Final assessment of the ambulatory transfer potential for the coronary angiography patients in a teaching hospital in Malaysia

Introduction

Our prior work over 2005 and 2006 validated the feasibility of our method to identify retrospectively transferable cases using IR DRG in-patient groups and exclusion criteria based mainly on severity of illness for same encounters. Now, we want to implement a method to identify behavioral changes required from the clinicians. In order to do so, we needed to demonstrate the full transferability and savings using our approach. To do so, we applied standard validity testing with a gold standard. We also tested additional global criteria of health status, taking into account all clinical conditions over the last three-year period, which would add to the retrospective predictions.

Methods

All 718 patients who underwent single-vessel angiography in either in-patient or out-patient settings at Hospital UKHM from 1 January 2003 through 31 December 2005 were included in the study. Out of these, 298 were in-patients and the rest were outpatients. The patients who were treated as in-patients are kept as the main object of our study.

Our initial retrospective ambulatory substitution criteria include patients that live less than 100 km from the hospital; had a length of stay of no more than two days; do not have an IR-DRG severity of illness level 2 or 3; and were discharged back home at the end of the episode of care (i.e., discharge status of 04). Our prospective criteria are the same as the retrospective criteria except that we do not consider length of stay. We apply these ambulatory substitution criteria on the in-patient group to assess how many of these patients treated in an in-patient setting could have been shifted to out-patient.

We summarize all the clinical information from all the in-patient encounters of these patients for the whole three-year period. All the diagnoses in these episodes of care are then coded and grouped with 3 M CRG (Clinical Risk Groups) software to rate their global health status (nine statuses and 27 subgroups using severity of illness).

Two cardiac surgeons independently assess the potential transferability of the in-patients. We show them only the clinical abstracted information for all the in-patients stays; differences are resolved by consensus, adding the author's judgment.

At the end, we are able to construct a screening-validity two-by-two table to assess the sensitivity, specificity, positive predictive value and negative predictive value of our tool. We also compare the above indicators, adding the additional criteria of exclusion of most severe health status and severity of illness based on CRG. Cost weights are assigned to these patients to estimate the savings if these patients are successfully transferred from in-patient care to out-patient care.

Results

Results will be presented in detail during the PCSI conference.

Conclusion

Based on this work, we will decide to include CRG retrospective criteria to monitor the performance of transferability of our angiography cases. This work will be used to successfully convince our clinicians not to automatically admit patients undergoing this procedure. The potential impact of this information will demonstrate the savings as a result of such a transfer.