A national electronic health record for primary care

There is much hope in being able to leverage technology to improve healthcare, and understandably much frustration when it does not deliver as quickly as one hopes. However, as with any cycle it is important to avoid the lows as well as the highs. We should not lose sight of the goal and not forget to understand the underlying issues.

The author’s assertion that interoperability is an “unsolvable problem” that could be solved with a single company health record is not borne out with fact in the Canadian system. Health Information Exchanges (HIEs) like those in Ontario (Clinicalconnect, Connnecting GTA) are proof that records can, and are, being accepted and transferred between different sites. The Ontario Laboratory Information System (OLIS) stores and transmits information to a multitude of inpatient and outpatient electronic health records. The challenge is not to provide a single technological platform, but to break down the barriers to interoperability that are created when business models are threatened and privacy laws are outdated.

Our local experience in the implementation of a commercial enterprise-wide EHR has taught us that customization of a local instance of software leads to forks in the software that makes interoperability with other instances difficult, not solving the interoperability problem. This is not unique to our hospital; given the countless differences in everything from nomenclature to business logic between medical organizations, there is simply no benefit in trying to employ “one-size-fits-all” solutions. In healthcare, forcing all practitioners to conform to a single product only serves to guarantee the inefficiency of this product.

The idea that open-source software could be the solution also ignores the real challenge of maintaining and developing a highly complex piece of software for all stakeholders in the healthcare system. Major university health systems in the United States have moved away from customizable software to commercial instances because of these challenges [1]. The current landscape of providers being stuck with a poor EHR would be magnified when all development is given to a monopoly.

Even if all healthcare providers did use the same EHR product, their data would not suddenly become interoperable. In fact, interoperability of healthcare data has existed for decades with HL7. This is why HIEs are effective – each provider is able to use technology in the way that is most beneficial to their practice, and is responsible only for outputting the requested data in a standardized fashion. This approach is accepted in other professions: for example shareholder documents (balance sheet, income statement) follow a standard format but internally each company is able to track their expenses in a way that best suits their specific business.

The problem is not each individual EHR platform, the problem is the lack of incentive alignment and paucity of regulation to enable interoperability. Standards already exist for semantic transfer of information. What we need is not another large public failure of IT implementation [2], but a bold push towards regulated interoperability requirements and modernization of privacy laws.

References

[1] P. D. McCluskey, "Partners’ $1.2b patient data system seen as key to future," 01 June 2015. [Online]. Available: https://www.bostonglobe.com/business/2015/05/31/partners-launches-billio....

[2] Centre for Public Impact, "The Electronic Health Records System In the UK," 3 April 2017. [Online]. Available: https://www.centreforpublicimpact.org/case-study/electronic-health-recor....

I recently read the editorial by Dr. Nav Persaud on a national electronic health record. I share his perspective in support of an evolutionary change in digital health within Canada.
As Dr. Persaud noted, health practitioners and service providers need better technology & collaboration to deliver better healthcare. “The lack of timely access to information means patients are at risk of harm from incorrect diagnoses and avoidable side effects on everything from prescription medications to the intravenous dyes radiologists inject when they do CT scans and other X-rays.” [1]
On behalf of innovators in the cybersecurity space, I want to let readers know that healthcare practitioners are not alone. Cybersecurity engineers and scientists with a mission to enable digital innovation have devoted time to developing identity and access management technology that radically moves digital healthcare forward as it solves the integration gap to connecting patients and service providers in Canada.
A federated identity solution differentiates by putting patients in control of their digital identity, and enabling secure access to their records. By putting patients first, an identity and access management solution can accelerate the speed and accuracy of healthcare delivery for patients without bringing hospital workflows and operations to a halt.
This type of solution provides the ability for citizens’ existing identity providers (e.g., Health Canada, Ministry of Health) and service providers (e.g., Hospitals, Physicians, Clinics, Research/Lab Technicians etc.) to federate identities digitally to serve citizens with a single view of their records, while keeping privacy and consent in the hands of the citizen. Therefore, their medical data would be complete and in full view to patients, even if they’re coming from different hospitals, provinces or jurisdictions - and access to service providers can be granted by patients from any trusted identity provider in seconds, eliminating serious health risks from lack of data and administrative hours for record requests.
Please consider attending IdentityNORTH (https://www.identitynorth.ca/), where important innovators in this space are fielding discussions as they work to integrate cybersecurity solutions within healthcare, with open standards and a foundation for secure digital collaboration.
Mike Cook
CEO, IDENTOS
References
1. https://www.cbc.ca/radio/whitecoat/a-national-electronic-health-record-f...

