Garrett M. Fitzmaurice, ScD

Biostatistics is key to the rigorous research for which McLean Hospital is known. Founded in 2007 and led by Dr. Fitzmaurice, the Psychiatric Biostatistics Laboratory collaborates with investigators across McLean’s research programs from the early stages of developing grant proposals through to the publication of findings. Dr. Fitzmaurice’s expertise includes experimental study design, statistical methodologies, and the analysis and interpretation of results.

Short-term biostatistical support is available to McLean investigators through the lab’s biostatistical consulting service.

In addition, Dr. Fitzmaurice’s group pursues their own research interests in statistical methods. These include methods for analyzing longitudinal data (e.g., repeated measures of depression on individuals over time) and social network data, methods for causal mediation analysis (e.g., methods to better understand the mechanism of action through which a treatment affects some mental health outcome of interest), general missing data problems, and methods for detecting and adjusting for overdispersion (i.e., adjusting for the presence of greater variability in a data set than would be expected based on a chosen statistical model for the data). The lab’s research in psychiatric epidemiology has concentrated on methodological problems surrounding the use of multiple informant data (e.g., parents’, teacher’s and child’s reports of a child’s own levels of anxiety) in mental health surveys.

Dr. Fitzmaurice and his staff develop statistical collaborations with McLean Hospital investigators on a wide variety of clinical and basic research projects in psychiatry and neuroscience. The following are examples of collaborative projects.

Over two decades, McLean scientists have built a rich database that is advancing the understanding of borderline personality disorder. Statistical analysis of these data is helping to clarify the factors affecting outcomes and the presentation of symptoms over time.

Using a factorial design, the lab has collaborated on a study of alternative interventions for alcohol misuse in TB patients, which examines both behavioral and pharmacological interventions simultaneously. This is an efficient and cost-effective approach in terms of sample size, enabling more than one intervention to be assessed in the same study participants.

Using novel causal mediation analysis approaches, the lab has examined the mediating and interactive role of symptoms and side effects in explaining the effect of antipsychotics on social functioning among schizophrenic patients. These analytic methods can provide insights on the complex interactions between patients’ symptoms improvements and side effects that commonly occur and thereby better inform clinical practice and new drug development.