- by Dr. Emory Fry CMD MC and Terry L. Wiechmann

When assessing your organizational business needs, do you totally rebuild your heritage applications or do you incrementally migrate to state-of-the-art technology?

Fundamental to the healthcare of military personnel is the clinical information that should follow service personnel throughout their career and, indeed, beyond. Central to healthcare in the military is the Composite Heath Care System (CHCS). The CHCS system, the organizations that support it and the personnel that run and use it on a daily basis form clinical informational backbone of military healthcare. Although vital to the daily care of U.S. service personnel and their dependents, it is showing its age. This paper describes how the legacy CHCS system has been rejuvenated, saving years of investment and, more importantly, improving patient healthcare in the DOD.

Background

The Military Health System (MHS) acknowledges its mandate to support our country’s Armed Forces and to provide quality care to all its beneficiaries. It recognizes that medical information systems are integral to achieving that goal and has funded countless programs to improve our ability to manage the world’s largest health care delivery system. Because of the will, dedication, and devotion of countless men and women there have been many successes.

Nevertheless, we are currently adrift. Strip away the hype, the vaporware, and the PowerPoint slides depicting our best intentions, and one unveils a lumbering juggernaut. Women are still missing their annual mammograms, pregnant mothers still don’t see the same physician twice, and we still inadvertently give nephrotoxic drugs to patients with renal disease. Our enormous investment in alternative information technology has, but incrementally improved the ability of our providers to impact the daily lives of the patients they serve. We recognize that medical decision support, population health, and disease management could revolutionize the Federal Health Care System, but realizing that potential remains frustratingly beyond our grasp.

We have all experienced the sensation of running in place at some point in our daily lives. But there comes a time when we settle down, take stock of our situation, reassess our priorities and summon the courage to refocus our energies in more efficient and constructive ways. Those individuals, communities, or organizations that fail to recognize when it its time, burn out and fade into oblivion. For medical information management in the MHS that time is now.

No one would argue that medical information should flow seamlessly from one application to another, from one user to another, without hindrance or duplicate entry, and should be presented in whatever manner enhances our ability to provide quality patient care with measurable improvements in outcome. Indeed, this consensus belief motivated the MHS to abandon “legacy” systems that historically have been difficult to integrate in favor of “modern” solutions promising “drag and drop” data mining, web-enabled efficiency, and fully computerized patient medical records accessible by anyone from anywhere. But what has been the health care sector’s experience during this journey?

First, CHCS is based on time proven MUMPS (M) systems. M is a computer language that supports an integrated multi-dimensional (hierarchical) database. Migrating hierarchically structured CHCS data to a relational database model is extremely expensive and fraught with a myriad of data integrity/synchronization concerns.

Second, no system can hope to replace the existing infrastructure in one fell swoop, thereby mandating the need to maintain two separate and extremely expensive solutions simultaneously.

Third, replacing the established system requires rewriting countless business rules and debugging millions of lines of new code just to get back to where the MHS spent millions of dollars to get to in the first place. Only then can new functionality and capabilities be added.

Fourth, the new system must gain the trust and confidence of an inherently conservative and skeptical community, a process that took years to accomplish with CHCS and is not inherently transferable.

Finally, reliance on proprietary technology and standards with their associated licensing fees has driven the cost of implementing and enhancing the next generation of systems into the stratosphere. All to often one hears vendors and system integrators say their solution is cheap because we have an enterprise license for this or that, but when all is said and done, somebody had to pay for it and that dollar was a dollar that went to a vendor instead of patient care. In this day and age of budget constraints and limited resources, the “different pot of money” argument rings hollow, especially to Military Treatment Facilities (MTF) scraping to replace antiquated equipment and improve business processes.

These lessons have slowed the progression of new deployable, sustainable Information Management (IM) tools to a crawl. Without centrally provided clinical tools in the immediate future, several grass roots initiatives have sprung up to fill the void. While comparatively low budget, many of these projects have been supported with considerable local programming talent and clinical expertise. There is, however, an inherent danger to such local initiatives, namely ending up with multiple, non-integrated, stovepipe solutions. Most solutions, which were designed as enterprise solutions, ultimately rely on separate conjoined databases feeding off of CHCS and are plagued by data integrity issues, synchronization concerns, and exorbitant expense.

Seeking an alternative strategy, Naval Medical Center San Diego (NMCSD) reexamined the assumption that architecturally CHCS was obsolete. Despite its many shortcomings, CHCS is based on a time-tested, industrial strength multi-dimensional database that as a transactional system is second to none. It is the central repository for most clinical data, and has been debugged and tested at considerable expense over the years. In short, it works reliably and is an integral part of the patient care business process. Criticism of CHCS is largely centered on its archaic terminal interface, the difficulty of integrating it with other data systems, and the expense of building new functionality.

