Financial consequences of the implementation of a rapid response system on a surgical ward

Rationale, aims and objectives Rapid response systems (RRSs) are recommended by the Institute for Healthcare Improvement and implemented worldwide. Our study on the effects of an RRS showed a non-significant decrease in cardiac arrest and/or unexpected death from 0.5% to 0.25%. Unplanned intensive care unit (ICU) admissions increased significantly from 2.5% to 4.2% without a decrease in APACHE II scores. In this study, we estimated the mean costs of an RRS per patient day and tested the hypothesis that admitting less severely ill patients to the ICU reduces costs. Methods A cost analysis of an RRS on a surgical ward, including costs for implementation, a 1-day training programme for nurses, nursing time for extra vital signs observation, medical emergency team (MET) consults and differences in unplanned ICU days before and after RRS implementation. To test the hypothesis, we performed a scenario analysis with a mean APACHE II score of 14 points instead of the empirical 17.6 points for the unplanned ICU admissions, including 33% extra MET consults and 22% extra unplanned ICU admissions. Results Mean RRS costs were (sic)26.87 per patient-day: implementation (sic)0.33 (1%), training (sic)0.90 (3%), nursing time spent on extended observation of vital signs (sic)2.20 (8%), MET consults (sic)0.57 (2%) and increased number of unplanned ICU days after RRS implementation (sic)22.87 (85%). In the scenario analysis mean costs per patient-day were (sic)10.18. Conclusions The costs for extra unplanned ICU days were relatively high but the remaining RRS costs were relatively low. The 'APACHE II 14' scenario confirmed the hypothesis that costs for the number of unplanned ICU days can be reduced if less severely ill patients are referred to the ICU. Based upon these findings, our hospital stimulates earlier referral to the ICU, although further implementation strategies are needed to achieve these aims.