Key Points
Question
Does a 30-day automated text messaging program to support primary care patients after hospital discharge reduce acute care revisits?
Finding
In a randomized clinical trial of 4736 individuals discharged from an acute care hospitalization, 30-day emergency department visits and readmissions did not differ significantly between patients who received an intensive automated texting intervention and those who did not.
Meaning
The findings of this study suggest that although automated texting as a standalone strategy appears to be insufficient to impact acute care revisits, it may serve as an operationally efficient adjunct or replacement for other commonly used strategies.
Abstract
Importance
Postdischarge outreach from the primary care practice is an important component of transitional care support. The most common method of contact is via telephone call, but calls are labor intensive and therefore limited in scope.
Objective
To test whether a 30-day automated texting program to support primary care patients after hospital discharge reduces acute care revisits.
Design, Setting, and Participants
A 2-arm randomized clinical trial was conducted from March 29, 2022, through January 5, 2023, at 30 primary care practices within a single academic health system in Philadelphia, Pennsylvania. Patients were followed up for 60 days after discharge. Investigators were blinded to assignment, but patients and practice staff were not. Participants included established patients of the study practices who were aged 18 years or older, discharged from an acute care hospitalization, and considered medium to high risk for adverse health events by a health system risk score. All analyses were conducted using an intention-to-treat approach.
Intervention
Patients in the intervention group received automated check-in text messages from their primary care practice on a tapering schedule for 30 days following discharge. Any needs identified by the automated messaging platform were escalated to practice staff for follow-up via an electronic medical record inbox. Patients in the control group received a standard transitional care management telephone call from their practice within 2 business days of discharge.
Main Outcomes and Measures
The primary study outcome was any acute care revisit (readmission or emergency department visit) within 30 days of discharge.
Results
Of the 4736 participants, 2824 (59.6%) were female; the mean (SD) age was 65.4 (16.5) years. The mean (SD) length of index hospital stay was 5.5 (7.9) days. A total of 2352 patients were randomized to the intervention arm and 2384 were randomized to the control arm. There were 557 (23.4%) acute care revisits in the control group and 561 (23.9%) in the intervention group within 30 days of discharge (risk ratio, 1.02; 95% CI, 0.92-1.13). Among the patients in the intervention arm, 79.5% answered at least 1 message and 41.9% had at least 1 need identified.
Conclusions and Relevance
In this randomized clinical trial of a 30-day postdischarge automated texting program, there was no significant reduction in acute care revisits.
Trial Registration
ClinicalTrials.gov Identifier: NCT05245773
This randomized clinical trial examines the efficacy of an automated text messaging program to reduce acute health care resource utilization among primary care patients who have been discharged from an acute care hospitalization.
Introduction
Transitional care management after hospital discharge is both a priority and challenge for health systems. Nearly 15% of all discharged inpatients are readmitted to the hospital within 30 days; including emergency department (ED) returns, the rate of overall acute care revisits is even higher.1,2 For patients, the period of time after discharge is challenging, as they adjust to new medication regimens, coordinate follow-up care, and recover strength.3,4 A variety of transitional care management programs have aimed to provide patients with enhanced support during this time and mitigate gaps in care, often with the overall aim of reducing subsequent acute care use.5,6
The most common approach has been to use nurse-led telephone calls to identify needs shortly after discharge.7,8 While these calls have proven effective in some settings, they are limited in scope and operationally burdensome.9,10 We hypothesized that an automated texting platform could improve efficiency and effectiveness beyond calls alone by automating large-scale outreach (requiring staff involvement only after a need has been identified), offering a low-friction medium for patient engagement, and enabling asynchronous communication.
A 30-day automated texting program was designed and piloted wherein patients received check-in messages from their primary care practice on a tapering schedule after hospital discharge. In a difference-in-differences analysis, implementation of this program at a single primary care practice saw high levels of engagement and was associated with a significant reduction in acute care revisits.11 In the present study, we tested the effectiveness of this program in a randomized clinical trial conducted across 30 primary care practices in the University of Pennsylvania Health System (UPHS).
