Variation in outpatient emergency department utilization in Texas Medicaid: a state-level framework for finding “superutilizers”

Table 2 Adult emergency department (ED) visits in index (2012) and subsequent years (2013, 2014) in Texas Medicaid

	Number of ED visits in index year (2012)^a
			Range of potential overutilization
							EF^b
	1	2	3–4	5–6	7–9	10–14	15+	All
Annualized ED visits, index year	2.43	3.71	5.03	7.07	9.69	13.78	27.47	3.94
Annualized ED visits, 1 year after	0.91	1.42	2.21	3.49	5.28	8.03	16.36	1.72
Annualized ED visits, 2 years after	0.74	1.15	1.73	2.78	4.00	5.94	11.46	1.35
Number of ED providers, index year^c	1.00	1.34	1.65	2.01	2.39	2.88	4.26	1.32
Primary diagnosis, visits in index year^d (%)
Chronic condition	3.79	4.04	4.39	5.05	5.76	6.49	9.73	5.02
Substance use disorder	0.62	0.61	0.71	0.76	1.02	1.14	1.33	0.79
Mental illness	2.00	2.12	2.37	2.66	3.16	3.55	4.15	2.59
Diagnoses 1–3, visits in index year^e (%)
Substance use disorder	7.03	7.43	8.17	8.97	9.82	10.47	10.73	8.41
Mental illness	5.33	5.93	6.87	7.99	9.25	10.71	11.74	7.40
NYU algorithm preventable events^f (%)
Injury	14.07	13.02	12.60	12.51	12.92	13.04	12.16	13.01
Non-emergent	24.92	25.28	26.00	26.00	25.29	25.09	22.21	25.16
Emergent, primary care treatable	22.53	23.04	22.79	22.64	22.38	22.18	23.14	22.72
Emergent, preventable, or avoidable	5.12	5.34	5.62	6.02	6.21	6.12	5.94	5.62
Emergent, not preventable, or avoidable	12.66	12.63	12.18	12.01	12.13	12.65	15.53	12.67
Frequent users, index year and 2 years after (%)
3+ visits each year	NA	NA	10.98	21.68	34.07	47.04	60.24	5.01
5+ visits each year	NA	NA	NA	9.29	19.68	32.42	50.47	1.91
Enrolled in Medicaid in the subsequent 2 years	51.00	57.34	63.03	68.89	72.25	75.99	76.05	56.42

Authors’ analysis of Texas Medicaid claims, encounter, and enrollment data. This analysis excludes dual-eligible enrollees
NA not applicable
^aThe index year is set to 2012 to provide a 2-year prospective window
^bExtremely frequent ED users
^cCalculated using the total number of unique national provide identifier (NPI) numbers; unique NPIs, however, may be attributed to facilities or physicians
^dThe ICD-9-CM codes used to define chronic condition, SUD, and mental illness were provided by Billings and Raven [7]
^eIncludes primary, secondary, and tertiary diagnoses
^fThe NYU ED profiling algorithm is used here

ISSN: 1865-1380