Comments

Health Affairs encourages readers to engage in discussion via comments on our Web site.

• To RESPOND to a particular article: Click on the link "Submit a response to this article" in the box at the top right-hand corner of the article.
• To READ responses to a particular article: Click on the link "View responses" in the box at the top right-hand corner of the article.

Comments to:

Jan Walker, Eric Pan, Douglas Johnston, Julia Adler-Milstein, David W. Bates, and Blackford Middleton, The Value Of Health Care Information Exchange And Interoperability, Health Affairs Web Exclusive, January 19, 2005 [Abstract] [PDF] [HTML Version] [Reprints & Permissions]

Comments:Submit a response to this article

Comments published:

Cautious Interpretation of Findings

Michael R. Mittleman   ( 7 February 2005 )

Will Interoperable HIT Lead to a Net Gain or to a Net Loss for Physicians?

Peter Basch, M.D.   ( 23 February 2005 )

The Value of Health Care Information Exchange and Interoperability: The Authors Respond

Jan Walker, Eric Pan, Douglas Johnston, Julia Adler-Milstein, David W. Bates, and Blackford Middleton   ( 16 March 2005 )

A Cancellation Effect?

Jaleann M. Matos-McClurg   ( 6 April 2007 )

Cautious Interpretation of Findings 7 February 2005
Michael R. Mittleman, Deputy Chief Information Officer State of New York Send comment to journal: Re: Cautious Interpretation of Findings mrmittleman{at}optonline.net Michael R. Mittleman	The recent article on health care information exchange and interoperability (HIEI) presents a business case largely built on technology and cost arguments. The authors conclude that there is a potential for realizing a net savings of $77.8 billion per year after a standardized HIEI is fully deployed. Lengthy experience dealing with health care information technology issues and large projects leads to questions about the process and model used in the article to derive the initiative’s cost and risks. Failing to assess factors such as risks; privacy; legal environment; organizational territorialism; competing agendas; management capability; and executive sponsorship, along with technology and cost considerations, is a recipe for certain failure. As students of enterprise architecture learn early on, the technology part of multi-organizational projects is comparatively easy, collaboration and governance are hard. Large technology projects are viewed with deep suspicion by administrators, as should be the case: they frequently fail to deliver on their promise. The view expressed in my eLetter is my personal opinion; it does not necessarily represent the position of my employer.
Will Interoperable HIT Lead to a Net Gain or to a Net Loss for Physicians? 23 February 2005
Peter Basch, M.D., Medical Director, eHealth MedStar Health Send comment to journal: Re: Will Interoperable HIT Lead to a Net Gain or to a Net Loss for Physicians? peter.basch{at}medstar.net Peter Basch, M.D.	The Center for Information Technology Leader’s (CiTL’s) recent report on the value of health information exchange and interoperability posits that the highest level of health information technology (HIT) interoperability will result in a $78 billion global net benefit/year, with over 40% of that accruing to providers ($33.5 billion or approximately $75,000 net gain per doctor/year). For physicians, this financial return is postulated to come from administrative savings – less “paper work” for staff, resulting in the ability of even a small practice to downsize. While the CiTL report is based on two uncontestable truths (that HIT has great potential for improving quality, safety, and efficiency, and that interoperability increases the value of HIT); its quantitative conclusions as to return on investment, at least for providers, is contrary to most other published reports as well as to prevailing wisdom. For example, the multi-stakeholder Connecting for Health Workgroup on Financial, Legal, and Organizational Sustainability recently concluded that adoption of an interoperable electronic health record (EHR) by physicians in a small practice would result in an approximate $20,000 annual loss/doctor/year.[1] In fact, it is this widespread consensus regarding the misalignment of costs and benefits of HIT for providers that is driving the ongoing work in both the public and private sectors on HIT incentives programs as well as on reimbursement reform. If the CiTL were correct, and every practicing physician could expect an approximate $75,000 annual net benefit from adopting an interoperable EHR (over 100% annual ROI!), not only would these incentive and reimbursement changes be unnecessary; physicians could even stop worrying about the upcoming scheduled annual Medicare cuts and the flawed and unsustainable “Sustainable Growth Rate” formula. Unfortunately for physicians, at least for those in small practices (about one-half of those in ambulatory practice today); there is good reason to view the CiTL conclusions with extreme skepticism. While this skepticism can only be generally fueled by the brief version of the CiTL report contained in Health Affairs, the entire report has now been published, and all of its underlying assumptions and formulas are now available for close scrutiny.