[Show abstract][Hide abstract] ABSTRACT: Positive continuous outcomes with a point mass at zero are prevalent in
biomedical research. To model the point mass at zero and to provide
marginalized covariate effect estimates, marginalized two part models (MTP)
have been developed for outcomes with lognormal and log skew normal
distributions. In this paper, we propose MTP models for outcomes from a
generalized gamma (GG) family of distributions. In the proposed MTP-GG model,
the conditional mean from a two-part model with a three-parameter GG
distribution is parameterized to provide regression coefficients that have
marginal interpretation. MTP-gamma and MTP-Weibull are developed as special
cases of MTP-GG. We derive marginal covariate effect estimators from each model
and through simulations assess their finite sample operating characteristics in
terms of bias, standard errors, 95% coverage, and rate of convergence. We
illustrate the models using data sets from The Medical Expenditure Survey
(MEPS) and from a randomized trial of addictive disorders and we provide SAS
code for implementation. The simulation results show that when the response
distribution is unknown or mis-specified, which is usually the case in real
data sets, the MTP-GG is preferable over other models.
[Show abstract][Hide abstract] ABSTRACT: Rationale:
Smoking is the largest contributor to lung cancer risk, those who continue to smoke post-diagnosis have a worse survival. Screening for lung cancer with low-dose computed tomography (LDCT) reduces mortality in high-risk individuals. Smoking cessation is an essential component of a high-quality screening program.
To quantify the effects of smoking history and abstinence on mortality in high-risk individuals who participated in the National Lung Screening Trial (NLST).
This is a secondary analysis of a randomized controlled trial (NLST).
Measurements included self-reported demographics, medical and smoking history, lung cancer-specific and all-cause mortality. Cox regression was used to study the association of mortality with smoking status and pack-years. Kaplan-Meier survival curves were examined for differences in survival based on trial arm and smoking status.
Current smokers had an increased lung cancer-specific (HR range 2.14-2.29) and all-cause mortality (HR range 1.79-1.85) compared to former smokers irrespective of screening arm. Former smokers in the control arm abstinent for seven years had a 20% mortality reduction comparable to the benefit reported with LDCT screening in the NLST. The maximum benefit was seen with the combination of smoking abstinence at 15 years and LDCT screening, which resulted in a 38% reduction in lung cancer-specific mortality (HR 0.62, 95% CI 0.51 - 0.76).
Seven years of smoking abstinence reduced lung cancer-specific mortality at a magnitude comparable to LDCT screening. This reduction was greater when abstinence was combined with screening highlighting the importance of smoking cessation efforts in screening programs.
American Journal of Respiratory and Critical Care Medicine 10/2015; DOI:10.1164/rccm.201507-1420OC · 13.00 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Introduction:
Adverse childhood experiences (ACEs) are associated with early mortality and morbidity. This study evaluated the association among ACEs, high-risk health behaviors, and comorbid conditions, as well as the independent effect of ACE components.
Data were analyzed on 48,526 U.S. adults from five states in the 2011 Behavioral Risk Factor Surveillance System survey. Exposures included psychological, physical, and sexual forms of abuse as well as household dysfunction such as substance abuse, mental illness, violence, and incarceration. Main outcome measures included risky behaviors and morbidity measures, including binge drinking, heavy drinking, current smoking, high-risk HIV behavior, obesity, diabetes, myocardial infarction, coronary heart disease, stroke, depression, disability caused by poor health, and use of special equipment because of disability. Multiple logistic regression assessed the independent relationship between ACE score categories and risky behaviors/comorbidities in adulthood, and assessed the independent relationship between individual ACE components and risky behaviors/comorbid conditions in adulthood controlling for covariates.
A total of 55.4% of respondents reported at least one ACE and 13.7% reported four or more ACEs. An ACE score ≥4 was associated with increased odds for binge drinking, heavy drinking, smoking, risky HIV behavior, diabetes, myocardial infarction, coronary heart disease, stroke, depression, disability caused by health, and use of special equipment because of disability. In addition, the individual components had different effects on risky behavior and comorbidities.
In addition to having a cumulative effect, individual ACE components have differential relationships with risky behaviors, morbidity, and disability in adulthood after controlling for important confounders.
