Nathaniel Erskine

Data extraction is a crucial, yet labor‐intensive and error‐prone part of evidence synthesis. To date, efforts to harness machine learning for enhancing efficiency of the data extraction process have fallen short of achieving sufficient accuracy and usability. With the release of large language models (LLMs), new possibilities have emerged to increase efficiency and accuracy of data extraction for evidence synthesis. The objective of this proof‐of‐concept study was to assess the performance of an LLM (Claude 2) in extracting data elements from published studies, compared with human data extraction as employed in systematic reviews. Our analysis utilized a convenience sample of 10 English‐language, open‐access publications of randomized controlled trials included in a single systematic review. We selected 16 distinct types of data, posing varying degrees of difficulty (160 data elements across 10 studies). We used the browser version of Claude 2 to upload the portable document format of each publication and then prompted the model for each data element. Across 160 data elements, Claude 2 demonstrated an overall accuracy of 96.3% with a high test–retest reliability (replication 1: 96.9%; replication 2: 95.0% accuracy). Overall, Claude 2 made 6 errors on 160 data items. The most common errors (n = 4) were missed data items. Importantly, Claude 2's ease of use was high; it required no technical expertise or labeled training data for effective operation (i.e., zero‐shot learning). Based on findings of our proof‐of‐concept study, leveraging LLMs has the potential to substantially enhance the efficiency and accuracy of data extraction for evidence syntheses.

Background: Cardiac rehabilitation programs, consisting of exercise training and disease management interventions, reduce morbidity and mortality after acute myocardial infarction. Objective: In this pilot study, we aimed to developed and assess the feasibility of delivering a health watch–informed 12-week cardiac telerehabilitation program to acute myocardial infarction survivors who declined to participate in center-based cardiac rehabilitation. Methods: We enrolled patients hospitalized after acute myocardial infarction at an academic medical center who were eligible for but declined to participate in center-based cardiac rehabilitation. Each participant underwent a baseline exercise stress test. Participants received a health watch, which monitored heart rate and physical activity, and a tablet computer with an app that displayed progress toward accomplishing weekly walking and exercise goals. Results were transmitted to a cardiac rehabilitation nurse via a secure connection. For 12 weeks, participants exercised at home and also participated in weekly phone counseling sessions with the nurse, who provided personalized cardiac rehabilitation solutions and standard cardiac rehabilitation education. We assessed usability of the system, adherence to weekly exercise and walking goals, counseling session attendance, and disease-specific quality of life. Results: Of 18 participants (age: mean 59 years, SD 7) who completed the 12-week telerehabilitation program, 6 (33%) were women, and 6 (33%) had ST-elevation myocardial infarction. Participants wore the health watch for a median of 12.7 hours (IQR 11.1, 13.8) per day and completed a median of 86% of exercise goals. Participants, on average, walked 121 minutes per week (SD 175) and spent 189 minutes per week (SD 210) in their target exercise heart rate zone. Overall, participants found the system to be highly usable (System Usability Scale score: median 83, IQR 65, 100). Conclusions: This pilot study established the feasibility of delivering cardiac telerehabilitation at home to acute myocardial infarction survivors via a health watch–based program and telephone counseling sessions. Usability and adherence to health watch use, exercise recommendations, and counseling sessions were high. Further studies are warranted to compare patient outcomes and health care resource utilization between center-based rehabilitation and telerehabilitation.

BACKGROUND Cardiac rehabilitation programs, consisting of exercise training and disease management interventions, reduce morbidity and mortality after acute myocardial infarction. OBJECTIVE In this pilot study, we aimed to developed and assess the feasibility of delivering a health watch–informed 12-week cardiac telerehabilitation program to acute myocardial infarction survivors who declined to participate in center-based cardiac rehabilitation. METHODS We enrolled patients hospitalized after acute myocardial infarction at an academic medical center who were eligible for but declined to participate in center-based cardiac rehabilitation. Each participant underwent a baseline exercise stress test. Participants received a health watch, which monitored heart rate and physical activity, and a tablet computer with an app that displayed progress toward accomplishing weekly walking and exercise goals. Results were transmitted to a cardiac rehabilitation nurse via a secure connection. For 12 weeks, participants exercised at home and also participated in weekly phone counseling sessions with the nurse, who provided personalized cardiac rehabilitation solutions and standard cardiac rehabilitation education. We assessed usability of the system, adherence to weekly exercise and walking goals, counseling session attendance, and disease-specific quality of life. RESULTS Of 18 participants (age: mean 59 years, SD 7) who completed the 12-week telerehabilitation program, 6 (33%) were women, and 6 (33%) had ST-elevation myocardial infarction. Participants wore the health watch for a median of 12.7 hours (IQR 11.1, 13.8) per day and completed a median of 86% of exercise goals. Participants, on average, walked 121 minutes per week (SD 175) and spent 189 minutes per week (SD 210) in their target exercise heart rate zone. Overall, participants found the system to be highly usable (System Usability Scale score: median 83, IQR 65, 100). CONCLUSIONS This pilot study established the feasibility of delivering cardiac telerehabilitation at home to acute myocardial infarction survivors via a health watch–based program and telephone counseling sessions. Usability and adherence to health watch use, exercise recommendations, and counseling sessions were high. Further studies are warranted to compare patient outcomes and health care resource utilization between center-based rehabilitation and telerehabilitation.

