Overcoming barriers to the adoption and implementation of predictive modeling and machine learning in clinical care: what can we learn from US academic medical centers?

There is little known about how academic medical centers (AMCs) in the US develop, implement, and maintain predictive modeling and machine learning (PM and ML) models. We conducted semi-structured interviews with leaders from AMCs to assess their use of PM and ML in clinical care, understand associated challenges, and determine recommended best practices. Each transcribed interview was iteratively coded and reconciled by a minimum of 2 investigators to identify key barriers to and facilitators of PM and ML adoption and implementation in clinical care. Interviews were conducted with 33 individuals from 19 AMCs nationally. AMCs varied greatly in the use of PM and ML within clinical care, from some just beginning to explore their utility to others with multiple models integrated into clinical care. Informants identified 5 key barriers to the adoption and implementation of PM and ML in clinical care: (1) culture and personnel, (2) clinical utility of the PM and ML tool, (3) financing, (4) technology, and (5) data. Recommendation to the informatics community to overcome these barriers included: (1) development of robust evaluation methodologies, (2) partnership with vendors, and (3) development and dissemination of best practices. For institutions developing clinical PM and ML applications, they are advised to: (1) develop appropriate governance, (2) strengthen data access, integrity, and provenance, and (3) adhere to the 5 rights of clinical decision support. This article highlights key challenges of implementing PM and ML in clinical care at AMCs and suggests best practices for development, implementation, and maintenance at these institutions.

Early Child Care and Weight Status in a Cohort of Predominantly Black Infants in the Southeastern United States.

Background: Previous studies show inconsistent relations between child care and obesity, but few assessed longitudinal associations during infancy and even fewer included racially diverse children. We examined associations of time infants spent in child care, both overall and in different types of care, with weight status at 6 and 12 months. Methods: We examined 664 infants living in central North Carolina. We conducted adjusted multivariable linear regressions examining (1) child care from birth to 6 months and 6-month weight-for-length (WFL) z-score, and (2) child care from birth to 12 months and 12-month WFL z-score. We assessed any child care and child care by type, including relative care, informal care by a nonrelative, formal child care, and a combination of care (e.g., relative and informal care). Results: Nearly 70% of infants were black and 49% were female. After adjustment for potential confounders, any child care was not associated with WFL z-score at 6 months (0.07; 95% confidence intervals [CI] -0.02 to 0.16; p = 0.13) or 12 months (0.05; 95% CI -0.02 to 0.12; p = 0.19). However, greater combination care was associated with higher WFL z-score at 6 months (0.68; 95% CI 0.23-1.13; p = 0.003) and greater care by a relative was associated with higher WFL z-score at 12 months (0.16; 95% CI 0.05-0.26; p = 0.005). Conclusions: Although we did not observe associations with any child care, combination care and relative care during infancy were associated with higher weight. Interventions aimed at preventing excessive weight gain in early life may target relatives who provide regular care for infants.

Barriers to Engagement in a Workplace Weight Management Program: A Qualitative Study.

PURPOSE: To investigate (1) why some participants in a workplace weight management program were more engaged in the program, (2) specific barriers and facilitators for engagement and weight loss, and (3) suggest how workplaces may better engage employees in these programs to improve their effectiveness. DESIGN: Qualitative study (8 focus groups). SETTING: A large academic university and medical system. PARTICIPANTS: Twenty-six (5%) of the 550 employees who participated in a weight management program as part of the Steps to Health study. MEASURES: A trained moderator guided the audio-recorded focus groups. ANALYSIS: Transcripts were analyzed using the directed content analysis approach. RESULTS: Participants faced numerous barriers to engagement in workplace weight management programs, both within and outside the workplace. Participants viewed the coaches positively and reported that the coaches had a strong influence on their engagement in the program. Participants suggested increased frequency and variety of contact by coaches, on-site group exercise classes, and tailored educational materials. CONCLUSION: Workplace weight management programs may be improved by being more flexible around participants' schedules and changing needs, by increasing access to affordable, convenient exercise facilities, and by implementing institutional changes that encourage healthy eating and physical activity during the workday. Employers should measure program engagement and solicit participant feedback to ensure that the programs are appropriate and delivered in an optimal manner.

