Hospital Population Health Program Reports

Overview

In CY 2021, the Centers for Medicare and Medicaid Innovation requested HSCRC staff to develop one or more measures that would incentivize hospitals to make improvements in population health. The HSCRC works with various workgroups and stakeholders to develop policies focused on encouraging population focused care at Maryland hospitals.

Below is a list of the programs under the Hospital Population Health Initiatives to date:

• Hospital Inpatient Diabetes Screening – The intent of this policy is to encourage testing in accordance with the time intervals recommended by the ADA (e.g., every three years for those whose most recent test was normal). However, the HSCRC are not currently able to remove patients from the denominator if they were recently tested and thus do not require additional screening. We are working to obtain data that would allow us to do this and will update the measure specification when these data become available. In the meantime, hospitals should provide screening in accordance with the ADA guidelines. Because the measure is focused on improvement, and it is unlikely that 100% of eligible patients will be screened, hospitals will still be able to perform well even when recently screened individuals appear in the denominator, but are excluded from inpatient screening and hence do not appear in the numerator.

Hospital Inpatient Diabetes Screening

The policy measures the proportion of eligible inpatient stays during which the patient received a hemoglobin A1c test, and is defined as follows:

Denominator: Inpatient claims records with a discharge date occurring during the measurement period. Claims for patients who died during the inpatient stay, patients who were transferred, those who left against medical advice, and those under 35 years old are excluded.

Numerator: Patients in the denominator with an A1c lab result in the CRISP hospital lab feed. Lab records are matched on CRISP EID, admission date, discharge data, and hospital ID. Those records with test units in milligrams per deciliter reflect estimated average glucose (EAG) rather than A1c), and are converted to A1c using the following formula : A1c = (EAGmg/dl + 46.7)/28.7

Records with a lab value of less than 3 or over 21 are excluded, as are those with a lab service date outside of the inpatient stay period.

HSCRC obtains A1c test information from lab data feeds reported to CRISP by hospitals. Details on the inpatient stay, such as vital status and whether the patient left against medical advice, are obtained from HSCRC inpatient casemix data.

Report Documentation:

• Policy Memo
• Data Dictionary