National Cohort Of Veterans Research Articles

A reduced risk of Alzheimer's disease is associated with the majority of cancers in a national cohort of veterans

A reduced risk of Alzheimer's disease is associated with the majority of cancers in a national cohort of veterans

Impact of diabetes control on mortality by race in a national cohort of veterans

PurposeThe association between glycated hemoglobin (HbA1c), medication use/adherence, and mortality stratified by race/ethnicity was examined in a national cohort of veterans with type 2 diabetes. MethodsA total of 892,223 veterans with diabetes in 2002 were followed through 2006. HbA1c category was the main exposure (i.e., HbA1c <7%, HbA1c 7%–8% [reference], HbA1c 8%–9%, and HbA1c >9%). Covariates included age, sex, marital status, rural/urban residence, geographic region, number of comorbidities, and diabetes medication use/adherence (i.e., adherent, medication possession ratio ≥80%; nonadherent; and nonusers). HbA1c and medication use/adherence varied over time, and Cox regression models accounting for time-varying variables were used. ResultsIn nonmedication users, HbA1c greater than 9% predicted higher mortality risk relative to HbA1c of 7%–8% in non-Hispanic whites (hazard ratio [HR], 1.55; 95% confidence interval [95% CI], 1.43–1.69), non-Hispanic blacks (NHB) (HR, 1.58; 95% CI, 1.34–1.87), and Hispanics (HR, 2.22; 95% CI, 1.75–2.82). In contrast, in nonadherent medication users, HbA1c less than 7% predicted higher mortality risk in NHB (HR, 1.12; 95% CI, 1.05–1.20), whereas HbA1c greater than 9% only predicted mortality in non-Hispanic whites (HR, 1.11; 95% CI, 1.06–1.16). In adherent medication users, HbA1c less than 7% predicted higher mortality in NHB (HR, 1.18; 95% CI, 1.07–1.31), whereas HbA1c greater than 9.0% predicted higher mortality risk across all race/ethnic groups. ConclusionWe found evidence for racial/ethnic differences in the association between glycemic control and mortality, which varied by medication use/adherence.

Fitting parametric random effects models in very large data sets with application to VHA national data

BackgroundWith the current focus on personalized medicine, patient/subject level inference is often of key interest in translational research. As a result, random effects models (REM) are becoming popular for patient level inference. However, for very large data sets that are characterized by large sample size, it can be difficult to fit REM using commonly available statistical software such as SAS since they require inordinate amounts of computer time and memory allocations beyond what are available preventing model convergence. For example, in a retrospective cohort study of over 800,000 Veterans with type 2 diabetes with longitudinal data over 5 years, fitting REM via generalized linear mixed modeling using currently available standard procedures in SAS (e.g. PROC GLIMMIX) was very difficult and same problems exist in Stata’s gllamm or R’s lme packages. Thus, this study proposes and assesses the performance of a meta regression approach and makes comparison with methods based on sampling of the full data.DataWe use both simulated and real data from a national cohort of Veterans with type 2 diabetes (n=890,394) which was created by linking multiple patient and administrative files resulting in a cohort with longitudinal data collected over 5 years.Methods and resultsThe outcome of interest was mean annual HbA1c measured over a 5 years period. Using this outcome, we compared parameter estimates from the proposed random effects meta regression (REMR) with estimates based on simple random sampling and VISN (Veterans Integrated Service Networks) based stratified sampling of the full data. Our results indicate that REMR provides parameter estimates that are less likely to be biased with tighter confidence intervals when the VISN level estimates are homogenous.ConclusionWhen the interest is to fit REM in repeated measures data with very large sample size, REMR can be used as a good alternative. It leads to reasonable inference for both Gaussian and non-Gaussian responses if parameter estimates are homogeneous across VISNs.

Regional, Geographic, and Racial/Ethnic Variation in Glycemic Control in a National Sample of Veterans With Diabetes