It is great to see all of the responses to this article. My feeling is that rather than choosing a single vendor solution the thing to do is to standardize on the data and the architecture of the records. A lot of work has been done on this by international organizations like HL7 and ISO. Canada participates at these meetings but the challenge seems to be on how to get that knowledge to the level of the EMR developers and to the EMR users. I have been to many meetings related to computer based health records where no one had heard of ISO 10781, 13606 or 18308. These documents provide extremely detailed information on health information standards. Physicians in general are only interested in having something that works. The problem is that while they be unhappy about how their system work they seem to be unwilling to spend any time informing their vendors about problems. There are many simple things that could be done to reduce some of the stress that these EMRs are causing but it seems to take decades to get anything significant done. A major problem are the organizations that produce the documents that are imported into physicians' EMRs. These systems were developed in the last century and there seems to be little pressure to get them updated.

I agree that physicians should not have to be involved with the deep in the weeds technical details. The problem is that the EMR developers have not really been involved either. There seems to be a preference to make things up as they go along. There is a need for at least a few physicians to provided guidance on a number of important issues like what a Problem List should look like and how it should function, how should lab result be displayed and grafted and how to deal with prescriptions that are written by never filled. These are just a few examples that need to have clinician input. Of course I have witnessed day long arguments on Problem List issues.

I think that the solution does not rest with either a single vendor or with multiple vendors. Having had the experience of switching from one EMR vendor to a different one I can say that doing this was difficult and will become more difficult as more data is accumulated. I believe that the solution is to get an agreement on what the EMR architecture should be, standardized the data elements and clinical terminology.

Solutions to these issues is long overdue. Considering that the major EMR vendors have users across the country what are our national organizations doing to get things to move forward?

I agree fully with Dr. Persaud that health data should be stored and administered at a national level, or failing that at a provincial level. This would eliminate costly retesting, reduce medical error, improve the efficiency of health care provision, and allow for improved research and quality monitoring in the health care system. It is truly unfortunate that the greatest promise of electronic health records were never realized due the diverse set of incompatible systems implemented by siloed IT departments and care providers with no national or provincial direction.

The solution proposed by Dr. Persaud, however, is fundamentally flawed and lacks insight into the realities and complexities which have precluded such an approach in the past. For instance, requiring a single national EHR system would hold hostage all Canadians, health care providers, and institutions to a single corporate entity. Consider how all the varying requirements of every stakeholder would be collected and negotiated into a manageable service contract? Who would pay for voiding all existing EHR service contracts with providers who were not selected for the single national service?

We need to work towards alternative but feasible solutions to achieve the stated goals.

In my opinion, the best solution is centrally managed data with interoperability standards and requirements for querying and storing health data. EHR systems implemented across the country would be required and certified to query and store a comprehensive set of health data attributes using the national database. This approach permits local customization of EHR systems, and safe migration from siloed local databases to the national database without forcing all stakeholders out of their existing service contracts.

Database technology has advanced tremendously in the last decade, with flexible horizontally scaling NoSQL systems becoming commonplace in the tech world. These systems permit centralization of data without compromising query performance on enormous datasets or database architecture flexibility.

Significant capital investment and technical leadership will still be required to achieve this. I fear that in the current environment of increasing austerity, gaining traction on this initiative will remain a challenge even if cost-effective in the mid-long term.

What can the wealthiest person in the world teach Canada about the design of electronic medical record systems?

A lot.

Around 2002, when Amazon was still more or less an online bookstore squeezed by shrinking margins, Jeff Bezos issued a single mandate via email (https://plus.google.com/+RipRowan/posts/eVeouesvaVX) that set Amazon on the path to become the dominant online retailing platform in the world. The keyword here is platform: Amazon doesn’t sell everything itself, but it provides the interface between buyers and sellers. In one email he changed the culture of the company from one focused on products to one focused on services. In doing so he also made the company vastly more efficient, leveraged the ingenuity of third party developers, and turned cost items into revenue streams.

