Open Source EHR in Practice

Open-Source Electronic Health Records: Practice Implications

Samuel Bowen MD, Robert Hoyt MD, Ladeana Glenn RN, BSN Donald McCormick, Xavier Gonzalez,

Abstract

This paper discusses the practice implications of open-source electronic health records (EHRs), in light of the recent proposed reimbursement for EHRs by the federal government. Open-source software has been accepted by consumers and healthcare organizations as an equivalent and cost-effective alternative to proprietary software. Less expensive alternatives to commercial EHRs are necessary for community health centers (CHCs), small private clinics and overseas clinics . We report on the costs, benefits and limitations of an open-source EHR and practice management system implemented in CHCs located in the United States and overseas.

The current electronic health record climate

In 2008 DesRoches and co-workers reported that the adoption rate of non-federal ambulatory electronic health records (EHRs) in the United States was 4% for an extensive fully functional EHR and 13% for a basic EHR. They defined a fully functional EHR as one that possessed clinical and demographic data, the management of lab and imaging results, management of order entry and clinical decision support. 1 A second study in 2009, by the same group, reported that only 1.5% of non-federal hospitals had a comprehensive EHR, while 7.6% owned a basic system. 2 Both of the EHR adoption rate studies, as well as a 2003 Commonwealth Fund survey, confirmed that EHR adoption is much lower in small medical practices. 1-3 The adoption rate of fully functional EHRs in physician groups of 1-5 members was 2-3%, compared to 17% for groups with more than 50 physician members. 1 In a national survey of 725 community health centers (CHCs), reported in 2007, 28% of had some type of EHR while only 8% self-reported a "full EHR". 3 These adoption rate numbers are much lower than anticipated by groups such as the Institute of Medicine that recommended EHRs in 1991. 4

To improve the low adoption rate of EHRs the Health Information Technology for Economic and Clinical Health (HITECH) Act, that is part of the American Recovery and Reimbursement Act (ARRA), allocated money to promote the adoption of electronic health records by physicians and hospitals. Beginning in 2011, physicians who participate in Medicare or Medicaid may be eligible to receive reimbursement for a certified EHR for "meaningful use". Physicians who qualify could receive $44,000 over five years from Medicare, with an additional 10% if they practiced in an under served area. Physicians whose practice consists of more than 30% Medicaid could be eligible for $65,000 over five years. Physicians can not be reimbursed for both Medicare and Medicaid. Beginning in 2015, if physicians do not opt to adopt EHRs, they will be subject to a penalty. The definition of "meaningful use" should be finalized by December 2009, after which it will be open for public feedback for 60 days. 5

The reasons less expensive EHR alternatives are necessary

In spite of federal reimbursement for "meaningful use" of EHRs, it is likely that this will not improve the adoption rate across the board. First, practices will need to purchase the system and long term maintenance with up front capital, which will be a challenge for many small practices. Second, clinicians who treat small numbers of Medicare and or Medicaid patients will not be eligible for reimbursement. Third, the amount of reimbursement may not be adequate in the long run if expensive commercial systems are purchased. Proprietary EHR systems are not only expensive to purchase and maintain, they may require in-house technology support. In an important study by Miller et al in 2005 of fourteen solo or small private practices using two commercial EHRs, the average cost per practitioner to purchase a client-server EHR system was $44K (range $14.4K-$66.6K) and the annual amount thereafter was $8.5 K (range $6K-$12K) for maintenance and support. In Miller’s article it was estimated that the EHR would save, on average, $32K (range $6.6K-$56.1K) per practitioner per year due to improved coding and other efficiencies. 27 It was clear from this study that practices varied tremendously, in terms of return on investment. In a second article by Miller in 2007 he estimated the value of four commercial EHRs to community health centers (CHCs) in six states. Both the initial costs and maintenance estimates were higher for the CHCs than the private practices previously reported. Unlike private practices, the financial benefits were far less due to reimbursement differences, making commercial EHRs for CHCs unsustainable. 28 In neither article did Miller mention open-source EHRs as an alternative.

