The Blue Coats program at Penn Medicine is a systemwide initiative designed to amplify the voices of clinical teams to understand and support workforce well-being. Recognizing that traditional system-level strategies often cannot address real-time frontline team needs, the Blue Coats program, launched in August 2022, is a locally embedded group of well-being ambassadors and operational consultants. These trained team members engage directly with clinicians and staff within their local clinical environments - observing, shadowing, and listening to all team members - to identify pain points and bring forth opportunities for improvement for local and senior leadership. The Blue Coats bridge the gap between frontline experience and leadership decision-making by operating within departments and aligning workforce concerns with leadership decision-making. The program utilizes a discovery-to-delivery model: Blue Coats gather actionable insights from clinicians and staff, then collaborate with department leadership to implement customized interventions that enhance workforce well-being and trust, ultimately working to improve retention. Individual solutions informed by Blue Coats' insights are variable by clinical site and have included hiring supply and support staff, redesigning workflows to balance clinical duties, enhancing physical safety infrastructure, and launching local recognition and small division-based reward programs. Early outcome data demonstrated meaningful gains across the domains of hope, trust, and belonging on the part of outpatient clinical staff; the largest improvements, as seen in survey results scored on a five-point Likert scale, include an 11.8% increase in trust (from 3.4 to 3.8) and 7.4% improvement in hope (from 3.4 to 3.65) among nursing staff, and a 24.6% increase in trust among support staff (from 2.85 to 3.55). The Blue Coats initiative has been deployed across diverse clinical units ranging from emergency departments to outpatient clinics. Staff expressed appreciation for being "seen, heard, and supported" - qualities often lost, or difficult to capture, in large health system surveys. Senior clinical leadership recognizes the Blue Coats as a feedback loop and a strategic tool to inform resource allocation, staffing decisions, and cultural investment. Notably, departments engaged with Blue Coats could move quickly from problem discovery to implementation, shortening the traditional lag between insight and action. Perhaps most importantly, the program restored a sense of agency among clinical teams. Many expressed that the presence of Blue Coats signaled a culture shift toward listening, transparency, and shared ownership of workplace challenges. This shift has led to higher trust scores, improvements in self-reported workforce retention, and renewed collaboration and engagement with staff and leadership. The Blue Coats approach offers a scalable, in-house framework for health systems nationally: embed, listen, act, and evolve. As health care continues to confront burnout and operational strain, this model demonstrates how relational, human-centered design can fuel the future of the workforce.

Background:The "fourth trimester," or postpartum time period, remains a critical phase of pregnancy that significantly impacts parents and newborns. Care poses challenges due to complex individual needs as well as low attendance rates at routine appointments. A comprehensive technological solution could provide a holistic and equitable solution to meet care goals. Objective:This paper describes the development of patient engagement data with a novel postpartum conversational agent that uses natural language processing to support patients post partum. Methods:We report on the development of a postpartum conversational agent from concept to usable product as well as the patient engagement with this technology. Content for the program was developed using patient- and provider-based input and clinical algorithms. Our program offered 2-way communication to patients and details on physical recovery, lactation support, infant care, and warning signs for problems. This was iterated upon by our core clinical team and an external expert clinical panel before being tested on patients. Patients eligible for discharge around 24 hours after delivery who had delivered a singleton full-term infant vaginally were offered use of the program. Patient demographics, accuracy, and patient engagement were collected over the first 6 months of use. Results:A total of 290 patients used our conversational agent over the first 6 months, of which 112 (38.6%) were first time parents and 162 (56%) were Black. In total, 286 (98.6%) patients interacted with the platform at least once, 271 patients (93.4%) completed at least one survey, and 151 (52%) patients asked a question. First time parents and those breastfeeding their infants had higher rates of engagement overall. Black patients were more likely to promote the program than White patients (P=.047). The overall accuracy of the conversational agent during the first 6 months was 77%. Conclusions:It is possible to develop a comprehensive, automated postpartum conversational agent. The use of such a technology to support patients postdischarge appears to be acceptable with very high engagement and patient satisfaction.