OBJECTIVEWe performed a retrospective analysis of a national cohort of veterans with diabetes to better understand regional, geographic, and racial/ethnic variation in diabetes control as measured by HbA1c.RESEARCH DESIGN AND METHODSA retrospective cohort study was conducted in a national cohort of 690,968 veterans with diabetes receiving prescriptions for insulin or oral hypoglycemic agents in 2002 that were followed over a 5-year period. The main outcome measures were HbA1c levels (as continuous and dichotomized at ≥8.0%).RESULTSRelative to non-Hispanic whites (NHWs), HbA1c levels remained 0.25% higher in non-Hispanic blacks (NHBs), 0.31% higher in Hispanics, and 0.14% higher in individuals with other/unknown/missing racial/ethnic group after controlling for demographics, type of medication used, medication adherence, and comorbidities. Small but statistically significant geographic differences were also noted with HbA1c being lowest in the South and highest in the Mid-Atlantic. Rural/urban location of residence was not associated with HbA1c levels. For the dichotomous outcome poor control, results were similar with race/ethnic group being strongly associated with poor control (i.e., odds ratios of 1.33 [95% CI 1.31–1.35] and 1.57 [1.54–1.61] for NHBs and Hispanics vs. NHWs, respectively), geographic region being weakly associated with poor control, and rural/urban residence being negligibly associated with poor control.CONCLUSIONSIn a national longitudinal cohort of veterans with diabetes, we found racial/ethnic disparities in HbA1c levels and HbA1c control; however, these disparities were largely, but not completely, explained by adjustment for demographic characteristics, medication adherence, type of medication used to treat diabetes, and comorbidities.

Risk of non-melanoma skin cancer in a national cohort of veterans with rheumatoid arthritis

To determine the incidence of and risk factors for non-melanoma skin cancer (NMSC) in a national cohort of veterans with RA. We examined skin cancer risk in a cohort of 20 648 patients with RA derived from the Department of Veterans' Affairs (VA) national administrative databases. The cohort was divided into two medication groups: patients treated with non-biologic and TNF-α antagonist DMARDs. We defined skin cancer as the first occurrence of an International Classification of Disease, Version 9, Clinical Modification (ICD-9-CM) code for NMSC after initiation of a DMARD. Outcome risk was described using hazard ratios (HRs) with Cox proportional hazards regression for time-to-event analysis and logistic regression. We performed medical record review to validate the diagnosis of NMSC. Incidence of NMSC was 18.9 and 12.7 per 1000 patient-years in patients on TNF-α antagonists and non-biologic DMARDs, respectively. Patients on TNF-α antagonists had a higher risk of developing NMSC (HR 1.42; 95% CI 1.24, 1.63). Risk factors for NMSC included older age, male gender, NSAID and glucocorticoid use and a history of prior malignancies. There was substantial agreement between ICD-9-CM diagnosis of NMSC and medical record validation (κ = 0.61). TNF-α antagonist therapy in veterans with RA may be associated with an increased risk of NMSC, compared with therapy with non-biologic DMARDs. Rheumatologists should carefully screen patients receiving TNF-α antagonists for pre-cancerous skin lesions and skin cancer.

TNF-α Antagonist Use and Risk of Hospitalization for Infection in a National Cohort of Veterans With Rheumatoid Arthritis

Medications used to treat rheumatoid arthritis (RA) may confer an increased risk of infection. We conducted a retrospective cohort study of veterans with RA followed in the United States Department of Veterans Affairs health care system from October 1998 through September 2005. Risk of hospitalization for infection associated with tumor necrosis factor (TNF)-α antagonists therapy was measured using an extension of Cox proportional hazards regression, adjusting for demographic characteristics, comorbid illnesses, and other medications used to treat RA. A total of 20,814 patients met inclusion criteria, including 3796 patients who received infliximab, etanercept, or adalimumab. Among the study cohort, 1465 patients (7.0%) were hospitalized at least once for infection. There were 1889 hospitalizations for infection. The most common hospitalized infections were pneumonia, bronchitis, and cellulitis. Age and several comorbid medical conditions were associated with hospitalization for infection. Prednisone (hazard ratio [HR], 2.14; 95% confidence interval [CI], 1.88-2.43) and TNF-α antagonist use (HR, 1.24; 95% CI, 1.02-1.50) were associated with hospitalization for infection, while the use of disease-modifying antirheumatic drugs (DMARDs) other than TNF-α antagonists was not. Compared to etanercept, infliximab was associated with risk for hospitalization for infection (HR, 1.51; 95% CI, 1.14-2.00), while adalimumab use was not (HR, 0.95; 95% CI, 0.68-1.33). In all treatment groups, rate of hospitalization for infection was highest in the first 8 months of therapy. We conclude that patients with RA who are treated with TNF-α antagonists are at higher risk for hospitalization for infection than those treated with other DMARDs. Prednisone use is also a risk factor for hospitalization for infection.

