Commentary: The ICD-10 Saga—Lost Years and Hard Lessons

The August issue of the Journal of AHIMA features an article by Sue Bowman that chronicles the dire predictions of negative consequences that would ensue if the antiquated 30-year-old ICD-9-CM coding system was replaced with the modern and up-to-date ICD-10 coding system. This routine administrative update became such a highly charged political issue that Congressional hearings were held with demands for Congressional action to stop ICD-10. In the end, it took 20 years and three delays before ICD-10 was implemented.

As Bowman observes in “A Look Back on the ICD-10 Transition: Crisis Averted or Imaginary?”, the transition to ICD-10 has largely turned out to be a Y2K-like non-event with none of the dire predictions actually occurring. She appropriately asks whether ICD-10 implementation was a crisis averted, a crisis that was imagined, or a crisis that was intentionally fabricated. Any major change to the basic healthcare data infrastructure will have stakeholders motivated by their own special interests.

If the transition to ICD-10 is any indication of the course of future updates to our basic healthcare data infrastructure, it is important to understand the tactics and motivation behind the policy debate and the campaign to stop ICD-10 implementation.

Proponents tended to argue the merits of ICD-10 with an academic approach that focused on healthcare policy. They made their case in scholarly journals that reached like-minded readers and not necessarily mainstream audiences. Proponents cited objective evaluations such as an exhaustive 2004 RAND study analyzing the cost benefit of ICD-10 implementation, but they initially failed to communicate and promote the value of ICD-10 in a way that ensured all healthcare industry participants understood its importance.

Opponents of ICD-10 took a different approach, one that was more like a political campaign with unrelenting sound bites of misinformation that sought to alarm and create opposition among key policy and Congressional leaders. As documented by Bowman, opponents said there were “too many ICD-10 codes,” which would be impossible to learn and use. The new codes would create havoc with the claims process and cause increased claims denials, cash flow disruptions and revenue shortfalls. The expense of implementing ICD-10 would put small physician practices out of business, they warned, and for those practices that survived ICD-10 would be so burdensome that patient care would suffer.

Hospitals also would be negatively impacted, opponents claimed. They predicted a huge drop in coder productivity, which would lead to slowdowns in payment that would cause major business disruption. Finally, opponents said the Centers for Medicare and Medicaid Services (CMS) wouldn’t be ready by the implementation date, nor would payers. “The healthcare industry needs more time,” was their message to Congress, an argument repeated time and again over two decades.

While ICD-10 proponents focused on the policy issues, opponents issued grim warnings, designed to earn media attention in today’s fast-paced news cycle. Unsubstantiated by independent research and analysis, their warnings were reported across US media with virtually no investigation or fact checking. With few exceptions, even healthcare industry journalists avoided any in-depth evaluation of the case against ICD-10. Instead, publications featured stories about “ridiculous” ICD-10 codes, which presumably attracted readership and drove increased traffic to media websites. Biased opinions with no factual foundation were portrayed as undisputed truths.

The existence of ICD-10 codes for an alligator bite or being pecked by a turkey, for example, were continually held up as laughable and an example of the government demanding overwhelming detail and creating a huge burden for providers. This ignored the fact that similar codes have always existed without causing a burden, that such codes are not generally required to be reported except for infrequent situations such as a worker compensation claim, and most important, that the codes could become essential from a public health perspective given the risk of animal-to-human disease transmission (i.e., swine flu, avian flu, etc.)

At the outset, proponents of ICD-10 mistakenly believed that all healthcare industry stakeholders would view ICD-10 as a necessary administrative update to the US coding system, not dissimilar to a routine software update. Most other industrialized countries had implemented ICD-10 a decade earlier without issues, and it was assumed the US would follow a similar path. The intensity of opposition caught proponents off-guard and they were unprepared with an effective response. ICD-10 opponents had set the tone and it was difficult to undo the misinformation.

It was not until the third ICD-10 delay that the Coalition for ICD-10 (made up of AHIMA, AHA, HFMA, AHIP, National Blue Cross Blue Shield, and other organizations) began addressing negative messaging with more confrontational rebuttals along with aggressive outreach to the press. The result was more balanced media reporting.

Still missing from the debate, however, was any examination of motive. Among the industry groups and stakeholders that opposed ICD-10 were strong advocates of SNOMED. Their published opposition to ICD-10 was never evaluated in the context of their support for SNOMED.

