Dr Sudipto Mukherjee: Identifying Patients With ISM in the US

It can be difficult to conduct population studies on patients with rare diseases, especially when using a health claims-based dataset that relies on diagnostic codes.

We feel relatively certain that we were able to capture most, if not all patients with indolent systemic mastocytosis (ISM) living in the US, says Sudipto Mukherjee, MD, PhD, MPH, hematology and medical oncology at Cleveland Clinic.

Transcript

How confident are you in the accuracy of the identification methods for patients with ISM and the algorithm for disease progression, considering the limitations associated with health claims data?

In addition to the diagnosis codes, we put in several layers of criteria to enhance and enrich our diagnostic certainty. And that included having at least 2 systemic mastocytosis (SM) ICD-10 diagnosis codes that were more than 30 days apart. This was done specifically to make sure that SM was not there as a rule-out diagnosis that a lot of times the physicians would code to be able to bill those visits.

We made a particular criterion where the SM ICD-10 diagnosis code was used only after the patient had a bone marrow biopsy. As you can imagine, for having a definitive confirmation of systemic mastocytosis, you have to have a bone marrow biopsy as part of the diagnostic workup. And then in addition to that, not just relying on the diagnosis code, but making sure that the diagnosis code occurred at least 2 times more than 30 days apart.

Second, we made sure that the diagnosis code for SM was captured following a bone marrow biopsy that the patient may have gone through for the diagnosis. So, we incorporated a procedure code, which is the bone marrow biopsy. We also included treatment codes that were geared to specific treatments that these patients may have received that are systemic mastocytosis directed, like tyrosine kinase inhibitors, which could be midostaurin or avapritinib, or having received chemotherapy drugs that are routinely used for systemic mastocytosis, like cladribine, interferon therapy, or supportive treatments that these patients may have received, like cromolyn, sodium, and omalizumab, and several other supportive treatment codes that are typically used in supportive care of systemic mastocytosis.

In addition, we make it so that any diagnostic codes corresponding to advanced cases of systemic mastocytosis preceded the index date of diagnosis of ISM. We believe that this is fully a very stringent rigorous set of criteria that—to the fullest extent that we can—captures the diagnosis accurately. If you look at it from the 320 million patients who were in the Komodo dataset, we were able to identify just based on the ICD-10 code for ISM, we were able to identify 25,000 patients with some evidence of SM. And after applying not just the diagnosis codes but introducing a time limit for 2 diagnosis codes and including the procedure code and the treatment codes, we were finally able to zero down and have what we believe is roughly 8400 patients with ISM. So that's a very small number, starting with 320 million patients. There is no way to be absolutely certain in the absence of pathological reports and clinical notes, but we feel reasonably certain based on this very rigorous set of criteria that we captured if not all, most of all the ISM cases in the US.