OUTCOMES AMONG NSTEMI PATIENTS PRESENTING TO HOSPITALS WITH

OUTCOMES AMONG NSTEMI PATIENTS PRESENTING TO HOSPITALS WITH AND WITHOUT BACKUP CARDIAC SURGERY From the National Registry of Myocardial Infarction Presented by Yuri B. Pride, MD

Background � Among patients with NSTEMI, early angiography and, if warranted, PCI have been associated with improved outcomes � Because of this, there has been a proliferation of cardiac catheterization laboratories with PCI capability � Many such laboratories have been established at hospitals without backup cardiac surgery (No-OHS hospitals)

Background � Elective PCI at No-OHS hospitals is currently not recommended in the most recent ACC/AHA guidelines � This is based largely on a Medicare study reporting higher mortality among patients undergoing non-primary/rescue PCI at No-OHS hospitals Anderson et al. , Circulation 116 e 148 Wennberg et al. , JAMA 292 1961

Hypothesis � Improved outcomes at hospitals with backup cardiac surgery (OHS hospitals) may be the result of its use after PCI complications or may simply be a marker of improved quality of care

Objective � The goals of the current analyses were two-fold 1) To determine the outcomes NSTEMI patients presenting to hospitals with and without backup cardiac surgery 2) To specifically evaluate the outcomes among such patients who underwent PCI � In order to evaluate these objectives, we used Phase 5 of the National Registry of Myocardial Infarction, which enrolled patients from 2004 -2006

Methods � Three analyses were performed 1) Unadjusted 2) Propensity-matched for baseline characteristics, including transfer-in status, in the entire cohort, and then further adjusted for differences in treatment within 24 hours 3) Propensity-matched in the groups undergoing primary and elective PCI

Results � There were 100, 071 NSTEMI patients, of whom 9, 189 presented to 52 No-OHS hospitals and 90, 872 presented to 214 OHS hospitals � Owing to the large number of patients, there were many differences in baseline characteristics in the unadjusted populations � In the propensity-matched analysis, baseline characteristics were well-matched Pride et al. JACC Cardiovasc Interv 2: 944

Results � Patients presenting to No-OHS hospitals were significantly less likely to receive aspirin, beta-blockers or statins within 24 hours of arrival, both in the unadjusted and propensity-matched analyses � Patients presenting to No-OHS hospitals were also less likely to undergo angiography and PCI both in unadjusted and propensity-matched analyses Pride et al. JACC Cardiovasc Interv 2: 944

Results p=0. 043 p<0. 001 p=0. 64 n=94, 817 Pride et al. JACC Cardiovasc Interv 2: 944

Results p=0. 63 p<0. 001 p=0. 50 n=14, 651 Pride et al. JACC Cardiovasc Interv 2: 944

Results Adjustment Hazard ratio 95% CI p-value None 0. 70 0. 64 -0. 77 <0. 001 Medications within 24 hours * 0. 81 0. 74 -0. 89 <0. 001 Above + hospital characteristics ^ 0. 89 0. 79 -1. 00 0. 050 * Aspirin, clopidogrel, Gp. IIb/IIIa, statins, beta-blockers ^ Region, teaching hospital, urban setting, size, MI volume Pride et al. JACC Cardiovasc Interv 2: 944

Results � In addition, NSTEMI patients presenting to No-OHS hospitals were less likely to be prescribed aspirin, beta-blockers or statins at discharge in both unadjusted and propensity-matched analyses Pride et al. JACC Cardiovasc Interv 2: 944

Results, PCI � The propensity-matched analysis was then restricted to patients who underwent PCI, who were generally of lower risk than the overall population � There was no significant difference in the administration of aspirin, betablockers or statins, but patients presenting to No-OHS hospitals were more likely to receive clopidogrel in the first 24 hours. Pride et al. JACC Cardiovasc Interv 2: 944

Results, PCI p=0. 62 p=0. 51 p=0. 047 p=0. 11 n=2, 513 Pride et al. JACC Cardiovasc Interv 2: 944

Limitations � NRMI enrolled far more patients at OHS hospitals � The outcomes of patients who were transferred are unknown � Long-term outcomes are unknown

Conclusions � STEMI and NSTEMI patients presenting to No-OHS hospitals have significantly worse in-hospital outcomes, even after adjusting for differences in baseline characteristics � After further adjusting for differences in hospital characteristics and adherence to guideline recommendations, the difference in outcomes was attenuated

Conclusions � Furthermore, among patients undergoing PCI, there were actually differences favoring No-OHS hospitals in the administration of some medications as well as some clinical outcomes � These results suggest that efforts to increase adherence to guideline recommendations are warranted