Friday, October 16, 2009

Fuzzy Math Indeed: Rising Costs in Government's Digital Health Stimulus

Seen in the blog of the Huffington Post Investigative Fund:

Fuzzy Math? Rising Costs in Government's Digital Health Stimulus
Spending Could Be Double The Obama Administration's Public Estimate of $19 Billion

By Fred Schulte
Huffington Post Investigative Fund

Creating digital medical records for every American within the next five years – a key provision of President Obama’s stimulus package -- could cost more than twice the $19.5 billion figure that has been cited by federal officials.

Federal budget documents show that actual spending for the plan, which will use stimulus money to help doctors and hospitals defray the cost of installing high-tech records systems, could hit nearly $47 billion.

The discrepancy between the Obama administration’s $19.5 billion public estimate and the entry in the budget of the U.S. Department of Health and Human Services mostly arises from the government’s calculation that it will recoup some of the cost of subsidizing digital systems through billions of dollars in reduced federal health spending. But many health analysts are highly skeptical that such savings can be accurately predicted.

It's worse. Far worse.

I believe the government is underestimating EHR costs by a factor of 5 to 10, due to implementation failure and remediation costs.

I'm a long term optimist on health IT, but a short term realist. I believe we will be having an avalanche of failed/stalled projects such as this one in years to come: "Another Major HIT Project Setback at UCSF".

The incompetence in this industry is worse than I imagined when I started writing about HIT dysfunction a decade ago. Worse are the coverups of problems that occur in implementation and in use.

At least in the U.S. military, there is transparency about health IT problems (e.g., see here). In the private sector, forget about publicity of what actually goes on in the U.S. (The U.K. has increased its own transparency to some extent.)

A spokesman for the department [U.S. Department of Health and Human Services], Nicholas Papas, said that the $47 billion figure is “undergoing revision and we anticipate it will decrease.” He said the agency expects to generate a “firmer figure” by the end of the year, but declined to say how much it would be.

Papas said there was no attempt to lowball the number for public consumption. Calculating both cost and benefits in the government’s estimate is a sound way to explain the price of the plan, he said.


Unfortunately, these calculations do not factor in the cost of dyscompetence, incompetence and failures created by lack of essential cross disciplinary expertise and talent in the HIT industry.

Many health policy experts agree that electronic health records can cut costs by reducing harmful medical errors and wasteful spending.

I would bet these "policy experts" have never tried to implement clinical IT.

The $19 billion figure has appeared in numerous news articles and been widely picked up and repeated by many bloggers covering the fast-growing health information technology industry. Earlier this month President Obama’s chief technology officer, Aneesh Chopra, mentioned the $19 billion figure twice in a live video cast from the White House.

But it’s “totally hypothetical,” said Sharona Hoffman, a professor of law and ethic at Case Western Reserve University School of Law. She said it counts on savings “with no concrete evidence they will occur.”

It's worse than "totally hypothetical." A major confounding variable is simply missing - the cost of HIT failures, false starts, delays, and remediation. More can be read about those risks here. That site has been online for a decade now, and ironically is nearly the only relevant "hit" on a search engine query on "healthcare IT failure" or similar concept.

Has anyone in government been paying attention?

A slide deck I recently presented entitled "Health IT Promises and Threats" provides a dose of reality on supposed cybernetic miracles. It is at this link (PPT).

-- SS