scholarly journals Developing a Primary Care EMR-based Frailty Definition using Machine Learning

2018 ◽  
Vol 3 (4) ◽  
Author(s):  
Sylvia Aponte-Hao ◽  
Bria Mele ◽  
Dave Jackson ◽  
Alan Katz ◽  
Charles Leduc ◽  
...  
Keyword(s):  
Machine Learning ◽  
Primary Care ◽  
Positive Predictive Value ◽  
Predictive Value ◽  
Laboratory Data ◽  
Screening Tools ◽  
Geriatric Syndrome ◽  
Surveillance Network ◽  
Clinical Frailty Scale ◽  
Definition Of

IntroductionFrailty is a geriatric syndrome that is predictive of heightened vulnerability for disability, hospitalization, and mortality. Annually an estimated 250,000 frail Canadians die, and this estimate is expected to double in the next 40 years, as Canadians grow older. Currently there is no single accepted clinical definition of frailty. Objectives and ApproachThe objective of this study was to develop an operational definition of frailty using machine learning that can be applied to a primary care electronic medical record (EMR) database. The Canadian Primary Care Sentinel Surveillance Network (CPCSSN) is a pan-Canadian network of primary care practices that collect de-identified patient information (such as encounter diagnoses, health conditions, and laboratory data) from EMRs. 780 patients from CPCSSN have were randomly selected and assessed by physicians using the Rockwood Clinical Frailty Scale (as frail or not frail), and their clinical characteristics from CPCSSN used to develop the definition using machine-learning. ResultsA total of 8,044 clinical features were extracted from these tables: billing, problem list, encounter diagnosis, labs, medications and referrals. A chi-squared automatic interaction detector (CHAID) approach was selected as the best approach. The bootstrapping process used a cost matrix that prioritized high sensitivity and positive predictive value. 10-fold cross validation was used for validity measures. Key features factored into the algorithm included: diagnosis of dementia (ICD-9 code 290), medications furosemide and vitamins, and use of key word “obstruction” within the billing table. The validation measures with 95% confidence intervals are as follows: sensitivity of 28% (95% CI: 21% to 36%), specificity of 94% (95% CI: 93% to 96%), positive predictive value of 53% (95% CI: 42% to 64%), negative predictive value of 86% (95% CI: 83% to 88%). Conclusion/ImplicationsNo other primary care specific frailty screening tools have sufficient validity. These results suggest heterogeneous diseases require clearly defined features and potentially more sophisticated algorithms to account for heterogeneity. Further research utilizing continuous features and continuous frailty scores may be more suitable in the creation of a case detection algorithm.

scholarly journals EVALUATING THE PREDICTIVE VALIDITY OF THE NIPISSING DISTRICT DEVELOPMENTAL SCREEN IN PRIMARY CARE SETTINGS AT THE 18-MONTH HEALTH SUPERVISION VISIT

2018 ◽  
Vol 23 (suppl_1) ◽  
pp. e37-e37
Author(s):  
Vinusha Gunaseelan ◽  
Patricia Parkin ◽  
Imaan Bayoumi ◽  
Patricia Jiang ◽  
Alexandra Medline ◽  
...  
Keyword(s):  
Primary Care ◽  
Positive Predictive Value ◽  
Diagnostic Accuracy ◽  
Developmental Delay ◽  
Predictive Validity ◽  
Predictive Value ◽  
Screening Tools ◽  
Developmental Screening ◽  
Research Personnel ◽  
Health Supervision

