Wolfson: In pursuit of a world-class health data system

Good arguments, challenge given complexity and jurisdictional issues. Some progress through programs like MyChart which makes managing my hospital-based visits and tests so much easier:

Canadians see new and increasingly powerful computerization in almost every facet of their day-to-day lives—everywhere, that is, except for something as fundamental as health care, where systems are too often stuck in the past.

Canadian governments have invested heavily to advance the use of health data, most often without a clear national vision. An incoherent approach for health data is hurting health outcomes, escalating sector costs and expanding inequities.

The health sector relies on data to guide almost every decision, from the choice of an antibiotic for treating an individual’s simple urinary tract infection, to national pandemic policies that affect millions of Canadians.  The challenge is that health and health-related data are so poorly organized and managed in Canada.

While health-care providers create and have responsibility for safe management of records of health care encounters, personal health information in Canada should be available to patients as well.

The reality is that patients continue to have limited access to or insight about data about them, as that remains under the de facto control of myriad and siloed health-service providers. The inevitable result of this scattered provider-centric rather than patient-centric approach is that patient data are typically spread among their various health care providers’ uncoordinated and unstandardized data systems. The result: both providers and patients have to work with incomplete and incoherent information.

The health sector has invested for years in digital technology in the mistaken belief it would immediately solve our health data woes. We have failed to realize the true obstacle to effective health data collection and use is not mainly technological, but a matter of policy and governance.

To realize its tremendous promise, health data in Canada must become centred around the individual. Services would be designed around people and by people. The “life flows” of patients and families would mesh with the “work flows” of providers and institutions. There would be one centrally accessible virtual digital record for each of us.

As patients move from home to care setting, and from provider to provider, data would remain seamlessly accessible for those in the patient’s “circle of care”; no need to repeat your health history for each new provider. Data flows about a person—who sees what—would be transparent to the person involved or to their proxy.

Health-care providers would continue to access personal information for the purposes of individual care, while a new role—health-data stewards—would be mandated to curate population-based data for public good while ensuring privacy and confidentiality.

How can we make this vision a reality in Canada?

This will only happen with a fundamentally reimagined approach to health data policy and governance for the digital age, grounded in mutual trust. Governments need to trust the public by providing clear and complete information, with people in Canada trusted to act for the good of the community. The public would be involved in policies regarding health data collection, sharing, use and communication. And governments need to make real their commitments to respect Indigenous data governance.

Governments must also trust each other and recognize the over-riding importance of coherent pan-Canadian data.  At present, each province and territory has its own health data policies, standards and governance, with very little active coordination across borders. Even in the area of death certification, where the vital statistics registrars have worked together for many decades, we have the current failure to agree on and report COVID-related causes of death in a standardized manner.

Accessing the routinely collected data that do exist is a marathon for fully authorized researchers, for example to study post-marketing drug safety and effectiveness. Even with all the available approval processes in place, it has still proven impossible to analyze these kinds of data for the country as a whole.

These pervasive blockages to generating coherent and timely pan-Canadian information seriously impairs our ability to respond to public health threats and generate insights that could improve health outcomes for all.

Ultimately, we need to rethink—not rejig—how we manage and use health data. The Pan-Canadian Health Data Strategy report, authored by an Expert Advisory group, points out the need for a culture shift in health data use. Public input has to be part of the transformation to person-centred health care and health data systems.

Of course, updating health data systems has costs. But given the tens of billions of health-care dollars the federal government is providing to the provinces through fiscal transfers, it is long past time the federal government leveraged this clout—using both carrots and sticks—so that people in Canada can finally have informed, accessible health data when and where they need it most.

Michael Wolfson is a former assistant chief statistician at Statistics Canada, and a current member of the University of Ottawa’s Centre for Health Law, Policy and Ethics. Vivek Goel is president and vice-chancellor of the University of Waterloo. Both are members of the Expert Advisory Group on the Development of the Pan-Canadian Health Data Strategy.

Source: In pursuit of a world-class health data system

Canadians’ health data are in a shambles

Unfortunately, all too true, with too few exceptions, based upon my admittedly anecdotal experience in Ottawa:

Canadians see new and increasingly powerful computerization in almost every facet of their day-to-day lives – everywhere, that is, except for something as fundamental as our health care, where systems are too often stuck in the past.

