Individualized Prospective Prediction of Opioid Use Disorder.

OBJECTIVE: Opioid use disorder (OUD) is a chronic relapsing disorder with a problematic pattern of opioid use, affecting nearly 27 million people worldwide. Machine learning (ML)-based prediction of OUD may lead to early detection and intervention. However, most ML prediction studies were not based on representative data sources and prospective validations, limiting their potential to predict future new cases. In the current study, we aimed to develop and prospectively validate an ML model that could predict individual OUD cases based on representative large-scale health data. METHOD: We present an ensemble machine-learning model trained on a cross-linked Canadian administrative health data set from 2014 to 2018 (n = 699,164), with validation of model-predicted OUD cases on a hold-out sample from 2014 to 2018 (n = 174,791) and prospective prediction of OUD cases on a non-overlapping sample from 2019 (n = 316,039). We used administrative records of OUD diagnosis for each subject based on International Classification of Diseases (ICD) codes. RESULTS: With 6409 OUD cases in 2019 (mean [SD], 45.34 [14.28], 3400 males), our model prospectively predicted OUD cases at a high accuracy (balanced accuracy, 86%, sensitivity, 93%; specificity 79%). In accord with prior findings, the top risk factors for OUD in this model were opioid use indicators and a history of other substance use disorders. CONCLUSION: Our study presents an individualized prospective prediction of OUD cases by applying ML to large administrative health datasets. Such prospective predictions based on ML would be essential for potential future clinical applications in the early detection of OUD.

Implementation of Enhanced Recovery After Surgery (ERAS) Across a Provincial Healthcare System: The ERAS Alberta Colorectal Surgery Experience.

BACKGROUND: Enhanced recovery after surgery (ERAS) colorectal guideline implementation has occurred primarily in standalone institutions worldwide. We implemented the guideline in a single provincial healthcare system, and our study examined the effect of the guideline on patient outcomes [length of stay (LOS), complications, and 30-day post-discharge readmissions] across a healthcare system. METHODS: We compared pre- and post-guideline implementation in consecutive elective colorectal patients, ≥ 18 years, from six Alberta hospitals between February 2013 and December 2014. Participants were followed up to 30 days post discharge. We used summary statistics, to assess the LOS and complications, and multivariate regression methods to assess readmissions and to estimate cost impacts. RESULTS: A total of 1333 patients (350 pre- and 983 post-ERAS) were analysed. Of this number, 55 % were males. Median overall guideline compliance was 39 % in pre- and 60 % in post-ERAS patients. Median LOS was 6 days for pre-ERAS compared to 4.5 days in post-ERAS patients with the longest implementation (p value <0.0001). Adjusted risk ratio (RR) was 1.71, 95 % CI 1.09-2.68 for 30-day readmission, comparing pre- to post-ERAS patients. The proportion of patients who developed at least one complication was significantly reduced, from pre- to post-ERAS, difference in proportions = 11.7 %, 95 % CI 2.5-21.0, p value: 0.0139. The net cost savings attributable to guideline implementation ranged between $2806 and $5898 USD per patient. CONCLUSION: The findings in our study have shown that ERAS colorectal guideline implementation within a healthcare system resulted in patient outcome improvements, similar to those obtained in smaller standalone implementations. There was a significant beneficial impact of ERAS on scarce health system resources.

Nurses' intention to support informed decision-making about breast cancer screening with mammography: a survey.

There is growing interest in informed decision-making about breast cancer screening with mammography and growing advocacy for the provision of balanced information about potential benefits and harms. The authors report on a survey evaluating nurses' intention to support women targeted by the Quebec Breast Cancer Screening Program in making informed decisions about breast cancer screening with mammography. Of the 840 questionnaires completed, 618 were included in the data analysis. The mean +/- standard deviation score for intention was 1.7 +/- 1.2 on a 6-point Likert scale ranging from -3 to +3, indicating strong intention to support the targeted women. Perceived behavioural control was the theory-based variable most strongly associated with intention, followed by attitude and social norm. These results can be used to develop interventions to train nurses in integrating informed decision-making about breast cancer screening with mammography into their practice and to design relevant decision support tools.

Risk factors for developmental vulnerability: Insight from population-level surveillance using the Early Development Instrument.

Objectives: Population-level studies may elucidate the most promising intervention targets to prevent negative outcomes of developmental vulnerability in children. This study aims to bridge the current literature gap on identifying population-level developmental vulnerability risk factors using combined social and biological/health information. Methods: This study assessed developmental vulnerability among kindergarten children using the 2016 Early Development Instrument (EDI) and identified risk factors of developmental vulnerability using EDI data cross-linked to a population-wide administrative health dataset. A total number of 23,494 children aged 5-6 were included (48% female). Prenatal, neonatal, and early childhood risk factors for developmental vulnerability were investigated, highlighting the most important ones contributing to early development. Results: The main risk factors for developmental vulnerability were children with a history of mental health diagnosis (risk ratio = 1.46), biological sex-male (risk ratio = 1.51), and poor socioeconomic status (risk ratio = 1.58). Conclusion: Our study encompasses both social and health information in a populational-level representative sample of Alberta, Canada. The results confirm evidence established in other geographic regions and jurisdictions and demonstrate the association between perinatal risk factors and developmental vulnerability. Based on these results, we argue that the health system should adopt a multilevel prevention and intervention strategy, targeting individual, family, and community together.

Family physicians' intention to support women in making informed decisions about breast cancer screening with mammography: a cross-sectional survey.

BACKGROUND: The net benefits of routine breast cancer screening with mammography have been questioned, and there is evidence to indicate that supporting women to make an informed decision about breast cancer screening with mammography is preferable. The aims of this study were to assess the intention of family physicians to provide women with this support and the determinants of this intention, and to identify factors that might influence family physicians adopting this behavior. METHODS: Family physicians from the province of Quebec, Canada, attending a 45-min lecture on informed decision making and cancer screening were asked to complete a questionnaire after the lecture regarding their intention to adopt the behavior. The questions, based on the Theory of Planned Behavior, measured physicians' intention and its determinants (attitude, perceived behavioral control, and socio-professional norm) regarding supporting women to make informed decisions about breast cancer screening with mammography. Open-ended questions were also used to explore complementary factors influencing their intention. RESULTS: Out of 800 questionnaires distributed, 301 (38 %) were returned and 288 were included in data analysis. The mean ± standard deviation and median score for intention were respectively 1.9 ± 1.2 and 2.0 on a 6-point Likert scale (-3 to +3). Perceived behavioral control was the variable most strongly associated with intention (high versus low score, odds ratio = 15.7, 95 % CI 6.7-36.6), followed by attitude (high versus low score, odds ratio = 7.5, 95 % CI 3.3-16.8), then social norm (high versus low score, odds ratio = 5.8, 95 % CI 2.6-12.9). The most-reported barrier to adopting the behavior was time constraints (41 %) while the most-reported facilitator was availability of relevant decision support tools (29 %). CONCLUSIONS: Respondents showed strong intention to support women in informed decision-making about breast cancer screening, the strongest predictor being perceived behavioral control. These results could contribute to training physicians to integrate this behavior into their practices and to designing relevant decision support tools.