Patients with cancer spend considerable time commuting to, waiting for, and receiving health care. Patient-reported outcomes have been collected electronically to monitor patients for toxicity related to treatment, but, to the authors' knowledge, they have not been used as a strategy to minimize patients' time spent on cancer care by streamlining care delivery. Researchers at Penn Medicine set an objective to assess the effectiveness and implementation of a text message-based symptom reporting electronic triage (e-triage) versus usual care to minimize the time toxicity associated with ambulatory cancer care. The methods employed included a hybrid type 1 effectiveness-implementation, unblinded, randomized controlled trial and sequential mixed-methods study, which was conducted between December 1, 2021, and December 31, 2022, with a follow-up period of 3 months or three visits (whichever came first, but all within the 2-year window). Adult patients with solid tumors receiving single-agent immune checkpoint inhibitors (ICIs) with access to a text-messaging device were enrolled, with a target sample size of 176. The intervention was a symptom-based e-triage via mobile text messaging combined with routine laboratory testing. Participants in the e-triage group with normal bloodwork and no symptoms of drug toxicity on e-triage were eligible to fast-track to ICI infusion, bypassing the pretreatment office visit. The primary end point was total time per ambulatory encounter; secondary end points included wait time, ED or hospital visits, health-related quality of life, patient satisfaction, and implementation (reach and fidelity). Implementation readiness (acceptability, appropriateness, and feasibility), barriers, and facilitators were evaluated in a mixed-methods analysis among treating oncologists, measured via surveys and focus groups. For the study, 40 patients were randomly assigned, of which 31 were evaluated for the primary end point; the median age among the 40 participants was 67.5 years of age (interquartile range 59.5-71.5 years of age), 80.0% were male, and 84.6% were white. Those randomly assigned to the e-triage group of the pilot randomized controlled trial (n=19, n=16 evaluable) had an average of 66.0 minutes less total time (95% confidence interval [CI], -123.7 to -8.08 minutes; P=0.03) and 30.1 minutes less wait time (95% CI, -60.9 to 1.1 minutes; P=0.08) per encounter, than those in usual care (n=21 randomly assigned, n=15 evaluable). ED or hospital visits, health-related quality of life, and patient satisfaction scores were similar. In the mixed-methods study, oncologists (n=31, 17 completed the survey) found the e-triage acceptable (mean 3.8, standard error [SE] 0.1), appropriate (mean 3.8, SE 0.1), and feasible (mean 3.9, SE 0.1) on a 5-point Likert scale of agreeability. Perceived barriers to uptake included challenges in patient identification, potential for drug toxicity underreporting, and reimbursement concerns. The authors conclude that the results of this pilot randomized controlled trial of a text message-based e-triage supports further investigation into the use of text message-based symptom reporting by patients as a strategy to safely assess readiness for treatment and thus reduce the time toxicity associated with cancer care.

This article describes how Penn Medicine navigated complex information technology challenges involved in processing electronically faxed patient information to improve workforce well-being and patient care. It details the decision-making process in choosing whether to build digital tools internally versus partnering with third-party vendors. Penn Medicine launched its homegrown technology platform, called coordn8, in October 2022. It automates the processing of electronic faxes received from other providers and health systems and integrates them into Penn Medicine's electronic health record, significantly reducing administrative burden. The platform has processed more than 370,000 faxes, saving the health system more than 8500 hours of staff time to date. Staff satisfaction improved markedly according to surveys, rising from 35% to 60%. The freed capacity was redirected to higher-value clinical tasks. This clinician-driven initiative demonstrates how automating repetitive high-volume tasks such as patient identification and record indexing can reduce workforce burden, enhance the accuracy and timeliness of filing electronic faxes, and create a scalable system that minimizes manual effort, streamlines workflows, and ensures timely access to critical patient information.

