Predictive validity of clinical AUDIT-C alcohol screening scores and changes in scores for three objective alcohol-related outcomes in a Veterans Affairs population.

AIMS: To evaluate the association between Alcohol Use Disorder Identification Test-Consumption (AUDIT-C) alcohol screening scores, collected as part of routine clinical care, and three outcomes in the following year (Aim 1), and the association between changes in AUDIT-C risk group at 1-year follow-up and the same outcomes in the subsequent year (Aim 2). DESIGN: Cohort study. SETTING: Twenty-four US Veterans Affairs (VA) healthcare systems (2004-07), before systematic implementation of brief intervention. PARTICIPANTS: A total of 486 115 out-patients with AUDIT-Cs documented in their electronic health records (EHRs) on two occasions ≥ 12 months apart ('baseline' and 'follow-up'). MEASUREMENTS: Independent measures were baseline AUDIT-C scores and change in standard AUDIT-C risk groups (no use, low-risk use and mild, moderate, severe misuse) from baseline to follow-up. Outcome measures were (1) high-density lipoprotein cholesterol (HDL), (2) alcohol-related gastrointestinal hospitalizations ('GI hospitalizations') and (3) physical trauma, each in the years after baseline and follow-up. FINDINGS: Baseline AUDIT-C scores had a positive association with outcomes in the following year. Across AUDIT-C scores 0-12, mean HDL ranged from 41.4 [95% confidence interval (CI) = 41.3-41.5] to 53.5 (95% CI = 51.4-55.6) mg/l, and probabilities of GI hospitalizations from 0.49% (95% CI = 0.48-0.51%) to 1.8% (95% CI = 1.3-2.3%) and trauma from 3.0% (95% CI = 2.95-3.06%) to 6.0% (95% CI = 5.2-6.8%). At follow-up, patients who increased to moderate or severe alcohol misuse had consistently higher mean HDL and probabilities of subsequent GI hospitalizations or trauma compared with those who did not (P-values all < 0.05). For example, among those with baseline low-risk use, in those with persistent low-risk use versus severe misuse at follow-up, the probabilities of subsequent trauma were 2.65% (95% CI = 2.54-2.75%) versus 5.15% (95% CI = 3.86-6.45%), respectively. However, for patients who decreased to lower AUDIT-C risk groups at follow-up, findings were inconsistent across outcomes, with only mean HDL decreasing in most groups that decreased use (P-values all < 0.05). CONCLUSIONS: When AUDIT-C screening is conducted in clinical settings, baseline AUDIT-C scores and score increases to moderate-severe alcohol misuse at follow-up screening appear to have predictive validity for HDL cholesterol, alcohol-related gastrointestinal hospitalizations and physical trauma. Decreasing AUDIT-C scores collected in clinical settings appear to have predictive validity for only HDL.

Thiazolidinediones are associated with a reduced risk of COPD exacerbations.

BACKGROUND: Thiazolidinediones (TZDs) are oral antihyperglycemic medications that are selective agonists to peroxisome proliferator-activated receptor gamma and have been shown to have potent anti-inflammatory effects in the lung. OBJECTIVE: The purpose of this study was to assess whether exposure to TZDs is associated with a decreased risk of chronic obstructive pulmonary disease (COPD) exacerbation. METHODS: A cohort study was performed by collecting data on all US veterans with diabetes and COPD who were prescribed oral antihyperglycemic medications during from period of October 1, 2005 to September 30, 2007. Patients who had two or more prescriptions for TZDs were compared with patients who had two or more prescriptions for an alternative oral anti-hyperglycemic medication. Multivariable negative binomial regression was performed with adjustment for potential confounding factors. The primary outcome was COPD exacerbations, including both inpatient and outpatient exacerbations. RESULTS: We identified 7,887 veterans who were exposed to TZD and 42,347 veterans who were exposed to non-TZD oral diabetes medications. COPD exacerbations occurred in 1,258 (16%) of the TZD group and 7,789 (18%) of the non-TZD group. In multivariable negative binomial regression, there was a significant reduction in the expected number of COPD exacerbations among patients who were exposed to TZDs with an incidence rate ratio of 0.86 (95% CI 0.81-0.92). CONCLUSION: Exposure to TZDs was associated with a small but significant reduction in risk for COPD exacerbation among diabetic patients with COPD.