Methods
Overview
We conducted a 2-arm randomized clinical trial between March 29, 2022, and January 5, 2023. Patients were enrolled after discharge from an acute care hospitalization. Participants were assigned either to usual postdischarge transitional care management or usual care plus receipt of a 30-day automated text messaging program. The texting program was built and managed by Way to Health, a platform created with National Institutes of Health funding to provide automated technology infrastructure in support of clinical care and care delivery innovation research. Electronic medical record (EMR) data were collected for 60 days after discharge. The complete protocol is available in Supplement 1. This study was reviewed and approved by the University of Pennsylvania Institutional Review Board. We followed the Consolidated Standards of Reporting Trials Extension (CONSORT Extension) reporting guideline (with pragmatic extension) and Standards for Quality Improvement Reporting Excellence 2.0 (SQUIRE) reporting guideline. The trial operated under a waiver of informed consent given that (1) both arms continued to receive usual care, (2) text outreach involved no more than minimal risk to patients, (3) once enrolled in the texting program, patients were free to opt out at any time, and (4) requiring informed consent would substantially limit the external validity of the study.
Participants and Study Sites
The patient population comprised adults (age ≥18 years) who received care in 30 participating University of Pennsylvania Health System (Penn Medicine) primary care practices in the Philadelphia region. Included were all patients discharged from any acute care hospital in the region as identified in HealthShare Exchange, even if not a Penn Medicine hospital, and identified as medium to high risk by a UPHS risk score of 4 or above at the time of discharge. The UPHS risk score is an Epic Systems Corporation developed and validated point score, which is used to estimate a patient’s risk of adverse health events in the next year based on clinical information presented in prior literature and generally available in the EMR. eTable 2 in Supplement 2 provides details regarding the calculation of this score.12,13,14 Patients were excluded only if they were discharged to home hospice care or discharged from (1) planned chemotherapy admissions, (2) scheduled surgeries, and (3) obstetrics admissions. Patients discharged more than once during the study period were enrolled only for their first discharge.
Enrollment Procedures and Randomization
Patients were identified for inclusion via daily HealthShare Exchange reports, a health information exchange for the greater Philadelphia region that provides practices with information on discharges from all regional hospitals.15 These reports were then screened for exclusion criteria based on data available in both the reports themselves and, where available, EMR data. The filtered reports were then uploaded into the Way To Health platform for enrollment and randomization. All eligible patients were automatically enrolled and were randomized 1:1 to intervention or control, stratified by practice. Investigators and data analysts were blinded to assignment, but patients and clinic staff were not.
Control Arm
Patients in the control arm received the standard transitional care management telephone call from their practice within 2 business days of discharge. This call, placed by a nurse, is meant to identify any needs soon after discharge and consists of a set of questions related to follow-up appointments, medications, symptoms, and home care needs. Patients are typically scheduled for a postdischarge visit during this call, if they do not already have one. If they do not answer the first call, another attempt is made; further outreach is left to the nurses’ discretion.
Intervention Arm
Patients in the intervention arm received the transitional care management telephone call following the same procedures as those for the control arm. In addition, they received automated text messages from their practice, via the Way To Health platform. On enrollment (day 0), patients received an introductory message describing the program and advising them how to reach out or opt out at any time.
Beginning the day after enrollment (day 1), patients received check-in messages on a tapering schedule (eTable 1 in Supplement 2). These messages asked, “Is there anything we can help you with today?” If they answered no, there was no further action. If they answered yes, a follow-up message asked them to further categorize their need (eg, “I need help with my medicines”). The eMethods in Supplement 2 provides the full script of messages. Patients could also reach out anytime outside of a scheduled check-in context by texting in “Call.”
Escalations (instances in which a patient identified a need) were routed to an inbox in the EMR that was monitored by practice staff during business hours. Patients would receive a follow-up telephone call from the practice care manager (a registered nurse) within 1 business day (generally the same day).
Patients were given the opportunity to opt out at any time. Patients who did not respond to 3 consecutive check-in messages received an additional message asking whether they needed further help or no longer wanted to receive the messages; if they did not respond to this, messages would continue according to the usual schedule.
Measures and Outcomes
The primary outcome was any acute care visit (ED visit or readmission) within 30 days of discharge. Secondary prespecified outcomes included any acute care visit within 7 and 60 days of discharge; an ED visit or readmission (analyzed separately) within 7, 30, and 60 days of discharge; total days spent in the hospital (either in the ED or as inpatient) within 30 days of discharge; time from discharge to first acute care visit; scheduling and completion of a primary care postdischarge follow-up visit within 14 days; and, as a surrogate measure of staff time, total number of nonvisit interactions between the patient and the primary care practice (a composite of telephone calls [including those resulting from program escalations for patients in the intervention arm], refill requests, and patient portal messages) within 30 days of discharge. Outcomes and patient demographic characteristics were obtained from the EMR, which also records acute care visits in regional systems outside of Penn Medicine.
