Association between chiropractic care and use of prescription opioids among older medicare beneficiaries with spinal pain: a retrospective observational study

Table 3 Adjusted risk of opioid prescription fill among recipients and subgroups vs. non-recipients

Cohort: n (%)	Fill: n (%)	No Fill: n (%)	Hazard Ratio	95% CI
Total	33,031 (59)	22,918 (41)	–	–
Non-recipients 46,593 (83)	29,371 (52)	17,222 (31)	(referent)	–
Recipients 9356 (17)	3660 (7)	5696 (10)
Unadjusted model			0.46	0.44–0.47
Adjusted for state; without propensity score			0.49	0.47–0.50
Full model: with propensity score, without state			0.39	0.36–0.43
Full model with both state and propensity score			0.44	0.40–0.49
Recipients, Early 8,161 (15)	2,938 (5)	5,223 (9)	0.38	0.34–0.42
Recipients, Delayed 937 (2)	564 (1)	373 (1)	0.90	0.80–1.03
Recipients, Late 258 (< 1)	158 (< 1)	100 (< 1)	0.93	0.77–1.13

Results of Cox proportional hazards modeling for isk of opioid prescription fill for up to 365 days following index. Full model controlled for for patient age, sex, race/ethnicity, state of residence, spinal pain diagnosis category, and Charlson comorbidity score. Non-recipients = subjects who received Primary Care only for spinal pain; Recipients = subjects who received both Primary Care and Chiropractic Care for spinal pain; % = percentage of total study sample; HR = Hazard Ratio; The hazard ratios quantify risk of receiving a prescription opioid within 365 days of initial visit. A hazard ratio of 1.0 signifies equal risk between groups: as the number decreases from 1.0 it signifies decreased risk of filling an opioid prescription. LCL = lower confidence limit; UCL = upper confidence limit; Early = subjects who received chiropractic care within 30 days following diagnosis of spinal disorder; Delayed = subjects who received chiropractic care at 31–90 days following diagnosis of spinal disorder; Late = subjects who received chiropractic care at 91–120 days following diagnosis of spinal disorder

ISSN: 2045-709X