Evaluating accuracy and fairness of clinical decision support algorithms when health care resources are limited.

OBJECTIVE: Guidance on how to evaluate accuracy and algorithmic fairness across subgroups is missing for clinical models that flag patients for an intervention but when health care resources to administer that intervention are limited. We aimed to propose a framework of metrics that would fit this specific use case. METHODS: We evaluated the following metrics and applied them to a Veterans Health Administration clinical model that flags patients for intervention who are at risk of overdose or a suicidal event among outpatients who were prescribed opioids (N = 405,817): Receiver - Operating Characteristic and area under the curve, precision - recall curve, calibration - reliability curve, false positive rate, false negative rate, and false omission rate. In addition, we developed a new approach to visualize false positives and false negatives that we named 'per true positive bars.' We demonstrate the utility of these metrics to our use case for three cohorts of patients at the highest risk (top 0.5 %, 1.0 %, and 5.0 %) by evaluating algorithmic fairness across the following age groups: <=30, 31-50, 51-65, and >65 years old. RESULTS: Metrics that allowed us to assess group differences more clearly were the false positive rate, false negative rate, false omission rate, and the new 'per true positive bars'. Metrics with limited utility to our use case were the Receiver - Operating Characteristic and area under the curve, the calibration - reliability curve, and the precision - recall curve. CONCLUSION: There is no "one size fits all" approach to model performance monitoring and bias analysis. Our work informs future researchers and clinicians who seek to evaluate accuracy and fairness of predictive models that identify patients to intervene on in the context of limited health care resources. In terms of ease of interpretation and utility for our use case, the new 'per true positive bars' may be the most intuitive to a range of stakeholders and facilitates choosing a threshold that allows weighing false positives against false negatives, which is especially important when predicting severe adverse events.

Deploying a national clinical text processing infrastructure.

OBJECTIVES: Clinical text processing offers a promising avenue for improving multiple aspects of healthcare, though operational deployment remains a substantial challenge. This case report details the implementation of a national clinical text processing infrastructure within the Department of Veterans Affairs (VA). METHODS: Two foundational use cases, cancer case management and suicide and overdose prevention, illustrate how text processing can be practically implemented at scale for diverse clinical applications using shared services. RESULTS: Insights from these use cases underline both commonalities and differences, providing a replicable model for future text processing applications. CONCLUSIONS: This project enables more efficient initiation, testing, and future deployment of text processing models, streamlining the integration of these use cases into healthcare operations. This project implementation is in a large integrated health delivery system in the United States, but we expect the lessons learned to be relevant to any health system, including smaller local and regional health systems in the United States.

Development of a 3-Step theory of suicide ontology to facilitate 3ST factor extraction from clinical progress notes.

OBJECTIVE: Suicide risk prediction algorithms at the Veterans Health Administration (VHA) do not include predictors based on the 3-Step Theory of suicide (3ST), which builds on hopelessness, psychological pain, connectedness, and capacity for suicide. These four factors are not available from structured fields in VHA electronic health records, but they are found in unstructured clinical text. An ontology and controlled vocabulary that maps psychosocial and behavioral terms to these factors does not exist. The objectives of this study were 1) to develop an ontology with a controlled vocabulary of terms that map onto classes that represent the 3ST factors as identified within electronic clinical progress notes, and 2) to determine the accuracy of automated extractions based on terms in the controlled vocabulary. METHODS: A team of four annotators did linguistic annotation of 30,000 clinical progress notes from 231 Veterans in VHA electronic health records who attempted suicide or who died by suicide for terms relating to the 3ST factors. Annotation involved manually assigning a label to words or phrases that indicated presence or absence of the factor (polarity). These words and phrases were entered into a controlled vocabulary that was then used by our computational system to tag 14 million clinical progress notes from Veterans who attempted or died by suicide after 2013. Tagged text was extracted and machine-labelled for presence or absence of the 3ST factors. Accuracy of these machine-labels was determined for 1000 randomly selected extractions for each factor against a ground truth created by our annotators. RESULTS: Linguistic annotation identified 8486 terms that related to 33 subclasses across the four factors and polarities. Precision of machine-labeled extractions ranged from 0.73 to 1.00 for most factor-polarity combinations, whereas recall was somewhat lower 0.65-0.91. CONCLUSION: The ontology that was developed consists of classes that represent each of the four 3ST factors, subclasses, relationships, and terms that map onto those classes which are stored in a controlled vocabulary (https://bioportal.bioontology.org/ontologies/THREE-ST). The use case that we present shows how scores based on clinical notes tagged for terms in the controlled vocabulary capture meaningful change in the 3ST factors during weeks preceding a suicidal event.

Suicide theory-guided natural language processing of clinical progress notes to improve prediction of veteran suicide risk: protocol for a mixed-method study.