For the intervention arm, additional feasibility and acceptability measures were obtained from the Way To Health platform, including satisfaction, measured at the conclusion of the 30-day program with a 1-item Net Promoter Score (NPS) question: “On a scale from 0 (unlikely) to 10 (extremely likely), how likely are you to recommend Penn Medicine’s discharge follow-up program to a friend or colleague?”16 The NPS varies from −100 to 100, calculated as the percentage of scores of 9 or above reduced by the percentage of scores 6 or below (eMethods in Supplement 2 provides further details).
Statistical Analysis
The target sample size was 5000 patients, based on 80% power and a 2-sided significance level of .05, to detect a risk ratio (RR) of 0.8 for the primary outcome of any acute care visit within 30 days. This estimate was based on pilot data, which showed an approximate event rate of 20% and, in the pilot study, an adjusted odds ratio of 0.6 associated with the primary outcome; the trial was designed to have 80% power to detect a 4% absolute reduction in the primary event rate.
All analyses were conducted using an intention-to-treat approach. Continuous and binary outcomes were compared between intervention and control arms using 2-sample t tests and Pearson χ2 tests. Prespecified subgroup analyses included risk group (UPHS risk score of 4-5 [medium] or ≥6 [high]) and age (<65 or ≥65 years). To assess subgroup effects, we conducted a stratified analysis by subgroup using Pearson χ2 and, in a logistic regression model, interacted the arm with a binary indicator for the baseline factor of interest. Additional prespecified exploratory analyses included a logistic regression model with fixed effects at the practice level, and adjusting for demographic (age, gender, self-reported race and ethnicity and insurance type) and clinical (length of stay and UPHS risk score) covariates. Race and ethnicity data were considered important in evaluating any baseline demographic variation in the intervention and control populations. All statistical analyses used Stata software, version 16.1 (StataCorp LLC).
Results
Study Population
A total of 4736 patients were randomized to the intervention (2352) or control (2384) arms, with a mean (SD) age of 65.4 (16.5) years. Of these, 2824 (59.6%) were female, 1912 (40.4%) were male, 66 (1.4%) were Asian, 2322 (49.0%) were Black, 2128 (44.9%) were White, 164 (3.5%) were Hispanic or Latino, 2768 (8.5%) were insured by Medicare, and 911 (19.2%) were insured by Medicaid (Table 1). The mean (SD) risk score was 5.5 (1.6), length of stay was 5.5 (7.9) days, and the number of acute care visits in the past 12 months was 3.3 (5.2). Of the 5048 patients who were initially enrolled at the time of upload of the HealthShare Exchange reports, 174 in the intervention group and 138 in the control group were recognized as ineligible after automated randomization based on repeat discharges during the study period (2), low risk score (1), and discharge from a non–acute care facility (309) (Figure).
Table 1. Patient Characteristics.
| Characteristic | No. (%) | |
|---|---|---|
| Control | Intervention | |
| Observations | 2384 | 2352 |
| Age, mean (SD), y | 65.2 (16.6) | 65.6 (16.4) |
| Gender | ||
| Female | 1409 (59.1) | 1415 (60.2) |
| Male | 975 (40.9) | 937 (39.8) |
| Racea | ||
| American Indian, Alaskan Native, Native Hawaiian, or Other Pacific Islander | 12 (0.5) | 9 (0.4) |
| Asian | 33 (1.4) | 33 (1.4) |
| Black | 1173 (49.2) | 1149 (48.9) |
| Hispanic | 4 (0.2) | 3 (0.1) |
| White | 1078 (45.2) | 1050 (44.6) |
| Unknown | 84 (3.5) | 108 (4.6) |
| Ethnicitya | ||
| Hispanic or Latino | 90 (3.8) | 74 (3.1) |
| Not Hispanic or Latino | 2281 (95.7) | 2263 (96.2) |
| Unknown | 13 (0.5) | 15 (0.6) |
| Payer | ||
| Commercial | 406 (17.0) | 421 (17.9) |
| Medicaid | 466 (19.5) | 445 (18.9) |
| Medicare | 1398 (58.6) | 1370 (58.2) |
| Self-pay | 0 | 2 (0.1) |
| Unknown | 114 (4.8) | 114 (4.8) |
| Acute care visits past 12 mo, mean (SD) | 3.2 (4.7) | 3.3 (5.7) |
| Risk profile score, mean (SD) | 5.5 (1.6) | 5.5 (1.6) |
| Length of stay, mean (SD) | 5.5 (6.3) | 5.6 (9.1) |
Self-reported race and ethnicity data were collected as measures of baseline demographic variation in the intervention and control populations.