Distracted driving is responsible for nearly 1 million crashes each year in the United States alone, and a major source of driver distraction is handheld phone use. We conducted a randomized, controlled trial to compare the effectiveness of interventions designed to create sustained reductions in handheld use while driving (NCT04587609). Participants were 1,653 consenting Progressive® Snapshot® usage-based auto insurance customers ages 18 to 77 who averaged at least 2 min/h of handheld use while driving in the month prior to study invitation. They were randomly assigned to one of five arms for a 10-wk intervention period. Arm 1 (control) got education about the risks of handheld phone use, as did the other arms. Arm 2 got a free phone mount to facilitate hands-free use. Arm 3 got the mount plus a commitment exercise and tips for hands-free use. Arm 4 got the mount, commitment, and tips plus weekly goal gamification and social competition. Arm 5 was the same as Arm 4, plus offered behaviorally designed financial incentives. Postintervention, participants were monitored until the end of their insurance rating period, 25 to 65 d more. Outcome differences were measured using fractional logistic regression. Arm 4 participants, who received gamification and competition, reduced their handheld use by 20.5% relative to control ( P < 0.001); Arm 5 participants, who additionally received financial incentives, reduced their use by 27.6% ( P < 0.001). Both groups sustained these reductions through the end of their insurance rating period.

PURPOSE Capecitabine is an oral chemotherapy used to treat many gastrointestinal cancers. Its complex dosing and narrow therapeutic index make medication adherence and toxicity management crucial for quality care. METHODS We conducted a pilot study of PENNY-GI, a mobile phone text messaging-based chatbot that leverages algorithmic surveys and natural language processing to promote medication adherence and toxicity management among patients with gastrointestinal cancers on capecitabine. Eligibility initially included all capecitabine-containing regimens but was subsequently restricted to capecitabine monotherapy because of challenges in integrating PENNY-GI with radiation and intravenous chemotherapy schedules. We used design thinking principles and real-time data on safety, accuracy, and usefulness to make iterative refinements to PENNY-GI with the goal of minimizing the proportion of text messaging exchanges with incorrect medication or symptom management recommendations. All patients were invited to participate in structured exit interviews to provide feedback on PENNY-GI. RESULTS We enrolled 40 patients (median age 64.5 years, 52.5% male, 62.5% White, 55.0% with colorectal cancer, 50.0% on capecitabine monotherapy). We identified 284 of 3,895 (7.3%) medication-related and 13 of 527 (2.5%) symptom-related text messaging exchanges with incorrect recommendations. In exit interviews with 24 patients, participants reported finding the medication reminders reliable and user-friendly, but the symptom management tool was too simplistic to be helpful. CONCLUSION Although PENNY-GI provided accurate recommendations in >90% of text messaging exchanges, we identified multiple limitations with respect to the intervention's generalizability, usefulness, and scalability. Lessons from this pilot study should inform future efforts to develop and implement digital health interventions in oncology.

A patient's hospital stay is too often wrapped in fear and worry. Healthcare leaders have emphasized the immense need to improve patient experience and address patients’ individual needs. In addition to helping with the medical aspect of healing, we believe health systems can encourage and empower providers to perform acts of kindness to help elevate the otherwise stressful experience of being hospitalized. We describe an initiative focused on tailoring joyful surprises, like unexpected gifts, to help support both patients and clinicians, aiming to improve patient experience and satisfaction while reducing provider burnout. In sharing the stories of the interactions between the providers and patients, it is clear that not only has this program brought joy to our patients, but that it has also helped reconnect our providers with their sense of meaning and purpose in caring for people and meeting their needs. Thus, we herein describe a patient-centered initiative that enables healthcare providers to provide unique and joyful surprises for their patients in a manner that is readily scalable, cost-effective, accessible, and deeply impactful.

