Switch It Off! Carbon, Financial and Health Service Impacts of Switching Off a Computed Tomography Scanner: A Quality Improvement Study

Medical imaging has been identified as a carbon hotspot in health care, and demand for imaging services is increasing. This study investigated switching off a surplus computerised tomography (CT) scanner when not clinically required as a possible simple and scalable intervention to reduce healthcare emissions. This before-after quality improvement study introduced a 'Switch it off' intervention where radiography staff switched off the power to a surplus CT scanner after hours (17:00-08:00) for 7 days ('intervention period': 07/7/2023-13/07/2023). Using a power data logger, power consumption (kilowatt hours, KWh) during the intervention period was compared to 7 days without the switch-off practice ('control period': 24/07/2023-31/07/2023). Financial and carbon emission impacts were calculated based on energy consumption. All CT radiographers working in the department were invited to undertake a pre and post intervention survey. Differences in quantitative data pre- and postintervention were analysed using chi-squared test for independent proportions. Free text survey responses were summarised into themes. Compared with energy use in the control period (433.96 kWh), there was a reduction in 139.79 kWh during the intervention period (294.17 kWh): 32% relative reduction. Extrapolation to 12 months found potential savings of 7280 kWh in energy use, $1381 to the hospital budget, and 5.5 T CO2e to the carbon budget. Of the 22 CT radiographers invited, 10 (45%) completed the survey, reporting no or trivial clinical impacts from switching off. The proportion of radiographers reporting switching off the scanner when not in use increased by 70% (95% CI: 39%, 100%; p = 0.002) from 10% (1/10) pre- to 80% (8/10) postintervention. Identifying and switching off surplus CT scanners in low use times is a simple and scalable intervention that can achieve significant power, financial and carbon savings with little to no impact on clinical workflow.