Publications

BACKGROUND: The development of delirium is very common in terminally ill patients. However, risk factors for terminal delirium in the veteran population are poorly identified. The purpose of this study was to (1) Identify risk factors for terminal delirium in a US Department of Veterans Affairs inpatient hospice population; (2) Assess usage patterns of antipsychotics for treatment of terminal delirium; and (3) Describe nursing assessment, nonpharmacologic interventions, and documentation of terminal delirium.

METHODS: This was a retrospective case-control study of veterans who expired while admitted into hospice care at a long-term care hospice unit during the period of October 1, 2013 to September 30, 2015. Veterans' medical records were reviewed for the 2 weeks prior to the recorded death.

RESULTS: Of 307 veterans admitted for hospice care, 67.4% required antipsychotics in the last 2 weeks of life for the treatment of terminal delirium. The average number of antipsychotic doses given was 14.9 doses per patient. The risk factors that were identified included the use of steroids, opioids, or anticholinergics; Vietnam-era veterans with liver disease; veterans with cancer and a comorbid mental health disorder; and veterans with a history of drug and/or alcohol abuse.

CONCLUSIONS: More than half of veterans admitted for hospice care experienced terminal delirium requiring treatment with antipsychotics. The identification of veterans most likely to develop terminal delirium will allow for early nonpharmacologic interventions and potentially decrease the need for treatment with antipsychotic medications.

INTRODUCTION: The objective of this study is to identify county-level characteristics that may be high-impact targets for opioid and antibiotic interventions to improve dental prescribing.

METHODS: Prescriptions during 2012-2017 were extracted from the IQVIA Longitudinal Prescription database. Primary outcomes were yearly county-level antibiotic and opioid prescribing rates. Multivariable negative binomial regression identified associations between prescribing rates and county-level characteristics. All analyses occurred in 2020.

RESULTS: Over time, dental opioid prescribing rates decreased by 20% (from 4.02 to 3.22 per 100 people), whereas antibiotic rates increased by 5% (from 6.85 to 7.19 per 100 people). Higher number of dentists per capita, higher proportion of female residents, and higher proportion of residents aged <65 years were associated with increased opioid rates. Relative to location in the West, location in the Northeast (59%, 95% CI=52, 65) and Midwest (64%, 95% CI=60, 70) was associated with lower opioid prescribing rates. Higher clinician density, median household income, proportion female, and proportion White were all independently associated with higher antibiotic rates. Location in the Northeast (149%, 95% CI=137, 162) and Midwest (118%, 95% CI=111, 125) was associated with higher antibiotic rates. Opioid and antibiotic prescribing rates were positively associated.

CONCLUSIONS: Dental prescribing of opioids is decreasing, whereas dental antibiotic prescribing is increasing. High prescribing of antibiotics is associated with high prescribing of opioids. Strategies focused on optimizing dental antibiotics and opioids are needed given their impact on population health.

AIMS: Registries and health plans estimate insulin need for population health metrics. We sought to identify how such estimates affect population- and individual-level estimates of over- and under-treatment.

METHODS: We developed a microsimulation comparing estimated insulin need to dispensation using the National Health and Nutrition Examination Survey (NHANES, 2005-2016, N= 2832) and Medical Expenditure Panel Survey (MEPS, 2005-2016, N = 29,615).

RESULTS: From NHANES,  21.6% of people with type 2 diabetes would require insulin to achieve a HbA1c target of 7% after maximum titration of two non-insulins (60.7 IU/person/day, or 84,629,833 vials of 1000 IU in the US). From MEPS, we observed 57.4 IU/person/day of insulin dispensed (81,585,842 vials). About 29% of people were dispensed at least two standard deviations less than their estimated need, and 22% at least two standard deviations more than estimated need. Population-level need estimates reduced 39.4% if liberalizing HbA1c targets to 8% for people ≥75 years old.