Survey nonresponders incurred higher medical utilization and lower medication adherence.

OBJECTIVES: To compare healthcare costs, utilization, and medication adherence between diabetic responders and nonresponders to a patient satisfaction survey. STUDY DESIGN: We performed a retrospective cohort study of 40,766 patients with diabetes who had been randomly selected to receive the 2006 Veterans Affairs' Survey of Healthcare Experiences of Patients. Outcomes were measured during the following year. METHODS: We used multivariable models to compare healthcare costs (generalized linear models), utilization (negative binomial regression), and adherence to oral hypoglycemic medications (logistic regression) between survey responders and nonresponders. RESULTS: There were 26,051 patients (64%) who responded to the survey. Survey nonresponders incurred significantly higher healthcare costs (incremental effect, $792; 95% CI, $599-$986; P < .01). Nonresponders had a modest increase in primary care (incidence rate ratio [IRR], 1.06; 95% CI, 1.05-1.08; P < .01) and specialty care visits (IRR, 1.17; 95% CI, 1.12-1.22; P < .01), but more substantial increases in mental health visits (IRR, 1.74; 95% CI, 1.62-1.87; P < .01) and hospitalizations (IRR, 1.60; 95% CI, 1.46-1.75; P < .01). Medication adherence was significantly lower among survey nonresponders (odds ratio, 0.68; 95% CI, 0.65-0.74; P < .01). CONCLUSIONS: Nonresponders to a patient satisfaction survey incurred higher healthcare costs and utilization, but had lower medication adherence. Understanding these characteristics helps to assess the impact of nonresponse bias on patient satisfaction surveys and identifies clinical practices to improve care delivery.

Associations between lipodystrophy or antiretroviral medications and cirrhosis in patients with HIV infection or HIV/HCV coinfection.

BACKGROUND: Many HIV antiretroviral medications have been associated with chronic liver injury. HIV-infected patients frequently develop HIV and highly active antiretroviral treatment-associated lipodystrophy syndrome (HALS), characterized by accumulation of intra-abdominal fat, insulin resistance, and hepatic steatosis. We sought to determine whether long-term exposure to specific antiretroviral medications or the presence of HALS predispose HIV-infected patients to the development of cirrhosis. METHODS: HIV-infected patients with cirrhosis who received care in the Veterans Affairs Healthcare System nationally in 2009 were matched by hepatitis C virus (HCV) coinfection status and year of first visit for HIV to the Veterans Affairs Healthcare System with HIV-infected patients without cirrhosis in a 1 : 3 ratio. RESULTS: Among HIV/HCV coinfected patients (593 with cirrhosis and 1591 matched controls), HALS was associated with a significantly increased risk for cirrhosis (adjusted odds ratio 1.6, 95% confidence interval 1.1-2.3), especially among Black patients (adjusted odds ratio 2.9, 95% confidence interval 1.6-5.2). In addition, among HIV/HCV coinfected patients, longer cumulative exposures to all antiretroviral medications, all nucleoside reverse transcriptase inhibitors, all protease inhibitors, and selected individual medications (didanosine, stavudine, and nelfinavir) were found to be significantly associated with cirrhosis. In contrast, among HIV-infected patients not coinfected with HCV (245 with cirrhosis and 658 matched controls), HALS or exposure to antiretroviral medications was found not to be significantly associated with cirrhosis, with the exception of didanosine. CONCLUSION: HALS and cumulative exposure to nucleoside reverse transcriptase inhibitors and protease inhibitors, especially stavudine, didanosine, and nelfinavir, were found to be associated with the development of cirrhosis in HIV/HCV coinfected patients, but not in HIV-monoinfected patients.

Factors affecting medication adherence: patient perspectives from five veterans affairs facilities.