Abstract BACKGROUND The Canadian Paediatric Society (CPS) recommends that every Canadian physician caring for young children provide an enhanced 18-month well-baby visit including the use of a developmental screening tool, such as the Nipissing District Developmental Screen (NDDS). The Province of Ontario implemented an enhanced 18-month well-baby visit specifically emphasizing the NDDS, which is now widely used in Ontario primary care. However, the diagnostic accuracy of the NDDS in identifying early developmental delays in real-world clinical settings is unknown. OBJECTIVES To assess the predictive validity of the NDDS in primary care for identifying developmental delay and prompting a specialist referral at the 18-month health supervision visit. DESIGN/METHODS This was a prospective longitudinal cohort study enrolling healthy children from primary care practices. Parents completed the 18-month NDDS during their child’s scheduled health supervision visit between January 2012 and February 2015. Using a standardized data collection form, research personnel abstracted data from the child’s health records regarding the child’s developmental outcomes following the 18-month assessment. Data collected included confirmed diagnoses of a development delay, specialist referrals, family history, and interventions. Research personnel were blind to the results of the NDDS. We assessed the diagnostic test properties of the NDDS with a confirmed diagnosis of developmental delay as the criterion measure. The specificity, sensitivity, positive predictive value, and negative predictive value were calculated, with 95% confidence intervals. RESULTS We included 255 children with a mean age of 18.5 months (range, 17.5–20.6) and 139 (55%) were male. 102 (40%) screened positive (1+ flag result on their NDDS). A total of 48 (19%) children were referred, and 23 (9%) had a confirmed diagnosis of a developmental delay (speech and language: 14; gross motor: 4; autism spectrum disorder: 3; global developmental delay: 1; developmental delay: 1). The sensitivity was 74% (95% CI: 52–90%), specificity was 63% (95% CI: 57–70%), positive predictive value was 17% (95% CI:10–25%), and the negative predictive value was 96% (95% CI: 92–99%). CONCLUSION For developmental screening tools, sensitivity between 70%-80% and specificity of 80% have been suggested. The NDDS has moderate sensitivity and specificity in identifying developmental delay at the 18-month health supervision visit. The 1+NDDS flag cut-point may lead to overdiagnosis with more children with typical development being referred, leading to longer wait times for specialist referrals among children in need. Future work includes investigating the diagnostic accuracy of combining the NDDS with other screening tools.

scholarly journals Predicting a diagnosis of ankylosing spondylitis using primary care health records – a machine learning approach

2021 ◽  
Author(s):  
Jonathan Kennedy ◽  
Natasha Kennedy ◽  
Roxanne Cooksey ◽  
Ernest Choy ◽  
Stefan Siebert ◽  
...  
Keyword(s):  
Machine Learning ◽  
Primary Care ◽  
Back Pain ◽  
Ankylosing Spondylitis ◽  
General Population ◽  
Positive Predictive Value ◽  
Predictive Value ◽  
Principal Component ◽  
Good Prediction ◽  
X Rays

AbstractAnkylosing spondylitis is the second most common cause of inflammatory arthritis. However, a successful diagnosis can take a decade to confirm from symptom onset (via x-rays). The aim of this study was to use machine learning methods to develop a profile of the characteristics of people who are likely to be given a diagnosis of AS in future.The Secure Anonymised Information Linkage databank was used. Patients with ankylosing spondylitis were identified using their routine data and matched with controls who had no record of a diagnosis of ankylosing spondylitis or axial spondyloarthritis. Data was analysed separately for men and women. The model was developed using feature/variable selection and principal component analysis to develop decision trees. The decision tree with the highest average F value was selected and validated with a test dataset.The model for men indicated that lower back pain, uveitis, and NSAID use under age 20 is associated with AS development. The model for women showed an older age of symptom presentation compared to men with back pain and multiple pain relief medications. The models showed good prediction (positive predictive value 70%-80%) in test data but in the general population where prevalence is very low (0.09% of the population in this dataset) the positive predictive value would be very low (0.33%-0.25%).Machine learning can be used to help profile and understand the characteristics of people who will develop AS, and in test datasets with artificially high prevalence, will perform well. However, when applied to a general population with low prevalence rates, such as that in primary care, the positive predictive value for even the best model would be 1.4%. Multiple models may be needed to narrow down the population over time to improve the predictive value and therefore reduce the time to diagnosis of ankylosing spondylitis.

scholarly journals Developing and validating a primary care EMR-based frailty definition using machine learning

2020 ◽  
Vol 5 (1) ◽  
Author(s):  
Tyler Williamson ◽  
Sylvia Aponte-Hao ◽  
Bria Mele ◽  
Brendan Cord Lethebe ◽  
Charles Leduc ◽  
...  
Keyword(s):  
Machine Learning ◽  
Primary Care ◽  
Sensitivity And Specificity ◽  
Health Spending ◽  
Case Definition ◽  
Complex Nature ◽  
Case Definitions ◽  
Surveillance Network ◽  
Clinical Frailty Scale ◽  
Cross Sectional

