Introduction

Malaria remains a major global public health challenge, with an estimated 282 million cases and 610,000 deaths globally in 2024 [1]. Indonesia is the second-largest contributor to malaria cases in the WHO Western Pacific region, accounting for 25% of all regional cases [1]. Indonesia recorded 543,965 cases in 2024, with Papua contributes more than 93% of national malaria cases despite comprising only a small proportion of the national population [2,3]. This figure indicates a 30% increase in malaria cases in Indonesia compared to 2023, despite the launch of a malaria elimination program in 2009 with a target of elimination by 2030 [3,4]. To achieve the district and city target of malaria elimination, several activities are being carried out including improving the quality of malaria diagnostics, the use of artemisinin-based combination therapy (ACT), malaria surveillance, and vector control [5]. However, the percentage of positive malaria patients who received appropriate treatment may not have reached the target because the budget was unavailable from the beginning of the year, resulting in a stockout of antimalarial drugs [6].

Despite substantial progress in some regions, Papua continues to experience persistent transmission [7]. Control efforts are complicated by the presence of multiple Plasmodium species; i.e., P. falciparum, P. vivax, P. malariae, and P. ovale [8]. Because species determination in routine surveillance often relies on microscopy and/or RDTs, misclassification can occur and rare-species reports should be interpreted cautiously [9,10]. Importantly, the predominance of P. vivax in many Papuan settings poses additional operational challenges due to relapse risk and the need for effective radical cure and adherence to primaquine regimens [8].

Indonesia’s malaria treatment guidelines recommend dihydroartemisinin–piperaquine (DHP) for three days plus primaquine for uncomplicated malaria, with species-specific dosing regimens [11]. Uncomplicated P. falciparum is treated with DHP and a single, low-dose (0.25 mg/kg on day 1) primaquine, while P. vivax and P. ovale require a 14-day primaquine regimen for radical cure. Relapsing P. vivax infections require higher primaquine dosing (0.5 mg/kg/day for 14 days) after G6PD testing. Severe malaria is treated with intravenous artesunate until the patient can take oral medication. The use of primaquine is avoided in pregnancy and infancy.

Sorong City in Southwest Papua (previously West Papua) province, remains a high-incidence setting. The annual parasite incidence (API) in the province is 7.56 in 2021 and increased to 12.68 in 2022 and 12.47 in 2023 [12–14]. There was an increase of P. vivax infection in this area, replacing the previously high burden of P. falciparum. Falciparum malaria contributed to 48.56% of total cases in Sorong City in 2021, then declined to 41.65% in 2022 and 36.81% in 2023. In contrast, P. vivax infection increased from 39.58% in 2021 to 44.91% in 2022 and 58.16% in 2023 [15]. Patients often experience mild symptoms and preferentially seek care at community health centers (CHCs) [16]. However, limited evidence exists on the factors influencing health-facility attendance (e.g., primary care vs hospital) or predictors of hospitalization using routine surveillance data, particularly in settings where cases may be detected through both passive case detection (PCD) and CHC-led active case detection (ACD). Understanding pathways to care, clinical management, and the current malaria burden, including associated risk factors in these settings is essential for informing malaria control and elimination strategies. This study therefore aimed to analyze national surveillance data from Sorong City to: (1) assess malaria prevalence and severity across healthcare facilities, (2) describe diagnostic and treatment practices, and (3) identify predictors of hospital attendance and hospitalization among malaria patients.

Methods

Results

Demographic characteristics of patients

A total of 3,953 laboratory-confirmed malaria cases were identified in SISMAL during the study period, including one follow-up case. Most cases were reported by CHCs (n = 2,438; 61.7%), followed by hospitals (n = 1,011; 25.6%) and clinics (n = 504; 12.7%). Because CHCs report cases detected through both PCD (facility presentation) and ACD (outreach screening), CHC notifications reflect both detection pathways.

Among the 3,953 patients, most were aged ≥15 years old (n = 2,607; 65.9%), males (n = 2,349; 59.4%), of Papuan ethnicity (n = 2,142; 54.2%), and students (n = 1,446; 36.6%). The majority of patients lived in highly endemic areas (n = 3,084; 78%). Uncomplicated and severe malaria accounted for 99.9% (n = 3,951) and 0.1% (n = 2), respectively. There were statistically significant differences in the demographic characteristics and disease severity of malaria patients attending clinics, CHCs, and hospitals (p = 0.001 and 0.003) (Table 1).