Background: Suboptimal care transitions from the hospital to home (transitions) may lead to adverse patient outcomes. As reimbursement becomes increasingly linked with clinical and patient-centered outcomes, providers have strong incentives to better manage transitions. Patients may offer key insights to better assess and improve transitions.Methods: In this qualitative study, we performed semi-structured concept elicitation interviews with patients experiencing a transition following an unplanned hospitalization in medical and surgical wards at a single academic medical center. Three analysts independently analyzed transcripts to identify themes within and outside the medical centers that influenced patients’ transition.Results: Our analyses of interview transcripts of patients (n=22) identified 4 themes related to actions of discharging medical centers and 5 themes outside of the direct purview of the discharging hospital that influence transitions. In the medical center, participants generally described positive roles for quality patient-centered care, opportunities for patient participation, comprehensive discharge education and coordination of medical services on their transitions. Outside of the medical center, participants reported that having caregiver support, social support, health literacy, adequate health insurance and accessibility to healthcare and non-healthcare resources following hospital discharge assisted their transitions.Conclusions: This qualitative study suggests that successful transitions result from actions of medical centers to engage and assist patients, as well as from patients’ access to sociodemographic and psychosocial resources. Future work on strategies to develop better patient-centered transitions and their impact on patient outcomes is warranted.

Background: Patient activation comprises the knowledge, skills, and confidence for self-care and may lead to better health outcomes. Objectives: We examined the relationship between patient activation and changes in health-related quality of life (HRQOL) after hospitalization for an acute coronary syndrome (ACS). Methods: We studied patients from 6 medical centers in central Massachusetts and Georgia who had been hospitalized for an ACS between 2011 and 2013. At 1 month after hospital discharge, the patients completed the 6-item Patient Activation Measure and were categorized into 4 levels of activation. Multinomial logistic regression analyses compared activation level with clinically meaningful changes (>= 3.0 points, generic; >= 10.0 points, disease-specific) in generic physical (SF-36v2 Physical Component Summary [PCS]), generic mental (SF-36v2 Mental Component Summary [MCS]), and disease-specific (Seattle Angina Questionnaire [SAQ]) HRQOL from 1 to 3 and 1 to 6 months after hospitalization, adjusting for potential sociodemographic and clinical confounders. Results: The patients (N = 1042) were, on average, 62 years old, 34% female, and 87% non-Hispanic white. A total of 10% were in the lowest level of activation. The patients with the lowest activation had 1.95 times (95% confidence interval, 1.05-3.62) and 2.18 times (95% confidence interval, 1.17-4.05) the odds of experiencing clinically significant declines in MCS and SAQ HRQOL, respectively, between 1 and 6 months than the most activated patients. The patient activation level was not associated with meaningful changes in PCS scores. Conclusions: Hospital survivors of an ACS with lower activation may be more likely to experience declines in mental and disease-specific HRQOL than more-activated patients, identifying a group at risk of poor outcomes.

Patients who develop herpes zoster or herpes zoster ophthalmicus may be at risk for cerebrovascular and cardiac complications. We systematically reviewed the published literature to determine the association between herpes zoster and its subtypes with the occurrence of cerebrovascular and cardiac events. Systematic searches of PubMed (MEDLINE), SCOPUS (Embase) and Google Scholar were performed in December 2016. Eligible studies were cohort, case-control, and self-controlled case-series examining the association between herpes zoster or subtypes of herpes zoster with the occurrence of cerebrovascular and cardiac events including stroke, transient ischemic attack, coronary heart disease, and myocardial infarction. Data on the occurrence of the examined events were abstracted. Odds ratios and their accompanying confidence intervals were estimated using random and fixed effects models with statistical heterogeneity estimated with the I² statistic. Twelve studies examining 7.9 million patients up to 28 years after the onset of herpes zoster met our pre-defined eligibility criteria. Random and fixed effects meta-analyses showed that herpes zoster, type unspecified, and herpes zoster ophthalmicus were associated with a significantly increased risk of cerebrovascular events, without any evidence of statistical heterogeneity. Our meta-analysis also found a significantly increased risk of cardiac events associated with herpes zoster, type unspecified. Our results are consistent with the accumulating body of evidence that herpes zoster and herpes zoster ophthalmicus are significantly associated with cerebrovascular and cardiovascular events.