External validation of the M-CARS score for mortality risk in coronary intensive care units

Abstract

Abstract Background The M-CARS (Mayo-CICU Admission Risk Score) score was recently developed and validated for coronary intensive care units (CICU) for the estimation of mortality risk at admission. Most existing scores were developed and validated for general intensive care unit patients with a distinct profile, and need a myriad of both clinical and laboratory data from the first 24 hours after admission and constant updates. The M-CARS is based on only 7 data points collected at admission: the presence of cardiac arrest, respiratory failure and/or any type of shock, the Branden score, blood urea nitrogen > 23mg/dL, anion gap > 14 and red blood cell distribution width (RDW) > 14.3. The scale varies from zero to 10 points and three risk categories were described: low (0 to 2 points), intermediate (from 2 to 4), and high (>4 points). Purpose To provide external validation of the M-CARS score. Methods Retrospective cohort of patients admitted to two CICUs in the same academic hospital. One is linked to the emergency ward and mainly admits acute cardiac care patients. The second one is linked to the cardiology clinical and surgical wards, the outpatient care center, and the interventional and surgical suites. Demographic, clinical, and laboratory data were collected during admission to obtain the M-CARS score. Our primary endpoint was death within 30 days of admission to CICU. Readmissions for the same time interval were also evaluated. Statistical analysis used parametric and non-parametric tests as needed to describe the data and for comparison between groups. P-value < 5% was established as significant. Results A total of 408 consecutive patients were included: 60% men, 81% white and the median age was 64 years. The mean stay duration was 6±6 days. The total mortality at 30 days was 9.8% (40 patients), with no age (62.4±14,6years - alive vs 66.4 ± 17.0years - dead; p=0.10), gender (p=0.31), and CICU location (10.9%- CICU linked to the emergency ward vs 8.3% - CICU linked to cardiology wards; p=0.38) differences. The M-CARS score was obtained in 363 patients, and the lack of blood chloride levels needed for anion gap calculation was the reason for the 45 patients not being included in this analysis. Within this group, 38 (10.5%) patients died. The M-CARS score was significantly higher in the group of patients who died within 30 days (4.4±2.6 vs 2.0±1,6; p<0.0001). Considering the risk categories, mortality was significantly distinct between all mortality risk categories (Low = 3.4%, Intermediate = 10.2%, and High =40.5%; p<0.0001, and P<0.0001 for all Bonferroni post-tests). Readmissions with 30 days had similar M-CARS score (P=0.64).CONCLUSIONS: The M-CARS score was adequate in identifying the mortality risk category in CICUs from outside the country it was developed. It could not predict readmissions. It may be incorporated into the admission proceedings and be used to guide medical care and instruct relatives about the risks of the patient.