Dataset Information

Development of an Artificial Intelligence-Based Automated Recommendation System for Clinical Laboratory Tests: Retrospective Analysis of the National Health Insurance Database.

ABSTRACT:

Background

Laboratory tests are considered an essential part of patient safety as patients' screening, diagnosis, and follow-up are solely based on laboratory tests. Diagnosis of patients could be wrong, missed, or delayed if laboratory tests are performed erroneously. However, recognizing the value of correct laboratory test ordering remains underestimated by policymakers and clinicians. Nowadays, artificial intelligence methods such as machine learning and deep learning (DL) have been extensively used as powerful tools for pattern recognition in large data sets. Therefore, developing an automated laboratory test recommendation tool using available data from electronic health records (EHRs) could support current clinical practice.

Objective

The objective of this study was to develop an artificial intelligence-based automated model that can provide laboratory tests recommendation based on simple variables available in EHRs.

Methods

A retrospective analysis of the National Health Insurance database between January 1, 2013, and December 31, 2013, was performed. We reviewed the record of all patients who visited the cardiology department at least once and were prescribed laboratory tests. The data set was split into training and testing sets (80:20) to develop the DL model. In the internal validation, 25% of data were randomly selected from the training set to evaluate the performance of this model.

Results

We used the area under the receiver operating characteristic curve, precision, recall, and hamming loss as comparative measures. A total of 129,938 prescriptions were used in our model. The DL-based automated recommendation system for laboratory tests achieved a significantly higher area under the receiver operating characteristic curve (AUROCmacro and AUROCmicro of 0.76 and 0.87, respectively). Using a low cutoff, the model identified appropriate laboratory tests with 99% sensitivity.

Conclusions

The developed artificial intelligence model based on DL exhibited good discriminative capability for predicting laboratory tests using routinely collected EHR data. Utilization of DL approaches can facilitate optimal laboratory test selection for patients, which may in turn improve patient safety. However, future study is recommended to assess the cost-effectiveness for implementing this model in real-world clinical settings.

SUBMITTER: Islam MM

PROVIDER: S-EPMC7710445 | biostudies-literature | 2020 Nov

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Development of an Artificial Intelligence-Based Automated Recommendation System for Clinical Laboratory Tests: Retrospective Analysis of the National Health Insurance Database.

Islam Md Mohaimenul MM Yang Hsuan-Chia HC Poly Tahmina Nasrin TN Li Yu-Chuan Jack YJ

JMIR medical informatics 20201118 11

<h4>Background</h4>Laboratory tests are considered an essential part of patient safety as patients' screening, diagnosis, and follow-up are solely based on laboratory tests. Diagnosis of patients could be wrong, missed, or delayed if laboratory tests are performed erroneously. However, recognizing the value of correct laboratory test ordering remains underestimated by policymakers and clinicians. Nowadays, artificial intelligence methods such as machine learning and deep learning (DL) have been ...[more]

PMID: 33206057

Similar Datasets

Project description:ObjectivesDementia is common in people over the age of 65 years, with 80% of people with dementia older than 75 years. Previous studies have linked dementia to late-life depression, but the association between dementia and mid-life depression is poorly understood. Depression is a preventable and treatable medical condition, which means it is a modifiable factor that can potentially prevent or delay dementia. This study aimed to identify the association between dementia and depression within the life course.DesignA nationwide, retrospective propensity score matched cohort study associating dementia with depression. Depression diagnosed between the ages of 45 and 64 years was classified as 'mid-life' and 'late-life' if diagnosed at 65 years or older. Patients were considered to have depression when one or more International Statistical Classification of Diseases and Related Health Problems, 10th revision codes for depression were recorded as primary or secondary diagnosis.SettingNational Health Insurance Service-National Sample Cohort database of the National Health Insurance Service in South Korea, containing patient data from 2002 to 2013.ParticipantsThe study included 1824 and 374 852 patients in the case and control groups, respectively. A logistic regression analysis with complex sampling design was performed after adjusting for covariates, using the propensity score matching method without callipers, with a 1:1 nearest neighbour matching algorithm.Primary and secondary outcome measuresThe association of mid-onset and late-onset depression with dementia in terms of sociodemographic characteristics, such as sex and age, within the Korean population.ResultsDementia was significantly associated with the presence of depression (OR=2.20, 95% CI=1.53-3.14); in particular, female patients with depression and patients aged 45-64 years with depression had increased odds of dementia (OR=2.65, 95% CI=1.78-3.93 and OR=2.72, 95% CI=1.41-5.24, respectively) CONCLUSION: Depression is an associated factor for dementia, especially among people aged 45-64 years (mid-life).

Project description:OBJECTIVE:Statins have been commonly used to treat patients with hypercholesterolaemia and to prevent cardiovascular disease (CVD) worldwide. This study examined trends in use of statins in Taiwan from 2002 to 2011. DESIGN:This is a retrospective observational study focusing on the utilisation of statins. SETTING:The monthly claims data for statins between 2002 and 2011 were retrieved from Taiwan's National Health Insurance Research Database. MAIN OUTCOME MEASURES:We calculated the yearly prescription rate per new user for each statin. Products were classified as high-intensity/moderate-intensity/low-intensity statins by type of statin and dosage. Users were also classified based on disease histories. RESULTS:The number of statin users increased from 10?299 (~1.4% of adults) in 2002 to 50?687 (~6.3% of adults) in 2011. Atorvastatin was the most commonly used agent (28.4%-36.7%) during the study period. After 2007, simvastatin ranked second with 21.7% market share, followed by rosuvastatin, a newer agent that exhibited a substantial growth in prescription rates (3.4% in 2005 and 19.5% in 2011). In 2011, 94.0% of new statin users used statin monotherapies, and 6.0% used combination therapies. Use of moderate-intensity statins increased from 49.0% in 2002 to 71.0% in 2011, while high-intensity statins remained low. Patients with history of coronary events or cerebrovascular events were more likely to be prescribed higher intensity statins compared with those without. Prescribing of higher intensity statins was not greater among people with diabetes compared with those without during 2007-2011. Selection of statins did not differ between people with versus without history of myopathy or liver injury. CONCLUSION:Atorvastatin was the most commonly used statin in Taiwan during 2002-2011. While patients with history of CVD were more likely to be prescribed higher intensity statins compared with those without, this difference was not found comparing those with and without diabetes.

Dataset Information

Development of an Artificial Intelligence-Based Automated Recommendation System for Clinical Laboratory Tests: Retrospective Analysis of the National Health Insurance Database.

Background

Objective

Methods

Results

Conclusions

Publications

Development of an Artificial Intelligence-Based Automated Recommendation System for Clinical Laboratory Tests: Retrospective Analysis of the National Health Insurance Database.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure