinic Pharmacy/Traditional healer/local practitioner Use local herbs/home remedies/medicines at home Don’t know Nowhere(n = 1031)761 (73.8%)263 (25.5%)4 (0.4%)01 (0.1%)2 (0.2%)(n = 114)99 (86.8%)15 (13.2%)0000(n = 722)493 (68.3%)226 (31.3%)1 (0.1%)002 (0.3%)(n = 195)169 (86.7%)22 (11.3%)3 (1.5%)01 (0.5%)0Ill with malaria past 12 mos Yes No(n = 1030)84 (8.2%)946 (91.8%)(n = 114)12 (10.5%)102 (89.5%)(n = 721)17 (2.4%)704 (97.6%)(n = 195)55 (28.2%)140 (71.8%)Place where participant sought malaria treatment, if diagnosed in past 12 months Public health sector Private clinic Pharmacy/Traditional healer/local practitioner Use local herbs/home remedies/medicines at home Other^(n = 84)77 (91.6%)5 (5.9%)01 (1.2%)1 (1.2%)(n = 12)10 (83.3%)2 (16.7%)000(n = 17)17 (100%)0000(n = 55)50 (91%)3 (5.5%)01 (1.8%)1 (1.8%)Location of activities before sleeping Inside the house Outside the house Location changes(n = 1028)982 (95.5%)41 (3.4%)5 (0.5%)(n = 114)109 (95.6%)2 (1.7%)3 (2.6%)(n = 722)692 (95.8%)29 (4.0%)1 (0.1%)(n = 192)181 (94.3%)10 (5.2%)1 (0.5%)Sleeping location Inside the house(n = 1031)1,028 (99.7%)(n = 114)111 (97.4%)(n = 722)722 (100%)(n = 195)195 (100%)Location of early morning activities Inside the house Outside the house Location changes(n = 359)271 (75.5%)84 (23.4%)4 (1.1%)(n = 73)66 (90.4%)5 (6.8%)2 (2.7%)(n = 251)194 (77.3%)55 (21.9%)2 (0.8%)(n = 35)11 (31.4%)24 (68.6%)0*Others: Self- employed/trade#Public health sector includes Govt. local health centre, ASHA^Other includes Missionary dispensary Socio-demographic and behavioural characteristics of study participants in three districts of Meghalaya state, India Age (years) 0–4 5–14 15–44 45–59 60 above 81 (7.8%) 271 (26.4%) 537 (52.2%) 101 (9.8%) 38 (3.7%) (n = 113) 5 (4.4%) 18 (15.9%) 71 (62.8%) 12 (10.6%) 7 (6.2%) (n = 721) 66 (9.1%) 198 (27.5%) 359 (49.8%) 75 (10.4%) 23 (3.2%) (n = 194) 10 (5.1%) 55 (28.3%) 107 (55.2%) 14 (7.2%) 8 (4.1%) Gender Males Females 472 (45.8%) 558 (54.2%) 63 (55.3%) 51 (44.7%) 325 (45.1%) 396 (54.9%) 84 (43.1%) 111 (56.9%) Education (for participants aged ≥ 18 years) No formal education Primary Middle Secondary/matric (class X) Higher secondary (class XII) Diploma above (n = 641) 309 (48.2%) 129 (20.1%) 70 (10.9%) 69 (10.7%) 43 (6.7%) 21 (3.3%) (n = 80) 14 (17.5%) 42 (52.5%) 13 (16.2%) 6 (7.5%) 4 (5.0%) 1 (1.2%) (n = 445) 282 (63.4%) 70 (15.7%) 29 (6.5%) 29 (6.5%) 20 4.5%) 15 (3.4%) (n = 116) 13 (11.2%) 17 (14.6%) 28 (24.1%) 34 (29.3%) 19 (16.4%) 5 (4.3%) Occupation Agriculture-related Daily wage-labour Housewife Salaried service Student Unemployed Others (n = 641) 321 (50.0%) 56 (8.7%) 128 (20.0%) 34 (5.3%) 57 (8.9%) 27 (4.2%) 18 (2.8%) (n = 80) 62 (77.5%) 6 (7.5%) 4 (5.0%) 5 (6.3%) 1 (1.2%) 2 (2.5%) 0 (n = 445) 251 (56.4%) 42 (9.4%) 84 (18.9%) 15 (3.4%) 33 (7.4%) 6 (1.4%) 14 (3.4%) (n = 116) 8 (6.9%) 8 (6.9%) 40 (34.5%) 14 (12.1%) 23 (19.8%) 19 (16.4%) 4 (3.5%) Travel history (past 14 days) Yes No (n = 1031) 93 (9.0%) 938 (91.0%) (n = 114) 9 (7.9%) 105 (92.1%) (n = 722) 76 (10.5%) 646 (89.5%) (n = 195) 8 (4.1%) 187 (95.9%) Stayed in field for ≥ 1 nights Yes No (n = 845) 120 (14.2%) 725 (85.8%) (n = 96) 4 (4.2%) 92 (95.8%) (n = 679) 112 (16.5%) 567 (83.5%) (n = 70) 4 (5.7%) 66 (94.2%) Take ITN while in field Yes No (n = 120) 85 (70.8%) 35 (29.2%) (n = 4) 3 (75%) 1 (25%) (n = 112) 79 (70.5%) 33 (29.5%) (n = 4) 3 (75%) 1 (25%) Knowledge of signs/symptoms Yes No (n = 1031) 357 (34.6%) 674 (65.4%) (n = 114) 40 (35.1%) 74 (64.9%) (n = 722) 181 (25.1%) 541 (74.9%) (n = 195) 136 (69.7%) 59 (30.3%) First choice of treatment Public health sector# Private clinic Pharmacy/Traditional healer/local practitioner Use local herbs/home remedies/medicines at home Nowhere Don’t know (n = 1031) 1,002 (97.2%) 25 (2.4%) 3 (0.3%) 1 (0.1%) 0 0 (n = 114) 104 (91.2%) 9 (7.9%) 1 (0.9%) 0 0 0 (n = 722) 705 (97.6%) 14 (1.9%) 2 (0.3%) 1 (0.1%) 0 0 (n = 195) 193 (99%) 2 (1.0%) 0 0 0 0 Second choice of treatment Public health sector Private clinic Pharmacy/Traditional healer/local practitioner Use local herbs/home remedies/medicines at home Don’t know Nowhere (n = 1031) 761 (73.8%) 263 (25.5%) 4 (0.4%) 0 1 (0.1%) 2 (0.2%) (n = 114) 99 (86.8%) 15 (13.2%) 0 0 0 0 (n = 722) 493 (68.3%) 226 (31.3%) 1 (0.1%) 0 0 2 (0.3%) (n = 195) 169 (86.7%) 22 (11.3%) 3 (1.5%) 0 1 (0.5%) 0 Ill with malaria past 12 mos Yes No (n = 1030) 84 (8.2%) 946 (91.8%) (n = 114) 12 (10.5%) 102 (89.5%) (n = 721) 17 (2.4%) 704 (97.6%) (n = 195) 55 (28.2%) 140 (71.8%) Place where participant sought malaria treatment, if diagnosed in past 12 months Public health sector Private clinic Pharmacy/Traditional healer/local practitioner Use local herbs/home remedies/medicines at home Other^ (n = 84) 77 (91.6%) 5 (5.9%) 0 1 (1.2%) 1 (1.2%) (n = 12) 10 (83.3%) 2 (16.7%) 0 0 0 (n = 17) 17 (100%) 0 0 0 0 (n = 55) 50 (91%) 3 (5.5%) 0 1 (1.8%) 1 (1.8%) Location of activities before sleeping Inside the house Outside the house Location changes (n = 1028) 982 (95.5%) 41 (3.4%) 5 (0.5%) (n = 114) 109 (95.6%) 2 (1.7%) 3 (2.6%) (n = 722) 692 (95.8%) 29 (4.0%) 1 (0.1%) (n = 192) 181 (94.3%) 10 (5.2%) 1 (0.5%) Sleeping location Inside the house (n = 1031) 1,028 (99.7%) (n = 114) 111 (97.4%) (n = 722) 722 (100%) (n = 195) 195 (100%) Location of early morning activities Inside the house Outside the house Location changes (n = 359) 271 (75.5%) 84 (23.4%) 4 (1.1%) (n = 73) 66 (90.4%) 5 (6.8%) 2 (2.7%) (n = 251) 194 (77.3%) 55 (21.9%) 2 (0.8%) (n = 35) 11 (31.4%) 24 (68.6%) 0 *Others: Self- employed/trade #Public health sector includes Govt. local health centre, ASHA ^Other includes Missionary dispensary Reported malaria diagnosis and risk behaviours In GH, 28.2% of participants (55/195) reported being diagnosed with malaria in the past 12 months, while the proportions were 10.5% (12/114) in KH and 2.4% (17/722) in JH (Table 1). A history of travel in the past 14 days was reported by 10.5% of the participants (76/722) in JH, 7.9% (9/114) in KH, and 4.1% (8/195) in GH. Staying overnight in the field for one or more days was reported by 16.5% (112/679) of participants in JH, 4.2% (4/96) in KH, and 5.7% (4/70) in GH. The average (SD) number of days spent overnight in the field was 5.9 (4.5) days in JH, 3.5 (0.7) days in KH, and 4.0 (1.7) days in GH. Outdoor activities before 6 a.m. ‘wake-up’ time were reported by 68.6% (24/35) of participants in GH, 22.7% (55/251) in JH, and 6.8% (5/73) in KH. Knowledge of malaria signs and symptoms was reported by 69.7% (136/195) of participants in GH, 35.1% (40/114) in KH, and 25.1% (181/722) in JH. Health-seeking behaviour Nearly all respondents (97.2%) reported that a public healthcare facility (i.e., primary health care centre or district hospital) was their first choice for seeking malaria treatment. If ineffective, 25.5% stated they would then seek