Figure. Patient Flowchart.
Primary Outcome
There were 557 (23.4%) acute care revisits in the control group and 561 (23.9%) in the intervention group within 30 days of discharge (RR, 1.02; 95% CI, 0.92-1.13) (Table 2). There was similarly no significant difference in the rate of acute care revisits between arms at 7 and 60 days (eTable 3 and eTable 4 in Supplement 2). There were no subgroup differences based on age or patient risk score, and there was no significant difference in the adjusted regression model (eTable 5 and eTable 6 in Supplement 2).
Table 2. 30-Day Health Care Use.
| Variable | No. (%) | RR, absolute difference (95% CI) | P value | |
|---|---|---|---|---|
| Control | Intervention | |||
| Acute carea | 557 (23.4) | 561 (23.9) | 1.02 (0.92 to 1.13)b | .70 |
| ED visit | 474 (19.9) | 463 (19.7) | 0.99 (0.88 to 1.11)b | .86 |
| Readmission | 376 (15.8) | 370 (15.7) | 1.00 (0.87 to 1.14)b | .97 |
| Total days in hospital, mean (SD) | 0.75 (2.75) | 0.74 (2.61) | −0.01 (−0.14 to 0.16)c | .89 |
Abbreviations: ED, emergency department; RR, risk ratio.
A composite measure of whether an individual had an ED visit or a readmission within 30 days.
Risk ratio.
Mean difference.
Secondary Outcomes
There was no significant difference in the likelihood of an ED visit (RR, 0.99; 95% CI, 0.88-1.11) or readmission (RR, 1.00; 95% CI, 0.87-1.14) at 30 days (Table 2). There were similarly no significant differences in these outcomes at 7 and 60 days (eTable 3 and eTable 4 in Supplement 2). Total days spent in the hospital within 30 days of discharge was similar between the groups (absolute difference, −0.01 days; 95% CI, −0.14 to 0.16 days). There was no significant difference in time to the first acute care visit (eTable 5 in Supplement 2). The percentage of patients scheduling a primary care follow-up visit within 14 days was also similar (50.2% vs 47.9% in the control arm: RR, 1.05; 95% CI, 0.99-1.11) (Table 3). The rate at which primary care follow-up visits were completed within 14 days was similar (RR, 1.02; 95% CI, 0.94-1.11). Total interactions between the patient and the practice within 30 days were higher in the intervention vs control arm (3.06 vs 2.29; absolute difference, 0.77; P < .01).
Table 3. Primary Care Access and Use.
| Variable | No. (%) | RR or absolute difference (95% CI) | P value | |
|---|---|---|---|---|
| Control | Intervention | |||
| Scheduled visit within 14 d | 1142 (47.9) | 1181 (50.2) | 1.05 (0.99-1.11)a | .11 |
| Completed visit within 14 d | 762 (32.0) | 768 (32.7) | 1.02 (0.94-1.11)a | .61 |
| Total interactions within 30 d, mean (SD) | 2.29 (2.63) | 3.06 (3.15) | 0.77 (0.60-0.94)b | <.01 |
Abbreviation: RR, risk ratio.
Risk ratio.
Absolute difference in means.
Intervention Feasibility and Acceptability Measures
Within the intervention group, 92 patients (3.9%) had a nonworking number (as these were cell phone numbers pulled directly from the EMR) (Table 4). An additional 129 individuals (5.5%) opted out of the program before the end of the 30 days. Among 2205 patients receiving at least 1 check-in message, the mean rate of message response was 56.1%, and 79.5% of patients responded to at least 1 message. A total of 985 patients (41.9% of the intervention group) had at least 1 escalation during the 30-day period (eTable 7 in Supplement 2). Among the 2130 patients who continued to the end of the 30-day program, 27.9% answered the NPS question, for a score of 69.
