Defense Date

2007

Document Type

Thesis

Degree Name

Master of Public Health

Department

Epidemiology & Community Health

First Advisor

Dr. Jennifer Marwitz

Abstract

Objectives: Identifying reliable, practical and easy to use tools to assess depression in patients with traumatic brain injury (TBI), is a necessary first step to addressing a high incidence problem. The intention of this study is to validate depression screening measures with a criterion-based structured interview among people with TBI. The final outcome will identify which measure is the best indicator of depression diagnosis in this population. Methods: 112 participants with TBI were administered the Beck Depression Inventory (BDI), Neurobehavioral Functioning Inventory (NFI), The Hamilton Rating Scale for Depression (Ham-D), and the Structured Clinical Interview for DSM-IV TR Axis I Disorders (SCID) in an outpatient neuropsychology clinic at a university medical center, outpatient physical medicine and rehabilitation clinics, and a long-term specialized living assistance program. Screening measure results were compared with SCID results to determine sensitivity, specificity, positive predictive value and negative predictive value of each screening measure.Results: The prevalence of depression as measured by the SCID was 3 1.53%. Depression was more likely among those who were unemployed, within 12 months of their injury, experienced a length of coma between 1-6 days following injury, and had an acute care stay between 2-6 days. The BDI showed sensitivity of 86% and specificity of 79%. Calculated sensitivity and specificity for the NFI was 62% and 92%, respectively. Based on specificity and sensitivity, the Ham-D missed the most cases in this sample, showing 46% sensitivity. However, the specificity for the Ham-D was highest of the three at 99%. The BDI had a positive predictive value (PPV) of 65%, highlighting the fact that 16 of the cases that the BDI identified as depressed were not actually depressed according to the SCID. The negative predictive value (NPV) for the BDI was 92%. The Ham-D showed the highest PPV at 94%, indicating that of those that tested positive on the Ham-D, 94% were rated as depressed on the SCID. NPV for the Ham-D was 80%. The NFI showed a PPV of 78% and NPV of 84%.Conclusions: Sensitivity was considered the most important aspect to evaluate, due to the need to minimize the number of cases of depression missed. From the results of the sensitivity and specificity calculations, the BDI appears to be the most appropriate screening measure for identifying cases needing further clinical evaluation.

Rights

© The Author

Is Part Of

VCU University Archives

Is Part Of

VCU Theses and Dissertations

Date of Submission

June 2008

Included in

Epidemiology Commons

Share

COinS