Importance Handheld phone use while driving is a major factor in vehicle crashes. Scalable interventions are needed to encourage drivers not to use their phones. Objective To test whether interventions involving social comparison feedback and/or financial incentives can reduce drivers' handheld phone use. Design, Setting, and Participants In a randomized clinical trial, interventions were administered nationwide in the US via a mobile application in the context of a usage-based insurance program (Snapshot Mobile application). Customers were eligible to be invited to participate in the study if enrolled in the usage-based insurance program for 30 to 70 days. The study was conducted from May 13 to June 30, 2019. Analysis was completed December 22, 2023. Interventions Participants were randomly assigned to 1 of 6 trial arms for a 7-week intervention period: (1) control; (2) feedback, with weekly push notification about their handheld phone use compared with that of similar others; (3) standard incentive, with a maximum $50 award at the end of the intervention based on how their handheld phone use compared with similar others; (4) standard incentive plus feedback, combining interventions of arms 2 and 3; (5) reframed incentive plus feedback, with a maximum $7.15 award each week, framed as participant's to lose; and (6) doubled reframed incentive plus feedback, a maximum $14.29 weekly loss-framed award. Main Outcome and Measure Proportion of drive time engaged in handheld phone use in seconds per hour (s/h) of driving. Analyses were conducted with the intention-to-treat approach. Results Of 17 663 customers invited by email to participate, 2109 opted in and were randomized. A total of 2020 drivers finished the intervention period (68.0% female; median age, 30 [IQR, 25-39] years). Median baseline handheld phone use was 216 (IQR, 72-480) s/h. Relative to control, feedback and standard incentive participants did not reduce their handheld phone use. Standard incentive plus feedback participants reduced their use by -38 (95% CI, -69 to -8) s/h (P = .045); reframed incentive plus feedback participants reduced their use by -56 (95% CI, -87 to -26) s/h (P < .001); and doubled reframed incentive plus feedback participants reduced their use by -42 s/h (95% CI, -72 to -13 s/h; P = .007). The 5 active treatment arms did not differ significantly from each other. Conclusions and Relevance In this randomized clinical trial, providing social comparison feedback plus incentives reduced handheld phone use while individuals were driving.

Objectives Acute care ophthalmic clinics often suffer from inefficient triage, leading to suboptimal patient access and resource utilization. This study reports the preliminary results of a novel, symptom-based, patient-directed, online triage tool developed to address the most common acute ophthalmic diagnoses and associated presenting symptoms. Methods A retrospective chart review of patients who presented to a tertiary academic medical center's urgent eye clinic after being referred for an urgent, semi-urgent, or nonurgent visit by the ophthalmic triage tool between January 1, 2021 and January 1, 2022 was performed. Concordance between triage category and severity of diagnosis on the subsequent clinic visit was assessed. Results The online triage tool was utilized 1,370 and 95 times, by the call center administrators (phone triage group) and patients directly (web triage group), respectively. Of all patients triaged with the tool, 8.50% were deemed urgent, 59.2% semi-urgent, and 32.3% nonurgent. At the subsequent clinic visit, the history of present illness had significant agreement with symptoms reported to the triage tool (99.3% agreement, weighted kappa = 0.980, p < 0.001). The triage algorithm also had significant agreement with the severity of the physician diagnosis (97.0% agreement, weighted kappa = 0.912, p < 0.001). Zero patients were found to have a diagnosis on exam that should have corresponded to a higher urgency level on the triage tool. Conclusion The automated ophthalmic triage algorithm was able to safely and effectively triage patients based on symptoms. Future work should focus on the utility of this tool to reduce nonurgent patient load in urgent clinical settings and to improve access for patients who require urgent medical care.

301 Background: Innovative strategies to mitigate the time toxicity of cancer therapy are desperately needed. We have shown that a text-based e-triage can utilize patient-reported outcomes to identify patients without immune checkpoint inhibitor (ICI) toxicity who could safely fast-track to ICI infusion without a pre-infusion office visit. We report the efficacy of e-triage versus usual care to minimize the time toxicity of cancer care. Methods: This hybrid type 1 effectiveness-implementation randomized controlled trial was conducted at Penn Medicine between December 2021 and December 2022. Eligible patients spoke English, were receiving single agent ICI for a solid tumor, and had access to a mobile device with text messaging. The e-triage arm included ICI toxicity symptom assessment via two-way text messaging 96 hours prior to scheduled ICI and routine laboratory testing. Patients on the e-triage arm with normal bloodwork and no symptoms identified by the e-triage were eligible to fast-track to ICI infusion, bypassing the pre-treatment office visit. Usual care was standard office visits. Primary endpoint was care time (total time per ambulatory encounter including commute-, wait-, infusion-, and lab-times). Secondary endpoints were patient wait time per encounter, incident emergency department (ED) or hospital visits during follow-up, health related quality of life (HRQOL) measured by The Functional Assessment of Cancer Therapy-General, and patient satisfaction measured by PSQ-18. Differences in proportions of ED or hospital visits between arms were evaluated by the Fisher’s exact test. Linear mixed-effects models with random intercepts for each individual evaluated differences in all other endpoints between arms, accounting for within-patient correlation. Implementation outcomes were adoption (# patients who participated in the trial out of # approached) and fidelity (# patients on the intervention arm who followed their triage assignment). Acceptability, feasibility, and appropriateness outcomes were reported previously. Results: Among 152 eligible patients, 51 consented onto study (adoption rate 33.6%) and 40 were randomized (n=21, usual care; n=19, e-triage). Patient characteristics will be presented. Of 52 encounters on the e-triage arm, 23 adhered to their e-triage assessment (fidelity rate 44.2%). Compared with the usual care arm, patients on the e-triage arm had an average of 70 minutes less care time per encounter (95% CI -123.7 to –8.08 minutes, p=0.03) and 30.1 minutes less wait time (95% CI -61 to 1.1 minutes, p=0.08). The incidence of ED or hospital visits did not differ by treatment arm (usual care, n=2, 12.5% vs. intervention, n=3, 20%, p=0.65). HRQOL and patient satisfaction scores were similar by treatment arm. Conclusions: Results from this pilot trial support further work to optimize the design and implementation of a mobile e-triage program to personalize cancer care delivery and minimize time toxicity. Clinical trial information: NCT05134636 .

