Effect of weather on GP home visits: a cross-sectional study

Background GPs in the UK conduct > 13 million home visits each year. The visits, which are resource intensive, are usually to the frailest patients who are least resilient to adverse weather. Aim To explore the relationship between meteorological variables (temperature, rainfall, sunshine) and temporal variables (day of the week, season) with GP home visits (HVs). Design and setting A cross-sectional study using data provided by Herts Urgent Care for its GP acute in-hours visiting service and UK Meteorological (Met) Office weather data for the Herts & South East region of the UK. Method The association between the number of GP HVs and weather and temporal variables was explored using univariable and multivariable negative binomial regression. Results There was a significant 0.4% decrease in HVs per degrees Celsius increase in minimum temperature (incidence rate ratio [IRR] 0.996, 95% confidence interval [CI] = 0.993 to 0.999), and a 0.4% decrease per hour increase in sunshine (IRR 0.996, 95% CI = 0.992 to 1.000), as well as significant decreases in weekday HVs compared with Mondays (Thursday IRR 0.824, 95% CI = 0.790 to 0.859). There were 6.2% fewer HVs in summer compared with winter (IRR 0.938, 95% CI = 0.902 to 0.975). Multivariable negative binomial regression showed nonsignificant relationships between meteorological variables and HVs, but a significant day-of-theweek relationship. Conclusion GP HVs increased on cold days and fell on sunnier days. The effect sizes were small so it is unlikely that there is any clinically significant effect of weather on HVs in this acute GP visit setting. A tentative conclusion might also be that GPs in this system can deliver care to frail housebound patients in most weather conditions.