INTRODUCTION: The state-of-the-art 3-step Theory of Suicide (3ST) describes why people consider suicide and who will act on their suicidal thoughts and attempt suicide. The central concepts of 3ST-psychological pain, hopelessness, connectedness, and capacity for suicide-are among the most important drivers of suicidal behaviour but they are missing from clinical suicide risk prediction models in use at the US Veterans Health Administration (VHA). These four concepts are not systematically recorded in structured fields of VHA's electronic healthcare records. Therefore, this study will develop a domain-specific ontology that will enable automated extraction of these concepts from clinical progress notes using natural language processing (NLP), and test whether NLP-based predictors for these concepts improve accuracy of existing VHA suicide risk prediction models. METHODS AND ANALYSIS: Our mixed-method study has an exploratory sequential design where a qualitative component (aim 1) will inform quantitative analyses (aims 2 and 3). For aim 1, subject matter experts will manually annotate progress notes of clinical encounters with veterans who attempted or died by suicide to develop a domain-specific ontology for the 3ST concepts. During aim 2, we will use NLP to machine-annotate clinical progress notes and derive longitudinal representations for each patient with respect to the presence and intensity of hopelessness, psychological pain, connectedness and capacity for suicide in temporal proximity of suicide attempts and deaths by suicide. These longitudinal representations will be evaluated during aim 3 for their ability to improve existing VHA prediction models of suicide and suicide attempts, STORM (Stratification Tool for Opioid Risk Mitigation) and REACHVET (Recovery Engagement and Coordination for Health - Veterans Enhanced Treatment). ETHICS AND DISSEMINATION: Ethics approval for this study was granted by the Stanford University Institutional Review Board and the Research and Development Committee of the VA Palo Alto Health Care System. Results of the study will be disseminated through several outlets, including peer-reviewed publications and presentations at national conferences.

Using natural language processing to construct a metastatic breast cancer cohort from linked cancer registry and electronic medical records data.

OBJECTIVES: Most population-based cancer databases lack information on metastatic recurrence. Electronic medical records (EMR) and cancer registries contain complementary information on cancer diagnosis, treatment and outcome, yet are rarely used synergistically. To construct a cohort of metastatic breast cancer (MBC) patients, we applied natural language processing techniques within a semisupervised machine learning framework to linked EMR-California Cancer Registry (CCR) data. MATERIALS AND METHODS: We studied all female patients treated at Stanford Health Care with an incident breast cancer diagnosis from 2000 to 2014. Our database consisted of structured fields and unstructured free-text clinical notes from EMR, linked to CCR, a component of the Surveillance, Epidemiology and End Results Program (SEER). We identified de novo MBC patients from CCR and extracted information on distant recurrences from patient notes in EMR. Furthermore, we trained a regularized logistic regression model for recurrent MBC classification and evaluated its performance on a gold standard set of 146 patients. RESULTS: There were 11 459 breast cancer patients in total and the median follow-up time was 96.3 months. We identified 1886 MBC patients, 512 (27.1%) of whom were de novo MBC patients and 1374 (72.9%) were recurrent MBC patients. Our final MBC classifier achieved an area under the receiver operating characteristic curve (AUC) of 0.917, with sensitivity 0.861, specificity 0.878, and accuracy 0.870. DISCUSSION AND CONCLUSION: To enable population-based research on MBC, we developed a framework for retrospective case detection combining EMR and CCR data. Our classifier achieved good AUC, sensitivity, and specificity without expert-labeled examples.

Enhanced Quality Measurement Event Detection: An Application to Physician Reporting.

The wide-scale adoption of electronic health records (EHR)s has increased the availability of routinely collected clinical data in electronic form that can be used to improve the reporting of quality of care. However, the bulk of information in the EHR is in unstructured form (e.g., free-text clinical notes) and not amenable to automated reporting. Traditional methods are based on structured diagnostic and billing data that provide efficient, but inaccurate or incomplete summaries of actual or relevant care processes and patient outcomes. To assess the feasibility and benefit of implementing enhanced EHR- based physician quality measurement and reporting, which includes the analysis of unstructured free- text clinical notes, we conducted a retrospective study to compare traditional and enhanced approaches for reporting ten physician quality measures from multiple National Quality Strategy domains. We found that our enhanced approach enabled the calculation of five Physician Quality and Performance System measures not measureable in billing or diagnostic codes and resulted in over a five-fold increase in event at an average precision of 88 percent (95 percent CI: 83-93 percent). Our work suggests that enhanced EHR-based quality measurement can increase event detection for establishing value-based payment arrangements and can expedite quality reporting for physician practices, which are increasingly burdened by the process of manual chart review for quality reporting.