Study Design
As part of a program evaluation, we conducted a retrospective chart to evaluate a community-based serious-illness care program, the REACH program (Reaching out to Enhance the care of Adults in their Communities and their Homes) based at an academic medical center. Data collection included chart reviews of all patients enrolled from August 2014 to March 2016. Two research assistants independently completed a subset of chart reviews of visit notes and patient data in the electronic health record (EHR), reconciled differences until reviews were consistent, and conferred regularly to prevent drift and maintain data quality. Using the EHR EPIC, the initial clinical visit note was templated into a semi-structured format, though other elements could be added separately to the note, and the plan of care was untemplated. Clinicians were not aware they were being evaluated during the time period of the study. A formal chart review guide was developed to standardize the data abstraction process.
Community-based Serious-Illness Care Model
The REACH program was a community-based serious-illness care program that extends both medical and palliative care to seriously ill and medically complex adults in their homes or assisted living communities.(16-18) The REACH program was designed to meet the needs of patients and primary care providers and thus offered an array of services from short-term consultation to ongoing co-management to assuming primary care if needed. The REACH program was interdisciplinary, with a team comprised of: (1) clinicians (MD, DO, nurse practitioner) with expertise in geriatrics and/or palliative care; (2) a clinical pharmacist (PharmD); (3) a care manager (Master’s of Social Work); and (4) a nurse coordinator (RN) who trained in the Guided Care model, a model designed to coordinate care across healthcare settings.(19) The program’s goals were for the patients to be seen at least quarterly, or more often as needed based on the evidence that frequent visits were needed to improve care outcomes.(20) Patients are seen by clinicians in their homes for an initial visit with support from the other team members remotely. The nurse coordinator occasionally accompanied the clinician to these visits and family members or caregivers could be present. The pharmacist and social worker would help provide input in medication management or social support resources, respectively, as needs identified by the visiting clinicians arose. The nurse coordinator would coordinate care with home health agencies, hospice, medication refills, radiologic orders and results, and orders for durable medical equipment. Patients had direct access to the nurse coordinator during business hours for acute issues and to an after-hour nurse triage system if needed. The nurse coordinator made infrequent visits to patient homes. Urgent visits could be made upon request.
Participants & Eligibility
Patients were eligible if they were seen for an initial visit by a REACH clinician in their home or assisted living community during the study period. Patients were referred by their primary care clinicians to enroll in REACH. Patients were eligible if they had become unable to access office-based care or had serious medical illnesses that made coordinated primary care challenging, were adults living within 30 miles of the medical campus, and agreed to home visits. Nursing home patients were not eligible.
Study Measures
Chart review collected demographic information and days enrolled in the program. We also quantified the degree of patients’ serious illness using the Charlson Comorbidity Index (CCI).(21) The CCI is comprised of 19 specific medical conditions and their severity.(21) The CCI yields a score from 0 to 35, where higher numbers represent worse overall health, and was assessed from available records within three months after the REACH initial visit. We also assessed functional status via the presence of the Katz instrumental activities of daily living (IADLs),(22) and basic ADLs.(23) Higher IADLs and ADLs represent better function.(22, 23) IADLs measure more complicated aspects of daily function such as managing finances, medications, and transportation, ADLs measure basic functions such as toileting, walking, or eating. We recorded independence in IADLs and ADLs (Yes/No). Other measures such as hospitalization and palliative care measures are reported elsewhere.(16, 18)
To assess the quality-of-care delivered by the community-based serious-illness care program, we operationalized standards based on seven of ten quality domains proposed by the National Home-Based Primary and Palliative Care Network.(8) We assumed that if a patient warranted admission into the community-based serious-illness care program that they should be assessed for all of these standards. The program was developed to adhere to standards of care. At the time of the program evaluation, no standards had been explicitly published, though the standards we used were very similar to current standards.(24) We attempted to find a standard that could be completed by chart review for each quality-of-care domain. Our team collaboratively mapped each of these standards onto the clinical visit notes. We were not able to operationalize three domains to chart review (patient education, patient and caregiver experience, and cost) and did not look at frequency of visits or other important quality metrics such as emergency department visits or hospitalizations, which need alternative methods of assessment. Though the proposed standards include more than those we assessed, we developed the following chart-based standards for seven domains, abstracted from the initial visit:
Medical assessment: percent of initial visits with a) structured functional status assessment,(22, 23) b) structured cognitive screening (via the Veterans Affairs St. Louis University Mental Status Exam (VA-SLUMS, range 0-30, with higher scores being better cognition)(25), c) assessment of communication barriers (e.g., confusion/sedation, dementia, language barrier, or other), and d) assessment of spiritual needs
Care coordination: percent of initial visits with documented communication with EITHER a primary care clinician, specialty clinician, or REACH team social worker or pharmacist.
Safety: percent of initial visits with a) medication review, b) assessment for falls, and c) assessment of behavioral issues
Quality of life: percent of initial visits with comprehensive symptom assessment inclusive of pain, dyspnea, constipation, poor appetite, nausea, depressive symptoms, anxiety, fatigue, and psychosocial distress
Clinician competency: percent of initial visits with a) an assessment or administration of the influenza vaccine, b) prescription for antidepressant among those with depression, c) opiate prescription (and bowel regimens), d) antipsychotic and benzodiazepine medication prescription (for each, lower frequency is better quality)
Goal attainment: percent of initial visits with documented discussions of advance directives, goals of care, and surrogate decision-makers
Access: percent of initial visits with new referrals to hospice, home health, or home caregiver agencies; percent of initial visits with the prescription of durable medical equipment.
Analysis
We used descriptive statistics to report the demographics and quality standards for all patients, including the serious illness within the population and how well the REACH program functioned across the seven domains. Each standard was noted as completed only if it was noted to be completed in its entirety upon chart review, e.g., the entire VA-SLUMS was performed, all IADLS/ADLS were evaluated, a symptom was documented in the note as assessed. We noted the presence or absence of each standard within each domain, to capture the breadth of the domains. All statistical analyses were performed using STATA/SE 11.2 (StataCorp, College Station, TX) or SPSS for Windows Version 24 (IBM Corp., Armonk, NY).