Presented by:
Nelly Leon-Chisen, RHIA
Director, Coding and Classification
American Hospital Association
Chicago, IL 60606
INTRODUCTION
My name is Nelly Leon-Chisen, director of coding and classification for the American Hospital Association (AHA). Many of you may know my role as the director of the Central Office on ICD-9-CM, editor of Coding Clinic, and one of the Cooperating Parties. I’m here today representing the AHA. On behalf of our nearly 5,000 member hospitals, health systems, networks, and other providers of care, I would like to thank you for the opportunity to provide comments on the RAND study, The Costs and Benefits of Moving to the ICD-10- Code Sets. I would also like to thank the AHA’s Member Advisory Panel on ICD-10 Implementation. This is a group made up of hospital health information management professionals, corporate compliance officers, clinical and financial data leaders from state hospital associations, and many others who would be impacted by a move to ICD-10. We sought this panel’s advice and suggestions in preparing our statements today.
We commend Dr. Martin Libicki, principal investigator of the study, and RAND on the thorough and balanced study of a subject not easily quantifiable. For America’s hospitals, migrating to a newer version of clinical code sets, capable of accurately reporting better information on diseases and procedures, is essential. Such a decision should not be based on financial considerations alone, but should take into account the shortcomings of ICD-9-CM and the consequences of “limping along” on a broken clinical code set incapable of recognizing advances in medical knowledge and technology. We agree with the study’s conclusion that there will be benefits associated with migration to ICD-10-CM and ICD-10-PCS, and that these benefits exceed the initial break-in costs within a few years of implementation.
COSTS
We agree with the report’s estimate that the costs of implementing ICD-10-CM and ICD-10-PCS can be classified into three categories: training, productivity losses and system changes. We believe, however, that, based on our own studies, the study overstates the costs of training and productivity losses for hospital coders.
It is important to note that hospital coders as a rule already receive training on a regular basis. They do so to sharpen their coding skills, keep up with code changes, meet requirements of corporate compliance programs, as well as for continuing education to maintain their professional credentials. Such training consists of between 10 to 30 hours per year on coding alone. In fact, many large health systems have staff at the corporate level to provide system-wide coding training in order to ensure the accuracy and integrity of their coding process. Following this model, it may not be necessary to incur a great deal of money to train America’s coders. Train-the-trainer programs could be effectively utilized to train coding leaders who would then disseminate the information to their colleagues at the hospital. As the RAND report states, most of today’s coders have been expecting clinical versions of ICD-10 to be implemented. To prepare, many hospitals have initiated training to bolster their coders’ knowledge of anatomy and physiology. The cost of ICD-10-CM and ICD-10-PCS training would, in many ways, replace costs already being incurred by hospitals to keep up with ICD-9-CM. Based on the advice from the AHA’s Member Advisory Panel on ICD-10 Implementation, the costs of training hospital coders, as reported in the study, could therefore be cut by at least half to three-fourths.
With regards to productivity losses, it is true that with any coding system change there will be a period of adjustment until coders become proficient with the new systems’ rules and guidelines. However, we believe that the RAND study’s estimate of productivity losses is an overestimate. Based on the recent AHA and American Health Information Management Association (AHIMA) field-testing of ICD-10-CM, in more than half the records coded, coders saw no increase in the time it took to code with ICD-10-CM compared to ICD-9-CM. This despite the fact that the testing was performed after only two hours of non-interactive training, using a cumbersome paper-based index file format, and with coders having to review the ICD-10-CM guidelines (more than likely they would not have needed to review the guidelines for ICD-9-CM). With the majority of hospitals using encoders, and with additional training and practice, the coding productivity under ICD-10-CM has the potential to be greater than under ICD-9-CM. As the data quality analyst at a large teaching hospital so aptly put it, “Coders waste a lot of time today trying to find codes for new procedures.”
With regards to the cost to change systems, we agree with the study’s assessment that most providers buy their systems from vendors and most of the provider-vendor relationships have a software maintenance component that would include regulatory updates to the systems.