[2] I believe that the CiTL authors arrived at their unprecedented (and at least for physicians in small practice) and erroneous conclusions based on five mistaken assumptions or flaws in logic. First, the baseline economic assumptions of the administrative costs associated with certain practice activities may be accurate for large practices but are vastly inflated and even unbelievable in the small-practice setting. For example, their use of $10 as the physician’s administrative cost for ordering one lab (they presume an equivalent cost on the lab end of the exchange) would translate to almost 45 minutes of staff time. This error is compounded when they calculated that a second and third lab would double or triple that time and cost. As inefficient as paper lab requisitions are, once a form is completed (it takes 1-2 minutes for one lab order), further orders on the same patient are accomplished by checking a checkbox (taking an additional few seconds at most). Thus, a typical lab requisition for a liver test and a cholesterol panel (which may take 2 minutes to do on paper) is budgeted by the CiTL as resulting in about 90 minutes of staff effort. Second, activity-based costing is used without acknowledging that for most small practices, staff typically multitask, making the sum of multiple activity-based costing invalid. For example, in most small practices, the same person filing lab reports in paper charts (budgeted at $9.25/report by the CiTL) is also typically simultaneously answering the phone and making appointments. This misapplication of activity-based costing resulted in the CiTL coming up with a total cost for practice overhead for just six of about twenty administrative activities (management of labs and imaging studies, inter-provider communications, prescriptions, and public health and payer communication – just a fraction of what staff do all day) that is greater than the average total practice overhead. Third, realizing administrative savings (such that they are more than a paper figure) for small practices requires downsizing staff. To actually achieve the savings for providers claimed in the CiTL report would require that a practice reduce overhead by >100%, and reduce staff to <0. Fourth, while they acknowledge that the standard of care in most small-group practices is to ask the patient, or to make educated guesses, when a particular piece of information is missing from the record (such as the results of a recent ER visit), they then assign a dollar value to this activity (using activity-based costing), as if it were done routinely. Applying their formula of 95% reduction in costs to the same activity done using level 4 interoperable HIT (instead of paper) created what they determined to be an annual savings of over $13B. While one could argue whether this level of information is truly always needed to improve care (as opposed to sometimes relying on the patient for a brief summarization of an encounter), a more accurate way to attribute these new costs and savings (instead of mislabeling them as provider savings) would be to calculate their presumed benefit for quality, safety, and the reduction of redundancy, and then to determine the cost of supplying the new solution using interoperable HIT instead of using paper. This cost/benefit analysis should result in a powerful economic argument for payers to build and incent the use of such a system. Finally, they reuse an implied benefit to providers that they developed in an earlier report that attributes 11.6% of the financial benefit of the reduction in redundant or unneeded testing or treatments back to providers.[3] This figure was based on a 1999-2000 AMA-based estimate of the percentage of income derived from capitation (a figure that overestimates capitation in 2005). However, the CiTL further incorrectly uses this figure as equivalent to at-risk capitation – which is the only capitation that would return financial benefit directly to doctors on the basis of not performing unneeded services. The capitation agreements that apply to most small practices is not at-risk, thus making the potential return to providers from avoiding unnecessary or redundant services as closer to zero dollars. Recalculating the net gain to providers in small practices based upon the five suggested corrections above results not only in no net gain, but in an approximate $4 billion annual net loss, a figure that is closer to the Connecting for Health estimate. Whether interoperable HIT results in a net gain or a net loss to providers is of more than academic interest; as this estimate will be used by payers as a rationale to create HIT incentive programs (or not) and could be used as an excuse for policymakers, assuming huge financial returns from the integration of HIT into practice, not to address even more difficult issues that will confront our health care system in the upcoming years. Notes 1. Working Group on Financial, Organizational, and Legal Sustainability of Health Information Exchange. Financial, Legal and Organizational Approaches to Achieving Electronic Connectivity in Health Care. Connecting for Health-Markle Foundation. October 2004. Downloaded free from: http://www.markle.org/downloadable_assets/flo_sustain_healtcare_rpt.pdf on February 20, 2005. 2. Center for Information Technology Leadership. The Value of Healthcare Information Exchange and Interoperability. Partners HealthCare System. 2005 3. Center for Information Technology Leadership. The Value of Computerized Provider Order Entry in Ambulatory Settings. Partners HealthCare System. 2003.
The Value of Health Care Information Exchange and Interoperability: The Authors Respond 16 March 2005
