Patient-Physician Agreement Using Summary Outcome Determination Scores

To determine whether the Summary Outcome Determination (SOD) score had exhibited a high level of physician-patient agreement in surgical patients.

The medical records of 320 postoperative patients were reviewed, of whom 164 patients were included in the study. Patients were included if both physician-assigned and patient-assigned SOD scores had been recorded. The SOD is administered as follows: the patient is asked "Compared to before surgery, is your elbow/shoulder better, worse or no different?" If better: "Is it improved, greatly improved, almost normal or normal?" If worse: "Is it worse or profoundly worse, or as bad as dying?" Each category is associated with a numerical value and definition for further clarification. The patient is asked to assign a category and a numerical value after the physician has already done so. These categories and values were evaluated between raters (ie, physician and patient) to assess reliability.

The intraclass correlation coefficient of physician-patient numerical ratings was "excellent" (0.93). The Bland-Altman 95% limits of agreement on the differences between the physician and the patient ranged from -1.3 to 1.3. The physician and patient numerical rankings matched exactly in 118 patients (72%) or differed by a factor of no more than 1 (26%) in 161 (98%) patients.

The SOD score can be used as both a surgeon-based and a patient-based outcome score, given the high level of agreement. Given its brevity, ease of understanding, and high interrater reliability, the SOD has the potential to be used across multiple specialties to rate outcomes.