Use of Outpatient Rehabilitation Among Adult Stroke Survivors — 20 States and the District of Columbia, 2013, and Four States, 2015

Carma Ayala, PhD; Jing Fang, MD; Cecily Luncheon, MD, DrPH; Sallyann Coleman King, MD; Tiffany Chang, MPH; Matthew Ritchey, DPT; Fleetwood Loustalot, PhD

Disclosures

Morbidity and Mortality Weekly Report. 2018;67(20):575-578.

Discussion

Overall, approximately one third of stroke survivors reported participating in stroke outpatient rehab. Although outpatient rehab use increased significantly in the four states that collected data in both 2013 and 2015, it remained suboptimal,^[3] highlighting missed opportunities to reach stroke survivors. Stroke recovery can be a long and complex process, involving multiple domains of therapy (e.g., physical, occupational, communication, and cognitive) and occurs in inpatient rehabilitation facilities, skilled nursing facilities, and outpatient rehabilitation facilities. Benefits of stroke outpatient rehab have been determined to improve patient functional status, survival, cardiovascular risk profiles, and quality of life and reduce risks for recurrent strokes and psychological or stress disorders.^[3,4,8,9] Generally, stroke outpatient rehab participation is underutilized,^[3,8] which this study found to be true for all subgroups and states included in the analysis. No subgroup had outpatient rehab use rates >40%, and no state had use rates >50%. Although the overall prevalence of outpatient rehab use was low, disparities in use were evident. Younger adults, women, non-Hispanic persons of other than black or white races, Hispanics, and adults with less than a high school education were less likely to use stroke outpatient rehab than their counterparts. Disparities in stroke outpatient rehab at the state level were also apparent. For example, adjusted outpatient rehab use prevalence in Minnesota (43.6%) was almost twice that in Oregon (23.1%).

Increasing participation in stroke outpatient rehab has been recognized as a national priority. Healthy People 2020** aims to increase the proportion of adult stroke survivors who are appropriately and effectively assessed and referred for rehabilitation services. The estimates from the Healthy People 2020 objective are high (90% during 2008–2011); however, they are reflective of assessment or referral, not participation.^[4] Improving coordination of care to support assessment, referral, and, ultimately, participation in rehab is needed. The continued underutilization of outpatient stroke rehab might be related to lack of patient access to outpatient facilities, ineffective referral from health care providers, high out-of-pocket costs, lack of health insurance coverage, or lack of knowledge and awareness of benefits of outpatient rehab for stroke survivors.^[4,6] The CDC-supported Paul Coverdell National Acute Stroke Program^†† seeks to better understand the care provided to stroke survivors to identify disparities and support quality improvement around the assessment for, effective referral to, and provision of stroke rehab services. Experiences from such programs can support system-level changes that encourage use of stroke rehab services across all subgroups and geographies.

The findings in this report are subject to at least five limitations. First, BRFSS data are self-reported and subject to recall bias. Moreover, recall bias might lead to participants inaccurately reporting the type of stroke rehab they used (i.e., outpatient rehab versus inpatient rehabilitation facilities, skilled nursing facilities, and home health rehab).^[10] Second, the survey does not capture stroke severity, variations in rehabilitation needs, or information about why participants did not participate in outpatient rehab. Third, the optional module was only used by selected states, and the findings should not be considered as nationally representative. Fourth, with few respondents reporting a history of stroke (162 in the District of Columbia to 1,618 in Florida), some state-level confidence intervals were wide, and results should be interpreted with caution. Finally, only participation in outpatient rehab was included in the module, limiting the ability to assess participation in other rehab modalities.

