Hospitalization Costs of Primary Stenting vs.Thrombolysis in Acute Myocardial Infarction: CostAnalysis of the Canadian STAT StudyLe May MR, Davies RF, Labinaz M, et al. Circulation 2003;108:2624 –30.

Study Question: The investigators compared 6-month hospi-talization costs associated with primary stenting with thosefor accelerated tissue plasminogen activator (tPA) in theStenting vs. Thrombolysis in Acute myocardial infarctionTrial (STAT) trial.Methods: The study design was a posthoc analysis of arandomized clinical trial. Initial and 6-month hospitaliza-tion costs were computed for all patients randomly assignedto primary stenting (n�62) or accelerated tPA (n�61) inthe Stenting vs. Thrombolysis in Acute myocardial infarc-tion Trial (STAT). Costs and resource usage were collectedin detail for each patient. Physician fees were obtaineddirectly from billings to the Ontario Health Insurance Plan.Results: The length of initial hospitalization was 6.7�11.3days in the stent group and 8.7�6.7 days in the tPA group(p�0.001). Total hospitalization days at 6 months were8.3�13 days in the stent group and 12.1�14.0 days in thetPA group (p�0.001). Hospitalization costs were less in thestent group for the initial hospitalization, $6354�6382 vs.$7893�4429 (p�0.001) and at 6 months, $7100�7111vs. $9559�6933 (p�0.001).Conclusions: The authors concluded that in centers in whichfacilities and experienced interventionists are available, pri-mary stenting is less costly and more effective than throm-bolysis.Perspective: The study demonstrates that primary stentingis cost saving compared with thrombolysis. Compared withaccelerated tPA, primary stenting was associated with sig-nificantly lower hospitalization costs than accelerated tPA.Primary stenting has also been shown to be more effectivethan thrombolysis in a number of previous studies, makingit a dominant strategy whenever feasible and appropriatefacilities and personnel are available. DM

Simplified Scoring System for Predicting MortalityAfter Percutaneous Coronary InterventionQureshi MA, Safian RD, Grines CL, et al. J Am Coll Cardiol2003;42:1890 –5.

Study Question: The investigators sought to develop a sim-plified scoring system based on pre-intervention clinicalcharacteristics to predict in-hospital mortality after percu-taneous coronary intervention (PCI).Methods: The study design was a prospective cohort design.Crude mortality and univariate odds ratios (ORs) for mor-tality associated with multiple clinical characteristics werecalculated for 9954 patients undergoing PCI at the WilliamBeaumont Hospital during 1996–1998. Based on the OR,each factor was assigned a weighted score. Using these

scores, a classification was constructed to determine theprobability of death after PCI, with classes I through IVrepresenting an increasing probability of procedural mor-tality. This classification was subsequently validated in aseparate group of patients.Results: The factors with the highest univariate odds ofdying and their scores were: myocardial infarction �14days �7; elevated creatinine �4; multivessel disease �4and age �65 years �3. Classes were created based on thepresence of these factors in a given patient. The odds ofdying and mortality increased significantly with each class.These results were reproduced in the validation subset.Conclusions: The authors concluded that preproceduralclinical risk factors have a differential influence on theprobability of death after PCI. Risk classification based onthese factors can be used to accurately predict the proce-dural outcome. This simple classification can be used byinterventionalists to assist in management decisions, toprovide an estimate of procedural risk to the patients andrelatives and for quality assurance.Perspective: The objective for this and any risk-predictionor adjustment tool should be ideally to foster continuousquality improvement. Although simple bedside scoring asproposed by this study and as developed by the BMC2

consortium may be useful classifying patients into broadrisk categories, public presentation of the results and scorecarding must be undertaken cautiously and with adequateexplanation of limitations. DM

Analysis of the Institutional Volume-OutcomeRelations for Balloon Angioplasty and Stenting inthe Stent Era in CaliforniaBrown DL. Am Heart J 2003;146:1071– 6.

Study Question: The investigator separately analyzed theoutcomes after coronary angioplasty and stenting in a large,unselected population to determine the effect of annualinstitutional angioplasty or stent volume on in-hospitaloutcomes after balloon angioplasty or stenting.Methods: The study is a retrospective analysis of an admin-istrative database. The California Office of Statewide HealthPlanning and Development database was queried to obtaindischarge data on all patients in 1997 treated with percuta-neous transluminal coronary angioplasty (PTCA) or a stentprocedure. Hospitals were divided into low-volume (�200procedures), intermediate-volume (200–400 procedures)and high-volume (�400 procedures) institutions as a func-tion of their 1997 PTCA or stent volumes to assess the effectof volume on outcome.Results: There were 44,276 percutaneous revascularizationprocedures performed in California in 1997, of which 57%involved coronary stent placement. Mortality rates afterPTCA were 2.6%, 1.9% and 1.4% in low-, intermediate-and high-volume PTCA hospitals, respectively (p�0.001).The need for same-day coronary artery bypass grafting

ACC CURRENT JOURNAL REVIEW Mar 2004

52