Proceedings of the Subcommittee on Population Health
Issue 5 - Evidence, April 30, 2008
OTTAWA, Wednesday, April 30, 2008
The Subcommittee on Population Health of the Standing Senate Committee on Social Affairs, Science and Technology met this day at 4:18 p.m. to examine and report on the impact of the multiple factors and conditions that contribute to the health of Canada's population — known collectively as the social determinants of health.
Senator Wilbert J. Keon (Chair) in the chair.
[English]
The Chair: We are now duly constituted and have a quorum. Senator Fairbairn will have to leave us early tonight because she is going to the opening of the new heart institute, the Mazankowski Alberta Heart Institute in Edmonton. I am green with envy because I was also invited but I cannot go. It will be a wonderful event and she will represent us well.
We are delighted with the panel we have this afternoon. We have you here to address the questions on pages 11 to 13 in our Population Health Policy: Issues and Options report. If you would like a copy of that report, we can get you one.
From Statistics Canada, we have Mr. Michael Wolfson back again. From the Institute of Population and Public Health, we have Dr. John Frank. From Human Resources and Social Development Canada, we have Mr. Cliff Halliwell, and from Indian and Northern Affairs Canada, Mr. Marc Brooks.
We have a truly outstanding panel and we will exploit them to the full. The initial reason we have requested you to appear is to address the issues and options from page 11 to page 13. Once we discuss those, if you will all collectively forgive me, I will lead you into other, broader discussions, since the committee has this unique opportunity to have your expertise in front of us.
Michael Wolfson, Assistant Chief Statistician, Analysis and Development, Statistics Canada: Thank you for this opportunity to help inform the very important work of your committee. There should be a complete copy of my text and overheads in your package.
[Translation]
At the invitation of the committee, and given the role of Statistics Canada, I will focus my remarks on the data and research portions of the first report. The committee has rightly noted that high-quality data and information are fundamental to appropriate and effective health policy.
As other witnesses have noted, Canada already has quite good and often excellent health data and information by international standards. However, by the standard of what is feasible, and what we need as a country, my view is that we still have major gaps. So what I would like to do in the next four slides and two diagrams is briefly outline the most important of these gaps.
[English]
The first basic information need is the actual health status of Canada's population. We have made major progress in the last decade, in particular with the Canadian Community Health Survey. At the time we conceived this survey in 1999, we thought it was huge, on the same order as the survey that gives us the monthly unemployment rate. By the second cycle, we have had provinces purchasing additional sample and increasing demands from health regions for even more geographically detailed data.
As you heard last Friday, just under two weeks ago, one key aspect of effective population health strategies is intersectoriality. Much of the intersection of policy domains happens at the local level. While we have excellent national and provincial data, we lack strong data at the local level.
Moreover, we still largely have two data solitudes, what I call the clinical and the vernacular. We rarely collect data on the whole person. Instead we have fragmented data collections that in most cases focus on health status defined in biomedical terms and less often in terms of functioning and handicap. Ideally these would be brought together.
The second major gap is in the area of health outcomes, which Mel Cappe mentioned two weeks ago. I think clinicians will agree that much of what is done in our health care system is not routinely evaluated even at the most basic level of whether or not the intervention actually improved the patient's health. There seems to be considerable confusion about the term "health outcome,'' but it is really very simple: we assess an individual's health status both before and then some time after the intervention. If health status has improved, then we have a good outcome. The bad news is that when these kinds of data are collected, we sometimes find non-trivial examples where health outcomes defined in this way are nil or even, unfortunately, negative.
Beyond regularly describing the health of Canadians and routinely assessing health outcome, another major gap in information is cohort data, longitudinal data that follow individuals over time to track how their health status is changing, and what factors — both upstream risk factors and health care treatments — seem most strongly associated. I recall a conversation about umbrellas and rain from David Butler-Jones last time, and causality and association. We need these kinds of data to disentangle and to help support that kind of determination.
This kind of information is fundamental to understanding the most important determinants of health and which interventions are actually having beneficial effects. I would note that there are cohort studies and cohort studies, some for population health and some more oriented towards genetics.
The committee has also asked about health goals, on page 15. Let me offer two thoughts. First, a set of national health goals, provided they are not simply generalities but actually have some substance, would play a major and highly beneficial role in focusing health information development. A notable example is the decadal "Healthy People'' exercise in the U.S., though it has its pluses and minuses.
Second, it would be extremely beneficial if one of these health goals was an overarching summary measure of the population's health — in essence, the overall bottom line for assessing health system performance in a way that encompasses the full breadth of health determinants, an analogue of the role profit plays for individual firms and GDP plays for the economy. Such a measure exists and has been growing in acceptance internationally. It is called health- adjusted life expectancy, HALE, or something similar.
There is a graph on page 5, which I showed this committee last year, illustrating the idea of HALE. Paraphrasing the Quebec Rochon commission of quite a few years ago, HALE is concerned not only with adding years to life, which is ubiquitously measured using life expectancy, but also with adding life to years. HALE combines length of life with the healthiness of those years of life. Given the interest of this committee in disparities, you will note that the graph on the right shows HALE broken down by income thirds, and the little brackets show that the difference between the top and bottom thirds is on the same order as that attributable to heart disease and lung cancer combined. It is not small.
The next graph, which could be called a butterfly graph, shows further that HALE in some cases gives a profoundly different impression of where the real burdens of ill health lie. For example, the bars at the top left show the effects of eliminating ischemic heart disease only on mortality: it would add 2.4 years to the life expectancy of men and 1.8 years to that of women. The bars on the right show the effects of eliminating a disease on both mortality and health status, in other words on HALE. The results are similar in the case of eliminating ischemic heart disease: HALE would increase by a bit less: 2.2 years for men and 1.5 years for women.
Generally, cause-deleted life expectancy and cause-deleted HALE look similar, with heart disease at the top and cancers following in importance. However, let me draw your attention to the second last row, where osteoarthritis has essentially zero impact on life expectancy on the left side of the graph, but has moved to number one in terms of HALE for women at 2.4 years, well ahead of heart disease, lung cancer and breast cancer. Also, mental disorders have almost tripled, from a reduction of 0.4 years in terms of life expectancy to a decline of 1.1 years when measured using HALE.
This chart and the previous one illustrate that HALE, as a kind of GDP for health, could provide not only an overarching goal but also a concept that can be broken down in a number of illuminating ways, including equity.
It is one thing to have information on the past; it is another to have projections and simulations of what would happen if one or another policy were adopted. Simulation models have been a fundamental part of economic, tax and social policy information systems since the 1960s. No self-respecting finance minister brings down a budget without having run the numbers through a simulation model. I know; I used to do that.
However, I am sorry to say that this kind of quantitative analysis remains sorely lacking in the health sector. While a profound synergy has evolved between data systems and simulation models in these other areas — for example, with the national accounts — it is just barely beginning in population health, led by researchers, not governments.
One of the key but still unrealized synergies, both for simulation analysis and more broadly, is to ensure that health care and clinical data systems, as they evolve, are embedded in and have a full population health perspective. Billion dollar investments — for example the Canada Health Infoway — are underway in electronic health record systems, but very little thought is being given to their tremendous potential for population health.
I do not want to overstate it. There are the beginnings of thinking, and I think Ms. Glenda Yeates emphasized that point before this committee two weeks ago.
