RESULTS
The extraction process identified 49,223 ESS procedures during the study period. The procedures were performed across 129 hospital trusts. Day-case surgery was conducted for 35,494 (72.1%) patients. The characteristics and outcomes for patients seen as day-case and with overnight stay are summarised in Table 1 . Patients with an overnight stay were on average three years older than day-case patients and more likely to have some level of frailty. Outcomes tended to be slightly poorer for overnight stay patients, most notably for complications (including haemorrhage) recorded during the index admission. Such complications are likely to have necessitated overnight stay in many cases. Although the one-year mortality rate was higher for patients with an overnight stay, this partly reflects the older age structure of the population. Mortality rates were low.
Of the 121 trusts that conducted more than 50 procedures across the five-year study period, 42 trusts (14,186 patients) performed ≥80% of procedures as day-case and 15 trusts (4,634 patients) performed <50% of procedures as day-case. In these 121 trusts, the highest rate of day-case surgery was 100% in three trusts (561 patients) and the lowest rate was 20.6% (n = 170 patients) in one trust. The distribution of day-case rates for each trust is summarised as a funnel plot in Figure 1 .
There was a substantial increase in the proportion of procedures conducted as day-case across the study period, from 64.0% in financial year 2014/15 to 78.7 % in financial year 2018/19. Despite this increase in the overall rate of day-case surgery, in financial year 2018/19, seven trusts still performed fewer than 50% of ESS procedures as day-case. The proportion of patients seen as day-case and their outcomes are shown in Table 2 . Patient outcomes were generally stable across the five-year period, with no clear trend.
The results of the multilevel, multivariable models comparing outcomes for day-case and overnight stay patients are presented in Table 3 . The odds of poorer outcomes were generally lower for day-case patients after adjusting for covariates and in the case of 30-day emergency readmission, one-year return for ENT surgery and 30-day complications significant. The results of the analysis comparing trusts with high and low rates of day-case surgery are presented inTable 4 . There was no significant difference in outcomes between high and low day-case rate trusts.