If you’re looking for evidence-based health information in Canada, a lot depends on who – and where – you are. A physician in a teaching hospital? No problem. But family doctors in rural areas, nurses or physiotherapists have a much harder time accessing up to date materials. And the general public is more likely to find themselves turning to Dr. Google than trustworthy sources.
In Norway, these differences disappear. Rural physicians and motivated patients can read the same high-level information a doctor in the largest hospital enjoys. That’s thanks to the Norwegian Electronic Health Library, an eight-year-old project that allows anyone in the country to log in to over 3,000 titles, including the most important medical journals, Cochrane Reviews and clinical resources like UpToDate.
A similar initiative for a pan-Canadian health library recently failed. But Norway’s egalitarian approach offers a glimpse into what it could have been.
- Should clinical practice guidelines consider value for money?
- Doctor smartphone use: distraction or good medicine?
- OPINION: Peering into peer-review – implications for web-enabled scientific debate
The wealth, the will, and the way
Thanks to its oil reserves, Norway is one of the richest countries in the world. Despite that, finding funding for the library was a challenge. The money, which comes from the government, was diverted from hospital budgets, and the institutions were resistant to give up the income as well as the autonomy. (Some, like university hospitals, still supplement the electronic library with their own additional subscriptions for researchers.) “Everyone thinks it’s a good idea, but it’s hard to get someone to give up some of their own money,” says Magne Nylenna, the project’s editor-in-chief. He believes the library saves money overall. In a 2010 article in The Lancet, he writes that it cost $9 million Cdn, or $44 per health-care professional. Institutional licences would have cost three to four times that.
Canada could also benefit from the bargaining power that comes with buying in bulk. And a national licence would prevent duplication, where journals charge multiple times for the same physicians, who have access through several places, like a hospital and a university.
Recently, the Canadian Health Libraries Association tried to create the Canadian Virtual Health Library. The library would have offered all health-care workers access to a network of libraries across the country through a bilingual site. The project received an $800,000 grant from the Canadian Institutes of Health Research over three years, starting in 2010. But without federal support, it didn’t take root, and has since shut down. “The logistics of trying to get to a pan-Canadian licence on anything – just to get all the people at the table is crazy,” says Lee-Anne Ufholz, president of the Canadian Health Libraries Association. “Norway’s project is beautiful and I’m totally jealous, but they just didn’t have the complexity that we were looking at.”
‘The democratization of knowledge’
The Norwegian electronic library initiative is based on the principle that everyone deserves equal care, and that requires equal access to health information. “We call it the democratization of knowledge,” says Nylenna. “If [rural and urban areas] aim at the same sort of quality of care, we should have the same access to the same quality of information and knowledge.”
The electronic library is part of The Norwegian Knowledge Centre for the Health Services, which helps promote evidence-based health information throughout the country. The e-library lets users view information by medical specialty and type, or by using a bilingual search function that includes Norwegian and English. About 200,000 to 250,000 unique users access it per month, in a country of just 5 million.
Canadian physicians across the country have access to some resources through the Canadian Medical Association. Its searchable online databases offers access to resources including the Canadian Medical Association Journal, clinical practice guidelines and DynaMed, as well as dozens of medical journals and textbooks. “Twenty years ago, a family doctor would keep a copy of Harrison’s textbook of medicine and a couple of [other] textbooks on the shelf, and now that just doesn’t work,” says Sam Shortt, director for quality initiatives at the Canadian Medical Association. Its goal is to “give every doc coast to coast access to the key materials.”
However, some things are still missing. Cochrane Reviews used to be available, but the CMA found they weren’t well read and stopped offering them. UpToDate, which many physicians believe is the best resource on the web, and many high-end journals, including The New England Journal of Medicine, aren’t offered because of their higher costs. “There are some journal gaps,” says Shortt.
Cochrane summaries and abstracts are also available to all health-care providers and the public. But there isn’t a national licence for the full text. Cochrane Canada director Jeremy Grimshaw estimates a national licence would cost about $550,000 to $650,000, and that Canada is currently spending about $50,000 more than that on individual licences. “Patients expect that our family doctors have the most up to date information – the trust in family practitioners is very much based on that idea. And we’re not making it easy for family doctors to have access to that knowledge,” says Grimshaw.
The Canadian Nurses Association has a similar portal as the CMA’s, called NurseONE. It offers members access to resources including Cochrane reviews, 3,700 journals, including the BMJ and CMAJ, DynaMed and over 1,000 books. It averages 11,700 unique visitors a month.
Nurses, along with psychiatrists and physiotherapists, are one of the groups who benefited the most from the Norweigan electronic library. They’re also one group the Canadian Virtual Health Library was trying to reach. “There is a need for access for not just physicians …. but for nurses, or even social workers, rehabilitation sciences people, who are underserved in how they access information,” says Ufholz.
Quality over quantity
Norway’s electronic library focuses on quality, using only articles from trustworthy journals. One way they do that is by subscribing to McMaster PLUS, a service from Hamilton’s McMaster University that acts as a filter for high-quality sources. Staff review more than 120 journals to see if articles are scientifically sound, relevant and interesting. That helps make keeping up more manageable for doctors, who report being overwhelmed by information. “It’s this incredible reduction in the noise-to-signal ratio,” says John Lavis, director of the McMaster Health Forum. Norway’s version has been customized for them, but the service is free to anyone globally.