Table 4. Engagement and Satisfaction in the Intervention Arm.
| Variable | No. (%) |
|---|---|
| Observations | 2352 |
| Nonworking telephone number | 92 (3.9) |
| Opt-out | 129 (5.5) |
| Message response rate, mean (SD)a | 0.56 (0.39) (or 56.1%) |
| Engaged (answered ≥1 messages)a | 1752 (79.5) |
| NPS response rateb | 594 (27.9) |
| NPS score | 69 |
Abbreviation: NPS, Net Promoter Score.
Of those who received at least 1 check-in message (n = 2205).
Of those to whom the NPS question was sent, which were those who made it to the end of the 30 days (ie, the total in the intervention arm minus those who opted out or did not have working number) (n = 2131).
Discussion
We built a 30-day automated texting program with the aim of providing an efficient approach to support patients after discharge and, in turn, reduce acute care revisits. The program was integrated within existing practice workflows, demonstrated high levels of engagement and satisfaction, and identified needs in a substantial number of patients. Nonetheless, the patients in the intervention arm were no less likely to be readmitted or need other acute care services.
Enrollment was generally smooth. Patients were identified through an existing workflow, which generates reports that could be uploaded to the platform for automated enrollment; the waiver of research consent allowed us to emulate the experiences in standard clinical deployment. A total of 3.9% of the participants had a nonworking telephone number and 5.5% of the participants opted out of the program; these latter groups were included in an intention-to-treat analysis, which might dilute potential treatment effects, but this was a pragmatic trial (designed to evaluate interventions in routine clinical practice conditions) and most clinical deployments are likely to face similar figures. Engagement with the program was high (79.5% responded to at least 1 message) and consistent with both the pilot study and other work that has compared text messaging with traditional call-based approaches.17
The intervention was conceptually appealing and supported by the encouraging results of the pilot study in which the same intervention in a largely identical population was associated with a 41% reduction in the odds of a 30-day acute care visit in a difference-in-differences analysis.11 Nevertheless, when evaluated in this prospective randomized trial, the intervention showed no effect relative to control in any of the prespecified measures. Why did the intervention work in the pilot study and not in the randomized trial? The results of the pilot study might have been spurious, perhaps because of unmeasured confounding residual to the difference-in-differences approach. Or perhaps the pilot study, run at a single practice, benefitted from idiosyncratic operational elements (eg, a care managers’ unique relationship with the patients, patients’ trust) not well represented in the much wider implementation.
This is why we conduct larger prospective trials rather than rely on either our loyalty to an appealing clinical design or the results of an observationally controlled pilot study. Despite the high engagement and satisfaction with the program, it was nonetheless insufficient to alter patients’ clinical trajectories. That is disappointing, but the reason to approach the problem was that it was already known to be challenging: acute care revisits are the outcome of a complex set of clinical and social forces. An emergent theme of transitional care research has been that the most successful programs have been multimodal and intensive. Remote postdischarge contact—most often via telephone—is a common component of these programs,18,19 but they may also include predischarge education, dedicated transitional care coordinators, home visits, and remote physiologic monitoring.20,21,22,23,24 For those that offer benefit, the active ingredients or the right combination of active ingredients are hard to identify and likely harder to generalize across patients.25,26,27
There is still potential value in an automated texting approach, given that call-based programs are operationally burdensome. We did not test a call replacement approach, and as such saw an overall increase in interactions with the practice during the 30-day period (0.77). Future work should compare a call only–based and text only–based approach directly, with the goal of comparing effectiveness and operational efficiencies. The text-based approaches may have value even if they offer no value beyond human calls.
Strengths and Limitations
This study had strengths, including streamlined enrollment of a large number of patients in a relatively short period of time. The program was well integrated into the staff’s preexisting workflows, and the escalations were routed directly into the EMR where they could be easily seen.
The limitations of this study include that it took place within a single academic health system, although across a large number of practices. It also targeted patients discharged from a wide range of hospitals, and patients—including those in the control arm—may have received overlapping outreach from other health systems. There was no significant difference in effect of the intervention, however, between patients discharged from internal and external hospitals. While external encounter data for the patients in this study were generally input into the EMR as they became available, it is possible that some external visits were missed by this process; this, however, would have occurred randomly across arms. Excluding low-risk patients may have removed a subset of patients who stood to benefit from this type of intervention. The demographic distribution of the population studied, while reflective of the population seen by the health system practices, differs in meaningful ways from national distributions, with higher proportions of Black patients and publicly insured patients. The NPS question was subject to reporting bias, and the response rate to this question was overall low.
