Setting Clear Expectations
Do you have a job description? When you read it, does it adequately describe what is expected of your hospitalists? Do you have an orientation for new members to your group? How long does it last? Is additional training offered? Are there outcomes that you expect from this training? And once you have oriented, trained, and offered a job description, does the actual work environment support or negate your efforts―i.e., does culture trump your formal process?
The cycle of setting clear expectations about work performance begins during the recruitment phase. Being absolutely forthcoming about what it is like to work in your group and what you expect from each and every member is paramount to allow both you and the candidate to determine a good fit. Once the physician has joined your group, orientation and training should hardly be a 1-, 2- or 3-day exercise. These are continuous and ongoing processes, given our rapidly changing practice environment. In fact, change is one of the only reliable characteristics of what we do, and extending the welcome “The job you take today is unlikely to be the job you will have next year” is hardly inappropriate. Be mindful that setting clear expectations with all of your hospitalists is the bedrock of a functional performance management system. Defining expectations alone will often improve performance, vis-à-vis the Hawthorne effect.
Expectations should always be depersonalized and focus on behavior. Behavior itself may be regarded in 2 distinct domains: those behaviors that are observed, and those outcomes that are measurable. Examples of observable behaviors include interpersonal interactions with nurses and consultants, pager response times, and attendance at monthly team meetings. Measurable outcomes include work RVU productivity, patient satisfaction, readmission rates, and compliance with coding and documentation guidelines. There are many ways to organize dimensions of performance that you may expect from your physicians―the 6 aims of quality (safe, timely, effective, efficient, equitable, and patient centered, as outlined in the IOM report Crossing the Quality Chasm), maintenance of a healthy workplace, citizenship, relationships with others, etc.―yet the key is to define and communicate them, then check often for understanding.
Measuring Actual Performance
Be the first to admit “the numbers are wrong,” and you will save hearing it from many others. There are many inherent problems in measuring actual performance, and the data may never be perfect. As an exercise, try assigning individual readmission rates within your group, and you will find that because of handoffs within the group and lack of precision in identifying who actually discharged the patient, there will be many arguments over whether the data is valid. However, in most circumstances, if the data is flawed, it still may serve a strong purpose to highlight the relative variation within the group. Searching for quantifiable systemic data and being transparent about the limitations of the data will be an exercise worth undertaking. In like manner, behavioral observation data are potentially fraught with conflict if the data are focused on judgment of character traits (I believe this hospitalist has a good bedside manner) rather than on observable behaviors (This hospitalist always/sometimes/never comes to meetings on time). Measures are best when they are objective, relevant to the position, and interpretable. Remember: All measures are flawed; some are useful.
Aligning Compensation With Expectations
Conventional wisdom states that people will do more of what they are incentivized to do. The corollary to this is to be sure what you incentivize is actually what you want. For the group that is trying to improve individual productivity and reduce length of stay, providing financial rewards for work RVU’s alone may result in less assertiveness in managing timely discharges and bickering over who picks up the 11 p.m. vs. 2 a.m. overnight admission the following morning. Ultimately, compensation must be intimately linked with the mission of the group, and tremendous care must be taken in determining the construct of any system. Although it is well beyond the scope of this article to detail the many considerations of designing a compensation system, one must understand that it is only one component―and not the most important component―of a performance management program.