care at a private clinic. This preference was reported regardless of whether the individual had previously experienced malaria. The proportion of respondents diagnosed with malaria who preferred to seek treatment from government healthcare facilities ranged from 83.3 to 100% across KH, GH, and JH. Reported malaria risk reduction Use of methods to prevent mosquito bites among participants was also analysed (Table 2). Almost all individuals (96.3%) reported sleeping under a bed net with virtually all reporting usage either every night (83.9%) or most of the time (4- 6 times a week) (11.4%). However, only 75.6% of the bed nets had been treated with insecticide. About half of respondents (55.8%) used clothing that covered their arms and legs to avoid mosquito bites, but 86.1% did not apply insecticide cream to prevent bites. Evening activities were reported by 95.5% of participants, which was usually held within the house. Virtually everyone ate their dinner (99.7%) and slept inside their houses. One quarter (24.5%) of activities recorded in the early morning occurred outside the house. Table 2Malaria prevention methods reportedly used by study participants from each of three districts of MeghalayaBehaviour or characteristicAll districtsWest Khasi hillsWest Jaintia hillsSouth Garo hillsSlept under bed net at night Yes No(n = 1031)993 (96.3%)38 (3.7%)(n = 114)110 (96.5%)4 (3.5%)(n = 722)694 (96.1%)28 (3.9%)(n = 195)189 (96.9)6 (3.1%)If yes, bed net treated with insecticide Yes No(n = 993)596 (60.1%)190 (19.1%)(n = 110)103 (93.6%)6 (5.4%)(n = 694)351 (50.6%)151 (21.8%)(n = 189)142 (75.1%)33 (17.5%)If yes, frequency of bed net use Every night Most of the times Sometimes Rarely Only in rainy season(n = 993)834 (83.9%)113 (11.4%)28 (2.8%)10 (1%)8 (0.8%)(n = 110)95 (86.3%)1 (0.9%)7 (6.4%)07 (6.4%)(n = 694)579 (83.4%)100 (14.4%)13 (1.9%)1 (0.1%)1 (0.1%)(n = 189)160 (84.6%)12 (6.3%)8 (4.2%)9 (4.7%)0Used bed net preceding night Yes No(n = 990)910 (91.9%)80 (8.2%)(n = 109)103 (94.5%)6 (5.5%)(n = 692)632 (91.3%)60 (8.7%)(n = 189)175 (92.6%)14 (7.4%)Cover arms and legs to prevent bites Always Sometimes Rarely Never(n = 1031)243 (23.6%)332 (32.2%)142 (13.8%)314 (30.5%)(n = 114)24 (21.1%)31 (27.2%)10 (8.8%)49 (43.0%)(n = 722)199 (27.6%)211 (29.2%)87 (12.0%)225 (31.2%)(n = 195)20 (10.3%)90 (46.1%)45 (23.1%)40 (20.5%)Used insecticide cream Always Sometimes Rarely Never(n = 1031)30 (2.2%)53 (5.1%)60 (5.8%)888 (86.1%)(n = 114)06 (5.3%)2 (1.7%)106 (93%)(n = 722)28 (3.9%)31 (4.3%)30 (4.2%)633 (87.7%)(n = 195)2 (1.0%)16 (8.2%)28 (14.4%)149 (76.4%)Other products used to prevent bites Coil Vaporizer Mats Burn other materials(n = 795)653 (63.3%)101 (9.8%)37 (3.6%)4 (0.4%)(n = 45)33 (28.9%)6 (5.3%)4 (3.5%)2 (1.7%)(n = 602)489 (67.7%)80 (11.1%)33 (4.6%)0(n = 148)131 (67.2%)15 (7.7%)02 (1.0%) Malaria prevention methods reportedly used by study participants from each of three districts of Meghalaya Slept under bed net at night Yes No (n = 1031) 993 (96.3%) 38 (3.7%) (n = 114) 110 (96.5%) 4 (3.5%) (n = 722) 694 (96.1%) 28 (3.9%) (n = 195) 189 (96.9) 6 (3.1%) If yes, bed net treated with insecticide Yes No (n = 993) 596 (60.1%) 190 (19.1%) (n = 110) 103 (93.6%) 6 (5.4%) (n = 694) 351 (50.6%) 151 (21.8%) (n = 189) 142 (75.1%) 33 (17.5%) If yes, frequency of bed net use Every night Most of the times Sometimes Rarely Only in rainy season (n = 993) 834 (83.9%) 113 (11.4%) 28 (2.8%) 10 (1%) 8 (0.8%) (n = 110) 95 (86.3%) 1 (0.9%) 7 (6.4%) 0 7 (6.4%) (n = 694) 579 (83.4%) 100 (14.4%) 13 (1.9%) 1 (0.1%) 1 (0.1%) (n = 189) 160 (84.6%) 12 (6.3%) 8 (4.2%) 9 (4.7%) 0 Used bed net preceding night Yes No (n = 990) 910 (91.9%) 80 (8.2%) (n = 109) 103 (94.5%) 6 (5.5%) (n = 692) 632 (91.3%) 60 (8.7%) (n = 189) 175 (92.6%) 14 (7.4%) Cover arms and legs to prevent bites Always Sometimes Rarely Never (n = 1031) 243 (23.6%) 332 (32.2%) 142 (13.8%) 314 (30.5%) (n = 114) 24 (21.1%) 31 (27.2%) 10 (8.8%) 49 (43.0%) (n = 722) 199 (27.6%) 211 (29.2%) 87 (12.0%) 225 (31.2%) (n = 195) 20 (10.3%) 90 (46.1%) 45 (23.1%) 40 (20.5%) Used insecticide cream Always Sometimes Rarely Never (n = 1031) 30 (2.2%) 53 (5.1%) 60 (5.8%) 888 (86.1%) (n = 114) 0 6 (5.3%) 2 (1.7%) 106 (93%) (n = 722) 28 (3.9%) 31 (4.3%) 30 (4.2%) 633 (87.7%) (n = 195) 2 (1.0%) 16 (8.2%) 28 (14.4%) 149 (76.4%) Other products used to prevent bites Coil Vaporizer Mats Burn other materials (n = 795) 653 (63.3%) 101 (9.8%) 37 (3.6%) 4 (0.4%) (n = 45) 33 (28.9%) 6 (5.3%) 4 (3.5%) 2 (1.7%) (n = 602) 489 (67.7%) 80 (11.1%) 33 (4.6%) 0 (n = 148) 131 (67.2%) 15 (7.7%) 0 2 (1.0%) Plasmodium infection prevalence Plasmodium infections were detected by RDT and/or PCR and/or light microscopy in 4.4% (45/1,031) of individuals. Plasmodium infection was higher in GH (24/195, 12.3%) than in JH (19/722, 2.6%) and KH (2/114, 1.8%) (P < 0.001). Of the positives, 10 (22%) were positive by RDT only, 2 (4%) by PCR only, 14 (31%) by both RDT and PCR, 1 (2%) by both RDT and light microscopy, and 18 (40%) by all three methods (e.g., RDT, PCR, and light microscopy); no infections were determined solely by microscopy (Fig. 2). Of the 43 RDT-positive samples, one was positive by both conventional and ultrasensitive RDT. A slightly higher proportion of PCR-positive infections was reported from GH (18/34; 40%) than from JH (15/34; 33.3%) and KH (1/34; 2%) (P < 0.001). Plasmodium falciparum was the only species detected in 60% (27/45) of positive samples, while 38% (17/45) were P. vivax, and 2% (1/45) were P. vivax/P. falciparum co-infections (Fig. 3). Most of the P. vivax infections (17/19; 89.4%) were reported from JH. A case of malaria was defined as a positive Plasmodium infection based on RDT and/or PCR results. Malaria positivity by RDT, PCR and Microscopy Species distribution of Malaria parasites Risk factors for Plasmodium infection The factors associated with the risk of Plasmodium infection were assessed collectively and separately for the three districts (Table 3). Statistically significant increased risk was associated with reported history of malaria in the previous 12 months (OR = 9.2, P < 0.001), activity outside the house before bedtime (OR = 2.0, P = 0.023), and activity outside the house before 6 am (OR = 3.1, P = 0.010). None of the other factors assessed (e.g., age, gender, education, occupation, IRS spraying, and LLIN use) were statistically significant. Table 3Risk factors for testing positive for Plasmodium spp. infection overall, and by study areaCharacteristic or BehaviourReference categoryAll sites(n = 1031)South Garo hills(n = 114)West Jaintia hills(n = 722)West Khasi hills(n = 195)OR (95% CI)p- valueOR (95% CI)p- valueOR (95% CI)p- valueOR (95% CI)p- valueAgeBelow 18 years1.7 (0.9–3.2)0.0660.8 (0.3–2.0)0.7323.5 (1.3–9.5)0.0102.3 (0.1–38.9)0.547GenderMale0.9 (0.5–1.6)0.6200.5 (0.1–1.2)0.1471.6 (0.6–4.2)0.2600.8 (0.04–13.2)0.880EducationNo formal education0.5 (0.2–1.0)0.0380.4 (0.1–3.3)0.4120.8 (0.3–2.2)0.806––OccupationAgricultural related/Daily wage/labor0.5 (0.3–1.1)0.0782.6 (0.7–9.0)0.1180.6 (0.2–1.7)0.3940.6(0.03–10.6)0.760Travelled for 14 days or moreYes1.2 (0.5–3.3)0.6172.5 (0.4–13.1)0.2801.6 (0.4–5.6)0.453##Stay a night in the fieldYes2.0 (0.9–4.8)0.08019.0 (1.7–203.2)0.0151.9 (0.6–5.7)0.199##Malaria reported in past 12 monthsYes9.2 (4.9–17.7) < 0.0016.7 (2.6–16.9) < 0.0012.3 (0.2–18.9)0.412##Activity outside house before bedtimeYes2.0 (1.1–3.7)0.0234.8 (1.3–18.1)0.018####Wake up time before 6amBefore 6 am0.4 (0.1–1.8)0.2432.1 (0.8–5.8)0.1120.4 (0.1–1.9)0.288##Before 6am activity outside houseOutside the house3.1 (1.3–7.4)0.0103.3 (0.3–31.7)0.2952.6 (0.5–12.1)0.213##Do not use bed net at nightNo1.6 (0.5–5.8)0.3961.4 (0.1–12.9)0.7433.0 (0.6–13.9)0.148##Use bed net sometimesNo0.3 (0.1–1.3)0.1210.4 (0.1–2.2)0.3540.3 (0.04–2.3)0.256##Cover arms and legsNo1.4 (0.7–2.7)0.372##0.4 (0.1–1.2)0.129##Use coilNo0.4 (0.2–0.7)0.004##0.3 (0.1–0.8)0.021##Use matsNo0.5 (0.1–4.5)0.618##1.1 (0.1–9.0)0.884##Use insecticide creamNo1.1 (0.5–2.6)0.7381.1 (0.4–2.9)0.8620.3 (0.01–2.9)0.360##Household sprayedNo1.1 (0.6–2.1)0.7291.7 (0.6–4.4)0.2310.7 (0.2–1.8)0.527### Odds ratio (95% CI) and p-value could not be estimated because of separation due to small number of participants with Plasmodium infection Risk factors for testing positive for Plasmodium spp. infection overall, and by study area # Odds ratio (95% CI) and p-value could not be estimated because of separation due to small number of participants with Plasmodium infection In GH, the odds of Plasmodium infection were significantly higher for individuals with a history of malaria in the past 12 months (OR = 6.7, P < 0.001). Staying a night in the field (OR = 19.0, P = 0.015) and engaging in activities outside before bedtime (OR = 4.8, P = 0.018) were also significantly associated with increased odds of malaria in this district. In JH (OR = 3.5, P = 0.010) and KH (OR = 2.3, P = 0.547), younger people (< 18 years) had significantly higher odds of Plasmodium infection this was not statistically significant in KH. Discussion This facility-based surveillance study investigated the prevalence and risk factors of Plasmodium infection among suspected malaria cases over a three-year period to gain deeper insights into the epidemiology of malaria in three districts of Meghalaya. The overall malaria prevalence among participants who were visiting health centres was 4.4%, which is much lower than the 18.2% prevalence reported from an active fever surveillance study conducted in West Garo Hills district of Meghalaya in 2007 [6, 10], but higher than another community-based cross-sectional survey conducted in 2018–2019 in KH and JH districts that reported a Plasmodium infection prevalence of 1.2% [7]. Collectively, these findings reinforce the declining trend in malaria cases in Meghalaya over the past decade, as observed through an analysis of the National Vector Borne Disease Control Programme data [1]. This decline can likely be attributed to the shift from sulfadoxine-pyrimethamine treatment to artemether-lumefantrine artemisinin combination therapy in 2013, and the state-wide distribution of LLINs in 2016, 2019 and 2020 [2, 13, 14]. Additionally, advancements in diagnostic and