Conclusions
In this randomized clinical trial of a 30-day postdischarge automated texting program, we found no significant reduction in acute care revisits. While the program did not improve the clinical outcomes tested, patient engagement and satisfaction were high. Future study should test a call replacement approach to evaluate clinical and operational efficiencies.
Trial Protocol
eTable 1. Schedule of Outreach
eMethods. Script of Messages and Net Promoter Score
eTable 2. Calculation of UPHS Risk Score
eTable 3. 7-Day Utilization
eTable 4. 60-Day Utilization
eTable 5. Regression Model and Time to Acute Care Use
eTable 6. Subgroup Analyses, Primary Outcome
eTable 7. Escalations
Data Sharing Statement
References
- 1.Weiss AJ, Jiang HJ. Overview of clinical conditions with frequent and costly hospital readmissions by payer, 2018. July 20, 2021. Accessed February 9, 2024. https://www.ncbi.nlm.nih.gov/books/NBK573265/ [PubMed]
- 2.Wadhera RK, Joynt Maddox KE, Kazi DS, Shen C, Yeh RW. Hospital revisits within 30 days after discharge for medical conditions targeted by the Hospital Readmissions Reduction Program in the United States: national retrospective analysis. BMJ. 2019;366:l4563. doi: 10.1136/bmj.l4563 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Forster AJ, Murff HJ, Peterson JF, Gandhi TK, Bates DW. The incidence and severity of adverse events affecting patients after discharge from the hospital. Ann Intern Med. 2003;138(3):161-167. doi: 10.7326/0003-4819-138-3-200302040-00007 [DOI] [PubMed] [Google Scholar]
- 4.Dharmarajan K, Hsieh AF, Kulkarni VT, et al. Trajectories of risk after hospitalization for heart failure, acute myocardial infarction, or pneumonia: retrospective cohort study. BMJ. 2015;350:h411. doi: 10.1136/bmj.h411 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Van Spall HGC, Rahman T, Mytton O, et al. Comparative effectiveness of transitional care services in patients discharged from the hospital with heart failure: a systematic review and network meta-analysis. Eur J Heart Fail. 2017;19(11):1427-1443. doi: 10.1002/ejhf.765 [DOI] [PubMed] [Google Scholar]
- 6.Hochman M, Bourgoin A, Saluja S, et al. Environmental Scan of Primary Care–Based Efforts To Reduce Readmissions. Agency for Healthcare Research and Quality; 2019. [Google Scholar]
- 7.Mistiaen P, Poot E. Telephone follow-up, initiated by a hospital-based health professional, for postdischarge problems in patients discharged from hospital to home. Cochrane Database Syst Rev. 2006;2006(4):CD004510. doi: 10.1002/14651858.CD004510.pub3 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Kind AJ, Jensen L, Barczi S, et al. Low-cost transitional care with nurse managers making mostly phone contact with patients cut rehospitalization at a VA hospital. Health Aff (Millwood). 2012;31(12):2659-2668. doi: 10.1377/hlthaff.2012.0366 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Costantino ME, Frey B, Hall B, Painter P. The influence of a postdischarge intervention on reducing hospital readmissions in a Medicare population. Popul Health Manag. 2013;16(5):310-316. doi: 10.1089/pop.2012.0084 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Jayakody A, Bryant J, Carey M, Hobden B, Dodd N, Sanson-Fisher R. Effectiveness of interventions utilising telephone follow up in reducing hospital readmission within 30 days for individuals with chronic disease: a systematic review. BMC Health Serv Res. 2016;16(1):403. doi: 10.1186/s12913-016-1650-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Bressman E, Long JA, Honig K, et al. Evaluation of an automated text message–based program to reduce use of acute health care resources after hospital discharge. JAMA Netw Open. 2022;5(10):e2238293. doi: 10.1001/jamanetworkopen.2022.38293 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Hippisley-Cox J, Coupland C. Predicting risk of emergency admission to hospital using primary care data: derivation and validation of QAdmissions score. BMJ Open. 2013;3(8):e003482. doi: 10.1136/bmjopen-2013-003482 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Billings J, Georghiou T, Blunt I, Bardsley M. Choosing a model to predict hospital admission: an observational study of new variants of predictive models for case finding. BMJ Open. 2013;3(8):e003352. doi: 10.1136/bmjopen-2013-003352 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Hasan O, Meltzer DO, Shaykevich SA, et al. Hospital readmission in general medicine patients: a prediction model. J Gen Intern Med. 2010;25(3):211-219. doi: 10.1007/s11606-009-1196-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.About HSX. HealthShare Exchange. About HSX. Accessed August 19, 2022. https://healthshareexchange.org/about-hsx/
