Publications
Department of Medicine faculty members published more than 3,000 peer-reviewed articles in 2022.
2015
OBJECTIVE
To inform development of intensive management programs for high-cost patients, this study investigated the relationship between psychiatric diagnoses and patterns of health care utilization among high-cost patients in the Department of Veterans Affairs (VA) health care system.
METHODS
The costliest 5% of patients who received care in the VA in fiscal year 2010 were assigned to five mutually exclusive hierarchical groups on the basis of diagnosis codes: no mental health condition, serious mental illness, substance use disorder, posttraumatic stress disorder (PTSD), and depression. Multivariable linear regression was used to examine associations between diagnostic groups and use of mental health and non-mental health care and costs of care, with adjustment for sociodemographic characteristics. The proportion of costs generated by mental health care was estimated for each group.
RESULTS
Among 261,515 high-cost VA patients, rates of depression, substance use disorder, PTSD, and serious mental illness were 29%, 20%, 17%, and 13%, respectively. Individuals in the serious mental illness and substance use disorder groups were younger and had fewer chronic general medical conditions and higher adjusted rates of mental health care utilization; they also had a greater proportion of costs generated by mental health care (41% and 31%, respectively) compared with individuals in the PTSD and depression groups (18% and 11%, respectively).
CONCLUSIONS
Optimal management of high-risk, high-cost patients may require stratification by psychiatric diagnoses, with integrated care models for patients with multiple chronic conditions and comorbid mental health conditions and intensive mental health services for patients whose primary needs stem from mental health conditions.
View on PubMed2015
OBJECTIVE
There is a need to determine which response measures in lupus nephritis trials are most predictive of good long-term renal function. We used data from the Euro-Lupus Nephritis Trial to evaluate the performance of proteinuria, serum creatinine (Cr), and urinary red blood cells (RBCs) as predictors of good long-term renal outcome.
METHODS
Patients from the Euro-Lupus Nephritis Trial with proteinuria, serum Cr, and urinary RBC measurements at 3, 6, or 12 months and with a minimum of 7 years of followup were included (n = 76). We assessed the ability of these clinical biomarkers at 3, 6, and 12 months after randomization to predict good long-term renal outcome (defined as a serum Cr value ≤1.0 mg/dl) at 7 years. Receiver operating characteristic curves were generated to assess parameter performance at these time points and to select the best cutoff for individual parameters. Sensitivity and specificity were calculated for the parameters alone and in combination.
RESULTS
A proteinuria value of <0.8 gm/day at 12 months after randomization was the single best predictor of good long-term renal function (sensitivity 81% and specificity 78%). The addition of serum Cr to proteinuria as a composite predictor did not improve the performance of the outcome measure; addition of urinary RBCs as a predictor significantly decreased the sensitivity to 47%.
CONCLUSION
This study demonstrates that the level of proteinuria at 12 months is the individual best predictor of long-term renal outcome in patients with lupus nephritis. Inclusion of urinary RBCs as part of a composite outcome measure actually undermined the predictive value of the trial data. We therefore suggest that urinary RBCs should not be included as a component of clinical trial response criteria in lupus nephritis.
View on PubMed2015
2015
2015
2015