Comprehensive Post-Acute Stroke Services (COMPASS) Study, North Carolina, 2016-2018 (ICPSR 38185)

Version Date: Oct 7, 2021 View help for published

Principal Investigator(s): View help for Principal Investigator(s)
Pamela W. Duncan, Wake Forest University. Baptist Medical Center; Wayne D. Rosamond, University of North Carolina-Chapel Hill; Cheryl D. Bushnell, Wake Forest University. Baptist Medical Center

Version V1

Slide tabs to view more

The Comprehensive Post-Acute Stroke Services (COMPASS) Study is a pragmatic cluster-randomized clinical trial that evaluated the real-world effectiveness of the COMPASS transitional care (COMPASS-TC) model compared to usual care among adult stroke and transient ischemic attack (TIA) patients discharged home between 2016 and 2018. In Phase 1, 40 North Carolina hospital units were randomized 1:1 to the COMPASS-TC intervention or usual care, stratified by stroke patient volume and stroke center certification. In Phase 2, hospitals randomized to usual care crossed over to implement COMPASS-TC, and hospitals randomized to the intervention sustained COMPASS-TC. The intervention was patient-centered and assessed social and functional determinates of health to inform individualized care plans for secondary prevention, recovery, and referrals to services and community-based resources. COMPASS-TC was consistent with Centers for Medicare and Medicaid Services (CMS) TC management reimbursement requirements.

The primary outcome was functional status (Stroke Impact Scale-16; SIS-16) at 90 days; secondary outcomes were mortality, disability, medication adherence, depression, cognition, self-rated health, fatigue, care satisfaction, home blood pressure monitoring, falls, and caregiver strain. Telephone interviewers, blinded to treatment assignment, assessed these outcomes at 90 days.

Duncan, Pamela W., Rosamond, Wayne D., and Bushnell, Cheryl D. Comprehensive Post-Acute Stroke Services (COMPASS) Study, North Carolina, 2016-2018. Inter-university Consortium for Political and Social Research [distributor], 2021-10-07.

Export Citation:

  • RIS (generic format for RefWorks, EndNote, etc.)
  • EndNote
Patient-Centered Outcomes Research Institute (PCORI) (PCS-1403-14532)


Access to the data in this collection is restricted. Users interested in obtaining these data must complete a Restricted Data Use Agreement, specify the reason for the request, and obtain IRB approval or notice of exemption for their research.

Inter-university Consortium for Political and Social Research

2016-07-01 -- 2018-03-31
2016-06 -- 2018-03, 2017-11 -- 2019-01
  1. Clinical patient data were collected by post-acute care nurse (PAC) at baseline among patients in the intervention arm. Clinical data were abstracted from the medical record and collected from patient report. At the COMPASS study clinic visit, which occurred approximately 14 days post-discharge, the PAC and advanced practice provider (APP; nurse practitioner or physician's assistant) conducted standardized assessments using an iPad application entered into a secure web-based data-entry system. Patient-reported measures captured as part of these assessments include, among others, functional and social determinants of health, cognitive status, health data (e.g., blood pressure, medication reconciliation), lifestyle management (e.g., alcohol, smoking, physical activity), and knowledge of stroke risk factors.
  2. Patient-reported outcomes, including the primary outcome (SIS-16) and other standardized assessments, were collected at approximately 90 days post-discharge via telephone administration by trained interviewers. Interviewers were blinded to treatment group and use standardized scripts and interviewing guidelines. We utilize reminder letters, additional phone contacts, mailed surveys, and proxy interviews to increase follow-up rates. All surveys were available in Spanish and in English.
  3. Caregiver-reported outcomes were collected through self-administered paper survey (including the Modified Caregiver Strain Index), which was mailed to them at approximately 95 days after the patient's hospital discharge. Caregivers who did not respond to the first mailed questionnaire were contacted by telephone and sent an additional paper questionnaire. All surveys were available in Spanish and in English.
  4. Hospital-level data were collected from hospitals and were obtained from public sources. Annual stroke volume was provided directly by hospitals. Primary/Comprehensive Stroke Center certification status was obtained from public Joint Commission data. Teaching status was obtained from public Centers for Medicare and Medicaid Services (CMS) data. Rural/urban status was determined by Rural Urban Commuting Area (RUCA) codes, available in public US Census data.
  5. For more information on the COMPASS Study, please visit the COMPASS Study website .
  6. Medical Subject Heading (MeSH) terms that apply to this study include: health services accessibility, patient care, stroke, transient ischemic attack, transitional care.


