Commoditising Healthcare IT

The Next Wave

Werner van Huffel

Werner van Huffel

Health and Social Services Industry Strategist, Regional Public Sector Group Microsoft Asia Pacific, Singapore.

quotes

With the costs of healthcare rapidly increasing, the monolithic model of HIT is no longer sustainable. HIT commodity capability that provides a new level of convenience and serviceability to the healthcare environment while being cost-effective.

quotes

Healthcare and Information Technology (IT) have been linked for decades. Some of the earliest uses of IT in healthcare were for the creation of Artificial Intelligence and Expert Systems—such as MYCIN and CADEUS—in the 1970s. These early implementations were heavily influenced by mathematics. Their implementation was labour-intensive as they utilised bespoke development techniques. They were complex to use and monolithic in structure. Monolithic systems are generally unique to clinical domains and are:

  • Costly to develop and maintain
  • “Closed loop”: not designed for easy integration with other systems
  • Tightly-bound to the requirements of the clinical domain they service: it would be easier to write an entirely new application than to reapply the functionality
  • Monolithic systems also have other, less obvious, implications:
  • They impede the diffusion of technology innovation within healthcare because systems development is slow and expensive
  • They waste capital expenditure as they cannot easily be reused
  • Patient safety is compromised because it is difficult to change system interfaces to comply with clinical domain or area standardisation requirements

As the field of Healthcare Information Technology (HIT) has progressed, our ability to manage, manipulate and invent Information Technology (IT) solutions outside of HIT has grown at a rate comparable to Moores’ Law. Our understanding of architectural frameworks, such as Enterprise Application-centric schemas (e.g. Zachman, TOGAF and others), has grown to incorporate more loosely defined capabilities such as those covered within Service Oriented Architecture frameworks. These architectural strategies are based on the implemented technologies: rules processing engines, workflow and document process management systems, databases and development environments. While these should all be part of the core IT infrastructure, many domain-specific solutions are still developed within the monolithic model (i.e. reproducing these core software capabilities as bespoke components).

To the objective observer, HIT has progressed very slowly in most areas and very quickly in others. As the clinical environment has become more complex, the data storage and processing requirements of the clinical domains have also increased. The data explosion caused by the advent of genomics, large scale requirements for electronic medical / health records and the widespread use of electronic documents is straining existing monolithic systems.

No one system will do everything for healthcare, ever. One has only to review the sheer number and the complexity of systems in a healthcare institution’s infrastructure to realise that “one size does not fit all”. Furthermore, the rapidly increasing costs of healthcare means that the monolithic model of HIT is no longer sustainable.

What is a HIT commodity?

A HIT commodity is a cost-effective capability that provides a new level of convenience and service ability to the healthcare environment. From general experience, a commodity is also a capability which has become ubiquitous, to the extent that it becomes background “noise” to the standard operation of an environment. In other words, successfully commoditised IT implementation becomes transparent to the people using its functionality. The transition from monolithic to commodious models is an ongoing, ever-increasing, drive within HIT.

In HIT, the technology implemented in healthcare has been undergoing continuous commoditisation since the earliest initial bespoke implementations. This is primarily because, in many cases, we understand the requirements of pure technology implementations better than we understand those of healthcare. Much of HIT operates within silos—information exchange is disjointed and collaboration is limited. This is a legacy of the history of healthcare itself, as well as the uses and implementation of IT within healthcare.

Where are we going?

In some respects, technology in healthcare has not come very far in the past 40-odd years. There are many reasons for this; some studies have shown that poor communication between healthcare professionals and the IT systems they use may be a contributor. Many developments within healthcare still utilise the legacy method of monolithic, bespoke development similar to the era of MYCIN and CADUES—consuming substantial development and infrastructure resources and delivering little interoperability capability.

Generalised technologies such as processing engines, databases and unified communications have begun to penetrate healthcare. There is a growing understanding that technological rigidity is not always the best approach to solving healthcare problems. To be fair, IT is not as mature as healthcare and is still coming to terms with the mechanics of change in areas other than itself (such as clinical domains). Standards can facilitate transformation, but standards change. When this happens, standards can initiate the very confusion that they attempt to alleviate.

In order to be effective in the future and have an impact in healthcare IT, systems need to be:

  • modular (be able to plug-and-play with varying clinical domain requirements)
  • user friendly (to increase their reach and minimize pushback from clinicians around data collection and monitoring systems perceived as ‘Orwellian’ initiatives)
  • commoditised (cost-effective with high societal usefulness)

The work done by standards bodies is an attempt to bring collaboration and information interchange capabilities to the healthcare domain in usable chunks. In many respects, standards bodies are leading attempts to drive commoditisation of IT in healthcare. Global bodies such as the IHE, HL7 and others are actively attempting to facilitate interoperability through the definition of standards. However, implementation of collaborative capability in HIT still depends on technologists and technology companies working in concert with healthcare professionals.

The key to commoditised HIT is for technology companies to further embed HIT requirements as capabilities within their products and architectures. Those requirements will include:

  • increased use by vendors of frameworks that support dislocated applications, such as Software as a Service(SaaS) and Service Oriented Architecture
  • creation of standards-agnostic (not atheistic) systems that extend capability directly to the desktop through the addition of modular, user-friendly functionality to enable the clinical
  • information worker to exchange clinical documents without the need for costly and complex software
  • introduction and penetration of systems which deliver high-end value-added capabilities, such as Electronic Medical / Health Records, as a commoditised unit of production rather than an ad-hoc amalgam of systems
  • healthcare domain-enabled search, as opposed to generalised search
  • creation of platforms for interoperability in which the entire background infrastructure of an operating system or physical architecture can be geared to facilitate healthcare information exchange

Many enterprise-scale and niche providers are working in collaboration with clinicians to understand the requirements for such healthcare solutions. This will deliver greater choice in HIT and can improve the chances of an organic resolution of the issues plaguing HIT, rather than the present dipolar battle that appears to be raging.

As more and more healthcare domain requirements are embedded onto the operational substrate of IT, proof of their applicability and suitability increases. This will in turn create more of a market for commoditisation. This ability to meet and further commoditise HIT requirements, without detrimentally impacting the delivery of care, is the next wave.

References:

  1. Moore GE. Available Online [ftp://download.intel.com/research/silicon/moorespaper.pdf]. Accessed 21 September 2007
  2. Comparison of the Top Four Enterprise Architecture Methodologies. Available Online [http://msdn2.microsoft.com/en-us/library/bb466232.aspx]. Accessed 10 June 2007
  3. IT in health care. Available Online [http://www.medpac.gov/publications/congressional_reports/June04_ch7.pdf ]. Accessed 19 September 2007.
  4. Nadkarni PM, Miller RA. Service-oriented Architecture in Medical Software - Promises and Perils. J Am Med Inform Assoc. 2007;14:244 –246
  5. Rogers EM. (2003). Diffusion of Innovations, Fifth Edition. New York, NY: Free Press
  6. Lairson DR, Chang Y, Bettencourt JL, Vernon SW, Greisinger A. Estimating Development Cost for a Tailored Interactive Computer Program to Enhance Colorectal Cancer Screening Compliance. J Am Med Inform Assoc. 2006;13(5):476-484
  7. Coiera E. When communication is better than computation. J Am Med Inform Assoc. 2000;7:277 -86
  8. IHE, Integrating the Healthcare Enterprise. Available Online [www.ihe.net]. Accessed 24 September 2007
  9. HL7, Health Level 7, Available Online [www.hl7.com]. Accessed 24 September 2007

TOP