American journal of preventive medicine 10/2015; DOI:10.1016/j.amepre.2015.07.022 · 4.53 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Background:
Socioeconomic status (SES) is a significant determinant of health outcomes and may be an important component of the causal chain surrounding racial disparities in kidney transplantation. The social adaptability index (SAI) is a validated and quantifiable measure of SES, with a lack of studies analyzing this measure longitudinally or between races.
Longitudinal cohort study in adult kidney transplantation transplanted at a single-center between 2005 and 2012. The SAI score includes 5 domains (employment, education, marital status, substance abuse and income), each with a minimum of 0 and maximum of 3 for an aggregate of 0 to 15 (higher score → better SES).
One thousand one hundred seventy-one patients were included; 624 (53%) were African American (AA) and 547 were non-AA. African Americans had significantly lower mean baseline SAI scores (AAs 6.5 vs non-AAs 7.8; P < 0.001). Cox regression analysis demonstrated that there was no association between baseline SAI and acute rejection in non-AAs (hazard ratio [HR], 0.92; 95% confidence interval [95% CI], 0.81-1.05), whereas it was a significant predictor of acute rejection in AAs (HR, 0.89; 95% CI, 0.80-0.99). Similarly, a 2-stage approach to joint modelling of time to graft loss and longitudinal SAI did not predict graft loss in non-AAs (HR, 1.01; 95% CI, 0.28-3.62), whereas it was a significant predictor of graft loss in AAs (HR, 0.23; 95% CI, 0.06-0.93).
After controlling for confounders, SAI scores were associated with a lower risk of acute rejection and graft loss in AA kidney transplant recipients, whereas neither baseline nor follow-up SAI predicted outcomes in non-AA kidney transplant recipients.
[Show abstract][Hide abstract] ABSTRACT: Background
Studies have shown that community and neighborhood characteristics can impact health outcomes of those with chronic illness, including T2DM. Factors, such as crime, violence, and lack of resources have been shown to be barriers to optimal health outcomes in diabetes. Thus, the objective of this study is to assess the effects of neighborhood factors on diabetes-related health outcomes and self-care behaviors.
Adult patients (N = 615) with type 2 diabetes mellitus (T2DM) were recruited from an academic medical center and a Veterans Affairs medical center in the southeastern United States. Validated scales and indices were used to assess neighborhood factors and diabetes-related self-care behaviors. The most recent HbA1c, blood pressure, and LDL cholesterol were abstracted from each patients’ electronic medical record.
In the fully adjusted model, significant associations were between neighborhood aesthetics and diabetes knowledge (β = 0.141) and general diet (β = -0.093); neighborhood comparison and diabetes knowledge (β = 0.452); neighborhood activities and general diet (β = -0.072), exercise (β = -0.104), and foot care (β = -0.114); food insecurity and medication adherence (β = -0.147), general diet (β = -0.125), and blood sugar testing (β = -0.172); and social support and medication adherence (β = 0.009), foot care (β = 0.010), and general diet (β = 0.016). Significant associations were also found between neighborhood violence and LDL Cholesterol (β = 4.04), walking environment and exercise (β = -0.040), and social cohesion and HbA1c (β = -0.086).
We found that neighborhood violence, aesthetics, walking environment, activities, food insecurity, neighborhood comparison, social cohesion and social support have statistically significant associations with self-care behaviors and outcomes to varying degrees. However, the key neighborhood factors that had independent associations with multiple self-care behaviors and outcomes were food insecurity, neighborhood activities and social support.
This study suggests that food insecurity, neighborhood activities, aesthetics, and social support may be important targets for interventions in individuals with T2DM.
Electronic supplementary material
The online version of this article (doi:10.1186/s12913-015-1086-7) contains supplementary material, which is available to authorized users.
BMC Health Services Research 10/2015; 15(1). DOI:10.1186/s12913-015-1086-7 · 1.71 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Aims:
Socioeconomic, psychosocial, and neighborhood factors influence clinical outcomes and self-care behaviors in diabetes; however, few studies simultaneously assessed the impact of multiple social determinant of health factors on glycemic control. We used an explanatory model to examine the differential contribution of social determinants and clinical factors on glycemic control. Secondarily, we examined the contribution of mutable and immutable factors to identify meaningful future interventions.