A New Approach

Following a comprehensive technical review, NMCSD concluded that ESI Technology Corporation offered not only a sophisticated technical architecture, but also an ideal migration path from our current legacy systems to a promising future of open standard distributed computing capable of addressing legacy CHCS shortcomings. Similar to the way the C programming language evolved into C++ and then Java, ESI provides language extensions to the M language that transform it into a powerful, modern, object oriented development language known as EsiObjects. It fully supports client side development tools via a diverse choice of middleware options including TCP/IP, COM+, ASP, Java and XML. It includes industry standard CORBA and COAS implementations, which combined with native XML capability, enables sophisticated three-tiered client-server enterprise solutions capable of integrating multiple legacy systems. Indeed, EsiObjects is a complete, standards-based, object oriented database system intended for creating complex applications in a time-critical environment. Furthermore, the latest version of EsiObjects has been released under Open Source licensing, enabling the MHS to build on a truly cost effective architectural solution.

The potential cost infrastructure savings are enormous, especially when compared to the expense of other commercially available systems. EsiObjects eliminates the need for expensive database licenses, enables existing hardware to host new enterprise applications without astronomical infrastructure upgrades, and provides for a distributed computing model that optimizes enterprise bandwidth and network resources.

A Technical Framework For The Future

Evolving CHCS to a modern, distributed system is essential to improving the business processes that use it. Web and wireless technology make CHCS accessible. This section will briefly explain how that was accomplished.

Figure 1: CHCS Legacy Tiers

All of the CHCS modules are implemented using M (MUMPS) technology. It is a highly integrated application. However, as illustrated in Figure 1 above, it can be logically divided into three tiers:

1. The Database Tier is responsible for maintaining data and data structures. Within CHCS systems, the File Manager DBMS is responsible for these tasks. Each file listed above defines metadata, that is, definitional information about the data that resides in the CHCS Database. The individual data element's business rules (validation, input transforms, output transforms) reside at this level also.

2. The Application Tier contains the application functions (code) for each of the CHCS modules.

3. The Presentation Tier consists of software modules that implement screens that are displayed on dumb terminals. Each screen displays data as well as collecting it for validation and storage.

Figure 2: EsiObjects File Wrappers

As part of a previous effort by DOD Health Affairs, ESI was contracted to wrap and expose the CHCS data as objects to the enterprise using EsiObjects. The CHCS modules wrapped were: Patient Demographics, Laboratory, Pharmacy and Radiology. EsiObjects exposes these objects to the Easy CHCS Provider Portal via its Java Gateway middleware. See the Figure 2 above.

Figure 3: Wrapping Appointment Scheduling Code

To solve a critical problem at NMCSD, new appointment scheduling functionality was added to the Easy CHCS web pages using existing CHCS functionality. As illustrated in Figure 3 above, an Appointment Scheduling wrapper was created to access the Appointment Functions code. Accessing existing functionality reuses existing business rules and gets the application up and running quickly. Over time this functionality can be transferred to the new object layer transparently.

A New Beginning

The fundamental goal at NMCSD is to improve and streamline the patient care process. To accomplish this goal, the business process had to be improved. Improving that process meant being able to access the CHCS data and functionality via the ubiquitous web browser found on every provider's desk.

A Problem To Be Solved

Currently, patients must schedule subspecialty appointments by going to the clinics in person, waiting for mail notification of a potentially inconvenient appointment time, or waiting for phone contact from the clinic. Poorly coordinated appointments result in multiple trips, parking difficulties, additional time off work and increased patient dissatisfaction. Patients are frustrated because they often do not get to select their appointment at a time that is convenient for them and if the subspecialty clinic does not make contact with the patient, they may be lost to follow up. Providers are frustrated because they do not have access to historical clinical information at their desktop and do not have time to look for it due to the short patient appointment time. Consequently, important information may be missed or clinical studies may be repeated because the information is unavailable. Figure 4 below illustrates this process.

Figure 4: Old Appointment Scheduling Process

The Solution

Having the Easy CHCS web portal available at the provider's desk streamlines the process. Upon entry to the PCM Clinic, the patient is checked using the CHCS system. During the appointment the doctor can access all the patients data via the Easy CHCS web portal. If a follow-up appointment is needed, the provider can make it immediately based on the patient's availability. This approach eliminates the confusion and possible legal problems resulting from rescheduling appointments, missed appointment, etc. Figure 5 below illustrates the new process.

Figure 5: New and Improved Appointment Scheduling Process

Other Business Process Improvements – Proposed and Real

The new CHCS framework provides an environment that lends itself to supporting new business process improvements that were difficult, if not impossible, in the old legacy framework. Some of these improvements are listed below.

Clerk Scheduling

Because of the success of the appointment-scheduling project, another area of improvement is to enhance the web portal for clerk-scheduling functionality. This will enable scheduling clerks to book patient appointments via the web interface.

Surgical Assist

The current Surgical Assist application is used to build and track surgical procedures for a specific patient. The application is built on a stand-alone Access database. It does not access the CHCS system where the patient’s clinical record is actually stored. The Surgical Assist database duplicates data and does not scale to support the workload. Moving this functionality to a web-based environment that accesses the demographic and clinical data of CHCS will overcome many of the problems inherent to stand-alone database applications.

