A small number of surgeons outside the control-limit: an observational study based on 9,482 cases and 208 surgeons performing primary total hip arthroplasties in western Sweden

Abstract

Background and purpose — Feedback programs relating to surgeon levels have been introduced in some orthopedic quality registers around the globe. The aim of an established surgeon feedback program is to help surgeons understand their practice and enable an analysis of their own results. There is no surgeon feedback program in Sweden in the orthopedic quality registers and there is a fear that a feedback system might pinpoint surgeons as poor performers, partly due to patient case mix. As a step prior to the introduction of a future possible feedback program in Sweden, we assessed the variation in the occurrence of adverse events (AE) within 90 days and reoperations within 2 years between surgeons in western Sweden and explored the number of surgeons outside the control-limit following primary total hip arthroplasties (THAs).Patients and methods — Patient data, surgical data, and information on the surgeons, relating to surgeries performed in 2011–2016, were retrieved from 9 publicly funded hospitals in western Sweden. Data from medical hospital records, the Swedish Hip Arthroplasty Register (SHAR) and a regional patient register located in western Sweden were linked to a database. Funnel plots with control-limits based on upper 95% and 99.8% confidence intervals (CI) were used to illustrate the variation between surgeons in terms of the outcome and to explore the number of surgeons outside the control-limit. Both observed and standardized proportions are explored. The definition of surgeons outside the control-limit in the study is a surgeon above the upper 95% CI.Results — The study comprised 9,482 primary THAs due to osteoarthritis performed by 208 surgeons, where 91% of the included primary THAs were performed by orthopedic specialists and 9% by trainees. The mean overall annual volume for all surgeons was 27. The observed overall mean rate for AEs within 90 days for all surgeons was 6.2% (5.8–6.7) and for reoperations within 2 years 1.8% (1.7–2.2). The proportion of surgeons outside the 95% CI was low for both AEs (0–5%) and reoperations within 2 years (0–1%) in 2011–2016. The corresponding numbers were even lower for AEs (0–3%) but similar for reoperations (0–1%) after standardization for differences in case mix. In a sub-analysis when the number of surgeries performed was restricted to more than 10 primary THAs annually to being evaluated, almost half or more of all the surgeons were excluded from the annual analysis. The result of this restriction was that all surgeons outside the control-limit disappeared after standardization for both AEs and reoperations for all the years investigated. Considering the complete period of 6 years, less than 1% (1 high-volume surgeon for AEs and 2 high-volume surgeons for reoperations) after risk adjustments were outside the 95% CI, and no surgeons were outside the 99.8% CI.Interpretation — In a Swedish setting, the variation in surgeon performance, as measured by AEs within 90 days and reoperations within 2 years following primary THA, was small and 3% or less of the surgeons were outside the 95% CI for the investigated years after adjustments for case mix. The risk for an individual surgeon to be regarded as having poor performance when creating surgeon-specific feedback in the SHAR is very low when volume and patient risk factors are considered.