Annals of Family Medicine最新文献

Purpose: Factors influencing missed appointments are complex and difficult to anticipate and intervene against. To optimize appointment adherence, we aimed to use personalized machine learning and big data analytics to predict the risk of and contributing factors for no-shows and late cancellations in primary care practices.

Methods: We conducted a retrospective longitudinal study leveraging geolinked clinical, care utilization, socioeconomic, and climate data from 15 family medicine clinics at a regional academic health center in Pennsylvania from January 2019 to June 2023. We developed multiclass machine learning models using gradient boost, random forest, neural network, and logistic regression to predict appointment outcomes, followed by feature importance analysis to identify contributing factors for no-shows or late cancellations at the population and patient levels. We performed stratified analysis to evaluate the prediction performance by sex and race/ethnicity to ensure the fairness of the final model among sensitive features.

Results: The analysis consisted of 109,328 patients and 1,118,236 appointments, including 77,322 (6.9%) no-shows and 75,545 (6.8%) late cancellations. The gradient boost model achieved the best performance with an area under the receiver operating characteristic curve of 0.852 for predicting no-shows and 0.921 for late cancellations. No bias against patient characteristics was detected. Schedule lead time was identified as the most important predictor of missed appointments.

Conclusions: Missed appointments remain a challenge for primary care. This study provided a practical and robust framework to predict missed appointments, laying the foundation for developing personalized strategies to improve patients' adherence to primary care appointments.

Purpose: Hearing loss is a common deficit that remains underdiagnosed. To address this issue, automatic self-hearing tests have been developed. These tools are based on pure-tone detection and speech-in-noise evaluation. The present study evaluated the acceptability and the feasibility of hearing screening for patients consulting in private family practice medicine.

Methods: Data were collected in 3 French medical care centers from May through November 2022. Fast pure-tone (SoTone) and speech-in-noise (SoNoise) tests were available on the SONUP application. Three parameters were measured: (1) duration of the protocol; (2) pertinence of performing both pure-tone and speech-in-noise tests; and (3) number of hearing-impaired patients detected and their follow-up (ie, consultation with an ear, nose, and throat [ENT] specialist, and hearing aid fitting).

Results: Of the 516 eligible patients, 219 (42%) were able to perform both tests. Among the screened patients, 161 (74%) had negative test results, while 59 (27%) had positive results indicating hearing loss. Although patients were encouraged to consult an ENT specialist, only 14 did so, and 8 agreed to be fitted with hearing aids. The average duration of the tests, including the explanation (1 minute 43 seconds), was 6 minutes 8 seconds. Interestingly, the SoTone (1 minute 10 seconds), appears to be sufficient for detecting hearing loss.

Conclusions: This study supports integration of app-based hearing screenings into family medical care, as it is compatible with routine consultations. The use of tablet-based applications may assist general practitioners by enhancing the diagnosis of hearing disorders.

Purpose: Identifying cardiovascular disease before conception and in early pregnancy can better inform obstetric cardiovascular care. Our main objective was to evaluate the diagnostic performance of artificial intelligence (AI)-enabled digital tools for detecting left ventricular systolic dysfunction (LVSD) among women of reproductive age.

Methods: In a pilot cross-sectional study, we enrolled an initial cohort of 100 consecutive women aged 18-49 years who had a primary care physician and a scheduled echocardiography at Mayo Clinic Florida (Jacksonville) (cohort 1). Twelve-lead electrocardiography (ECG) and digital stethoscope recordings (single-lead ECG + phonocardiography) were performed on the date of echocardiography. We used deep learning to generate prediction probabilities for LVSD (defined as left ventricular ejection fraction <50%) for the 12-lead ECG (AI-ECG) and stethoscope (AI-stethoscope) recordings. In a second cohort of 100 participants, we enrolled consecutive women seen in primary care to estimate the prevalence of positive AI screening results when deployed for routine use (cohort 2).

Results: The median age of participants was 38.6 years (quartile 1: 30.3 years, quartile 3: 45.5 years), and 71.9% identified as part of the non-Hispanic White population. Among cohort 1, 5% had LVSD. The AI-ECG had an area under the curve of 0.94, and the AI-stethoscope (maximum prediction across all chest locations) had an area under the curve of 0.98. Among cohort 2, the prevalence of a positive AI screen was 1% and 3.2% for AI-ECG and the AI-stethoscope, respectively.

Conclusion: We found these AI tools to be effective for the detection of cardiomyopathy associated with LVSD among women of reproductive age. These tools could potentially be useful for preconception cardiovascular evaluations.

We, as the current and immediate-past NAPCRG Trainee Committee, share our perspectives as an international and diverse group of primary care research trainees. In this essay, we discuss the challenges and opportunities for achieving a more diverse, equitable, and inclusive primary care workforce by reflecting on 2 main challenges: (1) insufficient support for underrepresented identities in medicine, and (2) inadequate integration within existing primary care teams. Within each of these challenges, we pose potential opportunities for improvement using a trainee lens.

Purpose: This study aimed to assess how agile implementation-driven iterative processes and tailored workflows can facilitate the implementation of a digital cognitive assessment (DCA) tool for patients aged 65 years or older into primary care practices.

Methods: We used agile implementation principles to integrate a DCA tool into routine workflows across 7 primary care clinics. The intervention involved a structured selection process for identifying an appropriate DCA tool, stakeholder engagement through iterative sprints (structured, time-bound cycles), and development of tailored workflows to meet clinic-specific needs. A brain health navigator role was established to support patients with positive or borderline screenings, and assist primary care clinicians with follow-up assessment. We used the reach, effectiveness, adoption, implementation, and maintenance (RE-AIM) framework to evaluate the intervention's performance over a 12-month period.

Results: The intervention engaged 69 (63.8%) of 108 clinicians across the 7 clinics. DCA screening was completed in 1,808 (10.8%) of 16,708 eligible visits. We selected the Linus Health Core Cognitive Evaluation tool as our DCA tool based on stakeholder evaluations. Screening workflows were tailored to each clinic. The brain health navigator received 447 referrals for further assessment of a positive or borderline screening result. Four clinics fully adopted the intervention, achieving a DCA completion rate of at least 20%, and 5 clinics were still routinely using the DCA tool at 12 months.

Conclusions: Agile implementation effectively helped integrate the DCA tool into primary care workflows. Customized workflows, stakeholder engagement, and iterative improvements were crucial for adoption and sustainability. These insights can guide future efforts for early detection and management of cognitive impairment in primary care, ultimately improving patient outcomes and easing the burden on health care professionals.

Promoting adherence to medical recommendations remains one of the oldest yet most persistent challenges of modern clinical practice. Although increasingly sympathetic to structural forces that affect health behavior, standard models frequently conceptualize nonadherence as a phenomenon of patient behavior, a self-evident quality belonging to patients that is responsible for a myriad of undesired outcomes. We contend, however, that this approach not only fails to consider the role of the clinician in the concept's origins in clinical encounters, but also has facilitated the use of adherence terms (eg, nonadherent, noncompliant, treatment resistant) as pejorative social labels to the detriment of the physician-patient relationship. Used without care, such terminology can alter the meaning assigned to patients' behaviors so that structural barriers to care such as poverty and systemic racism are reframed as problems of poor attitude or effort. This article explores the functions of adherence terms as social labels by reviewing their underlying logic in clinical settings and outlining pitfalls in the pathologization of nonadherence in research and practice. We propose the concept of adherence labeling-the assessment, classification, and dissemination of clinicians' perceptions of patients' adherence through social labels-as an alternative model to understand how adherence terms may inadvertently obstruct the care of marginalized patients.