This is important as 33% of ambulatory practices in the United States consist of fewer than 5 members. 5 This is particularly true of primary care practices. The American College of Physicians estimates that 50% of their members are in practices of 5 or fewer physicians. 6 The statistics for community health centers are probably worse [ref]Many clinics lack the capital to purchase, train, implement and sustain propriety systems. The most common reason cited by physicians and healthcare organizations to explain the low adoption rate is the high cost associated with proprietary EHR systems. 24-25

In addition to ambulatory clinics there are multiple other communities that could benefit from a more cost effective alternative to proprietary EHRs such as public health clinics, nursing homes, assisted living facilities, home health, high school and college medical clinics.

The Open Source movement

The classic proprietary software distribution denotes that users pay for highly restrictive software licenses without the ability to access or share the source code. Prior to the coining of the term “open-source”, developers at academic centers were beginning to share source code as part of projects such as Unix and the two ambulatory electronic health records COSTAR and the Veteran’s DHCP/VISTA. 1-2

According to the Open Source Initiative (OSI) the term “open-source” was coined in 1998 and is generally defined as the ability to copy, modify, use and distribute software source code. In order to receive open-source license approval OSI lists ten criteria for the distribution of open source software. Although GPL license is by far the most common license, sixty five different types of open-source licenses exist, that vary in terms of rights and responsibilities. 3 Importantly, the term “free” and open-source software (FOSS) refers to the freedom to distribute the source code and not the price. 4

Open-source software such as Linux (operating system), Apache (web server), MySQL (web based database), OpenOffice (Microsoft Office alternative) and Mozilla Firefox (web browser) have been widely accepted and adopted by consumers and businesses. 5-9 All of these programs have been in existence long enough to be tested and found to be reliable and comparable to their proprietary counterparts.

It is important to stress that FOSS is an international movement with adoption by the governments of Brazil, Argentina, Venezuela, China, Malysia, South Africa and Viet Nam. These countries have been seeking a more cost-effective alternative to build technological infrastructure. 10 In the United States there are over 900 federally qualified CHCs that provide medical and dental care to the under and non-insured, particularly in rural and inner city areas. 28 Clinicians in under served rural areas tend to see more patients who older and more complex, have reduced access to specialists and have higher turnover. 39

As open-source software became widely available and acceptable to the masses, it did not take long for open-source software to be adopted in the healthcare sector. Applications such as OpenVistA and the Canadian Health Infoway (electronic health records) and Care2x (integrated practice management system) have gained notoriety.11-13 Several resources list the extensive inventory of open-source software currently available in the healthcare field. 2, 14

With healthcare costs soaring, civilian and federal healthcare executives and information technology leaders are looking at open-source software as a way to afford newer technologies. Healthcare organizations could avoid the high costs of proprietary licenses, have the option to customize the software and instead pay for implementation and support costs. According to Goulde and Brown, open-source offers healthcare organizations multiple advantages: anyone can use or modify the software, the cost is minimal, it promotes public and private funding of software development and ensures data standards. 1 The American Medical Informatics Association (AMIA), the International Medical Informatics Association (IMIA), the Health Information Management Systems Society (HIMSS) and the Open Bioinformatics Foundation (OBF) have all established open-source working groups in support of this movement.15-18

Open-source software has been used by multiple US government agencies, to include the Department of Defense and the National Security Agency. 19 One of the most publicized recent open-source initiatives related to healthcare is CONNECT, a gateway for private and public agencies to connect to the Nationwide Health Information Network (NHIN). 20 Another important FOSS project is Mirth Connect, an integration engine that transforms HL7 messages so legacy systems become interoperable with electronic health records and other technologies, without the need to create multiple expensive commercial interfaces. 21 Commercial companies are also seeing the merit in open-source, as evidenced by Misys that offers a health information exchange architecture based, in part, on open source. 22 (Table 1)

The open source EHR movement

As previously noted, two electronic health record initiatives began releasing their source code in the 1970s. Most open-source electronic health records (EHRs), however, are less than a decade old. Table 2 lists some of the more common open-source EHR initiatives and their attributes. Table 3 lists FOSS EHR deployments by platform, to include the number of patient records in the systems. 23

