scholarly journals Reducing False-Positive Results in Newborn Screening Using Machine Learning

2020 ◽  
Vol 6 (1) ◽  
pp. 16 ◽  
Author(s):  
Gang Peng ◽  
Yishuo Tang ◽  
Tina M. Cowan ◽  
Gregory M. Enns ◽  
Hongyu Zhao ◽  
...  
Keyword(s):  
Machine Learning ◽  
Random Forest ◽  
Newborn Screening ◽  
False Positive ◽  
Clinical Laboratory ◽  
False Positives ◽  
Complex Data ◽  
Interpretive Analysis ◽  
Chain Acyl ◽  
Positive Results

Newborn screening (NBS) for inborn metabolic disorders is a highly successful public health program that by design is accompanied by false-positive results. Here we trained a Random Forest machine learning classifier on screening data to improve prediction of true and false positives. Data included 39 metabolic analytes detected by tandem mass spectrometry and clinical variables such as gestational age and birth weight. Analytical performance was evaluated for a cohort of 2777 screen positives reported by the California NBS program, which consisted of 235 confirmed cases and 2542 false positives for one of four disorders: glutaric acidemia type 1 (GA-1), methylmalonic acidemia (MMA), ornithine transcarbamylase deficiency (OTCD), and very long-chain acyl-CoA dehydrogenase deficiency (VLCADD). Without changing the sensitivity to detect these disorders in screening, Random Forest-based analysis of all metabolites reduced the number of false positives for GA-1 by 89%, for MMA by 45%, for OTCD by 98%, and for VLCADD by 2%. All primary disease markers and previously reported analytes such as methionine for MMA and OTCD were among the top-ranked analytes. Random Forest’s ability to classify GA-1 false positives was found similar to results obtained using Clinical Laboratory Integrated Reports (CLIR). We developed an online Random Forest tool for interpretive analysis of increasingly complex data from newborn screening.

scholarly journals Reduction in Newborn Screening False Positive Results Following a New Collection Protocol: a Quality Improvement Project

2021 ◽  
Vol 26 (7) ◽  
pp. 723-727
Author(s):  
May Kamleh ◽  
Julia Muzzy Williamson ◽  
Kari Casas ◽  
Mohamed Mohamed
Keyword(s):  
Quality Improvement ◽  
Newborn Screening ◽  
False Positive ◽  
Improve Patient Care ◽  
False Positives ◽  
Sample Collection ◽  
Improvement Project ◽  
Period 2 ◽  
Amino Acid Profiles ◽  
Positive Results

OBJECTIVE Premature infants are known to have a higher rate of false positive newborn screening (NBS) results, with TPN as a contributing factor. The purpose of this quality improvement (QI) project is to reduce false positive NBS results via a TPN interruption protocol METHODS A multidisciplinary team reviewed the literature and developed a new NBS collection protocol, which was implemented in 2 periods. In period 1, TPN was interrupted for 4 hours before NBS sample collection and initiation of carnitine supplements was avoided. In period 2, TPN was interrupted for 6 hours for infants birth weight (BW) < 1000 g, carnitine supplementation continued to be avoided. The rates of false positives NBS results were compared pre- and post-interventions in periods 1 and 2. RESULTS Four hundred twelve neonates were evaluated prior to implementation of this QI project (July 2013–June 2014) and 414 during period 1 intervention (July 2014–June 2016). False positive results decreased from 20.6% to 11.4% (p < 0.001) among all BW categories following the 4-hour TPN interruption. The rate of false positives was further reduced among infants < 1000 g (p = 0.035) in period 2 (n = 112), including a significant reduction in false positive results with elevated amino acid profiles (p = 0.005). CONCLUSIONS The implementation of a strict NBS collection protocol reduced false positive NBS results, which potentially can improve patient care by reducing unnecessary laboratory draws, pain, and parental anxiety. Interruption of TPN for 6 hours was significant in reducing NBS false positive results in neonates < 1000 g.

scholarly journals Psychological Impact of NBS for CF

2020 ◽  
Vol 6 (2) ◽  
pp. 27 ◽  
Author(s):  
Jane Chudleigh ◽  
Holly Chinnery
Keyword(s):  
Cystic Fibrosis ◽  
Newborn Screening ◽  
Health Professionals ◽  
False Positive ◽  
Psychological Impact ◽  
Confirmatory Test ◽  
Test Results ◽  
Burden Of Care ◽  
Positive Results ◽  
Screening Result

Newborn screening for cystic fibrosis has resulted in diagnosis often before symptoms are recognised, leading to benefits including reduced disease severity, decreased burden of care, and lower costs. The psychological impact of this often unsought diagnosis on the parents of seemingly well children is less well understood. The time during which the screening result is communicated to families but before the confirmatory test results are available is recognised as a period of uncertainty and it is this uncertainty that can impact most on parents. Evidence suggests this may be mitigated against by ensuring the time between communication and confirmatory testing is minimized and health professionals involved in communicating positive newborn screening results and diagnostic results for cystic fibrosis to families are knowledgeable and able to provide appropriate reassurance. This is particularly important in the case of false positive results or when the child is given a Cystic Fibrosis Screen Positive, Inconclusive Diagnosis designation. However, to date, there are no formal mechanisms in place to support health professionals undertaking this challenging role, which would enable them to meet the expectations set out in specific guidance.