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Demographic profile and disease severity of 3,953 malaria cases recorded in SISMAL across five clinics, 10 community health centers (CHCs), and seven hospitals in Sorong City from 1 January to 31 December 2024.

https://doi.org/10.1371/journal.pone.0350434.t001

Diagnosis and treatment of patients

Microscopy was the most commonly used diagnostic method across all healthcare facilities (n = 2,772; 70.1%, p = 0.001). Parasite density was recorded for 799 of 3,953 cases (20.2%), with the highest completeness among clinics. Among those with available parasite density data, CHC patients predominantly had low-to-moderate parasitemia, while high parasitemia was more common among patients attending clinics and hospitals (p = 0.001) (Table 2). Given the substantial missingness, parasite-density comparisons should be interpreted cautiously. Furthermore, since parasite load was only available in 3.4% (83/2,438) of CHC patients, a formal statistical comparison between ACD and PCD subgroups was deemed inappropriate.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Diagnosis and treatment of 3,953 malaria cases recorded in SISMAL across five clinics, 10 community health centers (CHCs), and seven hospitals in Sorong City from 1 January to 31 December 2024.

https://doi.org/10.1371/journal.pone.0350434.t002

There was a statistically significant difference in the distribution of Plasmodium species across healthcare facilities (p = 0.001). Regarding species distribution, most cases were P. vivax (n = 2,588; 65.5%), followed by P. falciparum (n = 1,222; 30.9%) which accounted for approximately half of clinic cases (263/504; 52.2%). Two patients were infected with P. ovale, and both were treated with DHP monotherapy. A single case was identified by microscopy as “probable knowlesi” in a 37-year-old Papuan male (parasite density: 3,636 parasites/µL). Another case was reported as “P. vivax, probable knowlesi” infection in an 8-year-old Papuan student. However, this remains an unconfirmed morphological diagnosis. Both cases were treated as outpatients with a 3-day course of DHP plus 14 days of primaquine.

The combination of DHP and primaquine was the most common treatment regimen, administered to almost 90% of patients with P. falciparum, P. vivax, and mixed infections. Primaquine monotherapy was given to 5.4% (66/1,222) of P. falciparum, 6.4% (166/2,588) of P. vivax, and 7.9% (11/140) of mixed infection cases. In contrast, 5.6% (144/2,588) of P. vivax patients did not receive primaquine (data not shown). Overall, 79.1% (3,128/3,953) of cases received standard treatment according to national malaria guidelines. These estimates reflect the appropriateness of treatments based on surveillance data documented in SISMAL.

Predictors of hospital attendance

Several demographic, occupational, and epidemiological factors were significantly associated with healthcare attendance (Table 3). There was a total of 3,208 PCD cases reported by CHCs, clinics, and hospitals. Predictors for hospital attendance include older age (aOR: 1.02; 95% CI: 1.00–1.02; p = 0.001), non-Papuan ethnicity (aOR: 1.77; 95%CI: 1.51–2.07; p = 0.001), being housewife (aOR: 1.57; 95%CI: 1.02–2.42; p = 0.041) or student (aOR: 1.56; 95%CI: 1.05–2.31; p = 0.029), and living in moderate- (aOR: 2.01; 95%CI: 1.69–2.40; p = 0.001) or low- (aOR: 4.57; 95%CI: 2.01–10.39; p = 0.001) endemicity areas.

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Predictors of hospital attendance among 3,208 passively detected malaria cases presenting to healthcare facilities in Sorong City from 1 January to 31 December 2024.

https://doi.org/10.1371/journal.pone.0350434.t003

Occupation also influenced hospital attendance. Office workers (aOR 5.2; 95% CI 3.21–8.44; p = 0.001) and unemployed individuals (aOR 3.3; 95% CI 2.15–5.06; p = 0.001) were more likely to attend hospitals. Endemicity levels of residence were also associated with facility choice: individuals living in moderate (aOR 2.84; 95% CI 2.38–3.4; p = 0.001) or low (aOR 6.08; 95% CI 3.54–10.44; p = 0.001) endemicity areas were more likely to seek care at hospitals. Those who were not pregnant were also more likely to visit hospitals (aOR 8.14; 95% CI 1.07–61.72; p = 0.043).

