Uncommon combinations of ICD10-PCS or ICD-9-CM operative procedure codes account for most inpatient surgery at half of Texas hospitals

Study objective Recently, there has been interest in activity-based cost accounting for inpatient surgical procedures to facilitate “value based” analyses. Research 10–20 years ago, performed using data from 3 large teaching hospitals, found that activity-based cost accounting was practical and useful for modeling surgeons and subspecialties, but inaccurate for individual procedures. We hypothesized that these older results would apply to hundreds of hospitals, currently evaluable using administrative databases. Design Observational study. Setting State of Texas hospital discharge abstract data for 1st quarter of 2016, 4th quarter of 2015, 1st quarter of 2015, and 4th quarter of 2014. Patients Discharged from an acute care hospital in Texas with at least 1 major therapeutic (“operative”) procedure. Measurements Counts of discharges for each procedure or combination of procedures, classified by ICD-10-PCS or ICD-9-CM. Main results At the average hospital, most surgical discharges were for procedures performed at most once a month at the hospital (54%, 95% confidence interval [CI] 51% to 55%). At the average hospital, approximately 90% of procedures were performed at most once a month at the hospital (93%, CI 93% to 94%). The percentages were insensitive to the quarter of the year. The percentages were 3% to 6% greater with ICD-10-PCS than for the superseded ICD 9 CM. Conclusions There are many different procedure codes, and many different combinations of codes, relative to the number of different hospital discharges. Since most procedures at most hospitals are performed no more than once a month, activity-based cost accounting with a sample size sufficient to be useful is impractical for the vast majority of procedures, in contrast to analysis by surgeon and/or subspecialty.