Machine learning-based clinical decision support using laboratory data.

Artificial intelligence (AI) and machine learning (ML) are becoming vital in laboratory medicine and the broader context of healthcare. In this review article, we summarized the development of ML models and how they contribute to clinical laboratory workflow and improve patient outcomes. The process of ML model development involves data collection, data cleansing, feature engineering, model development, and optimization. These models, once finalized, are subjected to thorough performance assessments and validations. Recently, due to the complexity inherent in model development, automated ML tools were also introduced to streamline the process, enabling non-experts to create models. Clinical Decision Support Systems (CDSS) use ML techniques on large datasets to aid healthcare professionals in test result interpretation. They are revolutionizing laboratory medicine, enabling labs to work more efficiently with less human supervision across pre-analytical, analytical, and post-analytical phases. Despite contributions of the ML tools at all analytical phases, their integration presents challenges like potential model uncertainties, black-box algorithms, and deskilling of professionals. Additionally, acquiring diverse datasets is hard, and models' complexity can limit clinical use. In conclusion, ML-based CDSS in healthcare can greatly enhance clinical decision-making. However, successful adoption demands collaboration among professionals and stakeholders, utilizing hybrid intelligence, external validation, and performance assessments.

Training and Competency Strategies for Point-of-Care Testing.

The increased availability and use of POCT are being influenced by many factors, such as; industry trends to move towards patient-centered care and healthcare decentralization, the increasing prevalence of infectious diseases also including the current use of Rapid SARS-CoV-2 Testing, a growing incidence of lifestyle diseases such as diabetes, heart disease, and hypertension, as well as advances in in-vitro diagnostic medical technologies. The use of POCT can increase the efficiency of services and improve outcomes for patients. However, the variability of the testing environment and conditions as well as the competency of staff performing the tests may have a significant impact on the quality and accuracy of POCT results. A majority of the staff who perform POCT are not trained laboratory staff and may not be as knowledgeable about the processes involved in testing, such as patient preparation, sample collection, management of equipment and supplies, instrument calibration and maintenance, the performance of the test, quality control, interpretation of the results, and reporting/documentation of results in each patient's context. Therefore, staff performing POCT must have the proper training and experience to ensure test results are accurate and reliable. This short communication outlines the specific requirements for staff training based on international standards which need to be considered to ensure the quality of test results and describes competency criteria required for compliance with POCT.

Autoverification of test results in the core clinical laboratory.

Verification of laboratory test results represents the last opportunity to identify errors before they become part of the electronic medical record. Manual verification of test results places significant reliance on the experience and attentiveness of individual observers to identify errors and is vulnerable to errors through omission and neglect. Peer-reviewed publications have documented gains in process efficiency and quality improvement by use of middleware or laboratory information systems to autoverify test results based on pre-defined acceptability criteria. This review evaluates the acceptability of autoverification (AV) as a safe and reliable alternative to total manual review of laboratory test results. AV schemes developed in accordance with international guidelines and standards are applied throughout the laboratory. Careful design of AV systems involves using multidisciplinary teams to develop test-specific decision algorithms, to assist with programming, to verify programming, and validate programmed algorithms prior to use in evaluation of patient test result profiles. Development of test specific decision algorithms makes use of criteria based on instrument messages and flags, quality control status, result limit checks, delta checks, critical values, consistency checks, and patient-related clinical information. Monitoring of the performance of AV parameters, and regular audits of the AV system integrity is recommended in both the literature and guidelines. The potential for gains to process efficiency, error detection and patient safety, through adoption of AV as part of a laboratories quality assurance tool-case, is well supported in published literature.

Delta Checks in the clinical laboratory.

International standards and practice guidelines recommend the use of delta check alerts for laboratory test result interpretation and quality control. The value of contemporary applications of simple univariate delta checks determined as an absolute change, percentage change, or rate of change to recognize specimen misidentification or other laboratory errors has not received much study. This review addresses these three modes of calculation, but in line with the majority of published work, most attention is focused on the identification of specimen misidentification errors. Investigation of delta check alerts are time-consuming and the yield of identified errors is usually small compared to the number of delta check alerts; however, measured analytes with low indices of individuality frequently perform better. While multivariate approaches to delta checks suggest improved usefulness over simple univariate delta check strategies, some of these are complex and not easily applied in contemporary laboratory information systems and middleware. Nevertheless, a simple application of delta checks may hold value in identifying clinically significant changes in several clinical situations: for acute kidney injury using changes in serum creatinine, for risk of osmotic demyelination syndrome using rapid acute changes in serum sodium levels, or for early triage of chest pain patients using high sensitivity troponin assays. A careful and highly selective approach to identifying delta check analytes, calculation modes, and thresholds before putting them into practice is warranted; then follow-up with careful monitoring of performance and balancing true positives, false negatives, and false positives among delta check alerts is needed.

