User Story: Women’s Mental Health

Mytilee Vemuri, MD, Clinical Associate Professor, Psychiatry and Behavioral Sciences

How did you hear about the informatics consult program?

I heard about it during a presentation by the Stanford Medical Informatics Directors to Psychiatry clinicians.

What is your primary area of interest?

My clinical focus is in women’s mental health — managing psychiatric conditions related to times of hormonal change such as pregnancy or perimenopause.

In addition to my clinical practice, I serve as our department’s physician improvement leader. 

In mental health care, it can be challenging to identify, measure and track meaningful clinical outcomes, and to understand the relationship between interventions and outcomes.  I am interested in learning how we can make use of real world clinical data to better understand treatment outcomes and improve quality of care.  When I first learned about Atropos, I had recently contributed to a paper about estrogen treatment and mood, and was aware of some gaps in the literature, leading me to ask Atropos a question.

What question(s) did you ask the Atropos informatics consult service?

There is inconsistent data regarding the association between estrogen treatment and mood.. Some population studies seem to show an association between estrogen treatment and the incidence of depression in peri/post-menopausal women.  However, there are also several RCTs that show estrogen treatment can improve or have minimal impact on mood. So we used the service to ask some initial questions  to determine what the real-world data would show:

Among women aged 40-60, what is the likelihood of subsequent development of depressive disorders in those treated with estrogen therapy versus those who were not treated with estrogen therapy?

How was your experience working with our service?

My first prognostogram came very quickly, within 1-2 days, and was much better than expected: I thought the propensity score matching was an interesting methodology and relatively easy for me as a non-statistician to understand. 

It would be nice to have a short video explaining how we can use the service most effectively and how to frame questions in order to make most use of the available data. Average clinicians just want to know how this can be helpful and what the gaps might be when we’re using observational data.

Any closing thoughts?

As physicians at Stanford, we get a lot of startups that come through our doors, but Atropos is one of the few that I am excited about using and can see direct applications to clinical care!

Because most clinicians, even at academic centers, don’t have their own labs or big budgets to do large trials, having a service like this to answer clinical questions that are typically not addressed in the literature, or provide data to support a publication has the potential to be very useful to clinician educators. 

There are a lot of unanswered questions clinically, especially in mental health, and your service can help us mine the data to get closer to usable answers to those questions.

Myteile Vemuri, MD, Clinical Associate Professor, Psychiatry