Is it really "Buy vs. Build"? Part II
In this two-part article, we examine the question of buy vs. build in healthcare. Part 1, which ran in Healthcare Informatics' November issue, discussed the traditional buy vs. build argument and proposed a more nuanced approach of buy+build and build+build. In reality, the degree of customization required in today's healthcare applications often creates a custom application sitting on top of a standard framework (buy+build): whereas build+build is internal or contracted development. Part 2 presents a real-world case study of the decision process that MedStar Health underwent in deciding to build a clinical Web portal (build+build).
MedStar Health is a not-for-profit, community-based healthcare organization comprised of 25 integrated businesses including seven major hospitals in the Baltimore/Washington D.C. area. The hospitals, which include both teaching and community facilities, are Franklin Square Hospital Center, Good Samaritan Hospital, Harbor Hospital, and Union Memorial Hospital in Baltimore, Maryland; and Washington Hospital Center, Georgetown University Hospital, and National Rehabilitation Hospital in Washington, D.C.
MedStar Health hospitals and healthcare organizations serve more than 500,000 patients each year. With 22,000 employees and 4,000 affiliated physicians, it is one of the largest health systems in the mid-Atlantic region. MedStar Health's services include primary, urgent and subacute care, medical education; and research. Other health-related services include assisted living, home health, hospice, and long-term care.
MedStar is an extensive healthcare organization with many healthcare technology needs. Because the organization is the result of a series of healthcare mergers and acquisitions, there are a large number of healthcare applications from a variety of vendors. This makes for a challenging environment in which to find technology solutions that can work well across the enterprise. MedStar e-Health, a division of MedStar Clinical Informatics, was created by the MedStar board to provide MedStar clinicians with information about available e-health applications and emerging technology.
Projects focus on finding and testing new technology to increase clinician productivity, office efficiency, practice quality and safety, and access/convenience for patients. Specific areas include applications for outpatient practice and bi-directional connectivity to MedStar hospitals (in particular, enhanced scheduling requests and results reporting).
Selection process
A key task in achieving the MedStar board's mandates was the establishment of a clinician portal, a Web-based destination for clinicians to access products and services provided by the MedStar Health System. The provision of a single, standard, and personalized destination on the clinician's desktop minimizes the complexity of accessing the system's services for the end-user while enhancing the MedStar Health brand. The major intent of the clinician portal was to be a gateway that securely extended the reach of many sources of patient and clinical information available from within MedStar. The portal could present information from other sources, aggregate multiple sources of information and present new sources of information exclusively available to the portal.
A clinician-centric approach was developed for the evaluation of portals. Applications or strategies under consideration had to meet criteria for added value, plus relevant, appropriate, robust, and low cost/low risk applications. Also, they had to enhance throughput and output, be attractive to clinicians, and create a short-term and long-term return on investment (ROI). This clinician-centric approach is in keeping with sound business principles within and outside of healthcare.
Additionally, if the areas of initial focus for physician process enhancements also were in keeping with patient needs and further had the ability to improve quality and safety, then delivery of these applications could gain traction among all stakeholders and create the basis for a win-win-win situation.
MedStar e-Health conducted months of research on clinician portals and while many vendors had high-quality software, there were no clear winners. Whereas vendors and software may change frequently, a vision and strategy for finding value and success using e-health is more constant. Many vendors offered clinician portals that were expensive and lacked flexibility.
No single vendor, in our review, had a complete toolset that fit our vision and strategy that included a staged approach to introducing features that we ranked as most important. Interestingly, at about this same time, CareGroup Healthcare System in Boston also decided to build its own portal, PatientSite.
From gateway to portal
The establishment of the MedStar clinician portal was not made by a straightforward declaration, but rather an iterative process resulting from e-Health's simultaneous work on its other mandates. Initially e-Health avoided the concept of internal development as it was believed that there surely had to be a vendor clinician portal that could meet MedStar's diverse needs.
However, while we conducted the portal vendor analysis, new solutions and technology pilots needed an aggregation point for participant involvement. Specific needs included user support, pilot requests and an online home for medical content important to MedStar clinicians. This need led to the establishment of a static Web site originally called the interim clinician portal.
The decision to initially consider the portal as an interim strategy allowed us maximum flexibility in determining future design decisions. Over the next two years, while simultaneously conducting the vendor portal search, we continued to add to the capabilities of the static Web site and the interim solution started to gain momentum with MedStar clinicians. As the interim portal expanded functionality through a clinician-centric design philosophy, so did the user base — expanding from a few hundred users to the majority of the 4,000-plus clinician community having clinician portal accounts.
As we continued to look at vendor portals during this time, the cost and degree of customization (buy+build) necessary for one of them to match our successfully growing interim solution pushed the vendor option farther and farther out of our reach. In many cases, the vendor portals did not even offer the ability to customize to the degree needed to match our clinician requirements.
Major milestones in the Web site evolution to a clinician portal included: the creation of a login-required area to display restricted-access MedStar information; the addition of the MedStar online clinical library; the addition of a custom clinical messaging application; and the creation of the clinician portal single sign-in strategy, which declared that all portal functionality would not require an additional sign-in for the end user.
Under this strategy a number of clinical applications have been integrated into the portal, including Alpharetta, Ga.-based McKesson Horizon Patient Folder, and, most notably, the Web-based medical results viewer, webAzyxxi, which is based on the Azyxxi Medical Information System, now owned by Microsoft (Redmond, Wash.). Azyxxi, created by Mark Smith, M.D., and Craig Feied, Ph.D., M.D., at MedStar's Washington Hospital Center, is another example of a self-built application having the flexibility to bring together disparate information sources from MedStar's many vendor systems. Azyxxi has since been purchased by Redmond, Wash.-based Microsoft.
MedStar's decision process evolved from pure buy to buy+build to build+build over the course of this project. A key finding was discovering and measuring the level of customization required to make the portal successful, no matter what the platform. The future of this strategy will be determined by the maturation of vendor systems, decreasing cost, and true interoperability.
Jason Kreuter Ph.D., is associate director, MedStar e-Health; Allison Stover is a healthcare consultant; and Peter Basch M.D., is medical director, MedStar e-Health.
