Designing electronic health records versus total digital health systems: a systemic analysis.(Research Paper)(Report)

INTRODUCTION

Until today, the development of electronic health records (EHRs) applications for the health care industry has been constrained by two factors: the design concept is limited to monolithic systems and designs appear to have been technology/ vendor driven (Kuhn and Giuse, 2001; Berg et al., 2003; Blobel, 2006). Nearly all current systems are simple relational database applications, not unlike databases found in other industries. They are used for patient data entry (forms), asking questions and producing reports. Not one of the EHR applications currently in use have the capacity to scale up to fully functional applications (Institute of Medicine, 2003), nor are they commercially viable as interoperable elements in a large-scale distributed environment (Hippisley-Cox et al., 2003; Koontz, 2005; Powner, 2005; Powner and Koontz, 2005; Powner, 2006). Moreover, as vendor driven entities, they do not adhere to generalize standards for portability. These shortcomings are due primarily to a fundamental design flaw: the failure to reflect fully on the diverse systemic features of the health care process (Koontz, 2005; Powner, 2005; Powner and Koontz, 2005; Powner, 2006).

To produce viable EHRs, designers need to take a holistic approach to their work. One example is a distributed systems approach, such as a network-centric approach (von Lubitz and Wickramasinghe, 2006) or service-oriented architecture (SOA) with web interactivity (Raghupathi and Kesh, 2007). Following more flexible modelling methods, such as the model-driven architecture (MDA) (Raghupathi and Umar, 2008), will also help designers. What will vault health information technology design into its rightful place in the 21st century is designers taking a holistic approach to development, an approach that considers what technology can do but more importantly how the health care industry operates and how the people involved in it will want to use that technology.

A second area of health care information systems, the one we call the total digital health system (TDHS), has the potential to take these systems to the next paradigmatic level, giving them the properties of openness, dynamism and interoperability (Rollins, 2005). Federation, too, will be a key property of the TDHS as health care delivery participants (physician clinics, HMOs, hospitals, diagnostic labs, pharmacies and so on) strive to share health information within the context of ethics, privacy and security (Mandl, 2000).

These new TDHSs offer design improvements that differ from those of the traditional EHRs. For instance, they typically allow for (1) the residence of data in multiple locations (e.g. insurance company, primary care physician's office, hospital, lab, pharmacy, government entities such as Medicare and Medicaid); (2) the distributed nature of various processes (e.g. scheduling a patient visit, ordering lab work, writing prescriptions, capturing diagnoses and treatment data); (3) usage (by different participants involved in the care such as physician, nurse, hospital discharge planner, case manager); (4) interoperability across multiple platforms (obviating incompatibility problems among software and legacy systems housed with the various health care delivery participants) and (5) real-time processing (making patient data available anywhere, anytime). In the not-so-distant future, it is conceivable that advanced, high-end EHRs will evolve into TDHSs.

This paper discusses TDHS design issues arising from systemic differences (Boulding, 1956; Churchman, 1971; Van Gigch, 1978; Van Gigch, 1991) between TDHSs and EHRs. We propose that different design methodologies be used in the two domains. First, though, we review EHRs, and then compare and contrast the design issues in both types of applications. We highlight key modelling issues and discuss, in the context of the two domains, the systemic properties of softness, openness, complexity, generality and purpose. We identify additional domain issues, give operational examples of TDHS and EHR applications in context, and discuss the requirements of the TDHS. We also propose a soft systems methodology (SSM) for the high-level design of the TDHS. Finally, we offer our conclusions.

ELECTRONIC HEALTH RECORDS