I'm very happy to say that I received a CIHR Project Grant with Andrea Benedetti (McGill) for our project:

"Evaluating treatment effect heterogeneity and optimal regimens in multidrug resistant tuberculosis using causal inference modeling".

Résumé en Français

Lay Abstract:
With the increase in antibiotic resistance, the treatment of certain strains of tuberculosis (TB) is becoming more challenging. TB caused 1.4 million deaths in 2014 making it the single most deadly bacterial infection worldwide. Multidrug resistant tuberculosis (MDR-TB) is a condition afflicting an estimated 480 000 globally, and is defined as TB that is resistant to the primary antibiotics used to fight the disease. MDR-TB is treated using different combinations of alternative antibiotics. What is not known is which types of patients respond better to which combinations of antibiotics, and which patient characteristics are most closely related to successful treatment. In order to use statistics to answer this question, a large amount of data is needed. Previous studies (with limited numbers of patients sharing their data) did not have enough information for investigators to unravel the relationships between patient characteristics and the probability of curing the disease. The objectives of our study are to 1) Learn which patient characteristics affect treatment success for different combinations of antibiotics, and 2) Design treatment strategies to maximize the success of treatment for MDR-TB. Our team has previously performed a large review of existing studies and has combined all available data into a much larger dataset. This dataset will allow us to investigate which patient characteristics (such as age, sex and antibiotic resistance) can help predict the success of specific treatments. We will use this information to statistically construct patient-specific treatment guidelines, which will ideally be used by doctors to treat their patients.

We greatly appreciate the support of CIHR and are excited to continue our work on this project!

I am currently seeking a statistical research assistant. Please see the job ad, attached. I will be accepting applications until February 27, 2017.

Our recent article (with lead author Ashley Naimi) describes the usage of causal mediation methods to evaluate potential interventions to reduce health disparities between racial groups in the US. In this article, Dr. Naimi describes how elevated non-Hispanic Black infant mortality rates can be largely (but not entirely) explained by lower rates of breastfeeding prior to hospital discharge. However, the importance of this potential intervention is seemingly overestimated using standard regression methods compared to double robust approaches.

I was happy to participate in this very interesting project, which is currently in press at the American Journal of Epidemiology.

In November I gave a seminar for the Canadian Network for Observational Drug Effect Studies (CNODES) entitled

"Targeted Learning for the Estimation of Drug Safety and Effectiveness: Getting Better Answers by Asking Better Questions."

​The video presentation synced with the overhead slides is now online on the CNODES website.

In this talk I demonstrate the importance of taking a marginal approach to the estimation of the effects of medications. I focus on the "Targeted Learning" roadmap (van der Laan and Rose, 2011) which is used to construct robust estimators of causal effects.

Photo credit: Christina Esteban Photography, http://cepstudio.com