How did he do it? And what can we learn from this case to improve the efficiency of our healthcare system?

Essentially, the genius of the mandate was to make all of the teams of programmers working for Amazon communicate via the same interface, and to make the internal services that the teams provide to each other accessible to third party programmers outside the company at the flip of a switch. In this way Bezos drastically reduced the amount of redundant work the independent teams had to do. Instead of building it yourself, just look for it in the directory of services — chances are someone has done it already. The fundamental resources that Amazon programmers needed — things like computing power, online storage, database management, data analytics — were now services that could be offered to the outside world for profit. In 2017 this collection of online services, known as Amazon Web Services (AWS), generated revenues of $17.4 billion USD, some 10% of Amazon’s total. AWS remains the dominant player in cloud computing, taking more market share than Microsoft, Google, and IBM combined (https://awsinsider.net/articles/2017/08/01/aws-market-share-3x-azure.asp...). Before the famous Mandate, none of this would have been possible — the massive amounts of computational resources that Amazon consumed would have purely been an expense on the balance sheet.

There are obvious parallels to be drawn between the Amazon of the early aughts and Canada’s healthcare system. Instead of teams of programmers, we have teams of physicians. Instead of a standardized communication interface, we have a balkanized system of electronic medical records (EMRs), none of which communicate with each other. Instead of asking a colleague to fax over a patient’s test results, it’s often easier to just repeat the investigation. In an era of non-stop hand-wringing about ballooning healthcare costs, this strikes me as irresponsible at best.

The Ontario EHealth debacle (http://www.cbc.ca/news/canada/toronto/ehealth-scandal-a-1b-waste-auditor...) is a prime example of how a focus on the perfect product (in this context, the perfect EMR) is doomed to fail. There is simply no way to predict and satisfy the needs of all stakeholders — our needs are too disparate. The path to success lies in establishing a set of standards to which all EMRs must adhere, and opening the market to the ingenuity of third-party developers to cater to the specific needs and whims of each customer, all while maintaining interoperability.

In 1983 Pierre Trudeau introduced The Canada Health Act, which enshrined portability as a central tenet of our healthcare delivery. While the letter of the law focuses on financial portability, the current barriers to effective communication between providers in different provinces impedes care delivery to the extent that it infringes on the idea of portability in spirit.

We need a Mandate, one from the federal government, that forces the provinces and the EMR providers to adhere to a common standard. This would go a long way towards reducing the costs of repeated tests, reduce errors in communication between practitioners, and serve Canadians in the way the Canada Health Act intended.

I was pleased to see the commentary by Dr. Persaud - A national electronic health record for primary care - and the resultant discourse that has arisen. Thoughtful debate about the best means of achieving health information interoperability is of central importance to the Canadian health service, as patient centric informational integrity is foundational to the delivery of quality care.

The notion of a single enterprise EMR as a means of solving the health information interoperability issue is not a theoretical premise in Canada. One need not look to the United States or Singapore to find such an instance. Just look north.

The Northwest Territories leads Canada as the most integrated health information system in the nation (see: https://cmajnews.com/2017/11/09/northwest-territories-leads-canada-in-el...). This was accomplished by deploying a single EMR on a single database through all services and all communities. In fact, this project goes beyond what is proposed by Dr. Persaud in that it unites not only primary health care providers, but specialists and allied health care service providers as well; virtually all health care providers in the Northwest Territories. And it works.

Although some may dismiss this success as a function of deployment to a small population of providers, the constraints faced in deploying the solution were much the same as those found in larger jurisdictions; fractured governance, software limitations, legacy systems, funding shortfalls, disparate workflow, strategic vacuity, antiquated policy, and transcendent obstruction arising from an opaque organizational culture. Further there were unique constraints that the Northwest Territories faced; a suboptimal network patched together with fiber and satellite and limited bandwidth, a health service that relies heavily on locum providers (nurses, physicians and allied providers), necessitating continuous training (often virtually in remote locations) of new or transient staff, limited funding, and the fact that the north in a university free zone and suffers from a lack of research and academic capacity and support.