Six hundred and fifteen adults with type 2 diabetes in the southeastern United States were recruited. A hierarchical model was run with HbA1c as the dependent variable and independent variables entered in blocks: demographics (block 1), socioeconomic (block 2), psychosocial (block 3), built environment (block 4), clinical (block 5), and knowledge/self-care (block 6).
Significant associations for HbA1c included self-efficacy (β=-0.10, p<0.001), social support (β=0.01, p<0.05), comorbidity (β=-0.09, p<0.05), insulin use (β=0.95, p<0.001), medication adherence (β=-0.11, p<0.05), and being a former smoker (β=0.34, p<0.05); accounting for 24.4% of the variance.
Important factors that drive glycemic control are mutable, and amenable to health interventions. Greater attention should be given to interventions that increase self-efficacy and social support, reduce the burden of comorbidities, and enhance medication adherence and smoking cessation.
Diabetes research and clinical practice 09/2015; DOI:10.1016/j.diabres.2015.09.007 · 2.54 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: This study assessed the prevalence and specific costs associated with discrete multimorbid mental health disease clusters in adults with diabetes mellitus (DM).
Longitudinal analysis of a retrospective cohort.
We performed a 5-year longitudinal analysis of a retrospective cohort of 733,071 patients with DM from the US Veterans Health Administration (VHA) between 2002 and 2006. The mental health comorbidities (MHCs) examined included depression, substance abuse, and psychosis. Our primary outcomes of interest were total inpatient, outpatient, and pharmacy costs measured in 2012 US$ from the perspective of the VHA.
DM was present with comorbid depression, substance abuse, and psychosis in 12.1%, 3.7%, and 4.2% of patients, respectively. Overall, 13.5% of patients had 1 MHC, 2.5% had 2 MHCs, and 0.5% had all MHCs. Total inpatient ($1,435,651,415), outpatient ($366,137,435), and pharmacy ($90,064,725) costs were highest for patients with DM and comorbid depression alone. At the per-patient level, DM plus psychosis and substance abuse had the highest inpatient costs ($35,518), DM plus all MHCs had the highest outpatient costs ($6962), and DM plus depression and psychosis had the highest pharmacy costs ($1753).
DM with comorbid depression is the most prevalent MHC combination and is associated with the highest total VHA healthcare costs. However, other comorbidity clusters are associated with higher mean per patient costs, and may therefore benefit from more intensive intervention. Analysis of healthcare expenditures by multimorbid disease clusters can be a useful tool for healthcare policy planning.
The American journal of managed care 08/2015; 21(8):535-544. · 2.26 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Objective:
Primary: To examine Veterans Administration (VA) utilization and other potential mediators between racial/ethnic differentials and mortality in veterans diagnosed with traumatic brain injury (TBI).
A national cohort of veterans clinically diagnosed with TBI in 2006 was followed from January 1, 2006 through December 31, 2009 or until date of death. Utilization was tracked for 12 months. Differences in survival and potential mediators by race were examined via K-Wallis and chi-square tests. Potential mediation of utilization in the association between mortality and race/ethnicity was studied by fitting Cox models with and without adjustment for demographics and co-morbidities. Poisson regression was used to study the association of race/ethnicity with utilization of specialty services potentially important in the management of TBI.
United States (US) Veterans Administration (VA) Hospitals and Clinics.
14, 690 US veterans clinically diagnosed with TBI in 2006.
Not Applicable. The study is a secondary data analysis.
Main outcome measures:
Hispanic veterans were found to have significantly higher unadjusted mortality (6.69%) than Non-Hispanic White veterans (2.93%). Hispanic veterans relative to Non-Hispanic White were found to have significantly lower utilization of all services examined, except imaging. Neurology was found to be the utilization mediator with the highest percent of excess risk (3.40%) while age was the non utilization confounder with the highest percent of excess risk (31.49%). In fully adjusted models for demographics and co-morbidities, Hispanic veterans relative to Non-Hispanic Whites were found to have less total visits (IRR 0.89), TBI clinic (IRR 0.43), neurology (IRR 0.35), rehabilitation (IRR 0.37), and other visits (IRR 0.85) with only higher mental health visits (IRR 1.53).
