Date of Award


Degree Type


Degree Name

Doctor of Philosophy in Pharmaceutical Sciences


Biomedical and Pharmaceutical Sciences

First Advisor

Brian J. Quilliam



National guidelines suggest 12-18 months of antidepressant treatment for depressed patients to maximize the benefits of treatment; however, patterns of medication use according to the guidelines are less well understood, especially in older adults who are at higher risk of adverse drug events. In addition, selective serotonin reuptake inhibitors (SSRIs) are recommended as a first-line therapy for older adults because of their favorable adverse event profile; however, several observational studies have suggested an association between SSRI use and an increased risk of gastrointestinal (GI) bleeding. While this link remains controversial, it is important to evaluate the risk of GI bleeding in association with SSRI use, especially in older adults who take multiple medications that may increase the risk of GI bleeding.


The purpose of this dissertation is to describe the patterns of antidepressant medication use by older adults and quantify the adverse events associated with antidepressant use. The third manuscript focuses on methodology to perform a systematic literature review of published articles reporting effect measure modification (EMM) and characterize the sample size considerations related to EMM.


In the first manuscript, we performed a cross-sectional study using the Medicare Current Beneficiary Survey (MCBS) data from 2004 to 2008. We estimated a 6-month discontinuation rate of antidepressant medications in older adults who initiated antidepressant treatment following diagnosis with depression. We further developed a multivariable logistic regression model to identify predictors of discontinuation of antidepressant medication.

For the second manuscript, we conducted a nested case-control study using the MCBS data from 2004-2008. We identified cases as all older adults diagnosed with upper GI bleeding. Using incidence density sampling, we randomly selected up to 6 controls who had no evidence of upper GI bleeding after matching on age (+/- 5 years), gender, calendar year, and Charlson comorbidity score. We developed a conditional logistic regression model to estimate the risk of upper GI bleeding associated with SSRI use, simultaneously adjusting for potential confounders. Furthermore, we evaluated whether use of non-steroidal anti-inflammatory drugs (NSAIDs) modified the effect of SSRI use on upper GI bleeding.

In the third manuscript, we systematically reviewed the published EMM literature by searching the PubMed articles published between January 2008 and June 2013. The included studies evaluated EMM by medication use, employed an observational study design, and published in English. We characterized these publications and calculated frequencies to summarize the percentage of studies identified, including specific sample size calculations.


We found that less than 5% of older adults in Medicare programs were diagnosed with major depression between 2006 and 2008. Nearly 1 in 2 depressed older adults were treated with antidepressant medications and 19.2% initiated medication after diagnosis. Of these new users of antidepressant medications, 30.3% discontinued medication treatment within 180 days of starting. Living in a metropolitan area was a significant predictor of antidepressant discontinuation. In addition, SSRI use did not significantly increase the risk of upper GI bleeding in older adults (Adjusted Odds Ratio [AOR]=1.3; 95% Confidence Interval [CI], 0.7-2.5). Furthermore, after adjusting for confounding factors, use of SSRIs along with NSAIDs was not significantly associated with an increased risk of upper GI bleeding (AOR=1.8; 95% CI, 0.5-6.7). Through a systematic literature review, we found that none of the studies performed sample size calculations specifically related to the EMM, although 14.3% of the identified studies mentioned sample size related issues in the study.


Older adults tend to persist in antidepressant medication use. Information obtained from this study can be used by both primary care physicians and policy administrators to improve care for older Medicare beneficiaries. Physicians and other healthcare providers can utilize the information obtained from this study to more fully evaluate the risk-benefit ratio of prescribing specific antidepressants to older adults. In addition, the descriptive information obtained in this study (e.g., treatment patterns, average length of treatment) can provide points of discussion for physicians and other healthcare providers when they are working with older adults regarding barriers to persistence in antidepressant use and adverse events the beneficiary may be experiencing. The third methodologically focused manuscript provided important insights into the correct (or incorrect) reporting of sample size related issues in the medical literature by elucidating current practices.



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.