A major reason to evaluate open-source EHRs is that they have the potential to increase EHR adoption rates, particularly in medical communities least likely to adopt. The Community Health Network of West Virginia implemented the Indian Health's System EHR known as Resource and Patient Management System (RPMS) EHR in 2007. RPMS is a variant of the Veteran's VistA EHR that targets ambulatory care and excels in disease management. They were able to document that the implementation costs were 51% less and the maintenance costs 66% less than comparable commercial EHR systems costs reported by Miller. 29

Two open-source EHR systems have been installed the most in the United States and overseas.

1) OpenEMR is an EHR and practice management system intended for small ambulatory practices. The system operates on Linux, Free BSD, MacOS X and Windows operating systems. It is based on the LAMP stack (Linux, Apache, MySQL and PHP) and is available as client-server and ASP models. OpenEMR exists as a commercial and community model and we will be discussing the latter. Technological features and other details are available in a 2008 monograph and the OpenEMR web site. 30-31

2) The Veterans Health Information and Technology Architecture (VistA) is available in the following formats: A) FOIA OpenVista and WorldVistA B) Commercial through DSS, Medshere and ClearHealth. C) Modified by the Indian Health System and available as Resource and Patient Management System (RPMS). Currently VistA has been deployed in Hawaii, West Virginia, Germany, Finland, Malaysia, Brazil, India and Jordan. 30, 32-38

Examples of open source implementations

We studied the impact of implementing an open-source electronic health record by an organization in Houston, Texas called, Tomorrow’s Bread Today (TBT). 40 They are a non-profit patient association whose main business activity is the management of an Independent Physician Association (IPA). The IPA consists of 520 physician practices of which only a small number have EHRs. The practices that have EHR paid the market prices of 40K plus for the proprietary systems and are generally very satisfied with the improvement in their medical record keeping and billing systems. However, most of the practices did not want to pay the cost or take the time to go through a conversion to EHR. TBT, as a manager of the IPA, has the responsibility to see that the care of Medicare Advantage Member patients seen by the physicians under the HMO contracts is well managed and accurately reported because reimbursement is based on the morbidity of the individual patients. The differences in reporting the patient’s medical condition completely and in a timely manner and not doing so are as much as 15% of the revenue paid for health services. The health plans pay incentive bonuses of as much as 50% of their profits from the hospital and professional and pharmacy pools directly to the IPA for distribution to the member physicians. Without any EHR in the past these bonuses have increased the physician reimbursement by an average of 33% for the Medicare Advantage Member patients. With an EHR we know the reporting improves and that the profit margins will improve because of increased revenue paid to the health plans. Beginning in 2010 the rates of reimbursement from CMS to the health plans will be cut by 4% and then is expected to be reduced 3% each subsequent year for four years. Therefore, failure to report completely and accurately on every patient encounter and diagnosis and treatment will eliminate the profit margins and bonuses. If every member of the IPA had an EHR and that information resided in the cloud then performance would be an easy matter and the patient outcomes would also improve because the physician’s could more easily act as a team in the patient care process.

TBT knew that the cost barrier and conversion problems that blocked EHR in most practices could possibly be overcome by using an Open Source EHR, such as OpenEMR or World VistA. In 2007 OpenEMR was installed in a community health center in Pearland, Texas that had approximately 4,000 patient encounters over the next year. Care was provided by a nurse practitioner who used a laptop computer and data was backed up to a Linux-based server. During the year TBT’s IT department made several changes to the data entry templates in OpenEMR at recommendation of the Nurse-Practitioner. In 2009 TBT opened two more clinics, but this time made them patient cooperatives in which members paid monthly dues to cover primary medical care and where some of the members were covered by Medicare and Medicaid programs. This required that OpenEMR document membership eligibility and provide billing and electronic claims processing. The clinic converted from a local server to a web-based server that was automatically backed up continuously locally and remotely. Claims to the HMOs and to Medicaid Plans were sent electronically from OpenEMR to Availity, as a paperless operation. There is no paper in the operations. TBT’s demonstration should be complete by 2010 with plans and they will begin to introduce OpenEMR to the IPA practices, however there is much more involved in getting it accepted and used than the fact that it works and is useful.

