E 2 635 harsha
Kudos to Robert Kaplan and Michael Porter for their illustration of careful process analysis and cost accounting in health care. Their idea is terrific—but it is hardly novel. Geisinger Health System in Pennsylvania and Intermountain Healthcare in Utah have long employed process analysis, and in my 1997 book, Market-Driven Health Care, I advocated activity-based costing of medi- cal care bundles, which I called “health care focused factories.”
Kaplan and Porter barely mention, how- ever, that people are not fungible, so their costs cannot be measured like the costs of
widgets in manufacturing. Some patients are much sicker than others. A process fre- quently called risk adjustment accounts for these differences. The impact of risk ad- justments on costs can be enormous. In a risk-adjustment and activity-based costing analysis that my students and I performed for a total-knee-replacement procedure, we found that while the average payment was $35,000, the top decile averaged $615,000. Useful risk adjustment requires complex statistical analysis. If it were done as the authors suggest, by simply adding comorbidities, the analysis could create
thousands of different products, each requiring separate analysis.
Furthermore, the implementation of activity-based costing can “solve the U.S. health care cost crisis” only if, as the authors suggest, payers switch to reim- bursing providers for value and bundles of care. But those adopting the Kaplan-Porter methods could repeat the mistakes of the California health care providers that used process analysis to price their offerings to insurers in the 1990s: Many suffered sub- stantial financial losses, in part because their prices lacked risk adjustments and reinsurance to protect them against ad- verse selection by very sick patients. Regina E. Herzlinger, Nancy R. McPherson Professor of Business Administration, Harvard Business School
HBR article by Robert S. Kaplan and Michael E. Porter, September 2011
“U.S. health care costs currently exceed 17% of GDP and continue to rise,” say Harvard Business School professors Kaplan and Porter. They trace spending to its source: health care providers. Doctors, nurses, and specialists do not understand the value of medical care to the consumer; they overspend because they can’t accurately measure health outcomes. The authors take a look at providers that are measuring costs the right way, and then prescribe a cost-measurement system based on individual patient conditions.
How to Solve the Cost Crisis in Health Care
Why Your IT Project May Be Riskier Than You Think HBR article by Bent Flyvbjerg and Alexander Budzier, September 2011 The authors’ research found that one in six IT projects finishes wildly over budget, with an average cost overrun of 200%.
The article avoids the root cause of IT failures: the lack of a framework for effective collaboration. Far too many organizations can’t do joined-up think-
ing across the enterprise. I see only one universal language that can fa- cilitate this essential communication: the language of end-to-end business
process management. Combine that with the ubiquitous need for compliance, and it’s clear why effective process manage- ment is emerging as a core capability for Enterprise 2.0. Michael Gammage, VP product marketing, Nimbus Partners
Outlining how the project will potentially affect both internal and external people (for example, partners and customers) is essential to understanding the project’s
22 Harvard Business Review November 2011
Interaction
The authors respond: We are pleased that Professor Herzlinger endorses our approach of building cost models based on good process maps, but her letter con- tains numerous errors. Like many in the health care sector, she confuses charges with costs (Myth #1 in our article) when she describes her students’ analysis of the variation in total-knee-replacement payments. She errs in her assertion that we introduce an adverse selection problem. We explicitly solve the problem by capturing costs at the individual pa- tient level. Providers and payers can use patient-specifi c information to explain and predict the cost and outcome varia- tions caused by some being “much sicker than others.” Herz linger asserts that we assume all patients are as alike as widgets. Just the opposite! Our approach captures how costs vary when providers respond to the diff erent medical needs of individual patients. And her concern that costing thousands of diff erent patient types would be complex is unfounded. The scale of the model is driven by a provider’s diff erent processes and resources, not by the number of diff erent types of patients it serves.
The best way to comment on any article is on HBR.ORG. You can also reach us via E-MAIL: [email protected] FACEBOOK: facebook.com/HBR TWITTER: twitter.com/HarvardBiz Correspondence may be edited for space and style.
Interact with Us
risk factors. And incorporating them into a risk-management plan helps companies better navigate through tricky project waters. Savvy IT and project sponsors acknowledge that human factors must be considered when building the business case. Alicia A. Lewis, manager, PricewaterhouseCoopers
HBR.ORG
Available in two scheduling options:
Four Consecutive Weeks April 29 – May 25, 2012
2x2 Option April 29 – May 11 and July 15 – 27, 2012
To learn more, visit:
RealizeYourLeadership.com
Columbia Senior Executive Program
How to realize leadership potential
2. Call for Help
4. Apply Resolution
3. Address the Issue
1. Recognize Organizational Problem
Practical skills for the business-minded
1377 Nov11 Interaction Layout [S];25.indd 231377 Nov11 Interaction Layout [S];25.indd 23 9/27/11 2:13 PM9/27/11 2:13 PM
Copyright 2011 Harvard Business Publishing. All Rights Reserved. Additional restrictions may apply including the use of this content as assigned course material. Please consult your institution's librarian about any restrictions that might apply under the license with your institution. For more information and teaching resources from Harvard Business Publishing including Harvard Business School Cases, eLearning products, and business simulations please visit hbsp.harvard.edu.