In this connection, let me highlight the longitudinal health and administrative data, LAHD, initiative, which Statistics Canada is undertaking in partnership with the provinces both to bridge the clinical and vernacular solitudes, which I mentioned earlier, to create more powerful longitudinal cohort data, and more generally to begin building out from the routinely collected administrative data to embed these data in an explicit population health framework. We are doing this by linking, where our survey respondents have agreed, our hundreds of thousands of survey results with their provincial health care records.
To conclude, instead of more information gaps I will highlight three major opportunities.
First, the electronic health record, or EHR, offers tremendous potential for population-level data. However, to realize this opportunity fully, there are two basic needs: to strike an appropriate balance with regard to the protection of privacy and confidentiality and to broaden beyond the biomedical clinical framework. On the privacy point, I worry that the pendulum is too far over on the side of restricting access to these data from bona fide and indeed highly beneficial uses.
Second, as in the LHAD initiative I just mentioned, Statistics Canada, with its constitutional and legislative foundation, offers the potential to form the hub for privacy-sensitive but much more aggressive development and utilization of emerging population health information.
Finally, I have a comment about the way we fund population health research. In a phrase, I find it too "lone ranger.'' We have not created the funding institutions that treat this area as big science, the way we do high energy physics with its cyclotrons or genomics with its very large investments in sequencing and bioinformatics. Just as there is a hot new field of systems biology, I think there would be tremendous benefits to a systems approach to population health research.
With that, I have touched lightly on a variety of points. I will be happy to answer questions. Thank you.
Cliff Halliwell, Director General, Strategic Policy Research Directorate, Human Resources and Social Development Canada: At the outset, I want to say that I am speaking here today as a policy research executive in a department that does not often deal directly with population health per se. With that being said, I am a policy research director in a department that faces issues that are not dissimilar from those that arise when looking at the field of population health. While the objects of the policy tool kit — the outcomes we wish to influence — are different, many policy tools in the tool kit are similar and, in particular, many of the feasible research and data tools are similar in character.
Just as investing in human capital, achieving better health outcomes means changing something that someone is doing through some policy instrument. Knowing how to do that requires more than just measurement and more than just research on the determinants of outcomes. It requires information on the effectiveness of interventions — what works.
I want to start with the first topic for discussion from the subcommittee's issues and options report: whether to expand and enrich the population health database in Canada. I have several short points here.
First, we have had a recent surge in available Canadian data ensuing from the three rounds of the Health Information Roadmap. We are vastly better off than we used to be. Second, Canada's health data is under-researched as it is. More data without more research would likely exacerbate this outcome.
Third, when looking at population health data, more and more cross-sectional
data — the data that captures a different sample of the population every year or
so — will only enable us to learn more and more about the largely contemporary
correlates of health outcomes. It will not enable us to look at what caused
those outcomes sometimes in the immediate past and sometimes in the long ago
past. For that, as
Mr. Wolfson has already mentioned, we need the longitudinal data that follows
individuals over their life course. That enables us to see what caused the good
or bad outcomes.
Fourth, reinforcing Mr. Wolfson's message, such longitudinal data would ideally be a mix of survey data and the administrative data captured within our health care system. This would be and is an enormous challenge both because of the complexity of administrative data, often not designed with this purpose in mind, and the privacy provisions that govern such data.
Finally, I would like to flag that we are facing a window right now that is currently open that could enrich our health data. That is, as Mr. Wolfson mentioned, the work going on at Canada Health Infoway to try to create a pan- Canadian electronic health record. If the potential use of that kind of information for research was left as a side issue, it would be an opportunity lost. I commend that to the subcommittee, although it sounds like you have been exploring that as well. It is an important window of opportunity.
Now I want to turn to the second issue of investing in more population health research and enhancing the translation of knowledge. Here I want to stress that there are two things we can do, one of which I think would be a mistake. The mistake would be to invest mostly in learning about the general determinants of health outcomes. That is the easy road, but we already know a lot about this. What we need to do is fill the enormous gap in our knowledge of what public policy interventions work.
This starts to sound like program evaluation, which it largely is, but it is unbiased program evaluation adhering to high standards of quality. It is also done to consistent standards of methodology so that one can have confidence in relative benefit-cost ratios of different interventions. In Canada we have underinvested by a substantial margin in unbiased, high quality, peer-reviewed, dispassionate effectiveness evaluation, especially in the population health field.
In particular, we know little about how to compare the benefits and costs of one intervention over another. Health advocates are often reluctant to admit that thinking about cost-benefit analysis is appropriate in the health field, because better health is often thought to be an outcome beyond price. However, asking whether existing resources would have a better health return when used in other ways is a legitimate question. Without this knowledge of what works, a good understanding of the general determinants of health outcomes will be intellectually interesting but not that applicable at the "coal face'' of public policy decision making.
In that vein, what strikes me is the contrast between biomedical and population health research. In biomedical research, as everyone here knows, the norm for research is to have randomly selected intervention and control groups, administer the intervention to the former only and then test what difference it makes by comparing the outcomes for the two groups. This tends to lead to robust results, especially when done with large enough samples. It is also easy to explain to decision makers because it is effectively a variant on the "Pepsi Challenge.''
It also enables identification of dose-response relationships, which is how much the outcome depends on the dose of the intervention. We do not typically look at that in the public policy context, but one cannot imagine not having that kind of information in the context of deciding how many antibiotics a patient should take.
In the U.S., there is a long history of applying such methodologies to labour market, education, skills, social policy interventions and even health promotion. These are often referred to as social experiments. In many respects, in social sciences research they are the new norm or standard for research worldwide, what an iPod is to a ghetto blaster.
In Canada, my department pioneered the use of such experiments in our policy domains in the 1990s, but only to a limited degree. I have often wondered why Canada does not have a greater hunger to try such experiments in public policy. I have been especially mystified by why we have not seen the population health field and the upstream intervention field as one that lends itself to applying these techniques.
Instead, even when we do thorough quantitative program evaluations of our interventions, we struggle to identify what the appropriate control group is, the group that did not get the intervention, a group we are not directly observing. We try to do that econometrically, with results that, when done through different approaches, prove to be highly sensitive to the chosen approach and subsequently very fragile. Of course, the results do not address this question of dose-response. For sure, trying to explain the econometrics to decision makers is a challenge of enormous scale.
These experimental approaches are not found only in medicine; they are increasingly found in the business world. I recently read the book Super Crunchers by Ian Ayres in the U.S.; it which deals with how the business sector is using more of this kind of data and randomly shocking their customers into seeing how they respond to different interventions. Their purpose is somewhat different from ours, but the power of the technology is there.
Finally, on the translation of knowledge, I know that researchers often complain about the uptake into decision making, but that will be inevitable if largely they can talk about determinants but cannot talk about what interventions will work. I would flag that decision makers themselves do need their own research capacity to help to be able to talk to the researchers, in a sense to serve as their Rosetta Stone and translate research into a known official language, either English or French.
I would also flag that we must treat the knowledge as a stock. Quite often in the research community we have created a conception that we need to firehose our research at decision makers as we discover it. In reality, we need to treat it as a stock and have it there readily accessible, in the sense that you have a repository in the Cochrane Collaboration, as an example.