scholarly journals Use of a Disposable Water Filter for Prevention of False-Positive Results due to Nontuberculosis Mycobacteria in a Clinical Laboratory Performing Routine Acid-Fast Staining for Tuberculosis

2007 ◽  
Vol 73 (19) ◽  
pp. 6296-6298 ◽  
Author(s):  
Hui-Zin Tu ◽  
Chiao-Shan Chen ◽  
Tsi-Shu Huang ◽  
Wen-Kuei Huang ◽  
Yao-shen Chen ◽  
...  
Keyword(s):  
False Positive ◽  
Clinical Laboratory ◽  
Tap Water ◽  
False Positive Rate ◽  
Acid Fast Bacillus ◽  
Water Filter ◽  
Point Of Use ◽  
Positive Rate ◽  
Positive Results ◽  
Laboratory Water

ABSTRACT A point-of-use 0.2-μm filter was evaluated for elimination of nontuberculosis mycobacteria in laboratory water to reduce false-positive acid-fast bacillus staining results. Use of the point-of-use filter can significantly reduce the false-positive rate to 1.2% compared to samples treated with tap water (10.7%) and deionized water (8.7%).

Reduction in False-Positive PF4/Heparin Antibody Testing Following Implementation of a Latex Immunoturbidimetric Assay

2019 ◽  
Vol 152 (Supplement_1) ◽  
pp. S34-S34
Author(s):  
Chiraag Gangahar ◽  
Daniel Webber ◽  
Ronald Jackups
Keyword(s):  
Clinical Judgment ◽  
Test Performance ◽  
False Positive ◽  
Clinical Laboratory ◽  
Turnaround Time ◽  
Serotonin Release ◽  
Reference Laboratory ◽  
Antibody Testing ◽  
Release Assay ◽  
Positive Results

Abstract Background Heparin-induced thrombocytopenia (HIT) is a life-threatening complication of exposure to heparin that is caused by autoantibodies against heparin-PF4 complexes. We recently changed our in-house HIT screening platform from a manual, daily batched ELISA (Stago-Asserochrom HPIA Immunoassay) to an automated, on-demand latex immunoturbidimetric assay (LIA, HemosIL HIT-Ab) and have also implemented a reflex from a positive LIA result to the confirmatory serotonin release assay (SRA). We compared the two methods in terms of utilization, test performance, and turnaround time. Methods Data were collected retrospectively from a 7-month period before (June-December 2017) and after (June-December 2018) implementation of the HemosIL LIA in the clinical laboratory at a large academic institution. This study includes consecutive test results from adults (median age: 64 years, range: 19-98 years) seen at our 1,300-bed main hospital. Test utilization, turnaround time (sample receipt to verification), and test performance characteristics were compared between the two methods. Repeat testing was excluded from the analysis. Samples with a positive result on the HemosIL LIA were reflexed to a serotonin release assay (SRA), performed at a large reference laboratory, whereas samples tested with the earlier ELISA assay were referred for SRA testing based upon clinical judgment. When performed, SRA was considered the gold standard for diagnosis of HIT. Results During the 7 months before and after switching methods, there were 109 of 594 (18.4%) positive ELISA results and 45 of 523 (8.6%) positive LIA results. Only 90 of 109 (82%) of the positive results from the ELISA HIT Ab test were sent out by clinicians for SRA testing, whereas 45 of 45 (100%) of the positive results from LIA testing were reflexed to SRA per protocol. Although fewer LIA tests were sent out for SRA testing, there were an equal number of SRA-confirmed cases of HIT with the ELISA (PPV: 16/90 [17.8%]) and LIA methods (PPV: 16/45 [35.6%]), resulting in a high positive predictive value (PPV) with the newly implemented method. Not only was the PPV higher with the LIA test, but it had a significantly shorter mean turnaround time of 96 minutes compared to the ELISA TAT of 1,234 minutes (P < .0001). Conclusions With the new testing protocol, patients received results faster (average 96-minute TAT) and had fewer false-positive results (74/594 pre vs 29/523 post), with no apparent reduction in detection of true-positive cases of HIT (16/594 pre vs 16/523 post).

scholarly journals The Prediction of Essential Medicines Demand: A Machine Learning Approach Using Consumption Data in Rwanda

Processes ◽  
2021 ◽  
Vol 10 (1) ◽  
pp. 26
Author(s):  
Francois Mbonyinshuti ◽  
Joseph Nkurunziza ◽  
Japhet Niyobuhungiro ◽  
Egide Kayitare
Keyword(s):  
Machine Learning ◽  
Random Forest ◽  
Predictive Modeling ◽  
Essential Medicines ◽  
Complex Data ◽  
Global Business ◽  
Oral Rehydration ◽  
Consumption Data ◽  
Machine Learning Approach ◽  
And Performance

