UW Shoulder and Elbow Academy: How surgeons can improve their surgical outcomes.

This post was prompted by a recent article, Strategies for Effective Implementation of Patient-Reported Outcome Measures in Arthroplasty Practice, that highlighted some of the challenges of using information technology to capture meaningful data on the comfort and function of patients before and sequentially after treatment.

Now may be an opportune time to re-examine the purpose and practicality of documenting the patient's condition over time.

For us, the most important reason for recording patient-reported data is to learn for different groups of patients what is and what is not working in our hands. Only our patients can teach us if we're going about treatment in the most effective way. But in order to gather actionable intelligence, we must strive to "follow each of our patients long enough to learn if the treatment was successful and if not to ask "why not"?" (see A Study in Hospital Efficiency: As Demonstrated by the Case Report of First Five Years of Private Hospital). Whatever method is applied, the principal priority is to capture data on the highest possible percentages of patients being treated. Patients at increased risk for treatment failure - for example, those who are socially disadvantaged, less well educated, who are not fluent in English, who live in remote rural settings, who have co-morbidities - are subject to systematic exclusion if followup depends on hooking up with a computer or tablet or traveling to the office for an examination of range of motion.

By contrast the most user-friendly and inclusive approach is a short questionnaire that can be filled either in the office or at home (see this link). This approach eliminates the travel and technology hurdles that can create a non-response bias in the data.

For shoulder surgery, we've found that the 12 "yes" or "no" questions of the Simple Shoulder Test (SST - see this link) enable the capture the information from a high percentage of the patients at a year after surgery: in a recent comparison the 2 year non-response rate for the SST was 13% while the 1 year non-response rate for the PROMIS system was 33% (see this link).

The SST does not require scoring but rather enables the surgeon and the patients to identify the specific functions that were improved after treatment - e.g. the ability to sleep on the affected side, or the ability to reach over head - in contrast to a "score", the significance of which is more difficult to communicate.

While the minimal clinically important difference (MCID) for the SST has been well studied (see this link), again this concept is harder to discuss with patients. Instead, we've found that the percent of maximal possible improvement (%MPI) is more straightforward: "our experience with shoulders like yours is that we can improve comfort and function X% of the way to normal" (see MCID vs %MPI at this link).

The bottom line is that by engaging our patients by emphasizing the importance of their providing us with longitudinal updates on their comfort and function, we can learn which treatments work best for which patients with which diagnoses and which do not.

Only our patients can teach us why those in the red rectangle did not get the outcome that they and we desired. For our learning, the individuals in the red rectangle have more to teach than the curves showing the mean ± standard deviation.