Finally, subcommittee staff wanted HRSDC to comment on its website on the indicators of well-being in Canada. Right now that would not directly advance population health research within policy research and policy development elites. They already know everything that you would find there, all of which is findable elsewhere with little additional effort and is an ongoing part of health system reporting elsewhere.
However, this kind of site can indirectly advance the policy agenda by means of making interest groups, the media and the general public more factually informed and more outcomes-oriented. This in turn helps raise the standard of debate. It also helps generate support for the data, as data and research are like potato chips, where once you have opened the bag it is hard to just eat one.
Dr. John Frank, Scientific Director of CIHR, Institute of Population and Public Health: It is delightful to be here again. I will try to comment succinctly on at least three areas in which I think better data capture would be both doable in the span of the next half decade or so and something that the subcommittee could specifically advocate.
I found everything that you wrote totally compelling and I do not disagree with anything in your draft. I certainly think you could even call the two possible options, as you so quaintly put it, essential actions, because that is what they are: to expand and enrich the database for population health in Canada and to invest in more appropriate research in knowledge transmission. Those will come up again.
My slides start by reiterating three basic public health principles, the first thing I teach to 120 incoming master's and doctoral students in public health every September at the University of Toronto. First, seek the root causes of why people get unwell. You have heard that we need to bring longitudinal data to bear on that, particularly throughout the stages and phases of human life.
Second, think of whole populations. A great deal of conversation goes on in Canada about the people who come to the clinic door. Many of the problems lie with the people who do not come to the clinic door. About one in two type 2 diabetics over age 70 are known to the system. One in two are not known.
Third, understand and apply the principles of social change over the life course. If you take the example of smoking, you know full well that it is social change that has made the difference. The issue is how to engineer that, how to make that happen for other issues, other kinds of determinants.
The first example is children and their level of function, their level of well-being and health, as opposed to how frequently they die and how frequently they are admitted.
The next slide is entitled "What Data We Collect — and Don't.'' The truth is that we are fantastic at the same things you can learn in a cemetery. We know when you came into the world and when you left it. We are not too bad at why you left it. We are fantastic at the two, three or four times you are hospitalized in a lifetime now, on average. Many admissions occur in a very small group, but for most of us we are hospitalized about that many times until we are elderly. It is just all the rest, that vast mass of daily existence, that we do not know much about. That is a problem.
We still do not have any national system of collecting function of children. I know Dr. Hillel Goelman came from Clyde Hertzman's Human Early Learning Partnership. I do not know whether he presented to you their approach in B.C., which is to have every kindergartener assessed every year by a simple instrument the kindergarten teacher fills out. It takes about a day of the teacher's time. They got it through the collective bargaining, which was amazing, in 59 school districts, including all First Nations communities. The beauty of it is that it is so smart because it is done before the school gets the kids. Consequently, the schools love it because they will not be held to blame for the results. Everyone loves it. It is not used to manage individual kids. That is an important point. It is used to develop the means and distributions for each community, and then they are fed back that information. Clyde Hertzman and his staff go all over B.C., school board by school board, and pick comparator communities at the same level of socio-economic status. They can pick communities even within the First Nations group at the same level so that you can see how you are doing compared to appropriate comparators.
The examples I have given show you that they can map at a very fine level how kids are doing going into kindergarten. That predicts almost everything else that goes on in your life, I am sorry to say. You can see that some parts of Vancouver are much worse off than others, and the range is enormous. There are areas with only 6 per cent, 7 per cent or 8 per cent of children scoring in the bottom 10 per cent of one of the five scales on this little instrument. There are other communities inside Vancouver where over 35 per cent of the kids are scoring in the bottom 10 per cent on one or more scales. Human disadvantage and human function are clustered by how well off you are, how educated your parents are, and how much stimulation you get, as the subcommittee has heard from many people.
On the next page, you see where the licensed child care and preschool spaces are. Where do you think they are? There is a very famous law of health care, which is shown here to be also true for early childhood education, called the inverse care law. It was written by a famous general practitioner in a Welsh coal mining village, Dr. Julian Tudor Hart. Dr. Hart coined the inverse care law by pointing out that the amount of care people get, unless there is a carefully managed public sector system to redistribute it, is inversely proportional to the amount of care they need. That is the same with daycare.
Turning to the next slide, the map is to remind me again to tell you that this is throughout B.C. after more than a half decade of tremendous effort by the health group at the University of British Columbia.
Here are several slides from the fabulous Manitoba centre. I know that Noralou Roos came here to present to you. They have the best longitudinal data linkage capacity across every sort of record, although B.C. is coming up quickly to match them. Theirs has been in place longer. They have records going back almost 30 years for some people, and they can link all of those databases.
This is the work of Marni Brownell, junior scientist in the group there. Ms. Brownell asked the following questions: How steep do the socio-economic differentials in school achievement look if you take what the school test results look like and you map them against how well-off the families are in each of the areas using the mean local area family income, just like many other analyses in Canada? Then how steep do the socio-economic differentials look for the same test results if you include all the kids who were not in the test-writing exercise because they were not in school anymore, by the same socio-economic groups?
Looking at the dark graph with the bar graphs on the left, we see that for Grade 12 the gradient does not look too bad between the poor and the rich parts of Winnipeg, I think it is. We go from 70 per cent passing in the low SES group to 90 per cent and a little more passing the Grade 12 standardized language arts test. Too bad it is completely unreflective of the kids born that year. The real truth is that when you include all kids born that year, which they can do in Manitoba by linking up the data sets, the percentage who passed the grade 12 test by that age is 27 per cent in the low SES area and about 77 per cent in the high SES area.
Do not get too fixated on the debate about how we are doing compared to last year, because school performance statistics are a function of who is sitting the tests. That is a function of who is in the class. That is a function of who is still in school.
The amazing thing to me is that Manitoba people showed it is almost as bad by Grade 3. The next graph shows that the gradient is four times as steep, if you have all the kids in the birth cohort instead of just the kids sitting in the Grade 3 classes sitting the tests. The gaming like that is not all deliberate. I am just telling you that you cannot believe what the school system tells you about the generation. It is not telling you the truth. It is not entirely the school's fault.
Cardiovascular disease is next. Here my plea is simple. Dr. Keon knows this. He has heard me say this before, but people still do not get it. We are fabulous at the people who shuffle off the mortal coil with heart disease and stroke. We could tell you where they died, and we are pretty good at telling you the details of how they shuffled away. We can even map these things by the mean family income of their neighbourhood, which Russell Wilkins at Statistics Canada has done brilliantly.
You will see that for men in the bottom graph, the gap has gotten narrower, but it stopped getting narrow for the last 10 years of the last 25-year period; it stalled.
For women the system does not work. It shows no gap because the mean family income misclassifies women's real social class. It does not tell you their real social class because it has mostly been reflective, in the people dying of heart disease, of male income for the first half of the last century. That is who is dying, so that is who makes up the statistics.
We must do it better for women. That means you have to collect individual-level data, as Mr. Wolfson and Mr. Halliwell indicated. You have to do so in real time, and the data must include more than family income. It has to include something about the person's education and, preferably, occupational status, for which we have many standard scales.
However, the real problem is that we have heart disease programs to prevent and promote health after program going back 25 years in this country and we still have no idea whether they are reducing the rate at which heart disease occurs, because we collect no data on the rate at which it occurs. No data. It is not that hard to do. It requires a linkage between all the sudden deaths out of hospital, which are held by the death registry people, and all the hospitalizations, which are held by a different ministry in Ontario with a very different set of motivations and kept at the Canadian Institute for Health Information, CIHI, and it requires linkage. There are people standing in the way. There are little ministers and little ministries standing in the way of the linkage.
Let me ask you a question. If you do not know how much heart disease is happening every year and you only know the death rate, do you know whether we should be spending more money on prevention or on diagnosis and treatment of people who come to the emergency with chest pains, whom Dr. Keon spent so much of his life taking care of? You do not have a clue. You do not have a clue whether we are behind on prevention, and you certainly do not know whether the health disparities, which are still significant for this, the major killer — it still will kill a third of us in this room — are because we are bad at prevention for the disadvantaged groups or because we are bad at diagnosis and treatment. We have some data on the latter, thanks to the good work at ICIS and others, but we could fix this. This is not difficult to do record linkage on. A registry would be better, but cancer got to the trough first on registries.
By the way, do you know why that is? Do you know why we collect current data on only one major disease, cancer? The pathologists realized it would actually be a bad thing for there not to be quality control on people having a cancer diagnosis. They very far-sightedly insisted 30 years ago that registries be created so that the evidence that a person has cancer would be standardized and pulled in from every part of their medical chart: pathology, surgical biopsy specimen, lab tests, imaging tests. It is brilliant; however, it is not done for any other major disease. If the disease is not always admitted with the person alive, you cannot use hospitalization data.
What would it look like if we did put the data together and got some sense of how many people have heart disease in a given moment, which is slightly different from how many people are getting it, which we call prevalence? The Manitoba centre has tried this. They can do that because they can pull out the drug data, as many drugs are specific for cardiovascular disease. They show that those death rates have been coming down for almost 60 years, but the number of people each year with the disease is the same every year. That is because, of course, we are keeping people alive longer.
The bill cannot go down because the bill is driven by all these drugs. They are the number one class of drugs, costing us more than everything else put together. You are on hypertensives for the rest of your life, any hypertensives. You are on that that statin for the rest of your life.
The bottom line is, we are not even in the driver's seat on this, and it is mostly a federal-provincial-territorial dispute. It needs to be pulled into some kind of a common, overarching structure. Statistics Canada could do a lot, and CIHI can do a lot, but they need to be given the power to do so.
I will finish with this: It is great that we do have very good quality data on smoking — self-reported, but pretty good. We are now getting measures in the Canadian Health Measures Survey; we will have some other measures from people, urinary cotinine and other things, to make sure we are getting the full picture of their smoking status. Smoking is by far the most important premature unnecessary cause of death. It is important that we understand not only that but also the dynamics of who is quitting and starting every year by age, by gender, by social class and by location. Those are the determinants.
If you could stop people from smoking until the age of 21, 95 per cent of them will never smoke through the rest of their lives. That is not such a hard window to hit, is it? Surely we could get that right. However, we do not disaggregate the data by who is starting and quitting and what the recent dynamics have been. That is the first problem.
The second problem is that we have not systematically collected what the British have done. The reference is the very last graph. They have 50 years of time series by social class. The picture it shows, in this case for women, is not a pretty one. It shows that whereas women in the various social classes in the U.K. all smoked about the same in 1955, poor women and women who are less educated now smoke four times as much as the most educated and wealthy women. I am sure it is no different here, and maybe worse in some communities.
Those are the kinds of data we need. Without them, and without your support as a subcommittee to make the right sorts of investment, we will not be able to steer the ship.
Marc Brooks, Director General, Community Development Branch, Socio-economic Policy and Regional Operations, Indian and Northern Affairs Canada: I would like to thank the chair and committee members for the opportunity to provide an update from my last report that I gave last April. I will primarily touch on housing and water as they relate to the social determinants of health for First Nations people living on reserve and will comment on how I believe our work contributes to tracking health outcomes and supporting research on interventions to enhance the health of the population.
Perhaps one of the ways in which Indian and Northern Affairs Canada, INAC, most directly influences population health stems from our continued effort to improve access to safe drinking water in First Nations communities. I recognize there are many other determinants, but I will focus on the basic ones that I am involved with.
Last year, when I addressed the committee I spoke about our department's activities to improve access to clean, safe drinking water in First Nations communities. I reported at that time that the number of First Nations communities with water treatment systems classified as high risk had been reduced from 193, in the year 2003, to 97. In that context, high risk means primarily that there is no built-in redundancy or backup, so if something goes wrong, the probability of being able to correct it to ensure that the drinking water is safe is extremely low. Over the past year we have been able to further reduce this number to 85. In addition, the number of First Nations operators who have achieved what we refer to as first level or greater certification for water treatment operators has increased to 41 per cent. I realize this number looks a little low, but this is actually starting to be an improvement. This is one area that is in extremely high demand across the country. Water treatment operators are in extremely high demand, and we are having a rough time retaining them in many communities. They are leaving for greener pastures.
[Translation]
At the beginning of April, the Minister of Indian and Northern Affairs announced the next step in our plan to ensure that safe drinking water is made available to all First Nations. He outlined a plan, and $330 million in additional funding, for the First Nations Water and Wastewater Action Plan.
[English]
A key part of that plan will be to conduct a national assessment of all water and waste water systems in First Nations communities across the country. This assessment will help provide us with a big-picture view of all drinking water systems and help determine what type of investments need to be made in communities. The results of the assessment will be made public next year and will provide the foundation upon which future water and waste water investment decisions will be made.
[Translation]
We all know how important it is to have properly trained and certified individuals operating these water treatment facilities. Qualified operators not only ensure that residents receive a safe supply of water, but they also play an important role in preserving the infrastructure assets. Consequently, we have doubled the scope of the Circuit Rider Training Program, which provides hands-on training for water operators in their community by certified professionals.
[English]
Moving forward, INAC plans to promote the construction of water systems that are within the technical, financial and managerial capacities and capabilities of the community to operate them, thereby reducing the overall health risk long term.
Many rural and small-town Canadian homes rely upon individual wells and septic systems for their water and waste water needs. Over the past several years, the trend in First Nations communities had been more to deploy centralized piped systems. My department is currently developing clear standards to guide First Nations communities in the planning, design and operation of smaller water and waste water systems, such as individual wells and septic systems, where appropriate and feasible. Onsite water and waste water systems also provide a more cost-effective solution, which is critical given the finite resources available for community infrastructure.
Finally, INAC and Health Canada have committed to engaging First Nations communities, the regional and national organizations, and provincial and territorial governments on the creation of a federal legislative framework for drinking water and waste water on reserve. Legislation would provide the Government of Canada with the authority to fill the regulatory gap that currently exists between on- and off-reserve water systems. This important step will provide First Nations communities with the same level of water system protection that is currently available to all other Canadians.
[Translation]
Now I would like to speak about the Government of Canada's role in First Nations' housing, which includes new construction, renovations, social housing, capacity development and other housing-related activities. The Government of Canada is an important player in this area, but it is also working to make its contribution more effective.
[English]
The government has committed to reviewing INAC's 1996 On-Reserve Housing Policy. Key to this review is the recognition that communities vary in their ability to provide good housing to their members. The goals of the renewed policy are to ensure that housing support reaches those in greatest need and that the full range of housing opportunities, including market-based housing or private home ownership, becomes more attainable to many First Nations members.
