Scalable and interpretable alternative to chart review for phenotype evaluation using standardized structured data from electronic health records.

OBJECTIVES: Chart review as the current gold standard for phenotype evaluation cannot support observational research on electronic health records and claims data sources at scale. We aimed to evaluate the ability of structured data to support efficient and interpretable phenotype evaluation as an alternative to chart review. MATERIALS AND METHODS: We developed Knowledge-Enhanced Electronic Profile Review (KEEPER) as a phenotype evaluation tool that extracts patient's structured data elements relevant to a phenotype and presents them in a standardized fashion following clinical reasoning principles. We evaluated its performance (interrater agreement, intermethod agreement, accuracy, and review time) compared to manual chart review for 4 conditions using randomized 2-period, 2-sequence crossover design. RESULTS: Case ascertainment with KEEPER was twice as fast compared to manual chart review. 88.1% of the patients were classified concordantly using charts and KEEPER, but agreement varied depending on the condition. Missing data and differences in interpretation accounted for most of the discrepancies. Pairs of clinicians agreed in case ascertainment in 91.2% of the cases when using KEEPER compared to 76.3% when using charts. Patient classification aligned with the gold standard in 88.1% and 86.9% of the cases respectively. CONCLUSION: Structured data can be used for efficient and interpretable phenotype evaluation if they are limited to relevant subset and organized according to the clinical reasoning principles. A system that implements these principles can achieve noninferior performance compared to chart review at a fraction of time.

A metadata framework for computational phenotypes.

With the burgeoning development of computational phenotypes, it is increasingly difficult to identify the right phenotype for the right tasks. This study uses a mixed-methods approach to develop and evaluate a novel metadata framework for retrieval of and reusing computational phenotypes. Twenty active phenotyping researchers from 2 large research networks, Electronic Medical Records and Genomics and Observational Health Data Sciences and Informatics, were recruited to suggest metadata elements. Once consensus was reached on 39 metadata elements, 47 new researchers were surveyed to evaluate the utility of the metadata framework. The survey consisted of 5-Likert multiple-choice questions and open-ended questions. Two more researchers were asked to use the metadata framework to annotate 8 type-2 diabetes mellitus phenotypes. More than 90% of the survey respondents rated metadata elements regarding phenotype definition and validation methods and metrics positively with a score of 4 or 5. Both researchers completed annotation of each phenotype within 60 min. Our thematic analysis of the narrative feedback indicates that the metadata framework was effective in capturing rich and explicit descriptions and enabling the search for phenotypes, compliance with data standards, and comprehensive validation metrics. Current limitations were its complexity for data collection and the entailed human costs.

A generalizable physiological model for detection of Delayed Cerebral Ischemia using Federated Learning.

Delayed cerebral ischemia (DCI) is a complication seen in patients with subarachnoid hemorrhage stroke. It is a major predictor of poor outcomes and is detected late. Machine learning models are shown to be useful for early detection, however training such models suffers from small sample sizes due to rarity of the condition. Here we propose a Federated Learning approach to train a DCI classifier across three institutions to overcome challenges of sharing data across hospitals. We developed a framework for federated feature selection and built a federated ensemble classifier. We compared the performance of FL model to that obtained by training separate models at each site. FL significantly improved performance at only two sites. We found that this was due to feature distribution differences across sites. FL improves performance in sites with similar feature distributions, however, FL can worsen performance in sites with heterogeneous distributions. The results highlight both the benefit of FL and the need to assess dataset distribution similarity before conducting FL.

Racial/ethnic and socioeconomic disparities in the use of newer diabetes medications in the Look AHEAD study.

Background: Among patients with type 2 diabetes, minority racial/ethnic groups have a higher burden of cardiovascular disease, chronic kidney disease, and hypoglycaemia. These groups may especially benefit from newer diabetes medication classes, but high cost may limit access. We examined the association of race/ethnicity with the initiation of newer diabetes medications (GLP-1 receptor agonists, DPP-4 inhibitors, SGLT-2 inhibitors). Methods: We conducted a secondary analysis of the Look AHEAD (Action for Health in Diabetes) trial including participants with at least one study visit after April 28, 2005. Cox proportional hazards models were used to estimate the association between race/ethnicity and socioeconomic factors with time to initiation of any newer diabetes medication from April 2005 to February 2020. Models were adjusted for demographic and clinical characteristics. Findings: Among 4,892 participants, 63.6%, 15.7%, 12.6%, 5.2%, and 2.9% were White, Black, Hispanic, American Indian or Alaskan Native (AI/AN), or other race/ethnicity, respectively. During a median follow-up of 8.3 years, 2,180 (45.2%) participants were initiated on newer diabetes medications. Race/ethnicity was associated with newer diabetes medication initiation (p=.019). Specifically, initiation was lower among Black (HR 0.81, 95% CI 0.70 -0.94) and AI/AN participants (HR 0.51, 95% CI 0.26-0.99). Yearly family income was inversely associated with initiation of newer diabetes medications (HR 0.78, 95% CI 0.62-0.98) comparing the lowest and highest income groups. Findings were mostly driven by GLP-1 receptor agonists. Interpretation: These findings provide evidence of racial/ethnic disparities in the initiation of newer diabetes medications, independent of socioeconomic factors, which may contribute to worse health outcomes.

