Health Information Technology, Electronic Medical and Health Records Definitions
Health information technology can be defined as the systems health stakeholders (e.g., patients, doctors, pharmacists, etc.) use to store, view, and interact with health information. Electronic health records (EHRs) and electronic medical records (EMRs) are two related but different HIT systems. Important differences between EHRs and EMRs are that: 1) an EHR is a full health record; an EMR is partial health record, 2) an EHR is “person-centric”; an EMR is “provider-centric”, and 3) an EHR interacts with many separate HIT systems; an EMR may or may not interact with separate HIT systems [16].
Healthcare and Health Information Technology in Canada
Canada’s healthcare system is decentralized, universal, and publicly funded. Most healthcare is administered sub-nationally by the provincial and territorial governments, including most public HIT, EHRs, and EMRs systems—this results in systems that are 1) non-interoperable, 2) expensive, and 3) inconsistent [17]. The Canada Health Act states that healthcare should be portable, but because these HIT systems are separate, it is not [18]. In 2001, the federal government budgeted $2.1 billion for a new independent HIT agency to address these problems, Canada Health Infoway (CHI). In 2009, the federal government budgeted a further $500 million for CHI [16]. In 2010, the Auditor General of Canada and 6 other provinces released a report on EHRs following a HIT spending scandal in Ontario [19], [20]. For global context, in 2004 the United Kingdom budgeted £12.4 billion for its HIT program, the National Programme for Information Technology (NPfIT). The Programme was later abandoned. In 2009, the United States budgeted USD$19 billion for HIT as part of the American Recovery and Reinvestment Act. These figures make HIT some of the largest IT projects ever [21]. There is no widely accepted, periodically published, country HIT ranking, but countries are often compared by rates of EMR adoption. In 2021, 86% of primary care physicians in Canada used an EMR compared to the average of 93% among 38 OECD member countries [22].
The National Health Information Technology System Debate
There is existing literature on the importance of, and how to adopt, HIT [23], [24]. One debate is whether or not to use open-source development models, which have proven incredibly successful for driving software innovation as a whole [25], [26], [27]. In addition, using open-source approaches can accrue substantial economic savings through a range of contexts [28], [29], [30], [31]. Substantial research has focused on comparisons of open-source HIT [32], [33], [34], [35], [36], [37], [38], and particularly in low-resource settings [39], [40]. In addition, there has been studies on CHI’s impact [17], [41] and the future development for HIT in Canada. The latter topic includes: 1) calls for restarting HIT development from scratch [18]; 2) modifying existing systems for common standards [42], [24], [41]; and 3) abandoning a national strategy [43]. No literature, however, has estimated the costs and savings of these options. An analysis of all options is beyond the scope of a single paper, so a cost and saving estimation was performed for the modification option. The modification option is likely the most feasible, considering the money already spent on existing systems and the benefits of a national system. The Canadian representatives to a 2017 OECD survey responded that Canada was pursuing “a system of exchange of patient data at the sub-national level only”. Canada was one of 10 OECD countries that reported only partial or no ability for patients to view health information, and one of 15 that reported only partial2 or no ability for patients to interact with their health information out of 273 [44].
Globally, implementation of nationwide EHRs has progressed slower than expected [45]. The advantages of a national system include reduced costs for: 1) training, 2) data migration, 3) data transfers, and 4) procurement. Training costs include the time stakeholders spend learning new HIT software. Data migration costs include the additional time stakeholders spend on tasks when they are unable to use their HIT software because they are waiting for data migration to complete and the cost of the data migration. Transfer costs include the additional time stakeholders spend transferring or waiting for an information transfer when they this could be done automatically. In health research, obtaining access to national data is often cumbersome and inhibits research [46], [47], [48], especially from the primary care level. This inhibition is unfortunate because there can be over 300 primary care interactions for every inpatient admission to an academic health sciences center [49]. Procurement costs include the additional time stakeholders spend procuring and managing software and the additional cost of not being able to negotiate better rates due to a lack of scale. EMR vendors often keep pricing private [50], but costs of $15 to $50 thousand per year per clinic are common which can be significant to independent stakeholders [51].
The disadvantages of a national HIT system primarily involve the cost of organizing the building and maintenance of a system. In Canada, because each province is primarily responsible for its own HIT, there is no organization that has the authority to develop national systems. Even CHI, as one stakeholder states, “funds things but doesn’t manage them” [52] because the majority of healthcare responsibilities are provincial and territorial. Additionally, national interoperability is often not prioritized because the majority of a patient’s care usually occurs in their home province; in 2019, reciprocal payments4 to physicians were $181 million, 0.6% of total physician billings, for 2.6 million services [53].
Open-Source Software and Open-Source Health Information Technology
The Open Source Initiative defines "open source” with 10 attributes [54]; in summary, OSS projects make source code freely available for anyone to store, view, or interact with. These attributes prevent vendor lock-in, and monopolistic companies, common with a closed source software (CSS) model; for example, 90% of EMRs in Canada are the products of three companies [55]. Preventing vendor lock-in lowers barriers to entry and promotes competition. Competition can result in divergent standards, but the most successful OSS utilizes common standards, solving this problem. This makes OSS often more reliable, secure, intuitive, performant, flexible, interoperable, and inexpensive than proprietary software because of these attributes [21], [36], [51], [56], [57]. These attributes make OSS an anti-rival good—the more people that share it, the more utility each person receives [25].
The most popular open-source projects (e.g., Linux) are for developers and essentially run the backbone of the Internet and the thousands of companies that benefit from it. There are, however, also many other popular OS projects for non-developers; LibreOffice, a suite similar to Microsoft Office, has 200 million active users [58]; Blender, a 3D modelling application, was downloaded 20 million times in 2020 [59]. The later example further demonstrates that OSS can succeed in specialized domains, like healthcare, where there are relatively few qualified programmers [60].
VistA is one of the most frequently cited examples of OS HIT. VistA has been used by the United States’ Veterans Health Administration for over 40 years. A 2010 study estimated VistA provided over $3 billion in cumulative benefits, after accounting for investment costs [61]. Despite the clear benefit of the OSS approach, HIT managers often choose CSS over OSS for organizational reasons, instead of technical ones. These barriers include: 1) HIT managers not being familiar with OSS concept and benefits, 2) misconceptions about OSS (e.g., the erroneous belief that because the source code is open, it is less secure), 3) lobby and sales teams of CSS vendors, and 4) lack of third-party support for OSS [51], [62].
[2] Canada was the only country that indicated partial ability to view and interact with health information because this functionality is only available in some provinces and territories.
[3] The United Kingdom responded individually for England, Northern Ireland, and Scotland—these responses were grouped using the AND operator.
[4] Payments made to physicians for services to people who are outside of the province where they hold a health card.