Objective To investigate the prevalence of non-occupational carbon monoxide poisoning in Pudong New Area, Shanghai, China in recent years, and analyze the epidemiological characteristics, spatial-temporal distribution patterns, and correlation with meteorological factors.

Methods Data on non-occupational carbon monoxide poisoning cases and concurrent meteorological data in Pudong New Area, Shanghai from November 2022 to September 2025 were collected. Descriptive analysis and Spearman rank correlation analysis were applied, combined with ArcGIS visualization, to explore the spatial-temporal distribution characteristics of poisoning cases and their correlation with meteorological factors.

Results A total of 308 cases of non-occupational carbon monoxide poisoning were reported in Pudong New Area, Shanghai from November 2022 to September 2025. The male-to-female ratio was 1∶1.26, and the high-risk age group was 18-44 years (65.9%). Most cases were mild poisoning (75.0%). The main causes of poisoning were improper use of coal/charcoal hot pot (35.7%) and heating with fuels such as coal, charcoal, and straw (32.5%). Household poisoning events accounted for 83.3%, mostly sporadic and involving 1-2 persons (1.3 cases per event). The cases generally exhibited a U-shaped distribution, peaking in winter (December-February) and concentrating in the central region. Poisoning events at food service establishments accounted for 8.6%, and they were prone to clustered poisoning (3.4 cases per event). The cases showed a fluctuating pattern with a peak in summer (July). Spatial autocorrelation analysis showed that the Global Moran's I value was 0.34 (P < 0.001), and there was a spatial positive correlation in the poisoning distribution. Monthly average temperature (r_s=-0.421, P=0.013), monthly maximum temperature (r_s=-0.411, P=0.016), monthly minimum temperature (r_s=-0.418, P=0.014), and relative humidity (r_s=-0.506, P=0.002) were negatively correlated with poisoning incidence, while monthly average atmospheric pressure (r_s=0.437, P=0.010) was positively correlated.

Conclusion Non-occupational carbon monoxide poisoning demonstrates distinct spatial-temporal distribution patterns and is associated with meteorological factors. Future prevention efforts should integrate surveys of resident health-related behaviors and meteorological early warnings to establish a targeted prevention and control system for reducing the incidence of poisoning.