The primary objective of the COMPASS Study is to evaluate the comparative effectiveness of the COMPASS Intervention compared to usual care with respect to improving stroke survivor functional status (measured by the Stroke Impact Scale) at 90 days post-discharge. Intention-to-treat principles will be used to determine all primary and secondary outcomes.

The Principal Investigator's primary research question was, Does implementation of the COMPASS for stroke patients discharged directly home, improve functional outcomes (measured by the Stroke Impact Scale-16) at 90 days post-stroke?

The COMPASS Study is a pragmatic, cluster-randomized trial of 50 hospitals in North Carolina designed to determine the effectiveness of a model of post-acute stroke care compared with usual care. Individual hospitals were randomized to either receive the COMPASS intervention at the beginning of the study (Phase 1) or in Phase 2 (referred to as the control group). A timeline of the study can be found on page 11 of the Initial Protocol document provided by the P.I.

In 2013, data from hospitals in the North Carolina Stroke Care Collaborative (NCSCC) indicated that 46% of patients were discharged directly home from the hospital after a stroke (our proposed study population). Using this data, an estimated sample of approximately 6,000 potentially eligible participants was anticipated. In Phase 1, hospital staff enrolled 6024 adults discharged home after stroke or transient ischemic attack (TIA). Twenty-five hospitals (3,771 participants) were included in the Phase 2 analysis.


Hospitals: All North Carolina hospitals were eligible to participate if they had an emergency department that treated patients with stroke and could identify stroke and TIA patients concurrent with care.

Patients: Eligible patients were aged 18 years old or older; English or Spanish speaking; diagnosed with ischemic stroke, hemorrhagic stroke (excluding subdural or aneurysmal hemorrhage), or TIA; and discharged directly home.

Caregivers: Unpaid family member or friend who provided assistance to the patient after returning home.

Individual, Hospital

The study is comprised of five parts:

  • Hospital-Level Data (12 variables, 40 cases; 3 string variables): Includes variables on volume of stroke patients, area characteristics (rural v. urban), and how the hospital participated in the study.
  • Baseline Data (76 variables, 10,090 cases; 48 string variables): Includes variables on patient demographics such as age and race, and medical and hospitalization history.
  • Intervention-Specific Data (14 variables, 6,817 cases; 8 string variables): Includes variables on hospital-patient communication and clinic visits.
  • Patient-Reported Outcome Data (103 variables, 9,894 cases; 3 string variables): Includes patient survey data on medical needs and general health post-treatment.
  • Caregiver-Reported Data (101 variables, 4,208 cases; 1 string variable): Includes variables on types of care provided to patients before and after stroke.

Hospitals: Among 95 eligible hospitals, 41 agreed to participate. These were randomized as 40 units due to shared staff for two hospitals. Thirty-nine hospitals completed at least 12 months of participation between July 2016 and March 2018; 1 hospital enrolled no patients and was excluded from analyses.

Patients: Of the 9,894 eligible patients enrolled in the study across Phases 1 and 2, 5,748 (58.1%) successfully completed the 90-day outcomes survey. An additional 3,969 (40.1%) were lost to follow-up, and 177 (1.8%) died prior to 90 days. The primary outcome (SIS-16) was ascertained for 59% of patients at 90 days in Phase 1 (61% [n=1644] intervention, 57% usual care [n=1832]). Thirty-five percent (n=928) of patients at intervention hospitals attended a COMPASS clinic visit within 30 days. Seventy-nine percent of these visits occurred within the 14-day window required for CMS transitional care management (TCM) reimbursement. Receipt of the intervention per protocol ranged from 6% to 70% across intervention arm hospitals in Phase 1.