Preexisting Comorbid Psychiatric Conditions and Mortality in Nonsurgical Intensive Care Patients

To examine the effects of preexisting comorbid psychiatric conditions on mortality in a large cohort of patients admitted to a nonsurgical intensive care unit. This retrospective cohort study involved 66,672 consecutive eligible nonsurgical patients admitted to intensive care units in 129 Veterans Health Administration hospitals during 2005 and 2006. Preexisting comorbid psychiatric conditions were identified by using diagnoses from outpatient encounters in the prior year for depression, anxiety, psychosis, bipolar disorders, and posttraumatic stress disorder. Generalized estimating equations were used to adjust the risks of in hospital and 30-day mortality for demographics, comorbid medical conditions, markers of severity, and abnormal findings on laboratory tests at admission. Comorbid psychiatric conditions were identified in 28% (n = 18 698) of patients. Patients with preexisting comorbid psychiatric conditions had lower (P < .001) unadjusted in hospital mortality (7.3% vs 8.7%) and 30-day mortality (10.0% vs 12.8%) than did patients without such conditions. After demographics, comorbid medical conditions, and severity were adjusted for, risk of in-hospital mortality among patients with comorbid psychiatric conditions was somewhat higher (odds ratio, 1.07, 95% confidence interval, 1.01-1.14; P = .02), although differences in 30-day mortality (odds ratio, 1.01, 95% confidence interval, 0.94-1.08; P = .70) were no longer significant. Preexisting comorbid psychiatric conditions are common among intensive care patients, but after comorbid medical conditions and severity were adjusted for, preexisting comorbid psychiatric conditions were not associated with a higher risk of 30-day mortality in a large national cohort of veterans.

Herpes Zoster Risk Factors in a National Cohort of Veterans with Rheumatoid Arthritis

Herpes zoster occurs more commonly in patients taking immunosuppressive medications, although the risk associated with different medications is poorly understood. We conducted a retrospective cohort study involving 20,357 patients who were followed in the Veterans Affairs healthcare system and treated for rheumatoid arthritis from October 1998 through June 2005. Cox proportional hazards regression was used to determine risk factors for herpes zoster and herpes zoster-free survival. Chart review was performed to validate the diagnosis of herpes zoster. The incidence of herpes zoster was 9.96 episodes per 1000 patient-years. In time-to-event analysis, patients receiving medications used to treat mild rheumatoid arthritis were less likely to have an episode of herpes zoster than patients receiving medications used to treat moderate and severe rheumatoid arthritis (P < .001). Independent risk factors for herpes zoster included older age, prednisone use, medications used to treat moderate and severe rheumatoid arthritis, malignancy, chronic lung disease, renal failure, and liver disease. Among patients receiving tumor necrosis factor-alpha antagonists, etanercept (hazard ratio, 0.62) and adalimumab (hazard ratio, 0.53) were associated with a lower risk of herpes zoster. There was excellent agreement between the International Classification of Diseases, Version 9, Clinical Modification diagnosis of herpes zoster and diagnosis by chart review (kappa = 0.92). Risk factors for herpes zoster included older age, prednisone use, medications used to treat moderate and severe rheumatoid arthritis, and several comorbid medical conditions. These results demonstrate that the Department of Veterans Affairs' national administrative databases can be used to study rare adverse drug events.

The Effectiveness of Inpatient Rehabilitation in the Acute Postoperative Phase of Care After Transtibial or Transfemoral Amputation: Study of an Integrated Health Care Delivery System

Stineman MG, Kwong PL, Kurichi JE, Prvu-Bettger JA, Vogel WB, Maislin G, Bates BE, Reker DM. The effectiveness of inpatient rehabilitation in the acute postoperative phase of care after transtibial or transfemoral amputation: study of an integrated health care delivery system. Objective To compare outcomes between lower-extremity amputees who receive and do not receive acute postoperative inpatient rehabilitation within a large integrated health care delivery system. Design An observational study using multivariable propensity score risk adjustment to reduce treatment selection bias. Setting Data compiled from 9 administrative databases from Veterans Affairs Medical Centers. Participants A national cohort of veterans (N=2673) who underwent transtibial or transfemoral amputation between October 1, 2002, and September 30, 2004. Interventions Not applicable. Main Outcome Measures One-year cumulative survival, home discharge from the hospital, and prosthetic limb procurement within the first postoperative year. Results After reducing selection bias, patients who received acute postoperative inpatient rehabilitation compared to those with no evidence of inpatient rehabilitation had an increased likelihood of 1-year survival (odds ratio [OR]=1.51; 95% confidence interval [CI], 1.26–1.80) and home discharge (OR=2.58; 95% CI, 2.17–3.06). Prosthetic limb procurement did not differ significantly between groups. Conclusions The receipt of rehabilitation in the acute postoperative inpatient period was associated with a greater likelihood of 1-year survival and home discharge from the hospital. Results support early postoperative inpatient rehabilitation following amputation.

