Characteristics and Influencing Factors of Healthcare Utilization in Post-conflict Primary Care Attendees in Northern Sri Lanka

**Background:** Sri Lanka provides a publicly funded universal healthcare system, yet little is known about actual healthcare utilization patterns, particularly in the Northern Province heavily affected by 30 years of civil war. With rising non-communicable diseases (NCDs) and mental health issues, understanding who uses services and why is critical for efficient resource allocation. This study aimed to explore characteristics of healthcare utilization by conflict-affected adult primary care attendees and examine service utilization at facilities providing outpatient and inpatient care.

**Methods:** This quantitative analysis used data from the COMGAP-S study, an epidemiologic survey conducted among adults attending 25 randomly selected public primary care facilities across all five districts of Northern Sri Lanka. Clusters were allocated proportionally by internally displaced population numbers. Within each facility, 41 individuals were recruited through systematic selection of every third attendee. Of 1,025 approached, 1,019 completed the Health Service Use survey (modified Client Service Receipt Inventory). Data were analyzed using SPSS version 26.0 with Chi-square, Fisher's Exact, and Monte Carlo tests, followed by binary logistic regressions on significant associations.

**Key Results:** Over half of participants (59.2%) were older than 50 years; 47.5% were male, 52.5% female; 89.2% were Tamil; 84.2% lived in rural areas; 42.8% were employed, 40.2% were housewives; 72.7% had education below O/L level. For outpatient services, 205 (20.1%) attended for chronic disease, with significant associations for age 50+ (80.5% of chronic disease visits; OR 3.412; 95% CI 2.31–5.03; p<0.000) and Tamil ethnicity (OR 0.430; 95% CI 0.27–0.67; p<0.000). Injury visits were associated with male gender (68.3%; p<0.001), Tamil ethnicity (OR 2.161; 95% CI 1.20–3.88; p<0.010), and being unemployed/student/retired (OR 0.541; 95% CI 0.40–0.72). Acute condition visits were associated with age below 50 (OR 0.525; 95% CI 0.37–0.74; p<0.000), education below O/L (OR 0.634; 95% CI 0.44–0.91; p<0.014), and rural residence (OR 2.332; 95% CI 1.41–3.84; p<0.001). For outpatient outcomes, 49.55% attended for drug prescriptions; 10.9% for follow-up (associated with age 50+: OR 3.044; 95% CI 1.87–4.94; p<0.000). Notably, 98.0% sought treatment at divisional hospitals rather than primary medical care units. A total of 161 participants (15.8%) reported paying fees at outpatient services, with rural residents most likely to pay 101–500 Sri Lankan rupees (p<0.046). For inpatient services, 378 (37%) stayed overnight. Inpatient chronic disease admissions were associated with age 50+ (OR 4.857; 95% CI 2.52–9.33; p<0.000), Tamil ethnicity (OR 0.340; 95% CI 0.19–0.60; p<0.000), and being unemployed/student/retired (OR 1.875; 95% CI 1.12–3.12; p<0.016). Overnight admission was associated with age 50+ (OR 1.676; 95% CI 1.27–2.2; p<0.000) and lower education (OR 1.542; 95% CI 1.12–2.10; p<0.007).

**Clinical Implications:** The strong preference for divisional hospitals over local primary medical care units suggests underutilization of decentralized primary care, potentially due to limited hours, lack of supplies, or perceived quality differences. Out-of-pocket payments contradict universal healthcare principles and risk catastrophic spending for poorer households. With NCDs accounting for 40% of premature deaths in Sri Lanka and an aging population, these findings underscore the urgent need for: (1) health education programs targeting older adults and rural populations about appropriate facility use; (2) formalized referral pathways to improve continuity of care; (3) expanded services and hours at primary medical care units; and (4) policies to reduce out-of-pocket expenditures. The results are relevant to other post-conflict and developing country settings facing similar challenges in healthcare resource allocation.