9115 Background: Helping patients to manage symptoms and adhere to oral anticancer agents (OACAs) is a major challenge in oncology. OACAs cause side effects that can lead to suboptimal adherence if not optimally managed, contributing to decreased effectiveness. Low-cost, text messaging approaches have shown promise, but have not been robustly studied in oncology. Guided by principles from implementation and behavioral science, we developed and tested the effect of an augmented intelligence chatbot on OACA adherence and symptom burden in patients with advanced lung cancer. Methods: We conducted a two-arm pilot randomized trial (NCT04347161) to evaluate the effect of the chatbot on OACA adherence and symptom burden compared to usual care. The chatbot engages patients via text messaging and applies natural language processing and machine learning to learn from interactions. Core functionalities include: 1) motivational daily adherence reminders, 2) longitudinal symptom monitoring with self-management support, and 3) bidirectional communication with clinical teams. Participants included English-speaking patients with advanced lung cancer treated with any of 9 OACAs targeting EGFR, ALK, or ROS-1. The primary outcome was 12-week adherence, measured using the microelectronic monitoring system (MEMS) and defined dichotomously if the patient achieved ≥95% adherent days or not. Secondary outcomes were assessed at baseline and 12 weeks using validated survey instruments, including symptom burden using the Edmonton Symptom Assessment System Total Distress Score (range 0-90), health-related quality of life (HRQOL) using EQ-5D-3L (range 0-100), and usability using Health-ITUES (range 1-5). We used multivariable logistic regression adjusting for stratification variables to test the chatbot’s effect on adherence (intent-to-treat analysis) and evaluated mean differences (by arm) in secondary outcomes using Fisher’s exact test. Results: From February 2021 to August 2022, 75 patients across 4 sites enrolled (median age 65 years, 64% female, 88% White, 21.3% high school education or lower); 50.7% (n=38) were randomized to intervention. Compared to usual care, we observed no significant differences in adherence in the intervention arm (78.8% vs 81.8%; aOR=1.7 95% CI: 0.3-9.4). However, in contrast to those in usual care, participants in the intervention arm had significantly greater decreases in symptom burden (mean difference: -2.7 vs 2.6; p<0.05) and increases in HRQOL (mean difference: 4.1 vs -4.8; p<0.05) from baseline to 12 weeks. Overall chatbot usability was high (mean score=3.9). Conclusions: In this pilot randomized trial, an augmented intelligence chatbot successfully reduced symptom burden and improved HRQOL but did not significantly alter OACA adherence. Chatbots are a potentially scalable strategy for improving symptom management that warrant study in larger randomized trials. Clinical trial information: NCT04347161 .

With an increase in teledermatology, accelerated by the Covid-19 pandemic, leaders at Penn Medicine recognized a need to improve the value of videoconference-based visits by developing an innovative process to simplify the method by which patients can share photos in advance of the visit while automating instructions and reminders for patients to reduce the burden on staff. The process enables patients to submit photos via cell phone text attachments rather than using the traditional patient portal, if they prefer. Initial results show improvements in patient compliance and engagement, photo quality, and teledermatology encounters.

