Machine learning models for predicting one-year survival in patients with metastatic gastric cancer who experienced upfront radical gastrectomy.

Tumor metastasis is a common event in patients with gastric cancer (GC) who previously underwent curative gastrectomy. It is meaningful to employ high-volume clinical data for predicting the survival of metastatic GC patients. We aim to establish an improved machine learning (ML) classifier for predicting if a patient with metastatic GC would die within 12 months. Eligible patients were enrolled from a Chinese GC cohort, and the complete detailed information from medical records was extracted to generate a high-dimensional dataset. Appropriate feature engineering and feature filter were conducted before modeling with eight algorithms. A 10-fold cross validation (CV) nested in a holdout CV (8:2) was employed for hyperparameter tuning and model evaluation. Model selection was based on the area under the receiver operating characteristic (AUROC) curve, recall, and precision. The selected model was globally explained using interpretable surrogate models. Of the total 399 cases (median survival of 8.2 months), 242 patients survived less than 12 months. The linear discriminant analysis (LDA), support vector machine (SVM), and random forest (RF) model had the highest AUROC (0.78 ± 0.021), recall (0.93 ± 0.031), and precision (0.80 ± 0.026), respectively. The LDA model created a new function that generally separated the two classes. The predicted probability of the SVM model was interpreted using a linear regression model visualized by a nomogram. The predicted class of the RF model was explained using a decision tree model. In summary, analyzing high-volume medical data by ML is helpful to produce an improved model for predicting the survival in patients with metastatic GC. The algorithm should be carefully selected in different practical scenarios.

Use of Middleware Data to Dissect and Optimize Hematology Autoverification.

BACKGROUND: Hematology analysis comprises some of the highest volume tests run in clinical laboratories. Autoverification of hematology results using computer-based rules reduces turnaround time for many specimens, while strategically targeting specimen review by technologist or pathologist. METHODS: Autoverification rules had been developed over a decade at an 800-bed tertiary/quarternary care academic medical central laboratory serving both adult and pediatric populations. In the process of migrating to newer hematology instruments, we analyzed the rates of the autoverification rules/flags most commonly associated with triggering manual review. We were particularly interested in rules that on their own often led to manual review in the absence of other flags. Prior to the study, autoverification rates were 87.8% (out of 16,073 orders) for complete blood count (CBC) if ordered as a panel and 85.8% (out of 1,940 orders) for CBC components ordered individually (not as the panel). RESULTS: Detailed analysis of rules/flags that frequently triggered indicated that the immature granulocyte (IG) flag (an instrument parameter) and rules that reflexed platelet by impedance method (PLT-I) to platelet by fluorescent method (PLT-F) represented the two biggest opportunities to increase autoverification. The IG flag threshold had previously been validated at 2%, a setting that resulted in this flag alone preventing autoverification in 6.0% of all samples. The IG flag threshold was raised to 5% after detailed chart review; this was also the instrument vendor's default recommendation for the newer hematology analyzers. Analysis also supported switching to PLT-F for all platelet analysis. Autoverification rates increased to 93.5% (out of 91,692 orders) for CBC as a panel and 89.8% (out of 11,982 orders) for individual components after changes in rules and laboratory practice. CONCLUSIONS: Detailed analysis of autoverification of hematology testing at an academic medical center clinical laboratory that had been using a set of autoverification rules for over a decade revealed opportunities to optimize the parameters. The data analysis was challenging and time-consuming, highlighting opportunities for improvement in software tools that allow for more rapid and routine evaluation of autoverification parameters.

Impact of Daylight Saving Time on the Clinical Laboratory.

Daylight saving time is a practice in some countries and local regions to set clocks forward (typically 1 hour) during the longer days of summer and back again in autumn. Time changes resulting from daylight saving time have the potential to impact clinical laboratory instruments, computer interfaces, and information systems. We analyzed turnaround time data for an academic medical center clinical laboratories (chemistry, hematology, blood gas analyzer, and transfusion medicine), examining how turnaround time was impacted by the daylight saving time shifts in 2017. We also determined whether the daylight saving time shift on November 5, 2017 ("fall back" by 1 hour) resulted in any "absurd" time combinations such as a receipt time occurring "before" a normally later time such as final result. We also describe challenges resulting from daylight saving time changes over a 5-year period. The only significant impact on turnaround time was for clinical chemistry samples during the autumn daylight saving time change, but the overall impact was low. Four instances of absurd time combinations occurred in the autumn time change with only a transfusion medicine example resulting in an interface error (a Type and Screen resulted "before" receipt in laboratory). Over a 5-year period, other daylight saving time impacts included problems of reestablishing interface to instruments, inadvertent discrepancies in manual time changes at different points of the core laboratory automation line, and time change errors in instruments with older operating systems lacking patches that updated daylight saving time rules after 2007. Clinical laboratories should be aware that rare problems may occur due to issues with daylight saving time changes.

