HCC Suspecting

Methods

Code on Github

The HCC Suspecting data mart identifies patients suspected of having a chronic condition but doesn't have the HCC recorded in the payment year. We use the following methods to identify these suspected conditions:

HCC Recapture: uses billed claims history to evaluate whether recurring diagnoses from prior years were captured during the current payment year.

HCC Capture/Discovery: uses all available sources from clinical and claims data to evaluate a patient's medical history, problems, comorbidities, lab results, medications, or observations to capture new HCCs that have not been coded before.

The 2024 CMS HCC model has 115 HCCs. Each condition category requires careful logic to identify suspecting conditions for capture. So far, we have built out the logic for the following conditions:

• Chronic Kidney Disease (HCC 326-329) using eGFR lab results.
• Depression (HCC 155) using medications and PHQ-9 assessments.
• Diabetes (HCC 37) using comorbidity of CKD Stage 1 or 2.
• Morbid Obesity (HCC 48) using a combination of vital signs and comorbidities Diabetes, Hypertension, or Obstructive Sleep Apnea.

Coding System Map: the terminology set SNOMED-CT to ICD-10-CM Map is used to capture additional suspecting conditions coded in a system not part of the CMS HCC model. This use case follows the default mapping guidance from NLM, which specifies that the map priority rule of “TRUE” or “OTHERWISE TRUE” should be applied if nothing further is known about the patient’s condition.

Data Dictionary

list

This final model displays the list of suspecting conditions per patient with the reason and contributing factors.

Column	Data Type	Description	Terminology

list_rollup

This final model displays the list of suspecting conditions per patient and hcc with the latest contributing factor rolled up.

Column	Data Type	Description	Terminology

summary

This final model displays a rollup of suspecting conditions per patient.

Column	Data Type	Description	Terminology

Example SQL

Total Suspected HCCs

select
      hcc_code
    , hcc_description
    , count(*) as gap_count
from hcc_suspecting.list
group by
      hcc_code
    , hcc_description
order by
      hcc_code
    , hcc_description;

Total Suspected HCCs by Reason Category

select
      reason
    , count(*) as gap_count
from hcc_suspecting.list
group by reason
order by reason;

Actionable Patient List

select
      patient_id
    , patient_birth_date
    , patient_age
    , patient_sex
    , suspecting_gaps
from hcc_suspecting.summary
order by suspecting_gaps desc;