Risk factors of pulmonary embolism in senile and non-senile inpatients and the predictive value of Caprini risk assessment model in these two populations

Abstract

Objective: To investigate the risk factors of pulmonary embolism (PE) in senile and non-senile inpatients, and evaluate the predictive value of Caprini risk assessment model in these two populations. Methods: Case control study design was used in this study. All the PE patients diagnosed in West China Hospital of Sichuan University between January 2012 and December 2014 was included and divided into senile PE group (age ≥65 years old) and non-senile PE group (age <65 years old). Age matched controls (senile control group and non-senile control group) were selected from the patients admitted into the same departments during the same time period as PE patients, at a ratio of 2∶1. The risk factors of PE for senile and non-senile inpatients were investigated through comparing senile or non-senile PE patients with corresponding controls. All the subjects were retrospectively evaluated by Caprini risk assessment model, and the associations between risk stratifications and PE risk were analyzed. Results: A total of 135 senile PE patients and 212 non-senile PE patients were finally included in this study, with average age of (73.58±6.66) years old and (45.60±13.11) years old, respectively. And 233 age-matched senile controls and 418 non-senile controls were also included. The multivariate analysis showed diabetes (OR=4.08, 95%CI: 1.58-10.51, P=0.004) , heart failure (OR=3.67, 95%CI: 1.10-12.20, P=0.034) , swollen legs (OR=10.50, 95%CI: 5.57-19.79, P<0.001) , severe lung disease (OR=2.05, 95%CI: 1.08-3.90, P=0.028) , patient confined to bed (>72 h) (OR=58.33, 95%CI: 7.46-456.17, P<0.001) were independent risk factors of PE in senile patients, while obesity[body mass index (BMI)≥25 kg/m(2)](OR=2.72, 95%CI: 1.42-5.24, P=0.003), history of deep venous thrombosis (DVT)/PE (OR=17.54, 95%CI: 2.74-112.19, P=0.002) , hip, pelvis, or leg fracture (OR=18.31, 95%CI: 1.97-170.11, P=0.011) , swollen legs (OR=18.53, 95%CI: 11.29-30.40, P<0.001) , severe lung disease ( OR=4.11, 95%CI: 2.41-7.00, P<0.001) , patient confined to bed (>72 h) (OR=4.04, 95%CI: 2.03-8.04, P<0.001) were independent risk factors of PE in non-senile patients. Among the senile patients, the risk of PE increased with the increase of Caprini risk levels; compared with Caprini moderate risk, classifications of high risk and highest risk were associated with 4.64-fold (95%CI: 1.05-20.44, P=0.043) and 10.74-fold (95%CI: 2.46-46.94, P=0.002) increased risk of PE, respectively; within the highest subgroup, the per 2-score increase of Caprini score was associated with 3.02-fold (95%CI: 1.76-5.19, P<0.001) increased risk of PE. Among those non-senile patients, the risk of PE for Caprini low risk and high risk patients was not significantly different, compared with Caprini moderate risk patients; however, the highest risk was still associated with 3.94-fold (95%CI: 2.39-6.51, P<0.001) increased risk of PE compared with moderate risk; within this subgroup, the per 2-score increase of Caprini score was associated with 2.13-fold (95%CI: 1.21-3.73, P=0.008) increased risk of PE. Conclusions: Swollen legs, severe lung disease, confined to bed (>72 h) are common PE risk factors among both senile and non-senile inpatients. Diabetes, heart failure are unique PE risk factors for senile inpatients, while obesity (BMI≥25 kg/m(2)), history of DVT/PE, hip, pelvis, or leg fracture are unique PE risk factors for non-senile inpatients. The Caprini risk assessment model has better predictive value in senile patients than non-senile patients, while Caprini highest risk classification is companied by significantly increased risk of PE in both populations.