Can collaborative learning, confidential performance feedback, and benchmarking help hospitals improve outcomes after heart bypass surgery?

The regional bypass surgery death rate fell to a level 24 percent lower than expected among five northern New England hospitals collaborating in a quality improvement intervention that included performance feedback, training in quality improvement, and benchmarking activities.

Why is this important?

Heart bypass surgery—one of the most frequently performed surgical procedures—has received much attention in quality improvement efforts. One of the most rigorous is the Northern New England Cardiovascular Disease Study Group (NNECDSG), a voluntary consortium of hospitals founded in 1987 when the Medicare program was publicly reporting hospital mortality rates. This group was a pioneer in the field and its efforts have served as an example to stimulate other similar initiatives.

Interventions

By collecting detailed clinical outcomes data over time, the NNECDSG discovered wide variation in bypass surgery death rates, ranging from 2 percent to 9 percent among surgeons (O'Connor et al. 1991). In response, the group implemented a cooperative quality improvement intervention consisting of:

• regular confidential feedback of outcomes data to each surgeon and hospital in the region;
• training in quality improvement techniques; and
• a series of site visits by multidisciplinary teams to discover best practices at each institution.

The complete surgical process was delineated to identify variations in process flow among the five participating hospitals (Kasper et al. 1992). Numerous changes were initiated at each hospital based on a better understanding about the relationship between processes and outcomes of care (O'Connor et al. 1996).

Findings

After implementation of the intervention, the regional in-hospital bypass surgery death rate fell to 3.3 percent, 24 percent lower than the expected rate of 4.3 percent based on historical regional data. This improvement translated to 74 fewer deaths from 1991 to 1993 (the original post-intervention study) (O'Connor et al. 1996).

By 2002, the regional mortality rate had fallen to 2.1 percent, 62 percent lower than an expected rate of 5.5 percent given patient characteristics. This difference translates to 811 fewer deaths than predicted if the historical trend had continued (NNECDSG 2003).

Implications

Rigorous outcomes-based clinical benchmarking offers a data-driven approach to quality improvement and clinical learning. Its potential urges that policymakers—public and private—continue to promote its adoption elsewhere. For example, Blue Cross Blue Shield of Michigan defrays a portion of the cost for affiliated hospitals to participate in the National Surgical Quality Improvement Program (described below).

Improvement Ideas and Resources

Additional resources are available from the Northern New England Cardiovascular Disease Study Group.

The American College of Surgeons offers a National Surgical Quality Improvement Program (NSQIP), modeled on a successful program developed by the Veterans Administration. NSQIP allows participating hospitals to compare standard surgical outcomes so that they can identify problems and make improvements (McCarthy 2005).

Measure:

Clinical data for this time series were collected on 15,095 consecutive patients undergoing isolated coronary artery bypass graft (CABG) surgery at five hospitals from July 1987 through July 1993 (with updated analysis for a total of 44,121 patients treated from 1987 through 2002). A multivariate regression model, which controlled for changes in patient characteristics over time, was used to predict hospital mortality during the post-intervention period in comparison with observed mortality rates (O'Connor et al. 1992). Results were similar when the analysis was limited to surgeons who were members of the consortium from the beginning of the study (O'Connor et al. 1996).

An analysis of modes of death associated with CABG surgery found that 80 percent of the variation in surgeon-specific mortality rates was attributable to fatal heart failure, but this variation was not explained by measurable differences in patients' clinical conditions (preoperative left ventricular function). The authors concluded that "the most likely explanation for the differences in rates of fatal heart failure in the current study is that different processes and systems of clinical care yield different results" (O'Connor et al. 1998).

Limitations:

The study was not a controlled trial; other factors may have contributed to the observed improvements. However, the magnitude of change exceeded the trend in surgical mortality at the time of the study (O'Connor et al. 1996).

Source:

Clinical registry data were collected and analyzed by the Northern New England Cardiovascular Disease Study Group (NNECDSG 2003; O'Connor et al. 1996). Participating hospitals included Catholic Medical Center, Manchester, N.H.; Dartmouth-Hitchcock Medical Center, Hanover, N.H.; Eastern Maine Medical Center, Bangor; Fletcher-Allen Health Care, Burlington, Vt., and Maine Medical Center, Portland.

References:

* Indicates source of data used in the chart(s).Kasper, J. F., S. K. Plume, and G. T. O'Connor. 1992. A Methodology for QI in the Coronary Artery Bypass Grafting Procedure Involving Comparative Process Analysis. Quality Review Bulletin 18 (4): 129–33.

McCarthy, D. 2005. Case Study: The National Surgical Quality Improvement Program. Quality Matters (May). New York: The Commonwealth Fund.

* NNECDSG (Northern New England Cardiovascular Disease Study Group). 2003. A Regional Intervention to Improve In-Hospital Mortality Associated with CABG Surgery. Lebanon, N.H.: Northern New England Cardiovascular Disease Study Group.

O'Connor, G. T., J. D. Birkmeyer, L. J. Dacey et al. 1998. Results of a Regional Study of Modes of Death Associated with Coronary Artery Bypass Grafting. Northern New England Cardiovascular Disease Study Group. Annals of Thoracic Surgery 66 (4): 1323–8.

O'Connor, G. T., S. K. Plume, E. M. Olmstead et al. 1992. Multivariate Prediction of In-Hospital Mortality Associated with Coronary Artery Bypass Graft Surgery. Northern New England Cardiovascular Disease Study Group. Circulation 85 (6): 2110–8.

O'Connor, G. T., S. K. Plume, E. M. Olmstead et al. 1991. A Regional Prospective Study of In-Hospital Mortality Associated with Coronary Artery Bypass Srafting. The Northern New England Cardiovascular Disease Study Group. Journal of the American Medical Association 266 (6): 803–9.

* O'Connor, G. T., S. K. Plume, E. M. Olmstead et al. 1996. A Regional Intervention to Improve the Hospital Mortality Associated with Coronary Artery Bypass Sraft Surgery. The Northern New England Cardiovascular Disease Study Group. Journal of the American Medical Association 275 (11): 841–6.