Risk prediction of future cardiac arrest by evaluation of a genetic risk score alone and in combination with traditional risk factors.

OBJECTIVES: Coronary heart disease (CHD) is a leading cause of death globally, commonly through sudden cardiac death. Cardiac arrest of cardiac origin (CA) is associated with a poor prognosis and there is a great need for risk assessment and intensified preventive actions. In this study we aim to assess if a genetic risk score for CHD, composed of 50 common CHD susceptibility variants (GRS), predicts CA and to evaluate a novel composite risk score including traditional risk factors as well as GRS. METHODS: The GRS score alone and in combination with traditional CHD risk factors were examined in relation to CA incidence among 23 000 middle aged subjects during 18.9 years of follow-up. The cohort excluded patients with a diagnosed history of CHD, heart failure or stroke. RESULTS: Two-hundred-fifty-two patients suffered a cardiac arrest during the follow up, of which 181 were CA. In a multivariate model with CHD risk factors, high versus low genetic risk predicted CA with a hazard ratio (HR) of 2.49 {(95% CI 1.50-4.12) (P < 0.001)}, surpassed only by higher estimates for male sex {HR = 2.91 (95% CI 2.09-4.06) (P < 0.001)}, ages 50-65 {HR = 2.74 (95% CI 1.42-5.25) (P = 0.003)} and ages 65-74 {HR = 5.10 (95% CI 2.56-10.16) (P < 0.001)}. Smoking, dyslipidemia, hypertension and diabetes mellitus also predicted CA but with lower HRs than GRS. A novel composite risk score including CHD risk factors as well as GRS predicted CA with a HR = 110.81 {(95% CI 15.43-795.63) (P < 0.001)} for the highest (5) versus the lowest quintile (1) of the risk score. CONCLUSIONS: Genetic risk of CHD is strongly associated with incident CA and when combined with traditional CHD risk factors may identify individuals who benefit from intensified preventive pharmacological treatment.

Midlife risk factor exposure and incidence of cardiac arrest depending on cardiac or non-cardiac origin.

OBJECTIVE: Little is known about midlife risk factors of future cardiac arrest. Our objective was to evaluate cardiovascular risk factors in midlife in relation to the risk of cardiac arrest (CA) of cardiac and non-cardiac origin later in life. METHODS: We cross-matched individuals of the population based Malmö Diet and Cancer study (n=30,447) with the local CA registry of the city of Malmö. Baseline exposures were related to incident CA. RESULTS: During a mean follow-up of 17.6±4.6years, 378 CA occurred, of whom 17.2% survived to discharge. Independent midlife risk factors for CA of cardiac origin included coronary artery disease {HR 2.84 (1.86-4.34) (p<0.001)}, diabetes mellitus {HR 2.37 (1.61-3.51) (p<0.001)} and smoking {HR 1.95 (1.49-2.55) (p<0.001)}. Dyslipidemia and history of stroke were also significantly associated with an elevated risk for CA of cardiac origin. Independent midlife risk factors for CA of non-cardiac origin included obesity (BMI>30kg/m2) {HR 2.37 (1.51-3.71) (p<0.001)}, smoking {HR 2.05 (1.33-3.15) (p<0.001)} and being on antihypertensive treatment {HR 2.25 (1.46-3.46) (p<0.001)}. CONCLUSION: Apart from smoking, which increases the risk of CA in general, the midlife risk factor pattern differs between CA of cardiac and non-cardiac origin. Whereas CA of cardiac origin is predicted by history of cardiovascular disease, dyslipidemia and diabetes mellitus, the main risk factors for CA of non-cardiac origin are obesity and hypertension. In addition to control of classical cardiovascular risk factors for prevention of CA, our results suggest that prevention of midlife obesity may reduce the risk of CA of non-cardiac origin.

External Validation of Two Classification and Regression Tree Models to Predict the Outcome of Inpatient Cardiopulmonary Resuscitation.

