乌鸦传媒

As seen for hospital-based procedures, ambulatory surgical outcomes were significantly better on average at centers with relatively high volumes, Medicare data indicated.

Odds of hospital admission or emergency department visit within 7 days of an ambulatory surgical procedure were 21% greater in low- versus high-volume centers (95% CI 9%-36%), according to Jeffrey H. Silber, MD, PhD, of Children's Hospital of Philadelphia, and colleagues.

The disparity was even greater for patients with multiple comorbidities, with an odds ratio of 1.57 (95% CI 1.27-1.94), the researchers . Such patients had nearly double the odds for admission or emergency care when orthopedic procedures were involved (OR 1.84, 95% CI 1.36-2.50).

"Older patients with multimorbidity should discuss with their surgeon the optimal location of their care," Silber and colleagues advised.

Numerous studies had shown that fewer complications, lower mortality, and less need for readmission or emergency care, at least in the short term. Practice may not always make perfect, but it should (and apparently does) raise rates of favorable outcomes. But whether the same is true for ambulatory surgical centers -- which promise same-day discharge, less red tape, and often lower cost -- has not been studied before, the researchers explained.

Silber and colleagues took a first crack at it by analyzing Medicare claims data from 2018 and 2019 for patients older than 65 undergoing any of 165 procedures with unique CPT codes. More than 150,000 surgeries were performed, of which 4,751 involved hospital admission or emergency department visit during the following week. For the current analysis, each of these cases were matched with five not involving a short-term revisit. Controls had the same procedure and morbidity count as their corresponding case, with "close" matches for demographics and the types of comorbidities.

Mean patient age was 75; about 7.5% were 85 or older. Some 57% were men and 88% were non-Hispanic white. About 22% had multiple morbidities, with an average of nearly seven.

Among the different procedures, the most common involved the knee and leg (26%), followed by prostate surgeries (19%) and hernia repairs (10%).

Ambulatory centers were included in the analysis if they had conducted at least 11 procedures in one category and at least 100 of all kinds. That totalled to 2,328 centers in all. "Low volume" was defined more or less arbitrarily as fewer than 50 surgeries during the 2-year study period, and 747 centers were in this category; "high volume" was 50 or more, with 1,581 centers meeting this criterion.

Within the low-volume group, about one-third performed fewer than 20 procedures. In the high-volume group, two-thirds conducted 100 or more.

Something like a dose-response relationship was seen for volume versus risk for revisit. At the very low-volume centers (11-19 procedures total), odds for revisit were 50% greater than those for centers with ≥100; for the group with volumes of 20-49, the increase in odds was just over 10%, again versus centers with ≥100. There was no difference in revisit rates for centers performing 50-99 procedures versus those with ≥100.

Other factors that could be tracked in the Medicare data had no bearing on revisit risk, the researchers said. These included rural versus urban location, local poverty levels, accreditation status, numbers of surgical specialties represented or numbers of operating rooms, presence of nurse anesthetists, or distance from the nearest hospital.

Silber and colleagues also looked at data for individual surgeons practicing in ambulatory centers. These data also included procedures they performed on Medicare patients in regular hospitals. The overall trend toward higher revisit rates for those with low volumes was repeated, although when analyzed by volume quartile for all patients, the difference between the lowest and highest did not reach statistical significance (OR 1.13, 95% CI 0.97-1.31). On the other hand, when examining only patients with multiple morbidities, the difference was much greater (OR 1.57, 95% CI 1.17-2.11).

Limitations to the study were mainly those related to Medicare data, which may contain errors and don't include many potential confounding parameters, such that control patients might have differed significantly from cases.

Comment