IMPORTANCE COVID-19 vaccine uptake among urban populations remains low. OBJECTIVE To evaluate whether text messaging with outbound or inbound scheduling and behaviorally informed content might increase COVID-19 vaccine uptake. DESIGN, SETTING, AND PARTICIPANTS This randomized clinical trial with a factorial design was conducted from April 29 to July 6, 2021, in an urban academic health system. The trial comprised 16 045 patients at least 18 years of age in Philadelphia, Pennsylvania, with at least 1 primary care visit in the past 5 years, or a future scheduled primary care visit within the next 3 months, who were unresponsive to prior outreach. The study was prespecified in the trial protocol, and data were obtained from the intent-to-treat population. INTERVENTIONS Eligible patients were randomly assigned in a 1:20:20 ratio to (1) outbound telephone call only by call center, (2) text message and outbound telephone call by call center to those who respond, or (3) text message, with patients instructed to make an inbound telephone call to a hotline. Patients in groups 2 and 3 were concurrently randomly assigned in a 1:1:1:1 ratio to receive different content: standard messaging, clinician endorsement (eg, "Dr. XXX recommends"), scarcity ("limited supply available"), or endowment framing ("We have reserved a COVID-19 vaccine appointment for you"). MAIN OUTCOMES AND MEASURES The primary outcome was the proportion of patients who completed the first dose of the COVID-19 vaccine within 1 month, according to the electronic health record. Secondary outcomes were the completion of the first dose within 2 months and completion of the vaccination series within 2 months of initial outreach. Additional outcomes included the percentage of patients with invalid cell phone numbers (wrong number or nontextable), no response to text messaging, the percentage of patients scheduled for the vaccine, text message responses, and the number of telephone calls made by the access center. Analysis was on an intention-to-treat basis. RESULTS Among the 16 045 patients included, the mean (SD) age was 36.9 (11.1) years; 9418 (58.7%) were women; 12 869 (80.2%) had commercial insurance, and 2283 (14.2%) were insured by Medicaid; 8345 (52.0%) were White. 4706 (29.3%) were Black, and 967 (6.0%) were Hispanic or Latino. At 1 month, 14 of 390 patients (3.6% [95% CI. 1.7%-5.4%]) in the outbound telephone call-only group completed 1 vaccine dose, as did 243 of 7890 patients (3.1% [95% CI. 2.7%-3.5%]) in the text plus outbound call group (absolute difference, -0.5% [95% CI, -2.4% to 1.4%]; P = .57) and 253 of 7765 patients (3.3% [95% CI, 2.9%-3.7%]) in the text plus inbound call group (absolute difference. - 0.3% [95% CI, -2.2% to 1.6%]; P = .72). Among the 15 655 patients receiving text messaging, 118 of 3889 patients (3.0% [95% CI, 2.5%-3.6%)) in the standard messaging group completed 1 vaccine dose, as did 135 of 3920 patients (3.4% [95% CI, 2.9%-4.0%]) in the clinician endorsement group (absolute difference, 0.4% [95% CI, -0.4% to 12%1 P = .31), 100 of 3911 patients (2.6% [95% CI, 2.1%-3.1%]) in the scarcity group (absolute difference, -0.5% [95% CI, -1.2% to 0.3%]; P = 20), and 143 of 3935 patients (3.6% [95% CI, 3.0%-4.2%]) in the endowment group (absolute difference, 0.6% [95% CI, -0.2% to 1.4%]; P = .14). CONCLUSIONS AND RELEVANCE There was no detectable increase in vaccination uptake among patients receiving text messaging compared with telephone calls only or behaviorally informed message content.