Introduction. Individuals who have been identified as frail have an increased state of vulnerability, often leading to adverse health events, increased health spending, and potentially detrimental outcomes. Objective. The objective of this work is to develop and validate a case definition for frailty that can be used in a primary care electronic medical record database. Methods. This is a cross-sectional validation study using data from the Canadian Primary Care Sentinel Surveillance Network (CPCSSN) in Southern Alberta. 52 CPCSSN sentinels assessed a random sample of their own patients using the Rockwood Clinical Frailty scale, resulting in a total of 875 patients to be used as reference standard. Patients must be over the age of 65 and have had a clinic visit within the last 24 months. The case definition for frailty was developed using machine learning methods using CPCSSN records for the 875 patients. Results. Of the 875 patients, 155 (17.7%) were frail and 720 (84.2%) were not frail. Validation metrics of the case definition were: sensitivity and specificity of 0.28, 95% CI (0.21 to 0.36) and 0.94, 95% CI (0.93 to 0.96), respectively; PPV and NPV of 0.53, 95% CI (0.42 to 0.64) and 0.86, 95% CI (0.83 to 0.88), respectively. Conclusion. The low sensitivity and specificity results could be because frailty as a construct remains under-developed and relatively poorly understood due to its complex nature. These results contribute to the literature by demonstrating that case definitions for frailty require expert consensus and potentially more sophisticated algorithms to be successful

scholarly journals A machine learning approach for predicting suicidal thoughts and behaviours among college students

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Melissa Macalli ◽  
Marie Navarro ◽  
Massimiliano Orri ◽  
Marie Tournier ◽  
Rodolphe Thiébaut ◽  
...  
Keyword(s):  
Mental Health ◽  
College Students ◽  
Positive Predictive Value ◽  
Predictive Value ◽  
Screening Tools ◽  
Receiver Operating Curve ◽  
Depression Symptoms ◽  
Suicidal Thoughts ◽  
One Year

AbstractSuicidal thoughts and behaviours are prevalent among college students. Yet little is known about screening tools to identify students at higher risk. We aimed to develop a risk algorithm to identify the main predictors of suicidal thoughts and behaviours among college students within one-year of baseline assessment. We used data collected in 2013–2019 from the French i-Share cohort, a longitudinal population-based study including 5066 volunteer students. To predict suicidal thoughts and behaviours at follow-up, we used random forests models with 70 potential predictors measured at baseline, including sociodemographic and familial characteristics, mental health and substance use. Model performance was measured using the area under the receiver operating curve (AUC), sensitivity, and positive predictive value. At follow-up, 17.4% of girls and 16.8% of boys reported suicidal thoughts and behaviours. The models achieved good predictive performance: AUC, 0.8; sensitivity, 79% for girls, 81% for boys; and positive predictive value, 40% for girls and 36% for boys. Among the 70 potential predictors, four showed the highest predictive power: 12-month suicidal thoughts, trait anxiety, depression symptoms, and self-esteem. We identified a parsimonious set of mental health indicators that accurately predicted one-year suicidal thoughts and behaviours in a community sample of college students.

Calculated globulin as a screening tool for hypogammaglobulinaemia or paraproteins in hospitalized patients

2021 ◽  
pp. 000456322198973
Author(s):  
Teng Hoo ◽  
Ee Mun Lim ◽  
Mina John ◽  
Lloyd D’Orsogna ◽  
Andrew McLean-Tooke
Keyword(s):  
Positive Predictive Value ◽  
Predictive Value ◽  
Screening Tool ◽  
Haematological Malignancy ◽  
Laboratory Data ◽  
Liver Function Tests ◽  
Serum Globulin ◽  
Tertiary Hospitals ◽  
Low Sensitivity ◽  
Diagnostic Delays

