Information is power, and perhaps a seed of corruption. In Michael Millenson's paean to medicine in the information age, once corporations have confiscated power from clinicians it absolutely does not corrupt.
Millenson, a Chicago-based health reporter turned consultant, is a true believer in the virtue of the information revolution. His engaging account of the development of information science and quality assurance in medicine is enlivened by clinical vignettes and brief biographies of key actors. The first of his principal theses -- that doctors, unsupervised and unaccountable throughout most of this century, committed a myriad of sins -- is surely correct. The second -- that medicine's new masters, having gained leverage through access to medical information, will make beneficent use of their power -- is surely questionable.
Millenson is at his best when describing the early, rebel, period of medical information science -- when it was largely the province of insurgent academics and liberal reformers. His account of the Food and Drug Administration's role in improving pharmaceutical quality is superb. He gives a nice glimpse of Archie Cochrane's seminal observations on effectiveness and efficiency, though he gives short shrift to Cochrane's insistence that all effective treatments must be free. He offers an accessible review of the pioneering work of Bunker, Wennberg, Barnes, Gittelson, and Mosteller on variations in practice patterns and of W. Edward Deming, the American engineer whose statistical quality-control techniques were first operationalized in Japan and were instrumental in transforming Japan's manufacturing from merely cheap to high-quality. Millenson nicely narrates the adoption of Deming's precepts by Ford Motors and Donald Berwick's elaboration of these ideas in medicine, spreading the gospel of Continuous Quality Improvement.
Millenson's rosy portrayal of today's corporation-dominated medical accountability, and his "celebration" of medicine's future (he assumes that we will continue on the current trajectory) are less convincing. He trusts employers to monitor their workers' care, ignoring the potential for conflicts of interest in such arrangements. Sometimes care that helps an employee hurts the firm -- expensive mental health services that might keep a marginally performing worker on the job may fall into this category, as well as employer-funded care for retirees or the aggressive diagnosis of workplace-induced illness.
Moreover, even if purchasers of care genuinely seek quality, the tools currently or foreseeably available cannot reliably detect market-driven cheating. In most instances, clinicians and health maintenance organizations (HMOs) literally create the data that must be used to monitor them. When such data are used as the basis for reimbursement they may have the accuracy of a tax return. Rewarding physicians for good outcomes in hypertension, among other things, rewards overdiagnosis. White-coated doctors who screen for hypertension using too-small cuffs in harried surroundings will label many patients "hypertensive," garnering a higher capitation fee based on risk adjustment, while virtually ensuring that most of these "hypertensives" will suffer few bad outcomes.
Millenson also ignores the distortion of medical care that may result when financial incentives are tied to quantitative measures of quality. Quality monitoring -- e.g., with HEDIS (the Health Plan Employer Data Information Set) often gives undue weight to the few items that are readily measurable. Listening, learning, and caring are difficult to quantify, and as one HMO executive put it, "It doesn't count unless you can count it." Counting Pap smears is a fine idea, but scarcely an adequate measure of primary care. When financial reward for performance on such counts is combined with pressures to increase productivity, too many women spend much of their visits in the lithotomy position.
Millenson seems unaware that the current corporate context may sabotage quality initiatives. Instead of a uniform formulary to inform prescribing, we labor with five different formulary lists, each embodying the deals cut between an insurance plan and drug manufacturers.
The book is also marred by occasional inaccuracies and important omissions. Hepatic failure is misdefined as diseased kidneys. A patient is said to suffer from "high blood gases." The pioneering work of Howard Bleich and Warner Slack in clinical computing, decision support, and improving access to medical literature receives no mention, and the Regenstreif Institute's groundbreaking and meticulously evaluated system of computerized medical records and feedback are ignored. Millenson's readiness to trumpet implausible non-peer-reviewed accounts of spectacular quality improvements is also disturbing (isn't this the mirror image of doctors' shoddy use of anecdotal evidence, which Millenson condemns?); hospitals' claims to have reduced transfusion reactions to zero and drug reactions to nine per year are uncritically reported.
Millenson's portrait of medicine in the information age is detailed and artful, but it omits the shadows that would give it depth.
Reviewed by Steffie Woolhandler, M.D., M.P.H.
Copyright © 1998 Massachusetts Medical Society. All rights reserved. The New England Journal of Medicine is a registered trademark of the MMS.
A former Chicago Tribune reporter turned healthcare analyst, Millenson has done yeoman's work in amassing and understanding the avalanche of data that lies beneath most of the managed-care headlines, with such medical horror stories as "drive-through deliveries." What he finds is both important and well explained: inconsistency, overlap, and inattention to quality measures in medical treatment cost more and are more dangerous than most cost-cutting measures. Millenson uses examples to show that quality healthcare is in fact cheaper, obviating the need for mindless "savings" that can cost money and lives. To assure such care, he urges that the mountain of available statistical information be made accessible to healthcare providers. Though his book can make for tedious reading (every sentence is a statistic), it elevates the healthcare debate to a new level and deserves a wide readership. Essential for health science and academic collections and worthwhile for general collections with health-conscious readerships.?Mark L. Shelton, Univ. of Massachusetts Medical Ctr., Worcester
Copyright 1997 Reed Business Information, Inc.