Jan Walker, Executive Director, Center for IT Leadership Partners HealthCare System, Eric Pan, Douglas Johnston, Julia Adler-Milstein, David W. Bates, and Blackford Middleton Send comment to journal: Re: The Value of Health Care Information Exchange and Interoperability: The Authors Respond jwalker3{at}partners.org Jan Walker, et al.	We appreciate the opportunity to respond to the commentaries and letters written by colleagues with diverse experiences and perspectives. We agree with Brailer that we have a lot to learn from other industries.[1] Today, most banks offer their customers easy local access to ATMs and the opportunity to manage their affairs from anywhere in the country, and from many parts of the world. This is all available at a broadly accepted level of privacy and security. Banking offers important parallels to health care, as we link large institutions while offering individual customers -- patients and clinicians -- access to their unique data. Indeed, as we work to reach the broad agreements necessary to move interoperability forward, we should invite banking leaders to advise us on how to proceed. Brailer points to real-time video, home monitoring, and other care-transforming technologies that will proliferate in an interoperable health care system. His wide-ranging thinking gives further cause for optimism about the value of interoperability. Above all, patients will be the biggest winners in an interoperable world, as they are finally empowered with the information and communication they need and are released from many geographic constraints on their care. Baker cautions readers against being overly optimistic about the value of interoperability, given the nation’s spotty experience to date with EMRs and related tools.[2] He points out that the economic benefit of these systems has been less than impressive. However, many current "EMR" implementations do not satisfy the Institute of Medicine’s (IOM's) criteria for minimal functionality, and expectations of high returns may, therefore, be misplaced.[3] In contrast, the systems for both our Level 3 and Level 4 calculations include advanced technology that brings together EMRs, Computerized Provider Order Entry (CPOE), and Clinical Decision Support Systems (CDSS) that meet IOM criteria. Baker questions our assumptions that 14% of outpatient tests are redundant and avoidable and that 95% of those would be avoided with Level 4 HIEI. The rate of avoidable tests is based on two studies, and further research may indeed find different results.[4] But we agree with our expert panel that eliminating 95% of avoidable tests is an attainable goal. Tests are avoided through two mechanisms in Level 3 and Level 4 systems. First, clinicians know about tests done in other facilities and can avoid rework. Second, CDSS logic is applied to all orders. It would be unreasonable to expect high savings if providers were not also using CPOE with advanced CDSS applied to virtually all relevant information in Level 4, both from their own systems and from external systems. This is the most complete way to eliminate redundant utilization. Level 3 also assumes advanced CDSS, but its impact is attenuated by a lack of fully structured and coded messaging. Baker also suggests that we may have imputed too many administrative savings in laboratories and radiology centers. Our experts’ estimate of costs was consistent with interviews with laboratory managers who described the paper shuffling and phone time required. But, our information admittedly reflects experience in large hospitals and integrated delivery networks. Freestanding facilities may be more efficient, but we did not have access to specific cost data. Baker and Basch are skeptical about some of our provider projections.[5] However, both make an assumption that we find unrealistic: that paperwork and administrative costs are borne primarily by the lowest-paid office staff. Today's disjointed medical records reduce the productivity of the highest-paid clinicians, not just the lowest-paid office staff, and comprehensive, seamless integration of medical knowledge and patient information cannot be achieved by low-level office workers alone. At first glance, for example, one might say that clinician time is limited to the seconds required to check a box to order a test. But the full picture is far more complex. Without Level 4 interoperability, integrating potentially redundant information from multiple sources requires human clinical knowledge and expertise. Clinicians are the ones who have to juggle multiple pieces of paper and mentally integrate illegible handwritten notes, blurry faxed reports, and multiple past records. While clinicians employ and delegate information-seeking tasks to office staffs, the interruption in workflow and the suboptimal productivity are not lost by the $12.18 per hour staff, but by the physicians who may cost several hundred dollars per hour. Our expert panelists attempted to recognize the cognitive tasks as well as the obvious paperwork in their estimates of the costs of Level 1 interoperability. Basch makes the excellent and unfortunate point that providers today often do not -- and, we would add, cannot afford to – take time to integrate information from disparate sources. However, we disagree with his assertion that the value of complete medical record information is not established. Reports such as To Err is Human and Crossing the Quality Chasm have clearly established the danger of incomplete patient information, with patients paying a steep price for compromised quality of care.[6] Today, none of us would accept a bank that makes educated guesses about our current account balance based on an incomplete transactions history. How can we ask patients to accept the same from us? As the IOM firmly states in Patient Safety: Achieving a New Standard for Care, we need an information infrastructure that can “provide immediate access to COMPLETE (our emphasis) patient information.”