We found evidence that utilization is a partial mediator between race/ethnicity and mortality, especially neurology utilization. We also found that Hispanic veterans receive significantly less TBI clinic, neurology, rehabilitation and other types of utilization. The use of innovative system factors (decision aids, information tools, patient activation, and adherence support interventions) could be valuable in enhancing utilization of specific TBI related services, especially among ethnic minorities.
Global journal of health science 07/2015; 8(2). DOI:10.5539/gjhs.v8n2p260
[Show abstract][Hide abstract] ABSTRACT: Perceptions of control impact outcomes in veterans with chronic disease. The purpose of this study was to examine the association between control orientation and clinical and quality of life (QOL) outcomes in male veterans with type 2 diabetes (T2DM). Cross-sectional study of 283 male veterans from a primary care clinic in the southeastern US. Health locus of control (LOC) was the main predictor and assessed using the Multidimensional Health LOC Scale. Clinical outcomes were glycosylated hemoglobin A1c (HbA1c), systolic and diastolic blood pressure, and low-density lipoprotein cholesterol (LDL-C). Physical (PCS) and mental (MCS) health component scores for QOL were assessed using the Veterans RAND 12-Item Health Survey. Unadjusted and adjusted multivariate analyses were performed to assess associations between LOC and outcomes. Unadjusted analyses showed internal LOC associated with HbA1c (β = 0.036; 95 % CI 0.001, 0.071), external LOC:powerful others inversely associated with LDL-C (β = -0.794; 95 % CI -1.483, -0.104), and external LOC:chance inversely associated with MCS QOL (β = -0.418; 95 % CI -0.859, -0.173). These associations remained significant when adjusting for relevant covariates. Adjusted analyses also demonstrated a significant relationship between external LOC:chance and PCS QOL (β = 0.308; 95 % CI 0.002, 0.614). In this sample of male veterans with T2DM, internal LOC was significantly associated with glycemic control, and external was significantly associated with QOL and LDL-C, when adjusting for relevant covariates. Assessments of control orientation should be performed to understand the perceptions of patients, thus better equipping physicians with information to maximize care opportunities for veterans with T2DM.
[Show abstract][Hide abstract] ABSTRACT: We examined the association between traumatic brain injury (TBI) severity and combat exposure by race/ethnicity.
We estimated logit models of the fully adjusted association of combat exposure with TBI severity in separate race/ethnicity models for a national cohort of 132 995 veterans with TBI between 2004 and 2010.
Of veterans with TBI, 25.8% had served in a combat zone. Mild TBI increased from 11.5% to 40.3%, whereas moderate or severe TBI decreased from 88.5% to 59.7%. Moderate or severe TBI was higher in non-Hispanic Blacks (80.0%) and Hispanics (89.4%) than in non-Hispanic Whites (71.9%). In the fully adjusted all-race/ethnicity model, non-Hispanic Blacks (1.44; 95% confidence interval [CI] = 1.37, 1.52) and Hispanics (1.47; 95% CI = 1.26, 1.72) had higher odds of moderate or severe TBI than did non-Hispanic Whites. However, combat exposure was associated with higher odds of mild TBI in non-Hispanic Blacks (2.48; 95% CI = 2.22, 2.76) and Hispanics (3.42; 95% CI = 1.84, 6.35) than in non-Hispanic Whites (2.17; 95% CI = 2.09, 2.26).
Research is needed to understand racial differences in the effect of combat exposure on mild TBI and on interventions to prevent TBI across severity levels. (Am J Public Health. Published online ahead of print June 11, 2015: e1-e7. doi:10.2105/AJPH.2014.302545).
American Journal of Public Health 06/2015; 105(8):e1-e7. DOI:10.2105/AJPH.2014.302545 · 4.55 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Although the national cost of missed workdays associated with diabetes has been estimated previously, we use the most recent available national data and methodology to update the individual and national estimates for the U.S population.