When we go to the doctor, we get prescriptions printed on paper; lab results are sent via fax; and typically, medical offices have no direct links to any patient hospitalization data. And while the pandemic sparked a mad scramble to set up many new data systems – to track who was infected, where there were ventilators, who has been vaccinated and with which vaccine – this has happened in a largely unco-ordinated way, with Ottawa and provincial governments each developing systems separately.

As a result, even these newest computer systems are duplicative, and they do not communicate across provincial boundaries, or even within some provinces – not even, for example, to connect vaccinations, infections, the genotype of the virus, hospitalizations, other diseases and deaths so they are centrally accessible. And so Canada’s recent health-data efforts have wasted millions of dollars while failing to provide the evidence base needed for real-time effective responses to the fluctuating waves of COVID-19 infections..

This kind of failure is not new. Even before the pandemic, key kinds of data have long been imprisoned by data custodians who are excessively fearful of privacy breaches, even though the data are generally collected and stored in secure computer databases. A broad range of critical health care data remains unavailable – not only for patients’ direct clinical care, research and quality control, but also for tracking adverse drug reactions, showing unnecessary diagnostic imaging and drug over-prescribing. The result is that major inefficiencies in the systems remain hidden – and may actually cause health problems, and even deaths by medical misadventure.

There are many directions one could point the finger of blame, but as a new report from the Expert Advisory Committee to the Public Health Agency of Canada found, the root cause is a failure of governance. Federal and provincial governments have failed to agree on strong enforcement of common data standards and interoperability, though this is not only a problem of federalism. Health-data governance problems are also evident within provinces where one health agency’s data system is not connected to others within the same province.

What Canada and the provinces have now is essentially provider-centric health-data systems – not just one but many kinds for hospitals, others for primary care, and yet others still for public health. What Canadians want and need is patient- or person-centric health data. That way, no matter where you are in the countryyour allergies, chronic diseases and prescriptions can be known instantly by care providers.

Private vendor-centric health-data software also pose a threat, as do data collected by powerful tech companies from new wearable technologies that offer to collect your health data for you. If Canada does not act swiftly and decisively to establish the needed governance, competing vendor software and individual data will continue the rapidly growing cacophony of proprietary standards. This trend is raising new concerns about privacy, along with untracked increases in health care costs.

The fundamental importance of standardized, interoperable, securely protected health data has been known for decades. There have been repeated efforts to achieve a modern effective health-data system for Canada. But federal cajoling and even financial incentives have failed. Much stronger governance mechanisms are required, and urgently, as the global pandemic has revealed.

The federal government has the constitutional authority to play a much stronger role, given its powers in spending, public health, statistics, as well as “peace, order and good government.” It also has readily available regulatory powers under the Canada Health Act.

Of course, high-quality data collection and data software have costs. But given the tens of billions of health care dollars the federal government is providing to the provinces through fiscal transfers, it is long past time they leveraged this clout – using both carrots and sticks – so Canadians can finally have informed, accessible health data when and where they need it most.

Michael Wolfson is a former assistant chief statistician at Statistics Canada, and a current member of the University of Ottawa’s Centre for Health Law, Policy and Ethics. Bartha Maria Knoppers is a professor, the Canada Research Chair in Law and Medicine, and director of the Centre of Genomics and Policy at McGill University’s Faculty of Medicine. They are both members of the Expert Advisory Group for the Pan-Canadian Health Data Strategy.

Source: https://www.theglobeandmail.com/opinion/article-canadians-health-data-are-in-a-shambles/

Wolfson: Without good data, we’re flying blind on good health care

Hard to disagree:
Yet again, the provinces are wailing about the need for more federal money for health care, with no strings attached. These are the same provinces who have for decades grossly underfunded long-term care. And as we are seeing in real time, many of the provinces are scrambling last minute to have the data to understand and manage the pandemic, most recently in rolling out vaccinations.Of course, the federal government should not cave in to these unwarranted provincial demands. It is entirely within the federal government’s constitutional authority, despite what the provinces are saying, to play all kinds of roles in the health sector. The provinces never object to the billions of dollars the federal government pours into health research every year.

