Benchmark findings from a veteran electronic patient-reported outcomes evaluation from a chronic pain management telehealth program.

BACKGROUND: Chronic pain is a leading cause of disability and negatively impacts biological/physical, psychological, and social aspects of life resulting in significant pain interference or disability. This project was part of a longitudinal mixed-methods implementation evaluation of the TelePain-Empower Veterans Program (EVP), a non-pharmacological chronic pain intervention. The purpose of this quality management project was to examine electronic patient-reported outcome measures (ePROs) including primary pain-related (intensity, interference, catastrophizing, kinesiophobia) and secondary outcomes (physical, psychological, acceptance, social) to determine TelePain-EVP effectiveness. Secondary purpose was to examine dosing effects to better understand potential dose relationships between EVP use and ePROs. METHODS: Standardized ePRO measures were examined at week 1 (baseline), week 10 (post-EVP), and week 26 (follow-up). Qualtrics, a cloud-based platform was used to collect ePRO data at each time point. Veterans that completed at-least one survey at any specified time point were categorized as responders (n = 221). Linear-mixed models (LMMs) were fit to assess changes for each primary and secondary ePRO. RESULTS: Participants ranged from 24 to 81 years old; veterans were typically male (65.16%), black or African American (76.47%), married or partnered (41.63%), attended at-least some college or vocational school (67.87%), and reported low back as their primary pain location (29.41%). There was a significant decrease in pain catastrophizing from baseline to post-TelePain-EVP (p < .001). However, pain catastrophizing improvement from baseline was not present at week 26 (p = .116). Pain interference also decreased from baseline to post-treatment (p = .05), but this improvement did not exceed the adjusted significance threshold. Additional pre-post improvements were also observed for certain secondary ePROs: psychological (anxiety, depression), acceptance (activities engagement). Only the activities engagement effect remained 26 weeks from baseline. Mixed results were observed for EVP dose across primary and secondary outcomes. CONCLUSIONS: Evidence from this evaluation indicate that TelePain-EVP has positive outcomes for certain pain (catastrophizing), psychological (anxiety, depression), and acceptance (activities engagement) for veterans with chronic pain. More TelePain related studies and enterprise-wide evaluations are needed along with comparative and cost effectiveness methods to determine patient benefits and the economic value gained of treatment options such as TelePain-EVP.

Mobile and Web-Based Partnered Intervention to Improve Remote Access to Pain and Posttraumatic Stress Disorder Symptom Management: Recruitment and Attrition in a Randomized Controlled Trial.

BACKGROUND: Increasing access to nonpharmacological interventions to manage pain and posttraumatic stress disorder (PTSD) is essential for veterans. Complementary and integrative health (CIH) interventions can help individuals manage symptom burden with enhanced accessibility via remotely delivered health care. Mission Reconnect (MR) is a partnered, self-directed intervention that remotely teaches CIH skills. OBJECTIVE: The purpose of this paper is to describe the recruitment, onboarding phase, and attrition of a fully remote randomized controlled trial (RCT) assessing the efficacy of a self-directed mobile and web-based intervention for veterans with comorbid chronic pain and PTSD and their partners. METHODS: A total of 364 veteran-partner dyads were recruited to participate in a mixed methods multisite waitlist control RCT. Qualitative attrition interviews were conducted with 10 veterans with chronic pain and PTSD, and their self-elected partners (eg, spouse) who consented but did not begin the program. RESULTS: At the point of completing onboarding and being randomized to the 2 treatment arms, of the 364 recruited dyads, 97 (26.6%) failed to complete onboarding activities. Reported reasons for failure to complete onboarding include loss of self-elected partner buy-in (n=8, 8%), difficulties with using remote data collection methods and interventions (n=30, 31%), and adverse health experiences unrelated to study activities (n=23, 24%). Enrolled veterans presented at baseline with significant PTSD symptom burden and moderate-to-severe pain severity, and represented a geographically and demographically diverse population. Attrition interviews (n=10) indicated that misunderstanding MR including the intent of the intervention or mistaking the surveys as the actual intervention was a reason for not completing the MR registration process. Another barrier to MR registration was that interviewees described the mailed study information and registration packets as too confusing and excessive. Competing personal circumstances including health concerns that required attention interfered with MR registration. Common reasons for attrition following successful MR registration included partner withdrawal, adverse health issues, and technological challenges relating to the MR and electronic data collection platform (Qualtrics). Participant recommendations for reducing attrition included switching to digital forms to reduce participant burden and increasing human interaction throughout the registration and baseline data collection processes. CONCLUSIONS: Challenges, solutions, and lessons learned for study recruitment and intervention delivery inform best practices of delivering remote self-directed CIH interventions when addressing the unique needs of this medically complex population. Successful recruitment and enrollment of veterans with chronic pain and PTSD, and their partners, to remote CIH programs and research studies requires future examination of demographic and symptom-associated access barriers. Accommodating the unique needs of this medically complex population is essential for improving the effectiveness of CIH programs. Disseminating lessons learned and improving access to remotely delivered research studies and CIH programs is paramount in the post-COVID-19 climate. TRIAL REGISTRATION: ClinicalTrials.gov NCT03593772; https://clinicaltrials.gov/ct2/show/NCT03593772.