The recently announced $300-million First Nations Market Housing Fund will support First Nation home ownership on reserve and expand community members' control over housing. The fund will be governed by a board of trustees drawn from three areas: First Nations, the private sector financing community and the federal government. After a publicly advertised process, the trustees have now been appointed. They recently convened their first meeting and have approved the access criteria and operating parameters of the fund. The fund is slated to be open for business shortly. In fact, it will be announced next week, in May 2008.
Market housing can bring many benefits to First Nations communities and individuals, including employment growth, based on new home construction, a reduction in housing shortages, pride of ownership and the means to build equity and generate wealth. It can lead to improved housing quality and, therefore, improved quality of life. The market housing fund will assist with the generation of 25,000 housing units over the next 10 years.
[Translation]
Increasing the availability of homes in First Nations has to be supported by maintaining the existing housing stock. Mould in homes presents a great risk to First Nations' health and safety. In general, mould is more prevalent on reserve due mainly to overcrowding, poorly constructed houses, a lack of appropriate maintenance, chronic flooding, and inadequate ventilation.
[English]
The First Nations indoor air quality committee, with representatives from INAC, Health Canada, Canada Mortgage and Housing Corporation and the Assembly of First Nations, developed a national mould strategy in partnership with many key First Nations stakeholders. The strategy includes such activities as an awareness campaign, technical training, community self-assessment and an identification process to assist First Nations experiencing critical mould problems in their respective communities.
Just as the committee emphasizes the potential of raising public awareness about, and education on, population health issues, Indian and Northern Affairs Canada has been active in raising public awareness on the living conditions and challenges affecting the well-being of First Nations. Our department is proud to be one of the sponsors of Closer to Home, a six-part documentary series that takes viewers into reserves across Canada to experience home, housing and life from a uniquely First Nations point of view. One of the goals of this series is to generate among the public, First Nations and non-First Nations, a more balanced understanding of what conditions create and sustain good health.
[Translation]
I would also like to focus on how our current work contributes to strengthening the intersectoral action proposed by the committee, and to the establishment of the population health database.
[English]
Indian and Northern Affairs Canada recognizes the gravity of complex social, economic and health challenges experienced by many First Nations and the need for equally comprehensive interdisciplinary responses to these challenges. By partnering with other federal departments, such as Health Canada, and learning from successful, traditional, holistic approaches adopted by First Nations themselves, the department is developing innovative policy and comprehensive community planning approaches to provide effective supports to First Nations communities in crisis.
With respect to the recommendations pertaining to data and research, we fully endorse the idea that improved data transfer and housing management practices are critical to increasing the numbers of better-maintained, healthier and safer homes on reserves.
INAC has developed an integrated capital management system that enables more effective and efficient collection, monitoring and reporting of all capital activities on the reserve, including housing, water infrastructure, community- built buildings, and management of these assets. The system can feed into a comprehensive health database by contributing data on how community infrastructure and its management affect health and well-being of Aboriginal Canadians. It will also be a source of focused information to support First Nations health research in such areas as drinking water and indoor air quality.
Further, our research and analysis directorate will be updating the First Nations Community Well-Being Index, which uses data from the census. Currently, the First Nations Community Well-Being Index tracks trends at the community level from 1981 to 2001. It is currently being updated to reflect the findings of the 2006 Census. This cross- sectional and longitudinal index includes data on educational attainment, labour force activity, income and housing quality. Components of the index measure the social determinants of health and well-being.
In conclusion, our vision is that Aboriginal Canadians, including First Nations, will have the same opportunities as other Canadians. We realize that despite our progress, there remains much to be done. Through this presentation, I have shared what I believe will serve to bring this vision closer to reality and will contribute to improving population health among First Nations people.
The Chair: Thank you all very much. I will now bring you back to the first possible option. Perhaps my question should be broken into two parts, so just tell me that as I put the question forth.
When it comes to expanding and enriching the population health database, I see a couple of components. First, we have some great databases — there is no question about it — such as Statistics Canada and the new CIHI initiative with the Canadian Population Health Initiative. The new Public Health Agency of Canada has a lot of data. I had the privilege of seeing some of it. Then, of course, we have the Canada Health Infoway, which is still in a gestation period, and all the provincial databases. If we could find an overarching way to get these pockets of information together, we would have good information on population health. The health component of it seems doable, if we could find the right way to accomplish this.
Let us move over to the other determinants that Mr. Brooks was talking about, such as water, food, housing, and so on. That is why I said there are probably two components to this question. Maybe we should address the health component first and then come back and see how we can pull the other eleven determinants into the database.
Could all the witnesses have a go at the question of whether we need something new and whether the resources are there in the existing resources? Is it just a question of how they can be integrated and made to talk to each other? Who has the tremendous wisdom to tackle that question?
Mr. Halliwell: I will address something that has bothered me for some time here, which I think could be a framework for this question. I saw it when I was in the United States, in Washington, talking with them about the work they do. It deals with how we do risk management in privacy in the field of research in Canada. I have written about this in Horizons, a Policy Research Initiative publication. We fail to make a distinction between a data set that has many individual identifiers, which is how most people think of protected personal information, and a data set that is stripped of those kinds of identifiers but has a reasonable amount of detail about the individual records in that data set. There is a minuscule risk that, as researchers, we will look at an individual record and say, "I think I know who that is. I know more about them than I otherwise would.'' If I describe it that way, and if I were to show you the kind of data set that one is talking about, it would not look like personal information. However, we have interpreted personal information to encompass all of that in one standard of a high degree of risk protection. Therefore, we are under-exploiting this data.
In contrast, in the United States they will strip the identifiers out of the data and do some aggregation of the data, but they do not reduce literally to zero the risk that, poring through the 3,000 records in the Canadian Community Health Survey, for example, you would discover your neighbour. That we do so in Canada has become a barrier to our capacity to do research in this domain. We are not doing the risk management of de-identified data sets used for research purposes, and it becomes a big cross. It results in data sets that are aggregated in such a way, often by geography, that even the capacity to not necessarily match things up individual by individual but by smaller, finer units of geography is difficult to manage.
We do have one out in this regard when dealing with individual data sets — for example, data contained in the Canadian Community Health Survey — and that is the wonderful work that Statistics Canada and the Social Sciences and Humanities Research Council have done in getting those micro data sets into Statistics Canada university research data centres, where Canada's researchers now have access to a wealth of data that was almost unheard of 15 years ago and can work with the kind of micro-level data that Dr. Frank described. However, as we try to knit these disparate data sources together, if we have this incredibly low tolerance for any sense of a risk of privacy loss, it may be an insurmountable barrier.
The Chair: Mr. Wolfson, who is equipped to deal with that?
Mr. Wolfson: There is no simple answer to that. I will comment on that as it is posed on page 12 in two ways: first, with respect to existing information flows and the way the status quo is likely to evolve; and second, with emerging things that should occur.
I would like to say that Statistics Canada can do everything, but that is not true. If we go back before the birth of CIHI, to the task force report that gave rise to it, the task force concluded, first, that our health information system was in a mess. It has improved considerably since then, but the report proposed a health information coordinating council. For better or for worse, that has not come to pass. We have the board of the Canadian Institute for Health Information. The other organizational recommendation made was to create CIHI, which was an amalgam of what used to be at Statistics Canada and part of what used to be at Health Canada, plus two predecessor organizations, one dealing with hospital patient data and the other with financial data.