Predictors of hospitalization

Among patients who presented to hospitals, age < 15 years was the only significant predictor of hospitalization (aOR 1.9; 95% CI 1.18–3.06; p = 0.009) (Table 4). No other factors were significantly associated with inpatient admission.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Predictors of hospitalization among 1,011 patients presented to hospitals in Sorong City from 1 January to 31 December 2024.

https://doi.org/10.1371/journal.pone.0350434.t004

Discussion

Malaria remains a major cause of morbidity and mortality, particularly among vulnerable groups such as young children, pregnant women, and populations living in disadvantaged settings. Papua continues to experience persistent transmission, reflecting systemic challenges in malaria control [23]. In Sorong City, the API in 2024 reached 13.23 per 1,000 population at risk, far above Indonesia’s elimination target of <1 per 1,000, underscoring the severity of the malaria burden in this area [3].

This study found that most malaria cases occurred among individuals aged ≥15 years, with males slightly predominating, consistent with findings from Papua and Papua New Guinea, where both genders experience similar exposure patterns in high-transmission zones [24,25]. Students and unemployed individuals comprised a substantial proportion of recorded cases; however, this study did not assess occupation-specific malaria risk, and these patterns may reflect the underlying population structure and detection pathways. Although prior studies suggest that outdoor work, geographic access, and socioeconomic factors can influence malaria risk and healthcare use [26–29], these variables (e.g., travel time/distance, education, household income) were not reported in SISMAL and therefore cannot be evaluated as drivers in this dataset. Nevertheless, the high case burden recorded at CHCs supports strengthening community-based services and targeted engagement with groups frequently represented in notifications (e.g., students and those unemployed), alongside locally appropriate outreach to reduce barriers to timely diagnosis and treatment.

Microscopy remained the primary diagnostic tool in Sorong City so that the quality of malaria laboratory services is essential for establishing a diagnosis and depends heavily on the competence and performance of laboratory personnel at every level of health care facilities. The quality of examinations is monitored through cross-checking mechanisms at the district/city, provincial, and national levels and the competency of laboratory personnels is maintained through workshops and external assessment [5]. However, variations in diagnostic capacity across facilities and the absence of routinely documented external quality assurance in SISMAL may affect species attribution. In addition, “classification errors” in this context may reflect electronic reporting and/or data-entry issues rather than laboratory microscopy performance [30]. The dominance of P. vivax aligns with regional trends [31,32] and is programmatically important because effective control requires not only treatment of blood-stage infection but also high-quality radical cure and adherence support to prevent relapse. Although most patients received DHP plus primaquine as recommended [33], treatment gaps such as primaquine or DHP monotherapy indicate non-standard regimens recorded in surveillance data; the underlying causes (e.g., incomplete documentation, provider practice variation, or drug availability constraints) could not be distinguished in this study and should be interpreted cautiously [34]. Strengthening routine guideline training, supervision, and linkage between case data and commodity/stock monitoring could help reduce non-standard regimens.

Misdiagnosis risk remains a major challenge in regions where multiple Plasmodium species circulate. Light microscopy can fail to differentiate species because of morphological overlap [35], and misclassification has been widely reported [36,37]. While one case in this dataset was recorded as “probable knowlesi” and another case was recorded as “P. vivax, probable knowlesi” infection via microscopy, we emphasize that these were not molecularly confirmed. Given the morphological similarities between P. knowlesi and other Plasmodium species, this finding should be interpreted cautiously [38]. Also, given the very small number observed, we do not draw inferences about zoonotic transmission patterns in Sorong City from these data. Where zoonotic malaria is suspected, expanding diagnostic capacity using appropriate RDTs and/or molecular tools may improve ‌‌diagnostic certainty [38].

It has been reported that in 2024, the Southwest Papua province achieved 82.34% of malaria standard treatment [3]. Non-standard treatment records such as primaquine monotherapy or missing primaquine documentation warrant careful attention. Primaquine is not indicated as monotherapy for acute blood-stage malaria. For P. falciparum, it is administered as a single-dose gametocytocidal agent alongside ACT to reduce transmission. For P. vivax, it remains the primary agent for radical cure [39]. However, its programmatic success is often limited by poor adherence to the standard 14-day regimen and the risk of hemolysis in G6PD-deficient individuals, necessitating supervised administration or simplified high-dose regimens where feasible. Because this study used routine surveillance data where age and body weight were not verified, dose assessment should be interpreted as an approximation and may be influenced by data-entry error. In addition, clinical context was incompletely recorded, we cannot determine whether non-standard regimens reflect true clinical practice, justified clinical contraindications, incomplete documentation, or data-entry error. Untreated cases should be interpreted with caution as this may reflect missing treatment entries, incomplete recording in surveillance data, or drug stockouts. Nonetheless, routine treatment audits, supportive supervision, and strengthened recording of contraindications and follow-up plans for radical cure could help reduce inappropriate case management [40]. It is also important to note that standard treatment relies on drug availability, so good drug management is needed to ensure that there is no stock-out of malaria drugs.