Today’s global business trends are causing a significant and complex data revolution in the healthcare industry, culminating in the use of artificial intelligence and predictive modeling to improve health outcomes and performance. The dataset, which was referred to is based on consumption data from 2015 to 2019, included approximately 500 goods. Based on a series of data pre-processing activities, the top ten (10) essential medicines most used were chosen, namely cotrimoxazole 480 mg, amoxicillin 250 mg, paracetamol 500 mg, oral rehydration salts (O.R.S) sachet 20.5 g, chlorpheniramine 4 mg, nevirapine 200 mg, aminophylline 100 mg, artemether 20 mg + lumefantrine (AL) 120 mg, Cromoglycate ophthalmic. Our study concentrated on the application of machine learning (ML) to forecast future trends in the demand for essential drugs in Rwanda. The following models were created and applied: linear regression, artificial neural network, and random forest. The random forest was able to predict 10 selected medicines with an accuracy of 88 percent with the train set and 76 percent with the test set, and it can thus be used to forecast future demand based on past consumption data by inputting a month, year, district, and medicine name. According to our findings, the random Forest model performed well as a forecasting model for the demand for essential medicines. Finally, data-driven predictive modeling with machine learning (ML) could become the cornerstone of health supply chain planning and operational management.

scholarly journals NeoSeq: a new method of genomic sequencing for newborn screening

2021 ◽  
Vol 16 (1) ◽  
Author(s):  
Huaiyan Wang ◽  
Yuqi Yang ◽  
Lingna Zhou ◽  
Yu Wang ◽  
Wei Long ◽  
...  
Keyword(s):  
Newborn Screening ◽  
False Positive ◽  
Disease Risk ◽  
Amplicon Sequencing ◽  
Dried Blood Spots ◽  
Inborn Errors ◽  
Screening Cost ◽  
Diagnostic Time ◽  
Significant Difference ◽  
Positive Results

Abstract Objective To explore the clinical application of NeoSeq in newborn screening. Methods Based on the results obtained from traditional newborn screening (NBS) with tandem mass spectrometry (TMS), three cohorts were recruited into the present study: 36 true positive cases (TPC), 60 false-positive cases (FPC), and 100 negative cases. The dried blood spots of the infants were analyzed with NeoSeq, which is based on multiplex PCR amplicon sequencing. Results Overall, the sensitivity of NeoSeq was 55.6% (20/36) in the detection of TPC. NeoSeq detected disease-related genes in 20 of 36 TPC infants, while it could not identify these genes in eight children. Five cases (3.1%) with disease risk were additionally found in the FPC and NC cohorts. There was a significant difference in the diagnostic time between the two methods—10 days for NeoSeq vs. 43 days for traditional NBS. Conclusions NeoSeq is an economic genomic screening test for newborn screening. It can detect most inborn errors of metabolism, reduce the rate of false positive results, shorten the porting cycles, and reduce the screening cost. However, it is still necessary to further optimize the panel design and add more clinically relevant genomic variants to increase its sensitivity.

scholarly journals Score and Correlation Coefficient-Based Feature Selection for Predicting Heart Failure Diagnosis by Using Machine Learning Algorithms

2021 ◽  
Vol 2021 ◽  
pp. 1-16
Author(s):  
Ebrahim Mohammed Senan ◽  
Ibrahim Abunadi ◽  
Mukti E. Jadhav ◽  
Suliman Mohamed Fati
Keyword(s):  
Machine Learning ◽  
Heart Failure ◽  
Random Forest ◽  
Clinical Laboratory ◽  
Learning Algorithms ◽  
Machine Learning Algorithms ◽  
Common Causes ◽  
Chi Squared ◽  
Clinical Records ◽  
Heart Failure Diagnosis

Cardiovascular disease (CVD) is one of the most common causes of death that kills approximately 17 million people annually. The main reasons behind CVD are myocardial infarction and the failure of the heart to pump blood normally. Doctors could diagnose heart failure (HF) through electronic medical records on the basis of patient’s symptoms and clinical laboratory investigations. However, accurate diagnosis of HF requires medical resources and expert practitioners that are not always available, thus making the diagnosing challengeable. Therefore, predicting the patients’ condition by using machine learning algorithms is a necessity to save time and efforts. This paper proposed a machine-learning-based approach that distinguishes the most important correlated features amongst patients’ electronic clinical records. The SelectKBest function was applied with chi-squared statistical method to determine the most important features, and then feature engineering method has been applied to create new features correlated strongly in order to train machine learning models and obtain promising results. Optimised hyperparameter classification algorithms SVM, KNN, Decision Tree, Random Forest, and Logistic Regression were used to train two different datasets. The first dataset, called Cleveland, consisted of 303 records. The second dataset, which was used for predicting HF, consisted of 299 records. Experimental results showed that the Random Forest algorithm achieved accuracy, precision, recall, and F1 scores of 95%, 97.62%, 95.35%, and 96.47%, respectively, during the test phase for the second dataset. The same algorithm achieved accuracy scores of 100% for the first dataset and 97.68% for the second dataset, while 100% precision, recall, and F1 scores were reached for both datasets.

Sign in / Sign up

Export Citation Format

Share Document