Unlike other countries, the US uses different procedure coding systems for inpatient and outpatient care. Prior to ICD-10 implementation, existing procedure coding systems were loosely organized lists of unstructured codes with no terminology definitions, while the ICD-10 procedure codes are a modern, multi-axial seven-character system with clearly defined terminology. The contrast is so stark that the healthcare industry may eventually question the necessity of multiple procedure coding systems, potentially jeopardizing the continued use of separate systems for inpatient and outpatient care.

Was ICD-10 viewed as a threat to the continued use of two procedure coding systems, and was this a factor in galvanizing opposition to ICD-10? Again, motive was never explored in any media reporting and ICD-10 proponents were unwilling to raise the uncomfortable question of whether unspoken special interests were behind the opposition to ICD-10.

In the years ahead, the healthcare industry will continue to move forward with standardization of our basic healthcare data infrastructure, making sharing, interpreting, and interoperability of healthcare data more effective. Based on the ICD-10 experience, future infrastructure updates and policy changes may encounter similar aggressive opposition. The lessons from ICD-10 are clear. Proponents must clearly define and communicate the purpose and value of the change for all healthcare stakeholders. They must prevent opponents from owning the story and message. An aggressive public relations effort should be launched at the outset, with education and information that reaches healthcare audiences in a way that’s easily accessed and understood. Any misinformation should be quickly refuted. The public policy debate also requires a full understanding of the motives of all stakeholders and any potential conflicts of interest.

The transition to ICD-10 has turned out to be a relatively smooth process. None of the dire predictions have come true and the healthcare industry is ready to move on. Before we relegate the implementation of ICD-10 to history, however, policy makers should evaluate the credibility of ICD-10 opponents, especially as it relates to the debate on future healthcare data infrastructure changes.

Put simply, how could ICD-10 opponents have been so wrong? Even more important, how were the opponents of ICD-10 able to convince the media and members of Congress to accept their unfounded claims without anyone successfully discrediting them? There is much to learn from the ICD-10 experience. Someday, we may view the long and arduous road to ICD-10 as an anomaly that will never be repeated. If future policy debates become as contentious, however, proponents of change need to understand the lessons of the 20-year ICD-10 saga.

Richard F. Averill is principal at The Hesperium Group, and recently retired as senior vice president of clinical and economic research at 3M Health Information Systems.