The appreciation in the early 1990s was that the players are so disparate that they would include not only federal government departments but also provincial ministries and hospitals. Each group has important parts of the story but not the whole thing. Somehow, there is a need to bring them together.
The CIHI board has done a wonderful job, but it is legally the board of the private non-profit corporation that is CIHI. While it is good representation, I do not know whether it is the right organization to take that on. Consequently, we have a gap. There was the National Forum on Health and then a committee on information appointed by the Minister of Health in 1997, which led to the birth of the Canada Health Infoway. There were several efforts over the 1990s to come to grips with this, but none of them succeeded. Bringing all of these things together is an organizational challenge.
I agree with Mr. Halliwell that this privacy thing has swung too far over to one side. It is bottlenecking a lot of good potential work.
Statistics Canada used to work directly with the provinces and a lot of data flowed directly to us. With the creation of CIHI, however, the world changed. To eliminate overlap and duplication, they would flow the data to CIHI and then to us, but that has not worked famously. We have reopened discussions with the provinces. Notwithstanding what Mr. Halliwell said, to do the linkages we need the names and addresses. Once we have done the linkages, two or three people need to be able to see that; everyone else does not need to see it. We strip it off so that people have the substantive information without the nominal information. There is a subtle distinction between denominalizing — taking names off — and making things non-identifiable. That is where the risk management comes in. As I said in my remarks, I think Statistics Canada is well positioned, given the Statistics Act, to play an important role on the record linkage end of things and to bring together the vernacular and biomedical worlds.
Another concern is Infoway. I know a number of the people there fairly well, but until a year or so ago they were for good reasons focused entirely on the patient care perspective. It is only in the last year or so that we have had a number of discussions of increasing breadth about what some people call "secondary use.'' I do not like that phrase because it connotes that using the data for something other than immediate patient care is of secondary importance, which is not at all true. Because of their legitimate concerns about the public reaction on privacy grounds, they have been leery of venturing into talking about that.
There is a tremendous potential in the design of these systems and the incentives that Infoway is able to give the provinces as they build these systems to say that, "You must be able to do this not only for the immediate patient care but also bearing in mind the tremendous potential population health uses of these data.'' I think Infoway would will tell you that nobody has told them to do that, that it is not in their mandate. Statistics Canada cannot do that and CIHI cannot do that. The Infoway board presumably has to say that it is not what they were given money to do.
Who should do that? I think the first step is identifying that there is a question there.
We already have informal discussions — for example, amongst CIHI and Statistics Canada and the health council and in the future I think the Mental Health Commission of Canada — where we have tried to agree. We have indeed agreed so far that CIHI and Statistics Canada are principally in the data collection end of things and these other organizations, as they come into being, will not — except under particular circumstances — go out and collect their own data. Rather, they will use the existing infrastructure. Without any explicit coordination or direction, we have informally worked out a modus vivendi to try to minimize multiple data collection processes.
What you are hearing from me, Mr. Halliwell and Dr. Frank in particular, at least on the broader level — the Aboriginal issue is a special case — is that there are tremendous opportunities as we move toward more electronic information or as different things come down the road. However, there is no locus for the strategic planning of Canada's information system.
Dr. Frank: I am struggling with the same question. I think it is important to be clear about what is not in place. For example, almost every nation state in Europe, certainly those that do the best job of this sort of data collection, analysis and interpretation for policy purposes, has a national public health institute. They are very different than what I have run for seven and a half years at CIHR, because I have no scientists inside; I have four staff.
I have enjoyed it. It is the best job in Canada for someone with my credentials, but I do not get to do any research inside CIHR because it is not set up for that. Only 6 per cent of its budget is spent on anything other than flow through.
Where is the locus for pulling all this together? First, there is no legal framework to allow any agency or ministry to have the right level of authority to pull it together with the appropriate oversight, including ethical and legal oversight.
Second, it will not even be worth building if you do not attract the best minds. Right now, the best minds want to work in universities or hospital research institutes. They do not want to work in a federal agency. There are a few exceptions, including some at this table.
The younger generation worry that working for government will mean muzzling when there is a problem, when they find something that nobody wants to have aired. We have to change that by creating a structure — maybe a conjoint structure. It must have deep links to all of the structures that we have heard Mr. Wolfson and the others talk about.
We need a structure that has some autonomy. We need strong university appointments and then you can hire smart people to pull together in a mission-driven program — not single investigators at a time picking off what interests them. We do that great at CIHR. To get a mission-driven program, you need a governance structure in an actual institution whose job it is to do this for a nation state. I do not see any alternative.
The Chair: What do you mean by "institution'' exactly, Dr. Frank?
Dr. Frank: I use it in that broad way that Fraser Mustard uses it. You have to have a structure that has a legal basis for its authority and sufficient autonomy that it can tell the truth to the Canadian people and policy-makers; but it needs to be sufficiently arm's-length from ministerial work — that is, line ministerial reporting — so that it attracts really smart people who need autonomy or they will not work.
Mr. Wolfson: The system is clearly federal and provincial. That is a major stumbling block in terms of creating an overarching structure of any sort.
The Chair: That is the same as Infoway. You will recall the conversations we all listened to for days as we tried to create Infoway.
We really must address this. I am hoping that we come out of the meeting this evening with some concept that we can enrich. In other words, the resources seem to be there if we could get them together. Is that a reasonable observation? Are you unsure?
Mr. Wolfson: There are several different categories of resources. For example, several of us have made reference to the importance of longitudinal data. I understand we have a cancer cohort at the cost of $90 million. That is a large resource. It is not clear how much of it will be a population health rather than a genomics-oriented kind of cohort.
Infoway has money to get things started. They are paying a fraction, but the provinces are highly variable in their take-up and which areas they choose to do first.
Everyone will say that standardizing the information collected at the hospital, at the doctor's office and at the lab is a great idea; but what if you already have a substantial cost sunk into setting up your system in each jurisdiction, and nobody got there early enough to say you should collect the information in this common way?
We take for granted cause of death, because 100 years ago the World Health Organization developed a classification system. When you get into other areas — What is the reason you went to see the doctor? — there are many different classification schemes across the country. Manitoba can do something within Manitoba, but when you want to start comparing across provinces, those things become more challenging.
Money helps. Where is the incentive to say we would like, at the national level, to incent the provinces not only to develop these systems but to develop them in a way that has standards? Infoway is doing that at quite a range of levels, but from a population health point of view, I do not know that there is enough money. There is no money on the table for that.
In the research area, Mr. Halliwell was talking about doing experiments. Those are very expensive. There is no money for that.
Mr. Halliwell: As an observation on that, in a sense we are holding out this ideal that we would have a pan- Canadian system equivalent to the one that the Manitoba centre put together, but we are feeling like someone needs to take charge of it. If that happened and we could deal with some of the privacy restrictions, maybe we could make it.
My concern is why, with the wonderful example of what has been done in Manitoba, other provincial jurisdictions have not said they want one of those as well and made the investment to make it happen. It has happened in fits and starts; I think British Columbia probably has a good one, but other provinces have not really taken that step.
I am a bit concerned that even with national impetus and the privacy barrier in some sense surmounted, the failure to have 10 of these systems, which would be an enormous step forward from where we are now, might be symptomatic of something profound and deep about our lack of commitment to evidence-based decision making in this particular business.
