Table 3.

Unadjusted and adjusted estimates and 95% confidence intervals (CIs) of the main effects of cohort and time period for each type of healthcare use

Model	Effect	Category	Estimate	Significance 95% CI	QIC
Hospitalization
Adjusted model	Time period	Year in housing	0.79	0.70, 0.89	6711.98
		Year before	Ref	–
	Cohort	Public housing	1.39	1.21, 1.61
		Matched	Ref	–
Unadjusted model	Time period	Year in housing	0.82	0.74, 0.90	7661.38
		Year before	Ref	–
	Cohort	Public housing	1.66	1.45, 1.89
		Matched	Ref	–
GP visits
Adjusted model	Time period	Year in housing	1.04	1.01, 1.06	− 120,990.85
		Year before	Ref	–
	Cohort	Public housing	1.08	1.04, 1.13
		Matched	Ref	–
Unadjusted model	Time period	Year in housing	0.95	0.93, 0.97	− 129,531.57
		Year before	Ref	–
	Cohort	Public housing	1.25	1.18, 1.31
		Matched	Ref	–
Specialist visits
Adjusted model	Time period	Year in housing	1.00	0.94, 1.06	− 3265.49
		Year before	Ref	–
	Cohort	Public housing	0.98	0.88, 1.09
		Matched	Ref	–
Unadjusted model	Time period	Year in housing	0.91	0.87, 0.96	− 3078.40
		Year before	Ref	–
	Cohort	Public housing	1.11	1.01, 1.23
		Matched	Ref	–
Prescriptions
Adjusted model	Time period	Year in housing	1.04	1.02, 1.05	− 132,628.77
		Year before	Ref	–
	Cohort	Public housing	1.09	1.05, 1.13
		Matched	Ref	–
Unadjusted model	Time period	Year in housing	1.00	0.99, 1.02	− 155,558.88
		Year before	Ref	–
	Cohort	Public housing	1.27	1.21, 1.33
		Matched	Ref	–
ED visits
Adjusted model	Time period	Year in housing	0.90	0.82, 1.00	3998.54
		Year before	Ref	–
	Cohort	Public housing	1.18	1.01, 1.37
		Matched	Ref	–
Unadjusted model	Time period	Year in housing	0.90	0.82, 1.00	3180.39
		Year before	Ref	–
	Cohort	Public housing	1.59	1.36, 1.86
		Matched	Ref	–

Bold values denote significant results. Adjusted model covariates include residential mobility, income quintile, injury, diabetes, arthritis, hypertension, respiratory disease, cancer, affective disorders, schizophrenia, substance abuse disorders, and ADGs