Farther from care, more avoidable deaths

A first-draft write-up — the data are real; the words are a starting point to argue with.

Some of the clearest evidence for why place matters in health hides inside a single word: avoidable. Statisticians split deaths under age 75 into two buckets — preventable ones that better prevention (screening, risk-factor reduction, injury control) could have averted, and treatable ones that timely, effective care could have caught. Together they're a rough scorecard for how well a health system is serving a population. Sort Canada by how remote each community is, and that scorecard falls off a cliff.

The gradient

Read the bars from left to right — from the easiest-to-reach neighbourhoods to the most remote. Avoidable mortality rises from about 178 per 100,000 in the most accessible areas to 415 in the most remote: roughly 2.3 times as high. Both halves grow — the preventable share (the bigger one) and the treatable share alike — so this isn't only about hospitals being far away; it's about prevention, risk, and access all thinning out together. Flip to Males and the whole gradient lifts: avoidable death in the most remote areas approaches 490 per 100,000, well over double the accessible-area figure.

The word that matters is still avoidable. These are, by definition, deaths the right system could have prevented — which is exactly what makes the gap a research-and-innovation target rather than a fact of geography.

Why the research is harder out here

Generating that evidence is genuinely tougher in a rural system, for reasons that compound:

• Thin workforce. Only about 8% of Canada's physicians — and a far smaller share of specialists — practise in rural areas that are home to roughly a fifth of the population (CIHI). Fewer clinicians means fewer investigators, thinner subspecialty coverage, and less protected time for research.
• Smaller, more dispersed populations. Recruitment for a trial is slower when patients are spread across counties, and rare-condition studies may never reach their numbers locally.
• Distance and infrastructure. Travel burdens fall on participants and staff alike, and the research infrastructure — coordinators, ethics support, data systems — is sparse.
• Recruitment and retention. The same pressures that thin the clinical workforce thin the research one.

Why it's arguably more valuable

Here's the twist. Most health evidence is generated in large urban academic centres — and a protocol that works with a subspecialist down the hall and a subway to the clinic often doesn't transfer to a community two hours from the nearest MRI. The gradient above is partly a measure of that mismatch: interventions designed for the easy case, deployed into the hard one.

The places where research is hardest to do are often the places where locally-fit evidence would change the most outcomes. Avoidable deaths are, by name, the addressable ones.

That's the case for research embedded in a rural system rather than imported into it: it studies the real constraints — distance, workforce, the local mix of disease and risk — and builds solutions that survive contact with them. When the baseline is 2.3 times the avoidable-death rate of the easy areas, even modest, well-fitted improvements move a lot of lives.

The honest caveats

• These are area-level (ecological) rates. Remoteness travels with other things — lower income, less education, a larger Indigenous population share carrying the legacy of colonialism, different risk behaviours — and this chart doesn't disentangle them. Remoteness is a marker for a bundle of causes, not a lone one.
• “Avoidable” is a defined framework, not a verdict on any single death. The preventable/treatable split follows an international (OECD/CIHI) cause list; it's a population-level lens.
• Age-standardized, and from 2016. Rates are adjusted for age so places with different age structures compare fairly; the pattern is stable across the 2011–2016 series, but the levels are a few years old.
• The physician figures use a different rural definition than the mortality table — treat “~8% of physicians, ~18% of people” as the well-established shape of the mismatch, not a decimal-precise ratio.