Neonatal Wireless Application

Immediate access to patient clinical data is critical in an Intensive Care Unit. Currently, access to data is only available through a dumb terminal that is inconveniently located. The Neonatal application will be available on a handheld device. As the Neonatal staff makes its rounds, patient data can be readily made available on the device. The data transmission and storage will be HIPAA compliant. This project has already been funded and is underway. Fundamental to making this work is the secure wireless communications capability of the EsiObjects system.

Immunization Tools

No development in modern medicine is more important to the health of an individual or that of a population than the advent of immunizations. And yet the Navy has no reliable mechanism for entering or tracking the immunization status of all its members. No mechanism exists for immunization data to effectively participate in population health or disease management initiatives. This problem can be overcome by gaining access to the County of San Diego immunization records of all service personnel stationed at NMCSD. These records reside on an Oracle database. However, the EsiObjects middleware can access and merge this data with the patient's demographic and clinical data in real-time, presenting a more complete patient record.

VA and DOD Data Sharing

The Military Heath System (MHS) and the Veteran’s Administration (VA) are rapidly approaching a time where the integration of patient data stored in CHCS and VISTA will prove to be an absolute necessity for good patient care. Although they share a common code heritage, the two clinical repositories have diverged over the years and conventional wisdom would have you believe that integration would be extremely difficult at best.

The NMCSD web-based provider portal can be enhanced so that data from both VISTA and CHCS could be integrated into a single graphical interface. Alternatively, data could be pulled from CHCS and merged into the VA CPRS system. We also believe it possible to generate unique patient identifiers that will be assigned on the first day of active duty and would carry through into the Veteran’s Administration upon retirement. These unique identifiers will essentially enable the collation of a patient longitudinal medical data across all of DOD and the VA into a single interface.

Patient Safety and Adverse Event Tracking

Medical treatment facilities large and small frequently rely on inefficient paper based occurrence reporting. Additionally, a recent NMCSD survey demonstrates that hospital staff often views the process as punitive in nature, or are unclear on what should or shouldn’t be reported. Consequently, adverse events are vastly underreported. The literature suggests a medication error incidence of 3/10,000, but even conservative estimates indicate an actual adverse event rate of at least 3/1000. Unfortunately, this data is vital to continuous improvement and patient safety initiatives.

A secure reporting tool using EsiObjects™ is proposed that can be assessed from any web browser, wireless device, or clinical application, to allow for rapid, automated, and accurate reporting of adverse events. The data would be stored on the server where automated metrics surveillance would alert management if trends were developing. It would also allow for profiling and focused educational efforts.

Radiology Support

One area of concern at NMCSD is radiology support at isolated duty stations. In the current environment, studies are ordered and processed by the CHCS host at the remote site. The images are then mailed or otherwise transported to a referral center for definitive interpretation. The patient’s registration, demographics, study information must then be reentered by hand in the receiving facility. Once the report is available, a copy is then sent back to the remote facility where the interpretation is again manually reentered into the CHCS host. The implications for efficiency and data quality are ominous.

A more efficient solution is to create a web based application that would allow the consultant radiologist to access the remote CHCS host, access new studies, view the images over a secure web connection, enter their interpretation directly into the remote host, and then copy any demographic data needed for work load accounting back to the local host automatically.

Readiness Tracking and Prevention Notification

Field commanders are often limited in their ability to rapidly determine the medical readiness of the forces under their command. The available tools do not enable them to collate disparate information in a single location automatically.

We propose to create a “Readiness Dashboard” that will monitor desired medical metrics, will alert a commander when a parameter is out of range, and will enable the email notification of individuals identified as needing preventive health interventions. This medical dashboard could be combined with training, dental, and legal metrics to enable a complete over-view of a unit’s readiness to deploy.

The Virtual Case Manager and Consult Tracking

Providers are ill equipped to manage and track the numerous consults, studies, and lab mail-outs they submit on behalf of their patients. Consequently, patients slip through the cracks, appointments are missed, results not checked, and needed interventions delayed until it may be too late for effective disease prevention. A virtual case manager can be created that will keep track of all consults, specialty studies, etc, and will notify the PCM or the clinic manager if appointments are not kept, if results are not reviewed, or consults not completed. The virtual case manager promises to dramatically improve our ability to ensure effective patient follow-up.

Benefits of the Easy CHCS Framework

In all healthcare information technology endeavors, we must never forget the reason why we build computerized systems – to improve patient care. The Easy CHCS framework provides full access to CHCS data and supports web and wireless technology that can expedite and improve the process of patient care. In general, the one goal Easy CHCS accomplishes is to bring vital data directly to the provider. All patient data can be accessed using the ubiquitous Internet browser, which is on every desktop. Also, wireless technology will offer the provider full mobility while accessing the information needed to provide care.

Using Easy CHCS results in some real benefits to our dedicated service personnel and their dependents.

• Old and inefficient business processes can be streamlined by making the data readily available thereby improving care.

• Data residing on the CHCS system is easily accessible and available to the provider thereby expediting the decision-making process.

• Data from different sources, including remote CHCS systems, can be accessed and integrated resulting in a more complete patient record. Treatment decisions are potentially more accurate and complete which translates to enhanced patient safety.

Dr. Emory Fry is a Neonatal specialist stationed at the Naval Medical Center San Diego.