Machine learning to develop a predictive model of pressure injury in persons with spinal cord injury.

STUDY DESIGN: A 5-year longitudinal, retrospective, cohort study. OBJECTIVES: Develop a prediction model based on electronic health record (EHR) data to identify veterans with spinal cord injury/diseases (SCI/D) at highest risk for new pressure injuries (PIs). SETTING: Structured (coded) and text EHR data, for veterans with SCI/D treated in a VHA SCI/D Center between October 1, 2008, and September 30, 2013. METHODS: A total of 4709 veterans were available for analysis after randomly selecting 175 to act as a validation (gold standard) sample. Machine learning models were created using ten-fold cross validation and three techniques: (1) two-step logistic regression; (2) regression model employing adaptive LASSO; (3) and gradient boosting. Models based on each method were compared using area under the receiver-operating curve (AUC) analysis. RESULTS: The AUC value for the gradient boosting model was 0.62 (95% CI = 0.54-0.70), for the logistic regression model it was 0.67 (95% CI = 0.59-0.75), and for the adaptive LASSO model it was 0.72 (95% CI = 0.65-80). Based on these results, the adaptive LASSO model was chosen for interpretation. The strongest predictors of new PI cases were having fewer total days in the hospital in the year before the annual exam, higher vs. lower weight and most severe vs. less severe grade of injury based on the American Spinal Cord Injury Association (ASIA) Impairment Scale. CONCLUSIONS: While the analyses resulted in a potentially useful predictive model, clinical implications were limited because modifiable risk factors were absent in the models.

Primary care virtual resource use prior and post COVID-19 pandemic onset.

BACKGROUND: The COVID-19 pandemic has been a catalyst for rapid uptake of virtual care through the use of virtual health resources (VHR). In the Department of Veterans Affairs (VA) Healthcare System, virtual care has been critical to maintaining healthcare access for patients during COVID-19. In the current study we describe primary care patient aligned care team (PACT) VHR use patterns within one VA medical center (i.e., hospital facility and five community-based outpatient clinics) pre- and post-COVID-19 onset. METHODS: VHR provider and patient use data from 106 individual PACTs were extracted monthly between September 2019 to September 2020. Data were extracted from VHA web-based project application and tracking databases. Using longitudinal data, mixed effect models were used to compare pre- and post-COVID onset slopes. RESULTS: Findings highlight an increase in patient users of secure messaging (SM) and telehealth. The rate of utilization among these patients increased for SM but not for telehealth visits or online prescription refill (RxRefill) use. Finally, VetLink Kiosk check ins that are done at in person visits, diminished abruptly after COVID-19 onset. CONCLUSIONS: These data provide a baseline of VHR use at the PACT level after the initial impact of the COVID-19 pandemic and can inform healthcare delivery changes within the VA systems over time. Moreover, this project produced a data extraction blueprint, that is the first of its kind to track VA VHR use leveraging secondary data sources.

Preliminary Evidence on the Association of Complementary and Integrative Health Care Program Participation and Medical Cost in Veterans.

