Measure to Manage Healthcare


To paraphrase Peter Drucker’s famous aphorism, you can’t manage, much less improve, what you don’t measure. This applies in spades to Canadian healthcare. Despite reams of administrative data, we really don’t know much about individual transactions between patients and their providers and even less about their outcomes. We remain remarkably uninformed about both the efficiency and effectiveness of a program on which we spent about $228 billion in 2016, 70% of it in public funds. Our “system” is not quite flying blind but we have far from a clear view of where we are, where we’ve been, or where we’re going!

We look to CIHI[1], Statistics Canada, and provincial Quality Councils[2] for data on how Canada’s healthcare “systems” are meeting our needs. Perforce they all deal primarily with administrative and clinical data collected in and by hospitals to measure the incidence of failures of patient:provider interactions. These data are important but to manage the “system” well we need to be more broadly informed on what each of its elements is contributing to the health and well being of the population.

It is an iron law of statistics that measures of all transactions will be distributed throughout a spectrum ranging from, at the extremes, abject failures to outstanding successes. With respect to the outcomes of a hospital stay or treatment by any provider, we should expect that distribution to be strongly skewed toward a return to health and no recurrence. Important as measures of failures are, they are but one tail of the distribution of outcomes, hopefully a small tail at that. To get a real handle on what society is getting for the expenditure of some 11% of GDP on healthcare services, it is vital to measure the whole spectrum of the distribution of outcomes from the failures, through the so-sos, and goods, to the outstandings. The very shape of the distribution would tell us a lot about where to put our efforts to make healthcare more effective and less costly. Assessing the outcome of a procedure or stay in hospital, for example, requires more than measuring whether or not it was followed by an infection or readmission; of the care provided an elderly person by a personal support worker if a fall occurred subsequently; or of a visit to a family physician if the patient went later to an emergency room.

We do know from international comparisons[3] by an American Foundation (itself somewhat discomfiting) that Canada and the United States stand second-last and last of 11 countries respectively as “clear outliers” in their healthcare systems’ performance! These comparisons of quality, safety, efficiency, equity, cost, and the overall health of the populations served are based, in the main, on well-tested surveys of patients and physicians. While highly credible as sources for a high level report card of system performance, such surveys are insufficiently granular to inform governance or management decisions whether by a health authority, hospital, primary care team, or community service provider. Real measures of transactional efficacy and of the outcomes achieved are required.

The need for more and better measures is well recognized. Haj-Ali and Hutchison reported recently[4] on the creation of a framework to measure how primary care meets the needs of its providers, system planners, patients and their families, governments, and of the general public. There remains, however, a long way to go before this work and studies like it extending to the many other healthcare services can be converted into practical measures of effectiveness and efficiency. Conceptual frameworks are one thing; measures of individual transactions and outcomes are something else. To what extent and how cost-effectively do various tests and images improve the safety and effectiveness of a patient’s diagnosis and treatment? How much is added to a family’s quality of life by regular home care services? How near normal is the range of activities of patients with knee replacements and how long does it take to achieve it?

You can bet that the owner of every corner store right through to those responsible for the governance and management of our largest corporations can tell you to the penny the cost of their every transaction and the impact of each on customer appeal and the bottom line. Why is getting transactional and outcome measures so hard and taking so long in healthcare? For the most part records bearing on the interactions between providers and patients exist. To get measures of their efficiency they need to be digitized, converted to a common standard, compared, aggregated, and rendered anonymous for analysis; the technology is available as are Canadian agencies[5] capable of doing the work.

Statistics Canada and CIHI have reported[6] on the need for and ways to collect more and better outcome data. Using the now nearly ubiquitous cell phone, tablet, and personal computer technology, Patient-Reported Outcome Measures (PROMs) and their results (PROs) can be readily obtained from patients and their families. Could it be that our provincial governments don’t want their publics and providers and their patients and families to know the results – what the spectrum of Canadian healthcare’s outcomes really looks like?

We need but to agree to use standards in common, link together our software systems, share the clinical information all of us have in one form or another, and devise new methods to measure what we are getting as a society for our $228 billion. It’s time to get on with it!

[1] Canadian Institute for Health Information

[2] in Alberta, Ontario, and Saskatchewan


[4] Wissam Haj-Ali and Brian Hutchison. 2017. Establishing a primary care performance measurement framework for Ontario. Healthcare Policy, 12(3) February: 66-79

[5] CIHI, Statistics Canada, Quality Councils


Authored by members of the Queen’s Health Policy Council:

Don Drummond
Chris Simpson
Duncan G. Sinclair
David Walker
Ruth Wilson

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>