  1. AHIMA, Dr. Averill, and the rest of the ICD-10 Coalition continues to ignore the real issues as to why ICD-10-CM/PCS is so problematic and why physicians resisted it. First, and foremost, ICD-10-CM is not current with physician language, introducing concepts such as a “functional quadriplegia” (which, according the there is no clinical definition), “initial encounter” to mean “first diagnosis” or “active treatment phase” (not the first visit), “subsequent encounter” to mean “healing phase” (not the second or subsequent visit), “pulmonary insufficiency following surgery or trauma” (not respiratory insufficiency – what’s the difference between it and respiratory failure”), equating acute respiratory distress in the Index with acute respiratory distress syndrome (J80) – which is insane, and many, many others. Second is the requirement that ICD-10-CM code assignment be based only on explicit provider documentation of the codeable condition using ICD-10-CM language, not a reasonable interpretation of the patient’s clinical indictors, meaning words that physicians understand, such as “urosepsis”, “unresponsive”, “neurotoxicity”, “HFrEF” (gladly, this was resolved with the 1st Q 2016 Coding Clinic), “sepsis syndrome”, and the like require a burdensome query process that costs hospitals and practices at least a $1 billion/year to maintain (of note, 3M, for whom Dr. Averill worked, made a killing from their DRG Assurance product and selling the APR-DRG grouper). Third, CMS and the Federal Government require physicians to use SNOMED for their EHR problem lists which contains clinical terms that cannot be coded in ICD-10-CM, thus the physician must document in two languages, not just one. Finally, physicians are excluded in the control of the ICD-10-CM language, given that the CDC and CMS who have final authority over ICD-10-CM and PCS respectively make their final decisions of what codes are implemented behind closed doors (anyone know of where their minutes are?), that the Cooperating Parties (CDC, CMS, AHIMA, AHA) who write the guidelines and govern the interpretation do not include a physician group, and that these same Cooperating Parties actively don’t want a physician group to share control with them. The ICD-10-CM/PCS transition was an unfunded mandate geared toward providing administrative data that Sue Bowman has stated in articles in JAHIMA is not for clinical care (that’s SNOMED’s role), yet physicians are having to buy new software, document more than they would normally have to document using a language that’s not in their literature, and be excluded from the final decision making process that governs the code set. Who here can blame physicians for being wary of additional government regulations when our time is now being stolen to perform administrative duties? Sadly, I don’t think Ms. Bowman and Dr. Averill have learned from history, meaning that they are bound to repeat it when ICD-11-CM hits us in the next decade. In this article, Dr. Averill portrays the ICD-10 proponents as the thoughtful elite power brokers (which is arrogant, in my mind) and physicians as Trump-like self-serving manipulators. Its as if physician’s complaints about their cost in transitioning to ICD-10-CM didn’t matter, even though 3M and AHIMA made a lot of money in selling ICD-10 consulting and software. This attitude has to change. Physicians have a lot on their plate as we transition in a rapidly changing healthcare reimbursement and clinical care environment. The Cooperating Parties can do more, I believe, in bringing physicians to the table as ICD-11 approaches. The first thing they can do is invite the AMA to be a Fifth Cooperating Party so that the Guidelines and Coding Clinic advice can be vetted by a practitioner in the trenches before its publication. Next, they need to make minutes of the final code selection process and the Editorial Advisory Board of the Coding Clinic public; why keep them secret? Next, they need to eliminate the onerous documentation requirements acknowledged by the Coding Clinic in their discussion of excisional debridements in the 3Q 2015 issue and allow coders to make reasonable interpretations of physician intent without having to subject physicians to a difficult “non-leading” query process. Finally, they must publicly work together with physicians to have one language, optimally a blend between SNOMED and ICD-11, that can be used both for clinical care and for administrative purposes and that the costs of transition and implementation must be ameliorated. This work needs to start now, not 2 years before the ICD-11 transition is announced like they did with ICD-10-CM. I know that this opinion may cause heartburn in HIM leaderships who believe ICD-10 is the solution to the problems of the world today; however I don’t believe I’m the only person in the USA that believes it. Coding is supposed to be a partnership between the physician and the coder; what good is it if the physician does not share the control of his or her own language? Thank you so much.

    Post a Reply
    • Very interesting response. Thank you for providing another viewpoint. I have been involved in the implementation of ICD-10 as a long term inpatient coder. My impression so far is that while PCS is vastly superior to ICD-9 procedure coding, the ICD-CM changes are not as useful, and mainly in the area of adding laterality to fracture codes. If there truly was no practicing physician input, that could explain a lot.

      I suppose at this point they couldn’t go back to the drawing board and start again, but have already started to hope that ICD-11 might be implemented soon, and fix the issues?

      Post a Reply
    • Dr. Kennedy…Very well said. You have some very valid, common sense points in your reply. I think one thing Dr. Averill neglected to point out about the other countries moving to ICD-10-CM years ago is those other countries don’t use ICD-10-CM as a payment system and PCS was developed in the US by 3M. Those countries use ICD-10-CM for morbidity and mortality data. The US is the only country that uses ICD-10-CM/PCS as a payment system. Dr. Averill did say in the article one of the arguments against ICD-10 was “CMS wouldn’t be ready by the implementation date, nor would payers,” which as far as CMS is concerned, is partly true as evidenced by the issues with their non-existent or incorrect NCD’s or the claims that have had to be rebilled because of incorrect edits which Dr. Averill also failed to mention. He also said “opponents issued grim warnings, designed to earn media attention in today’s fast-paced news cycle” and other such harsh criticisms. Being that he is “recently retired as Senior Vice President of Clinical and Economic Research at 3M Health Information Systems,” and stood to make a tremendous amount of money from ICD-10 products and services, I don’t think he should be berating people who have a different opinion than his.

      Post a Reply
  2. Was just wondering if there are any plans to replace CPT with ICD-10-PCS? As the article pointed out, PCS is leaps and bounds better than CPT, which was better than ICD-9 procedure coding. I had always assumed that the AMA developed CPT as ICD-9 procedure coding was so inadequate, especially for billing.

    Post a Reply


  1. The ICD-10 Saga—Lost Years and Hard Lessons | Coalition for ICD-10 - […] commentary was originally published by the Journal of AHIMA on August 9, […]

Submit a Comment

Your email address will not be published. Required fields are marked *

Share This

Share This

Share this post with your friends!