Verification study on the NORIP LDH reference intervals with a proposed new upper reference limit.

We verified the lactate dehydrogenase (LDH) reference interval (RI) provided by the Nordic Reference Interval Project (NORIP). The serum LDH concentration was analysed on the Dimension Vista 1500 system with an IFCC method with a bias of +2.1 % and +2.7 % against NFKK Reference Serum X and ERM-AD453/IFCC, respectively, showing verification of transference of the NORIP RI. Selective data mining in clinical laboratory information systems for retrospective serum LDH test results was used to calculate an indirect RI. For the adult age group (18 to <70 years) the limits of the interval was 127 U/L (90 % CI: 123-132 U/L) and 240 U/L (90 % CI: 234-243 U/L). However, the NORIP upper limit for the adult age group is 205 U/L (90 % CI: 198-210 U/L). Accordingly, 25.1 % of LDH test results were above the NORIPs upper limit of 205 U/L. If LDH analysis was requested by the hospital's medical departments, outpatient clinics or general practitioners 29.2 %, 26.2 % and 20.9 %, respectively, were above the 205 U/L limit. Differences in transport time before centrifugation of blood, and different transport principles could not explain the relative high percent of test results above the NORIP 205 U/L limit. The indirect finding of an upper limit of 240 U/L (90 % CI: 234-243 U/L), and the relative high number of test result >205 U/L, suggests that the NORIP upper limit should be adjusted.

Variation in Results Release and Patient Portal Access to Diagnostic Test Results at an Academic Medical Center.

BACKGROUND: Electronic health records (EHRs) are commonplace in industrialized countries. Many hospitals are granting their patients access to their medical information through online patient portals. In this report, we describe a retrospective analysis of patient access to diagnostic test results released through the patient portal (MyChart; Epic, Inc.) at a state academic medical center. METHODS: We analyzed 6 months of data for anatomic pathology, clinical laboratory, and radiology test results to evaluate variations in results release (automated vs. manual) and subsequent patient access to the institutional patient portal. During this period, diagnostic test results were released for all patient encounters including inpatient units, outpatient clinics, and the emergency department. RESULTS: Manual results release by providers before automated release time occurred most commonly in the outpatient setting. The highest rates of access of diagnostic test results occurred for outpatients (about 30% overall view rate), females (two times or more compared to males in nearly every age bracket), and 20-45-year-old. Access rates of diagnostic tests in the emergency department or inpatient units were <10% across all populations. Access of diagnostic test results was very low for 12-17-year-old, likely influenced by institutional policies limiting parental proxy access within this pediatric age range. Approximately 20% of outpatient laboratory results were viewed by patients within 8 h of release from the EHR to the patient portal and 10% within 2 h of release. CONCLUSIONS: Patient accessing of diagnostic test results were generally higher for females, outpatients, and 20-45-year-old. Approximately, 20% of outpatient results were viewed quickly by patients after release to the EHR.

Preferred Names, Preferred Pronouns, and Gender Identity in the Electronic Medical Record and Laboratory Information System: Is Pathology Ready?

BACKGROUND: Electronic medical records (EMRs) and laboratory information systems (LISs) commonly utilize patient identifiers such as legal name, sex, medical record number, and date of birth. There have been recommendations from some EMR working groups (e.g., the World Professional Association for Transgender Health) to include preferred name, pronoun preference, assigned sex at birth, and gender identity in the EMR. These practices are currently uncommon in the United States. There has been little published on the potential impact of these changes on pathology and LISs. METHODS: We review the available literature and guidelines on the use of preferred name and gender identity on pathology, including data on changes in laboratory testing following gender transition treatments. We also describe pathology and clinical laboratory challenges in the implementation of preferred name at our institution. RESULTS: Preferred name, pronoun preference, and gender identity have the most immediate impact on the areas of pathology with direct patient contact such as phlebotomy and transfusion medicine, both in terms of interaction with patients and policies for patient identification. Gender identity affects the regulation and policies within transfusion medicine including blood donor risk assessment and eligibility. There are limited studies on the impact of gender transition treatments on laboratory tests, but multiple studies have demonstrated complex changes in chemistry and hematology tests. A broader challenge is that, even as EMRs add functionality, pathology computer systems (e.g., LIS, middleware, reference laboratory, and outreach interfaces) may not have functionality to store or display preferred name and gender identity. CONCLUSIONS: Implementation of preferred name, pronoun preference, and gender identity presents multiple challenges and opportunities for pathology.

Implementation of Epic Beaker Clinical Pathology at an academic medical center.