treatment strategies, particularly the involvement of Accredited Social Health Activists (ASHAs) in testing and treating individuals with Plasmodium infections at the community level, may have further contributed to decreased malaria prevalence [15]. This study also found that when participants were ill, most preferred to seek treatment from the healthcare facilities in the public sector. Over the past decade, health-seeking behaviour in Meghalaya has evolved significantly. Traditionally, communities relied on tribal medicines for personal health concerns [16]. However, with increasing government healthcare initiatives and growing public health awareness, there has been a shift toward utilising modern healthcare system facilities. Programmes under the National Health Mission (NHM), such as the Tribal Health Initiative, have played a significant role in improving access to healthcare services [17]. These changes reflect broader trends in health-seeking behaviour among India's tribal populations, where access to healthcare and growing awareness have driven this shift [17]. This suggests that the study likely captures an accurate estimate of malaria’s true point prevalence, as individuals actively presenting at health facilities allow for the detection of a broader spectrum of disease cases, ranging from mild to severe. More than half of infected individuals had sub-microscopic infections, detectable only by RDT and PCR-based methods. These results are similar to the earlier community-based, active surveillance in KH and JH [7]. As sub-microscopic carriers contribute to Plasmodium transmission [18], regular use of more sensitive and specific diagnostic tests, such as PCR, will improve diagnostic accuracy in low-prevalence settings [18], thereby enhancing case detection and further supporting the malaria elimination efforts. Plasmodium falciparum has been documented as the most common species in Meghalaya [2, 7]. In this study too, it represented 60% of all infections, with P. vivax present in 38% of detected cases Here, P. vivax infections were observed in one health centre in JH. Differences in patterns of Plasmodium species across regions may be important as understanding species distributions helps in identifying high-risk regions where tailored control measures may be implemented, such as antimalarial treatments for chronic P. vivax infection [19]. A history of malaria within the past 12 months was associated with increased risk of current Plasmodium infection, consistent with findings from a previous community-based study conducted in Meghalaya [7]. Recurrent episodes of malaria have previously been reported in a small proportion of individuals from endemic regions [20, 21]. Malaria recurrence can occur for several reasons, such as the return of parasites (recrudescence), dormant liver stages (relapses), or new infections [20]. Taken together, it highlights the need for ongoing monitoring and targeted interventions like Reactive Case Detection (RCD) and foci investigation for those with recent malaria case history [3, 22], along with improved access to treatment and strengthened preventive measures to reduce disease burden. In this study, people engaging in outdoor activities during the evening and early morning hours had a higher risk of Plasmodium infection. This finding aligns with another study in Odisha, India that showed a correlation