Information Technology Reference
In-Depth Information
inpatient.mdotherpv,
inpatient.mdotherinsur,
inpatient.mdotherpub,
inpatient.mdprivateins,
inpatient.mdselfpaid,
inpatient.mdstate,
inpatient.mdtotal,
inpatient.mdtricare,
inpatient.mdveterans,
inpatient.mdworkcomp,
inpatient.mdpayment,
inpatient.fcmedicaid,
inpatient.fcmedicare,
inpatient.fcotherfed,
inpatient.fcotherpv,
inpatient.fcotherinsur,
inpatient.fcotherpub,
inpatient.fcprivateins,
inpatient.fcselfpaid,
inpatient.fcstate,
inpatient.fctotal,
inpatient.fctricare,
inpatient.fcveterans,
inpatient.fcworkcomp,
inpatient.fcpayment,
inpatient.totalcharge,
inpatient.totalexpenditure,
inpatient.PERWT04F,
inpatient.PANEL,
inpatient.PERWT05F
FROM NIS.INPATIENT AS inpatient
LEFT JOIN NIS.ICD9CODES AS ICD9CODES ON (inpatient.DUPERSID = ICD9CODES.DUPERSID);
QUIT ;
Table 25 gives the results of this combination to define the New Index on the MEPS dataset.
Index 6 has slightly lower expenditures as well as charges compared to index 5, with a considerable
reduction for index 7, while increasing back for index 8. Figure 19 gives a kernel density comparison
for total charges; Figure 20 gives it for length of stay.
While not as regular for Figure 15, it still has a regular hierarchy. There is a small crossover between
NewIndex levels 5 and 7 that makes it slightly irregular. The same pattern is visible for the length of
stay in Figure 20.
These results show that the New Index can be transferred to other datasets. However, it also shows
that there is room for improvement in the development of a risk adjustment index. It shows the limita-
tions in the use of the Charlson Index when the ICD9 codes are incomplete.
Search WWH ::




Custom Search