Introduction
Suicide and mental health disorders are a recognized increasing public concern. Most suicide prevention rely on evidence from mortality data, although suicide attempts are a better predictor for completed suicides. Understanding spatio-temporal patterns and demographic profiles of people at risk can improve suicide prevention schemes, including for carbon monoxide (CO) poisoning, a common method for gas-related suicides.

Objective
Describe spatio-temporal patterns of intentional CO poisoning hospitalization rates in England between 2002 and 2016, and identify population sub-groups at risk.

Methods
We used NHS Digital's Hospital Episode Statistics (HES) routinely collected data on hospital admissions for intentional CO poisoning. We estimated age-standardised rates (ASR) by year, gender and residential small-area characteristics, including rural/urban, deprivation and ethnic composition. Temporal trends were assessed through linear regression and joinpoint regression analysis. Regional differences were explored.

Results
On average, we identified 178 hospital admissions for intentional CO poisoning per year. The ASR decreased substantially over the study period, particularly among males (average annual percent change of −7.8 % (95 % CI: −11.0; −4.6)), in comparison to 3.9 % (95%CI, −6.4; −1.4) among females. Most admissions (81 %) occurred in males. White men aged 35–44 years were particularly at risk. The ASR in London (0.08/100,000) was almost six times lower than in the South-West (0.47/100,000).

Conclusions
This study provides novel insights into attempted suicides by intentional CO poisoning. Further prevention interventions, targeting sub-groups at risk (i.e. white men in their 30s/40s), need to be developed and implemented to reduce the burden of suicides and of CO poisoning.