Clinical Risk Score for Postoperative Pneumonia Following Heart Valve Surgery
Postoperative pneumonia (POP) is one of the most common infections following heart valve surgery (HVS) and is associated with a significant increase in morbidity, mortality, and healthcare costs. The incidence of POP after cardiac surgery varies widely, ranging from 2.1% to 21.6% in published studies. Despite the progress in anesthesia and surgical techniques, the emergence of drug-resistant bacteria has markedly increased the risk of POP. Although numerous studies have attempted to identify predictors of POP after cardiac surgery, many were based on small sample sizes and were conducted decades ago. Furthermore, most studies included patients undergoing various types of surgery, such as coronary artery bypass grafting (CABG), but a convincing prediction model specific to POP following HVS is still lacking. Therefore, understanding the factors influencing POP after HVS and developing a validated prediction rule are essential.
This study aimed to identify the major risk factors associated with the occurrence of POP following HVS and to derive and validate a clinical risk score. The study enrolled adults who underwent open HVS between January 2016 and December 2019 at a single institution. Patients were randomly assigned to the derivation and validation sets at a 1:1 ratio. A prediction model was developed using multivariable logistic regression analysis in the derivation set, and points were assigned to independent risk factors based on their regression coefficients.
The overall incidence of POP in this study was 8.2%, occurring in 316 of the 3853 patients. Multivariable analysis identified ten significant predictors for POP in the derivation set: older age, smoking history, chronic obstructive pulmonary disease (COPD), diabetes mellitus, renal insufficiency, poor cardiac function, heart surgery history, longer cardiopulmonary bypass (CPB) time, blood transfusion, and concomitant coronary and/or aortic surgery. A 22-point risk score was generated based on these predictors, demonstrating good discrimination (C-statistic: 0.81) and calibration (Hosmer-Lemeshow chi-square = 8.234, P = 0.312). The prediction rule also showed adequate discriminative power (C-statistic: 0.83) and calibration (Hosmer-Lemeshow chi-square = 5.606, P = 0.691) in the validation set. Three risk intervals were defined as low-, medium-, and high-risk groups.
The study population included 3853 adults undergoing HVS, with a mean age of 51.3 years and 2077 male patients. The baseline characteristics and comorbidities were similar between the derivation and validation sets. Of the 3853 operations, 75.2% were performed for isolated valve surgery, 12.5% for concomitant CABG, 10.7% for concomitant aortic surgery, and 1.6% for concomitant CABG and aortic surgery. The mean CPB time was 118.0 minutes, and blood products were transfused in 84.1% of the cases.
The most common microorganism isolated in this study was Acinetobacter baumannii (37.9%), followed by Klebsiella pneumoniae (20.9%), Staphylococcus aureus (12.6%), and Pseudomonas aeruginosa (12.2%). Polymicrobial POP was detected in 26.9% of cases. The overall mortality was 2.9%, with a rate in patients with POP of 28.2% compared to 0.7% in those without POP. The duration of mechanical ventilation (MV) was longer in patients with POP than those without POP (7.4 days vs. 1.3 days), and similar results were observed for ICU stay (13.7 days vs. 3.2 days) and hospital stay (28.1 days vs. 14.8 days).
Univariate analysis of risk factors for POP in the derivation group identified several significant predictors, including older age, smoking history, hypertension, diabetes mellitus, COPD, cerebrovascular disease, peripheral vascular disease, renal insufficiency, heart surgery history, NYHA class III-IV, ejection fraction, white blood cell count, red blood cell count, hemoglobin, platelet count, creatinine, albumin, type of surgery, CPB time, aortic cross-clamp time, and blood transfusion. Multivariable logistic regression analysis identified ten independent predictors of POP: age >60 years, smoking history, diabetes mellitus, renal insufficiency, COPD, NYHA class III-IV, heart surgery history, CPB time >120 minutes, blood transfusion, and concomitant CABG and/or aortic surgery.
A simplified risk score of 22 possible points was generated by summing the point values of all the predictors. In the derivation group, scores ranged from 0 to 20 with a median of 5. The predicted probability of POP based on the risk score was calculated, and the occurrence of POP after HVS was significantly predicted in the multivariable model. The area under the ROC curve was 0.81, demonstrating reasonable discrimination. The correlation between the observed and expected events of POP was high, indicating good calibration.
When the prediction rule for POP was applied to the validation set, risk scores ranged from 0 to 16 with a median of 5. The discriminatory ability of the risk score was robust, with an area under the ROC curve of 0.83. No significant difference was found between the derivation and validation groups. The clinical risk score outperformed the Kilic risk score (C-statistic: 0.69) and the Allou risk score (C-statistic: 0.60) in predicting POP. The rule also indicated good calibration in the validation set.
Three risk intervals were identified as low-, medium-, and high-risk groups corresponding to scores of 0 to 6, 7 to 9, and ≥10. Approximately two-thirds of the patients were categorized at low risk, nearly a quarter at medium risk, and only about one-tenth at high risk. The population composition of each risk group and their corresponding observed POP rates in the derivation and validation groups were compared. Compared with the low-risk group, the odds ratios for the occurrence of POP were 6.99 for the medium-risk group and 18.98 for the high-risk group in the derivation set. The corresponding values in the validation set were 6.23 and 20.27, respectively.
To evaluate the clinical utility of the risk score, decision curve analysis was conducted. The decision curves of the model in the derivation and validation sets indicated that the risk score could obtain more clinical net benefits within a large range of risk thresholds compared with “no intervention” or “intervention for all” strategies. The clinical impact curves also demonstrated that the model had good clinical utility and excellent predictive power.
In conclusion, this study derived and validated a 22-point risk score for POP following HVS using ten significant predictors. The rule performed well in both discrimination and calibration, and three risk intervals were created. The risk score is easily calculable and the factors incorporated are readily available, making it applicable at the bedside. It may also have utility in risk stratification and preventive interventions.
doi.org/10.1097/CM9.0000000000001715
Was this helpful?
0 / 0