Transitions in healthcare are a high-risk event. Variation in the structure and quality of transitions and associations between transitions and adverse events (e.g. medical errors, readmission, mortality) have led to increased focus on assessing and improving transitions of care. Discharge from the hospital is one of the most error-prone transitions in healthcare, as increasingly ill patients, hospitalized for fewer days and receiving more complex care, create an environment where errors are increasingly likely.
The VHA is focused on and committed to improving transitions of care and patient-centered care, especially for vulnerable populations of Veterans. Studies have shown that collecting patient-centered data, including specific needs and functional status information post-discharge, may help improve care transitions and decrease adverse events. Further, using telehealth interventions to support care transitions is increasingly highlighted as an important future direction for healthcare. Although Veterans now are scheduled to receive a telephone call from a VA Patient Aligned Care Team (PACT) provider within two days of discharge, not all Veterans receive this call, a standardized needs checklist and functional assessment is not part of the call, and completing these calls takes substantial time for PACT providers.
We proposed the use of current VA automated telephone response (ATR) systems to collect a checklist of common post-discharge problems and patient-reported functional status items to demonstrate feasibility, reliability, and validity of the system for potential future use in interventions to improve transitions of care.
Our objectives were: 1) To assess the feasibility and reliability of the ATR system to collect patient-reported post-discharge needs and functional status data 1- and 30-days post-discharge; 2) To examine the use of a post-discharge needs checklist and functional status questions to identify actionable needs and predict subsequent adverse events; and 3) To explore the use of patient-reported instrumental activities of daily living (IADL) data post-discharge.
Over a 5-month period we enrolled 121 Veterans aged 65 and older who were discharged from the Medicine or Neurology service at the Roudebush VAMC. Baseline functional status, clinical, and social data were collected at enrollment, prior to discharge. The ATR system contacted the Veteran at 1 (+/- 2 days) and 30 days (+/- 5 days) post-discharge. Once a 1-day call is completed, study staff attempted a person-to-person call to collect identical survey data to assess reliability of the ATR method and to assess for any additional unmet needs. Each call included 23-26 questions including a 10-item post-discharge problems/needs checklist (yes/no), and PROMIS items assessing activities of daily living (ADLs) and instrumental ADLs (3-point Likert response for level of difficulty). We assessed the completion rate, number of calls required, patient versus caregiver response rates, and data completion and quality using the ATR system. We calculated the % accuracy for each questionnaire item, comparing data collected via ATR to that collected during the personal call. When data are available to complete 90-day post-discharge assessments, we will construct models of post-discharge events (mortality and readmission) examining the association of 1-day ATR call results with events.
Enrollment: Of 1063 patients screened, 610 were eligible and 121 were enrolled. Mean age was 72.3 years, 97% were male, 78% White and 20% Black.
ATR Call Response data: Over 75% of enrolled patients (94/121, 77.7%) completed a 1-day ATR call; 54% on the 1st attempt, 30% on the 2nd attempt, 16% on the 3rd attempt. No subjects required more than 3 call attempts to complete a call. Calls were completed relatively rapidly, with 2.7 mean days to call completion post-discharge; 23% were completed on post-hospital day 1, and 36% on post-hospital day 2. All 94 1-day calls had 100% complete data (no missing items).
For 30-day calls, 71% of those completing a 1-day call also completed the 30-day call (67/94). Most patient completed this call on the 1st (60%) or the 2nd (16%) call attempt. Mean time to post-hospitalization to call completion was 32.3 days and 66/67 (98.5%) of calls had complete data.
Reliability data: Nearly 2/3 of the patients completing a 1-day ATR call also completed a reliability (personal) call (60/94, 63.8%). Individual item agreement on the ATR and the reliability calls ranged from 73% to 97%. One needs checklist item had < 80% agreement (the question about whether things were more difficult to carry out than before). Two of six ADL items had < 80% reliability (bathing and toileting) and two of three IADL items had < 80% reliability (preparing food and overall caring for self). All ADL and IADL items scored higher (better) in the reliability call than the ATR call. Overall, however, the summed ADL scores from the ATR and the reliability call were significantly associated (p = 0.0012, Wilcoxon signed-rank test).
Patient response data: At both 1- and 30-days nearly half of all subjects reported at least one problem post-discharge, and 15% at 1-day and 18% at 30-days reported > 3 problems. At both time points the most common problems endorsed were problems with usual work/roles (21% at 1-day, 24% at 30-days) and coping with changes in health (20% at 1-day, 27% at 30-days). At 30-days, 21% also reported feeling anxious or depressed. The most common ADL problems at 1-day were continence (27% with some difficulty and 6% with much difficulty), and at 30-days were dressing (27% some difficulty and 5% much difficulty) and continence (22% some difficulty and 3% much difficulty). At 1-day, 18% reported some difficulty and 8% much difficulty with taking medications on time; this dropped slightly at 30-days to 16% and 3%. At the close of the call, 22% at 1-day and 19% at 30-days requested a call from their provider.
Patient satisfaction: Overall, 85% of Veterans responding to the reliability call said they would use the ATR system again to communicate with their provider. Among this group of medically complex, older Veterans, 95% preferred being contacted by telephone if their VA provider needed to reach them, and 91% preferred to use the telephone themselves if they wanted to communicate information to their care provider. While half of these patients had used MyHealthEVet, 49% said they do not use a computer at home, and many had accessed MyHealthEVet through someone else's computer or had another person access the system on their behalf.
An ATR call is feasible and reliable for collecting information about problems and functional limitations immediately post-hospital discharge. IADL items may be less reliable than ADL and problem checklist items in this format. Many Veterans report ongoing needs and functional limitations post-discharge, suggesting this system could be useful as an early indicator of needs that might be addressed by the PACT team to prevent complications and improve Veteran outcomes. Future analyses will compare the needs identified via ATR call to those identified in the PACT team post-hospital call, and will assess the association between needs, ADLs and 90-day utilization.
External Links for this Project
Grant Number: I21HX001805-01A1
None at this time.