Physical Therapy and Hospitalization Among Medicare Beneficiaries With Low Back Pain: A Retrospective Cohort Study.

STUDY DESIGN

A retrospective cohort study.

OBJECTIVE

The aim of this study was to evaluate associations between receipt and quantity of outpatient physical therapy (PT) during an episode of care and 30-day and 180-day hospital admissions for any condition and lumbar spine conditions.

SUMMARY OF BACKGROUND DATA

Low back pain (LBP) is a common cause of hospitalization and the most common reason Medicare beneficiaries utilize outpatient PT. The association between PT and hospitalization among patients with LBP is unknown.

METHODS

A national sample of Medicare Fee-for-Service claims included 413,608 beneficiaries with an International Classification of Disease 9th revision (ICD-9) code of LBP and 1,415,037 episodes of care between June 1, 2010, and June 30, 2011. Episodes were classified as PT episodes or non-PT episodes. Relative risk of hospitalization from the episode start date was caldulated, adjusting for health status (Charlson comorbidity index), prior care utilization (number of prior hospitalizations and total number of episodes), an indicator of LBP severity (number of LBP ICD-9 codes), and demographics (sex, race/ethnicity, age).

RESULTS

The proportion of 30-day hospitalization for any condition was 3.42% for PT episodes of care and 6.54% for non-PT episodes. For 180-day hospitalization, proportions were 15.45% (PT) and 21.65% (non-PT). The adjusted relative risk reduction of PT (vs. non-PT) was 41% for 30 days [99% confidence interval (CI) 38-44] and 22% for 180 days (20-24). For admitting diagnoses of lumbar spine, reductions were 65% at 30 days and 32% at 180 days. More PT treatment days showed greater 30-day risk reductions. For any condition, compared with non-PT, reductions were 24% for 1 to 2 treatment days (lowest tertile), 45% for 3 to 7 days, and 65% for more than 8 days (highest tertile). Stronger effects were found for lumbar spine admissions. Associations between PT quantity and 180-day hospitalization were less consistent. Limitations of Medicare claims include the potential for inaccuracies, limited knowledge about disease severity, and which PT interventions were conducted.

CONCLUSION

Receipt of PT during an episode had a 22% to 65% reduced relative risk of hospitalization, with greater short-term reductions for more PT treatment days.

LEVEL OF EVIDENCE