CONCLUSIONS: Estimated insulin needs of people with type 2 diabetes in the U.S. are consistent with their dispensed insulin at the population level, but are sensitive to HbA1c targets for older adults, and conceal under- and over-treated subpopulations.

INTRODUCTION: Benzodiazepines contribute to substance use disorder and are often part of polydrug abuse, most frequently with opioids. Although dental opioid prescribing differs significantly between countries, little is known about the patterns of dental benzodiazepine prescribing. The aim of this study is to compare dental prescribing of benzodiazepines among the U.S., England, and Australia in 2013-2018.

METHODS: Population-level data were accessed from national data sets for each country for dental benzodiazepine prescriptions. Outcome measures of dental benzodiazepine prescribing included: (1) prescribing rates by population for each year and (2) the quantity and relative proportion of benzodiazepines by type for each country. The analysis was conducted in 2020.

RESULTS: Between 2013 and 2018, U.S. dentists prescribed 23 times more than English dentists and 7 times more than Australian dentists by population. During the study period, the rate of dental benzodiazepine prescribing decreased in England and the U.S. but increased in Australia. Despite these trends, U.S. dental prescribing rates remained 28 times more than English dentists and 6 times more than Australian dentists in 2018 (U.S., 3.10 prescriptions/1,000 population; England, 0.11 prescriptions/1,000 population; Australia, 0.50 prescriptions/1,000 population). U.S. dentists prescribed a wider variety of benzodiazepines than English and Australian dentists. Diazepam was most commonly prescribed in all countries. In the U.S., triazolam, lorazepam, and alprazolam were next most commonly prescribed. Temazepam was next most frequent in England and Australia.

CONCLUSIONS: Significant variation in benzodiazepine prescribing rates and types were seen among the countries. To improve patient safety, further investigation into the appropriate use and choices of benzodiazepines in dentistry is needed.

This is an epidemiological study of carbapenem-resistant Enterobacteriaceae (CRE) in Veterans' Affairs medical centers (VAMCs). In 2017, almost 75% of VAMCs had at least 1 CRE case. We observed substantial geographic variability, with more cases in urban, complex facilities. This supports the benefit of tailoring infection control strategies to facility characteristics.

BACKGROUND: Infections caused by carbapenem-resistant Enterobacteriaceae (CRE) and carbapenemase-producing (CP) CRE are difficult to treat, resulting in high mortality in healthcare settings every year. The Veterans Health Administration (VHA) disseminated guidelines in 2015 and an updated directive in 2017 for control of CRE focused on laboratory testing, prevention, and management. The Consolidated Framework for Implementation Research (CFIR) framework was used to analyze qualitative interview data to identify contextual factors and best practices influencing implementation of the 2015 guidelines/2017 directive in VA Medical Centers (VAMCs). The overall goals were to determine CFIR constructs to target to improve CRE guideline/directive implementation and understand how CFIR, as a multi-level conceptual model, can be used to inform guideline implementation.

METHODS: Semi-structured interviews were conducted at 29 VAMCs with staff involved in implementing CRE guidelines at their facility. Survey and VHA administrative data were used to identify geographically representative large and small VAMCs with varying levels of CRE incidence. Interviews addressed perceptions of guideline dissemination, laboratory testing, staff attitudes and training, patient education, and technology support. Participant responses were coded using a consensus-based mixed deductive-inductive approach guided by CFIR. A quantitative analysis comparing qualitative CFIR constructs and emergent codes to sites actively screening for CRE (vs. non-screening) and any (vs. no) CRE-positive cultures was conducted using Fisher's exact test.

RESULTS: Forty-three semi-structured interviews were conducted between October 2017 and August 2018 with laboratory staff (47%), Multi-Drug-Resistant Organism Program Coordinators (MPCs, 35%), infection preventionists (12%), and physicians (6%). Participants requested more standardized tools to promote effective communication (e.g., electronic screening). Participants also indicated that CRE-specific educational materials were needed for staff, patient, and family members. Quantitative analysis identified CRE screening or presence of CRE as being significantly associated with the following qualitative CFIR constructs: leadership engagement, relative priority, available resources, team communication, and access to knowledge and information.

