Psychopharmacologic Services for Homeless Veterans: Comparing Psychotropic Prescription Fills Among Homeless and Non-Homeless Veterans with Serious Mental Illness

Using national Veterans Health Administration (VHA) administrative data, this study evaluated differences in psychotropic medication use between homeless and non-homeless adults with serious mental illness (SMI) who used VHA services in 2010. The adjusted mean number of psychotropic prescription fills associated with homeless individuals were identified using regression models adjusted for socio-demographics, diagnoses, and use of health services. Of the 876,989 individuals with SMI using VHA services, 7.2 % were homeless at some time during 2010. In bivariate analysis, homeless individuals filled more psychotropic medication prescriptions compared with non-homeless individuals. However, after adjusting for potentially confounding variables, homeless individuals were found to have filled 16.2 % fewer prescriptions than non-homeless individuals when all psychotropics were analyzed together (F = 6947.1, p < .001) and for most individual classes of psychotropics. Greater use of residential/inpatient mental health services by the homeless was the most important single factor associated with filling more psychotropic prescriptions than non-homeless individuals.

A Comparison of Surgical Quality and Patient Satisfaction Indicators Between VA Hospitals and Hospitals Near VA Hospitals

BACKGROUND

The Maintaining Internal Systems and Strengthening Integrated Outside Networks (MISSION) Act established a community care program allowing veterans to receive care outside Veteran Affairs Medical Centers (VAMCs). We sought to compare patient safety and satisfaction indicators from VAMCs and surrounding non-VAMCs (non-VAs).

METHODS

We identified VAMCs with at least one non-VA acute care hospital within 25 miles in three geographic regions (West/Southwest, New England, and Deep South). Children's, specialty, and critical access hospitals were excluded. Using publicly available Hospital Compare data, we analyzed VAMC and surrounding non-VA performance in postsurgical patient safety indicator (PSI) events and Hospital Consumer Assessment of Healthcare Providers and Systems patient satisfaction scores and hospital star ratings.

RESULTS

The 34 VAMCs performed better than 319 surrounding non-VAs in rates of wound dehiscence, accidental lacerations, and perioperative hemorrhage/hematoma as well as composite PSI rating (P < 0.05). VAMCs performed significantly better than non-VAs (18.0 versus 51.4 events per 1000 patients, P < 0.001) in composite surgery-specific PSIs. When comparing mean linear Hospital Consumer Assessment of Healthcare Providers and Systems score star ratings (1-5 scale), VAMCs had similar performance in overall hospital rating compared with non-VAs (3.28 versus 3.38, P = 0.48) and summary rating of hospital stays (2.87 versus 2.92, P = 0.69). When compiled patient satisfaction star ratings were compared, there was no difference (2.96 versus 2.97, P = 0.9). VAMCs performed worse than non-VAs in "would recommend" ratings (2.7 versus 3.13, P = 0.007).

CONCLUSIONS

Across disparate regions, VAMCs match or outperform neighboring non-VAs in surgical quality metrics and patient satisfaction ratings. Veterans receiving surgical care at VAMCs may receive equivalent or better care than at non-VAs.

What is the role of selection bias in quality comparisons between the Veterans Health Administration and community care? Example of elective hernia surgery

OBJECTIVE

To investigate the relationship between community care (CC) treatment and a postoperative surgical complication in elective hernia surgery among Veterans using multiple approaches to control for potential selection bias.

DATA SOURCES AND STUDY SETTING

Veterans Health Administration (VHA) data sources included Corporate Data Warehouse (VHA encounters and patient data), the Program Integrity Tool and Fee tables (CC encounters), the Planning Systems Support Group (geographic information), and the Paid file (VHA primary care providers).

STUDY DESIGN

Prior works suggest patient outcomes are better in VHA than in CC settings; however, these studies may not have appropriately accounted for the selection of higher-risk cases into CC. We estimated (1) a naïve logistic regression model to calculate the effect of CC setting on the probability of a complication, controlling for facility fixed effects and patient and procedure characteristics, and (2) a 2-stage model using the hernia patient's primary care provider's 1-year prior CC referral rate as the instrument.

DATA COLLECTION

We identified patients residing ≤40 miles from a VHA surgical facility with elective VHA or CC hernia surgery from 2018 to 2019.

PRINCIPAL FINDINGS

Of 7991 hernia surgeries, 772 (9.7%) were in CC. The overall complication rate was 4.2%; 286/7219 (4.0%) among VHA surgeries versus 51/5772 (6.6%, p < 0.05) in CC. We observed a 2.8 percentage point increase in the probability of postoperative complication given CC surgery (95% confidence interval: 0.7, 4.8) in the naïve model. After accounting for the VHA provider's historical rate of CC referral, we no longer observed a relationship between surgery setting and risk of postoperative complication.

CONCLUSIONS

After accounting for the selection of higher-risk patients to CC settings, we found no difference in hernia surgery postoperative complications between CC and VHA. Future VHA and non-VHA comparisons should account for unobserved as well as observed differences in patients seen in each setting.

Trends of opioid use following anterior cervical discectomy and fusion: A 10-year longitudinal study of the Veterans Health Administration

Background

The United States Veteran Health Administration (VHA) cares for a substantial group of patients who are at higher risk of substance abuse in comparison to the general population. The purpose of this study was to (1) examine opioid consumption in the veteran population both pre- and postoperatively to anterior cervical discectomy and fusion (ACDF) and (2) understand the risk factors that are associated with sustained postoperative opioid use.

Methods

A retrospective database study was conducted using the Veterans Affairs Informatics and Computing Infrastructure database. Patients who underwent ACDF between 2010 and 2020 were identified and stratified into 3 groups based on their preoperative opioid usage prior to the procedure: opioid naïve, low preoperative opioid use (1-3 preoperative claims), and high preoperative opioid use (≥4 preoperative claims). Cumulative pre- and postoperative opioid usage for each patient was calculated in Morphine Milligram Equivalents (MME).

Results

A total of 7,894 patients were identified with 3,929 (49.7%) opioid naïve, 1,813 (23.0%) low preoperative opioid use, and 2,152 (27.3%) high opioid usage. The proportion of patients in the opioid-naïve cohort, low preoperative usage cohort, and high preoperative opioid usage cohort, that remained on opioids 1 year postoperatively was 13.1%, 31.3%, and 77.8%, respectively. At 1 year postoperatively, the median opioid MME significantly decreased pre- to postoperatively (25.0 vs. 0, p<.006). High preoperative opioid consumption was found to be the greatest risk factor for continued chronic opioid use (OR 17.1, p<.001).

Conclusions

Following ACDF procedures, opioid consumption significantly decreased; however, at 1 year, over one-third of patients remained on opioid therapy. A disproportionate number of patients who remained on chronic opioid therapy had high preoperative opioid consumption. Notably, increased scrutiny and policy changes regarding opioids, which began around 2017, resulted in a significant reduction in preoperative opioid use by 2020 compared to a decade earlier.