International Planned Parenthood Federation (IPPF) has over 8,000 clinics in 180 countries, serving 31 million clients. 41 One the major reasons for the IPPF to investigate EHRs was the need for improved resource management (practice management system), in addition to information management (EHR). OpenEMR was selected because it was the right size for its IPPF clinics, was cost-effective, LAMP compliant and supported multiple languages. The IPPF Clinic in Barbados has 9,000 clients and has used OpenEMR for......months. Implementation and training required........months. The initial cost for implementation and training was $.... per clinician and $..... per clinician per month for maintenance and support.

Discussion: Practice Implications

Kantor was one of the first to comment on the potential benefit of open-source EHRs for primary care. 43 FOSS is a particularly attractive alternative for CHCs, small practices and hospitals who are under-capitalized in the US and overseas. The importance of practice management applications, as part of the EHR, can not be over-emphasized. Clinics that benefit from open-source EHRs also benefit from maximizing reimbursement from civilian and federal payers.

Numerous advantages of open-source software have been reported, in addition to the lack of licensing fees A. Software can be modified and customized to match unique requirements B. The size of the developer community is far greater than that associated with proprietary software, hence development and deployment, as well as security are enhanced. 42 C. Open-source software is more likely to embrace open standards 43 D. FOSS is felt to be as secure or more secure, compared to proprietary software. 44 This is also confirmed by searches on the National Vulnerability Database to compare open source operating systems with proprietary 45. E. Unlike proprietary software, if a vendor goes out of business, the source code is still available. F. Vendors of proprietary software are usually unwilling to share training or demonstration applications for students studying EHRs G. Open-source EHRs have greater potential for research because the source code can be modified and customized to participate in research. 46

Information technology staff must be familiar with open-source technologies and associated support. Open-source initiatives may be less well capitalized which could decrease their ability to upgrade the software. Marketing for FOSS EHRs is likely to be less than for commercial EHRs. Many open-source EHRs will need to find resources to make the necessary changes that will be stipulated by the federal government for reimbursement under Medicare and Medicaid. 47 However, the Certification Commission for Health Information Technology (CCHIT) modified their requirements in mid-2009 to accommodate more than commercial EHRs, which should allow for more FOSS EHR certifications. 48 Open-source EHRs would greatly reduce the initial and long term costs, but multiple other obstacles impede the adoption of EHRs besides cost. The major ones generally listed are lack of interoperability, unclear return on investment, decreased initial productivity, physician resistance and lack of skilled IT workforce. 24-25

The federal government is clearly warming to the idea of open source in several areas. As mentioned earlier, the open-source gateway known as CONNECT will be an integral part of the NHIN. In August 2009 the Department of Health and Human Services held a "code-a-thon" to invite programmers to work face-to-face to improve and collaborate on the gateway. 50 The HITECH ACT that is part of the American Recovery and Reimbursement Act, stipulates that the Secretary of Health and Human Services report on open-source EHRs by October 1 2010. Specifically, the report must include current availability of FOSS for safety net clinicians, cost comparisons with proprietary systems, the applicability to children and the disabled and the capacity for interoperability. 47

One could argue that this is the approach the government should take in regards to electronic health records. Build the platform based on open-source and open standards and let open-source and proprietary developers compete on the basis of value and cost. This would also encourage a modular approach to the technology so developers could focus on singular enhancements such as disease registries and electronic prescribing. Clinical decision support is still in its infancy with many obstacles reported. 51 Why not unleash a task force of developers along with end-users to help solve some of the current issues?

The above strategy would be in keeping with President Obama's Open Government Initiative that states "The Administration is committed to developing those recommendations in an open fashion. Consistent with the President’s mandate, we want to be fully transparent in our work, participatory in soliciting your ideas and expertise, and collaborative in how we experiment together to use new tools and techniques for developing open government policy". 52 This is quite different from the current policy that favors commercial vendors and deep pockets.