BENEFITS
We agree with the reports’ identified classes of benefits:
As reported in the Rand study, Australia has cited an impressive list of benefits that resulted from migration to more current and specific code sets. We also recognize the difficulty in quantifying improvements in health or cost savings, but intuitively we believe that better information to measure inputs and outputs makes for better decision-making. Clinical codes are an essential part of such activities. Maintaining a system that provides incomplete or vague information due to outdated codes hurts the entire health care industry.
An additional benefit not specifically identified in the report is the ability to improve community health based on coding consistency between mortality data and morbidity data. Cause of death data is currently reported using ICD-10. In many states, such as Illinois and Utah, mortality data is cross-analyzed with hospital data for community health to develop intervention strategies. Resources are wasted converting mortality data back to ICD-9-CM to correlate to hospital data. From a public health perspective, it is also more difficult to compare international health data when the United States is using a different coding system than the rest of the world.
The study offered several suggestions to facilitate the transition to ICD-10-CM and ICD-10-PCS. I’d like to address these suggestions one by one.
Transition Date
The study recommended selecting a certain date by which everyone must make the transition. Fundamentally, we agree. However, we disagree that it be date driven – that is, submit ICD-9-CM on one day and ICD-10-CM and ICD-10-PCS on the next. Hospitals, other providers, and payers currently base code version changes on the discharge date for inpatients, and service date for outpatient services. We support a similar transition date as we currently have for ICD-9-CM—namely October 1—regardless of the start of a hospital’s fiscal year. For example, a patient discharged on September 28 is assigned the codes effective on that date (and valid through September 30), whether the claim is submitted on September 30 or October 28. By contrast, a patient discharged on October 2, is coded using the code set version effective October 1 of that year. AHA supports the transition to ICD-10-CM for diagnosis coding and ICD-10-PCS for hospital inpatient reporting on a date certain, (preferably October 1), based on the discharge date for inpatients and the date of service for outpatients.
Transition Timeline
The timeline for implementing the new system should be carefully orchestrated to minimize the administrative burden to providers. The importance of this approach cannot be overestimated. We support an implementation date of two years from the publication of a final rule, for several reasons. First, it would allow sufficient time for organizations to incorporate these changes into their budgetary planning cycles. Second, it would allow system developers to make appropriate changes and test those changes before going live. As we are seeing today with the HIPAA transaction standards, testing is key to ensuring that both payers and providers can submit and accept claims using the new codes. AHA strongly agrees that migration to ICD-10-CM for diagnosis coding should be carried out in tandem with the migration to the ICD-10-PCS procedure codes, two years after publication of the final rule. The AHA would welcome the opportunity to work with NCVHS and the Centers for Medicare & Medicaid Services (CMS) to develop a detailed implementation timeline.
Crosswalks
We agree with the study’s suggestion that a reliable and readily understandable crosswalk be prepared and promulgated. AHA believes backward and forward electronic crosswalks between ICD-10-CM and ICD-9-CM diagnosis codes, and between ICD-10-PCS and ICD-9-CM procedure codes, should be made available by NCHS and CMS, free of charge, or at a reasonable cost. Such crosswalks would be useful not only in easing the transition into the new codes, but also in converting existing databases to compare clinical data pre- and post- ICD-10-CM and ICD-10-PCS implementation.
Comparison
Finally, we would like clarification on the RAND suggestion to have a major provider code diagnoses and procedures in both ICD-9-CM and ICD-10-CM and ICD-10-PCS, to determine which codes are interpreted similarly and what a crosswalk is in practice, as well as in theory. It is not clear what the objective of this effort would be. What is clear is that this would be a massive undertaking for any organization. An evaluation of the purpose of such an undertaking would be helpful.
Conclusion
The AHA strongly supports a move to ICD-10-CM for diagnosis coding and ICD-10-PCS for hospital inpatient procedure coding within two years of the publication of a final rule.
Again, thank you for the opportunity to provide our comments on this important transition of code set upgrades. I will be happy to answer any questions you may have.