Treatments and Health Outcomes of Medicare Patients With Back Pain.

ABSTRACT

Back pain treatments are costly and frequently involve use of procedures that may have minimal benefit on improving patients' functional status. Two recent studies evaluated adverse outcomes (mortality and major medical complications) following receipt of spinal surgery but neither examined whether such treatments affected functional ability. Using a sample composed of Medicare patients with persistent back pain, we examined whether functional ability improved after treatment, comparing patients treated with back surgery or spinal injections to nonrecipients. We analyzed four binary variables that measure whether the ability to perform routine tasks improved. We used instrumental variables analysis to address the nonrandom selection of treatment received due to unobservable confounding. Contrary to the observational results, the instrumental variable estimates suggest that receipt of either back surgery or spinal injections does not improve back patients' functional ability. Failure to account for selection into treatment can lead to overestimating the benefits of specific treatments.