Early Studies Show Promise of ICD-11

Two studies are extremely encouraging, in terms of the content coverage and feasibility of replacing ICD-10-CM with ICD-11.

In February, the World Health Organization (WHO) released the official version of ICD-11. With this, the newest edition of the International Classification of Diseases (ICD) officially came into effect, and the WHO is now encouraging all member countries to begin using it.

Anticipating this, the National Committee on Vital and Health Statistics (NCVHS) has been urging action that will enable the U.S. to make informed decisions regarding the adoption of ICD-11. The NCVHS first issued recommendations to the Secretary of Health and Human Services in November 2019, and did so again in September 2021. These recommendations include a research agenda to evaluate the use of ICD-11 in the U.S. Such research is just beginning, but early published studies are promising.

Kin Wah Fung, a data scientist at the National Library of Medicine (NLM), led a study comparing ICD-11 to both the WHO’s ICD-10 and the U.S. version, ICD-10-CM. Published in March 2020, it is the first broad-based comparison of ICD-11 to ICD-10 and ICD-10-CM, focusing on identifying the differences between them.

To compare ICD-11 with ICD-10-CM, they used the Unified Medical Language System (UMLS) lexical tool to normalize and match ICD-11 codes with ICD-10-CM codes. They also manually recoded a sample of 388 ICD-10-CM codes from six disease areas, recoding them in ICD-11 to determine whether the meaning could be fully represented with or without post-coordination. The six common conditions that they recoded included diabetes mellitus type 2, hypertension, polyhydramnios in pregnancy, tuberculosis, fracture of thumb, and skin cancer. They found that about 60 percent of the ICD-10-CM codes for these conditions could be represented fully by pre- or post-coordinated ICD-11 codes. But with the addition of just three episode-of-care extension codes, ICD-11 could represent 85 percent of the ICD-10-CM codes. In their analysis, they found that post-coordination reduced the total number of codes, while still making it possible to represent the extra detail in the U.S. clinical modification, which may, in their words, “obviate the need for” a U.S. clinical modification of ICD-11.

Fung proceeded with a larger team to carry out a more comprehensive coverage analysis to evaluate the feasibility of replacing ICD-10-CM with ICD-11. This study, published in August 2021, evaluated ICD-10-CM codes from each chapter representing the top 60 percent of the most frequently used codes in Medicare claims and hospital data. They manually recoded 943 representative ICD-10-CM codes in ICD-11 and evaluated the degree to which ICD-11 represented the ICD-10-CM codes. They found that 32 percent of ICD-10-CM codes could be represented fully by pre- or post-coordinated ICD-11 codes. But with the addition of just nine extension codes, ICD-11 could fully represent about 60 percent of the ICD-10-CM codes.

This is extremely significant, because at the time we implemented ICD-10-CM, only 24 percent of ICD-9-CM codes had an exact match in ICD-10-CM. This implies that moving from ICD-10-CM to ICD-11, without creating a clinical modification, would be less disruptive than the move to ICD-10-CM was. Fung et al. also stated that “Serious consideration should be given to using the ICD-11 for morbidity coding.”

So, these two studies are extremely encouraging, in terms of the content coverage and feasibility of replacing ICD-10-CM with ICD-11. At this point, it appears that ICD-11 can represent ICD-10-CM codes, without any modification, much better than ICD-10-CM was able to represent ICD-9-CM codes when the U.S. made that change. And this means our healthcare system should definitely invest resources to explore adopting ICD-11 to ensure that our standard reflects up-to-date medical knowledge. Remember, though we implemented ICD-10-CM seven years ago, it was first released nearly 25 years ago, and is based on the WHO’s ICD-10 system, which was created 32 years ago (and is now obsolete). Certainly, more research is needed, and NCVHS has outlined a comprehensive research agenda. It will be imperative to monitor ICD-11 developments, particularly now that the official version of ICD-11 is available.

Programming Note:

Listen to Mary Stanfill report this story live today on Talk Ten Tuesdays at 10 Eastern.