Senator Eggleton: Dr. Frank, regarding the institute you were talking about a few moments ago, is there a model? I am trying to get a clear understanding of how this would work.
Dr. Frank: I would visit Scandinavia. All the Scandinavian countries have such a structure, where there is a strong intramural program with some of the top minds and strong ties to the university.
I hesitate to add something to the mix. It is duck soup. The list of organizations, agencies, structures, commissions and offices is bewildering, even at the federal level. You can hardly figure out who does what. It is not set up in the best way to secure the right kind of minds, the right data with the links to the policy-makers to ask the right questions.
Have you visited a few Northern and Western European models? It would be worth having a look.
The Chair: We have interviewed them by video conference, but we have not visited.
Dr. Frank: Do not bother to going to Scotland; they do not have it figured out either, and they have only 5 million people.
The Chair: I get the feeling we are close, whether we advocate the adoption of the Manitoba model or whether we try to create some overarching structure. No delusions, by the way, Mr. Halliwell, I agree.
Senator Cook: From my understanding, you have all of this wonderful data but no one is sure how to use it. Am I correct in thinking that way? I will finish with the second part of my question. It seems there are barriers to progress. Would you like to elaborate what the barriers are and how you would eliminate them, if you had a wish list? We are in a hurry to improve public health for the population of this country to live better lives. We have to eliminate the barriers and develop some realistic processes.
Mr. Wolfson: On your first comment about how to use the data, Mr. Halliwell put his finger on it. It is one thing to be able to describe the problem. We have reasonably good data that tells us at a point in time about the distribution of health status and where the inequities are. It is quite another thing to know what to do about it. The gap that Mr. Halliwell identified in intervention research strikes me as the bottleneck.
One of the barriers is privacy. I hesitate to use the word, but I will say the phrase "privacy chill.'' There is a legitimate concern that we have not quite struck the right balance in research, policy analysis and effective communication of information to the Canadian public. I do not want for a moment to minimize the importance of privacy and confidentiality, and certainly at Statistics Canada it is a deep part of our culture to be extremely careful and protective of privacy. However, we have to be able to use these data. As a data analyst, I like to say, "Make the data sing and dance.''
Another barrier, which we just talked about, is the institutional structures. There are many players — federal, provincial and local — and some things are at the discretion of a hospital or regional health board as opposed to a provincial or federal government. All of these players have to have a common view and be willing to work together.
Mr. Halliwell: What you have been after, Dr. Keon, is whether we have the bricks for the wall or are we just missing the mortar. Perhaps that was your question as well, Senator Cook.
My own instincts are that relative to the situation prior to the first Health Information Roadmap, we now have a great deal of bricks. We also have a lot of public support for this kind of thing were we to work around the barriers that come from the elites. For this, I will need help from Mr. Wolfson on the specific statistic. When people fill out their questionnaire for the National Population Health Survey or for the Canadian Community Health Survey, they are asked at the very end whether they would share this record with Health Canada for research purposes. My recollection is that about 95 per cent of them said yes, Health Canada could have that micro-record. They supported it. That is perhaps indicative of the level of trust in Health Canada as an institution, but it is also indicative of the level of trust that Statistics Canada has.
Similarly, I believe the National Population Health Survey for years has been asking if they can link people's OHIP or Quebec health record with this data in order to conduct research. My recollection is that there was a very high check-off rate on that as well.
The vast majority of Canadians understand this, providing you are talking about institutions in which they have a level of trust. They are willing to do it and to support it but, somehow at some higher level, we are not getting the key ingredient to begin turning this pile of bricks into a wall. I am not quite sure I understand it, to be honest.
Senator Cook: Surely the people we are seeking to serve have done this. Where do the higher-ups that you reference get a mandate to put a brick in your way?
Mr. Wolfson: We are engaged in the longitudinal health and administrative data initiative. It is precisely to take the 300,000 to 400,000 health survey responses where the people interviewed said that we could link to their provincial health records and then go to the province and ask for the records. We have a constitutional mandate, the federal Statistics Act and the peoples' permission. Yet it has taken us two years, and we are just on the verge of signing memoranda of understanding with at least three or four provinces. For each province, our lawyer has had to talk to their lawyer and go down through the Privacy Act and their health act and explain that it is legal, but officials in these ministries are nervous and their best way to minimize risk is to not release the data.
Senator Cook: Yet the people we would serve have said yes to doing it. Is Newfoundland one of those three or four provinces?
Mr. Wolfson: Actually, at the meeting we had last week Newfoundland was one of two provinces that entered into a race to see who could be first to sign the MOU.
The Chair: We will move away from the broad organization, which we hope to promote for population health, and get down to the community that we are zeroing in on now: Aboriginal health. I am aware of many of the challenges. When it comes to an information system that would be helpful in Aboriginal health, some areas are already beginning to create some autonomous data banks.
What should be the information technology approach to Aboriginal health? You have to tell us about what you are doing to gather information on air, water and food.
Mr. Brooks: The information systems I was referring to, the Integrated Capital Management System, it is not a health system whatsoever. It is more of a life cycle system to gather information, which we believe can be used to help to figure some of the factors of the determinants of health.
We do have the First Nations Statistical Institute, but several First Nations communities or First Nations groups have refused to participate in the census, which has basically left a void in activities. Data gathering has been difficult.
I am positive that Mr. Wolfson can speak more eloquently than I as to what is and is not happening. Part of the purpose of the statistical institute, since it will be run by First Nations, is to gather the right information. We are working very closely with Statistics Canada on this. There is definitely a vehicle in the statistical institute that can help gather some of the data.
In terms of gathering all the determinants you are looking for, the system we currently have, which, as I mentioned, is more of a life cycle system, is narrowly focused in the area I am responsible for, physical assets. It does not look at other aspects, recognizing that if you do not have good physical infrastructure, you will probably not have successful healthy outcomes. There are such things as housing, water, waste water and things of that nature.
I realize I am not answering the question the way you would like me to answer it, Senator Keon, because it is a little beyond the scope of what I do. However, there are systems at play that I think, from our perspective as a department, we could work with along with my colleagues around the table to improve the First Nations outcomes. This is more specific than just Aboriginal writ large, as most Aboriginals are off reserve and would probably be captured by the majority of provincial systems right now.
Mr. Wolfson: This is a challenging area, as I am sure you appreciate, not least because quite a few on-reserve Aboriginal groups have articulated a principle called OCAP: ownership, control, access and possession. From the point of view of a national statistical system, it would be unfortunate if a completely independent statistical system for on-reserve individuals were to emerge.
From the point of view of a national statistical system, two key principles are fundamental. One is that there be at least some common core of information that is collected in exactly the same way as for the rest of the country so that one can make reasonably valid comparisons. The other is that, from a statistical point of view, there are internationally accepted principles about objectivity, no political interference and, in effect, no prior checking or censorship of results from statistical endeavours and the publication of results. Hopefully the First Nations Statistical Institute can play an important role there.
A final point is that roughly two thirds of the Aboriginal population are not living on reserves, so they are within the scope of the data that Statistics Canada already collects, for example, in the Canadian Community Health Survey, the Census of Population and the post-census surveys of Aboriginals and Aboriginal children. There is quite a lot of data there.