Depression and Other Mental Health Diagnoses After Stroke Increase Inpatient and Outpatient Medical Utilization Three Years Poststroke

Poststroke depression (PSD) has been linked to negative outcomes, including mortality and decreased functioning. However, the effect of PSD and other mental health conditions on inpatient and outpatient healthcare utilization after stroke has not been examined. The primary objective of this study was to evaluate the relationship between PSD and healthcare utilization after stroke. The secondary objective was to evaluate the relationship between other mental health diagnoses and medical utilization after stroke. We examined 3 years of poststroke healthcare utilization data in a national cohort of veterans with ischemic stroke. Mental health diagnoses were identified with inpatient International Classification of Diseases, 9th Revision codes. Multivariate analysis of variance was used to compare patient characteristics and multivariate linear regression to model utilization in: 1) patients with PSD, 2) patients diagnosed with other mental health conditions poststroke, and 3) patients with no mental health diagnosis. Of 51,119 veterans identified, 2405 (5%) received a PSD diagnosis and 2257 (4%) received a diagnosis of another mental health condition after their stroke. Patients with PSD had significantly more inpatient hospitalization days and outpatient visits than those without any mental health diagnosis, even after adjusting for the number of mental health clinic visits. Likewise, patients diagnosed with other mental health diagnoses had greater medical utilization than patients without any mental health diagnosis. Both PSD and other mental health diagnoses independently predicted medical utilization. Mental health diagnosis after stroke increases inpatient and outpatient healthcare utilization in the first 3 years poststroke. Additional biologic and psychosocial factors should be investigated.

Applying diabetes-related Prevention Quality Indicators to a national cohort of veterans with diabetes.

The Prevention Quality Indicators (PQIs) are measures of hospitalizations that reflect quality of ambulatory care. We applied the PQIs of metabolic decompensation to Veterans Health Administration (VHA) utilizers with diabetes. We identified patient-level characteristics associated with hospitalization for metabolic decompensation, developed a risk-adjustment model for the measures, and compared regional network performance using these PQIs. This was a retrospective cohort study of 406575 veterans with diabetes who used the VHA between 1997 and 1999. The outcomes were the PQIs of uncontrolled diabetes, short-term complications, or a combined measure. Patient-level variables were identified from administrative databases. Variation in performance of the networks was compared between full risk adjustment and age and sex adjustment only. In fiscal year 1999, there were 1719 VHA discharges (4.2 per 1000 cohort members) for uncontrolled and short-term complications of diabetes. A logistic regression model including age, sex, marital status, Charlson Comorbidity Index, mental health condition, insulin use, and oral antiglycemic medication use was developed for risk adjustment of the combined PQI. Full risk adjustment changed performance ranks of the networks using the combined PQI outcome relative to age and sex adjustment only. Ten networks remained in the same quartile of performance, five moved one quartile, and seven moved two or more quartiles. The PQIs of uncontrolled and short-term complications of diabetes are uncommon outcomes among veterans with diabetes and should be used only as a combined outcome. More complete risk adjustment should be used when comparing systems of care using the combined measure.

Variation in health care utilization and outcomes in ischemic stroke

P125 Objectives: Little data is available about regional variation in utilization and outcomes during and after ischemic stroke (IS). Examining these relationships can lead to better understanding of the patient and process characteristics that influence stroke outcomes. Methods: Index IS admissions were identified using primary position ICD-9 codes 434 and 436 in a national cohort of veterans hospitalized at any VAMC from 10/90 to 10/97. Demographic, admission, mortality and outpatient data were extracted from VA administrative databases for the regions Northeast (NE), South (S), Midwest (MW) and West (W). Discharge diagnoses were used to assess risk factors and to construct the Charlson Index (CI), a predictor of mortality risk (↑ CI = ↑ risk). Time to death was modeled with Cox regression. Results: We identified 55,094 IS patients (98% male, 73% white, mean age 68). Regional data are summarized in the Table. NE patients were sicker (higher CI), had longer LOS and higher in-hospital mortality. However, adjusted subsequent mortality was lower in the NE and W (hazard ratio NE = .82, W = .91, p&lt;.001 for both). More NE and W patients had a neurology or primary care visit within 60 days or 1 year. Conclusions: Regional variations exist in outcomes and patterns of care after IS. Adjusted subsequent mortality is lower in regions where more patients have follow-up. Prospective studies evaluating these variations are needed to determine the best strategies for providing optimal stroke care.

