In my post "Will the U.S. spend the Economic Recovery Act's $20 billion for Healthcare IT more wisely than the UK?", I expressed concern that the answer to the titular question might in fact be "no."
In the next few posts I present a few recent cases that stand as exhibits to my thesis above, the first being mine personally, the others being that of other individuals who shall remain nameless.The overarching theme is that hospitals are significantly behind other industries in IT strategy and leadership, an "IT backwater" if you will.
Congress, are you listening?
Several months ago (just before the appearance of the Joint Commission Sentintal Event Alert on HIT and the National Research Council report on HIT that stated approaches to HIT are "inadequate"), I received an inquiry from an executive recruiter.
The recruiter had been retained by a large East Coast hospital system to hire a Director of Informatics.
The recruiter told me three things: first, that I was the first candidate he was speaking with. Second, that he knew little of Biomedical Informatics, and could I please inform him about the field. Third, that the hospital had already selected an EHR vendor and was now looking for an expert to make it all work.
I refrained from asking the questions I ask here:
- Why does a hospital system retain a recruiter for a highly specialized executive position who knows nothing about the nature of the field?
- Why doesn't the NFL retain me, a physician informaticist, as a talent scout for football champions? ( the second question is sarcastic with reference to the first, of course.)
I did spend some time explaining to the recruiter than under the right circumstances and with the appropriate hospital leadership, I might be interested in the position, although the "system selection first, expert selection second" was a fundamental strategic error.
I then explained the challenges and issues in the field that led many hospitals astray and resulted in project difficulties and failures. Being new to the field, I'm not sure the recruiter believed me. In my own due diligence, I noted the CIO of the organization was a typical hospital MIS professional and had no biomedical or biomedical informatics experience.
At the conclusion of our conversation, the recruiter said that there were many more candidates he would talk to and after that, if the organization expressed interest in me, he would contact me again. I did not expect to hear from him again as I had deviated from the traditional "health-IT-it-must-be-good" religion.
Lo and behold, just a few days ago I got a call from the recruiter asking if I was still interested in the role. I asked for specifics on reporting and on resources the position would control.
The answer: The position reports to the CIO (a non medical professional). To the position reports 1.5 FTE's. That's one-and-a-half FTE's.
I responded to the recruiter along these lines (see my ten year old essay "Ten critical rules for applied informatics positions: What every Chief Medical Informatics Officer should know" for more on these issues):
- I ran a biomedical research library in pharma for drug discovery and had more than 50 staff reporting to me. How could I run an entire HIT implementation for a multi hospital system - with patient lives at stake - with 1.5 FTE at my disposal?
- How would going from 50 FTE to 1.5 FTE help my career?
- Why would the CIO and hospital believe an informatics domain expert who'd once been a CMIO at a large hospital should only have 1.5 FTE reporting to him/her?
I didn't ask these, but perhaps should have:
- What reasonable rationale does a non medical CIO have in requiring that a physician informaticist building clinical tools that happen to involve computers report to them, rather than to medical leadership?
- What kind of weak medical leadership would allow such a reporting relationship?
Indeed, such an organizational structure was a strategic error, out of the same playbook as the "select vendor then select expert" strategy as above. The position is actually a non-management non-executive "Director of Nothing" role, engineered for frustration and failure.
(Some would argue that dotted-line relationships are sufficient for such a Director of Informatics role, which in reality creates an "internal consultant" role at best. I would argue, then, that all formal hierarchical relationships in hospital MIS departments should be ended and replaced with dotted lines. I don't think that would sell well to a CIO.)
I can also make a few additional points to those who might opine that "doctors can't manage people":
- Medical residents have more direct reports, in far more complex and mission critical settings, than one and a half. I had ten or more med students, interns, jr. residents, pharmacy trainees, etc. during my own medical residency, in diverse settings including ICU's, intermediate units, and regular hospital floors.
- Private physicians have been managing staff since modern medicine began (and probably long before as well!)
- An enterprise clinical IT project is at least as complex an endeavor as, say, a clinical trials information system in pharma. You can be sure pharma does not put generalist MIS personnel in charge of clinical trials information systems and then appoint a single domain expert with only 1.5 FTE at his or her disposal.
The above points would be completely lost on most hospital IT backwater CIO's, who even if they did understand would probably be loathe to give up the territory to a "non IT professional" such as an informatics trained physician, anyway.
Rare as hen's teeth in hospitals is the appropriate position, i.e., true VP of Medical Informatics. In effect, hospitals put at risk millions of dollars of IT investment and patient well being due to the backwardness, reactionary bias and ego issues of their CIO's and other executives who believe that an empowered physician informaticist, even when available, "is not a sine qua non of health IT projects."
I told the recruiter I was not interested in this role as it was ill conceived.
In a stunning example of "Physicians' Expected Helplessness", though, the recruiter tried not once, but twice, to pick my brains on how he might advise his client hospital to structure the Director of Informatics role. I declined to provide free information, but did offer my services as a consultant on position structure to the hospital. The recruiter made as a condition to presenting me to the hospital as a possible consultant that I tell him some of my specific ideas. I declined.
Such scenarios are not uncommon. I know several physician informaticists holding such roles currently. Most are quite unhappy due to the imbalance between responsibility and authority in a most political type of job; those that claim not to be are either in denial or do not desire true leadership roles.
In summary, $20 billion is about to be given by the Federal Government to healthcare organizations who make major strategic blunders on HIT (and with their own precious capital, let alone government handouts!), hire retained recruiters who need education on HIT issues and biomedical informatics, and then compound their errors by structuring critical roles in a manner guaranteed to make maximal contributions from the experts either unpleasant or impossible.
It should be understood I am not "against" health IT nor a luddite. I completed a postdoctoral fellowship in Biomedical Informatics in 1994 out of love for the idea of improving healthcare through IT. It is clear, however, as these examples and others at my academic website "Common Examples of Healthcare IT Difficulty" illustrate, that significant further research is needed in order to determine how to best make this technology meet the needs of real world clinicians, and how to best implement it under real world conditions.
I do not advocate abandonment of health IT, only a return to the understanding that this technology remains largely experimental. This understanding was usurped in the past decade by an overaggressive and indeed opportunistic HIT industry, enamored of profit potential, and the irrational exuberance has now spread in a manner reminiscent of other speculative bubbles of late.
HIT should be treated as experimental, not as a drop-in panacea for healthcare's ills. In its present state is is perhaps as likely to exacerbate those ills. This is probably not technology that should be deployed en masse at present. We cannot afford as a society to learn how to do this by trial and error.
Do we today have a hospital IT backwater $20 billion abyss? I report, you decide.
Case two of this series is here, case three is here.