BACKGROUND: In the United States, more than 25 million people have diabetes. Medication adherence is known to be important for disease control. However, factors that consistently predict medication adherence are unclear and the literature lacks patient perspectives on how health care systems affect adherence to oral hypoglycemic agents (OHAs). This study explored facilitators and barriers to OHA adherence by obtaining the perspectives of Veterans Affairs (VA) patients with OHA prescriptions. METHODS: A total of 45 patients participated in 12 focus groups that explored a wide range of issues that might affect medication adherence. Participants were patients at clinics in Seattle, Washington; San Antonio, Texas; Portland, Oregon; Salem, Oregon, and Warrenton, Oregon. RESULTS: Key system-level facilitators of OHA adherence included good overall pharmacy service and several specific mechanisms for ordering and delivering medications (automated phone refill service, Web-based prescription ordering), as well as providing pillboxes and printed lists of current medications to patients. Barriers mirrored many of the facilitators. Poor pharmacy service quality and difficulty coordinating multiple prescriptions emerged as key barriers. CONCLUSIONS: VA patient focus groups provided insights on how care delivery systems can encourage diabetes medication adherence by minimizing the barriers and enhancing the facilitators at both the patient and system levels. Major system-level factors that facilitated adherence were overall pharmacy service quality, availability of multiple systems for reordering medications, having a person to call when questions arose, counseling about the importance of adherence and providing tools such as pillboxes and updated medication lists.

Massachusetts health reform and Veterans Affairs health system enrollment.

OBJECTIVES: Veterans Health Administration (VA) operates the largest integrated health system in the nation. The Affordable Care Act (ACA) does not require any changes to VA, but the individual mandate and expanded health insurance options may change veterans' preferences for coverage. We examined the impact of healthcare reform in Massachusetts, which also included these policy changes, on veterans' enrollment in VA, private insurance, and Medicaid. STUDY DESIGN: Massachusetts' healthcare reform in June 2006 served as a natural experiment. Using data from the 2004-2013 Current Population Surveys, we examined enrollment in VA, private insurance, and Medicaid, comparing veterans residing in Massachusetts with veterans residing in neighboring New England states that did not undergo health reform. METHODS: We estimated the probability of being enrolled in VA, private insurance, and Medicaid before and after healthcare reform, using multivariate probit models while adjusting for individual characteristics. Using a difference-in-difference approach, we compared pre-post changes in enrollment probability among Massachusetts and non-Massachusetts veterans, respectively. RESULTS: Compared with other New England veterans, Massachusetts veterans decreased their enrollment in VA and private insurance by 0.2 (P = .857) and 0.9 (P = .666) percentage points, respectively, following health reform. In contrast, Medicaid enrollment increased by 2.5 percentage points (P = .038). CONCLUSIONS: Healthcare reform in Massachusetts was associated with greater Medicaid enrollment, but was not significantly associated with VA and private insurance enrollment. Our results are significant for informing VA fiscal planning in the post ACA era.

Elements of the patient-centered medical home associated with health outcomes among veterans: the role of primary care continuity, expanded access, and care coordination.

Care continuity, access, and coordination are important features of the patient-centered medical home model and have been emphasized in the Veterans Health Administration patient-centered medical home implementation, called the Patient Aligned Care Team. Data from more than 4.3 million Veterans were used to assess the relationship between these attributes of Patient Aligned Care Team and Veterans Health Administration hospitalization and mortality. Controlling for demographics and comorbidity, we found that continuity with a primary care provider was associated with a lower likelihood of hospitalization and mortality among a large population of Veterans receiving VA primary care.

Thiazolidinediones and the risk of asthma exacerbation among patients with diabetes: a cohort study.