Although estimates of stroke outpatient rehab referral might be high, participation in stroke outpatient rehab remains suboptimal. Barriers to participation in stroke outpatient rehab are evident,^[3,8–10] but focused attention on system-level interventions that ensure participation is needed, especially among populations with lower levels of participation. Interventions that might improve outpatient rehab participation include increasing coverage for outpatient rehab services by health insurers, reducing copayments, extending rehab clinic hours to improve access availability, and implementing standardized assessments by health care professionals to guide appropriate referrals to outpatient rehab at hospital discharge.^[3–5,8] Stroke survivors should be educated about stroke outpatient rehab opportunities possibly available in their community that reduce barriers related to transportation and time (e.g., telehealth, mobileHealth, and home-based care).^{[3,5,6,8–10]}

1 2

Next Section

**Healthy People 2020 Heart Disease and Stroke Objectives (HDS-23). https://www.healthypeople.gov/node/4588/data_details.
^††The Coverdell program works with health systems across funded recipient states to gather data and drive quality improvement in the prehospital, in-hospital, and posthospital care settings. https://www.cdc.gov/dhdsp/programs/stroke_registry.htm.

Table 1. Unadjusted and adjusted* percentages of adults who survived a stroke and received outpatient stroke rehabilitation, by demographic characteristics and presence of cardiovascular disease risk factors — Behavioral Risk Factor Surveillance System (BRFSS), 20 U.S. states and the District of Columbia, 2013
Characteristic	Sample size	Unadjusted % (95% CI)	P-value^†	Adjusted* % (95% CI)	P-value
Total	6,743	31.2 (29.1–33.4)	—	31.2 (29.1–33.4)	<0.001
Sex
Men	2,616	33.7 (30.4–37.2)	0.038	33.8 (30.5–37.2)	0.030
Women	4,127	29.1 (26.5–31.9)	0.038	29.1 (26.4–31.8)	0.030
Age group (yrs)
18–64	2,468	30.9 (27.6–34.4)	0.798	30.6 (27.5–34.0)	0.622
≥65	4,275	31.5 (29.0–34.1)	0.798	31.7 (29.0–34.6)	0.622
Race/Ethnicity
White, non-Hispanic	5,132	30.3 (28.0–32.7)	0.019	30.0 (27.6–32.5)	0.013
Black, non-Hispanic	966	39.1 (32.9–45.7)		39.8 (33.6–46.2)
Hispanic	481	26.5 (20.0–34.1)		26.7 (20.3–34.3)
Other, non-Hispanic	164	24.4 (15.9–35.6)	—^§	25.8 (17.0–37.0)	—^§
Education
Less than high school	1,141	25.8 (21.5–30.6)	0.025	25.7 (21.4–30.5)	0.022
High school	2,255	32.5 (28.9–36.2)		32.3 (28.8–36.2)
Some college	1,923	31.9 (28.0–36.1)		32.1 (28.3–36.2)
College or higher	1,424	36.3 (31.5–41.2)		36.4 (31.5–41.5)
Insurance
Yes	6,276	31.6 (29.5–33.7)	0.492	31.3 (29.2–33.5)	0.836
No	467	28.2 (20.2–38.0)	0.492	30.3 (21.9–40.2)	0.836
CVD risk factors (no.)^¶
0	455	35.6 (27.0–45.3)	0.382	35.6 (27.4–40.2)	0.478
1	1,349	29.2 (24.7–34.2)		29.4 (24.9–44.7)
2	2,103	30.1 (26.5–33.9)		30.3 (26.7–34.4)
3	1,805	31.5 (28.0–35.2)		31.2 (27.7–34.9)
4	918	34.8 (28.7–41.5)		34.5 (28.5–41.1)
5	113	21.8 (12.6–35.1)	—^§	22.9 (13.3–36.5)	—^§

Abbreviations: CI = confidence interval; CVD = cardiovascular disease.
*Adjusted for age, sex, race/ethnicity, education, insurance status, and CVD risk.
^†P-values were obtained using Wald F test to identify statistically significant differences among subgroup.
^§BRFSS recommends that data be suppressed when relative standard error (RSE) is >30% or denominator <50; it is also suggested that if RSE is 20%–30%, the estimates are potentially unreliable.
^¶Selected self-reported CVD risk factors include hypertension, high blood cholesterol, diabetes, obesity, and current smoking. Categories were assigned based on the number of risk factors present: 0, 1, 2, 3, 4, or 5.

