Table 4.

Factors associated with a missed diagnosis of stroke (based on prior ED treat-and-release visits for dizziness or headache 30 days before stroke admission) among all patients^a aged > = 18 years admitted to inpatient care for stroke: generalized estimating equation (GEE) results, 2009.

Data element	Value	EST	SE	Z	p	OR	LCL	UCL
Patient characteristics
Sex	0: Male	–0.29	0.05	–6.16	<0.001	0.75	0.68	0.82
	1: Female
Age group, years	45–64	–0.85	0.07	–12.99	<0.001	0.43	0.38	0.49
	65–74	–1.26	0.11	–11.58	<0.001	0.28	0.23	0.35
	75 and over	–1.68	0.12	–14.56	<0.001	0.19	0.15	0.23
	18–44
Race/ethnicity	Black	0.17	0.07	2.39	0.02	1.18	1.03	1.35
	Hispanic	0.27	0.08	3.27	<0.001	1.30	1.11	1.53
	Asian/Pacific Islander	0.25	0.11	2.30	0.02	1.29	1.04	1.60
	White
Payer	Medicare	–0.41	0.10	–4.30	<0.001	0.66	0.55	0.80
	Medicaid	–0.36	0.09	–3.89	<0.001	0.70	0.58	0.84
	Uninsured	0.01	0.09	0.11	0.91	1.01	0.85	1.20
	Other	–0.46	0.15	–2.98	0.003	0.63	0.47	0.85
	Private insurance
Income quartile	Lowest income < 40K	0.06	0.08	0.75	0.45	1.06	0.90	1.25
	Low income 40K to < 50K	0.05	0.08	0.65	0.52	1.05	0.90	1.23
	Moderate income 50K– < 66K Highest income ≥ 66K	0.05	0.08	0.61	0.54	1.05	0.90	1.23
Comorbidities	Range: 0 – 13	–0.07	0.02	–4.00	<0.001	n/a	n/a	n/a
Hospital characteristics
Region	Midwest	–0.17	0.15	–1.1	0.27	0.84	0.62	1.14
	South	–0.12	0.13	–0.93	0.35	0.88	0.68	1.14
	West	–0.03	0.12	–0.24	0.81	0.97	0.77	1.22
	Northeast
Population size	Small metropolitan	–0.26	0.09	–2.94	0.003	0.77	0.65	0.92
	Micropolitan	0.21	0.15	1.41	0.16	1.23	0.92	1.64
	Rural	0.06	0.24	0.24	0.81	1.06	0.67	1.68
	Large metropolitan
Ownership	Public	–0.01	0.12	–0.10	0.92	0.99	0.78	1.25
	Private, for–profit	–0.22	0.12	–1.93	0.05	0.80	0.64	1.00
	Private, not-for-profit
Teaching Status	Nonteaching	0.37	0.11	3.24	<0.001	1.45	1.16	1.82
	Teaching
Hospital workflow (annual average)
Inpatient occupancy rate (annual)	Low ≤ 0.5	0.00	0.13	0.03	0.98	1.00	0.78	1.29
	Moderate > 0.5, < 0.7	0.11	0.11	0.94	0.35	1.11	0.89	1.39
	High ≥ 0.7
ED volume (annual)	Low ≤ 29,124	0.45	0.17	2.69	0.007	1.57	1.13	2.18
	Moderate 29,125–64,434	0.10	0.10	1.02	0.31	1.11	0.91	1.36
	High ≥ 64,435
Percent admitted from ED (annual)	Low ≤ 11.82%	0.44	0.15	2.88	.004	1.55	1.15	2.09
	Moderate > 11.82%, < 19.46%	0.21	0.11	1.95	0.05	1.24	1.00	1.54
	High ≥ 19.46%
ED visit characteristics (day of initial treat-and-release visit)
Weekend	Monday–Friday	0.11	0.05	2.09	0.04	1.11	1.01	1.23
	Saturday–Sunday
ED crowding on day of visit (percentile)	0–20th percentile	–0.02	0.07	–0.33	0.75	0.98	0.84	1.13
	21 – 40th percentile	0.04	0.07	0.5	0.62	1.04	0.90	1.19
	41 – 60th percentile	0.04	0.07	0.52	0.60	1.04	0.90	1.20
	61 – 80th percentile	0.08	0.07	1.18	0.24	1.08	0.95	1.23
	61 – 80th percentile
ED admit rate on day of visit (percentile)	0 – 20th percentile	1.85	0.16	11.72	<0.001	6.34	4.66	8.63
	21 – 40th percentile	0.91	0.11	8.03	<0.001	2.48	1.99	3.10
	41 – 60th percentile	0.61	0.10	6.05	<0.001	1.85	1.51	2.25
	61 – 80th percentile	0.34	0.08	4.04	<0.001	1.40	1.19	1.66
	81 – 100th percentile
Patient left against advice	Against medical advice	1.08	0.14	7.50	<0.001	2.94	2.22	3.89
	Not against advice

187,188 of 198,819 trials used; number of events used = 2088 of 2243 (records with missing data excluded); exchangeable correlation structure (working correlation = 0.002); 1016 clusters (facilities). EST, estimate; SE, standard error; Z, Z score; p, probability level; OR, odds ratio; LCL, lower confidence limit; UCL, upper confidence limit.

^aThis is a patient-level analysis of inpatient stroke admissions, with and without a prior treat-and-release ED visit for dizziness or headache within 30 days of the stroke admission; only a single ‘initial’ ED visit (the most proximate to the ‘index’ stroke admission) is considered.