| Medicare Data |
|
|
The Medicare data represents about 11 million admissions per year. Medicare data has a great deal of consistency and integrity. Under Medicare, acute medical/surgical admissions (with the exception of those in Maryland) are reimbursed by DRG. Most all hospitals have access to and use grouper software to help improve their coding and enhance their reimbursement. CMS’s Fiscal Intermediaries are charged with auditing the admissions data to make sure that the coding is complete, consistent and correct. This administrative data (based on UB-92 forms submitted by hospitals) includes: Medicare hospital ID number (sometimes several hospitals in a system use the same billing ID, which means their data is combined as one entity), some patient information, up to nine ICD-9 diagnoses, up to nine ICD-9 procedures and other information for individual admissions. This is the data that CMS uses for updating DRG payments each year and for other analytical studies.
Some of the non-medical/surgical acute Medicare admissions are not coded as consistently, since these are not currently reimbursed on a DRG basis (although this is changing.) For example, psychiatric hospitals appear to have significantly lower severities for their admissions than regular acute care full-service hospitals have for similar psychiatric DRGs. Since the full-service hospitals are required to fully code for their DRG reimbursements, they may also be more likely to fully code their psychiatric and other (non-DRG reimbursed) admissions as well. On the other hand, these acute medical/surgical hospitals may also get a higher mix of emergency room admissions, resulting in higher severities for their psychiatric admissions. Similar issues could apply to acute Long Term Care, Substance Abuse and Rehab admission categories. These issues must be considered in interpreting Hospital Efficiency results. |