Hardhats: Four Major Strategic VA Missteps

"VA reorganized and centralized its IT resources under the VA Office of Information & Technology (OIT). Rather than centralizing selected corporate wide functions and resources (e.g. planning, policy, security), the VA went further and centralized IT development and operations resources of all administrations under OIT. Rather than following best practices that have emerged and are being followed by many other organizations with regards to decentralizing management of these type of resources to the business side of the house, the VA chose to pursue an approach that was widely used in the past but which are no longer strongly advocated. The expected result will be a loss of innovation, lack of responsiveness to business needs, and a tendency to lose its leadership position in the healthcare industry, especially with regards to health IT systems.

The VA Office of Information & Technology (OIT) recently chose to award a contract to Cerner to replace the VistA laboratory module. This seems to be part of a pattern of steps being taken to unplug VistA software modules over time and replace them with commercial off the shelf (COTS) products. This strategy leads to a system composed of a mix of multiple, costly commercial systems that will all need to be interfaced - an approach that many private sector hospitals have put in place and been roundly criticized for following. The VA has consistently been praised for the tightly integrated VistA software modules that are written in the same language, use the same database, and require no extensive interface solutions. This appears to run contrary to recommended best practices in health IT.

The VA Office of Information & Technology (OIT) has begun to centralize its hospital information systems at four regional processing centers. Unfortunately, one of these centers recently went down basically disabling all the VA healthcare facilities on the west coast, Hawaii, and Pacific Islands. While on the surface this was an unfortunate incident and centralizing IT systems seems cost effective, that may in fact be way off base. If you think of centralizing data that will simply be viewed by users, like a personal health record (PHR), this makes sense. However, if you use a different analogy and realize that hospitals are very similar to factories, and hospital information systems are interfaced to laboratory equipment, radiology x-ray and MRI systems, pharmacy Rx dispensing units, bed control, and food handling systems this makes no sense. Patient care at a hospital is handled in a very systematic way, with patients moving down a production line with specialized components put in place to serve the needs of the patient and healthcare provider - much like a healthcare 'factory'. Ford or GM would not centralize the operation of its information systems and equipment located in a factory that make up the line used to produce and/or care for their product - cars. Similarly, centralizing a hospital system being used in the minute to minute care of a patient as they move through the hospital makes no sense. You can centralize a PHR or EHR data repositories to view data, but not a heavy duty industrial automated system being used in patient care.

Finally, the VA Office of Information & Technology (OIT) has begun to disengage from its health IT partnering activities with the larger 'open' VistA community that has grown quite large. This unfortunately leads to an inability of the VA and the federal government to tap into the knowledge and development resources that community is willing to make freely available to the VA. Given the continual change to healthcare and health IT systems, this is a potentially great loss. It slows down innovation, ability to rapidly change, increases costs, and harms efforts to help the country to move towards interoperable EHR systems and health information exchange (HIE) networks.

If I could think of four strategies I would think the VA would absolutely choose to avoid, the new OIT management team has chosen to pursue all four."

<  |  >