OBJECTIVE: To prospectively validate a previously developed classification and regression tree (CART) model that predicts the likelihood of a good outcome among patients undergoing inpatient cardiopulmonary resuscitation. DESIGN: Prospective validation of a clinical decision rule. SETTING: Skåne University Hospital in Malmo, Sweden. PATIENTS: All adult patients (N = 287) experiencing in-hospital cardiopulmonary arrest and undergoing cardiopulmonary resuscitation between 2007 and 2010. INTERVENTIONS: Patients from Skåne University Hospital who underwent CPR (N = 287) were classified using the CART models to predict their likelihood of survival neurologically intact or with minimal deficits, based on a cerebral performance category score of 1. Discrimination and classification accuracy of the score in the Swedish population was compared to that in the original (derivation and internal validation) populations. MEASUREMENTS AND MAIN RESULTS: For model 1, the area under the receiver-operating characteristic curve (AUROCC) was 0.77, compared with 0.76 and 0.73 in the original derivation and validation populations, respectively. Model 1 classified 71 (2.8%) of 287 patients as being at a very low risk of a good neurologic outcome compared with 157 (26.1%) of 287 patients predicted to be at an above average risk of a good neurologic outcome. Model 2 had a similar AUROCC as the original validation population of 0.71 but lower than the original derivation population. Model 2 performed similarly to Model 1 with regards to its ability to correctly classify patients as very low or higher than average likelihood of a good neurologic outcome. CONCLUSION: Two CART models validated well in a different population, displaying similar discrimination and classification accuracy compared to the original population. Although additional validation in larger populations is desirable before widespread adoption, these results are very encouraging.

Validation of the good outcome following attempted resuscitation score on in-hospital cardiac arrest in southern Sweden.

BACKGROUND: There is a great need for a simple and clinically useful instrument to help physicians estimate the probability of survival to discharge with a good neurological outcome (cerebral performance category, CPC=1) in cases of in-hospital cardiac arrest (IHCA). Our aim was to validate the "Good Outcome Following Attempted Resuscitation" (GO-FAR) score in a different country with different demographics than previously investigated. METHODS: A retrospective observational study including all cases of IHCA who were part of a cardiac arrest registry at Skåne University Hospital in Sweden 2007-2010. RESULTS: Two-hundred-eighty-seven patients suffered IHCA during the period. A majority were male and mean age was 70years. Overall survival to discharge independent of neurological function was 20.2%; 78% of the survivors had CPC=1 and survival to discharge with CPC=1 was 15.7%. The area under the receiver operating characteristics curve for the GO-FAR score was 0.85 (CI=0.78-0.91, p<0.001), consistent with very good discrimination. Patients in the group with low or very low probability of survival had a likelihood of 2.8% (95% CI 0.0-6.7), whereas the groups with average and above average probabilities had likelihoods of 8.2% (3.7-13) and 46% (34-58), respectively, for good neurological outcome. This compares with likelihoods of 1.6%, 9.2% and 27.8% in the original study. CONCLUSION: The GO-FAR score accurately predicted the probability of survival to discharge with CPC=1, even when applied to a different population in another country.

Evaluation of pre-arrest morbidity score and prognosis after resuscitation score and other clinical variables associated with in-hospital cardiac arrest in southern Sweden.

OBJECTIVE: To evaluate pre-arrest morbidity score (PAM), prognosis after resuscitation score (PAR) and to identify additional clinical variables associated with survival after in-hospital cardiac arrest (IHCA) treated with cardiopulmonary resuscitation (CPR). METHODS: A retrospective observational study involving all cases of IHCA at Skåne University Hospital Malmö 2007-2010. RESULTS: Two-hundred-eighty-seven cases of IHCA were identified (61.3% male; mean age 70 years) of whom 20.2% survived until discharge. The odds ratio (95% confidence interval) for death prior to discharge was 6.49 (1.50-28.19) (p=0.013) for PAM>6 and 3.88 (1.95-7.73) (p<0.001) for PAR>4. At PAM- and PAR-scores >5, specificity exceeded 90%, while sensitivity was only 20-30%. The odds ratio for in-hospital mortality was 0.38 (0.20-0.72) (p=0.003) for patients with cardiac monitoring, 9.86 (5.08-19.12) (p<0.001) for non-shockable vs shockable rhythm, 0.32 (0.15-0.69) (p=0.004) for presence of ST-elevation myocardial infarction (STEMI), 0.27 (0.09-0.78) (p=0.016) for patients with independent Activities of Daily Life (ADL) and 13.86 (1.86-103.46) (p=0.010) for patients with malignancies. Heart rate (HR) on admission (per bpm) [1.024 (1.009-1.040) (p=0.002)] and sodium plasma concentration on admission (per mmoll(-1)) [0.92 (0.85-0.99) (p=0.023)] were significantly associated with in-hospital mortality. CONCLUSION: PAM- and PAR-scores do not sufficiently discriminate between in-hospital death and survival after IHCA to be used as clinical tools guiding CPR decisions. We confirm that malignancy is associated with increased in-hospital mortality, and cardiac monitoring, shockable rhythm, STEMI and independent ADL, with decreased in-hospital mortality. Interestingly, our results suggest that HR and plasma sodium concentration upon admission may represent new tools for risk stratification.