A statistical analysis of pain relief after surgical operations.

OBJECTIVES Administering pain relief measures requires information about development of post-operative pain. Such information can be obtained from studying the patients' perception of pain. The aim of this study is to develop a statistical model that predicts the level of pain in different periods after the operation and the time effects of analgesics. The results can be used for an effective and timely application of pain medications. METHODS This paper applies an ordered response model to a sample of 392 observations from 49 patients undergoing orthopedic surgeries. The patients' subjective pain levels have been recorded for several intervals up to 24h after their respective operations. The adopted statistical model accounts for the unobserved heterogeneity among patients through random coefficients. Such heterogeneity could be due to differences in patients' perception of pain as well as their health status and sensitivity. RESULTS The analysis indicates that post-operative pain gradually increases over time but with a slightly diminishing rate. The results suggest that analgesics are quite effective in containing the development of pain. However, the analgesic effects manifest gradually, at a rate more or less similar to that of post-operative pain. CONCLUSION Highlighting the importance of pain from both clinical and economic perspectives, the analysis indicates that a timely application of analgesics is crucial for an effective pain relief after surgical operations. In particular, the results indicate that the optimal time of administering analgesics is immediately after the operation. The main policy implication of this finding is that preemptive measures of pain relief could be most effective and should be favored to similar analgesic treatments after the manifestation of pain. The results also show that the post-operative pain risks differ significantly across individual cases, which suggests that case-specific clinical assessments supersede any statistical analysis. The proposed statistical model can be used to identify high-risk categories and explore the variations across different analgesic types.