Impact of the announcement and implementation of the UK Soft Drinks Industry Levy on sugar content, price, product size and number of available soft drinks in the UK, 2015-19: A controlled interrupted time series analysis.

BACKGROUND: Dietary sugar, especially in liquid form, increases risk of dental caries, adiposity, and type 2 diabetes. The United Kingdom Soft Drinks Industry Levy (SDIL) was announced in March 2016 and implemented in April 2018 and charges manufacturers and importers at £0.24 per litre for drinks with over 8 g sugar per 100 mL (high levy category), £0.18 per litre for drinks with 5 to 8 g sugar per 100 mL (low levy category), and no charge for drinks with less than 5 g sugar per 100 mL (no levy category). Fruit juices and milk-based drinks are exempt. We measured the impact of the SDIL on price, product size, number of soft drinks on the marketplace, and the proportion of drinks over the lower levy threshold of 5 g sugar per 100 mL. METHODS AND FINDINGS: We analysed data on a total of 209,637 observations of soft drinks over 85 time points between September 2015 and February 2019, collected from the websites of the leading supermarkets in the UK. The data set was structured as a repeat cross-sectional study. We used controlled interrupted time series to assess the impact of the SDIL on changes in level and slope for the 4 outcome variables. Equivalent models were run for potentially levy-eligible drink categories ('intervention' drinks) and levy-exempt fruit juices and milk-based drinks ('control' drinks). Observed results were compared with counterfactual scenarios based on extrapolation of pre-SDIL trends. We found that in February 2019, the proportion of intervention drinks over the lower levy sugar threshold had fallen by 33.8 percentage points (95% CI: 33.3-34.4, p < 0.001). The price of intervention drinks in the high levy category had risen by £0.075 (£0.037-0.115, p < 0.001) per litre-a 31% pass through rate-whilst prices of intervention drinks in the low levy category and no levy category had fallen and risen by smaller amounts, respectively. Whilst the product size of branded high levy and low levy drinks barely changed after implementation of the SDIL (-7 mL [-23 to 11 mL] and 16 mL [6-27ml], respectively), there were large changes to product size of own-brand drinks with an increase of 172 mL (133-214 mL) for high levy drinks and a decrease of 141 mL (111-170 mL) for low levy drinks. The number of available drinks that were in the high levy category when the SDIL was announced was reduced by 3 (-6 to 12) by the implementation of the SDIL. Equivalent models for control drinks provided little evidence of impact of the SDIL. These results are not sales weighted, so do not give an account of how sugar consumption from drinks may have changed over the time period. CONCLUSIONS: The results suggest that the SDIL incentivised many manufacturers to reduce sugar in soft drinks. Some of the cost of the levy to manufacturers and importers was passed on to consumers as higher prices but not always on targeted drinks. These changes could reduce population exposure to liquid sugars and associated health risks.

Productivity of 18F-FDG-PET/CT Diagnostic Tool in the Management of Pediatric Lymphoblastic Lymphoma.

BACKGROUND: Lymphoblastic lymphoma (LL) comprises approximately 20% of childhood non-Hodgkin lymphoma (NHL); however, few studies had investigated the role of 18F-FDG-PET/CT in pediatric LL patients. We aim in this study to assess the role of 18F-FDG-PET/CT in the initial staging of newly diagnosed pediatric patients with LL as well as in the assessment of response after induction chemotherapy. PATIENTS AND METHODS: A prospective study enrolled biopsy proven newly diagnosed pediatric LL patients presenting in the Children Cancer Hospital Egypt (CCHE) during the period from October 2014 to October 2016. 18F-FDG-PET/CT was done initially before therapy and after induction chemotherapy in all patients. The patients were followed until the end of April 2018 (mean 23.5 months). RESULTS: All lymphoma involvement lesions (n = 43) were FDG avid and the intensity of nodal FDG uptake was variable. Two patients (11%) had bone marrow (BM) involvement by < 25% blast cells with corresponding positive BM focal uptake in 18F-FDG-PET/CT (SUVmax = 4 and 4.5). Evaluation post induction phase; CT detected 8 residual lesions in 8 patients (44.4%), while 18F-FDG-PET/CT detected only 3 Deauville-positive residual lesions in 3 patients (16.6%). No intensification of therapy was done in all post-induction positive patients. Repeated 18F-FDG-PET/CT at week 18 for post-induction patients revealed cleared all Deauville-positive residual lesions. On the other hand, repeated CT at week 18 detected regression but still residual in 4/8 (50%) post-induction CT lesions with clearance of the rest (50%). CONCLUSION: In initial staging, 18F-FDG-PET/CT is a useful tool for disease extent evaluation of pediatric LL. Moreover, it could provide a diagnostic hint for BM involvement. 18F-FDG-PET/CT done after induction therapy has a good negative predictive value with higher specificity than CT alone, but is not an indication for treatment intensification due to false positive results. However, larger sample size is required for better conclusion.

