Newfoundland Youth Hospitalized more than Canadian Average?

I vividly recall my first research methods and statistics course, which I took during the second year of my undergrad. I considered myself a math-liker–not a lover–who would like research and do well with statistics. Thus, I was excited to get into the nitty-gritty of research. The course kicked off with an intense lecture regarding the nature of knowledge and believing information. Why should we believe what someone tells us? Long story short, the idea is that we should read research, the news, and other sources of information with critical eyes. Not critical in a way that seeks to berate a researcher’s chosen method or approach, but a way that seeks to think about how a different approach may influence the results and implications. This lecture taught me that perspective matters. Why believe someone who says it’s raining outside when you could just look out the window? Perhaps, it’s simply drizzling. This is a dramatic rendition of the lessons learned from that first lecture, but the central tenants have stuck with me. One outcome of these lessons is that I love open data because, sometimes, perspective matters.

Recently, there were some media publications regarding data released by the Canadian Institute of Health Information (CIHI). These included articles by the CBC and some well-known Canadian psychologists, including some from Ontario. The main argument presented by these pieces was that youth in Newfoundland and Labrador (NL; a small eastern province in Canada, for any international readers!) are hospitalized for mental disorders more than the average Canadian youth. Specifically, “In 2018, there were 537 hospitalizations per 100,000 people in this province — a number that has gone up every year since 2007 — compared with the national average of 495.” (CBC; The hospitalization data is public (thanks to @njpollock for bringing this to my attention), so let’s have a look. You can access the data yourself:

The following Table is the data to which the CBC article and others were referring:

The number don’t lie! 537 (NL rate) is higher than 495 (Canadian rate). But, wait a minute, there are confidence intervals (CI)! This isn’t population data; rather, we are inferring the population parameters from a sample. Without going into much statistical detail, the CI represents a range of plausible or compatible values of the true population parameter (which we do not know). A visualization may help (I can send the R script used to create this, if interested).

The main takeaway is that NL isn’t that different (if at all) from the Canadian average. Contrastingly, there are other provinces, and especially the territories, that deserve greater attention regarding youth hospitalization rates for mental disorders and related challenges.

So why would the CBC and other psychologists draw attention to NL’s hospitalization rates? Well, I don’t know. But in likelihood, they didn’t directly look at CIHI data, or understand statistical inference. Likely, the psychologists committed the former and the CBC the latter. I am grateful to have been given a critical eye, thanks largely to my first research methods and statistics course. Otherwise, I may have just retweeted false information. I implore others to visit public data when available to view it from your own lens. Thanks Dr. Stewart (@pstewart72) for that first lecture.

POSTSCRIPT: I should note that I am not undermining the need of mental health resources in Newfoundland and Labrador. I wholeheartedly believe that there is a lack of resources in NL. Specialized and collaborative care for mental health services is wanting and there are incredibly long wait times. Furthermore, NL is replete with rural and remote communities with no access to mental health services. I am eager to complete my training and bring my services to my home province. Hopefully, I can help reduce the youth hospitalization rate, regardless of where it lies relative to a Canadian average.

Here are some Mental Health Resources for NL residents:

Tyler Pritchard
PhD Student, Sessional Professor, Researcher, and Clinician

My research interests include suicide and theory, research methods and statistics, and online activity.