Using an Automated Electronic Health Record Score To Estimate Life Expectancy In Men Diagnosed With Prostate Cancer In The Veterans Health Administration

Disparities With Systemic Therapies for Black Men Having Advanced Prostate Cancer: Where Do We Stand?

PURPOSE

Prostate cancer represents the most common cancer diagnosis in Black men and is the second leading cause of cancer death in this population. Multilevel disparities have been well-documented in Black men with prostate cancer and play a role in poorer survival outcomes when compared with White men with prostate cancer. In this review, we highlight the changing trend in disparities for systemic therapy outcomes in Black men diagnosed with metastatic prostate cancer.

METHODS

We reviewed data from real-world registries and prospective clinical trials with a particular focus on equal access settings to compare outcomes to systemic therapies between Black and White men with metastatic prostate cancer.

RESULTS

In metastatic prostate cancer, there is growing evidence to suggest that Black men may have similar, if not better, outcomes to systemic therapies than White men with advanced disease, as corroborated by prospective studies and clinical trials where health care delivery and follow-up are more likely to be standardized.

CONCLUSION

This review illustrates the importance of nonbiological drivers of racial disparities in Black men with advanced prostate cancer. Mitigating barriers to health care access and delivery as well as including participation in clinical trials will be pivotal to ongoing efforts to address disparities in systemic therapy outcomes for Black men with metastatic prostate cancer.

Adaption and National Validation of a Tool for Predicting Mortality from Other Causes Among Men with Nonmetastatic Prostate Cancer

BACKGROUND

An electronic health record-based tool could improve accuracy and eliminate bias in provider estimation of the risk of death from other causes among men with nonmetastatic cancer.

OBJECTIVE

To recalibrate and validate the Veterans Aging Cohort Study Charlson Comorbidity Index (VACS-CCI) to predict non-prostate cancer mortality (non-PCM) and to compare it with a tool predicting prostate cancer mortality (PCM).

DESIGN, SETTING, AND PARTICIPANTS

An observational cohort of men with biopsy-confirmed nonmetastatic prostate cancer, enrolled from 2001 to 2018 in the national US Veterans Health Administration (VA), was divided by the year of diagnosis into the development (2001-2006 and 2008-2018) and validation (2007) sets.

OUTCOME MEASUREMENTS AND STATISTICAL ANALYSIS

Mortality (all cause, non-PCM, and PCM) was evaluated. Accuracy was assessed using calibration curves and C statistic in the development, validation, and combined sets; overall; and by age (<65 and 65+ yr), race (White and Black), Hispanic ethnicity, and treatment groups.

RESULTS AND LIMITATIONS

Among 107 370 individuals, we observed 24 977 deaths (86% non-PCM). The median age was 65 yr, 4947 were Black, and 5010 were Hispanic. Compared with CCI and age alone (C statistic 0.67, 95% confidence interval [CI] 0.67-0.68), VACS-CCI demonstrated improved validated discrimination (C statistic 0.75, 95% CI 0.74-0.75 for non-PCM). The prostate cancer mortality tool also discriminated well in validation (C statistic 0.81, 95% CI 0.78-0.83). Both were well calibrated overall and within subgroups. Owing to missing data, 18 009/125 379 (14%) were excluded, and VACS-CCI should be validated outside the VA prior to outside application.

CONCLUSIONS

VACS-CCI is ready for implementation within the VA. Electronic health record-assisted calculation is feasible, improves accuracy over age and CCI alone, and could mitigate inaccuracy and bias in provider estimation.

PATIENT SUMMARY

Veterans Aging Cohort Study Charlson Comorbidity Index is ready for application within the Veterans Health Administration. Electronic health record-assisted calculation is feasible, improves accuracy over age and Charlson Comorbidity Index alone, and might help mitigate inaccuracy and bias in provider estimation of the risk of non-prostate cancer mortality.

Estimating remaining life expectancy in veterans with basal cell carcinoma using an automated electronic health record scoring system: A retrospective cohort study

BACKGROUND

Active surveillance may be considered for low-risk basal cell carcinomas (BCCs) in patients with limited life expectancy; however, estimates of life expectancy are not readily available. Veterans Health Administration's Care Assessment Need (CAN) score may address this problem.