We identified 14,429 employed individuals ≥ 18 years of age in 2011 Medical Expenditure Panel Survey (MEPS) data. Diabetes and missed workdays were based on self-report, and cost was based on multiplying the daily wage rate for each individual by the number of missed days. Adjusted total national burden of missed workdays associated with diabetes was calculated using a novel two-part model to simultaneously estimate the association of diabetes with the number and cost of missed workdays.
The unadjusted annual mean 2011 cost of missed workdays was $277 (95 % CI 177.0-378.0) for individuals with diabetes relative to $160 (95 % CI $130-$189) for those without. The incremental cost of missed workdays associated with diabetes was $120 (95 % CI $30.7-$209.1). Based on the US population in 2011, the unadjusted national burden of missed workdays associated with diabetes was estimated to be $2.7 billion, while the fully adjusted incremental national burden was estimated to be $1.1 billion.
We provide more precise estimates of the cost burden of diabetes due to missed workdays on the U.S population. The high incremental and total cost burden of missed workdays among Americans with diabetes suggests the need for interventions to improve diabetes care management among employed individuals.
Journal of General Internal Medicine 05/2015; DOI:10.1007/s11606-015-3338-y · 3.42 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Study Objective
To determine the effect of tacrolimus trough concentrations on clinical outcomes in kidney transplantation, while assessing if African-American (AA) race modifies these associations.DesignRetrospective longitudinal cohort study of solitary adult kidney transplants.SettingLarge tertiary care transplant center.PatientsAdult solitary kidney transplant recipients (n=1078) who were AA (n=567) or non-AA (n=511).ExposureMean and regressed slope of tacrolimus trough concentrations. Subtherapeutic concentrations were lower than 8 ng/ml.Measurements and Main ResultsAA patients were 1.7 times less likely than non-AA patients to achieve therapeutic tacrolimus concentrations (8 ng/ml or higher) during the first year after kidney transplant (35% vs 21%, respectively, p<0.001). AAs not achieving therapeutic concentrations were 2.4 times more likely to have acute cellular rejection (ACR) as compared with AAs achieving therapeutic concentrations (20.8% vs 8.5%, respectively, p<0.01) and 2.5 times more likely to have antibody-mediated rejection (AMR; 8.9% vs 3.6%, respectively, p<0.01). Rates of ACR (8.3% vs 6.7%) and AMR (2.0% vs 0.9% p=0.131) were similar in non-AAs compared across tacrolimus concentration groups. Multivariate modeling confirmed these findings and demonstrated that AAs with low tacrolimus exposure experienced a mild protective effect for the development of interstitial fibrosis/tubular atrophy (IF/TA; hazard ratio [HR] 0.78, 95% confidence interval [CI] 0.47–1.32) with the opposite demonstrated in non-AAs (HR 2.2, 95% CI 0.90–5.1).Conclusion
In contradistinction to non-AAs, AAs who achieve therapeutic tacrolimus concentrations have substantially lower acute rejection rates but are at risk of developing IF/TA. These findings may reflect modifiable time-dependent racial differences in the concentration-effect relationship of tacrolimus. Achievement of therapeutic tacrolimus trough concentrations, potentially through genotyping and more aggressive dosing and monitoring, is essential to minimize the risk of acute rejection in AA kidney transplant recipients.
[Show abstract][Hide abstract] ABSTRACT: Blacks with lung cancer (LC) experience higher mortality because they present with more advanced disease and are less likely to undergo curative resection for early stage disease. The National Lung Screening Trial (NLST) demonstrated improved LC mortality by screening high-risk patients with low-dose computed tomography (LDCT). The benefit of LDCT screening in Blacks is unknown.
Examine results of the NLST by race Methods: Secondary analysis of a randomized trial (NCT00047385) performed in 33 US centers.
Overall and lung cancer specific mortality Main Results: Screening with LDCT reduced LC mortality in all racial groups but more so in Blacks (HR 0.61 vs 0.86). Smoking increased the likelihood of death from LC and when stratified by race Black smokers were twice as likely to die compared to White smokers (HR 4.10 vs 2.25). Adjusting for socio-demographic and behavioral characteristics, Blacks experienced higher all-cause mortality than Whites (HR 1.35; 95% CI, 1.22-1.49), however Blacks screened with LDCT had a reduction in all-cause mortality. Blacks were younger, more likely to be current smokers, had more co-morbidities, and fewer years of formal education than Whites (P<0.05).
