Introduction

Chronic obstructive pulmonary disease (COPD), a complex multi-component disease with chronic systemic inflammation, is a common cause of death around the world [1, 2]. Acute exacerbations in COPD (AECOPD) are key events and they negatively influence the health status of patients, rates of hospitalization, readmission, and disease progression [3, 4]. However, effective methods to identify hospitalized patients with AECOPD with poor outcomes are still lacking. It is vital for clinicians to pay more attention to evaluating the severity and prognosis of patients with AECOPD in advance. Some severity scoring systems had been developed to assist with predicting the outcomes for AECOPD, such as the APACHE II score, DECAF score and SAPS II score [5,26]. In our cohort, we observed that those who died during the hospitalization had lower albumin levels than the survivors, which was similar to the results from a previous study [15]. Clinically, we should pay more attention to the nutritional status of patients with AECOPD.

Many studies have reported the prognostic value of BUN/ALB ratio in some respiratory diseases. Uga** et al. [25]conducted a prospective study with 175 patients and they previously found that an elevated BUN/ALB ratio was a simple but independent predictor of mortality and severity of community-acquired pneumonia. The BUN/ALB ratio could reflect the severity of the patients with pneumonia. For aspiration pneumonia patients, Ryu et al. [27]demonstrated that the BUN/ALB ratio was associated with increased 28-day mortality. In addition, Fang et al. [28]concluded that the BUN/ALB ratio could be a useful prognostic tool to predict mortality in critically ill patients with an acute pulmonary embolism. In our prospective cohort study, we first investigated the association between BUN/ALB ratio and in-hospital and 90-day all-cause mortality. Interestingly, after adjusting the confounding factors, we found that a BUN/ALB ratio ≥ 0.249 was a strong and an independent risk factor of in-hospital (RR = 15.08, 95% CI 3.80–59.78, P < 0.001) and 90-day all-cause mortality (HR = 5.34, 95% CI 1.62–17.57, P = 0.006) in patients with AECOPD. Therefore, we considered that our study might suggest a supplement to the BUN/ALB ratio as a prognostic factor for respiratory diseases. With stratified BUN/ALB ratios, we observed that those with a high BUN/ALB ratio level had increased in-hospital (30.5 vs. 1.9%, P < 0.001) and 90-day all-cause (17.1 vs. 2.7%, P < 0.001) mortality. These results indicated that higher BUN/ALB ratios meant patients with AECOPD had higher BUN levels and lower albumin levels in serum. Thus, a higher BUN/ALB ratio often reflects the severity and increased mortality of AECOPD. Many factors might explain the association between a high BUN/ALB ratio level and in-hospital and 90-day mortality in patients with AECOPD. First, once patients with COPD suffer acute exacerbations, they usually suffer from infection [16] and the inflammatory response will accelerate the process of proteolysis and the BUN levels will be elevated due to lower albumin levels in patients. Second, congestive heart failure is a common comorbidity in COPD [16]. Because of decreased cardiac output, the renin angiotensin aldosterone system (RAAS) and sympathetic nervous system are activated. Meanwhile, angiotensin and adrenergic stimulation cause renal vasoconstriction and decreased GFR and renal blood flow. This process can enhance the reabsorption of the urea, causing BUN levels to become elevated [29]. In our cohort, we found that the BUN/ALB ratio of  ≥ 0.249 group had more patients with AECOPD with renal dysfunction than the BUN/ALB ratio of  < 0.249 group (62.7% vs. 43.8%, P = 0.009). We suggest that renal dysfunction might lead to urea excretion obstruction. Specifically, reabsorption of the urea will be enhanced. However, in this study, we did not find renal dysfunction was associated with in-hospital mortality in AECOPD (RR = 0.32, 95% CI 0.09–1.13, P = 0.076), which was inconsistent with previous studies [30, 31]. We suggest that our small sample size or different statistical methods might be the reasons for the different results.