INTRODUCTION: Transforming Health and Resilience through Integration of Values-based Experiences (THRIVE) is a complimentary and integrative health program. THRIVE is delivered through shared medical appointments where participants engage in provider-led education and group discussion on wellness-related topics. THRIVE has been associated with improved patient-reported outcomes in a female veteran cohort. This quality improvement study evaluated the association between THRIVE participation and Veterans Health Administration (VHA) healthcare costs across a 1 year period. MATERIALS AND METHODS: A cohort study design (n = 184) used VHA administrative data to estimate the cost difference between 1 year pre- and post-THRIVE participation. The 1 year post-cost of the THRIVE cohort was then compared to the 1 year cost of a quasi-experimental waitlist control group (n = 156). Data sources included VHA administrative and electronic health records. RESULTS: Patients were roughly 51 years old, were typically White/Caucasian, and had a service priority level representing catastrophic disability. The adjusted post-THRIVE cost was $26,291 [95% confidence interval (CI): $23,014-29,015]; $1,720 higher than the previous year's cost but was not statistically significant (P = 0.289). However, a comparison between the THRIVE cohort and a group of waitlist THRIVE patients (n = 156) the intervention group on average was $8,108 more than the waitlist group (95% CI: $3,194-14,005; P < 0.01). CONCLUSIONS: In summary, data analysis of veterans' annual healthcare cost trajectories were inconclusive. This preliminary study produced mixed results requiring more research with larger samples and randomized control trial methodology. Evidence of whether the THRIVE intervention can maintain cost effectiveness while maintaining its supported evidence of healthcare quality is needed.

Mortality and Cause of Death in Younger Homeless Veterans.

OBJECTIVES: Increased mortality has been documented in older homeless veterans. This retrospective study examined mortality and cause of death in a cohort of young and middle-aged homeless veterans. METHODS: We examined US Department of Veterans Affairs records on homelessness and health care for 2000-2003 and identified 23 898 homeless living veterans and 65 198 non-homeless living veterans aged 30-54. We used National Death Index records to determine survival status. We compared survival rates and causes of death for the 2 groups during a 10-year follow-up period. RESULTS: A greater percentage of homeless veterans (3905/23 898, 16.3%) than non-homeless veterans (4143/65 198, 6.1%) died during the follow-up period, with a hazard ratio for risk of death of 2.9. The mean age at death (52.3 years) for homeless veterans was approximately 1 year younger than that of non-homeless veterans (53.2 years). Most deaths among homeless veterans (3431/3905, 87.9%) and non-homeless veterans (3725/4143, 89.9%) were attributed to 7 cause-of-death categories in the International Classification of Diseases, 10th Revision (cardiovascular system; neoplasm; external cause; digestive system; respiratory system; infectious disease; and endocrine, nutritional, and metabolic diseases). Death by violence was rare but was associated with a significantly higher risk among homeless veterans than among non-homeless veterans (suicide hazard ratio = 2.7; homicide hazard ratio = 7.6). CONCLUSIONS: Younger and middle-aged homeless veterans had higher mortality rates than those of their non-homeless veteran peers. Our results indicate that homelessness substantially increases mortality risk in veterans throughout the adult age range. Health assessment would be valuable for assessing the mortality risk among homeless veterans regardless of age.

Predictors of Mortality in Older Homeless Veterans.

OBJECTIVES: In this analysis of a cohort of older homeless veterans, we examined psychosocial, health, housing, and employment characteristics to identify predictors of mortality. METHOD: Our sample of 3,620 older veterans entered Veteran Affairs homeless programs in years 2000-2003. Fifteen variables from a structured interview described this sample and served as predictors. National Death Index data for years 2000-2011 were used to ascertain death. Survival table analyses were conducted to estimate and plot cumulative survival functions. To determine predictors and estimate hazard functions, Cox proportional hazards regression analysis was conducted. RESULTS: Five variables (presence of a serious health issue, hospitalization for alcohol abuse, alcohol dependency, unemployment for 3 years, and age 60+) were associated with increased risk of death; three (non-White, drug dependency, and dental problems) were associated with reduced risk. A risk score, based on total unit-weighted risk for all eight predictors, was used to identify three groups that were found to differ significantly in mortality. CONCLUSIONS: These analyses underline the jeopardy faced by older homeless veterans in terms of early death. We were able to identify several variables associated with mortality; more importantly, we were able to show that a risk score based on status for these variables was significantly related to survival.

