Physical activity (PA) is associated with improvement in breast cancer treatment-related symptoms and survival, yet most breast cancer survivors do not meet national PA guidelines. This study aimed ...to identify characteristics of participants that were associated with an increased likelihood of meeting PA guidelines. Adults with breast cancer seen at Mayo Clinic (Rochester, MN) were surveyed regarding their PA participation, and those who self-reported at least 150 minutes of moderate and/or strenuous aerobic PA weekly on average were considered to be "meeting guidelines". Three thousand participants returned PA data. Younger age, completion of the survey 7-12 years after diagnosis, absence of recurrence, no bilateral mastectomy, absence of metastatic disease, and lower BMI at the time of survey completion were associated with PA participation (P < .05 in univariate and multivariate analyses). Findings were similar when a threshold of 90 minutes was applied. These results may inform the development of targeted PA-facilitating interventions.
To examine predictors of understanding preemptive CYP2D6 pharmacogenomics test results and to identify key features required to improve future educational efforts of preemptive pharmacogenomics ...testing.
One thousand ten participants were surveyed after receiving preemptive CYP2D6 pharmacogenomics test results.
Eighty-six percent (n = 869) of patients responded. Of the responders, 98% were white and 55% were female; 57% had 4 years or more of post-secondary education and an average age of 58.9 ± 5.5 years. Twenty-six percent said that they only somewhat understood their results and 7% reported they did not understand them at all. Only education predicted understanding. The most common suggestion for improvement was the use of layperson's terms when reporting results. In addition, responders suggested that results should be personalized by referring to medications that they were currently using. Of those reporting imperfect drug adherence, most (91%) reported they would be more likely to use medication as prescribed if pharmacogenomic information was used to help select the drug or dose.
Despite great efforts to simplify pharmacogenomic results (or because of them), approximately one-third of responders did not understand their results. Future efforts need to provide more examples and tailor results to the individual. Incorporation of pharmacogenomics is likely to improve medication adherence.Genet Med advance online publication 05 January 2017.
•Falls are a leading cause of unintentional injury among older adults.•EHR documents contain many fall events, not captured by ICD-9/10 codes.•A BERT model can capture fall events that require ...context understanding.•A hybrid model further improves the performance of BERT through post-hoc rules.
Falls are a leading cause of unintentional injury in the elderly. Electronic health records (EHRs) offer the unique opportunity to develop models that can identify fall events. However, identifying fall events in clinical notes requires advanced natural language processing (NLP) to simultaneously address multiple issues because the word “fall” is a typical homonym.
We implemented a context-aware language model, Bidirectional Encoder Representations from Transformers (BERT) to identify falls from the EHR text and further fused the BERT model into a hybrid architecture coupled with post-hoc heuristic rules to enhance the performance. The models were evaluated on real world EHR data and were compared to conventional rule-based and deep learning models (CNN and Bi-LSTM). To better understand the ability of each approach to identify falls, we further categorize fall-related concepts (i.e., risk of fall, prevention of fall, homonym) and performed a detailed error analysis.
The hybrid model achieved the highest f1-score on sentence (0.971), document (0.985), and patient (0.954) level. At the sentence level (basic data unit in the model), the hybrid model had 0.954, 1.000, 0.988, and 0.999 in sensitivity, specificity, positive predictive value, and negative predictive value, respectively. The error analysis showed that that machine learning-based approaches demonstrated higher performance than a rule-based approach in challenging cases that required contextual understanding. The context-aware language model (BERT) slightly outperformed the word embedding approach trained on Bi-LSTM. No single model yielded the best performance for all fall-related semantic categories.
A context-aware language model (BERT) was able to identify challenging fall events that requires context understanding in EHR free text. The hybrid model combined with post-hoc rules allowed a custom fix on the BERT outcomes and further improved the performance of fall detection.
Monoclonal B-cell lymphocytosis (MBL) is a precursor to CLL. Other than age, sex, and CLL family-history, little is known about factors associated with MBL risk. A polygenic-risk-score (PRS) of 41 ...CLL-susceptibility variants has been found to be associated with CLL risk among individuals of European-ancestry(EA). Here, we evaluate these variants, the PRS, and environmental factors for MBL risk. We also evaluate these variants and the CLL-PRS among African-American (AA) and EA-CLL cases and controls. Our study included 560 EA MBLs, 869 CLLs (696 EA/173 AA), and 2866 controls (2631 EA/235 AA). We used logistic regression, adjusting for age and sex, to estimate odds ratios (OR) and 95% confidence intervals within each race. We found significant associations with MBL risk among 21 of 41 variants and with the CLL-PRS (OR = 1.86, P = 1.9 × 10
, c-statistic = 0.72). Little evidence of any association between MBL risk and environmental factors was observed. We observed significant associations of the CLL-PRS with EA-CLL risk (OR = 2.53, P = 4.0 × 10
, c-statistic = 0.77) and AA-CLL risk (OR = 1.76, P = 5.1 × 10
, c-statistic = 0.62). Inherited genetic factors and not environmental are associated with MBL risk. In particular, the CLL-PRS is a strong predictor for both risk of MBL and EA-CLL, but less so for AA-CLL supporting the need for further work in this population.