It’s even harder for the public to distinguish between high- and low-quality information. Information for patients needs to be written in plain English, like the Cochrane review’s accessible summaries. And screening out low-quality health information is crucial, as the media – and online sources – play up counterintuitive and sensational findings. In Norway, the electronic library helps solve that problem too, with increasing numbers of users looking at patient information sheets; many of the users are coming straight from Google.
“I often talk about diabetes, clinically, as starving in a banquet hall. Diabetes information is the same thing,” says Jan Hux, chief scientific advisor at the Canadian Diabetes Association. Despite being surrounded by millions of articles online, she says, “people are starving for information.”
The comments section is closed.
I agree it’s a very complex situation but curated access on a more equitable basis has to be a common denominator in our system. What I would really hope is that the Norwegian librarians have found a way to measure the before and after effects in health care costs – a challenging measurement to be sure but perhaps even at an overarching stats level like number of days in hospital and mortality it would be gratifying to see a difference.
As a member (albeit the least expert member) of the now-defunct Canadian Virtual Health Library board, I of course lament our inability to achieve a sustainable funding model for a valuable element of health system infrastructure. As the article implies, we ran up against an intractable game theory dilemma: while it was in everyone’s collective interest to take the padlock off the health information warehouse, it proved impossible to get all of the players to cooperate in lockstep to overcome the historical and structural barriers to advancing the cause.
That said, I don’t think universal access to the best available evidence is the fundamental problem, for several reasons. One is that while access to journals and related materials may be limited, access to high quality evidence summaries, consensus statements, and sound clinical advice is by and large not. There is a tonne of good stuff on the web, compiled by reputable organizations and searchable by condition or symptom. Indeed there is a lot of junk too, but it is not all that hard to identify the higher quality ones. Neither patients nor clinicians have to set foot in a physical library to access sound material, yet practice remains highly variable, safety is compromised, and resources are routinely wasted in Canadian health care. High performing systems do much better, and it’s not because everyone is up until midnight reading Cochrane reviews.
The health information production enterprise is vast – in fact, way too vast. Most of it is provisionally valid at best and the great bulk is eminently forgettable, deserving no impact whatsoever. If instead of over 20,000 health journals in the world there were 500 – roughly 10 each in every specialty area – there would easily be sufficient capacity to publish all of the research to which clinicians should pay attention. Securing universal access to 500 journals would be relatively affordable and, to use John Lavis’s analogy, the signal to noise ratio would be eminently higher than it is now. Both Norway and the McMaster service do some of this mining and filtering already. That this noble and thankless work needs to be done stems from a worrisome flaw in the health information production system. Volume trumps quality because of the incentives to publish that have nothing to do with the sharing of findings worth paying attention to: prestige, tenure, promotion, narcissism. And even first class science is by definition provisional and its application ambiguous. That seemingly powerful and persuasive RCT may have had study inclusion criteria so restrictive that the results may not mean much for my messy cohort of patients.
But the problem is deeper than that. The end goal is better decisions by both clinicians and patients. The presumption that easy access to a certified A grade electronic library will translate into effective action is wildly optimistic. If we want better decisions, we need better decision support, in real time, available with minimal effort. It has been well-documented that the best-motivated clinicians cannot possible keep up with the latest developments in their own sub-specialties, let alone the enormous terrain of family medicine or mental health or geriatrics. What people need is a well-organized, mainly automated intelligence function, supported by health information systems that document practice variations, relate inputs to outputs to outcomes, and support a culture of peer deliberation and support. In other words, a continuous, virtuous cycle of improvement.
Is there any doubt that making the use of drug decision support software a mandatory condition of the entitlement to write prescriptions would improve performance far more than giving every doctor, nurse practitioner and pharmacist access to the world literature on pharmacotherapy? The day is not far off when computer algorithms will do a better job of symptom- and history-based diagnostics than even the best human heuristic reasoning. Aviation in advanced countries is spectacularly safe not because all of the flight attendants and passengers have access to aviation journals, but because it has the right culture and the right processes.
Don’t get me wrong – I support the information liberation movement and am prepared to pay for it as a taxpayer. However, it’s not a utilitarian argument – it is based on the premise that the barrier-free sharing of information is an intrinsic good.
It is clear that access to high quality library resources is not the primary driver of quality health care. On the other hand, computer algorithms will not be a silver bullet for enhancing quality either – for many of the same reasons.
Does this negate the need to have greater access to the best possible knowledge readily available to people seeking health care, people providing care, people organizing care, our governments or individuals paying for these services, people struggling to establish proper policies and governance practices while being bombarded by low quality ‘information’, competing interests and values? Wouldn’t this enhance and improve the quality of some decision making across the vast network of health and social services provided across a Province or all of Canada – even though the diffuse impact may be hard to specifically attribute to the specific initiative?