Blacks screened with LDCT had decreased mortality from lung cancer. However the demographics associated with improved LC survival were less commonly found in Blacks. The overall mortality in the NLST was higher for Blacks than Whites, but improved in Blacks screened suggesting that this subgroup may have had improved access to care. To realize the reductions in mortality from LC screening, dissemination efforts need to be tailored to meet the needs of this community.
American Journal of Respiratory and Critical Care Medicine 04/2015; 192(2). DOI:10.1164/rccm.201502-0259OC · 13.00 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: To assess differences among USA ethnic groups in psychological status of adult family members (FMs) and their involvement with the diabetes of another adult.
Data are from the FM survey of the USA DAWN2 study, including 105 White non-Hispanics, 47 African Americans, 46 Hispanic Americans and 40 Chinese Americans. All FMs lived with and cared for an adult with diabetes. Analysis of covariance controlled for respondent and patient characteristics to assess ethnic group differences (P<0.05). Multiple regression analyses identified significant (P<0.05) independent correlates of psychological outcomes.
FM psychological outcomes measured include well-being, quality of life (QoL), impact of diabetes on life domains, diabetes distress, and burden.
White non-Hispanics reported less diabetes burden and distress, more negative life impact, and lower well-being than FMs from ethnic minority groups. African Americans reported the highest well-being and lowest negative life impact, Chinese Americans reported the most diabetes burden, Hispanic Americans reported the highest distress. There were no ethnic group differences in QoL. Ethnic minority FMs reported having more involvement with diabetes, greater support success, and more access to a diabetes support network than White non-Hispanics. Higher FM diabetes involvement was associated with negative psychological outcomes, while diabetes education, support success and diabetes support network size were associated with better psychological outcomes. Potential limitations are the sample sizes and representativeness.
Minority ethnic FMs experienced both advantages and disadvantages in psychological outcomes relative to each other and to White non-Hispanics. Ethnic minority FMs had more involvement in diabetes care, support success and support from others, with the first associated with worse and the latter two with better psychological outcomes. Additional studies are needed with larger samples and broader representation of ethnic groups to better understand these associations and identify areas for intervention.
Current Medical Research and Opinion 04/2015; 31(7):1-34. DOI:10.1185/03007995.2015.1043251 · 2.65 Impact Factor
[Show abstract][Hide abstract] ABSTRACT: Approximately 1 in 3 adults with diabetes have CKD. However, there are no recent national estimates of the association of CKD with medical care expenditures in individuals with diabetes. Our aim is to assess the association of CKD with total medical expenditures in US adults with diabetes using a national sample and novel cost estimation methodology.
Data on 2,053 adults with diabetes in the 2011 Medical Expenditure Panel Survey (MEPS) was analyzed. Individuals with CKD were identified based on self-report. Adjusted mean health services expenditures per person in 2011 were estimated using a two-part model after adjusting for demographic and clinical covariates.
Of the 2,053 individuals with diabetes, approximately 9.7% had self-reported CKD. Unadjusted mean expenditures for individuals with CKD were $20,726 relative to $9,689.49 for no CKD. Adjusted mean expenditures from the 2-part model for individuals with CKD were $8473 higher relative to individuals without CKD. Additional significant covariates were Hispanic/other race, uninsured, urban dwellers, CVD, stroke, high cholesterol, arthritis, and asthma. The estimated unadjusted total expenditures for individuals with CKD were estimated to be in excess of $43 billion in 2011.
We showed that CKD is a significant contributor to the financial burden among individuals with diabetes, and that minorities and the uninsured with CKD may experience barriers in access to care. Our study also provides a baseline national estimate of CKD cost in Diabetes by which future studies can be used for comparison.
Published by Elsevier Ireland Ltd.
Diabetes research and clinical practice 04/2015; 109(1). DOI:10.1016/j.diabres.2015.04.011 · 2.54 Impact Factor