Chemotherapy-induced peripheral neuropathy (CIPN) is a common and potentially permanent adverse effect of chemotherapeutic agents including taxanes such as paclitaxel and platinum-based compounds ...such as oxaliplatin and carboplatin. Previous studies have suggested that genetics may impact the risk of CIPN. We conducted genome-wide association studies (GWASs) for CIPN in two independent populations who had completed European Organisation for Research and Treatment of Cancer Quality of Life Questionnaire (EORTC QLQ)-CIPN20 assessments (a CIPN-specific 20-item questionnaire which includes three scales that evaluate sensory, autonomic, and motor symptoms). The study population N08Cx included 692 participants from three clinical trials (North Central Cancer Treatment Group (NCCTG) N08C1, N08CA, and N08CB) who had been treated with paclitaxel, paclitaxel plus carboplatin, or oxaliplatin. The primary endpoint for the GWAS was the change from pre-chemotherapy CIPN20 sensory score to the worse score over the following 18 weeks. Study population The Mayo Clinic Breast Disease Registry (MCBDR) consisted of 381 Mayo Clinic Breast Disease Registry enrollees who had been treated with taxane or platinum-based chemotherapy. The primary endpoint for the GWAS assessed was the earliest CIPN20 sensory score available after the completion of chemotherapy. In multivariate model analyses, chemotherapy regimen (
= 3.0 × 10
) and genetic ancestry (
= 0.007) were significantly associated with CIPN in the N08Cx population. Only age (
= 0.0004) was significantly associated with CIPN in the MCBDR population. The SNP most associated with CIPN was rs56360211 near
(
=7.92 × 10
) in N08Cx and rs113807868 near
in the MCBDR (
= 1.27 × 10
). Due to a lack of replication, we cannot conclude that we identified any genetic predictors of CIPN.
We investigated monogenic and polygenic causes of hypercholesterolemia in a population-based cohort, excluding secondary hypercholesterolemia, and using an established framework to identify ...pathogenic variants. We studied 1682 individuals (50.2 ± 8.6 years, 41.3% males) from southeast Minnesota with primary hypercholesterolemia (low-density lipoprotein cholesterol (LDL-C) ≥155 mg/dl in the absence of identifiable secondary causes). Familial hypercholesterolemia (FH) phenotype was defined as a Dutch Lipid Clinic Network (DLCN) score ≥6. Participants underwent sequencing of LDLR, APOB, and PCSK9, and genotyping of 12 LDL-C-associated single-nucleotide variants to construct a polygenic score (PGS) for LDL-C. The presence of a pathogenic/likely pathogenic variant was considered monogenic etiology and a PGS ≥90th percentile was considered polygenic etiology. The mean LDL-C level was 187.3 ± 32.3 mg/dl and phenotypic FH was present in 8.4% of the cohort. An identifiable genetic etiology was present in 17.1% individuals (monogenic in 1.5% and polygenic in 15.6%). Phenotypic and genetic FH showed poor overlap. Only 26% of those who met the clinical criteria of FH had an identifiable genetic etiology and of those with an identifiable genetic etiology only 12.9% met clinical criteria for FH. Genetic factors explained 7.4% of the variance in LDL-C. In conclusion, in adults with primary hypercholesterolemia, 17.1% had an identifiable genetic etiology and the overlap between phenotypic and genetic FH was modest.
PurposeThe depth and breadth of clinical data within electronic health record (EHR) systems paired with innovative machine learning methods can be leveraged to identify novel risk factors for complex ...diseases. However, analysing the EHR is challenging due to complexity and quality of the data. Therefore, we developed large electronic population-based cohorts with comprehensive harmonised and processed EHR data.ParticipantsAll individuals 30 years of age or older who resided in Olmsted County, Minnesota on 1 January 2006 were identified for the discovery cohort. Algorithms to define a variety of patient characteristics were developed and validated, thus building a comprehensive risk profile for each patient. Patients are followed for incident diseases and ageing-related outcomes. Using the same methods, an independent validation cohort was assembled by identifying all individuals 30 years of age or older who resided in the largely rural 26-county area of southern Minnesota and western Wisconsin on 1 January 2013.Findings to dateFor the discovery cohort, 76 255 individuals (median age 49; 53% women) were identified from which a total of 9 644 221 laboratory results; 9 513 840 diagnosis codes; 10 924 291 procedure codes; 1 277 231 outpatient drug prescriptions; 966 136 heart rate measurements and 1 159 836 blood pressure (BP) measurements were retrieved during the baseline time period. The most prevalent conditions in this cohort were hyperlipidaemia, hypertension and arthritis. For the validation cohort, 333 460 individuals (median age 54; 52% women) were identified and to date, a total of 19 926 750 diagnosis codes, 10 527 444 heart rate measurements and 7 356 344 BP measurements were retrieved during baseline.Future plansUsing advanced machine learning approaches, these electronic cohorts will be used to identify novel sex-specific risk factors for complex diseases. These approaches will allow us to address several challenges with the use of EHR.