But many of the provinces are still in the dark ages when it comes to collecting and making available for crucial health services research the kinds of data essential to understanding how well health care is being delivered.

Car dealers and airlines have for years had far better computer systems to keep track of the health of your car, and every airplane seat in the world. The Big Tech software companies not only collect humongous volumes of data on many of us, with their real-time technologies; they are continually doing experiments to see what will induce us to click more on their sites and advertisements.

Yet our public health authorities are having difficulties even connecting our COVID-19 tests, our vaccinations, and any hospital visits, as in many cases these software systems are completely separate silos.

So, what’s wrong with the federal government saying to the provinces, if you want more cash, you first have to implement decent real-time data systems? Indeed, the constitution expressly assigns jurisdiction for statistics to the federal government, hence the authority to play a strong leadership role in health data systems in the provinces.

The federal government has to be accountable to us in our role as federal taxpayers, not only as provincial taxpayers. It is incumbent on the federal government to ensure that any monies it transfers to the provinces are used for the purposes intended.

If the transfer is for health care, a province cannot use the cash received to finance tax cuts. If the transfer is to push the provinces to improve the generally awful state of long-term care, then it is entirely reasonable for the federal government to impose requirements on the ways the provinces spend the money, including collecting data and other reporting to ensure that provincial promises are more than rhetoric.

These requirements are more than just reasonable; they are based on the federal government’s constitutional authorities for the spending power, and for peace, order and good government.

At the same tine, the federal government sorely needs to up its game.  For example, the most recent speech from the throne committed the federal government to take the lead in developing national standards for long-term care. But one has to wonder how this can be achieved when there is virtually no nationally comparable data on one of the most crucial aspects of long-term care quality, namely staffing.

The federal government was negotiating the purchase of COVID-19 vaccines many months ago. That should have been more than sufficient lead time to ensure that there would also be a national system in place when the vaccines started arriving to track how the vaccine rollout was progressing.

Yes, it can and should be left up to each province to decide vaccine allocation. But with the current hodge-podge of computer, fax and email systems, there is no way to keep track in real time of what’s being done across the country. There’s also no way to connect vaccinations to rises and falls in outbreaks at the level of detail needed to inform lockdown policy.

Health care and public health are quintessentially knowledge industries. It should be obvious that they should be organizations that learn from experience. But it is impossible to learn from experience if you have no way of knowing just what you are experiencing.

Proper data collection and analysis are essential.

It is long past time that the federal government stiffened its spine and, in addition to saying it will work collaboratively with the provinces, put some muscle into meeting nation-wide concerns.

Michael Wolfson, PhD, is a former assistant chief statistician at Statistics Canada and a member of the Centre for Health Law, Policy and Ethics at the University of Ottawa.

Source: Wolfson: Without good data, we’re flying blind on good health care

Canada needs a national database to track COVID-19 vaccination in real time

Good analysis on the need by Michael Wolfson, a former assistant chief statistician at Statistics Canada.

One of the frustrations I encountered with the non-resident birth data was that Quebec does not automatically include its health data as a matter of course in the Canadian Institute of Health Information, with Quebec refusing to provide me with the comparable data (given health provincial jurisdiction, point of principle over-riding common sense).

So not sure how realistic Wolfson’s proposal is but better and consistent data helps all:

With Canada in the midst of rolling out the vaccines, the importance of effectively monitoring the immunization campaign is coming to the fore. The federal government has recognized the importance of monitoring data, at least within federal jurisdiction, and the prime minister himself recently emphasized the federal government will “be a partner with the provinces  … [for] better co-ordination of data.”

The government response nicely recognizes the lead role of the provinces in setting priorities for vaccination. And the federal government appears sanguine about the existing jumble of layers of vaccine-monitoring data systems, including for adverse reactions.

This co-operative federalism is wonderful—when it works. However, for anyone with experience in software, databases and statistical analysis, the vaccination monitoring described sounds like a dog’s breakfast. That’s not good enough when lives are on the line.

Standard adverse-event reporting systems in the U.S. and Canada missed the scandalous connection between Vioxx and heart attacks. Something more reliable is essential for COVID-19 vaccinations, not only for safety but to avoid misinformation from anti-vaxxers.

