Physician-Related Variability in the Outcomes of an Invasive Treatment for Neck and Back Pain: A Multi-Level Analysis of Data Gathered in Routine Clinical Practice
Ana Royuela Vicente | Francisco M. Kovacs | Jesús Seco-Calvo | Borja M. Fernández-Félix | Víctor Abraira | Javier Zamora
Neuro-reflexotherapy (NRT) is a proven effective, invasive treatment for neck and back pain. To assess physician-related variability in results, data from post-implementation surveillance of 9023 patients treated within the Spanish National Health Service by 12 physicians were analyzed. Separate multi-level logistic regression models were developed for spinal pain (SP), referred pain (RP), and disability. The models included all patient-related variables predicting response to NRT and physician-related variables. The Intraclass Correlation Coefficient (ICC) and the Median Odds Ratio (MOR) were calculated. Adjusted MOR (95% CI) was 1.70 (1.47: 2.09) for SP, 1.60 (1.38: 1.99) for RP, and 1.65 (1.42: 2.03) for disability. Adjusted ICC (95%CI) values were 0.08 (0.05: 0.15) for SP, 0.07 (0.03: 0.14) for RP, and 0.08 (0.04: 0.14) for disability. In the sensitivity analysis, in which the 6920 patients treated during the physicians’ training period were excluded, adjusted MOR was 1.38 (1.17: 1.98) for SP, 1.37 (1.12: 2.31) for RP, and 1.25 (1.09: 1.79) for disability, while ICCs were 0.03 (0.01: 0.14) for SP, 0.03 (0.00: 0.19) for RP, and 0.02 (0.00: 0.10) for disability. In conclusion, the variability in results obtained by different NRT-certified specialists is reasonable. This suggests that current training standards are appropriate.
显示更多 [+] 显示较少 [-]