Adapting a single platform to meet the needs, of an orthopedic surgeon in Yellowknife, a physiotherapist in Hay River and a nurse practitioner in Ulukhaktok was not a simple proposition, and one that required a measure of collective compromise and adaptation for the greater good; that being patient care. The entire project, virtually all services in all 33 communities (excepting Lutselk’e because of satellite bandwidth issues) in a jurisdiction twice the size of France on a patchwork network with a few dedicated staff took 17 years to complete, and at a minimal cost.

A shared patient centered charting system solves many problems; among the advantages:

• Reduction in operations and management - there is one system to pay for and maintain; 
• Ease of training – just one platform to learn;
• Ease of support – can be centralized and digitally distributed;
• Software upgrades (legendary disruptors of care) are minimized to one application;
• The chaos and cost of managing interoperability on a technical and personnel level is avoided;
• Intrinsic system e-referral / e-consultation is implicit;
• There is one jurisdictional clinical database which promotes analytics and conjoins population health to point of care service (closed-loop analytics);
• There is intrinsic chart based (secure) patient centric messaging across the scope of care.
• There is a reduction of redundant testing as all results are visible on a need to know basis.
​In short there is a network effect, from which the strategic model upon which this this system was designed gets its name; a Networked Health Ecosystem.

However, the system is not a panacea; there are issues that bear consideration. Handing over administration of all health information to one vendor, unmolested by competition, confers great power upon the vendor to dictate how information is managed, and to set fees. The corporatization of information, an insidious byproduct of the digital revolution (think Cambridge Analytica), should give us all pause in the health industry; corporatization conjoined with a monopoly is more concerning. Dr. Persaud suggests circumventing this concern by engaging open-source solutions, but this proposition has its limitations as well. Further, virtually all current marketed EMR’s, including that used in the Northwest Territories have poorly designed databases that do not support robust analytics, and were not designed to function on an enterprise level, having evolved (or not) as products for the independent primary care clinic market. Customization of the software for shared use by the broad cross-section of users can be a challenge, or necessitate costly modifications. A single enterprise EMR project was made easier for the Northwest Territories by the fact that eighteen years ago there were virtually no costly assets to sink; not the reality for the rest of Canada today. Further data interoperability capacity has now evolved, and with the advent of FHIR (Fast Health Interoperability Resources) there is optimism that achieving functional interoperability will become easier to achieve.

In the end, what the Northwest Territories experience demonstrated (and what may be lost in this discourse on the best software architecture to embrace), is that interoperability is ultimately not a technical feat. It took me years to recognize fully that the principle effort in achieving a measure of interoperability in the Northwest Territories was anthropological and sociological in nature. Almost to a person virtually no health care worker in the NWT (nurses, physicians, allied health care workers, administrators, deputy ministers), at least in the early years, wanted to share a chart across services. Everyone wanted their own, tailored to their own service, and their own perceived needs. Further not only did this feat required an endorsement of functional-interoperability by providers, but also oversight-interoperability by multiple levels of Territorial health governance (authorities, department of health, funding agencies, professional groups, etc., etc.), all with vested interests that were reflected in opaque policies. Virtually all were resistant. This was agonizingly slow work. That is why the project took seventeen years to complete; seventeen years to change the culture of care.

There are a variety of technical paths to Canadian health information interoperability; a single EMR, a ubiquitous and universally shared middle-ware platform, or discrete systems integration initiatives between existing resources. There is only one functional path; to behave ‘interoperably’. Interoperability is a way of being, and not a way of being that seems to come naturally to the Canadian health care industry, or manifest in our governance or policy strata. But it is essential, because without it, regardless of whether we have a single platform or not, we will not achieve interoperability. We will remain mired in a fractured health information dystopia, littered with shiny technology, broken promises, compromised patient outcomes, and a disaffected health workforce that has learned to distrust the promise of health information technology. A promise that is real, but needs to be rediscovered through properly architected interoperable health information.



Ewan Affleck CM. BSc. MDCM.
Past CMIO of the Northwest Territories
Senior Medical Advisor Health Informatics – College of Physicians and Surgeons of Alberta

The above editorial describes the author’s desire and need for a single national Electronic Medical Record (EMR) in primary care. He speculates that many of our problems in primary care could be solved by such an instance, from sharing of records between clinics, to research, to specialist access to notes. He cites single health systems in the US and Singapore as examples of how and why this could work. This is an interesting perspective, but one which may be overly simplistic and not shine light on the whole picture. There has been significant progress and development made in the digital health space in Ontario over the past ten years and this should be recognized and celebrated. This, coupled with a relentless focus on systems integration across the continuum of care, is where we should put our energy.