between outdoor exposure during these times and heightened malaria risk [4]. Traditionally, primary Anopheles vector species are considered to bite late at night [5], which meant evening activities were not expected to pose significant exposure risk. However, recent research indicates that some Anopheles species are changing their host-seeking behaviour – biting earlier in the evening, later in the morning, and outdoors [23]. This shift increases Plasmodium transmission risk during evening or morning activities such as fetching water, preparing food, or socializing [24, 25]. These periods are particularly concerning because many individuals remain unprotected by ITNs or indoor spraying when competent vectors are active [23, 26]. Collectively, these findings suggest that while the use of protective measures such as ITNs is effective indoors, individuals who are active outdoors during the evening and early morning remain at risk, underscoring the need for targeted interventions such as the implementation of larval source management [27] and risk communication to address this gap in malaria control. Almost all study participants reported using ITNs, consistent with findings from previous community-based studies in Meghalaya which also highlighted the widespread use of LLINs throughout the region [13] and that they were preferred given their perceived effectiveness and durability [28]. The current finding is noteworthy, as widespread use of LLINs is a cornerstone of India’s malaria prevention efforts [29]. LLINs not only serve as an effective physical barrier against mosquito bites but also help to reduce Plasmodium transmission [30], contributing to the broader goal of public health programs focused on controlling and ultimately eliminating malaria. The findings from this study have implications for both clinicians and public health policymakers. Clinicians must remain vigilant and proactive in malaria-endemic areas to ensure timely diagnosis and appropriate treatment, particularly in low-endemic areas where prevalence is lower but transmission still occurs [31]. For policymakers, these insights highlight the need for region-specific malaria control strategies, targeted resource allocation, and enhanced surveillance efforts to detect persistent low-level transmission, prevent outbreaks, and maintain progress in malaria elimination. This study has several limitations. As a facility-based study, it may have missed patients who sought care at other facilities or those who self-medicated. Additionally, the true prevalence of P. vivax malaria may have been underestimated as asymptomatic infections with low parasite densities, who are less likely to seek treatment [32], would have gone undetected. Since participants were enrolled from public health facilities, the findings may not fully represent the health-seeking behaviours of all individuals residing in the study area, although participants of qualitative study that interviewed members from the same communities reported similar health-seeking behaviour [28]. Overall, continued surveillance, targeted interventions, and improved access to prevention and treatment are crucial for reducing malaria in the region. These findings highlight the importance of accounting for local factors, particularly community-specific behaviours, in the design of effective malaria control measures. Outdoor