Are temporal trends in retained foreign object rates after surgery in Switzerland impacted by increasing coding intensity? A retrospective analysis of hospital routine data from 2000 to 2019.

Abstract

Retained foreign objects (RFOs) after surgery can cause harm to patients and negatively impact clinician and hospital reputation. RFO incidence based on administrative data is used as a metric of patient safety. However, it is unknown how differences in coding intensity across hospitals and years impact the number of reported RFO cases. The objective of this study is to investigate the temporal trend of RFO incidence at a national level and the impact of changes in coding practices across hospitals and years. Retrospective study using administrative hospital data. 21 805 005 hospitalisations at 354 Swiss acute-care hospital sites PRIMARY AND SECONDARY OUTCOME MEASURES: RFO incidence over time, the distribution of RFOs across hospitals and the impact of differences in coding intensity across the hospitals and years. The annual RFO rate more than doubled between 2000 and 2019 (from 4.6 to 11.8 with a peak of 17.0 in 2014) and coincided with increasing coding intensity (mean number of diagnoses: 3.4, SD 2.0 in 2000; 7.40, SD 5.2 in 2019). After adjusting for patient characteristics, two regression models confirmed that coding intensity was a significant predictor of both whether RFO cases were reported at the hospital level (OR: 12.94; 95% CI: 7.38 to 22.68) and the number of reported cases throughout the period at the national level (Incidence Rate Ratio (IRR): 5.95; 95% CI: 1.11 to 31.82). Our results raise concerns about the use of RFO incidence for comparing hospitals, countries and years. Utilising coding indices could be employed to mitigate the effects of coding intensity on RFO rates.