Table 2. Crude and adjusted percentages* of adults who survived a stroke and received stroke outpatient rehabilitation, by state and ascending adjusted percentage — Behavioral Risk Factor Surveillance System, 20 U.S. states and the District of Columbia (DC), 2013, and four U.S. states, 2015
Year/State^†	No.	Unadjusted % (95% CI)	P–value^§	Adjusted* % (95% CI)	P–value^§
2013
Total (20 states and DC)	6,743	31.2 (29.1–33.4)	0.003	31.2 (29.1–33.4)	0.012
Oregon	202	22.7 (16.8–29.9)		23.1 (17.1–30.5)
Georgia	306	24.8 (19.3–31.2)		23.7 (18.5–29.9)
Oklahoma	197	23.8 (17.0–32.3)		24.6 (17.3–33.5)
Hawaii	210	24.1 (17.4–32.3)		25.9 (17.8–36.0)
North Dakota	230	26.0 (19.5–33.8)		27.0 (20.3–34.9)
Maine	139	27.2 (19.4–36.6)		28.1 (20.2–37.6)
Tennessee	254	27.7 (20.8–35.9)		28.1 (21.0–36.5)
Arkansas	278	27.9 (20.9–36.2)		29.1 (22.3–37.0)
Florida	1,618	30.6 (25.8–35.8)		30.0 (25.4–35.0)
Washington	378	30.6 (24.5–37.5)		30.8 (24.5–37.9)
Arizona	170	30.7 (20.9–42.6)		32.1 (22.2–43.9)
Wisconsin	166	31.1 (21.2–42.9)		32.4 (22.3–44.4)
District of Columbia	162	39.0 (28.5–50.8)		32.7 (22.5–44.8)
Missouri	319	31.8 (24.8–39.7)		32.8 (25.8–40.7)
North Carolina	204	32.9 (24.5–42.6)		33.4 (24.8–43.2)
Mississippi	426	35.5 (28.4–43.2)		34.0 (26.8–42.0)
South Carolina	441	37.4 (31.1–44.1)		35.8 (29.6–42.5)
Massachusetts	112	38.9 (24.1–56.0)	—^¶	37.8 (24.8–52.9)	—^¶
Nebraska	291	38.9 (30.6–47.8)		39.6 (30.7–49.3)
Iowa	263	40.6 (33.4–48.1)		41.7 (34.5–49.4)
Minnesota	377	43.0 (31.5–55.2)		43.6 (31.0–57.1)
2013
Total (four states)	910	27.2 (23.5–31.3)	0.001	27.4 (23.5–31.3)	0.0004
Oregon	202	22.7 (16.8–29.9)		22.7 (16.2–30.8)
Georgia	306	24.8 (19.3–31.2)		24.2 (18.8–30.6)
Maine	139	27.2 (19.4–36.6)		28.4 (20.3–38.2)
Iowa	263	40.6 (33.4–48.1)		41.7 (34.4–49.4)
2015
Total (four states)	729	35.5 (29.6–41.8)	0.033	35.5 (29.6–41.8)	0.008
Maine	182	28.0 (21.0–36.3)		31.3 (23.1–41.0)
Georgia	201	34.0 (25.4–43.8)		31.8 (24.2–40.6)
Oregon	180	36.2 (26.8–46.7)		39.7 (29.6–50.8)
Iowa	166	46.1 (37.2–55.3)		49.8 (40.3–59.3)

Abbreviation: CI = confidence interval.
*Adjusted for age, sex, race/ethnicity, education, insurance status and cardiovascular disease risk factors.
^†States are listed in ascending order of adjusted percentages for outpatient stroke rehabilitation in 2013 and 2015.
^§P-values were obtained using Wald F test to identify statistically significant differences among states within each year; comparison of differences between 2013 and 2015 among four states only was p = 0.0289.