activities, especially during the evening and early morning when mosquitoes are most active, contribute significantly to malaria risk. By taking these local behaviours into account, targeted prevention strategies such as outdoor mosquito control and improved community education can be developed to protect more effectively those at greatest risk. Acknowledgements We are deeply grateful to the study participants for their willingness and involvement. Our sincere thanks go to the Department of Health and Family Welfare, Government of Meghalaya, and the Meghalaya State Programme Management Unit, National Vector-Borne Disease Control Programme, for granting permission to carry out this study and for their continuous support and collaborative efforts. We acknowledge the support of Anna Maria van Eijk for her help with the study design and Steven A. Sullivan for data management. We would also like to extend our appreciation to the IIPH-Shillong staff, especially Christine Manar, Manroi Challam, Enrichson Suting, Watson Siangshai, Jurysha Nongdhar, Charisma Khongwir, and Peter Marbaniang, for their diligent work in data collection and sample processing. A special thanks to Bandapkupar Mawkhlieng and Innang Sangniang for their valuable contributions in the laboratory analysis. Author contributions AK, JC and SA conceived and designed study and undertook the data collection; PL, RS and MLW cleaned the data and performed the data analyses; IL completed the DNA extractions and PCR amplifications for the study; PL, RS and MLW wrote the manuscript with input from all authors. All authors read and approved the final manuscript. Funding Research reported in this publication was supported by the US National Institute of Allergy and Infectious Diseases of the National Institutes of Health under Award Number U19AI089676. The content is solely the responsibility of the authors and does not necessarily represent the official views of the US National Institutes of Health. This manuscript is the result of funding in whole or in part by the National Institutes of Health (NIH). It is subject to the NIH Public Access Policy. Through acceptance of this federal funding, NIH has been given a right to make this manuscript publicly available in PubMed Central upon the Official Date of Publication, as defined by NIH. Data availability Data generated in this study are available through the open-access online resource for population-based epidemiological studies ClinEpiDB (https://clinepidb.org) at https://clinepidb.org/ce/app/record/dataset/DS_4670e06911. Declarations Ethics approval and consent to participate Approval to undertake the study was obtained from the Institutional Review Board at New York University, New York, NY, USA and the University Research Ethics Committee of Martin Luther Christian University, Shillong, Meghalaya, India. Written informed consent was obtained from all adult participants (≥ 18 years of age). Assent was obtained for participants aged 7–17 years, in addition to the parental consent. Consent for publication Not applicable. 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. References Associated Data Data Availability Statement Data generated in this study are available through the open-access online resource for population-based epidemiological studies ClinEpiDB (https://clinepidb.org) at https://clinepidb.org/ce/app/record/dataset/DS_4670e06911.