The Forgotten Argument For ICD-10

The following is a guest post by Eric Hodge, Service Line Executive for Revenue Cycle and ICD-10 at Encore Health Resources.

Yesterday evening, the United States Senate joined the House of Representatives in Washington D.C., voting to delay ICD-10 adoption until October 2015.  That’s no surprise.  Truth be told, the vast majority of discussion related to ICD-10 has been all about how difficult it will make our lives.

Providers are asking, “Why is HHS forcing this down our throats when it obviously won’t help me do my job any better?” The AMA is throwing out headlines like, “ICD-10 Compliance Costs Are Triple What Was Expected,” while reminding us that they warned us all along. Now, many commentators are declaring the whole shmeer a disaster before it even goes live.

This attitude has skewed the thinking on ICD-10. Few providers are asking how they will benefit from the new information; the vast majority are simply asking how they will survive getting ready to meet the requirement. And that’s too bad, because what we as providers, as an industry, and even as an economy will find that ICD-10 is a key step toward gigantic improvement in how healthcare works in the U.S.

I am not going to argue that the transition is coming without cost or discomfort. But I am saying that this is how large-scale improvement of a system (a broken system, don’t forget) works, and that the benefits are clear and significant, at least for those who get past our first reaction (“Change frightens me!”) and take the time to understand what kind of system this whole healthcare reform effort is trying to build.

Benefits that I have seen with my own two eyes include:

  • Dramatic improvement in the assignment of costs to procedures performed. Most industry observers agree that we ought to move toward rewarding activities that keep a population healthy instead of getting paid for how many times we can treat a patient. Most would also agree that identifying the costs associated with certain disease states or treatments is the key to figuring out economical ways to promote healthy populations. ICD-10 will allow us to develop meaningful estimates about what a disease state or a procedure costs us, while ICD-9 is limited in what it can do in this regard.

    For example, I was working with a well-regarded regional hospital in the Mid-Atlantic on an effort to improve their charge capture. They knew they were losing money in their obstetrics operating room, but they were having a hard time figuring out exactly what was going on. Using ICD-9 information, all we could tell was that there were wildly variable times that a patient would spend in the OR for a cesarean procedure, but we could not gather any more detail. ICD-9 diagnosis codes do not have very specific information about the severity of the condition or comorbidities. Fortunately, this hospital was dual-coding at the time, and we were able to take advantage of the severity information included in the ICD-10 codes to identify the fact that they had a relatively high percentage of moderate and severe diagnoses — complications that were likely to lead to longer OR times and higher resource consumption (costs) to the hospital.

    This information allowed them to build a business case for establish pricing tiers for their OB OR services and gave them the information they needed to turn obstetrics surgery into both a profitable activity center and one that could revise treatment protocols for high-resource-consumption patients (costs).

    Could this have been done without ICD-10 data? Probably. But it would have taken many hours of chart review and qualitative analysis instead of the several dozen key strokes of a database query.

  • Identify opportunities to avoid cost and improve lives. The additional information inherent in an ICD-10 diagnosis code includes severity and specific comorbidity, as illustrated in the OB OR example, but it can also include information about demographics and some of the underlying reasons for the diagnosis. All of this information can easily be combined to make decisions that will save lives while cutting costs for a provider.

    I was working with a multi-facility provider in New England on vendor selection for revenue cycle technology, and I visited the cancer clinic. In talking with the nurses there about the kind of data that would help them care for their patients, they let me know that they would like to be able to flag patients with a high chance of readmission. One of the nurses told me that after 22 years of experience, she knew that a patient who was over 80 with moderate or severe lung cancer and a history of mental illness was going to be readmitted within three weeks. “And wouldn’t it be nice,” she said, “if my new system could flag those patients when they came in for an appointment?”

    Well, only ICD-10 codes include severity of illness, age, and the latitude to include reasons for a diagnosis. In this case, included in the diagnosis code was the fact that the patient was non-compliant in taking his/her medication. We were able to model this scenario for ICD-10 and identify these patients with a simple data query – in minutes. This allowed the clinic to first confirm the nurse’s intuition about those high-risk patients, and second to identify those patients who could use a case manager’s involvement to ensure that they are compliant with their regimen, saving the costly readmission and improving the quality of the patient’s remaining life.

    Again, this sort of effort is possible with ICD-9, but it would take chart reviews, extensive manual analysis, and aggregation of data from several sources to model this type of patient for predictive purposes. This organization did not have the extra resources or the budget to undertake such an effort.

  • Share higher-quality data with other providers and partners. When I meet with providers who are trying to figure out whether to start or join an Accountable Care Organization (ACO), the first question is generally, “What is this big pile of aggregated data going to do for us?” Actually, that’s the second question after, “What incentive dollars am I going to get for doing this whole ACO thing?” But it should be the first question.

    As the data sets grow larger, the ability to parse information into meaningful subsets will become more important. ICD-10 increases the amount of specific information in every diagnosis code and actually makes these large, aggregated pools of data from many providers useful. For example, ICD-9 has a code for laceration of an artery. ICD-10 lets you know if that artery was in someone’s finger or in their heart. If I want to be able to pull meaningful information out of my ACO data sets, I need to have the information that is included in ICD-10.

    I have helped organizations use aggregated diagnosis data like this to decide whether pursuing certain services in certain markets will pay off for them. We helped a provider in Washington State decide to extend its diabetes education services into rural Oregon and Idaho by demonstrating that there were enough diagnosed patients to support that business. This type of analysis becomes much faster and easier with ICD-10 data.

There are dozens of other tangible benefits to ICD-10 analytics, but this is a blog entry, not a thesis. Briefly, some of the biggies:

  • Being able to aggregate our diagnosis and procedure information with the rest of the industrialized world, which has already demonstrated that the benefits of ICD-10 will significantly advance healthcare service in the US. There are lots of sick people outside America, too, so being able to combine our coding data with theirs for analysis would be most helpful.  For example, the US has benefited from the increased data collected about the Avian Flu and how to best treat the disease based on ICD-10-collected information.
  • Reimbursements will better align with activity and cost. Payers will reimburse severe and complex cases better and simple cases at lower rates – because now they will be able to identify them as simple or complex from the codes. Those providers whose costs are higher will get paid more. Those whose resource costs are lower based on actual services rendered will get paid less. This principle is how the rest of the free market works; it should also work well in healthcare.
  • Outcome analytics will become more accurate and more efficient. I can quickly determine what happened to my severe CHF cases without having to go back through every single one of their charts or pull in data from multiple sources to figure out which CHF patients were only moderate or mild.
  • Population-based projections will become much more possible. If you want to look at the incidence of advanced diabetes in the aged population in southeast Missouri so you know how to negotiate your value-based reimbursement contracts, you can use ICD-10 data or you can go do a lot of legwork.

The point here is that ICD-10 makes coding information detailed enough so that American providers and payers can make healthcare work in ways that it doesn’t work now: like a free market, with costing and pricing that accurately reflects the effort and the expense. Like a continuously improving system where better courses of treatment are developed for more specific populations. And like a system where we try to prevent high cost and lousy outcomes before they happen.

Looks like we’re going to have to wait until 2015 before we see many of these benefits.