Although we have much in common as family doctors across the country, we all ask very different things from our EMRs. Practice environments are not the same in inner city urban centres, suburban practices, rurally, in university health clinics or aboriginal care centres. This delivery complexity needs to be appreciated. Nationally over 80% of family doctors already have purchased an EMR that works for them and are using it fulsomely. In Ontario this is over 85%. Physicians own their systems. They have made significant investments in these … financially, in time spent recording information about their patients, and in blood sweat and tears producing clinically useful data. We should appreciate this effort and use the systems to their maximum. Much has been accomplished and advancing from where we are is a very tenable option.

Switching to one EMR solution that attempts to meet the needs of 43,500 family doctors nationally is an impossible task. And the disruption in care created by “ripping and replacing” would take decades to recover from. Canadian provinces have collectively invested billions of dollars in EMR development, deployment and mature use. We cannot afford as a society or at any level of government to start over again. That said, fundamental health system reform, advances in integrated care models and associated compensation reform should be the driver that defines how integrated digital health platforms can and should be presented at the local, regional, provincial and national level.

Satisfaction ratings with individual EMRs is actually quite high in provincial surveys. Even with multiple EMRs on the market, connectivity is improving year over year (witness Netcare in Alberta, Connecting Ontario and Clinical Connect in ON and Saskatchewan’s eHealth portal). In Ontario, information from virtually every hospital is pushed directly into EMRs within minutes of it being generated via Health Report Manager. Lab results from everywhere can be queried and downloaded through OLIS. Integration with drug and immunization repositories is occurring now. eReferral and eConsult systems are up and running and are becoming more and more integrated into the point of care.

As we evolve into an increasingly cloud-based environment the perceived advantages of a single EMR product or database are no longer compelling. What is most helpful is gaining access to data for clinical, research and system planning purposes. This is less a technical issue than a policy one. Integration and interconnectivity are the key. We are getting closer and closer to this daily, with the advent of single sign on, contextual launching of external digital tools from the EMR, data standards and data movement. The most unpredictable factor is a human one: even when standards exist, having people use them consistently is a challenge. A single EMR does not fix this. Endless dropdown lists and tick boxes do not either. In primary care nothing is more important than the patient narrative. We can never lose this.

Choosing a single EMR vendor has other dangers in creating a monopolistic environment. This makes us vulnerable as a health system and as a profession. Currently Ontario has certified 12 EMR vendors and 16 EMR products to a set of core requirements that improve constantly over time. This allows progress to occur incrementally on interconnectivity, data portability and system functionality. Having a single vendor control the entire market means that we are at the whim of one company whose business interests may not be aligned with those of clinicians or patients. We lose our collective influence.

Open source software is not the solution to this problem either. Although it has the advantage of being inexpensive and in some cases free, it has the disadvantage of needing as much if not more support as proprietary EMRs do. Most physicians lack the knowledge, skill and desire to program and produce changes in their EMRs themselves. They just want to get down to the work of looking after patients. Multiple different customized instances of an open source EMR do not improve the situation over current state in any way.

It may seem on the surface that having a selection of EMRs nationally is folly. But experience has shown that competition drives change and innovation. We do not disagree that there are aggravations in navigating from an EMR to a viewer or external portals, but this is a solvable problem. The key is to build bridges allowing access to data that is required for a clinician at the point of care. We maintain privacy, confidentiality and security more effectively this way. Data for secondary use can be liberated easily through these structures if we create the right policy and business drivers. All of this comes at far less risk and with a far better user experience for the average doctor. Banks have done it. Retailers have done it. We are doing it.

At OntarioMD we believe in more choice not less. We want to encourage new software products to enter the market to speed up the pace of innovation. Some of these are EMRs, some are apps that make an EMR fly. We want to let doctors be doctors, not computer engineers or data scientists. And ultimately, we think that patients should be the first and last point of approval regarding secondary use of their personal health data. These are our principles. We strongly believe that best way to accomplish them is via our current approach to EMR selection, certification and improvement. One EMR system for all is simply not a realistic option.

OntarioMD
Darren Larsen, CMO
Sarah Hutchison, CEO