Depression and Mental Health Disorders Increase Adjusted Risk of Death Post-Stroke

P140 Background/Objectives: Post-stroke depression (PSD) has been linked to higher mortality after stroke. Whether PSD and other mental health (MH) diagnoses are independent risks for mortality is not known. The purpose of this study was to examine the relationship of depression and other MH diagnoses to mortality post-stroke. Methods: We examined a national cohort of veterans hospitalized at any VA Medical Center from 10/90–10/98. Index ischemic stroke admissions were defined as the first admission with primary discharge ICD-9 diagnosis code 434 or 436. Demographic, admission and all-cause mortality data were abstracted from VA administrative databases. ICD-9 codes at discharge were used to assess risk factors and to construct the Charlson Index, a measure of mortality risk. PSD within 1-year post-stroke was identified using ICD-9 codes. Other MH diagnoses during the entire period were also ascertained. We analyzed the data with and without patients who died &lt; 365 days post-stroke, as patients with early death have less chance of receiving a 1-year PSD diagnosis. Hazard ratios (HR) were modeled using Cox regression. Results: We identified 55,094 patients with ischemic stroke, 40,171 survived &gt; 365 days post-stroke. Mean age = 67 and 30% were non-white. 2,537 (6%) had depression within 1 year and 6,761 (17%) had another MH diagnosis. PSD increased adjusted relative mortality risk by 25% (HR 1.25, 95% CI 1.16–1.34). This increase was exceeded only by congestive heart failure (HR 1.39 95% CI 1.31, 1.48), was equal to that of atrial fibrillation (HR 1.25, 95%CI 1.18,1.32) and exceeded age, hypertension, diabetes and coronary artery disease (all HR &lt; 1.21). Other MH diagnoses also increased adjusted relative mortality risk (HR 1.20, 95% CI 1.14, 1.26). Similar results were obtained with the model including patients who died &lt; 365 days after stroke. Conclusions: PSD independently increases adjusted risk of all-cause mortality post-stroke. The magnitude of this risk equals many chronic cardiovascular conditions. Other MH diagnoses also increase adjusted mortality risk. Future studies should address whether treatment of these conditions can decrease mortality risk after stroke.

Depression and Mental Health Disorders Increase Adjusted Risk of Death Post-Stroke

P140 Background/Objectives: Post-stroke depression (PSD) has been linked to higher mortality after stroke. Whether PSD and other mental health (MH) diagnoses are independent risks for mortality is not known. The purpose of this study was to examine the relationship of depression and other MH diagnoses to mortality post-stroke. Methods: We examined a national cohort of veterans hospitalized at any VA Medical Center from 10/90–10/98. Index ischemic stroke admissions were defined as the first admission with primary discharge ICD-9 diagnosis code 434 or 436. Demographic, admission and all-cause mortality data were abstracted from VA administrative databases. ICD-9 codes at discharge were used to assess risk factors and to construct the Charlson Index, a measure of mortality risk. PSD within 1-year post-stroke was identified using ICD-9 codes. Other MH diagnoses during the entire period were also ascertained. We analyzed the data with and without patients who died Results: We identified 55,094 patients with ischemic stroke, 40,171 survived > 365 days post-stroke. Mean age = 67 and 30% were non-white. 2,537 (6%) had depression within 1 year and 6,761 (17%) had another MH diagnosis. PSD increased adjusted relative mortality risk by 25% (HR 1.25, 95% CI 1.16–1.34). This increase was exceeded only by congestive heart failure (HR 1.39 95% CI 1.31, 1.48), was equal to that of atrial fibrillation (HR 1.25, 95%CI 1.18,1.32) and exceeded age, hypertension, diabetes and coronary artery disease (all HR Conclusions: PSD independently increases adjusted risk of all-cause mortality post-stroke. The magnitude of this risk equals many chronic cardiovascular conditions. Other MH diagnoses also increase adjusted mortality risk. Future studies should address whether treatment of these conditions can decrease mortality risk after stroke.