Background Calculated globulin fraction is derived from the liver function tests by subtracting albumin from the total protein. Since immunoglobulins comprise the largest component of the serum globulin concentration, increased or decreased calculated globulins and may identify patients with hypogammaglobulinaemia or hypergammaglobulinaemia, respectively. Methods A retrospective study of laboratory data over 2.5 years from inpatients at three tertiary hospitals was performed. Patients with paired calculated globulins and immunoglobulin results were identified and clinical details reviewed. The results of serum electrophoresis testing were also assessed where available. Results A total of 4035 patients had paired laboratory data available. A calculated globulin ≤20 g/L (<2nd percentile) had a low sensitivity (5.8%) but good positive predictive value (82.5%) for hypogammaglobulinaemia (IgG ≤5.7 g/L), with a positive predictive value of 37.5% for severe hypogammaglobulinaemia (IgG ≤3 g/L). Paraproteins were identified in 123/291 (42.3%) of patients with increased calculated globulins (≥42 g/L) who also had a serum electrophoresis performed. Significantly elevated calculated globulin ≥50 g/L (>4th percentile) were seen in patients with either liver disease (37%), haematological malignancy (36%), autoimmune disease (13%) or infections (9%). Conclusions Calculated globulin is an inexpensive and easily available test that assists in the identification of hypogammaglobulinaemia or hypergammaglobulinaemia which may prompt further investigation and reduce diagnostic delays.

scholarly journals Identification of newborns at risk for autism using electronic medical records and machine learning

2019 ◽  
Author(s):  
Rayees Rahman ◽  
Arad Kodesh ◽  
Stephen Z Levine ◽  
Sven Sandin ◽  
Abraham Reichenberg ◽  
...  
Keyword(s):  
Machine Learning ◽  
Autism Spectrum Disorder ◽  
Positive Predictive Value ◽  
Electronic Medical Records ◽  
Predictive Value ◽  
False Positive ◽  
Medical Records ◽  
False Positive Rate ◽  
Autism Spectrum ◽  
Positive Rate

AbstractImportanceCurrent approaches for early identification of individuals at high risk for autism spectrum disorder (ASD) in the general population are limited, where most ASD patients are not identified until after the age of 4. This is despite substantial evidence suggesting that early diagnosis and intervention improves developmental course and outcome.ObjectiveDevelop a machine learning (ML) method predicting the diagnosis of ASD in offspring in a general population sample, using parental electronic medical records (EMR) available before childbirthDesignPrognostic study of EMR data within a single Israeli health maintenance organization, for the parents of 1,397 ASD children (ICD-9/10), and 94,741 non-ASD children born between January 1st, 1997 through December 31st, 2008. The complete EMR record of the parents was used to develop various ML models to predict the risk of having a child with ASD.Main outcomes and measuresRoutinely available parental sociodemographic information, medical histories and prescribed medications data until offspring’s birth were used to generate features to train various machine learning algorithms, including multivariate logistic regression, artificial neural networks, and random forest. Prediction performance was evaluated with 10-fold cross validation, by computing C statistics, sensitivity, specificity, accuracy, false positive rate, and precision (positive predictive value, PPV).ResultsAll ML models tested had similar performance, achieving an average C statistics of 0.70, sensitivity of 28.63%, specificity of 98.62%, accuracy of 96.05%, false positive rate of 1.37%, and positive predictive value of 45.85% for predicting ASD in this dataset.Conclusion and relevanceML algorithms combined with EMR capture early life ASD risk. Such approaches may be able to enhance the ability for accurate and efficient early detection of ASD in large populations of children.Key pointsQuestionCan autism risk in children be predicted using the pre-birth electronic medical record (EMR) of the parents?FindingsIn this population-based study that included 1,397 children with autism spectrum disorder (ASD) and 94,741 non-ASD children, we developed a machine learning classifier for predicting the likelihood of childhood diagnosis of ASD with an average C statistic of 0.70, sensitivity of 28.63%, specificity of 98.62%, accuracy of 96.05%, false positive rate of 1.37%, and positive predictive value of 45.85%.MeaningThe results presented serve as a proof-of-principle of the potential utility of EMR for the identification of a large proportion of future children at a high-risk of ASD.

scholarly journals Machine Learning Models for Abnormality Detection in Musculoskeletal Radiographs

