[Predictability of Postoperative Pain and Satisfaction. How Precise Are We and What Is the Role of Professional Experience?]

Background The indication for surgery is justified by an expected improvement for the patient. To evaluate the probability and extent of individual postoperative patient benefit, the surgeon needs to elaborate numerous parameters of potential relevance for the outcome beyond his key competence, that is the technical dimension of the operation. Despite the highest medical standards, individual postoperative satisfaction with surgery is highly variable, even in cases with a technically good result. The aim of the present study was to investigate the individual predictability of postoperative pain and satisfaction in patients with elective musculoskeletal surgery. Moreover, it was analysed whether the quality of the prediction of the outcome depends on professional experience and if a better prediction can be obtained in such a highly standardised procedure as primary total arthroplasty. Patients/Material and Methods In our hospital on the day before surgery, patients with their medical history are presented to the head of department and a short clinical examination is performed as well as a joint analysis of radiographic images. During this grand round, doctors gave a written preoperative estimation of both expected postoperative satisfaction and pain at 6 months after surgery on a scale from 0 - 10. At 6 months postoperatively, patients were asked to give their actual level for these two parameters. Preoperative estimations were obtained from both senior and resident physicians and compared with the values actually reported by the patient. Results A total of 194 physicians' predictions of 63 patients were analysed. Preoperative pain levels were reduced markedly by surgery from a median NRS of 6.25 to 2.5 (p < 0.001). Median prediction for postoperative pain was 2 (IQR 2), which is only slightly more optimistic than reported. On an individual level, the discrepancy between prediction and actual outcome was, however, considerable - with a median absolute difference of 2 (IQR 3) values on the scale. This means that only 50% of all predictions were close enough to the actual value to be at least in the correct half of the entire scale. When looking at prediction precision as a function of professional experience, no difference could be observed between senior and junior doctors (p = 0.738 for postoperative pain and p = 0.370 for satisfaction with surgery). Even in primary arthroplasty patients (n = 17), precision of outcome prediction for pain was no better that in the remaining collective (p = 0.634). With respect to postoperative satisfaction, precision of prediction was even worse (p = 0.042), as satisfaction was slightly underestimated by the physicians in the primary arthroplasty group, by a median of 1. Conclusion While general prediction showed almost perfect agreement with actual postoperative values, individual predictability showed highly variable results. Even in such a standardised collective as primary arthroplasty, this scattering of deviation was observed. Since professional experience did not lead to improved results, it can be speculated that, beside the technical dimension of surgery, other factors such as patient expectation are of crucial relevance for postoperative outcome. To further improve outcome and patient satisfaction with surgery we therefore recommend developing an individualised and realistic prognosis together with each patient, but bearing in mind own limits of outcome prediction.