Statement of Purpose To compare the effectiveness of novel interventions aimed at building the habit of putting down one’s phone while driving, among drivers eligible for a smartphone telematics-based auto-insurance rate. Methods/Approach We enrolled 1,670 Progressive Snapshot usage-based auto insurance customers in a 10-week randomized trial (NCT04587609) to test the additive impact of interventions designed to reduce handheld phone use while driving. Arm 1 (control) educated participants about the risks of handheld use. Arm 2 also gave them a free phone mount. Arm 3 included goal commitment and habit tips. Arm 4 added gamification and social competition, and Arm 5 linked performance to financial incentives ($11 average/driver). Post-intervention, participants were monitored for 25–65 more days. Outcome differences were measured using fractional logistic regression with Holm adjustment for multiple comparisons. Results Participants had a mean age of 33 (18 to 77); 66% identified as white, 22% as Black, 4% as Asian, and 15% as Hispanic. Mean overall baseline handheld phone use was 388 sec/hr. During the intervention, Arm 2 (phone mount) had similar handheld use compared to control. Arm 3 (commitment + tips) had 26 sec/hr less use than control, Arm 4 (gamification + competition) had 51 sec/hr less, and Arm 5 (incentives) had 90 sec/hr less. After Holm adjustment, Arm 5 remained significantly different from control during the intervention (25% relative reduction, p < 0.001) and after (23%, p < 0.005). Subgroup analyses found that Arm 5 was successful across all ages. Conclusion A multi-component behavioral intervention focusing on habit formation led to a sustained, one-quarter decrease in a common form of distracted driving. Significance Given its successful implementation in a large usage-based auto insurance program, this intervention has significant potential for reducing a leading crash risk if brought to scale.

Couples with infertility experience psychological distress associated with the prolonged period of recognizing a need for fertility assistance and the long journey of evaluation and management. In August 2019, Penn Medicine began developing Fast Track to Fertility, a novel model of care delivery to improve access to fertility care and decrease time to completion of fertility workup and, by June 2021, deployed division-wide implementation of the new care model. Advanced practice providers conducted new-patient visits via telehealth, and patients were offered enrollment in a texting platform to assist with completion of a complex workup for both partners. Penn Medicine initially used a fake back-end texting platform (to test and authenticate various iterations) and then transitioned to an artificial intelligence-augmented semi-automation approach to facilitate timely completion of the workup. The redesigned model increased annualized new patient access by 23.8% (to 5,570 from 4,500) and decreased the time to initiation of treatment by approximately 50% (to 41 days from 97 days). The addition of educational materials and anticipatory guidance via text contributed to increased patient engagement (to 80% from 60%) and high satisfaction scores (measured by Net Promoter Scores greater than 70).

Automated chatbots offer the promise of reducing reliance on personnel while maintaining or increasing effective and comfortable customer engagement. But accuracy standards are higher and quality assurance more difficult in health care settings, where patients are the customers. The University of Pennsylvania Health System implemented a prototype chatbot to reduce message burden on clinicians in a text-based remote hypertension management program. The chatbot accurately triaged 99% of patient messages (1,379 of 1,393), with most messages, nearly 75% (1,073), not requiring escalation to the clinician. The authors share both their approach for building while doing using innovation methodology and the lessons learned in developing a chatbot for clinical care.

Gaulton JS, et al. BMJ Innov 2021;0:1–5. doi:10.1136/bmjinnov-2021-000791 Neonatology, Jefferson Health–Abington, Abington, Pennsylvania, USA HUP Obstetrics and Gynecology, Penn Medicine, Philadelphia, Pennsylvania, USA Center for Health Care Innovation, Penn Medicine, Philadelphia, Pennsylvania, USA Division of Neonatology, Department of Pediatrics, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania, USA Office of the Dean of the University of Pennsylvania, Penn Medicine, Philadelphia, Pennsylvania, USA Emergency Medicine, University of Pennsylvania Health System, Philadelphia, Pennsylvania, USA