Systematic review and meta-analysis of remotely delivered interventions using self-monitoring or tailored feedback to change dietary behavior.

Background: Self-monitoring (SM) of diet and tailored feedback (TF) have been suggested as tools for changing dietary behavior. New technologies allow users to monitor behavior remotely, potentially improving reach, adherence, and outcomes. Objective: We conducted a systematic literature review and meta-analysis to address the following question: are remotely delivered standalone (i.e., no human contact) interventions that use SM or TF effective in changing eating behaviors? Design: Five databases were searched in October 2016 (updated in September 2017). Only randomized controlled trials published after 1990 were included. Trials could include any adult population with no history of disordered eating which delivered an SM or TF intervention without direct contact and recorded actual dietary consumption as an outcome. Three assessors independently screened the search results. Two reviewers extracted the study characteristics, intervention details, and outcomes, and assessed risk of bias using the Cochrane tool. Results were converted to standardized mean differences and incorporated into a 3-level (individuals and outcomes nested in studies) random effects meta-analysis. Results: Twenty-six studies containing 21,262 participants were identified. The majority of the studies were judged to be unclear or at high risk of bias. The meta-analysis showed dietary improvement in the intervention group compared to the control group with a standardized mean difference of 0.17 (95% CI: 0.10, 0.24; P < 0.0001). The I2 statistic for the meta-analysis was 0.77, indicating substantial heterogeneity in results. A "one study removed" sensitivity analysis showed that no single study excessively influenced the results. Conclusions: Standalone interventions containing self-regulatory methods have a small but significant effect on dietary behavior, and integrating these elements could be important in future interventions. However, there was substantial variation in study results that could not be explained by the characteristics we explored, and there were risk-of-bias concerns with the majority of studies.

Health impact assessment of the UK soft drinks industry levy: a comparative risk assessment modelling study.

BACKGROUND: In March, 2016, the UK Government proposed a tiered levy on sugar-sweetened beverages (SSBs; high tax for drinks with >8 g of sugar per 100 mL, moderate tax for 5-8 g, and no tax for <5 g). We estimate the effect of possible industry responses to the levy on obesity, diabetes, and dental caries. METHODS: We modelled three possible industry responses: reformulation to reduce sugar concentration, an increase of product price, and a change of the market share of high-sugar, mid-sugar, and low-sugar drinks. For each response, we defined a better-case and worse-case health scenario. We developed a comparative risk assessment model to estimate the UK health impact of each scenario on prevalence of obesity and incidence of dental caries and type 2 diabetes. The model combined data for sales and consumption of SSBs, disease incidence and prevalence, price elasticity estimates, and estimates of the association between SSB consumption and disease outcomes. We drew the disease association parameters from a meta-analysis of experimental studies (SSBs and weight change), a meta-analysis of prospective cohort studies (type 2 diabetes), and a prospective cohort study (dental caries). FINDINGS: The best modelled scenario for health is SSB reformulation, resulting in a reduction of 144 383 (95% uncertainty interval 5102-306 743; 0·9%) of 15 470 813 adults and children with obesity in the UK, 19 094 (6920-32 678; incidence reduction of 31·1 per 100 000 person-years) fewer incident cases of type 2 diabetes per year, and 269 375 (82 211-470 928; incidence reduction of 4·4 per 1000 person-years) fewer decayed, missing, or filled teeth annually. An increase in the price of SSBs in the better-case scenario would result in 81 594 (3588-182 669; 0·5%) fewer adults and children with obesity, 10 861 (3899-18 964; 17·7) fewer incident cases of diabetes per year, and 149 378 (45 231-262 013; 2·4) fewer decayed, missing, or filled teeth annually. Changes to market share to increase the proportion of low-sugar drinks sold in the better-case scenario would result in 91 042 (4289-204 903; 0·6%) fewer adults and children with diabetes, 1528 (4414-21 785; 19·7) fewer incident cases of diabetes per year, and 172 718 (47 919-294 499; 2·8) fewer decayed, missing, or filled teeth annually. The greatest benefit for obesity and oral health would be among individuals aged younger than 18 years, with people aged older than 65 years having the largest absolute decreases in diabetes incidence. INTERPRETATION: The health impact of the soft drinks levy is dependent on its implementation by industry. Uncertainty exists as to how industry will react and about estimation of health outcomes. Health gains could be maximised by substantial product reformulation, with additional benefits possible if the levy is passed on to purchasers through raising of the price of high-sugar and mid-sugar drinks and activities to increase the market share of low-sugar products. FUNDING: None.