Table 2.

Association of insomnia with COPD-related healthcare utilization and costs in 12 months after index date

			Unadjusted analysis		Adjusted analysis
	Insomnia (n = 407,969)	No Insomnia (n = 603,677)	Difference^a	95% CI	IRR^b	95% CI
Healthcare use, mean ± SD
Outpatient visits	4.81 ± 8.64	3.78 ± 6.22	1.03**	1.00–1.06	1.17**	1.16–1.18
Hospitalizations	1.57 ± 1.15	1.46 ± 0.99	0.11**	0.10–0.12	1.02*	1.02–1.03
Hospitalization LOS (Days)^c	90.70 ± 114.60	65.51 ± 93.40	25.19**	22.67–27.71	1.26**	1.22–1.31
Prescription fills for steroids and/or antibiotics	4.80 ± 6.08	3.64 ± 4.90	1.16**	1.05–1.26	1.15**	1.13–1.18
Healthcare costs, median ± IQR					Difference at Median ^d	95% CI
Outpatient visit-related costs ($)	885.13 ± 1,276.18	806.54 ± 1,015.48	78.59**	75.74–81.44	3.46	-1.43–8.34
Hospitalization-related costs ($)	79,428.00 ± 209,460.00	69,068.00 ± 159,719.00	10,360.00**	6,936.54–13,783.00	10,344.51**	6,715.18–13,973.84

Definition of abbreviations: CI = confidence interval; COPD = chronic obstructive pulmonary disease; IRR = incidence risk ratio; LOS = length of stay

^a Calculated as the difference between patients with insomnia and patients without insomnia. Significance determined by Wilcoxon rank-sum test

^b Negative binomial regression model adjusted for age, sex, race, marital status, current smoker, service connection, body mass index, and comorbid conditions

^c Maximum 12-month inpatient length of stay set to 365 days

^d Quantile regression model adjusted for age, sex, race, marital status, current smoker, service connection, body mass index, and comorbid conditions. Wilcoxon rank test of differences in cost across percentile coefficients for insomnia

*p < 0.05, **p < 0.001