|
|
|
|
State Data
|
|
The state inpatient admission data sets (Commercial and Medicaid) contain information similar to the Medicare data described above, but some variables have state specific definitions that differ from each other. We judgmentally convert them to a consistent basis. The diagnoses and procedures are all coded using ICD-9 codes, the same as Medicare, although some states collect nine codes for diagnoses and procedures per admission, while others collect up to 25 separate codes. Because of these differences and because auditing of state collected data is also of variable quality, there are more likely to be inconsistencies in the state databases than in the Medicare database. On the other hand, hospital comparisons within states will avoid many of the inconsistencies that result from coding differences between states. However, the results of the HEI are very consistent between the Medicare and state databases.
The following is a list of issues that tend to make the state databases less reliable than Medpar data. In the state data:
| • | Some hospitals do not report insurer denied days (at least on some of their admissions) which lowers their LOS and artificially improves their efficiency.
| | • |
Some hospitals misreport some ambulatory cases as inpatient or various other coding differences may exist.
| | • |
There may not be consistent coding for DRG purposes from hospital to hospital when the outcome does not directly affect reimbursement. This would imply a different coding method for Medicare and non-Medicare admissions.
| | • |
No uniform method for auditing state records exists similar to CMS’s specifications for Medicare Intermediaries. Some states have no audits. Some records are submitted with no charges or such low charges that they are inconsistent with inpatient admission diagnoses and procedures.
| | • |
Mental Health (MH) or Sub-Abuse (SA) facilities are not required to submit data in some states. Sometimes this applies to units within a hospital where separate ownership or control applies. This can affect both hospital and statewide comparisons since MH and SA are among the least efficient services.
| | • |
The state data tends to show a higher percentage of ungroupable admissions or admissions with inconsistent diagnoses and procedures than the Medpar data. Sometimes these result from mechanical coding errors like wrong spacing or incorrect ordering of variables.
|
|
|