A Systems Science Approach to Understanding Polytrauma and Blast-Related Injury: Bayesian Network Model of Data From a Survey of the Florida National Guard.

We sought to further define the epidemiology of the complex, multiple injuries collectively known as polytrauma/blast-related injury (PT/BRI). Using a systems science approach, we performed Bayesian network modeling to find the most accurate representation of the complex system of PT/BRI and identify key variables for understanding the subsequent effects of blast exposure in a sample of Florida National Guard members (1,443 deployed to Operation Enduring Freedom/Operation Iraqi Freedom and 1,655 not deployed) who completed an online survey during the period from 2009 to 2010. We found that postdeployment symptoms reported as present at the time of the survey were largely independent of deployment per se. Blast exposure, not mild traumatic brain injury (TBI), acted as the primary military deployment-related driver of PT/BRI symptoms. Blast exposure was indirectly linked to mild TBI via other deployment-related traumas and was a significant risk for a high level of posttraumatic stress disorder (PTSD) arousal symptoms. PTSD arousal symptoms and tinnitus were directly dependent upon blast exposure, with both acting as bridge symptoms to other postdeployment mental health and physical symptoms, respectively. Neurobehavioral or postconcussion-like symptoms had no significant dependence relationship with mild TBI, but they were synergistic with blast exposure in influencing PTSD arousal symptoms. A replication of this analysis using a larger PT/BRI database is warranted.

Increased Mortality Among Older Veterans Admitted to VA Homelessness Programs.

OBJECTIVE: National Death Index data were examined to describe mortality patterns among older veterans who are homeless. METHODS: Homelessness and health care records from the U.S. Department of Veterans Affairs were used to identify old (ages 55-59) and older (ages ≥60) veterans who were (N=4,475) or were not (N=20,071) homeless. Survival functions and causes of death of the two samples over an 11-year follow-up period were compared. RESULTS: Substantially more veterans who were homeless (34.9%) died compared with the control sample (18.2%). Veterans who were homeless were approximately 2.5 years younger at time of death compared with the control sample. Older veterans who were homeless had the lowest survival rate (58%). No disease category appeared to be critical in reducing survival time. Suicide was twice as frequent in the homeless (.4%) versus the control (.2%) sample. CONCLUSIONS: Older veterans who were homeless experienced excess mortality and increased suicide risk.

Association between health literacy and medical care costs in an integrated healthcare system: a regional population based study.

BACKGROUND: Low health literacy is associated with higher health care utilization and costs; however, no large-scale studies have demonstrated this in the Veterans Health Administration (VHA). This research evaluated the association between veterans' health literacy and their subsequent VHA health care costs across a three-year period. METHODS: This retrospective study used a Generalized Linear Model to estimate the relative association between a patient's health literacy and VHA medical costs, adjusting for covariates. Secondary data sources included electronic health records and administrative data in the VHA (e.g., Medical and DCG SAS Datasets and DSS-National Data Extracts). Health literacy assessments and identifiers were electronically retrieved from the originating health system. Demographic and cost data were retrieved from the VHA centralized databases for the corresponding patients who had VHA use in all three years. RESULTS: In a study of 92,749 veterans with service utilization from 2007-2009, average per patient cost for those with inadequate and marginal health literacy was significantly higher ($31,581 [95 % CI: $30,186 - $32,975]; $23,508 [95 % CI: $22,749 - $24,268]) than adequate health literacy ($17,033 [95 % CI: $16,810 - $17,255]). Estimated three-year cost associated with veterans' with marginal and inadequate health literacy was $143 million dollars more than those with adequate health literacy. CONCLUSIONS: Analyses suggest when controlling for other person-level factors within the VHA integrated healthcare system, lower health literacy is a significant independent factor associated with increased health care utilization and costs. This study confirms the association of lower health literacy with higher medical service utilization and pharmacy costs for veterans enrolled in the VHA. Confirmation of higher costs of care associated with lower health literacy suggests that interventions might be designed to remediate health literacy needs and reduce expenditures. These analyses suggest 17.2 % (inadequate & marginal) of the Veterans in this population account for almost one-quarter (24 %) of VA medical and pharmacy cost for this 3-year period. Meeting the needs of those with marginal and inadequate health literacy could produce potential economic savings of approximately 8 % of total costs for this population.

