Association between albumin corrected anion gap (ACAG) and all-cause mortality in intensive care unit heart failure patients treated with inotropes and vasopressors

1. Introduction

Heart failure manifests as a clinical syndrome marked by impaired cardiac pump function, which can lead to poor prognosis for patients. In developed nations such as the United States, approximately 10%–15% of hospitalized heart failure cases require intensive care unit (ICU) admission for advanced critical care [1]. In the ICU, these patients often necessitate inotropes and vasopressors, indicating more severe disease states and heightened mortality risks [2]. Thus, accurate evaluation of adverse event risks is essential for efficiently allocating medical resources and implementing proactive interventions. To assess risk in critically ill patients, existing scoring systems such as the Sequential Organ Failure Assessment (SOFA) and Acute Physiology Score Ⅲ (APS Ⅲ) scores are commonly utilized [2, 3]. However, these scores exhibit significant heterogeneity, and their effectiveness in assessing conditions varies across different populations [4, 5, 6], thereby presenting challenges in clinical decision-making and highlighting the need for more reliable prognostic factors to enhance the risk prediction system.

Notably, heart failure patients frequently develop metabolic acidosis as a result of hemodynamic disruptions and diuretic usage, which may significantly contribute to their poor prognosis [7]. The anion gap is a laboratory parameter commonly used for assessing metabolic acidosis [8]. Albumin, the primary unmeasured anion in plasma, plays a crucial role in maintaining a normal anion gap [9]. To counteract the impact of decreased albumin levels on the anion gap, researchers have introduced the concept of the albumin-corrected anion gap (ACAG). Growing evidence indicates that ACAG serves as an indicator of poor prognosis in critically ill individuals. A retrospective analysis has demonstrated that ACAG levels could assess the risk of in-hospital mortality in ICU patients with sepsis, outperforming both the anion gap and albumin [10]. Another study indicated that elevated ACAG was a distinct risk factor for in-hospital mortality in ICU patients with cardiac arrest [11]. However, the association between ACAG and the risk of in-hospital mortality in ICU heart failure patients requiring inotropes and vasopressors remains uncertain.

In this study, we analyzed the clinical data from ICU patients diagnosed with congestive heart failure based on international classification of diseases (ICD-9 and ICD-10) codes and receiving inotropic and pressor therapy. We hypothesized that ACAG levels were associated with the risk of in-hospital mortality in those patients and may represent a valuable supplement to SOFA and APS Ⅲ scores. The potential integration of ACAG could help mitigate the variability of scores and improve the risk assessment system.

2. Materials and methods

2.2 Inclusion and exclusion criteria

Data on congestive heart failure patients were extracted from databases using ICD-9 and ICD-10 codes (Supplementary Table 1). The inclusion criteria comprised all patients diagnosed with congestive heart failure and treated with inotropes and vasopressors. The exclusion criteria were: (a) patients under the age of 18, (b) non-first ICU admissions, (c) ICU stays of less than 24 hours and (d) absence of treatment with inotropes and vasopressors. Ultimately, the eICU-CRD and MIMIC-Ⅳ databases were utilized to recruit 1622 and 4583 patients for this study, as illustrated in the study flowchart (Fig. 1). This study constituted a secondary analysis based on the datasets, with all patient privacy information de-identified, thereby obviating the need for obtaining patient informed consent.

Fig. 1.Flowchart of data selection from the eICU-CRD and MIMIC-IV datasets. eICU-CRD: The eICU Collaborative Research Database; MIMIC-IV: Medical Information Mart for Intensive Care Ⅳ.

3. Results

3.3 Subgroup analysis

The subgroup analysis assessed the association between ACAG and all-cause mortality in various subgroups stratified by age, gender, diabetes, hypertension, COPD, RRT, mechanical ventilation, SOFA and APS Ⅲ. Furthermore, interaction effect tests were conducted (Fig. 2B). The results revealed that elevated ACAG levels were linked with increased all-cause mortality, and this positive association remained consistent across all subgroups (p for interaction > 0.05).

Fig. 2.Association of ACAG with all-cause mortality: restricted cubic spline and subgroup analysis. (A) The adjusted cubic spline model illustrates the association between the ACAG and the risk of all-cause mortality. (B) Odds ratios and error bars representing 95% confidence intervals from Model 3 are presented by subgroups. Multiplicative interaction terms for each subgroup were assessed, with all subgroups showing no statistically significant differences (p > 0.05 for all). COPD: chronic obstructive pulmonary disease; RRT: renal replacement therapy; SOFA: the Sequential Organ Failure Assessment; APS Ⅲ: Acute Physiology Score Ⅲ; aOR: adjusted odds ratios; CI: confidence interval.

3.4 Additional prognostic value of ACAG

The results of ROC curve analysis demonstrated that the inclusion of ACAG to SOFA and APS Ⅲ scores significantly improved the accuracy of prognostic assessment (Fig. 3). Notably, ACAG provided a more substantial enhancement. Specifically, the area under the ROC curve (AUC) for ACAG was higher than for AG (0.719 vs. 0.690, p < 0.001). When AG or ACAG was combined with SOFA (AUC 0.740), the AUC increased to 0.763 or 0.772 (DeLong’s test for 0.763 and 0.772, p < 0.001), and when combined with APS Ⅲ (AUC 0.815), the AUC increased from 0.822 to 0.824 (DeLong’s test for 0.822 and 0.824, p = 0.005). Furthermore, decision curve analysis revealed similar results.

