Assessment of the Accuracy and Reliability of Vascular Surgery Quality Metrics
Zachary B. Fang, MD, MSc, Calvin L. Chao, BA, Andres Schanzer, MD, Edward J. Arous, MD, Elias J. Arous, MD, Danielle R. Doucet, MD, Dejah Judelson, MD, Louis M. Messina, MD, Jessica P. Simons, MD, MPH, Robert C. Steppacher, MD, Julie M. Flahive, MS, Dawn M. Durocher, BA, Francesco A. Aiello, MD.
University of Massachusetts Medical School, Worcester, MA, USA.
Objectives: Healthcare quality metrics are crucial to medical institutions, payers, and patients. Obtaining current and reliable quality data is challenging, as publicly reported databases lag by several years. Vizient Clinical Data Base (previously University Health Consortium (UHC)) is utilized by over 5000 academic and community medical centers to benchmark healthcare metrics with results based on predetermined Vizient service lines. We sought to assess the accuracy and reliability of vascular surgery service line metrics, as determined by Vizient.
Methods: Vizient utilizes encounter-data submitted by participating medical centers and generates a diverse array of health care metrics ranging from mortality to costs. All inpatient cases captured by Vizient under the vascular surgery service line were identified at the University of Massachusetts Medical Center (fiscal year 2016). Each case within the service line was reviewed and categorized as “vascular” or “non-vascular” based on care provided by UMass vascular surgery faculty: vascular=vascular surgery was integral part of care, non-vascular=vascular surgery had minimal or no involvement. Statistical analysis comparing length of stay (LOS), cost, readmission, mortality, and complication rates between vascular and non-vascular cohorts was performed. All inpatient cases discharged by a vascular surgeon NPI were also reviewed and categorized according to Vizient service lines.
Results: Vizient’s vascular surgery service line identified 696 cases, of which 556 (80%) were vascular and 140 (20%) were non-vascular. When comparing these two cohorts, vascular cases had a significantly lower LOS (3.4 vs. 8.7 days; p<0.0001), cost ($8,535 vs. $16,498; p<0.0001), and complication rate (6.5% vs. 18%; p<0.0001) than non-vascular. Mortality was also lower (1.6% vs. 5.7%; p<0.01), but after risk-adjustment, this difference was not significant. When discharging vascular surgeon NPI was used to identify vascular surgery cases, only 69% of these cases were placed within the vascular surgery service line.
Conclusions: Health care quality metrics play an important role for all stakeholders but obtaining accurate and reliable data to implement improvements is challenging. In this single institution experience, inpatient cases that were not under the direction or care of a vascular surgeon resulted in significantly negative impacts on LOS, cost, complication rate, and mortality to the vascular surgery service line, as defined by a national clinical database. Therefore, clinicians must understand the data abstracting and reporting process prior to implementing effective strategic plans.
Back to 2019 Abstracts