In the wake of news coverage and studies examining surgeons with overlapping surgical procedures and patient outcomes, Burns et al. examined anesthesiologists covering overlapping anesthesia cases and the staffing ratio impact on patient outcomes in this retrospective, matched cohort study of major noncardiac inpatient surgical procedures done in 23 U.S. academic and private hospitals (JAMA Surg 2022;157:807-15). The study used the Multicenter Perioperative Outcomes Group (MPOG) database, which is populated directly from the electronic medical records, including anesthesia records (asamonitor.pub/3ftbeZt). Only cases where the anesthesiologist working with one to four nurse anesthetists at same time were included. (Cases where an anesthesiology resident was in the case >25% of time were excluded.) Propensity score-matching methods were applied. Staffing ratio was based on the anesthesiologist in the staffing grid of the electronic record. The staffing ratio was a weighted average of the staffing ratios throughout the anesthesia case into four groups: Group 1 (only covering one anesthesia case), Group 1-2 (>1 and ≤2), Group 2-3 (>2 and ≤3), and Group 3-4 (>3 and ≤4). The major outcomes were 30-day mortality and major surgical morbidity in six areas (cardiac, respiratory, gastrointestinal, urinary, bleeding, and infectious complications). Major exclusions were personally performed cases, cases at night or weekend cases, and fixed staffing ratio cases (e.g., cardiac surgery). More than 575,000 cases were included. Compared to Group 1-2, Groups 2-3 and 3-4 had a 4% and 5.25% increase, respectively, in risk-adjusted mortality and morbidity.
Why it matters
In a strategy to reduce staffing costs, anesthesiology groups are pressured to increase staffing ratios and sometimes forced to create staffing models with high ratios. In contrast to previous studies on anesthesia outcomes using claims database, this study utilizes a large database that is populated by the EMR that provides more granular and detailed data.
The results are consistent with daily clinical experience. Throughout any given day, the anesthesiologist must evaluate each patient and determine what the best staffing ratio is for that patient and the planned surgery. When the staffing model allows for flexing up and down with the staffing ratio, then the anesthesiologist can change coverage to meet the patient's needs (what I call “demand matching”). Issues arise when the staffing model is already maxed out at 1:4 – that is, each anesthesiologist is assigned four different anesthetizing sites, then needs to provide closer care to one patient. Either the other patients may not receive the care they need, or the anesthesiologist's colleagues must cover more than four sites during this time.
This study provides evidence that this inability to flex staffing to meet “demand matching” leads to worse patient outcomes. In other words, staffing ratios need to be determined by patient needs, and staffing models need to take this into account further. Extrapolating the results, this study provides evidence that having no anesthesiologist involvement in anesthesia care will lead to worse patient outcomes.