Reports ◽  
2019 ◽  
Vol 2 (4) ◽  
pp. 26 ◽  
Author(s):  
Govind Chada
Keyword(s):  
Machine Learning ◽  
Primary Care ◽  
Deep Learning ◽  
Transfer Learning ◽  
Primary Care Physicians ◽  
Screening Tools ◽  
Learning Approaches ◽  
Limited Data ◽  
Learning Models ◽  
High Recognition Accuracy

Increasing radiologist workloads and increasing primary care radiology services make it relevant to explore the use of artificial intelligence (AI) and particularly deep learning to provide diagnostic assistance to radiologists and primary care physicians in improving the quality of patient care. This study investigates new model architectures and deep transfer learning to improve the performance in detecting abnormalities of upper extremities while training with limited data. DenseNet-169, DenseNet-201, and InceptionResNetV2 deep learning models were implemented and evaluated on the humerus and finger radiographs from MURA, a large public dataset of musculoskeletal radiographs. These architectures were selected because of their high recognition accuracy in a benchmark study. The DenseNet-201 and InceptionResNetV2 models, employing deep transfer learning to optimize training on limited data, detected abnormalities in the humerus radiographs with 95% CI accuracies of 83–92% and high sensitivities greater than 0.9, allowing for these models to serve as useful initial screening tools to prioritize studies for expedited review. The performance in the case of finger radiographs was not as promising, possibly due to the limitations of large inter-radiologist variation. It is suggested that the causes of this variation be further explored using machine learning approaches, which may lead to appropriate remediation.

Comparison of eight screening tools to detect interactions between herbal supplements and oncology agents

2020 ◽  
Vol 26 (8) ◽  
pp. 1843-1849
Author(s):  
Faisal Shakeel ◽  
Fang Fang ◽  
Kelley M Kidwell ◽  
Lauren A Marcath ◽  
Daniel L Hertz
Keyword(s):  
Drug Interaction ◽  
Positive Predictive Value ◽  
Drug Interactions ◽  
Predictive Value ◽  
Screening Tools ◽  
Herbal Supplements ◽  
Interaction Detection ◽  
Natural Medicine ◽  
Interaction Screening ◽  
Drug Drug Interaction

Introduction Patients with cancer are increasingly using herbal supplements, unaware that supplements can interact with oncology treatment. Herb–drug interaction management is critical to ensure optimal treatment outcomes. Several screening tools exist to detect drug–drug interactions, but their performance to detect herb–drug interactions is not known. This study compared the performance of eight drug–drug interaction screening tools to detect herb–drug interaction with anti-cancer agents. Methods The herb–drug interaction detection performance of four subscription (Micromedex, Lexicomp, PEPID, Facts & Comparisons) and free (Drugs.com, Medscape, WebMD, RxList) drug–drug interaction tools was assessed. Clinical relevance of each herb–drug interaction was determined using Natural Medicine and each drug–drug interaction tool. Descriptive statistics were used to calculate sensitivity, specificity, positive predictive value, and negative predictive value. Linear regression was used to compare performance between subscription and free tools. Results All tools had poor sensitivity (<0.20) for detecting herb–drug interaction. Lexicomp had the highest positive predictive value (0.98) and best overall performance score (0.54), while Medscape was the best performing free tool (0.52). The worst subscription tools were as good as or better than the best free tools, and as a group subscription tools outperformed free tools on all metrics. Using an average subscription tool would detect one additional herb–drug interaction for every 10 herb–drug interactions screened by a free tool. Conclusion Lexicomp is the best available tool for screening herb–drug interaction, and Medscape is the best free alternative; however, the sensitivity and performance for detecting herb–drug interaction was far lower than for drug–drug interactions, and overall quite poor. Further research is needed to improve herb–drug interaction screening performance.