OBJECTIVE

We examined the CAN score's performance in predicting 1-, 3-, and 5-year mortality in US veterans with BCC.

METHODS

This retrospective cohort study used national Veterans Health Administration's electronic medical record data. The CAN score's performance in the prediction of mortality in veterans with BCC was evaluated based on tests of goodness-of-fit, discrimination, and calibration.

RESULTS

For 54,744 veterans with BCC treatment encounters between 2013 and 2018, the CAN score performed well in the prediction of mortality based on multiple tests. A threshold CAN score of 90 had a positive predictive value of 55% for 3-year mortality, clinically useful in identifying patients with intermediate-term survival.

LIMITATIONS

The study relied upon the combination of diagnosis codes and procedure codes to identify BCC cases.

CONCLUSION

The CAN score has the potential to improve the quality of cancer care for veterans by providing clinicians with an estimate of life expectancy and facilitating conversations in cases where active surveillance can be considered.

Unhealthy alcohol use and brief intervention rates among high and low complexity veterans seeking primary care services in the Veterans Health Administration

INTRODUCTION

Brief intervention (BI) is recommended for all primary care (PC) patients who screen positive for unhealthy alcohol use; however, patients with multiple chronic health conditions who are at high-risk of hospitalization (i.e., "high complexity" patients) may face disparities in receiving BIs in PC. The current study investigated whether high complexity and low complexity patients in the Veterans Health Administration (VHA) differed regarding screening positive for unhealthy alcohol use, alcohol-use severity, and receipt of BI for those with unhealthy alcohol use.

METHODS

Patients were veterans receiving PC services at the VHA in a mid-Atlantic region of the United States. The study extracted VHA administrative and clinical data for a total of 282,242 patients who had ≥1 PC visits between 1/1/2014 and 12/31/2014, during which they were screened for unhealthy alcohol use by the Alcohol Use Disorders Identification Test-Consumption (AUDIT-C). The study defined high complexity patients as those within and above the 90th percentile of risk for hospitalization per the VHA's Care Assessment Need Score. Logistic regression models assessed if being a high complexity patient was associated with screening positive for unhealthy alcohol use (AUDIT-C ≥ 5), severity of unhealthy alcohol use in those who screened positive (AUDIT-C score range 5-12), and receipt of BI in those who screened positive.

RESULTS

Our sample was 94.5% male, 83% White, 13% Black, 4% other race, and 1.7% Hispanic. A total of 10,813 (3.8%) patients screened positive for unhealthy alcohol use from which we identified 569 (5.3%) high complexity and 10,128 (93.6%) low complexity patients (n = 116 removed due to missing complexity data). Relative to low complexity patients, high complexity patients were less likely to screen positive for unhealthy alcohol use (3.3% vs. 4.1%, AOR = 0.59, p < .001); however, in patients who screened positive, high complexity patients had higher AUDIT-C scores (Mean AUDIT-C = 7.75 vs. 6.87, AOR = 1.46, p < .001) and were less likely to receive a BI (78.0% vs. 92.6%, AOR = 0.42, p < .001).

CONCLUSIONS

Disparities in BI exist for highly complex patients despite having more severe unhealthy alcohol use. Future research should examine the specific patient- and/or clinic-level factors impeding BI delivery for complex patients.

Developing and Implementing Predictive Models in a Learning Healthcare System: Traditional and Artificial Intelligence Approaches in the Veterans Health Administration

Predicting clinical risk is an important part of healthcare and can inform decisions about treatments, preventive interventions, and provision of extra services. The field of predictive models has been revolutionized over the past two decades by electronic health record data; the ability to link such data with other demographic, socioeconomic, and geographic information; the availability of high-capacity computing; and new machine learning and artificial intelligence methods for extracting insights from complex datasets. These advances have produced a new generation of computerized predictive models, but debate continues about their development, reporting, validation, evaluation, and implementation. In this review we reflect on more than 10 years of experience at the Veterans Health Administration, the largest integrated healthcare system in the United States, in developing, testing, and implementing such models at scale. We report lessons from the implementation of national risk prediction models and suggest an agenda for research. Expected final online publication date for the Annual Review of Biomedical Data Science, Volume 5 is August 2022. Please see http://www.annualreviews.org/page/journal/pubdates for revised estimates.