Background:

Hospital medicine faculty performance is measured using a combination of process measures (ie. discharge times and length of stay) as well as outcomes including readmissions, patient satisfaction, and mortality.  Recognition and rewarding high performers is an important strategy to improve overall work culture for our group.  We developed a composite hospital medicine provider performance score, which takes into account a combination of these variables for the purpose of engaging faculty, acknowledging high performers, identifying best practice behaviors, and creating accountability.

Purpose:

The primary purpose of this quality improvement project will be to develop a composite performance score for hospital medicine providers using available clinical metrics. Secondary goals of this effort include:

1)    Assess the relative value practicing hospitalists assign to available performance measures as an indicator for the quality of their care.

2)    Evaluate hospital medicine and provider performance over time after development of this score

3)    Identify high performers for the purpose of recognition to help foster a positive work environment and to help in the identification of best practices

4)    Measure the effect of this recognition on the work culture of our group via the work culture survey

Description:

The project team developed a survey using Qualtrics to ascertain the perceived value of hospital medicine metrics currently available.  We used the survey data to assess the value each faculty member assigned to common performance measures, which are currently available for routine hospital medicine administrative needs.  These measures include readmission rate, mortality index, patient satisfaction scores (HCAHPS), length of stay (LOS), discharge by 11 AM rates, and case-mix index (CMI).  Based on this survey data, a relative weight from 1-6 was assigned to each measure for the purpose of developing a composite score.

The composite score was developed by calculating the historical performance (FY 2014) by provider for the measures in question and assigning the relative weight.  The historical data was used to calculate a group mean for each measure as well as standard deviation for the purposes of normalization. Individual performance tended to fall within 1.5 standard deviations of the mean.  A score of zero was assigned to providers who fell below 1.5 standard deviations of the mean and a maximum score was assigned to those who fell above 1.5 standard deviations of the mean.  50% of the maximum score was assigned to those at the mean.  Other scores were assigned within a range of 0-100% of the maximum score proportional to the number of standards of deviation above or below the mean. 

The composite score will be tracked by provider quarterly and will be used to assess the top performer and improvement over time – overall and for the individual measures. Individual scores will be shared directly with the faculty group and the highest performers will be acknowledged.

Conclusions:  We developed this composite scoring system to use the metrics available to identify our quarterly top performer in order to recognize the performer in a public forum.  As more metrics become available to us over time, we hope to incorporate these newer metrics into our model.  We hope that recognizing our top performing hospitalist on a regular basis will improve our overall work culture.  We plan on expanding this initiative to all three hospitals in our health system.