Given that BUN/ALB ratio was associated with in-hospital and short-term prognosis in patients with AECOPD. We recommend that BUN/ALB ratio should gain prominence in clinical work. How to quickly and accurately assess the severity of AECOPD in patients and give corresponding intervention measures to improve the prognosis and outcome is crucially important. One research from Zemans et al. [32]showed that combined clinical variables might be more highly predictive of outcomes than individual clinical variables in COPD. BUN/ALB ratio, combined with BUN and ALB, might play a better role in predicting the outcomes of patients with AECOPD (Fig. 2). Mathioudakis et al. [33] summarized novel data on the diagnosis, phenoty**, targeted treatment, and prevention of AECOPD and indicated that the development and validation of accurate biomarkers for early characterization of the different types of COPD exacerbations is still a challenge. Biomarkers should be measurable quickly and simple to perform. In this research, we showed that an elevated BUN/ALB ratio was associated with increased mortality in patients with AECOPD. BUN and ALB measurements are easy to obtain, convenient to use, and are not affected by the evaluator's subjectivity, which is more realistic in clinical work. Therefore, we suggest that BUN/ALB ratio might be a promising indicator for assessment of the severity of AECOPD in patients.

In this study, we also found that CHF was an independent risk factor for in-hospital mortality (RR = 4.36, 95% CI 1.07–17.85, P = 0.040), which was consistent with previous findings that CHF was a significant and independent risk factor for all-cause mortality in COPD and had an influence on the COPD course [34]. Stratified by the BUN/ALB ratio, compared with the BUN/ALB ratio of  < 0.249 group, there were more patients with congestive heart failure in the BUN/ALB ratio of ≥ 0.249 group (44.1% vs. 27.3%, P = 0.012). The pathophysiological mechanism behind the association between elevated BUN/ALB ratios and CHF has been previously described [25, 29]. This might indicate that those combined with CHF have a higher BUN/ALB ratio and more attention should be paid to those with CHF when considering adverse hospitalization outcomes.

Compared with COPD, the inflammation in AECOPD is amplified, causing a poor prognosis [35]. In our study, we found that those died in hospital had a higher neutrophil count level than survivors [10.27 (7.21–14.04) vs. 6.58 (4.58–9.04) × 109/L, P < 0.001]. Meanwhile, we also found that a higher neutrophil count was an independent predictor of in-hospital mortality in patients with AECOPD after adjustment in a multivariate logistic regression (RR = 1.15, 95% CI 1.02–1.30, P = 0.020). More importantly, from Table 3, in the BUN/ALB ratio ≥ 0.249 group, we observed that the neutrophil counts were higher than those with a BUN/ALB ratio of  < 0.249. This indicated that the systemic inflammatory response contributed to a decrease in the albumin levels in serum and increased the BUN levels [25]. As indicated above, a higher neutrophil count might predict adverse hospital outcomes of patients with AECOPD.

There are several limitations in our study. First, this study was a single-center study with a small sample size, especially a small number of patients who died in hospital. However, to the best of our knowledge, this study was the largest prospective cohort study that we know of that was applied to investigate the prognostic of BUN/ALB ratio in patients with AECOPD. Second, although it was the first study to reveal the relationship between BUN/ALB ratio and in-hospital and 90-day all-cause mortality of patients with AECOPD, there is still a lack of previous studies as references. Meanwhile, in the present study, we did not analyze the effect of therapeutic interventions on BUN levels of patients with AECOPD during hospitalization, which may influence the results. We also did not exclude the patients who had chronic kidney disease. As renal dysfunction is common in AECOPD, its inclusion in the cohort might have made our study closer to a real-world study. Third, patients with AECOPD often have severe hypoxia on admission and they may have been given oxygen treatment in the outpatient department before the first blood draw. To some extent, it may influenced the results of our study. Fourth, previous studies [25, 27] have suggested that BUN/ALB ratio was associated with pneumonia. However, clinically, it is difficult to accurately distinguish AECOPD from COPD combined with community-acquired pneumonia, although we only recruited the patients discharged from the hospital with AECOPD as the primary reason for hospitalization. Finally, in our study, we only explored the short-term prognosis and we do not know the association between BUN/ALB ratio and long-term mortality. Thus, the results of this study still need to be further verified by large-sample, multicenter prospective studies in the future.

Conclusion

In this cohort study, we found that an elevated BUN/ALB ratio was associated with in-hospital and 90-day all-cause mortality in patients with AECOPD.