BACKGROUND: Thiazolidinediones are oral diabetes medications that selectively activate peroxisome proliferator-activated receptor gamma and have potent anti-inflammatory properties. While a few studies have found improvements in pulmonary function with exposure to thiazolidinediones, there are no studies of their impact on asthma exacerbations. Our objective was to assess whether exposure to thiazolidinediones was associated with a decreased risk of asthma exacerbation. METHODS: We performed a cohort study of diabetic Veterans who had a diagnosis of asthma and were taking oral diabetes medications during the period of 10/1/2005 - 9/30/2006. The risk of asthma exacerbations and oral steroid use during 10/1/2006 - 9/30/2007 was compared between patients who were prescribed thiazolidinediones and patients who were on alternative oral diabetes medications. Multivariable logistic regression and negative binomial regression analyses were used to characterize this risk. A sensitivity analysis was performed, restricting our evaluation to patients who were adherent to diabetes therapy. RESULTS: We identified 2,178 patients who were on thiazolidinediones and 10,700 who were not. Exposure to thiazolidinediones was associated with significant reductions in the risk of asthma exacerbation (OR = 0.79, 95% CI, 0.62 - 0.99) and oral steroid prescription (OR = 0.73, 95% CI 0.63 - 0.84). Among patients who were adherent to diabetes medications, there were more substantial reductions in the risks for asthma exacerbation (OR = 0.64, 95% CI 0.47 - 0.85) and oral steroid prescription (OR = 0.68, 95% CI 0.57 - 0.81). CONCLUSIONS: Thiazolidinediones may provide a novel anti-inflammatory approach to asthma management by preventing exacerbations and decreasing the use of oral steroids.

Adoption of transradial percutaneous coronary intervention and outcomes according to center radial volume in the Veterans Affairs Healthcare system: insights from the Veterans Affairs clinical assessment, reporting, and tracking (CART) program.

BACKGROUND: Studies examining the association between radial approach and post-percutaneous coronary intervention (PCI) bleeding and mortality have reached conflicting conclusions. There are no current data about the use and outcomes of transradial PCI (r-PCI) in the Veterans Affairs system. METHODS AND RESULTS: Consecutive veterans (n=24143 patients) undergoing PCI in the Veterans Affairs between 2007 and 2010 were examined. On the basis of propensity to undergo r-PCI, 3 cohorts matched with veterans undergoing transfemoral access were constructed among sites performing ≥ 1 r-PCI, ≥ 50 r-PCI (high volume), and <50 r-PCI (low volume). Cox proportional hazard models were used to determine the association between PCI access site, blood transfusion, and mortality. The prevalence of r-PCI increased over time (2007=2.1%; 2010=8.8%). Overall, there was no difference in procedure success between matched groups (r-PCI 97.3% versus transfemoral PCI 96.6%; P=0.182), or in the risk of postprocedure transfusion or mortality. Among matched patients treated at high r-PCI volume sites, radial access was associated with a decreased risk of post-PCI blood transfusion (hazard ratio, 0.4; 95% confidence interval, 0.3-0.7; P<0.001), and no significant difference in the risk of mortality (hazard ratio, 0.7; 95% confidence interval, 0.4-1.3; P=0.279). CONCLUSIONS: Within the Veterans Affairs, the use of r-PCI increased over time. r-PCI may be associated with a significant decreased risk of post-PCI blood transfusion among higher volume r-PCI sites. These data demonstrate that potential benefits of r-PCI in terms of reduced post-PCI blood transfusions may be more pronounced at sites that routinely use radial access.

The patient-centered medical home in the Veterans Health Administration.

BACKGROUND: The Veterans Health Administration (VHA) is the largest integrated US health system to implement the patient-centered medical home. The Patient Aligned Care Team (PACT) initiative (implemented 2010-2014) aims to achieve team based care, improved access, and care management for more than 5 million primary care patients nationwide. OBJECTIVES: To describe PACT and evaluate interim changes in PACT-related care processes. STUDY DESIGN: Data from the VHA Corporate Data Warehouse were obtained from April 2009 (pre- PACT) to September 2012. All patients assigned to a primary care provider (PCP) at all VHA facilities were included. METHODS: Nonparametric tests of trend across time points. RESULTS: VHA increased primary care staff levels from April 2010 to December 2011 (2.3 to 3.0 staff per PCP full-time equivalent). In-person PCP visit rates slightly decreased from April 2009 to April 2012 (53 to 43 per 100 patients per calendar quarter; P < .01), while in-person nurse encounter rates remained steady. Large increases were seen in phone encounters (2.7 to 28.8 per 100 patients per quarter; P < .01), enhanced personal health record use (3% to 13% of patients enrolled), and electronic messaging to providers (0.01% to 2.3% of patients per quarter). Post hospitalization follow-up improved (6.6% to 61% of VA hospital discharges), but home telemonitoring (0.8% to 1.4% of patients) and group visits (0.2 to 0.65 per 100 patients per quarter; P < .01) grew slowly. CONCLUSIONS: Thirty months into PACT, primary care staff levels and phone and electronic encounters have greatly increased; other changes have been positive but slower.

