By Jason Clement
A native of Winnipeg, Manitoba, Dr. Robert Platt, Professor in the departments of Pediatrics and Epidemiology, Biostatistics and Occupational Health (EBOH) at McGill University’s Faculty of Medicine, and the inaugural Albert Boehringer (1st) Chair in Pharmacoepidemiology, has long been fascinated with numbers.
This fascination led Platt to McGill for his undergraduate studies where he obtained a BSc in Mathematics in 1990. He returned home to Manitoba and completed a Master’s degree in Statistics in 1993 before earning a second Master’s degree and a PhD, both in Biostatistics at the University of Washington in Seattle in 1995 and 1996 respectively. Uncertain what he wanted to do with his career, he considered options in the pharmaceutical industry and academia in the United States before an opportunity presented itself at McGill and the appeal of returning home to Canada lured him back.
Home is where the heart is
The position at McGill was based in the Department of Pediatrics, and a joint appointment in EBOH allowed provided Platt the opportunity to perform statistical research while working on child health with pediatricians. “I took the job at the time because it offered me a lot of flexibility,” he says now. “If I didn’t like it I could still go into the pharmaceutical industry. I could go back into a more traditional academic role. It gave me a lot of options.” Shortly thereafter he got involved working with people like Dr. Michael Kramer, a frequent collaborator, in pregnancy-related studies and after a few years he realized he was never going to leave.
“I think epidemiology and biostatistics at McGill offer a really dynamic environment where there’s a lot of interaction between very theoretical statisticians on one end, very clinically-oriented researchers on the other end and people in between,” notes Platt. “What I like is that I get to be somewhere in the middle where if I get involved in a pharmacoepidemiological study where I’m learning a lot about the content – I can be more on the clinical side, or I can be more on the methods side – there’s a lot of flexibility to do what you want and there are also a lot of smart people around.”
Verifying the safety of on-the-market medications
The study of both the good and bad effects of drugs as they’re used in real life, pharmacoepidemiology is a field with tangible impact on the general public. Drugs are typically approved for use based on randomized trials and a process that Health Canada and the Food and Drug Administration in the U.S. go through prior to their release on the market when physicians can start prescribing them. “In pharmacoepidemiology our job is to look first at how they’re used in practice in the real world and then whether there are any unexpected side effects or adverse events, or benefits, that occurred because of the drugs,” says Platt.
For an example of how these studies can make an impact, one need look no further than at the case of Vioxx, where everything looked fine during randomized trials but once it went to market an extra risk of cardiovascular disease was found and the drug was removed from the market. “It’s our job to look at these extra risks of those adverse events or serious side effects that might occur once a drug is on the market,” he explains. “If the randomized trials aren’t very big – a few thousand to tens of thousands of people – but if you have a rare event you’re not going to see it in a sample of that size – it’s not until millions of people start taking the drugs that you start seeing these associations with rarer side effects.”
In service to the nation
Platt notes that McGill has a long history of good work in pharmacoepidemiology – led by people like Dr. Samy Suissa – who really have driven the program for a long time. Along with several other colleagues, Platt and Suissa – who serves as principal investigator – now work as part of a national network known as the Canadian Network for Observational Effects of Drug Studies or CNODES, whose coordinating centre is located at McGill. Working together, members of CNODES typically look to answer questions that arise from Health Canada’s suspicion that there might be an association between a drug and an outcome. Among the many things that the group has studied to date, they’ve looked at the effects of statins on the kidney and on diabetes and have shown that at higher doses statins may have side effects that make calibrating the prescription a little better a useful process.
“The flipside of it is that we have also had a number of studies where we have shown that drugs are safe,” notes Platt. “So, one thing is to say that does this drug cause this side effect? The other is to show that, no, it doesn’t cause this side effect. The most recent example is the study on incretins, which are an advanced drug – a relatively new class of drugs for type-2 diabetes and there were some suspicions that they may have some adverse effects on heart disease and on the pancreas and we have a series of studies that more or less demonstrate that those drugs are safe, that they don’t cause an increased risk of heart disease or pancreatic cancer or pancreatitis. We can show that it’s safe to prescribe them (the drugs) and that the good things that they do outweigh the bad.”
A draw to big data
Platt was drawn to pharmacoepidemiology not through a specific study or drug or background but rather through the research methods. “For a long time I’ve worked as a collaborating biostatistician on these studies,” he explains. “So these are big data sets that require lots of statistical analyses and my role has been to try to improve the processes by which we do the studies to make the answers better. In other words, independent of whatever drug they are studying, let’s see how we can design the study and analyze the study to get the best possible answer.”
The field is an example of big data that presents some interesting combinations of scientific knowledge making it important to work with people that understand both the science and the clinical practice around these drugs but then also these massive amounts of data and the statistical and computational issues that relate to handling them. “It’s just a fun and challenging area,” says Platt.
The future is bright
Platt became the inaugural Albert Boehringer (1st) Chair in Pharmacoepidemiology at McGill when his Department Chair, Dr. Gilles Paradis, thought he would be a good fit and encouraged him to put his name in the hat for the role. Now that he’s in the position he’s excited because of the possibilities and the vast talent in pharmacoepidemiology around McGill and its affiliated research hospitals. “We’re a real leader in the field and the Chair gives me an opportunity to be a hub for pharmacoepidemiology on campus at McGill and to expand and extend my research program in terms of the methods work and the substantive work,” says Platt. “The generous gift from Boehringer Ingelheim is also allowing us to create a graduate option in pharmacoepidemiology that’s focused on training the students and to attract some new faculty who are being brought on board due to the Chair.”
For students who find themselves considering their futures just as Platt himself once did, he believes that pharmacoepidemiology offers a world of opportunities. “It’s a field in very high demand so we’d like to generate more graduates,” he says. “If someone is interested in health sciences, reasonably quantitatively comfortable and comfortable with statistics it’s a good spot to be looking at for future career prospects.”