Predictors of hospital attendance included older age, non-Papuan ethnicity, being housewife or student, and residence in low- or moderate-endemic areas. These associations should be interpreted cautiously because key determinants such as socioeconomic status, education, and geographic access were not measured in SISMAL, and markers of clinical severity were rare (only 2 severe malaria cases) and incompletely recorded (parasite density available for ~20%). Therefore, while the observed pattern is consistent with non-clinical factors influencing facility attendance, this study cannot directly attribute hospital attendance to socioeconomic or geographic drivers [41]. Among those who reached hospitals, only age < 15 years predicted hospitalization. Given that most cases were recorded as uncomplicated while hospitalization was common, additional unmeasured factors (e.g., local admission practices, clinician risk perception, comorbidities, referral protocols, or documentation practices) may contribute and should be explored in future work.

The data quality challenges identified in this study, such as gaps in non-mandatory fields and reporting inconsistencies, may be common limitations of routine surveillance systems [42]. Nevertheless, these findings provide a critical evidence base for health authorities to refine the SISMAL system. These findings suggest that moving toward malaria elimination in Sorong City will require a shift toward evidence-driven supervision. Specifically, our results can inform the development of clearer data entry standards and the implementation of periodic quality audits. By identifying these systemic gaps, this study provides the necessary justification for targeted training of facility-level malaria staff, ensuring that routine data can reliably guide future elimination strategies.

This study has several limitations related to the study setting and study period. The data was confined to Sorong City, limiting generalizability to other malaria endemic regions with differing vector ecologies and healthcare capacities. Furthermore, temporal trends and year-to-year variability could not be evaluated since the analysis was based on a single year of surveillance data.

A key limitation of this study is that surveillance data depend heavily on completeness and accuracy, restricting analysis of treatment adequacy and disease severity due to missing parasitemia or clinical variables. Missing information about prior malaria infection and its treatment, current clinical symptoms and laboratory results (e.g., parasitemia density, hemoglobin levels) restricted the evaluation of treatment appropriateness as well as disease severity and its association with demographic profile and laboratory findings. Although we did not analyze the clinical symptoms, active screening led by CHCs likely facilitated the detection of malaria among asymptomatic individuals. In endemic areas like Sorong City, partial immunity may allow individuals to remain asymptomatic while harboring parasites. These individuals serve as persistent reservoirs because they can still carry gametocytes capable of infecting mosquito vectors. Our findings suggest that malaria elimination in Papua cannot rely exclusively on hospital-based passive detection; instead, targeted active screening in transmission “hotspots” is essential to identify and clear asymptomatic reservoirs and interrupt local transmission. Misclassification of mixed or zoonotic infections likely occurred given reliance on microscopy without molecular confirmation [43] and reporting delays in SISMAL may contribute to discrepancies between medical records and surveillance data.

Last but not least, behavioral drivers of care seeking among passively detected cases could not be directly measured and warrant future qualitative inquiry. This study relies on national surveillance data, which lacks granular socioeconomic and cultural variables, such as education level, income, and distance to facilities, that are known to influence healthcare utilization. Consequently, while our quantitative analysis identifies broad patterns, further qualitative research is required to provide a deeper understanding of the barriers to treatment seeking in this region.

Future research should integrate molecular diagnostics, expand surveillance across Papua, and link SISMAL with medical records using unique identifiers to improve completeness of clinical variables and treatment documentation. Mixed-methods approaches (including qualitative work) could clarify why patients attend hospitals versus primary facilities and why hospitalization is frequent despite predominantly uncomplicated malaria classifications. Considering the lack of data in the national surveillance system, better management and auditing is urgently required to produce complete and reliable data. Strengthening digital infrastructure and promoting interoperable data governance will be essential to advancing Indonesia’s malaria elimination goals.

Conclusions

Malaria remains a major burden in Sorong, driven by P. vivax. Strengthening surveillance, diagnostics, and equitable access is essential. Improving adherence to national case-management guidelines, particularly documentation and delivery of effective P. vivax radical cure where safe and appropriate, and strengthening primary-care capacity may support timely diagnosis, reduce avoidable hospital attendance and admissions, and advance Indonesia’s 2030 elimination goals in endemic communities.

Acknowledgments

We thank the Ministry of Health of the Republic of Indonesia for providing data in this study.