The development of a logistic model for carpal tunnel syndrome risk assessment

Carpal tunnel syndrome (CTS), a specific cumulative trauma disorder involving hand dysfunction caused by compression of the median nerve at the wrist, is developed by personal factors, psychosocial factors, physical factors, or a combination of these. The purposes of this study were to: (1) provide a screening tool for CTS by examining the relationship between clinical symptom severity scales and nerve conduction studies on CTS, (2) develop a reliable risk assessment instrument specifically for CTS, and (3) develop a statistical risk assessment model containing the three risk factors, and (4) probe an effect of carpal tunnel release compared to non-surgical treatment methods; In this cross-sectional study, a group of 39 CTS patients and a group of 35 healthy workers from two manufacturing companies and a university library participated in the study. Clinical symptom severity and nerve conduction measures were correlated for 60 hands with CTS in the CTS group. The secondary symptoms (Pain, Clumsiness, and Weakness) and nocturnal symptoms of the primary condition did not show any statistical significance with the nerve conduction measures at α = .05 while numbness and tingling of the primary symptoms were highly correlated (p < .05) with the measures. The nocturnal symptoms could be more of a mechanical problem of wrist flexion at night and not a reflection of nerve slowing at night versus daytime but still was a good measure of CTS. The primary symptoms could be used as a potential screening tool for early CTS in the workplace. Post-surgery effect was examined with the same questionnaire in the CTS patients. A total of 23 hands had carpal tunnel release (CTR) and showed significant improvement in severity symptoms (p < .001 for all symptoms) whereas the rest treated by the conservative methods did not show much improvement (p > .20 for all symptoms).; Next, psychosocial and physical risk exposure information was analyzed between the two groups. The pseudo-univariate analysis was used to select competitive candidate factors from a total of 87 variables. From the selected factors, a multiple logistic regression model was developed including 7 factors: age, gender, right hand wrist ratio, body mass index (BMI), medical history of musculoskeletal disorders at hand/wrist, skill utilization, and unbalanced hand use on highly repetitive motions. The Hosmer-Lemeshow test shows H_c* = 1.536 (d.f. = 7, p = 0.981), indicating that the derived logistic regression equation adequately fits the data. The correct classification performance of the model is 91%.; The logistic regression model includes more personal factors than work-related factors. It's mainly due to two major reasons: (1) heavier physical workload of the healthy workers' than CTS patients', i.e., manual workers vs. office workers, or (2) higher survival rate of healthy workers, i.e., the stronger keeps the job and the weaker leaves the job. (Abstract shortened by UMI.)...