CONCLUSIONS: Effective CRE identification, prevention, and treatment require ongoing collaboration between clinical, microbiology, infection prevention, antimicrobial stewardship, and infectious diseases specialists. Our results emphasize the importance of leadership's role in promoting positive facility culture, including access to resources, improving communication, and facilitating successful implementation of the CRE guidelines.

BACKGROUND: Proton pump inhibitors (PPIs) are among the most prescribed medications and are often used unnecessarily. PPIs are used for the treatment of heartburn and acid-related disorders. Emerging evidence indicates that PPIs are associated with serious adverse events, such as increased risk of Clostridioides difficile infection. In this study, we designed and piloted a PPI de-implementation intervention among hospitalized non-intensive care unit patients.

METHODS: Using the Systems Engineering Initiative for Patient Safety (SEIPS) model as the framework, we developed an intervention with input from providers and patients. On a bi-weekly basis, a trainee pharmacist reviewed a random sample of eligible patients' charts to assess if PPI prescriptions were guideline-concordant; a recommendation to de-implement non-guideline-concordant PPI therapy was sent when applicable. We used convergent parallel mixed-methods design to evaluate the feasibility and outcomes of the intervention.

RESULTS: During the study period (September 2019 to August 2020), 2171 patients with an active PPI prescription were admitted. We randomly selected 155 patient charts for review. The mean age of patients was 70.9 ± 9 years, 97.4% were male, and 35% were on PPIs for ≥5 years. The average time (minutes) needed to complete the intervention was as follows: 5 to assess if the PPI was guideline-concordant, 5 to provide patient education, and 7 to follow-up with patients post-discharge. After intervention initiation, the week-to-week mean number of PPI prescriptions decreased by 0.5 (S<0.0001). Barriers and facilitators spanned the 5 elements of the SEIPS model and included factors such as providers' perception that PPIs are low priority medications and patients' willingness to make changes to their PPI therapy if needed, respectively. Ready access to pharmacists was another frequently reported facilitator to guideline-concordant PPI. Providers recommended a PPI de-implementation intervention that is specific and tells them exactly what they need to do with a PPI treatment.

CONCLUSION: In a busy inpatient setting, we developed a feasible way to assess PPI therapy, de-implement non-guideline-concordant PPI use, and provide follow-up to assess any unintended consequences. We documented barriers, facilitators, and provider recommendations that should be considered before implementing such an intervention on a large scale.

BACKGROUND: Advanced practice providers in the outpatient setting play a key role in antibiotic stewardship, yet little is known about how to engage these providers in stewardship activities and what factors influence their antibiotic prescribing practices.

METHODS: We used mixed methods to obtain data on practices and perceptions related to antibiotic prescribing by nurse practitioners (NP) and Veteran patients. We interviewed NPs working in the outpatient setting at one Veterans Affairs facility and conducted focus groups with Veterans. Emerging themes were mapped to the Systems Engineering Initiative for Patient Safety framework. We examined NP antibiotic prescribing data from 2017 to 2019.

RESULTS: We interviewed NPs and conducted Veteran focus groups. Nurse practitioners reported satisfaction with resources, including ready access to pharmacists and infectious disease specialists. Building patient trust was reported as essential to prescribing confidence level. Veterans indicated the need to better understand differences between viral and bacterial infections. NP prescribing patterns revealed a decline in antibiotics prescribed for upper respiratory illnesses over a 3-year period.

CONCLUSION: Outpatient NPs focus on educating the patient while balancing organizational access challenges. Further research is needed to determine how to include both NPs and patients when implementing outpatient antibiotic stewardship strategies. Further research is also needed to understand factors associated with the decline in nurse practitioner antibiotic prescribing observed in this study.