Automated screening tool for Subcutaneous Implantable Defibrillator in Brugada syndrome has a high eligibility rate which is predicted by simple electrocardiographic parameters

2020 ◽  
Vol 41 (Supplement_2) ◽  
Author(s):  
S Marelli ◽  
D Kukavica ◽  
A Mazzanti ◽  
T Chargeishvili ◽  
A Trancuccio ◽  
...  
Keyword(s):  
Positive Predictive Value ◽  
Brugada Syndrome ◽  
Predictive Value ◽  
Screening Tool ◽  
Screening Tools ◽  
Molecular Medicine ◽  
Funding Source ◽  
S Wave ◽  
High Positive Predictive Value ◽  
Automated Screening

Abstract Background Manual electrocardiographic (ECG) screening tools for the use of subcutaneous cardiac defibrillator (S-ICD) have been associated with high ineligibility rates in Brugada syndrome patients (BrS). Although recent works identified ECG parameters for S-ICD eligibility in general population, automated screening tool (AST) for S-ICD eligibility have not even been assessed in large series of patients with BrS. Purpose This study evaluates the AST-derived eligibility rates for an S-ICD in patients with BrS, and ECG parameters associated with S-ICD eligibility. Methods Screening for S-ICD eligibility was performed using AST in 194 consecutive patients with BrS. Eligibility was defined when at least one of the three vectors was acceptable both in supine and standing position. Twelve-lead ECGs were registered during the screening. ECG parameters associated with AST eligibility were identified using multivariable logistical regression. Results Our study population consisted of 194 patients, with male preponderance (n=165/194; 85%); and were 43±12 years old at the time of screening. Majority of patients presented a spontaneous type 1 pattern during screening (n=128/194; 66%), with an average pattern height of 3±3 mm. Remarkably, 93% of patients passed the screening with AST. No differences in eligibility rates in terms of gender (93% males vs. 93% females eligible; p=1) and age (48±9 years non-eligible vs. 42±12 eligible; p=0.07) existed. Notably, our eligibility rate was 2.5 times higher than rates reported in literature when using manual screening tools (p=0.023). Independent 12-lead ECG parameters (Table) associated with AST eligibility were duration of S wave &lt;80 ms in aVF and R/T ratio ≥3 in lead II (Figure), which have a high positive predictive value (97% and 99%, respectively) for screening eligibility. Conclusions Most BrS patients (93%) are eligible for S-ICD when AST is used. S wave &lt;80 ms in aVF, and R/T ratio ≥3 in lead II have a high positive predictive value for S-ICD eligibility. Funding Acknowledgement Type of funding source: Public grant(s) – National budget only. Main funding source(s): The Italian Ministry of Research and University Dipartimenti di Eccellenza 2018–2022 grant to the Molecular Medicine Department (University of Pavia)

scholarly journals The study of automatic machine learning base on radiomics of non-focus area in the first chest CT of different clinical types of COVID-19 pneumonia

2020 ◽  
Vol 10 (1) ◽  
Author(s):  
Hui-Bin Tan ◽  
Fei Xiong ◽  
Yuan-Liang Jiang ◽  
Wen-Cai Huang ◽  
Ye Wang ◽  
...  
Keyword(s):  
Machine Learning ◽  
Predictive Value ◽  
Laboratory Data ◽  
Texture Features ◽  
Chest Ct ◽  
Automatic Machine ◽  
Classification Model ◽  
Control Group ◽  
Ct Image ◽  
Severe Group

Abstract To explore the possibility of predicting the clinical types of Corona-Virus-Disease-2019 (COVID-19) pneumonia by analyzing the non-focus area of the lung in the first chest CT image of patients with COVID-19 by using automatic machine learning (Auto-ML). 136 moderate and 83 severe patients were selected from the patients with COVID-19 pneumonia. The clinical and laboratory data were collected for statistical analysis. The texture features of the Non-focus area of the first chest CT of patients with COVID-19 pneumonia were extracted, and then the classification model of the first chest CT of COVID-19 pneumonia was constructed by using these texture features based on the Auto-ML method of radiomics, The area under curve(AUC), true positive rate(TPR), true negative rate (TNR), positive predictive value(PPV) and negative predictive value (NPV) of the operating characteristic curve (ROC) were used to evaluate the accuracy of the first chest CT image classification model in patients with COVID-19 pneumonia. The TPR, TNR, PPV, NPV and AUC of the training cohort and test cohort of the moderate group and the control group, the severe group and the control group, the moderate group and the severe group were all greater than 95% and 0.95 respectively. The non-focus area of the first CT image of COVID-19 pneumonia has obvious difference in different clinical types. The AUTO-ML classification model of Radiomics based on this difference can be used to predict the clinical types of COVID-19 pneumonia.

Sign in / Sign up

Export Citation Format

Share Document