A greater analgesia, sedation, delirium order set quality score is associated with a decreased duration of mechanical ventilation in cardiovascular surgery patients.

OBJECTIVE: Protocols and order sets for the delivery of analgesia, sedation, and delirium care of the critically ill, mechanically ventilated patient have been shown to improve outcomes but are not uniform in hospitals across geographic areas. The extent to which greater order set quality is associated with improved patient outcomes is not known. We hypothesized that cardiac surgery patients cared for at hospitals with a greater analgesia, sedation, and delirium order set quality score (more guideline-concordant order sets) would have a shorter average duration of mechanical ventilation. DESIGN: Retrospective cohort study. SETTING: All Washington State non-federal hospitals providing cardiac surgery. PATIENTS: All mechanically ventilated cardiac surgery patients from January 1, 2008, until September 30, 2011. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: We created a multivariable linear regression model to assess the relationship between a hospital's pain, agitation and delirium order set quality, as assessed by an expert-validated order set quality score, and the average duration of mechanical ventilation of its cardiac surgery patients, independent of other hospital and patient factors. A total of 19,561 patients underwent cardiac surgery at 16 Washington state hospitals during the study period. The order set quality scores ranged from 4 to 19 with a mean of 11.8 ± 4.5. The mean duration of mechanical ventilation was 27.0 ± 196.6 hours. In the multivariable model, independent of other patient and hospital factors, a 1-point increase in the order set quality score was associated with a 3.3 ± 0.9 hour (p < 0.01) decrease in average duration of mechanical ventilation. CONCLUSIONS: Cardiac surgery hospitals with more guideline-adherent analgesia, sedation, and delirium order sets have patients with shorter mean durations of mechanical ventilation than hospitals with lower order set quality scores.

Patterns of preoperative consultation and surgical specialty in an integrated healthcare system.

BACKGROUND: Many patients scheduled for elective surgery are referred for a preoperative medical consultation. Only limited data are available on factors associated with preoperative consultations. The authors hypothesized that surgical specialty contributes to variation in referrals for preoperative consultations. METHODS: This is a cohort study using data from Group Health Cooperative, an integrated healthcare system. The authors included 13,673 patients undergoing a variety of common procedures-primarily low-risk surgeries-representing six surgical specialties, in 2005-2006. The authors identified consultations by family physicians, general internists, pulmonologists, or cardiologists in the 42 days preceding surgery. Multivariable logistic regression was used to estimate the association between surgical specialty and consultation, adjusting for potential confounders including the revised cardiac risk index, age, gender, Deyo comorbidity index, number of prescription medications, and 11 medication classes. RESULTS: The authors found that 3,063 (22%) of all patients had preoperative consultations, with significant variation by surgical specialty. Patients having ophthalmologic, orthopedic, or urologic surgery were more likely to have consultations compared with those having general surgery-adjusted odds ratios (95% CI) of 3.8 (3.3-4.2), 1.5 (1.3-1.7), and 2.3 (1.8-2.8), respectively. Preoperative consultations were more common in patients with lower revised cardiac risk scores. CONCLUSION: There is substantial practice variation among surgical specialties with regard to the use of preoperative consultations in this integrated healthcare system. Given the large number of consultations provided for patients with low cardiac risk and for patients presenting for low-risk surgeries, their indications, the financial burden, and cost-effectiveness of consultations deserve further study.

Predicting risk of hospitalization or death among patients receiving primary care in the Veterans Health Administration.

