The future of the mainframe in healthcare: Two scenarios

Mainframe use in the healthcare industry may look different depending out the outcome of the 2008 presidential election. Each candidate has a different healthcare system proposal that would be a change from the present system.

The healthcare IT mainframe market has grown and changed over the last seven years. Health-insurer pressure to contain costs has meant more hospital use of mainframes in areas such as billing and quality management. HIPAA regulations have demanded better security of patient records, an area in which the mainframe excels. The need to pass patient records from organization to organization has sparked a surge in demand for standardized patient Electronic Health Records (EHRs), which require mainframe-style increased storage as well as Linux-style Web standards. And doctors' realization that the ability to store and query on huge MRI and electrocardiogram image databases yields diagnoses in minutes instead of hours relates the mainframe directly to the healthcare industry's critical success factors.

Mainframe software advances medicine

Increases in the importance of the healthcare market, and of the mainframe within that market, are likely to continue. Imaging studies alone are driving increases of two petabytes a month in healthcare image storage needs -- the kind of scalability that the mainframe excels in providing. Upcoming potential trends such as sharing information across borders, remote health coaching, and partnering between clinics/pharmacies and hospitals will call for much higher distributed activity, which will require the scalability of the mainframe paired with the communications of distributed systems.

Meanwhile, IBM has some interesting mainframe software to handle these present and future trends. IBM wants to deliver "info-based medicine." On top of its present horizontal software stack -- which includes content management with Content Manager and FileNet -- IBM is adding Health Info Analytics and Workflow Health Integration software (i.e., data mining and workflow). An information lifecycle management (ILM) solution for healthcare includes Tivoli Storage Manager (TSM) HSM, which migrates inactive files to lower-cost storage using policy-based management, and General Parallel File System (GPFS), a high performance, shared-disk file system. A new IBM grid solution for healthcare, the Grid Medical Archive Solution, differs from the usual concept in being cross-site, and thus allows "virtual disaster recovery" and storage "capacity on demand." It's intended to be enterprise-wide rather than cross-organization. The grid assumes data input through a "gateway node", which feeds system, control, archive, and storage nodes.

Political change brings new mainframe utilization strategy

Imaging studies alone are driving increases of two petabytes a month in healthcare image storage needs -- the kind of scalability that the mainframe excels in providing.

However, the future of healthcare's use of mainframes is not entirely determined by, or predictable from its past. The present political campaign has surfaced two different visions of healthcare's future, with two different healthcare systems -- different from each other and from the present system -- as the result. How would this affect mainframe usage in healthcare?

The key differences between the candidates are in their approaches to achieving universal healthcare coverage. Senator McCain would increase use of private insurance plans. Increased private insurance, in turn, would increase the number/size of, and the patient health data in these insurers' databases. As a result, the mainframe's use as a repository of patient data would spread to the insurance companies, creating not only new privacy concerns but also new needs for standardized communications of new data types between insurers and healthcare providers. In other words, the mainframe would be asked to underpin a highly distributed dual coordination between insurer and provider, not only about costs but also about patient health data.

Senator Obama's plan, by contrast, aims for increased use of governmental insurance. Here the focus would be on regulatory and reporting requirements on healthcare providers, accompanying the government's need for more data to track insurance effectiveness. Superficially, this seems similar to McCain's effect of creating standardized communications between insurer and provider covering both costs and patient data. In the real world, this would probably mean a centralized repository on the government's side, greater emphasis on achieving EHR standardization, and a slower increase in the overall need for storage scalability. In other words, under Obama's plan, more centralization would mean greater use of the mainframe at centralized governmental sites and less need for the mainframe at insurers. In today's mainframe market, the hospital is a key customer; in McCain's scenario the insurer might also be a key customer; in Obama's, the government might also be a key customer.

The challenge: Harnessing the power of computing in healthcare

One caveat: it is clear that the challenge going forward is not just (or mostly) about designing the basic components and systems; IBM, among others, has useful technology galore today. The problem is getting providers to use the technology effectively, by standardizing records, identifying ways to speed processes and inter-hospital and provider procedures, and focusing on building massive rich-media databases. It's also getting insurers to change from penny-pinching critic to a data-collecting enabler, and providers shifting from the role of single incident sickness processors to group-driven ongoing healthcare managers. In other words, IT will happen, but cultural change is the needed ingredient to make it much more effective, much less costly, and much swifter to implement. So under any scenario, the healthcare market will change fundamentally, and the mainframe's role will change correspondingly; but whether that change is slow or fast really depends not on the candidates or vendors, but on the providers themselves.

About the author:
Wayne Kernochan is president of Infostructure Associates, an affiliate of Valley View Ventures. 

Dig Deeper on IBM system z and mainframe systems