Physical activity and sedentary behavior during pregnancy and postpartum, measured using hip and wrist-worn accelerometers.

BACKGROUND: Physical activity in pregnancy and postpartum is beneficial to mothers and infants. To advance knowledge of objective physical activity measurement during these periods, this study compares hip to wrist accelerometer compliance; assesses convergent validity (correlation) between hip- and wrist-worn accelerometry; and assesses change in physical activity from pregnancy to postpartum. METHODS: We recruited women during pregnancy (n = 100; 2014-2015), asking them to wear hip and wrist accelerometers for 7 days during Trimester 2 (T2), Trimester 3 (T3), and 3-, 6-, 9- and 12-months postpartum. We assessed average wear-time and correlations (axis-specific counts/minute, vector magnitude counts/day and step counts/day) at T2, T3, and postpartum. RESULTS: Compliance was higher for wrist-worn accelerometers. Hip and wrist accelerometers showed moderate to high correlations (Pearson's r 0.59 to 0.84). Hip-measured sedentary and active time differed little between T2 and T3. Moderate-to-vigorous physical activity decreased at T3 and remained low postpartum. Light physical activity increased and sedentary time decreased throughout the postpartum period. CONCLUSIONS: Wrist accelerometers may be preferable during pregnancy and appear comparable to hip accelerometers. As physical activity declines during later pregnancy and may not rebound post birth, support for re-engaging in physical activity earlier in the postpartum period may benefit women.

Perceived Barriers to Healthy Eating and Physical Activity Among Participants in a Workplace Obesity Intervention.

OBJECTIVE: To characterize barriers to healthy eating (BHE) and physical activity (BPA) among participants in a workplace weight management intervention. METHODS: Steps to health participants completed a questionnaire to ascertain barriers to physical activity and healthy eating faced. Exploratory factor analysis was used to determine the factor structure for BPA and BHE. The relationships of these factors with accelerometer data and dietary behaviors were assessed using linear regression. RESULTS: Barriers to physical activity included time constraints and lack of interest and motivation, and to healthy eating, lack of self-control and convenience, and lack of access to healthy foods. Higher BHE correlated with higher sugary beverage intake but not fruit and vegetable and fat intake. CONCLUSIONS: To improve their effectiveness, workplace weight management programs should consider addressing and reducing barriers to healthy eating and physical activity.

Cohort profile for the Nurture Observational Study examining associations of multiple caregivers on infant growth in the Southeastern USA.

PURPOSE: Childcare has been associated with obesity in children in cross-sectional and longitudinal studies, although some observed no association. Few studies have focused on care during infancy, a period when children may be especially vulnerable. PARTICIPANTS: The Nurture Study is an observational birth cohort designed to assess longitudinal associations of childcare and the presence of multiple caregivers on infant adiposity and weight trajectories throughout the first year of life. We examine as potential mediators feeding, physical activity, sleep and stress. We completed recruitment in 2015. Of the 860 women who enrolled during pregnancy, 799 delivered a single live infant who met our inclusion criteria. Of those, 666 mothers (77.4%) agreed to participate in the study for themselves and their infants. FINDINGS TO DATE: Among the 666 women in the study, 472 (71%) identified as black, 127 (19%) as white, 7 (1%) as Asian or Asian American, 6 (1%) as Native American and 49 (7%) as other race or more than one race; 43 (7%) identified as Hispanic/Latina. Just under half (48%) had a high school diploma or less, 61% had household incomes <$20 000/year and 59% were married or living with a partner. The mean (SD) infant gestational age was 41.28 weeks (2.29) and birth weight for gestational age z-score was -0.31 (0.93). Just under half (49%) of infants were females, 69% received some human milk and 40% were exclusively breast fed at hospital discharge. Data collection began in 2013, is currently underway, and is scheduled to conclude in late 2016. FUTURE PLANS: Results will help assess the magnitude of associations between childcare in infancy and subsequent obesity. Findings will also inform intervention and policy efforts to improve childcare environments and help prevent obesity in settings where many infants spend time. TRIAL REGISTRATION NUMBER: Clinicaltrials.gov, NCT01788644.