Canada has world-class potential for statistical surveillance of adverse health events in the electronic health databases housed in each province. But these data often reside in multiple impenetrable silos within each province.

The COVID-19 pandemic has dramatically increased the urgency of breaking down these data silos. One of the most important blockages has been provincial insistence that health care is their show; the only role for the federal government is to hand them more money, no strings attached.

This has to stop.

Specifically, for a vaccine registry and monitoring, the obvious solution is a single standardized system, mandated by the federal government using its constitutional jurisdiction for statistics. The federal government could commission an organization—Statistics Canada is an obvious choice—to immediately develop a secure, real-time data-collection portal or site for critical information on every person who is vaccinated for COVID-19.

This software system would be used in clinics, doctors’ offices, and drugstores. The nurses and other health professionals giving the vaccination would enter information, exactly as done for flu vaccinations. But now, some of the information would be federally mandated, over and above anything recorded for patients’ medical records and provincial billing purposes.

Decades of experience have shown that rhetoric about federal-provincial co-operation has continually failed, resulting in the patchwork of incoherent and incomplete data that have been limiting too much of the science for managing Canada’s pandemic, and the health-care sector more generally.

The federal government was successful in eliminating doctors’ extra billing by holding back transfers to the provinces. But with no strings attached, a number of provinces have been shamefully clawing back some of the COVID-19 cash payments the federal government has sent to the neediest Canadians by reducing or cancelling their social assistance. To ensure effective implementation of this monitoring solution, strong fiscal sanctions should be included if provinces do not co-operate.

Real-time, federally mandated vaccine monitoring will provide crucial information on vaccination uptake not only by province, but also by neighbourhood, type of vaccine, race/ethnicity and occupation —enabling provincial and local public-health authorities to target vaccinations to the vulnerable. This is not federal intrusion into provincial jurisdiction; it is simply the most efficient constitutionally enabled way to provide critical information.

There is no reason that this kind of software could not be adapted and made available across the country for vaccinations in a matter of weeks, along with speedy agreements on data standardization.

While confidential personal data are involved, Statistics Canada has, for decades, collected exactly such data in the monthly labour-force survey (recently doing so online), with exceptionally strong safeguards for security and confidentiality.

There are obvious privacy concerns. However, we must be careful not to allow them to overshadow the potentially huge benefits. The framers of Canada’s constitution, over a century and a half ago, recognized the fundamental importance of critical statistical information that is national in scope.

While the proposed data flows may raise concerns among provinces and territories regarding ownership, these can be ameliorated with clear ground rules on how they can access these data.

Privacy commissioners across Canada have adopted the principles of necessity and proportionality as the central criteria for data collections that raise privacy concerns. For pandemic vaccination, with the deaths of potentially thousands of Canadians in the balance, these criteria would clearly be met.

Now, more than ever, Canada needs a strong national approach for monitoring data to ensure vaccination proceeds effectively, fairly and safely.

Michael Wolfson is a former assistant chief statistician at Statistics Canada and a member of the Centre for Health Law, Policy and Ethics at the University of Ottawa.

Source: Canada needs a national database to track COVID-19 vaccination in real time

Ottawa must put data first and tie to health funding

Agree in principle but politically hard to achieve (Quebec doesn’t even automatically share its data with CIHI):

The federal government looks yet again about to transfer billions of dollars to the provinces with essentially no strings attached.

We’ve seen this before with $40 billion in the 2004 First Ministers’ Health Accord and then $11 billion in the 2017 Health Accord, both highlighting home care, without evidence of significant progress.

And the prime minister just announced $19 billion for the Safe Restart Program, though without any details, especially as to what the federal government receives in return.

One major quid pro quo could address Canada’s profound lack of high-quality data, especially highlighted by the COVID-19 pandemic. While U.S. analysts are able in near real time to estimate and project COVID cases, hospitalizations and deaths down to the county level, Canada is barely able to produce comparable data by province.

Some of this $19 billion is meant for COVID testing and tracing, and improvements in long-term care.

A major failing in the tragic and disproportionate COVID-19 mortality rates in nursing homes was due to poor staffing levels, an issue that has been known for decades and pointed out in myriad reports and studies. But there are essentially no comparable and complete national data in this area.