6527 Background: Patients with cancer spend substantial time receiving cancer care. There is a need for innovative strategies to decrease the time burden of cancer therapy. The current care model consists largely of in-person visits to assess treatment toxicity. Most patients treated with immunotherapy, however, do not experience substantial toxicity. We designed and evaluated a text-based instrument to identify patients without symptoms of immunotherapy toxicity. This instrument has the potential to be combined with lab assessment to identify individuals who can safely proceed directly to treatment, lessening the need for in-person office visits. Methods: This cross-sectional study evaluated the performance characteristics of a text-based instrument to identify patient-reported immunotherapy toxicity, against the gold standard in-person provider assessment documented in the electronic medical record (EMR). Those eligible for inclusion spoke English, were receiving single agent immune checkpoint blockade for a solid tumor, and had access to a mobile device with text messaging capabilities. The instrument contained 16 questions adapted from the NCI Pro-CTCAE and was administered via text-message 96 hours prior to the patient’s scheduled infusion visit. Patient perspectives were quantified via a 13-item questionnaire. Results: Between October 1 and November 25, 2021, 50 patients enrolled in the study, and 45 patients completed the instrument (90% response). The median age was 68 (IQR 60-72), 31 (62%) were male, and 44 (88%) were white. Most patients received either pembrolizumab (n=27, 54%) or nivolumab (n=17, 34%) in the palliative setting (n=37, 74%) for genitourinary (n=15, 30%), lung (n=13, 26%), or skin (n=11, 22%) cancer. Patients who completed the instrument were younger (median age 67 vs 76) than those who did not complete the instrument. The prevalence of immune related toxicity documented in the EMR was 57.8%. The sensitivity and negative predictive value of the instrument was 100% (95% CI 0.87-1.00) and 100% (95% CI 0.664-1.00), respectively; other accuracy parameters are presented in the Table. The patient user questionnaire revealed that visual impairment, lack of access to a smart phone, and lack of recognition of the instrument were barriers to completion. Conclusions: A text-based platform is both feasible and effective at identifying patients who are not experiencing symptoms of immune toxicity, and when combined with lab assessment, can eliminate office visits for up to 47% of patients. A prospective clinical trial to assess this is underway (NCT05134636). [Table: see text]

To develop a patient‐centered text message‐based platform that promotes self‐management of symptoms of interstitial cystitis/bladder pain syndrome (IC/BPS).

424 Background: Capecitabine (cape), an oral chemotherapy, is the treatment backbone for many GI cancers. Its complex dosing and narrow therapeutic index make medication adherence and toxicity management crucial for quality patient care. Methods: We conducted a feasibility study of “Penny,” an augmented intelligence mobile phone chatbot that leverages algorithmic surveys and natural language processing (NLP) to engage with patients in conversational, bi-directional text messages. Penny provides patients with medication reminders tailored to their prescribed doses and schedules, sends weekly check-in messages, manages low-grade symptoms in real time, and escalates high-grade symptoms for resolution by the clinical team. Patients ≥18 years old receiving cape for the treatment of a GI cancer were accrued in sequential cohorts of 20 for participation over a three-month period. Feasibility was assessed during planned interim analyses and was predefined as the completion of a 20-patient cohort without a safety event, defined as the communication of incorrect medication or symptom management recommendations as ascertained by two independent clinician reviewers (Κ = 0.89). Secondary outcomes included patient-reported adherence and engagement with the chatbot’s weekly check-in messages. At study completion, all patients were invited to participate in structured interviews to provide feedback on the platform. Results: The first cohort of 20 patients was enrolled from 8/2021 to 4/2022; the median age was 57 years, and patients were primarily female (55%), white (65%), commercially insured (55%), and had colorectal cancer (55%). Chemotherapy regimens included cape with oxaliplatin (50%), concurrent RT (30%), temozolomide (5%), and monotherapy (15%). A total of 2,149 text messaging exchanges were reviewed with 150 (7%) medication-related and 9 (0.4%) symptom-related safety events identified. Most medication-related safety events were due to misalignment with prescribed chemotherapy schedules (55%) and doses (32%). Symptom-related safety events were primarily due to the misinterpretation of patient messages by Penny’s NLP functionality (89%). Average patient-reported adherence was 67% (SD 27%), and patients engaged with 27% (SD 24%) of the chatbot’s weekly check-in messages. In post-study interviews with 12 patients, participants reported that the medication reminders were reliable and user-friendly, whereas the symptom management tool was too simplistic to be helpful. Conclusions: Although Penny has not yet met its feasibility endpoint, the lessons learned from this first cohort have informed further refinements to the platform. Ongoing efforts aim to integrate Penny with the electronic health record and further train the chatbot’s NLP functionality to minimize medication- and symptom-related safety events, respectively. Clinical trial information: NCT05113264.