- 16.Reichheld FF. The one number you need to grow. Harv Bus Rev. 2003;81(12):46-54, 124. [PubMed] [Google Scholar]
- 17.Cittanova ML, Chauvier S, Combettes E, et al. Association of automated text messaging with patient response rate after same-day surgery. JAMA Netw Open. 2021;4(1):e2033312. doi: 10.1001/jamanetworkopen.2020.33312 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Altfeld SJ, Shier GE, Rooney M, et al. Effects of an enhanced discharge planning intervention for hospitalized older adults: a randomized trial. Gerontologist. 2013;53(3):430-440. doi: 10.1093/geront/gns109 [DOI] [PubMed] [Google Scholar]
- 19.Koehler BE, Richter KM, Youngblood L, et al. Reduction of 30-day postdischarge hospital readmission or emergency department (ED) visit rates in high-risk elderly medical patients through delivery of a targeted care bundle. J Hosp Med. 2009;4(4):211-218. doi: 10.1002/jhm.427 [DOI] [PubMed] [Google Scholar]
- 20.Davis KK, Mintzer M, Dennison Himmelfarb CR, Hayat MJ, Rotman S, Allen J. Targeted intervention improves knowledge but not self-care or readmissions in heart failure patients with mild cognitive impairment. Eur J Heart Fail. 2012;14(9):1041-1049. doi: 10.1093/eurjhf/hfs096 [DOI] [PubMed] [Google Scholar]
- 21.Naylor MD, Brooten D, Campbell R, et al. Comprehensive discharge planning and home follow-up of hospitalized elders: a randomized clinical trial. JAMA. 1999;281(7):613-620. doi: 10.1001/jama.281.7.613 [DOI] [PubMed] [Google Scholar]
- 22.Kwok T, Lum CM, Chan HS, Ma HM, Lee D, Woo J. A randomized, controlled trial of an intensive community nurse-supported discharge program in preventing hospital readmissions of older patients with chronic lung disease. J Am Geriatr Soc. 2004;52(8):1240-1246. doi: 10.1111/j.1532-5415.2004.52351.x [DOI] [PubMed] [Google Scholar]
- 23.Rytter L, Jakobsen HN, Rønholt F, et al. Comprehensive discharge follow-up in patients’ homes by GPs and district nurses of elderly patients: a randomized controlled trial. Scand J Prim Health Care. 2010;28(3):146-153. doi: 10.3109/02813431003764466 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Kulshreshtha A, Kvedar JC, Goyal A, Halpern EF, Watson AJ. Use of remote monitoring to improve outcomes in patients with heart failure: a pilot trial. Int J Telemed Appl. 2010;2010:870959. doi: 10.1155/2010/870959 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Leppin AL, Gionfriddo MR, Kessler M, et al. Preventing 30-day hospital readmissions: a systematic review and meta-analysis of randomized trials. JAMA Intern Med. 2014;174(7):1095-1107. doi: 10.1001/jamainternmed.2014.1608 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Verhaegh KJ, MacNeil-Vroomen JL, Eslami S, Geerlings SE, de Rooij SE, Buurman BM. Transitional care interventions prevent hospital readmissions for adults with chronic illnesses. Health Aff (Millwood). 2014;33(9):1531-1539. doi: 10.1377/hlthaff.2014.0160 [DOI] [PubMed] [Google Scholar]
- 27.Hall KK, Shoemaker-Hunt S, Hoffman L, et al. Making healthcare safer III: a critical analysis of existing and emerging patient safety practices. agency for healthcare research and quality. March 2020. Accessed February 9, 2024. https://www.ncbi.nlm.nih.gov/books/NBK555526/ [PubMed]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Trial Protocol
eTable 1. Schedule of Outreach
eMethods. Script of Messages and Net Promoter Score
eTable 2. Calculation of UPHS Risk Score
eTable 3. 7-Day Utilization
eTable 4. 60-Day Utilization
eTable 5. Regression Model and Time to Acute Care Use
eTable 6. Subgroup Analyses, Primary Outcome
eTable 7. Escalations
Data Sharing Statement