Caregivers: A total of 4,208 patients identified an eligible caregiver, 1722 (64%) among intervention patients and 2486 (78%) among usual care patients. Thirty-three percent (n=569) of caregivers in the intervention arm and 27% (n=659) of caregivers in the usual care arm completed caregiver assessment.

  • Stroke Impact Scale (SIS-16) measured physical function (Primary Aim) at 90 days post-discharge. Each of the 16 items was measured using a 5-point Likert response scale. Scores were standardized to 0-100 for analysis, with higher scores corresponding to better function. Surveys with at least 12 completed items were scored. Mean SIS-16 scores were compared between treatment arms.
  • Self-reported general health score was measured on a 5-point scale from 1 (poor) to 5 (excellent) at 90 days post-discharge. Scores were analyzed as a continuous variable with responses transformed according to custom from poor to excellent, assigned as 15, 30, 80, 90, and 95. Mean self-rated health scores were compared between treatment arms.
  • Self-reported home blood pressure monitoring (yes or no) and frequency (daily, weekly, and monthly) was queried and dichotomized for analysis as monitoring with any frequency versus none. Odds of blood pressure monitoring at least monthly were compared between treatment arms.
  • Montreal Cognitive Assessment (MoCA) measured cognition using a 5-minute protocol that included 4 items examining attention, verbal learning and memory, executive functions/language, and orientation. Scores ranged from 0 and 30, with higher scores representing better cognition. Mean MoCA scores were compared between treatment arms.
  • Patient Health Questionnaire-2 (PHQ-2) queried the frequency of depressed mood and anhedonia over the past 2 weeks. Each item was answered using a Likert scale from 0 (not at all) to 3 (nearly every day). Total scores range from 0-6 and were dichotomized as less than 3 (no depression) and greater than or equal to 3 for analysis, following standard screening criteria. Odds of depression were compared between treatment arms.
  • Self-reported physical activity measured 10-minute bouts of walking and average number of days and minutes of walking. Values were dichotomized based on the total number of minutes walked during the past seven days as greater than or equal to 150 versus less than 150, based on CDC guidelines for physical activity for older adults. Odds of walking 150 minutes or more per week were compared between treatment arms.
  • PROMIS Fatigue Instrument ranked the level of fatigue over the past 7 days from 1 (not at all) to 5 (very much) using 4 items. Total raw scores range from 4-20, with lower scores corresponding to less fatigue. For analysis, raw scores were translated into T-scores using a conversion table that rescales to a standardized score with mean 50 and SD 10. Mean standardized PROMIS fatigue scores were compared between treatment arms.
  • Self-reported falls since discharge were measured at 90 days. Analysis was based on prevalence of any fall since hospital discharge (no falls versus at least one fall). Odds of no fall by 90 days were compared between treatment arms.
  • Modified Rankin Scale (mRS) measured the degree of disability or dependence in daily activities for people who suffered a stroke. The scale ranges from 0 (no symptoms) to 6 (dead). Patients confirmed to have died before completion of the 90-day survey received a score of 6. For analysis, values were categorized as 0, 1, 2-3, and 4-6. Odds of lower mRS category (lower disability) were compared between treatment arms.
  • Morisky Green Levine Scale (MGLS-4) measured medication adherence consists of four items with yes/no responses.



2021-10-07 ICPSR data undergo a confidentiality review and are altered when necessary to limit the risk of disclosure. ICPSR also routinely creates ready-to-go data files along with setups in the major statistical software formats as well as standard codebooks to accompany the data. In addition to these procedures, ICPSR performed the following processing steps for this data collection:

  • Checked for undocumented or out-of-range codes.


  • The public-use data files in this collection are available for access by the general public. Access does not require affiliation with an ICPSR member institution.

  • One or more files in this data collection have special restrictions. Restricted data files are not available for direct download from the website; click on the Restricted Data button to learn more.