Standardized combined outcome index as an instrument for monitoring performance after pulmonary resection

Modern healthcare systems demand more transparent and accurate monitoring of clinical performance with the purpose to improve standards of care in a cost-effective way. Outcomes, such as mortality, are still the most widely used quality indicators in our specialty. However, previous studies have shown that mortality alone does not reflect performance accurately in our specialty. Ideally, multiple risk-adjusted outcomes should be used for a more comprehensive assessment. The objective of this analysis was to develop and use an index combining multiple risk-adjusted outcomes to track down the performance of our thoracic surgery unit over time.

In all, 511 major lung resections (465 lobectomies, 46 pneumonectomies) performed from January 2005 through September 2010 were analyzed. Four risk-adjusted outcomes were considered: 30 days or in-hospital mortality, cardiopulmonary morbidity, unplanned/emergency intensive care unit admission, and prolonged length of stay (more than 14 days, prolonged hospital stay). Risk adjustment was performed using published regression models. Each indicator was converted into its opposite (ie, mortality rate to survival rate) so that higher scores reflected better performance. Moreover, to account for differences in measurement scales, the standardized outcomes were rescaled according to their mean total standard deviations. Finally, the individual rescaled indicators of each year were summed to generate a combined outcome index.

Mean cumulative observed mortality, morbidity, unplanned intensive care unit, and prolonged hospital stay rates were 1.8%, 23%, 6.6%, and 7.4%, respectively. The combined outcome index scores showed a progressive improvement of performance during the study period, progressing from -3.48 in 2005 to 2.87 in 2009. The combined outcome index was also used prospectively in a variable life-adjusted display chart to track down trends of practice variation in the last 6 months.

The present analysis is proposed as a methodologic template for developing a risk-adjusted index combining four different outcomes. It aims at overcoming inherent limitations of outcomes when used individually for performance assessment. This or similar combined indexes may be effective instruments of internal clinical audit and could be incorporated along with process indicators in composite performance scores to more comprehensively evaluate the postoperative domain of our practice.