BACKGROUND: Epic Beaker Clinical Pathology (CP) is a relatively new laboratory information system (LIS) operating within the Epic suite of software applications. To date, there have not been any publications describing implementation of Beaker CP. In this report, we describe our experience in implementing Beaker CP version 2012 at a state academic medical center with a go-live of August 2014 and a subsequent upgrade to Beaker version 2014 in May 2015. The implementation of Beaker CP was concurrent with implementations of Epic modules for revenue cycle, patient scheduling, and patient registration. METHODS: Our analysis covers approximately 3 years of time (2 years preimplementation of Beaker CP and roughly 1 year after) using data summarized from pre- and post-implementation meetings, debriefings, and the closure document for the project. RESULTS: We summarize positive aspects of, and key factors leading to, a successful implementation of Beaker CP. The early inclusion of subject matter experts in the design and validation of Beaker workflows was very helpful. Since Beaker CP does not directly interface with laboratory instrumentation, the clinical laboratories spent extensive preimplementation effort establishing middleware interfaces. Immediate challenges postimplementation included bar code scanning and nursing adaptation to Beaker CP specimen collection. The most substantial changes in laboratory workflow occurred with microbiology orders. This posed a considerable challenge with microbiology orders from the operating rooms and required intensive interventions in the weeks following go-live. In postimplementation surveys, pathology staff, informatics staff, and end-users expressed satisfaction with the new LIS. CONCLUSIONS: Beaker CP can serve as an effective LIS for an academic medical center. Careful planning and preparation aid the transition to this LIS.

Use of a data warehouse at an academic medical center for clinical pathology quality improvement, education, and research.

BACKGROUND: Pathology data contained within the electronic health record (EHR), and laboratory information system (LIS) of hospitals represents a potentially powerful resource to improve clinical care. However, existing reporting tools within commercial EHR and LIS software may not be able to efficiently and rapidly mine data for quality improvement and research applications. MATERIALS AND METHODS: We present experience using a data warehouse produced collaboratively between an academic medical center and a private company. The data warehouse contains data from the EHR, LIS, admission/discharge/transfer system, and billing records and can be accessed using a self-service data access tool known as Starmaker. The Starmaker software allows users to use complex Boolean logic, include and exclude rules, unit conversion and reference scaling, and value aggregation using a straightforward visual interface. More complex queries can be achieved by users with experience with Structured Query Language. Queries can use biomedical ontologies such as Logical Observation Identifiers Names and Codes and Systematized Nomenclature of Medicine. RESULT: We present examples of successful searches using Starmaker, falling mostly in the realm of microbiology and clinical chemistry/toxicology. The searches were ones that were either very difficult or basically infeasible using reporting tools within the EHR and LIS used in the medical center. One of the main strengths of Starmaker searches is rapid results, with typical searches covering 5 years taking only 1-2 min. A "Run Count" feature quickly outputs the number of cases meeting criteria, allowing for refinement of searches before downloading patient-identifiable data. The Starmaker tool is available to pathology residents and fellows, with some using this tool for quality improvement and scholarly projects. CONCLUSION: A data warehouse has significant potential for improving utilization of clinical pathology testing. Software that can access data warehouse using a straightforward visual interface can be incorporated into pathology training programs.

Autoverification in a core clinical chemistry laboratory at an academic medical center.

BACKGROUND: Autoverification is a process of using computer-based rules to verify clinical laboratory test results without manual intervention. To date, there is little published data on the use of autoverification over the course of years in a clinical laboratory. We describe the evolution and application of autoverification in an academic medical center clinical chemistry core laboratory. SUBJECTS AND METHODS: At the institution of the study, autoverification developed from rudimentary rules in the laboratory information system (LIS) to extensive and sophisticated rules mostly in middleware software. Rules incorporated decisions based on instrument error flags, interference indices, analytical measurement ranges (AMRs), delta checks, dilution protocols, results suggestive of compromised or contaminated specimens, and 'absurd' (physiologically improbable) values. RESULTS: The autoverification rate for tests performed in the core clinical chemistry laboratory has increased over the course of 13 years from 40% to the current overall rate of 99.5%. A high percentage of critical values now autoverify. The highest rates of autoverification occurred with the most frequently ordered tests such as the basic metabolic panel (sodium, potassium, chloride, carbon dioxide, creatinine, blood urea nitrogen, calcium, glucose; 99.6%), albumin (99.8%), and alanine aminotransferase (99.7%). The lowest rates of autoverification occurred with some therapeutic drug levels (gentamicin, lithium, and methotrexate) and with serum free light chains (kappa/lambda), mostly due to need for offline dilution and manual filing of results. Rules also caught very rare occurrences such as plasma albumin exceeding total protein (usually indicative of an error such as short sample or bubble that evaded detection) and marked discrepancy between total bilirubin and the spectrophotometric icteric index (usually due to interference of the bilirubin assay by immunoglobulin (Ig) M monoclonal gammopathy). CONCLUSIONS: Our results suggest that a high rate of autoverification is possible with modern clinical chemistry analyzers. The ability to autoverify a high percentage of results increases productivity and allows clinical laboratory staff to focus attention on the small number of specimens and results that require manual review and investigation.