In order to achieve universal EHRs in the next decade it will take more than financial reimbursement. We need an open and transparent strategy that is based on collaboration and our collective intelligence.

Table 1. Open-Source Electronic Health Records

Open-Source EHR	Web Address	Programming Language	License Type	Practice Management	ASP Model Offered	Hospital Based	Target Size
OpenVistA	http://medsphere.org	Mumps, C#	Gnu Affero GPL		yes	Hospital or large clinic	Large
OpenEMR	http://oemr.org	PHP, MySQL	Gnu GPL (v.2)	yes	both	Outpatient	Small to Medium
Clear Health	http://clear-health.com	PHP, MySQL	Gnu GPL	Yes	Yes	Outpatient	Small to Medium
OpenEHR	http://www.openehr.org		Mozilla tri-license

Table 2. Open-Source Electronic Health Exchanges

Open-Source Data Exchange Projects	Web Address	Data Exchange Type	License Type
NHIN-CONNECT	http://www.connectopensource.org	Connect VA, DoD, Community Exchanges
Mirth	http://www.mirthcorp.com	Deals with small discreet messages of data
Misys Open Source Solutions	http://www.misys.com/corp/OpenSource	Community Level Data Exchange

References

1.Goulde M, Brown E. Open-Source Software: A Primer for Healthcare Leaders. California Healthcare Foundation. March 2006. http://www.chcf.org/documents/healthit/OpenSourcePrimer.pdf (Accessed July 26 2009)

2.McDonald CJ, Schadow G, Barnes M, Dexter P, Overhage JM, Mamlin B, et al. Open-source software in medical informatics---why, how and what. Int J Med Info 2003;69:175-184

3.History of the OSI. Open Source Initiative. http://www.opensource.org (Accessed July 26 2009)

4.Free Software Foundation http://www.fsf.org (Accessed July 25 2009)

5.Linux http://www.linux.org (Accessed July 26 2009)

6.Apache http://www.apache.org (Accessed July 27 2009)

7.MySQL http://www.mysql.com (Accessed July 27 2009)

8. OpenOffice http://www.openoffice.org (Accessed August 2 2009)

9.Mozilla http://www.mozilla.org (Accessed July 29 2009)

10. Wong K. Free/Open Source Software. Government Policy. Asian-Pacific Development Information Programme. http://www.iosn.net/government/foss-government-primer/foss-govt-policy.pdf (Accessed August 13 2009)

11.OpenVistA http://sourceforge.net/projects/openvista/ (Accessed August 1 2009)

12.Care2x http://www.care2x.org (Accessed July 26 2009)

13.Canadian Health Infoway http://www.infoway-inforoute.ca/lang-en (Accessed July 26 2009)

14.Open-source software in healthcare. http://www.wikipedia.org (Accessed July 26 2009)

15.The American Medical Informatics Association. Open-source software working report monograph. http://www.amia.org (Accessed July 29 2009)

16.The International Medical Informatics Association OSWG http://www.chirad.info/imiaoswg (Accessed August 1 2009)

17. The Healthcare Information and Management Systems Society Open Source Technology Workgroup. http://www.himss.org/content/files/HIMSSOpenSource.pdf (Accessed August 1 2009)

18. Open Bioinformatics Foundation http://www.open-bio.org (Accessed August 1 2009)

19. Case Studies. Open Source for American. http://www.opensourceforamerica.org (Accessed August 10 2009)

20. CONNECT http://www.connectopensource.org (Accessed August 10 2009)

21. Mirth Connect http://www.mirthcorp.com/products/mirth-connect (Accessed August 10 2009)

22. Misys Open Source Solutions http://www.misys.com/corp/OpenSource (Accessed August 12 2009)

23. Valdes, I. Free and Open Source Software in Healthcare 1.0. Open Source Working Group White Paper. AMIA. November 2008. http://www.amia.org/.../Final-OS-WG%20White%20Paper_11_19_08.pdf (Accessed July 28 2009)

