Change in healthcare utilisation after surgical treatment: observational study of routinely collected patient data from primary and secondary care

Abstract

Background: Most patients fully recover after surgery. However, high-risk patients may experience an increased burden of medical disease.

Methods: We performed a prospectively planned analysis of linked routine primary and secondary care data describing adult patients undergoing non-obstetric surgery at four hospitals in East London between January 2012 and January 2017. We categorised patients by 90-day mortality risk using logistic regression modelling. We calculated healthcare contact days per patient year during the 2 yr before and after surgery, and express change using rate ratios (RaR) with 95% confidence intervals.

Results: We included 70 021 patients, aged (mean [standard deviation, sd]) 49.8 (19) yr, with 1238 deaths within 2 yr after surgery (1.8%). Most procedures were elective (51 693, 74.0%), and 20 441 patients (29.1%) were in the most deprived national quintile for social deprivation. Elective patients had 12.7 healthcare contact days per patient year before surgery, increasing to 15.5 days in the 2 yr after surgery (RaR, 1.22 [1.21-1.22]), and those at high-risk of 90-day mortality (11% of population accounting for 80% of all deaths) had the largest increase (37.0 days per patient year before vs 60.8 days after surgery; RaR, 1.64 [1.63-1.65]). Emergency patients had greater increases in healthcare burden (13.8 days per patient year before vs 24.8 days after surgery; RaR, 1.8 [1.8-1.8]), particularly in high-risk patients (28% of patients accounting for 80% of all deaths by day 90), with 21.6 days per patient year before vs 49.2 days after surgery; RaR, 2.28 [2.26-2.29].

Discussion: High-risk patients who survive the immediate perioperative period experience large and persistent increases in healthcare utilisation in the years after surgery. The full implications of this require further study.

Keywords: anaesthesia; complications; healthcare cost; peri-operative medicine; primary care; surgery.

Fig 1

Summary of inclusion in study cohort.

Fig 2

Cumulative rate of death, stratified by risk group among patients undergoing elective surgery. Shaded area indicates the 95% confidence interval; numbers in brackets are the proportion of patients undergoing elective surgery within each risk group.

Fig 3

Proportion of patient days in contact with healthcare in the 2 yr before and after surgery, stratified by risk group among patients undergoing elective surgery. Data aggregated by 7-day windows, each of which has 95% confidence intervals. Each point represents the proportion of patient days spent in contact with healthcare during the 7-day window. All patients contact healthcare on Day 0 (i.e. the date of surgery), and this was excluded from modelling to prevent distortion. Numbers in brackets are the proportion of patients undergoing elective surgery within each risk group.