[7] Allocating costs and benefits is tricky, as Basch suggests, and we agree it would be valuable to identify the incremental investment required. But, without an accurate characterization of today’s workflows (How much time do providers spend hunting for information from other organizations? How often do they not even try? How often do they make the wrong educated guess? How much harm and death are the direct result?), it is impossible to know how our current level of effort compares to what would be required to insure access to complete information in a paper-based system. Instead of guessing about the size of that gap, we adopt the IOM ideal and assume that the work necessary to share and integrate information is indeed accomplished. Our analysis of Levels 2, 3, and 4 interoperability quantifies benefits compared to that baseline, and it makes a clear case for Level 4 as the best way to deliver complete information. Basch compares these results about the most effective approach to interoperability to the Connecting for Health analysis of financial issues related to EHR adoption in small- and medium-sized practices.[8] We suggest that this comparison is off the mark. James describes the workflow benefits of EMRs with several vivid examples from clinical practice, benefits that would grow as communications with external care partners become electronic. His Microsoft Office analogy is an excellent one, and it offers a fine example of how difficult it is to place a value on interoperability.[9] No matter what we think of near monopolies, most of us use MS Word, PowerPoint, and other Office applications, and we share files with colleagues around the world with perfect communication of unique formats and content. Though none of us could quantify exactly how much time (and error) this saves, we would all agree that it makes us more efficient. At the same time, we share his concerns about the profitability of transition states in individual practices. And we would underscore Mittleman’s cautions about the importance of risk assessment and planning in multi-institutional projects.[10] Like many other complex aspects of interoperability, these sensitive issues must be considered in a policy context. We agree with James’s call for certification standards, and with Baker’s advice to first tackle one set of transactions -- such as those between providers and laboratories -- where widely accepted standards already exist, in order to learn progressively. James’s logical steps in EMR implementation may have correlates in the development of interoperability. But, as Brailer notes, the sequence of actions is not yet clear. We support his call for an open and vigorous debate to establish the framework. Notes 1. D.J. Brailer, “Interoperability: The Key to the Future Health Care System,” Health Affairs, January 19, 2005, 10.1377/hlthaff.w5.19. 2. L.C. Baker, “Benefits of Interoperability: A Closer Look at the Estimates,” Health Affairs, January 19, 2005, 10.1377/hlthaff.w5.22. 3. Committee on Data Standards for Patient Safety, Key Capabilities of an Electronic Health Record System, Institute of Medicine (October 2003). 4. D. Bates et al., “What Proportion of Common Diagnostic Tests Appear Redundant?” American Journal of Medicine 104, no.4 (1998): 361-368; D. Brailer et al., “Moving Toward Electronic Health Information Exchange: Interim Report on the Santa Barbara County Data Exchange.” (Oakland, CA: California HealthCare Foundation, July 2003). 5. P. Basch, “Will Interoperable HIT Lead to a Net Gain or to a Net Loss for Physicians?” Health Affairs Electronic Letter, February 23, 2005. 6. L. T. Kohn, J. M. Corrigan, M. S. Donaldson, Editors, Committee on Quality of Health Care in America, To Err Is Human: Building a Safer Health System Institute of Medicine, 2000; Committee on Quality of Health Care in America, Crossing the Quality Chasm: A New Health System for the 21st Century, Institute of Medicine, 2001. 7. P. Aspden, J. M. Corrigan, J. Wolcott, S. M. Erickson, Editors, Committee on Data Standards for Patient Safety, Patient Safety: Achieving a New Standard for Care, Institute of Medicine, 2004. 8. Working Group on Financial, Organizational, and Legal Sustainability of Health Information Exchange. Financial, Legal and Organizational Approaches to Achieving Electronic Connectivity in Health Care. Connecting for Health-Markle Foundation. October 2004. 9. B. James, “E-Health: Steps on the Road to Interoperability,” Health Affairs, January 19, 2005, 10.1377/hlthaff.w5.26. 10. M. R. Mittleman, “Cautious Interpretation of Findings,” Health Affairs Electronic Letter, February 7, 2005.
A Cancellation Effect? 6 April 2007
Jaleann M. Matos-McClurg, Utilization Analyst DAMA, AHIMA, AAG Send comment to journal: Re: A Cancellation Effect? jaleann_mcclurg{at}yahoo.com Jaleann M. Matos-McClurg	I believe that the impact depicted in this article from health care information exchange and interoperability (HIEI) between providers and payers is not leveraged for the fact of higher claim-processing analytics and development of standards at the payer level. I ask the authors the following question for future research: Would this yield a cancellation effect (less net dollar value) impact upon HIEI implementation? Note: While the payer side will boost their claim-processing analytics power pre- and postpayment during adjudication process, then providers will strive to document in EMRs the clinical guidelines required to justify billed codes for services rendered. I suggest that the savings equation given implementation will be the net effect of both outcomes. I am very interested to pursue further research for my applied project as part of degree requirement in health information management in this topic, as it raises very interesting and multicausal inputs yet to be considered in a benefit analysis, given standard developments to drive reimbursement more efficiently. For a future modeling effort, I recommend considering the dollar value involved prior to implementation, therefore the infrastructure/human resources investment vs. the costs saved. I envision this task as a very challenging quantitative excercise but not impossible.