BACKGROUND: Statistical models that identify patients at elevated risk of death or hospitalization have focused on population subsets, such as those with a specific clinical condition or hospitalized patients. Most models have limitations for clinical use. Our objective was to develop models that identified high-risk primary care patients. METHODS: Using the Primary Care Management Module in the Veterans Health Administration (VHA)'s Corporate Data Warehouse, we identified all patients who were enrolled and assigned to a VHA primary care provider on October 1, 2010. The outcome variable was the occurrence of hospitalization or death during the subsequent 90 days and 1 year. We extracted predictors from 6 categories: sociodemographics, medical conditions, vital signs, prior year use of health services, medications, and laboratory tests and then constructed multinomial logistic regression models to predict outcomes for over 4.6 million patients. RESULTS: In the predicted 95th risk percentiles, observed 90-day event rates were 19.6%, 6.2%, and 22.6%, respectively, for hospitalization, death, and either hospitalization or death, compared with population averages of 2.7%, 0.7%, and 3.4%, respectively; 1-year event rates were 42.3%, 19.4%, and 51.3%, respectively, compared with population averages of 8.2%, 2.6%, and 10.8%, respectively. The C-statistics for 90-day outcomes were 0.83, 0.86, and 0.81, respectively, for hospitalization, death, and either hospitalization or death and were 0.81, 0.85, and 0.79, respectively, for 1-year outcomes. CONCLUSIONS: Prediction models using electronic clinical data accurately identified patients with elevated risk for hospitalization or death. This information can enhance the coordination of care for patients with complex clinical conditions.

The prevalence of cirrhosis and hepatocellular carcinoma in patients with human immunodeficiency virus infection.

UNLABELLED: Cirrhosis is a leading cause of death among patients infected with human immunodeficiency virus (HIV). We sought to determine risk factors for and time trends in the prevalence of cirrhosis, decompensated cirrhosis, and hepatocellular carcinoma (HCC) among patients diagnosed with HIV who received care in the Veterans Affairs (VA) health care system nationally between 1996 and 2009 (n = 24,040 in 2009). Among patients coinfected with HIV and hepatitis C virus (HCV), there was a dramatic increase in the prevalence of cirrhosis (3.5%-13.2%), decompensated cirrhosis (1.9%-5.8%), and HCC (0.07%-1.6%). Little increase was observed among patients without HCV coinfection in the prevalence of cirrhosis (1.7%-2.2%), decompensated cirrhosis (1.1%-1.2%), and HCC (0.03%-0.13%). In 2009, HCV infection was present in the majority of patients with HIV who had cirrhosis (66%), decompensated cirrhosis (62%), and HCC (80%). Independent risk factors for cirrhosis included HCV infection (adjusted odds ratio [AOR], 5.82; 95% confidence interval [CI], 5.0-6.7), hepatitis B virus (HBV) infection (AOR, 2.40; 95% CI, 2.0-2.9), age (AOR, 1.03; 95% CI, 1.02-1.04), Hispanic ethnicity (AOR, 1.76; 95% CI, 1.4-2.2), diabetes (AOR, 1.79; 95% CI, 1.6-2.1), and alcohol abuse (AOR, 1.78; 95% CI, 1.5-2.1), whereas black race (AOR, 0.56; 95% CI, 0.48-0.64) and successful eradication of HCV (AOR, 0.61; 95% CI, 0.4-0.9) were protective. Independent risk factors for HCC included HCV infection (AOR, 10.0; 95% CI, 6.1-16.4), HBV infection (AOR, 2.82; 95% CI, 1.7-4.7), age (AOR, 1.05; 95% CI, 1.03-1.08), and low CD4+ cell count (AOR, 2.36; 95% CI, 1.3-4.2). Among 5999 HIV/HCV-coinfected patients, 994 (18%) had ever received HCV antiviral treatment, of whom 165 (17%) achieved sustained virologic response. CONCLUSION: The prevalence of cirrhosis and HCC has increased dramatically among HIV-infected patients driven primarily by the HCV epidemic. Potentially modifiable risk factors include HCV infection, HBV infection, diabetes, alcohol abuse, and low CD4+ cell count.

Initial validation of a self-report measure of the extent of and reasons for medication nonadherence.