Vitamin D levels in patients with breast cancer: importance of dressing style.

BACKGROUND: Vitamin D deficiency is a potentially modifiable risk factor that may be targeted for breast cancer (BC) prevention. It may also be related to prognosis after diagnosis and treatment. The aim of our study was to determine the prevalence of vitamin D deficiency as measured by serum 25-hydroxy vitamin D (25-OHD) levels in patients with BC and to evaluate its correlations with life-style and treatments. MATERIALS AND METHODS: This study included 186 patients with stage 0-III BC treated in our breast center between 2010-2013. The correlation between serum baseline 25-OHD levels and supplement usage, age, menopausal status, diabetes mellitus, usage of bisphosphonates, body-mass index (BMI), season, dressing style, administration of systemic treatments and radiotherapy were investigated. The distribution of serum 25-OHD levels was categorized as deficient (<10ng/ ml), insufficient (10-24 ng/ml), and sufficient (25-80 ng/ml). RESULTS: The median age of the patients was 51 years (range: 27-79 years) and 70% of them had deficient/insufficient 25-OHD levels. On univariate analysis, vitamin D deficiency/insufficiency was more common in patients with none or low dose vitamin D supplementation at the baseline, high BMI (≥25), no bisphosphonate usage, and a conservative dressing style. On multivariate analysis, none or low dose vitamin D supplementation, and decreased sun-exposure due to a conservative dressing style were found as independent factors increasing risk of vitamin D deficiency/insufficiency 28.7 (p=0.002) and 13.4 (p=0.003) fold, respectively. CONCLUSIONS: The prevalence of serum 25-OHD deficiency/insufficiency is high in our BC survivors. Vitamin D status should be routinely evaluated for all women, especially those with a conservative dressing style, as part of regular preventive care, and they should take supplemental vitamin D.

Implementing a resource management program for accreditation process at the medical laboratory.

OBJECTIVES: To plan for and provide adequate resources to meet the mission and goals of a medical laboratory in compliance with the requirements for laboratory accreditation by Joint Commission International. DESIGN AND METHODS: The related policies and procedures were developed based on standard requirements for resource management. RESULTS: Competency assessment provided continuing education and performance feedback to laboratory employees. Laboratory areas were designed for the efficient and safe performance of laboratory work. A physical environment was built up where hazards were controlled and personnel activities were managed to reduce the risk of injuries. An Employees Occupational Safety and Health Program (EOSHP) was developed to address all types of hazardous materials and wastes. Guidelines were defined to verify that the methods would produce accurate and reliable results. CONCLUSIONS: An active resource management program will be an effective way of assuring that systems are in control and continuous improvement is in progress.

Detection of epithelial cell transfer in spinal areas by light microscopy and determining any tissue coring via cell culture during combined spinal-epidural interventions.

BACKGROUND AND OBJECTIVES: Epithelial tissue coring by spinal needles during subarachnoid injections may cause intraspinal epidermal tumors. Previous studies have investigated tissue transfer with different needle types during subarachnoid or epidural injection. This study deals with the transfer of epithelial tissue during combined spinal-epidural (CSE) anesthesia. METHODS: We studied 68 American Society of Anesthesiologists I to III adult patients. CSE anesthesia was induced under aseptic conditions at the L2-3 or L3-4 interspace with patients in the lateral decubitus position. Cerebral spinal fluid, spinal needle stylet, fluid used to flush the interior of the spinal needle, fluid used to wash the exterior of the spinal needle, fluid used to flush the interior of the epidural needle, and fluid used to wash the exterior tip of the epidural needle were examined under light microscopy (n = 30 patients) or incubated in a cell-culture medium (n = 38 patients). Samples were incubated in cell-culture medium alone (n = 13) or in a cell-culture medium for 3 weeks and then in a medium with epidermal growth factor (n = 25). As a positive control, skin tissue samples were taken by punch biopsy from 10 randomly chosen patients who underwent CSE interventions. These samples were incubated in an enriched medium serum. RESULTS: Light microscopy revealed that there was cell transfer in all phases in various rates: samples 1, 2, 3, 4, 5, and 6 contained epithelial cells and debris in ratios of 6.9%, 20.7%, 6.9%, 20.7%, 26.7%, and 33.3%, respectively. Epithelial cell colonization was detected in the cell-culture samples taken from the control group but not in the samples taken from the CSE group. CONCLUSIONS: We could not reproduce the cells or cell debris obtained during the CSE interventions in vivo, which can be explained by a possible structural deformation of cells or the inadequacy of the amount of cells that were transferred.