Measuring health in patients with cervical and lumbosacral spinal disorders: is the 12-item short-form health survey a valid alternative for the 36-item short-form health survey?

OBJECTIVES: To determine the convergent validity of the 12-Item Short-Form Health Survey, version 2 (SF-12v2), with 36-Item Short-Form Health Survey, version 2 (SF-36v2), in patients with spinal disorders, and to determine other key factors that might further explain the variances between the 2 surveys.
DESIGN: Cross-sectional study.
SETTING: Orthopedic ambulatory care. PARTICIPANTS: Eligible participants (N=98; 24 with cervical, 74 with lumbosacral disorders) who were aged 18 years and older, scheduled to undergo spinal surgery, and completed the SF-36v2.
INTERVENTIONS: Not applicable. MAIN OUTCOME MEASURES: SF-36v2 and SF-12v2 (extracted from the SF-36v2).
RESULTS: The 2 summary scores, physical and mental component scores (r range, .88-.97), and most of the scale scores (r range, .81-.99) correlated strongly between the SF-12v2 and SF-36v2, except for the general health score (cervical group, r=.69; lumbosacral group, r=.76). Stepwise linear regression analyses showed the SF-12v2 general health scores (cervical: beta=.61, P<.001; lumbosacral: beta=.68, P<.001) and the level of comorbidities (cervical: beta=-.37, P=.014; lumbosacral: beta=-.18, P=.039) were significant predictors of the SF-36v2 general health score in both groups, whereas age (beta=.32, P<.001) and smoking history (beta=-.22, P=.005) were additional predictors in the lumbosacral group.
CONCLUSIONS: SF-12v2 is a practical and valid alternative for the SF-36v2 in measuring health of patients with cervical or lumbosacral spinal disorders. The validity of the SF-12v2 general health score interpretation is further improved when the level of comorbidities, age, and smoking history are taken into consideration.