Laura Langer is the lead author of a soon-to-be published study, “Increasing Incidence in Concussion: True Epidemic or Better Recognition,” that indicates at least 1.2 per cent of people in Ontario experience concussion each year, a number far higher than previously reported. Working with Dr. Mark Bayley and Charissa Levy, Laura examined data collected between 2008 and 2016.
What educational and research background do you bring to your work?
I have a degree in molecular neurobiology from the University of Toronto. I really enjoyed learning about both various types of neurodegeneration & neurotrauma and mathematical models.
You have been involved in a range of research regarding concussions children, youth and adults. What caused you to be interested in concussions?
I’m interested in trauma to the brain and each injury has its own unique characteristics. My early projects were in the area of stroke and epilepsy. This project allowed me to use both my knowledge of brain injuries as well as statistical modelling.
What have been the “aha moments” in your work?
For me it was the current concussion project. Looking at such a wide range of data from so many people has given me a great opportunity to see what has been going on in terms of treatment and outcomes. As I dove into the available data, I started to see some unexpected patterns. For instance, everyone seems to assume that concussion is a male-oriented injury but when you are older than 50 it becomes a “women’s injury”. This may mean more effort is necessary to understand how assessment and treatments are organized, and if current approaches may be affected by the previous assumptions.
There are also significant differences by region, both in the incidence of concussion, and the data shows not everyone has the same access to treatment. If someone lives in a rural part of the province the access to specialists and concussion-specific treatment is significantly reduced. Another interesting finding was the time between injury and death. Between 31 and 100 days after injury, those who died were usually over 80. There are some gaps in the data but it’s not clear why this is the case.
How do you manage the very large administratively-focussed data sets knowing that they do not necessarily present the complete picture of services used? For instance, not everything is or can be tracked in the system and there may be human error in the data entry, how do you account for that in your analysis.
For this project I happened to have access to ICES datasets which include both Emergency Department data and primary care information but of course nothing that identified individual patients. This means there is not as much information from primary care physicians and other health care providers. I definitely have had to increase details in the “limitations” section of my reports for this reason. For instance, I would love to have cause of injury but this is not available
The quality data of I get to work with It is very dependent upon the correct codes being submitted by the billing physicians, but I think most doctors do a good job. However, I think the numbers are sufficiently large enough to minimize billing errors and my models are internally validated. One recent model required almost 50 hours to bootstrap!
As physiotherapists, occupational therapists, psychologists, etc., do not bill to OHIP, their information is not tracked in the databases. There is an information gap there for sure.
What is the role of administrative data set analysis in driving policy and service funding decisions? Does this provide context for your efforts?
We believe that one of the outcomes of this project is information around access to appropriate health care between rural and urban. This information would definitely inform policy decisions. The outcome of the work can also support physician and care giver education.
We have been able to look at benchmarks regarding follow-up care. Seeing the actual rates between emergency department visits and those seen for a follow-up with a primary care physician within three days varies significantly depending upon location. For instance, there may be other mental health issues that arise from concussions but it is well known that access to psychiatric care is difficult across the province. Also, if someone chooses to see an allied health professional (not a physician) for follow up to an injury that information is not available.
Who do you see as your audience for your research findings?
I focus on clinicians. They need to know what the population they treat really “looks like.” The Ministry (of Health and Long-term Care) could also use this data to help determine how and where care should be adjusted across the province.
I would like it to be accessible to the general public given the existing stereotypes. People need to recognize the importance of seeing a physician soon after injury.
What do you see as the next steps for your concussion work?
There is already a follow-up to the current paper that will be focused on adults and their pre-morbid health conditions and demographic factors. This will focus on the current development of risk score to help identify adults who might be most at risk for developing ongoing symptoms following a concussion.
Based on our early modelling of a risk score tool we are finding that most people fall into the low risk category. There is a lot of information about concussion out there, and some of it may not be based on evidence leaving people with the wrong impression about concussion injuries. The limited testing of the risk tool has been well-received the physicians piloting the tool. At the end of the day this will help physicians better direct those who need the most care to get it, while supporting the recovery of all patients more effectively.