BACKGROUND: Self-report measures of medication nonadherence confound the extent of and reasons for medication nonadherence. Each construct is assessed with a different type of psychometric model, which dictates how to establish reliability and validity. OBJECTIVES: To evaluate the psychometric properties of a self-report measure of medication nonadherence that assesses separately the extent of nonadherence and reasons for nonadherence. RESEARCH DESIGN: Cross-sectional survey involving the new measure and comparison measures to establish convergent, discriminant, and predictive validity. The new measure was readministered 2-21 days later. SUBJECTS: A total of 202 veterans with treated hypertension were recruited from the Durham Veterans Affairs Medical Center. MEASURES: A new self-report measure assessed the extent of nonadherence and reasons for nonadherence. Comparison measures included self-reported medication self-efficacy, beliefs about medications, impression management, conscientiousness, habit strength, and an existing nonadherence measure. RESULTS: Three items assessing the extent of nonadherence produced reliable scores for this sample, α = 0.84 (95% confidence interval, 0.80-0.87). Correlations with comparison measures provided evidence of convergent and discriminant validity. Correlations with systolic ( r = 0.27, P < 0.0001) and diastolic (r = 0.27, P < 0.0001) blood pressure provided evidence of predictive validity. Reasons for nonadherence were assessed with 21 independent items. Intraclass correlations were 0.58 for the extent score and ranged from 0.07 to 0.64 for the reasons. CONCLUSIONS: The dual conceptualization of medication nonadherence allowed a stronger evaluation of the reliability and validity than was previously possible with measures that confounded these 2 constructs. Measurement of self-reported nonadherence consistent with psychometric principles will enable reliable, valid evaluation of interventions to reduce nonadherence.

Predicting risk of hospitalization or death among patients with heart failure in the veterans health administration.

Patients with heart failure (HF) are at high risk of hospitalization or death. The objective of this study was to develop prediction models to identify patients with HF at highest risk for hospitalization or death. Using clinical and administrative databases, we identified 198,460 patients who received care from the Veterans Health Administration and had ≥1 primary or secondary diagnosis of HF that occurred within 1 year before June 1, 2009. We then tracked their outcomes of hospitalization and death during the subsequent 30 days and 1 year. Predictor variables chosen from 6 clinically relevant categories of sociodemographics, medical conditions, vital signs, use of health services, laboratory tests, and medications were used in multinomial regression models to predict outcomes of hospitalization and death. In patients who were in the ≥95th predicted risk percentile, observed event rates of hospitalization or death within 30 days and 1 year were 27% and 80% respectively, compared to population averages of 5% and 31%, respectively. The c-statistics for the 30-day outcomes were 0.82, 0.80, and 0.80 for hospitalization, death, and hospitalization or death, respectively, and 0.82, 0.76, and 0.77, respectively, for 1-year outcomes. In conclusion, prediction models using electronic health records can accurately identify patients who are at highest risk for hospitalization or death. This information can be used to assist care managers in selecting patients for interventions to decrease their risk of hospitalization or death.

The incidence and timing of noncardiac surgery after cardiac stent implantation.

BACKGROUND: In November 2007, national guidelines were released recommending delay of elective noncardiac surgery after cardiac revascularization with drug eluting stents (DES) for 12 months compared with 6 weeks after implantation of bare metal stents (BMS). STUDY DESIGN: To determine the incidence of noncardiac surgery within 24 months after stent placement, national VA data on cardiac stent implantation were merged with data from the VA National Surgery Office and Medicare. Using chi-square tests and log-rank analyses, we measured the incidence of noncardiac surgery after BMS and DES in relation to guideline release and surgical characteristics. RESULTS: From 2000 to 2010, 126,773 stent procedures were followed by 25,977 (20.5%) noncardiac operations within 24 months. Overall, 11.8% of the BMS surgery cohort had early surgery (less than 6 weeks) compared with 46.7% of the DES surgery cohort, which had early surgery (less than 12 months). The incidence of surgery differed significantly by stent type (BMS 24.1% vs DES 17.5%, p < 0.001) and in relation to guideline release (pre- 24.6% vs postguideline 13.1%, p < 0.001). Higher complexity operations (work relative value units) were more likely to occur in the early period for both BMS (p < 0.0001) and DES (p < 0.003). After guideline release, the incidence of surgery within 12 months decreased from 16.7% to 10.0% (p < 0.0001). CONCLUSIONS: We found evidence that guidelines recommending delaying surgery appear to be effective in decreasing the incidence of early surgery; however, early surgery is still a frequent occurrence. Additional research is needed to better define the risk of cardiac and bleeding complications in patients undergoing subsequent noncardiac surgery, as well as the optimal time for surgery and method of antiplatelet management.