Sara Johnson

Sara Johnson


Sarasa (Sara) Johnson, MSc, completed her undergraduate degree at Queen’s University in Life Sciences, and as soon as she took her first epidemiology course, she was hooked! After conducting a mini-thesis in epidemiology and attending a seminar by Dr. Linda Levesque, Sara was convinced of the importance of observational drug effect studies. She moved to Montreal to do her master’s in epidemiology at McGill. Soon after graduating from the program in 2016, Sara accepted her current position as an analyst at the Centre de recherche du CHUM in Montreal. Here, she conducts analyses for the CNODES Quebec site, along with other research projects for CHUM.

Trainee Profile

CNODES Analyst from 2016 to present
Because CNODES keeps providing training opportunities and implementing new methods and different projects, I feel like I will continue to learn at CNODES for a long time.

In April, 2018, Sara presented a poster at the International Society for Pharmacoepidemiology (ISPE) Mid-Year Meeting, which was held in Toronto, Ontario. She kindly offered to share her research, experiences, and a bit about herself for the CNODES newsletter and website.

Could you summarize the research that you presented at the ISPE Mid-Year Meeting?

I presented the poster, “Impact of accounting for hospitalizations during follow-up on measures of medication adherence and persistence in prescription drug databases.” One of the challenges of doing drug utilization studies with administrative health data is when you only look at the dispensation of medications, anytime someone goes to hospital, it looks like they are not taking their medications. So when we try to measure medication adherence or persistence, we might end up falsely concluding that someone is not adherent, when in reality they are receiving care in hospital. For the study, we examined certain measures of medication adherence and persistence, taking into account the time spent in hospital and then without taking hospitalizations into account. In the populations we looked at – cardiovascular patients taking antiplatelets after an acute coronary syndrome (ACS) or direct oral anti-coagulants (DOACs) for atrial fibrillation – we found that accounting for hospitalizations didn’t really make a difference. This was good news because including hospitalization data can be difficult and time consuming. Our study indicates that, at least in these populations, hospitalizations are not necessarily something to be concerned about when looking at how consistently people take their medications. Where this could be an issue is in populations where people have long hospitalization times (for example, possibly with certain psychiatric populations).

We are hoping to have this research published in a journal very soon. Stay tuned for that!

Tell us about your experience presenting at ISPE.

It was great! People were much more interested in the poster and the project than I had expected. A lot of people came up to me at the conference and said that when they try to submit papers, the issue of hospitalization data comes up. Reviewers ask, “Why didn’t you look at the time people spent in hospital?” At the conference people were actually quite excited about the idea of having a study that they could possibly cite to support not accounting for hospitalizations, especially for those who don’t have access to hospitalization data. It is a more common problem than I had thought initially. I also made some great contacts at the conference, some of whom wanted to collaborate on projects in the future, so that was exciting as well.

What excites you the most about the research you are doing?

For me it’s the practicality of it. Being part of CNODES I’m doing pharmacoepi work and it’s really interesting to find out what’s actually happening within populations. Also, I quite enjoy working with these big data sets. It’s kind of like being a sculptor – like there’s something in there that you are trying to find and you’re chipping away to get at the information. The team that I’m working with now at the CHUM are focused on cardiology research. It’s not necessarily my direct interest but, to be honest, that’s why I like working at CNODES as an analyst; I’m not necessarily driven by my own research interests, but I enjoy participating in projects and being someone who collaborates and helps others achieve their research goals.

Are there any challenges you would like to share?

Having time in general is a bit of a challenge, and trying to balance the CNODES projects can be difficult. Sometimes for the CNODES projects we are using methods that are quite new to me, so it can be a bit of a challenge to get myself up to speed and ensure that I’m applying models appropriately. Definitely CNODES is building the infrastructure to have more support. Our site at CHUM is quite small and for my first year at CNODES, I was the only Quebec analyst. Now we have another analyst here. Also, I know that if I’m having difficulty, I can reach out to one of the other more experienced CNODES analysts. That’s a great resource because they have so much experience applying these methods for CNODES studies. Getting other analysts to send me pieces of code is helpful, and having access to the video presentations on the CNODES website is really helpful too because often they are talking about methods that we are applying directly in the CNODES projects.

Where would you like to be career-wise in five years?

I’m happy for now in my position at CNODES as an analyst in Quebec, and I’d like to stay here until I feel I’m no longer learning. Maybe that will be in five years, but maybe that will be in 20 years, who knows. Because CNODES keeps providing training opportunities and implementing new methods, and different projects, I feel like I will continue to learn at CNODES for a long time. One of the challenges of being a data analyst in an academic setting is that there isn’t a clearly defined career path or steps for advancement that you might see in a different setting. Since I want to continue doing this type of collaborative work, rather than do a PhD and direct my own research, it’s hard to predict what I will be doing next.

Outside of work and studies, what do you enjoy doing?

I’m someone who likes to do a lot of different activities. I quite enjoy sports even though I’m not good at them. Now that the weather is nicer, I’m hoping to be playing more tennis and beach volleyball. And I really like board games as well – it’s a hobby of mine.