Healthcare System Transparency and Patient Safety

When a few cautious experiments in the direction of greater health care provider transparency were proposed, some were alarmed. The worry was that so-called "Disclosure, Apology & Offer" (DA&O) practices would lead to increased litigation costs.

In a Perspective post in the NEJM, A. Kachalia discusses the immediate benefits for transparency, but also cites systemic obstacles to broader adoption of the practice.  (Kachalia is likely aware of the issues, guessing from the symbols listed after his name.) Citing the few known DA&O studies, he writes that:
. . . data from two pioneering programs have revealed improved liability outcomes, including a 60% decrease in legal and compensation costs in one program. Proponents of DA&O programs also tout downstream safety benefits from greater transparency. Early program successes have fueled extensive interest and a push for broader implementation, but there has not been immediate widespread adoption, so transparency is far from ubiquitous.
That's the promising news. Less hopeful: a complex mix of obstacles. Many of the impediments would be concerns in any setting, not only health care. They include:
  • A real or perceived tendency to single out a single provider's error rather than pursue a systemic solution
  • Unclear path to clear communication between providers, patients, family and public reportage
  • Even if organizations accept responsibility for "systems-level errors" (and if those can be defined), providers are singled out in reporting to state boards and quality monitoring data bases like the National Practitioner Data Bank. Worse, physicians may worry that, if caught up in systems-level errors, they could be "associated" with compensation to patients.
Kachalia also argues for legal reforms, such as greater use of "enterprise liability" instead of the current method which holds individual practitioners responsible. Another idea would be to have "administrative health courts."


The practice of medicine will become increasingly a systems-level enterprise. Transparency must become an integral part of those enterprises. The legal framework in which failures occur should take into account systems processes. Transparency should be fully integrated into the workflow -- both human and automated -- inside those enterprises, and nurtured by professional standards within health care related disciplines.

According to a 2012 DHS report, most medical errors are not reported, and even when they are, corrective actions are often not taken.

Increased transparency should be part of health reform. As more of the practice of medicine becomes software-assisted, e.g., through clinical decision support and workflow automation, initiatives such as those proposed by Kachalia will become increasingly important. Accountability in health care -- as with quality and cost containment --  must be seen as the corporate, enterprise effort that it has become.


Health Exchange Glitchware: Find the Responsible Contractor in the Haystack

Which among the numerous recipients of taxpayer-subsidized contracts to work with CMS worked on the glitch-plagued health care exchanges?  A Sunlight Foundation post lists many (all?) of the contractors associated with the Affordable Care Act (ACA). Unfortunately, this list is a superset of work performed for the agency. The New York Times reporting mentions CGI Group, Booz Allen Hamilton (Snowden's employer), Development Seed (home page) and Oracle (identity management component). 
An NPR blog post quotes Clay Johnson, a former government worker, who blames the contracting process. True -- this is an area overripe for reform. But this near-platitude misses the point. Some of the same contractors mentioned in the Sunlight report have performed satisfactorily on major complex efforts for the Department of Defense. For instance, Northrop Grumman worked on the X-47B UAV project, which succeeded in landing a large unmanned aerial vehicle (drone) on the aircraft carrier USS George H.W. Bush for the first time within two years of its first test flight. This system was smart enough to know when a landing had gone amiss and was able to fly itself back to home base instead of crashing in the ocean. These projects are not without problems -- including cost overruns and lobbying influences, but they show that positive results are possible at the same time as the inevitable taxpayer humiliations are endured.

The impression remains that had the health exchange project been properly managed and executed, success was within reach.

More facts need to come to light. Where is the information technology press on this? What problems were posed by the short timeframe, limited funding or other less obvious constraints?

In the meantime, here's a start at the Monday morning quarterbacking:
  • HHS should have enlisted help from major government contract managers, such as those at Mitre, government labs, the Defense Acquisition University, the CMMI Institute 
  • Great consideration for reuse of existing software and infrastructure, not only from commercial [Ed: and open source] sources as many have suggested, but from existing government portals. Consider, for instance, those that already provide important functions for millions of veterans and their families 
  • Use existing frameworks for management and development, such as DoD Architecture Framework or UPDM 
  • Provide greater design transparency to publicly vet the overall architecture with other interested parties, such as what NIST has done with cloud and Big Data technologies 
  • Hire an independent project journalist with access to all on the project team