The Washington Post asks some important questions about electronic medical records:
…bipartisan enthusiasm has obscured questions about the effectiveness of health IT products, critics say. Interviews with more than two dozen doctors, academics, patients and computer programmers suggest that computer systems can increase errors, add hours to doctors’ workloads and compromise patient care.
“Health IT can be beneficial, but many current systems are clunky, counterintuitive and in some cases dangerous,” said Ross Koppel, a sociologist at the University of Pennsylvania School of Medicine who published a key study on electronic medical records in 2005.
Under the stimulus program, hospitals and physicians can claim millions of dollars for IT purchases, and will be penalized if they do not go digital by 2015. Obama has said the changes will save billions in health-care costs and will minimize medication errors.
But health IT’s effectiveness is unclear. Researchers at the University of Minnesota found in March that electronic records prevented only two infections a year. A 2005 report in the journal Pediatrics found that deaths at the children’s hospital at the University of Pittsburgh Medical Center more than doubled in the five months after a computerized order-entry system went online. UPMC said the study had not found that technology caused the rise in mortality and maintained that medication errors were down 60 percent since computers were introduced in 2002.
I’ve written a great deal about success and failure rates in IT in addition to teaching courses specifically tailored to improving the likelihood that technology might deliver on its intended outcomes. What such discussions often miss is the notion of the stealth failure: systems implemented on time, on budget, and according to requirements that nevertheless complicate the processes they automate, costing user time and errors and spawning workarounds that may or may not address the reasons for developing the systems. In the aggregate, these stealth failures can bear a heavy cost, but they are almost never considered in the aggregate. Electronic medical records have a strong potential for becoming stealth failures, exacting a heavy toll not only in financial costs, but in quality and length of life.
Of course, these outcomes are not inevitable. Careful attention to use cases, outcome-based requirements, and relationships between interdependent purposes go a long way toward catching and addressing the elements of stealth failures. Unfortunately, political processes tend to undermine the communications and clarity necessary to realize better outcomes.