Fig. 3.Receiver operating characteristic (ROC) curves (A,B) and clinical decision curves (C,D) for AG, ACAG, SOFA, APS Ⅲ, AG or ACAG plus SOFA, and AG or ACAG plus APS Ⅲ in predicting all-cause mortality. ACAG: albumin-corrected anion gap; AG: anion gap; SOFA: sequential organ failure assessment; APS III: acute physiology score Ⅲ; AUC: area under the ROC curve.

4. Discussion

This study elucidated the association between ACAG and the risk of in-hospital mortality in ICU heart failure patients requiring inotropes and vasopressors, suggesting that ACAG could serve as a valuable supplement to the existing SOFA and APS Ⅲ scoring systems.

Metabolic acidosis is the most common acid-base imbalance in critically ill patients, with ACAG serving as a metric for assessing metabolic acid-base status. Previous evidence has indicated that ACAG acts as a prognostic marker for adverse outcomes in patients with cardiovascular diseases and critical illnesses. Jian et al. [13] identified high ACAG levels as a significant risk factor for 30-day all-cause mortality in ICU patients with acute myocardial infarction, demonstrating superior accuracy compared to the anion gap, which aligns with those reported by Sheng [12]. Additionally, Wang et al.’s [14] study suggested that elevated ACAG levels may correlate with more severe coronary artery stenosis and heart failure, thereby increasing the risk of postoperative systemic inflammatory response, microcirculation disorder and complications. Moreover, higher ACAG levels have been closely associated with in-hospital mortality in various diseases, including acute pancreatitis [15], asthma [16], acute renal failure [17], and sepsis [10]. Building on these investigations, the current study sought to investigate the association between ACAG levels and in-hospital mortality in ICU heart failure patients treated with inotropes and vasopressors. Notably, patients in the ICU requiring such treatment exhibited a markedly heightened risk of mortality compared to those not needing it, with observed mortality rates of 24.97% and 17.41%, respectively, in this population, contrasting with rates of 13.4% and 13.3% in all ICU heart failure patients [18]. Thus, our study held significant practical relevance in this regard. By conducting various analytical methods, including multivariate adjustment and curve fitting, all findings consistently revealed a linear relationship between ACAG levels and adverse outcomes, with the prognostic impact of ACAG surpassing that of AG and subgroup analysis results further underscoring the consistent relationship across all subgroups.

SOFA and APS Ⅲ, recognized as effective tools for critical care physicians to gauge the severity of patients’ conditions, are widely used in clinical practice [19]. In line with previous research, our findings underscored the prognostic value of these two scores. Importantly, the association of ACAG with all-cause in-hospital mortality risk remained independent of SOFA and APS Ⅲ scores. Furthermore, the incorporation of ROC curve analysis and decision curve analysis expanded the analytical scope of the data, offering additional confirmation from an alternative perspective. These results suggest that incorporating ACAG alongside SOFA and APS Ⅲ could significantly enhance the assessment of adverse outcomes, positioning ACAG as a potent adjunct to the SOFA and APS Ⅲ tools. To the best of our knowledge, our study represents the first endeavor to evaluate ACAG in conjunction with SOFA and APS Ⅲ. Nonetheless, acknowledging the inherent limitations of observational studies, further investigations, including randomized controlled trials, are imperative to validate the present findings.

Our study had some limitations. Firstly, being a retrospective study, inherent selection bias and regression bias were unavoidable. Nonetheless, to mitigate the influence of these biases, a sensitivity analysis was conducted on the population from an alternative center. Secondly, the absence of dynamic monitoring of ACAG, SOFA and APS Ⅲ may have limited the accurate reflection of potential changes in patient’s conditions and treatment circumstances, warranting further research to address this gap. Lastly, despite incorporating confounding factors in the multivariate logistic regression models, the possibility of residual and unmeasured confounders affecting the study outcomes remains.

5. Conclusions

ACAG was independently associated with the risk of in-hospital all-cause mortality in ICU heart failure patients treated with inotropes and vasopressors and could be considered an effective adjunct to the SOFA and APS Ⅲ scores, with its inclusion potentially enhancing the effectiveness of the risk warning system in this patient population.

Availability of data and materials

The authors declare that all data supporting the findings of this study are available within the paper and any raw data can be obtained from the corresponding author upon request.

Author contributions

JYW and KW—designed the study and carried them out. KW and GBF—supervised the data collection. JYW and YW—analyzed the data; interpreted the data. JYW, YW, GBF and KW—prepare the manuscript for publication and reviewed the draft of the manuscript. All authors have read and approved the manuscript.

Ethics approval and consent to participate

Not applicable.

Acknowledgment

Not applicable.

Funding

This research received no external funding.

Conflict of interest

The authors declare no conflict of interest.

Supplementary material

Supplementary material associated with this article can be found, in the online version, at https://oss.signavitae.com/mre-signavitae/article/1888051940889116672/attachment/Supplementary%20material.docx.