By Mary H. Stanfill, MBI, ACHIP, RHIA, CCS, CCS-P, FAHIMA

Content analysis is a key consideration in evaluating a new code set. This involves examining the content and level of detail represented in each code set to identify how specific code categories and codes are the same or different and how changes may impact continuity of data and statistical analyses. A thorough analysis, however, can be extremely resource intensive. Some preliminary work in this respect has been completed on ICD-11. So far, efforts to evaluate the content of ICD-11 compared to ICD-10-CM are promising. Keep reading for a summary of what the early comparisons have shown.

The National Library of Medicine (NLM) has taken the lead in ICD-11 content analysis, with studies conducted by the NLM lead health data scientist, Kin Wah Fung, and colleagues. They performed an initial, preliminary content analysis that was published in 2020. This work examined a small group of 388 ICD-10-CM codes, representing commonly reported conditions. For these 388 ICD-10-CM codes, they found approximately 60% could be fully represented in ICD-11 by using both pre- and postcoordinated codes.

If you’ve been following my blogs, you know that ICD-11 introduces a new feature, postcoordination, that is used to link codes together, in a code “cluster,” to capture all the details of a condition when those details are not included in a combination code, i.e. a “precoordinated” code. All ICD-10-CM codes are precoordinated. However, ICD-11 has both pre- and postcoordinated codes. The postcoordination feature presents challenges for implementation (uncertainties include how to select, capture, store, retrieve, and share codes that belong in a cluster) and it remains uncertain whether the US will implement this feature along with ICD-11.

Additional work by Kin Wah Fung and colleagues was published in 2021. This 2021 study evaluated the feasibility of replacing ICD-10-CM with ICD-11 by examining the most frequently used ICD-10-CM codes from each chapter. The authors evaluated 943 ICD-10-CM codes, which were manually recoded in ICD-11. Their results showed 23% of the ICD-10-CM codes were fully reflected in ICD-11 without postcoordination. But using postcoordination was a game changer. By using postcoordination and adding 9 additional extension codes, nearly 60% of the 943 ICD-10-CM codes could be fully represented by ICD-11. These results are astounding, statistically speaking, particularly when you recall that only 24% of the ICD-9-CM codes could be fully represented by ICD-10-CM codes when we made that transition in 2015. The authors therefore conclude that the move from ICD-10-CM to ICD-11 is not more disruptive than it was to move from ICD-9-CM to ICD-10-CM, which is very encouraging.

The most recent study, published in 2023, examined 1725 ICD-10-CM codes and found 35% were fully reflected with ICD-11 precoordinated stem codes. That expanded to 46% when additional ICD-11 Foundation entities were used, and to 89% using the ICD-11 postcoordination feature. The authors concluded the existing ICD-11 content can fully represent 89% of ICD-10-CM codes, if postcoordination is used.

While these results are promising, evaluations of the ICD-11 content compared to the ICD-10-CM content are still underway. More information is needed on the hierarchical structure of ICD-10-CM compared to ICD-11. We also need creative solutions to deploy ICD-11 in an automated fashion with postcoordination. So, stay tuned.