As strongly recommended in the recent Royal Society of Canada report, high quality data on current staffing levels, connected at the individual level to health outcomes, are essential, especially for the federal government to develop the evidence-based national standards for long-term care so many have been calling for.

The provinces have typically argued that health care is a provincial jurisdiction, so the federal government cannot compel them to provide sorely needed data. However, in another example, we have had almost two decades of cajoling the provinces with federally funded Canada Health Infoway paying at least half the cost to develop and implement standardized and interoperable software systems for electronic health records.

Most relevant for the current pandemic, Infoway was specifically tasked with producing a system for anticipating and dealing with infectious disease outbreaks. This system, had it been working even 15 years after its initial funding in 2004, would have enabled a very different outcome this year, likely with far fewer cases and deaths from COVID-19.

Paper agreements and cajoling the provinces with optional subsidies have clearly failed. It’s time for a much tougher stance.

The federal government has the necessary constitutional powers, including explicit jurisdiction for statistics, criminal law, spending powers, and the general peace, order and good government (POGG) power, to compel the collection and flows of 21st century kinds of data.

Monique Bégin, as federal minister of health, successfully ended the practice of physicians’ extra-billing by amending the Canada Health Act to deduct any extra billing from an offending province’s fiscal transfer. The Supreme Court has just upheld the federal government’s genetic privacy legislation as constitutional despite objections from Quebec.

In the current pandemic emergency, high-quality, standardized, real-time data on “excess deaths,” COVID cases and hospitalizations, and details on the operations of the thousands of nursing homes and retirement residences across Canada are essential.

For nursing homes, we need these data to learn why some were completely successful in avoiding any novel coronavirus cases amongst residents and staff, while others suffered tragically. In turn, such statistical information will provide the federal government the strong evidence base needed to take the lead in establishing national standards for nursing home staffing levels, though action on staffing must not wait for perfect data.

And once we have standardized individual-level data on COVID cases, including factors like age, sex, neighbourhood, other diseases, individuals’ household composition, race, hospitalization rates, disease severity, and deaths, as the U.K. has been able to do for 17 million of its residents in near real time, then Canada will be able to support far more sophisticated analysis and projections to deal with the current top pandemic issues — not least, whether to open bars or schools.

StatsCan takes criticism for cutting funding to LifePaths database

Another example of reduced government emphasis on evidence-based approaches:

Former Statistics Canada official Michael Wolfson is criticizing the agency’s decision to stop funding its LifePaths database, saying the program has been essential for researching the long-term impacts of policy decisions.

Mr. Wolfson, a professor at the University of Ottawa who was previously assistant chief statistician at Statistics Canada, has written a new paper on retirement adequacy, which includes comments about his disappointment over the decision to cut funding to the LifePaths database, which he has used in his own work on retirement income.

“As a result, discussion of multibillion-dollar policies – discussions that could be informed by far smaller investments in statistical infrastructure – can now be pursued in ignorance,” he writes.

Statistics Canada stopped supporting the LifePaths modelling tool at the end of 2014, which means the database is not being updated with new data.

Statistics Canada spokeswoman Nadine Lacroix said the resources required to update and maintain the model were too great, and it “was no longer feasible” to continue the program.

She said the agency is developing a new “dynamic socio-economic” modelling tool that will be structured to ensure “sustainability, efficiency and responsiveness to client needs.” Statistics Canada expects to solicit feedback from stakeholders on the proposal next year.

LifePaths is a complex modelling tool developed in the 1990s that contains data on Canadians starting from 1971. It was started during Mr. Wolfson’s time at Statistics Canada to project demographic trends for Canadians decades into the future. It was intended to help shape public policy in numerous areas – including pensions, education and health care – by modelling the impacts of various policy alternatives.

The decision to stop maintaining the model comes amid broader criticisms over cuts to Statistics Canada research, most notably the 2010 decision to eliminate the mandatory long-form census in Canada and replace it with a voluntary survey.

In an interview, Mr. Wolfson said he has not published comments critical of Statistics Canada in the past, and worries he is being disloyal to his former colleagues by speaking out now.

“But I felt it was sufficiently important that I really felt I had to do it,” he said Monday.

StatsCan takes criticism for cutting funding to LifePaths database – The Globe and Mail.