A potential opportunity is that by and large provincial health care systems do not have any kind of Aboriginal identifier on them. It is not easy to imagine how they would do that. If we were not quite so nervous about privacy — and Statistics Canada is very stringent on rules and policies for this — maybe we could imagine some sort of record linkage program that would allow one to begin to generate an Aboriginal health information system including, for example, the health care encounters that would allow some analysis based on Aboriginal identity, at least as well as we are able to capture it on our various surveys.
Mr. Halliwell: I will disagree somewhat with Mr. Wolfson, and I will use an illustrative point I got from him against his point here.
I will say first that the anecdote he shared with me took place in the late 1800s, with a doctor I believe in Chicago, who began keeping card files on his patients so that he could find out how they were doing a year later. In some sense, there was no information technology involved in that. It was a simple case of wanting to acquire information.
I have a bit of a concern — and this is a personal view — that our efforts to try to create a world standard, coherent and statistical system, especially on-reserve, have tended to flounder so far, yet maybe the kind of information we really want to obtain is by smaller units, instead of being able to do league table comparisons to see how we did next to a comparable band or how we did relative to Canada as a whole, which is often fraught with political danger.
Maybe what one would really want to know is how we are doing relative to last year or how we are doing relative to five years ago. As long as you begin doing it on a comparable basis, even if you are doing it within individual communities, it forces you to take ownership of how you are doing over time in your own area. You will not be able to compare with national data or other jurisdictions, and that is often advantageous in forcing people to focus on issues — for instance, our department's experience with international comparisons of adult literacy and the attention it brings to the file. However, maybe something that starts out simple and is built over time will be their first potato chip, to carry on with the analogy from my remarks.
Mr. Wolfson: I do not disagree with Mr. Halliwell in one sense. I am trying to remember the names of the two British Columbia researchers who have been doing seminal work in B.C. about suicide rates. Their names are Chandler and Lalonde. They went community by community and looked at the suicide rate, which varied from zero in some First Nations communities to very high in others. This is basic information, so one does not have to have a fancy gathering system.
It is crucial, however, not only to be able to measure that in a way that is standardized but also to begin to pick out what the co-variates are, what other things are going on between these communities that account for these dramatic differences in suicide rate. There again, it does not have to be a high-tech or totally electronic system. What is crucial is that it is done in a systematic and rigorous way with agreed-upon concepts.
Senator Callbeck: My brief question is on the electronic health records. I know that my own province of Prince Edward Island in the last two years has made great progress, and I understand many other provinces have as well. We have had witnesses refer to a pan-Canadian system. That has been brought again up today.
However, I think I heard one of you say that someone has to grab a hold of this. Did I hear properly? Who is responsible for this system? Who is leading the initiative?
Mr. Wolfson: I hesitate to tread here. Dick Alvarez is the CEO of Canada Health Infoway. He has a large sum of money from the federal government and a board of directors that includes provincial deputy ministers; and he, his staff and his organization make a series of bilateral arrangements, as I understand it, with this province, that health region or that physician group to help fund the development of particular kinds of software, the initial focus being on diagnostic imaging, moving into labs and pharmaceutical information systems.
Each province is autonomous. There is clearly an effort on Infoway's part to make a software system once and encourage other provinces, if they are willing, to make a copy of it and adapt it so that Canada can realize economies of scale. However, there is a clear requirement, as I understand it, for interoperability, though I think that is as far as Infoway goes.
From the point of view of using the data in a Manitoba centre-like way across the country, nobody is in charge of that. Both CIHI and Statistics Canada are keen because we see it as a tremendous strategic opportunity to greatly improve the health information of the country, both from a population health and a health services information point of view, but we have no sticks or even many carrots. It is moral suasion.
Senator Callbeck: Regarding the money given out to provinces, is there no condition at all to apply it to a pan- Canadian system?
Mr. Wolfson: There is an information architecture. However, in terms of standards, I am not familiar with details. You would be best to talk to Infoway directly on that.
Mr. Halliwell: In some sense, the key to it will be interoperability and standardization of the kinds of metrics that are exchanged. When you get that, you do achieve interoperability.
There is clearly a whole set of constraints on what you can share, but part of me that thinks that, with the proper protocols, et cetera, it might become a bit like the Internet. No one asks who is in charge of the Internet. There is just a whole set of protocols by which information flows back and forth between what amounts to disparate data browsers and different web-hosting software.
Under no circumstances do I want anyone to get a sense that we want health information to be like the Internet and so widely available. However, it may be that with the right attention to a minimum data set, the information can start to get glued back together again as a discrete exercise that does not have to be done in real-time like an actual patient care system.
That possibility is there. However, the window of opportunity potentially gets lost if you do not think through how you would do that at an early design phase. I can remember being at the Department of Finance when we all did what I called the Homer Simpson "D'oh.'' We realized when the GST was introduced and businesses first registered for GST numbers that we did not find out what kind of industry they were in as part of the form they were required to fill out. Therefore, we lacked a set of useful information that it was hard to put in place subsequently.
Even if you are thinking that we do not know whether we have approval to do this or whether people will tolerate this happening, it does not mean that you do not build in the capacity to do it, so that you subsequently have the choice to try to use this data for research purposes.
At the level of Infoway and the processes around that, this ought to be a priority, but putting it in place is hard. There is an old joke: Why was the Lord able to create the world in six days and rest on the seventh? There was no existing installed user base.
The problem is that you are starting from somewhere, not starting from nowhere. I think there are technical solutions. However, there is a window of opportunity that we are in danger of seeing disappear if someone does not pay even a modest amount of attention to it.
The Chair: Dr. Frank, I do not want you to go abroad without giving your input on the Aboriginal situation. Would you please tell us your views?
Dr. Frank: I do not have any magic solutions. You have to accept that the long historical wave of events will keep many First Nations communities of the view that data collection is an opportunity to fight a fight that has historical roots with justice issues involved in it. Therefore, to de-politicize the data collection exercise, I do not think there is any choice but to adhere to the broad principles that have been enunciated by the First Nations communities that Mr. Wolfson gave us the acronym for.
In the end, people only want to cooperate with data collection if they get something back. They have to get back useful summaries that will help them with their own problems as a community, say. I think that means they have to trust the people who are collecting, collating and analyzing the data to do that.
That would mean that it has to have Aboriginal governance, whatever it is. That is a tall order, as Mr. Brooks knows. There are 600 or more communities. You cannot go any other way. All over the world, data is collected in health care systems; people are filling out little tick boxes, and we know most of it is garbage because they do not have any reason to do it carefully. They never get anything back.
What does some nurse in an African health centre like I used to work in as a young doctor in Tanzania get back? Nothing. They do it when it is required, but it has been shown many times it is mostly noise and not much signal. There is no benefit.
Whatever we do, we have to constantly ask what people get back for their participation, both individuals and also their local and provincial governments. There would have to be a legally-binding agreement by the provinces, territories and the federal government that core data collection for the electronic health record and others was in everyone's interests and that they trusted a governance structure, which currently does not exist.
You must get the lawyers of the sort we already heard about — who represent each of these as autonomous, governing structures that are held up and given those rights by our Constitution — to agree. If you do not get the lawyers to agree on the precise boundaries before you start, do not set anything up. You will not get to first base.
The Chair: Our time is up, so unless someone has a further question or a final comment, we will have to sign off. Thank you all very much. You were very generous to come and give us your time.
The committee adjourned.