24. DesRoches CM, Campbell EG, Rao SR et al. Electronic Health Records in Ambulatory Care—A National Survey of Physicians. NEJM 2008;359(1):50-60

25. Jha AK, DesRoches CM, Campbell EG, et al. Use of Electronic Health Records in US Hospitals. NEJM 2009;360 (16):1628-1638

26. The Commonweath Fund National Survey of Physicians and Quality of Care 2005. http://www.commonwealthfund.org/usr_doc/Audet_physicianqualitysurvey_chartpack_823.ppt (Accessed August 13 2009)

27. Miller RH, Sim I. Physician’s Use of Electronic Medical Records: Barriers and Solutions. Health Affairs March/April 2004 http://content.healthaffairs.org/cgi/reprint/23/2/116 (Accessed July 27 2009)

28. Miller RH, West CE. The Value of Electronic Health Records in Community Health Centers: Policy Implications. Health Affairs 2007;26 (1):206-214

29. RPMS EHR for Community Health. Community Health Network of West Virginia. http://www.chnwv.org (Accessed August 13 2009)

30. Faus SA, Sujansky W. Open-Source EHR Systems for Ambulatory Care: A Market Assessment. http://www.chcf.org (Accessed July 20 2009)

31. OpenEMR http://www.oemr.org (Accessed July 3 2009)

32. OpenVistA http://sourceforge.net/projects/openvista (Accessed August 14 2009)

33. WorldVistA http://www.worldvista.org (Accessed August 14 2009)

34. DSS, Inc. http://www.dssinc.com/index.htm (Accessed August 14 2009)

35. Medsphere http://www.medsphere.com (Accessed August 14 2009)

36. ClearHealth http://clear-health.com/ (Accessed August 14 2009)

37. Resource and Patient Management System http://www.ihs.gov/Cio/RPMS/index.cfm?module=home&option=Index (Accessed August 14 2009)

38. Longman P. Code Red, how software companies could screw up Obama's health care reform. Washington Monthly July/August 2009. http://www.washingtonmonthly.com (Accessed July 15 2009)

39. Effken JA, Abbott P. Health IT-enabled Care for Underserved Rural Populations: The Role of Nursing. J Amer Med Info Assoc 2009;16 (4):439-445

40. Tomorrow's Bread Today http://www.tbt.org (Accessed August 1 2009)

41. International Planned Parenthood Federation http://www.ippf.org (Accessed August 1 2009)

42. HIMSS Open Source Software Fact Sheet. http://www.himss.org/content/files/HIMSSOpenSource.pdf (Accessed August 12 2009)

43. Kantor GS, Wilson WD, Midgley A. Open-source and the Primary Care EMR. Letter to the Editor. J Am Med Inform Assoc 2003;10:616

44. Boulanger, A. Open-source versus proprietary software is one more reliable and secure than the other? IBM Systems Journal. 2005. p 239. http://www.research.ibm.com/journal/sj44-2.html (Accessed August 1 2009)

45. National Vulnerability Database. http://web.nvd.nist.gov (Accessed July 31 2009)

46. Albritton DW. Using open-source solutions to teach computing skills for student research. Behavior Research Methods, Instruments and Computers: A Journal of the Psychonomic Society. 2003;35 (2):251-254

47. Public Law 111-5. American Recovery and Reinvestment Act of 2009. http://www.gpo.gov/fdsys/pkg/PLAW-111publ5/content-detail.html (Accessed August 13 2009)

48. Certification Commission for Health Information Technology. www.cchit.org (Accessed August 13 2009)

49. Mandl KD, Kohane IS. No Small Change for the Health Information Economy. NEJM 2009; 360 (13):1278-1281

50. Mosquera, M. Feds to host NHIN software code-a-thon. Government Health IT. July 31 2009. www.govhealthit.com (Accessed August 15 2009)

51. Schedlbauer A, Prasad V, Mulvaney C et al. What Evidence Supports the Use of COmputerized Alerts and Prompts to Improve Clinicians' Prescribing Behavior? JAMIA 2009;16:531-538

52. Open Government Initiative http://www.whitehouse.gov/Open (Accessed August 15 2009)