Improving identification of fall-related injuries in ambulatory care using statistical text mining.

OBJECTIVES: We determined whether statistical text mining (STM) can identify fall-related injuries in electronic health record (EHR) documents and the impact on STM models of training on documents from a single or multiple facilities. METHODS: We obtained fiscal year 2007 records for Veterans Health Administration (VHA) ambulatory care clinics in the southeastern United States and Puerto Rico, resulting in a total of 26 010 documents for 1652 veterans treated for fall-related injury and 1341 matched controls. We used the results of an STM model to predict fall-related injuries at the visit and patient levels and compared them with a reference standard based on chart review. RESULTS: STM models based on training data from a single facility resulted in accuracy of 87.5% and 87.1%, F-measure of 87.0% and 90.9%, sensitivity of 92.1% and 94.1%, and specificity of 83.6% and 77.8% at the visit and patient levels, respectively. Results from training data from multiple facilities were almost identical. CONCLUSIONS: STM has the potential to improve identification of fall-related injuries in the VHA, providing a model for wider application in the evolving national EHR system.

Posttraumatic stress disorder and use of psychiatric and alcohol related services: the effect of the 2004-2005 Florida hurricane seasons on veterans.

The purpose of this study was to document preliminary findings of the association between posttraumatic stress disorder (PTSD), mental health service use, and alcohol related health visits among veterans following 2004-2005 Florida hurricane seasons. A retrospective review of the Veterans Health Administration Medical SAS Outpatient Dataset was conducted to identify veterans residing in Florida during the 2004-2005 hurricane seasons with a history of PTSD and/or PTSD and a substance use disorder. It was found that veterans with PTSD residing in counties affected by hurricanes demonstrated an immediate 28 % increase in use of mental health services following hurricane landfall versus veterans residing in non-hurricane affected counties (+28.0 vs. -6.5 %, p = 0.001). Additionally, veterans residing in affected counties were found to use more group psychotherapy treatment sessions overall (30.3 vs. 27.2 %, p = 0.001). Of note, veterans with PTSD experienced a -0.16 per month (p = 0.114) decrease in alcohol related visits following the 2004 hurricane season. These findings provide insight into the mental health needs of veterans with PTSD following a disaster and can inform delivery of services to veterans with PTSD and alcohol related issues in disaster prone areas.

All-cause mortality rates of hip fractures treated in the VHA: do they differ from Medicare facilities?

PURPOSE: To estimate the 1-year all-cause mortality rates for hip fracture (HFx) patients hospitalized at Veterans Health Administration (VHA) facilities and compare with previous published mortality rates for veterans treated in Medicare facilities. METHODS: In total, 7 years of VHA discharge data on HFxs for 12,539 patients age 65 and older were combined with national death registry data. We performed a 1-year survival analysis using the Cox proportional hazard method. RESULTS: The adjusted rates for veterans treated in the VHA (30 days=9.3%, 90 days=17.5%, 180 days=23.3%, 365 days=29.8%) were similar to veterans treated in Medicare facilities (30 days=8.9%, 90 days=15.6%, 180 days=21.8%, 365 days=29.9%). For veterans treated for a HFx in Medicare facilities, the average length of stay was 7 days and 49% were discharged to a nursing home. Veterans treated in the VHA had an average length of stay of 14 days and only 35% were discharged to a nursing home. CONCLUSIONS: Our study suggests no difference in HFx-adjusted mortality rates between the VHA and Medicare facilities. Given the institutional factor differences between Medicare and the VHA, future study and comparison of health outcomes for nursing home HFx patients and related costs between these two health care programs may contribute to the on-going health care reform debate.