If it really takes just 500 high quality journals to synthesize the state of clinical knowledge how do we mobilize to get this ball rolling – 1 journal at a time, one specialty at a time, one organization or province at a time. In the end if people at all levels of society do not have a way of accessing the current state of knowledge how can they possibly entertain this information as they formulate their thoughts, plans, actions and ultimately overall functioning of health services as these pool into ever larger groups.
Sometimes the most important influences are foundational – such as wide access to the best possible information and knowledge at a critical decision point at all scales of decision making.
I think this thoughtful discussion reveals, among other things, that there is no shared view of what the problem is. My view is that any computer literate clinician in any discipline can very easily get access to high quality information and guidance, free, without having to go to the source journals and libraries. Granted, it is not perfect access to everything, but my assumption is that the “pull” community wants information for instrumental purposes – what do I do in circumstance X to make things better (for patients, for the system, whatever). Perhaps where we disagree is on the definition of (I’m quoting Andrew here) “wide access to the best possible information.” The question is whether this has to be direct access – do I need access to the NEJM to harvest the knowledge generated by the NEJM? For last week’s seminal review article in my field, perhaps; but generally, no, because some organization will have incorporated the relevant research into some sort of distillation or synthesis or evidence review. The difference between high quality, high efficiency health care organizations and others is not explained by differential access to good science. The contest is not between the status quo and some idealized state of evidence-informed perfection; it is between highly variable, often wasteful and harmful existing practices and doing a lot better. If we declared a moratorium on producing new knowledge until the existing base of compelling evidence is fully embedded into universal practice, there wouldn’t be any research for a century.
I don’t know if this is helpful, but I tend to look at this from the perspective of the end user. If I were a clinician with, say, 4 hours a week to update my knowledge base, what would I want? I think I would not want to be left to my own devices to find the needle of relevant insight in the haystack of a giant database. I would look to intermediary sources of distillation and synthesis, and would be wary of relying on a single clinical trial or a hopefully reported innovation. It is true that some research – a very small proportion – is so compelling that it deserves to be translated into practice right away. But most is not; it adds incrementally to what we know and needs to be expanded, replicated, and adapted to real-world settings that differ substantially from the pristine characteristics of the study population. In a sense the greatest value of the great bulk of even the best quality published research accrues to other researchers, not clinicians.
I have the highest respect for clinicians who assemble in journal clubs and devour source materials, but I don’t think this devotion, or lack thereof, explains much of the variation in what clinicians know and how they apply it. I think the best argument for making a vast health information library available universally is that if we’re going to pay enormous sums for the production of knowledge, it’s folly to restrict access. Perhaps the best solution is the obvious one: do what the NIH has done, and require research published with its money to appear in open access journals. The elephant in the room here is that giant journal publishing conglomerates and/or professional societies have been very successful in profiting from the (usually) public investment in science. They have privatized a public resource; the public pays the salaries of academics, funds the grants, the peer reviewers work for free, and the publisher walks off with the cash. It is a long-established business model, but a highly debatable one, which the open access movement is challenging. If every granting agency implemented the same policies as the NIH, the access problem would wither away in a decade.
I agree with much of what you say Steven.
In the age of Dr. Google maybe the problem we face is an agreed upon and reliable ‘source of truth’ … i.e. what is the best available evidence, knowledge and facts … how do we individually and collectively distinguish fact from fiction?
I agree there is no common understanding or definition of how an e-library is defined. For millenium, haven’t libraries been the repository of knowledge, arts, literature … collected and organized into easily retrievable forms and codified into distinct categories to help us educate ourselves about science, art, fiction, religion, politics, economics …
In the age of ‘Dr.Google’ how can we collectively leverage today’s technologies to make better use of ever greater amounts of information and knowledge based on their scientific rigor as distinct from works of fiction, artistic impressions, …. why not start with collecting the highest quality knowledge available relating to healthcare as one cornerstone of a public e-library?
Makes you wonder if professional associations, provincial governments, university libraries …. and other beneficiaries of a ‘public evidence-based health library’ pool a few million dollars per year for a core set of licenses for some key journals and outsource the administration on a cost recovery basis to McMasterPLUS as a start?
Surely this minimal cost for the muiti-billion dollar Provincial and National health sectors in Canada would allow an entirely new dialogue to emerge between health professionals, those patients seeking high quality health information and allow more refined healthy policy debates based on competing values and needs but also help to screen low-quality or mis-information for all. It also highlights were evidence and collective knowledge ends and new frontiers for health services research could be most productive and helpful for sorting through the many complex and interdependent issues facing health care today.
I am sure the leadership at McMaster or any other leading health institute has the capacity to help facilitate such an endeavour – much like the leadership role in Canada when the Cochrane Collaboration was being established in the early 1990’s.
Is this not one logical next step in making high quality health care information accessible and therefore usable in day to day health care practices?
McMaster PLUS FS is a provided by the Health Information Research Unit at McMaster University. An equivalent version is available for all Canadians: ACCESSSS http://hiru.mcmaster.ca/ACCESSSS. Registration is free.
Brian Haynes, McMaster University
Oops! The correct URL is ACCESSSS http://plus.mcmaster.ca/ACCESSSS