Abstract Aim Colorectal surgery is the highest contributor of surgical site infection (SSI) of any surgical specialty in England. SSI is an important metric for patient safety, yet there is significant variability in its reported incidence following colorectal surgery. National data is offered voluntarily and only includes diagnoses made in hospital. Enhanced recovery programmes mean that SSI frequently presents in primary care after discharge, meaning the true incidence is likely to be underestimated by hospital data alone. Using routinely collected data to estimate SSI is a potential alternative to resource intensive surveillance strategies. Comprehensive data regarding the frequency, distribution and risk factors for SSI may improve the recognition of at-risk groups and enable more targeted and effective prevention strategies. Methods A cohort of adult patients undergoing elective or emergency colorectal resection in England between 2014 and 2019 was identified. Patients were included from linked databases combining Hospital Episode Statistics (HES) data from NHS hospitals and Clinical Practice Research Datalink (CPRD) data from GP practices. Cases were stratified according to disease, operative and patient factors. The frequency of SSI was analysed using MedCodes or ICD-10 codes for wound infection within 30 days of surgery. A generalised estimating equation with clustering by GP practice was used to determine risk factors for SSI. Results 59 376 patients were included in the final analysis, with a median age of 67; 51.3% were male. 59.6% of operations were open procedures and 67.3% of cases were elective. The commonest procedures were rectal/anterior resection (39%) and right hemicolectomy (38%). The most frequent indication for surgery was cancer (55%). The overall SSI rate was 8.3%, strikingly similar to that reported by the UK government (8.7%) over this time period1. 3770 (76.3%) SSIs were diagnosed in hospital and 1168 (23.7%) were diagnosed in the community. Patients who developed an SSI had a significantly longer hospital stay (12 versus 8 days; P 0.001). Only 23.5% of patients diagnosed with SSI in primary care required readmission to hospital. On regression modelling, SSI was associated with smoking (OR 1.14, 95%c.i. 1.07–1.22), diabetes (OR 1.30, 95%c.i. 1.19–1.41) and diverticular disease (OR 1.47, 95%c.i. 1.33–1.66; all P 0.001). Laparoscopic approach (OR 0.63, 95%c.i. 0.59–0.67; P 0.001) and elective admission (OR 0.90, 95%c.i. 0.84–0.96; P = 0.002) were protective. Conclusion This study represents the largest national analysis of SSI incidence following colorectal surgery across primary and secondary care to date. With the necessary automation and data flows, existing linked population databases could provide consistent feedback on SSI incidence. This could form the basis of a coordinated local and national surveillance programme for colorectal SSI. Further validation studies are needed to explore differences in SSI recording between primary and secondary care, as well as following elective and emergency surgery.
Dean et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: