Skip to main content
Download PDF

Association between baseline serum bicarbonate and the risk of postoperative delirium in patients undergoing cardiac surgery in the ICU: a retrospective study from the MIMIC-IV database

BMC Anesthesiology volume 24, Article number: 347 (2024) Cite this article

Abstract

Background

Although serum bicarbonate is a reliable predictor of various disease complications, its relationship with postoperative delirium (POD) remains unclear. Our research aimed to assess the effect of baseline serum bicarbonate levels on the incidence of POD in cardiac surgery patients.

Methods

A retrospective analysis was conducted on cardiac surgery patients who met specific inclusion and exclusion criteria, using data from the Marketplace for Information in Critical Care Medicine (MIMIC-IV) database. Univariate and multivariate logistic regression models are employed to explore the correlation between serum bicarbonate levels and the risk of POD, and their predictive efficacy is assessed by means of restricted cubic spline regression models (RCS) and receiver operating characteristic curves (ROC). In addition, subgroup and sensitivity analyses are conducted to test the robustness of the results.

Results

In this study, 5,422 patients were included, where the incidence of POD was 13.0%. For each 1 mmol/L increase in bicarbonate, a 13% reduction in the risk of POD was observed in the fully adjusted model (OR = 0.87, 95% CI: 0.83–0.91, P < 0.001). The RCS model demonstrated a linear negative correlation between the level of bicarbonate and the risk of POD (P for nonlinearity = 0.987). The ROC curve analysis demonstrated that the bicarbonate level had moderate predictive efficacy (AUC = 0.629). Both subgroup and sensitivity analyses reaffirmed the robustness of these results.

Conclusions

Lower baseline serum bicarbonate levels in cardiac surgery patients are linked to a higher risk of POD. Monitoring and adjusting serum bicarbonate levels may help identify high-risk patients and potentially improve outcomes.

Peer Review reports

Introduction

Delirium is an acute cerebral dysfunction characterized by confusion, diminished attention, a fluctuating course, and perceptual deficits [1]. The reported incidence of postoperative delirium (POD) in cardiac surgery patients ranges from 9–53.5% [2,3,4], with a notable prevalence in the intensive care unit (ICU) [5]. The occurrence of delirium is attributed to various complex and variable pathophysiological factors, including neurotransmitter imbalance, inflammatory response, electrolyte disturbances, metabolic issues, physiological stress, and genetic factors [6]. The occurrence of POD not only extends the length of hospitalization and deteriorates the quality of life for patients but also potentially increases the risk of death [7,8,9,10]. Consequently, exploring simple and practical indicators for predicting POD risk in cardiac surgery patients in the ICU is of utmost clinical importance.

As a critical indicator of acid-base balance in the body, abnormalities in serum bicarbonate typically signify respiratory or renal dysfunction [11]. In the ICU setting, due to the widespread use of blood gas analyzers, the bicarbonate level has become a readily accessible parameter for all patients. Early studies focused on the link between bicarbonate concentrations and mortality rates in individuals with kidney disease [12,13,14]. Recently, significant correlations have been found between bicarbonate levels and complications of cardiovascular disease and mortality [15, 16]. Furthermore, the associations between bicarbonate levels and mental health conditions, such as depression, brain dysfunction, and Alzheimer’s disease, have received increasing attention [17,18,19,20]. While research has examined the link between bicarbonate concentrations and delirium rates among critically ill individuals [21, 22], these findings have been inconsistent, and investigations into a potential nonlinear relationship are limited. Consequently, the exact nature of the connection between bicarbonate concentrations and the risk of POD in ICU cardiac surgery patients remains unclear.

This study sought to clarify this connection through a retrospective analysis, utilizing data from the Medical Information Marketplace in Intensive Care (MIMIC-IV) database.

Methods

Data source

Information for this investigation was obtained from the publicly available database, MIMIC-IV (version: v2.2), documenting more than 50,000 ICU admissions at Beth Israel Deaconess Medical Center in Boston, Massachusetts, from 2008 to 2019 [23]. This database encompasses comprehensive demographic details, physiological monitoring metrics, laboratory results, and diagnoses according to the International Classification of Diseases, Ninth (ICD-9) and Tenth (ICD-10) Editions. All personal information was deidentified to protect patient privacy; therefore, this study did not require informed consent from the patients and ethical review board approval. After completing web-based training and passing the Protecting Human Research Participants examination, One of the authors, Panxu Guo, was granted permission to extract data from the MIMIC-IV database (Record ID: 58462281). The study was conducted following the STROBE guidelines [24].

Participant selection

Patients meeting specific criteria were screened using the MIMIC-IV database (version 2.2) for this study. The study cohort comprised individuals aged 18 years and older who were first admitted to the hospital for a single cardiac procedure, including coronary artery bypass grafting (CABG), heart valve surgery, or aortic surgery. Diagnoses were determined based on the International Classification of Diseases (ICD) codes (see Supplementary Table 1). The exclusion criteria were as follows: (1) ICU stays of less than 24 h, (2) a diagnosis of schizophrenia or dementia, (3) absence of documented delirium assessment, (4) a diagnosis of delirium prior to surgery or within 24 h before ICU admission, and (5) bicarbonate levels not measured on the first day of ICU admission. Following the application of these criteria, a total of 5,422 patients were eligible for inclusion in the study. The patient screening process is illustrated in Fig. 1.

Fig. 1
figure 1

The inclusion and exclusion criteria of study participants

Full size image

Data collection

Data were extracted from the MIMIC-IV database using PostgreSQL and Navicat 16.3.3 software. The collected information, recorded within 24 h of ICU admission, included patients’ age, gender, body mass index (BMI), ethnicity, vital signs, laboratory results, co-morbidities, and disease severity scores. Additionally, the use of statins and benzodiazepines in the ICU, along with hospital and ICU stays. To reduce data bias, variables with over 10% missing values were excluded. For variables with less than 10% missing data, multiple imputation was performed using the random forest algorithm provided by the R package “missForest” [25], as detailed in Supplementary Fig. 1.

Exposure

The primary outcome variable analyzed in the study was serum bicarbonate level, which was assessed as a continuous variable. It was also divided into four quartile groups based on admission values from the MIMIC-IV database: Q1 (11–22 mmol/L), Q2 (22–23 mmol/L), Q3 (23–25 mmol/L), and Q4 (25–37 mmol/L). Baseline serum bicarbonate concentration was determined by the initial measurement taken within 24 h of admission to the ICU.

Outcome

The primary outcome of this study involved determining the presence or absence of delirium during ICU hospitalization. The diagnosis of delirium was made using the CAM-ICU scoring tool, which includes four characteristics: (1) acute mental status changes or fluctuations; (2) inattention; (3) disorganized thinking; and (4) altered level of consciousness. The medical staff meticulously documented these characteristics on the “chartevents” form. According to the CAM-ICU scale, a diagnosis of delirium (CAM-ICU-positive) is confirmed when the patient displays any combination of Characteristics 1 and 2, along with either 3 or 4.

Statistical analysis

Participants’ baseline characteristics are divided into two groups based on the presence or absence of delirium. Normally distributed continuous variables are presented as mean ± standard deviation, while skewed continuous variables are reported as median (interquartile range [IQR]), and categorical variables are shown as percentages (%). Continuous data between groups were compared using the Student’s t-test and the Wilcoxon rank-sum test, depending on the data distribution, while categorical data were compared using the chi-square test (χ²).

Univariate and multivariate logistic regression analyses were employed to explore independent associations between baseline serum bicarbonate levels and the risk of developing POD. Confounders were selected based on clinical significance, prior scientific literature, and variables showing a ≥ 10% change in effect estimates. Serum bicarbonate concentration was evaluated as both a continuous and categorical variable across four adjusted models. The initial model included no adjustments for any covariates. In Model 1, adjustments were made for age, sex, BMI, and Ethnicity. Model 2 was further adjusted for comorbidities such as hypertension, myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, renal disease, and liver disease. Model 3 included adjustments for respiratory rate (RR), anion gap (AG), creatinine, sodium, PH, partial pressure of oxygen (PO2), partial pressure of carbon dioxide (PCO2), Benzodiazepine drugs, simplified acute physiology score II (SAPS II), sequential organ failure assessment (SOFA), and ICU length of stay. To prevent multicollinearity, variance inflation factors (VIF) were assessed, and variables with a VIF ≥ 5 were subsequently excluded (see Supplementary Table 2).

The potential linear relationship between serum bicarbonate concentration and the risk of POD occurrence was investigated using restricted cubic spline (RCS) modeling. Receiver operating characteristic (ROC) curves assessing the predictive value of bicarbonate concentration on the incidence of POD. Additionally, subgroup analyses were conducted to examine the consistency of POD prediction across various subgroups. Subgroups were defined according to age (< 65 years, ≥ 65 years), gender, ethnicity, surgical type, and comorbidities. Interactions between subgroup pairs were assessed using likelihood ratio tests.

To ensure the robustness of the study results, a series of sensitivity analyses were conducted. First, considering that cerebrovascular disease is a significant cause of delirium, patients with this condition were excluded to reduce potential confounding factors. Second, patients with renal disease were excluded to specifically examine the risk of postoperative delirium (POD) in cardiac surgery patients whose condition is not closely related to bicarbonate levels. Third, the analysis was performed using the original data set without applying multiple imputation to minimize potential biases introduced by the imputation process.

All statistical analyses were performed using R software (version 4.4.1). A two-sided p-value of less than 0.05 was considered statistically significant.

Results

Baseline characteristics

A total of 5,422 participants were enrolled in this study, with 3,873 (71.4%) being male and having a mean age of 66.9 ± 11.6 years. POD developed in 704 patients, which represents an incidence rate of 13.0% among cardiac surgery patients. Compared with patients who did not develop delirium, those with delirium tended to be older and exhibited higher rates of myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, renal disease, and liver disease. Additionally, these patients were more likely to undergo heart valve and aortic surgeries and exhibited higher rates of benzodiazepine use. Patients with delirium displayed significantly higher RR, AG, White blood cell count (WBC), creatinine, and lactate levels, along with elevated SAPS II, SOFA, and CCI scores, and prolonged hospital and ICU stays. In contrast, levels of systolic blood pressure (SBP), diastolic blood pressure (DBP), hemoglobin, chloride, and PO2 were significantly lower, as detailed in Table 1.

Table 1 The baseline characteristics of participants
Full size table

Incidence of POD in different groups

Figure 2A shows that the incidence of POD progressively decreases across quartiles of serum bicarbonate levels, with the highest incidence in the first quartile group (Q1) at 19.4% and the lowest in the fourth quartile group (Q4) at 8.0%. Figure 2B presents the incidence of POD stratified by age and gender, indicating a general trend of increasing POD incidence with advancing age. Notably, among patients older than 45 years, the incidence of POD is significantly higher in females than in males, while in patients over 75 years of age, the incidence of POD remained high in both males and females.

Fig. 2
figure 2

A: Bar graph depicting the incidence of POD across different bicarbonate quartiles; B: Comparative graph illustrating POD incidence across age groups

Full size image

Relationship between baseline bicarbonate levels and the incidence of POD in cardiac surgery patients

This study explored the correlation between baseline bicarbonate levels and the risk of postoperative delirium (POD) in cardiac surgery patients. Utilizing univariate logistic regression analysis, a significant association was confirmed between bicarbonate level and the risk of POD. Additionally, factors including age, gender, race, and various clinical and surgical parameters were also found to be significantly associated with the risk of POD, as detailed in Supplementary Table 3.

Further multivariate analyses revealed that a significant negative correlation between bicarbonate levels and risk of POD persisted (see Table 2). In the fully adjusted model (Model 3), a 1 mmol/L increase in bicarbonate was associated with a 13% reduction in the risk of POD (OR = 0.87, 95% CI: 0.83–0.91, P < 0.001). The highest bicarbonate level group (Q4) demonstrated a 59% reduction in POD incidence compared to the lowest group (Q1) (OR = 0.41, 95% CI: 0.29–0.59, P < 0.001). Trend tests across all models yielded p-values below 0.001, underscoring the robust negative association between bicarbonate levels and POD risk.

Table 2 Relationships between bicarbonate concentrations and the risk of postoperative delirium according to the different models
Full size table

Restricted cubic spline regression model

The relationship between serum bicarbonate levels and POD risk was elucidated using the RCS model. In the initial model (Fig. 3A) and subsequent adjustments in Model1 (Fig. 3B) and Model2 (Fig. 3C), a nonlinear relationship was observed between increasing serum bicarbonate levels and the risk of POD (P for nonlinearity < 0.05). However, in Model3 (Fig. 3D), which was fully adjusted, the relationship between bicarbonate levels and POD risk transitioned to linear, revealing a significant linear trend of decreasing POD risk with increasing bicarbonate levels (P for nonlinearity = 0.987).

Fig. 3
figure 3

Restricted cubic spline curves illustrating the relationship between serum bicarbonate levels and the risk of postoperative delirium. Panels A, B, C, and D represent the different multivariable adjustment models: Crude Model, Model 1, Model 2, and Model 3, respectively. Each model accounts for varying sets of confounding factors

Full size image

Receiver operating characteristic curves

Through ROC curve analysis, the predictive performance of serum bicarbonate levels and severity of illness scores (SAPS II and SOFA) was assessed for the occurrence of POD in patients after cardiac surgery. It was found that the serum bicarbonate levels (AUC = 0.629, 95% CI: 0.606–0.652) surpassed both SAPS II (AUC = 0.616, 95% CI: 0.594–0.639) and SOFA (AUC = 0.589, 95% CI: 0.564–0.611) in predicting POD, demonstrating moderate predictive accuracy. This observation underscores the potential of serum bicarbonate levels as a reliable predictor of POD in patients undergoing cardiac surgery. See Fig. 4 for details.

Fig. 4
figure 4

ROC curves to assess the predictive power of bicarbonate levels on the incidence of postoperative delirium. Abbreviation: ROC, receiver operating characteristic

Full size image

Sensitivity and subgroup analysis

To confirm the robustness of our main findings concerning the relationship between serum bicarbonate levels and the risk of postoperative delirium (POD), several sensitivity analyses were conducted. The results indicated that the significant association between serum bicarbonate levels and the risk of developing POD persisted across all analyses, as detailed in Supplementary Tables 4–6.

Furthermore, in conducting subgroup analyses, multiple variables were assessed, including age, gender, ethnicity, surgical type, and various comorbidities such as hypertension, myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, renal disease, and liver disease. Analysis results demonstrated that no significant interactions were observed between these subgroups (P for interaction > 0.05), as illustrated in Fig. 5. This further confirms that the association between the risk of POD and bicarbonate levels remained consistent across subgroups.

Fig. 5
figure 5

Subgroup analyses of the relationship between bicarbonate and postoperative delirium. Bicarbonate was analyzed as a continuous variable. The above models were adjusted for the variables in model 3. In each case, the models were not adjusted for stratification variables

Full size image

Discussion

In this study, we retrospectively analyzed data from 5,422 patients who underwent cardiac surgery. Findings indicated that lower bicarbonate levels were significantly correlated with an increased risk of developing POD; the robustness of these results was further confirmed through sensitivity and subgroup analyses. Additionally, ROC curve analysis revealed moderate predictive efficacy of bicarbonate for POD in cardiac surgery patients, highlighting its significance as a predictive tool in intensive care management.

The bicarbonate concentration is an essential indicator of the body’s acid-base balance status. It is usually measured as part of routine biochemical tests to monitor renal diseases [26]. In recent years, several clinical studies have shown that bicarbonate levels are crucial not only for patients with kidney disease but also for patients with other diseases. For example, low bicarbonate concentrations are associated with increased mortality after aortic coarctation, whereas high concentrations may increase the risk of cardiovascular disease complications [27]. However, studies on the relationship between bicarbonate levels and the risk of delirium remain limited. One known randomized controlled study conducted in Thailand involving 62 surgical ICU patients revealed that patients with delirium had significantly lower bicarbonate levels than those without delirium [21], revealing for the first time a strong correlation between low bicarbonate levels and the occurrence of delirium. However, in a follow-up study, the group performed a generalized estimating equation (GEE) analysis on 65 elderly patients who underwent undergoing hip fracture surgery, and showed that bicarbonate levels were negatively correlated with psychomotor abnormalities (agitation or retardation), but were not definitively correlated a the direct diagnosis of delirium [22]. These inconsistencies might partly stem from the use of venous rather than arterial blood samples to assess acid- base status and blood gas markers. Although analyses of both venous and arterial blood gases have been shown to yield similar outcomes [28], venous blood data might not precisely mirror arterial blood criteria. Additionally, differences in study populations may also be responsible. The previous study included all surgical ICU patients, whereas the subsequent study was limited to elderly patients undergoing hip fracture surgery. Although prior studies indicated a correlation between postoperative bicarbonate concentrations and delirium in ICU patients, the conclusions were limited by small sample sizes, hindering comprehensive adjustment for all confounders in multivariate analyses. Our present study analyzed the correlation between bicarbonate levels and the risk of POD in a continuous and categorical manner by including 5,422 patients who underwent undergoing cardiac surgery and adjusting for 22 potential variables.

Our study revealed that lower bicarbonate levels significantly increased the risk of POD development. The existence and stability of this relationship were further validated by sensitivity analyses of participants who did not experience in-hospital death and did not have cerebrovascular or renal disease. Despite the decrease in mortality among patients undergoing cardiac surgery [29], postoperative neurocognitive complications and delirium remain major clinical concerns due to their frequency. The pathophysiological underpinnings of delirium are not entirely understood; however, evidence indicates metabolic acidosis’s association with delirium development, providing a potential physiological rationale for the lower bicarbonate levels in POD patients [30, 31], and offering a possible physiological explanation for the decreased bicarbonate levels in POD patients. Low bicarbonate levels, indicative of metabolic acidosis, might initiate a systemic inflammatory response [32,33,34], impairing brain function and enhancing blood- brain barrier permeability, thus increasing patient vulnerability to circulating delirium factors [35, 36]. Animal experiments have also shown that metabolic acidosis may affect the release and reuptake of neurotransmitters [37,38,39], such as dopamine and acetylcholine, and changes in these neurotransmitters have been associated with the development of delirium [40, 41]. In particular, sustained reductions in bicarbonate levels may lead to vascular endothelial dysfunction, affecting hemodynamics and brain oxygen supply, thereby impairing neurological function [42]. These findings emphasize the significance of assessing bicarbonate levels in cardiac surgery patients and suggest targeted treatments for those exhibiting low concentrations.

To date, no single medication has been shown to effectively prevent delirium. Consequently, understanding the diverse risk factors for delirium is essential for facilitating its early detection and prevention, highlighting the primary purpose and clinical significance of our research. Therefore, we initiated this study to examine the relationship between baseline bicarbonate levels and the incidence of POD in patients undergoing cardiac surgery. Our research has several strengths. First, to the best of our knowledge, this is the first study to investigate the correlation between baseline bicarbonate levels and the risk of postoperative delirium (POD) in cardiac surgery patients. Second, this study utilized a large dataset with ethnically diverse participants, enhancing the generalizability and applicability of the findings. Furthermore, incorporating a wide range of covariates in multivariate regression analyses reduced potential selection bias, thereby enhancing the accuracy of the findings. Our findings indicate that cardiac surgery patients with low bicarbonate levels at ICU admission face an increased risk of developing POD. This highlights the importance of monitoring bicarbonate levels in cardiac surgery patients to help physicians identify those at high risk for POD and develop tailored treatment strategies for patients with low bicarbonate levels. Finally, sensitivity analyses of patients who did not experience in-hospital death or who suffered from cerebrovascular or renal disease confirmed the stability and reliability of our findings.

This study also has several limitations. First, as a retrospective observational study, it unavoidably suffers from selection bias. Nonetheless, we established strict inclusion criteria to closely mirror real-world scenarios. Second, delirium can present in various subtypes, such as hyperactive, hypoactive, and mixed forms, our inability to differentiate between these subtypes. Third, the study’s reliance on a single-center sample may restrict the generalizability of the findings. Finally, our data span from 2008 to 2019, a period marked by significant changes in delirium management. Due to data anonymization in the MIMIC-IV database, we could not evaluate the specific impact of these changes on postoperative delirium. Future studies could explore the effects of various stages of delirium management on patient outcomes.

Conclusion

Our study indicates that low serum bicarbonate levels constitute an independent risk factor for the increased incidence of POD following cardiac surgery. Being a simple, cost-effective, and readily accessible biomarker, serum bicarbonate offers potential clinical applications. To corroborate this finding and further investigate the underlying mechanisms of POD and its broader implications, further in-depth studies are essential.

Data availability

Publicly available datasets were analyzed in this study. These datasets can be found at https://physionet.org/content/mimiciv/2.2/. The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Abbreviations

POD:

Postoperative delirium

MIMIC-IV:

Medical Information Mart for Intensive Care IV

ICU:

Intensive Care Unit

CABG:

Coronary artery bypass grafting

CPB:

Cardiopulmonary bypass

BMI:

Body Mass Index

AG:

Anion Gap

WBC:

White blood cell count

PO2 :

Partial pressure of oxygen

PCO2 :

Partial Pressure of Carbon Dioxide

SBP:

Systolic blood pressure

DBP:

Diastolic blood pressure

SpO2 :

Oxygen Saturation

HR:

Heart rate

RR:

Respiratory rate

SAPS II:

Simplified Acute Physiology Score II

CCI:

Charlson Comorbidity Index

SOFA:

Sequential Organ Failure Assessment

References

  1. Mattison MLP. Delirium. Annals of Internal Medicine. 2020;173:ITC49–64.

  2. Brown CH, Probert J, Healy R, Parish M, Nomura Y, Yamaguchi A, et al. Cognitive decline after Delirium in patients undergoing cardiac surgery. Anesthesiology. 2018;129:406–16.

    Article  PubMed  Google Scholar 

  3. Cai S, Li J, Gao J, Pan W, Zhang Y. Prediction models for postoperative delirium after cardiac surgery: systematic review and critical appraisal. Int J Nurs Stud. 2022;136:104340.

    Article  PubMed  Google Scholar 

  4. Habeeb-Allah A, Alshraideh JA. Delirium post-cardiac surgery: incidence and associated factors. Nurs Crit Care. 2021;26:150–5.

    Article  PubMed  Google Scholar 

  5. Girard TD, Thompson JL, Pandharipande PP, Brummel NE, Jackson JC, Patel MB, et al. Clinical phenotypes of delirium during critical illness and severity of subsequent long-term cognitive impairment: a prospective cohort study. Lancet Respir Med. 2018;6:213–22.

    Article  PubMed  PubMed Central  Google Scholar 

  6. Inouye SK, Westendorp RG, Saczynski JS. Delirium in elderly people. Lancet. 2014;383:911–22.

    Article  PubMed  Google Scholar 

  7. Huang H, Li H, Zhang X, Shi G, Xu M, Ru X, et al. Association of postoperative delirium with cognitive outcomes: a meta-analysis. J Clin Anesth. 2021;75:110496.

    Article  PubMed  Google Scholar 

  8. Lin L, Zhang X, Xu S, Peng Y, Li S, Huang X, et al. Outcomes of postoperative delirium in patients undergoing cardiac surgery: a systematic review and meta-analysis. Front Cardiovasc Med. 2022;9:884144.

    Article  PubMed  PubMed Central  Google Scholar 

  9. Hoogma DF, Venmans E, Al Tmimi L, Tournoy J, Verbrugghe P, Jacobs S, et al. Postoperative delirium and quality of life after transcatheter and surgical aortic valve replacement: a prospective observational study. J Thorac Cardiovasc Surg. 2023;166:156–e1666.

    Article  PubMed  Google Scholar 

  10. Gou RY, Hshieh TT, Marcantonio ER, Cooper Z, Jones RN, Travison TG, et al. One-year Medicare costs Associated with Delirium in older patients undergoing major elective surgery. JAMA Surg. 2021;156:462.

    Article  PubMed Central  Google Scholar 

  11. Shrimanker I, Bhattarai S. In: StatPearls, editor. Electrolytes. Treasure Island (FL): StatPearls Publishing; 2024.

    Google Scholar 

  12. Yl C, Sc H, My H, Yr L, Cl Y, Kh C et al. Effect of sodium bicarbonate on cardiovascular outcome and mortality in patients with advanced chronic kidney disease. Front Pharmacol. 2023;14.

  13. Zhu A, Whitlock RH, Ferguson TW, Nour-Mohammadi M, Komenda P, Rigatto C, et al. Metabolic acidosis is Associated with Acute kidney Injury in patients with CKD. Kidney Int Rep. 2022;7:2219–29.

    Article  PubMed  PubMed Central  Google Scholar 

  14. Fukasawa H, Kaneko M, Uchiyama Y, Yasuda H, Furuya R. Lower bicarbonate level is associated with CKD progression and all-cause mortality: a propensity score matching analysis. Bmc Nephrol. 2022;23:86.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  15. Wigger O, Bloechlinger S, Berger D, Häner J, Zanchin T, Windecker S, et al. Baseline serum bicarbonate levels independently predict short-term mortality in critically ill patients with ischaemic cardiogenic shock. Eur Heart J: Acute Cardiovasc Care. 2018;7:45–52.

    PubMed  Google Scholar 

  16. Al-Kindi A SG, Sarode, Zullo M, Rajagopalan S, Rahman M, Hostetter T, et al. Serum bicarbonate concentration and cause-specific mortality: the National Health and Nutrition Examination Survey 1999–2010. Mayo Clin Proc. 2020;95:113–23.

    Article  CAS  PubMed  Google Scholar 

  17. Li X, Mao Y, Zhu S, Ma J, Gao S, Jin X, et al. Relationship between depressive disorders and biochemical indicators in adult men and women. BMC Psychiatry. 2023;23:49.

    Article  PubMed  PubMed Central  Google Scholar 

  18. Lemon N, Canepa E, Ilies MA, Fossati S. Carbonic anhydrases as potential targets against neurovascular unit dysfunction in Alzheimer’s Disease and Stroke. Front Aging Neurosci. 2021;13:772278.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  19. Maroofian R, Zamani M, Kaiyrzhanov R, Liebmann L, Karimiani EG, Vona B, et al. Biallelic variants in SLC4A10 encoding a sodium-dependent bicarbonate transporter lead to a neurodevelopmental disorder. Genet Med. 2024;26:101034.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  20. Alka K, Casey JR. Bicarbonate transport in health and disease. IUBMB Life. 2014;66:596–615.

    Article  CAS  PubMed  Google Scholar 

  21. Potharajaroen S, Tangwongchai S, Tayjasanant T, Thawitsri T, Anderson G, Maes M. Bright light and oxygen therapies decrease delirium risk in critically ill surgical patients by targeting sleep and acid-base disturbances. Psychiatry Res. 2018;261:21–7.

    Article  PubMed  Google Scholar 

  22. Thisayakorn P, Tangwongchai S, Tantavisut S, Thipakorn Y, Sukhanonsawat S, Wongwarawipat T, et al. Immune, Blood Cell, and blood gas biomarkers of Delirium in Elderly individuals with hip fracture surgery. Dement Geriatr Cogn Disord. 2021;50:161–9.

    Article  CAS  PubMed  Google Scholar 

  23. Johnson AEW, Bulgarelli L, Shen L, Gayles A, Shammout A, Horng S, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10:1.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. The strengthening the reporting of Observational studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Int J Surg. 2014;12:1495–9.

    Article  Google Scholar 

  25. Stekhoven DJ, Bühlmann P. MissForest–non-parametric missing value imputation for mixed-type data. Bioinformatics. 2012;28:112–8.

    Article  CAS  PubMed  Google Scholar 

  26. Nagami GT, Kraut JA. Regulation of acid-base balance in patients with chronic kidney disease. Adv Chronic Kidney Dis. 2022;29:337–42.

    Article  PubMed  Google Scholar 

  27. Tan L, Xu Q, Li C, Chen X, Bai H. Association between the Admission Serum Bicarbonate and short-term and long-term mortality in Acute Aortic Dissection patients admitted to the Intensive Care Unit. Int J Gen Med. 2021;14:4183–95.

    Article  PubMed  PubMed Central  Google Scholar 

  28. Ritter C, Tomasi CD, Dal-Pizzol F, Pinto B, Dyson A, De Miranda AS, et al. Inflammation biomarkers and delirium in critically ill patients. Crit Care. 2014;18:R106.

    Article  PubMed  PubMed Central  Google Scholar 

  29. Selnes OA, Gottesman RF, Grega MA, Baumgartner WA, Zeger SL, McKhann GM. Cognitive and neurologic outcomes after coronary-artery bypass surgery. N Engl J Med. 2012;366:250–7.

    Article  CAS  PubMed  Google Scholar 

  30. Zaal IJ, Devlin JW, Peelen LM, Slooter AJC. A systematic review of risk factors for Delirium in the ICU*. Crit Care Med. 2015;43:40–7.

    Article  PubMed  Google Scholar 

  31. Liang S, Chau JPC, Lo SHS, Bai L, Yao L, Choi KC. Validation of PREdiction of DELIRium in ICu patients (PRE-DELIRIC) among patients in intensive care units: a retrospective cohort study. Nurs Crit Care. 2021;26:176–82.

    Article  PubMed  Google Scholar 

  32. Poyourow C (Nina), Byham-Gray L, Leonberg K, Chung M, editors. The role of dietary acid load on progression of estimated glomerular filtration rate among individuals diagnosed with chronic kidney disease. J Ren Nutr. 2024;:S1051227624000372.

  33. Forget M-F, Del Degan S, Leblanc J, Tannous R, Desjardins M, Durand M, et al. Delirium and inflammation in older adults hospitalized for COVID-19: a Cohort Study. Clin Interv Aging. 2021;16:1223–30.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  34. Adamczak M, Surma S. Metabolic acidosis in patients with CKD: Epidemiology, Pathogenesis, and treatment. Kidney Dis. 2021;7:452–67.

    Article  Google Scholar 

  35. Katsumi Y, Racine AM, Torrado-Carvajal A, Loggia ML, Hooker JM, Greve DN, et al. The role of inflammation after surgery for elders (RISE) study: examination of [11 C]PBR28 binding and exploration of its link to post-operative delirium. NeuroImage: Clin. 2020;27:102346.

    Article  PubMed  Google Scholar 

  36. Schreuder L, Eggen BJ, Biber K, Schoemaker RG, Laman JD, De Rooij SE. Pathophysiological and behavioral effects of systemic inflammation in aged and diseased rodents with relevance to delirium: a systematic review. Brain Behav Immun. 2017;62:362–81.

    Article  CAS  PubMed  Google Scholar 

  37. Imomnazarov K, Torrence SE, Lindgren CA. Reduced plasma-membrane calcium ATPase activity and extracellular acidification trigger presynaptic homeostatic potentiation at the mouse Neuromuscular Junction. Neuroscience. 2023;532:103–12.

    Article  CAS  PubMed  Google Scholar 

  38. Sheikh IA, Malik A, AlBasri SFM, Beg MA. In silico identification of genes involved in chronic metabolic acidosis. Life Sci. 2018;192:246–52.

    Article  CAS  PubMed  Google Scholar 

  39. Celotto AC, Ferreira LG, Capellini VK, Albuquerque AAS, Rodrigues AJ, Evora PRB. Acute but not chronic metabolic acidosis potentiates the acetylcholine-induced reduction in blood pressure: an endothelium-dependent effect. Braz J Med Biol Res. 2016;49:e5007.

    Article  CAS  PubMed  Google Scholar 

  40. Maldonado JR. Neuropathogenesis of delirium: review of current etiologic theories and common pathways. Am J Geriatr Psychiatry. 2013;21:1190–222.

    Article  PubMed  Google Scholar 

  41. Thom RP, Levy-Carrick NC, Bui M, Silbersweig D, Delirium. AJP. 2019;176:785–93.

    Article  Google Scholar 

  42. Gamberale R, D’Orlando C, Brunelli S, Meneveri R, Mazzola P, Foti G, et al. Study protocol: understanding the pathophysiologic mechanisms underlying delirium in older people undergoing hip fracture surgery. BMC Geriatr. 2021;21:633.

    Article  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

The authors extend their gratitude to all participants and staff involved with the Medical Information Mart for Intensive Care (MIMIC-IV) database for their valuable contributions.

Funding

None.

Author information

Author notes

  1. Panxu Guo and Yue Ma contributed equally to this work.

Authors and Affiliations

  1. General Practice Center, The Seventh Affiliated Hospital of Southern Medical University, Foshan, Guangdong, 528244, People’s Republic of China

    Panxu Guo & Peixi Wang

  2. School of Nursing and Health, Henan University, Kaifeng, Henan, 475004, People’s Republic of China

    Panxu Guo, Danying Xie, Xiaowan Li, Ke Wang & Peixi Wang

  3. Affiliated Hospital of Guangdong Medical University, Zhanjiang, Guangdong, 524002, People’s Republic of China

    Yue Ma

  4. Operations Management Department, Shenzhen Longhua District People’s Hospital, Shenzhen, Guangdong, 518100, People’s Republic of China

    Wanmin Su

Authors
  1. Panxu Guo
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Yue Ma
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Wanmin Su
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. Danying Xie
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. Xiaowan Li
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. Ke Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

  7. Peixi Wang
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

PG, YM, and PW formulated the research questions and designed the study. PG extracted clinical data from the MIMIC-IV database. DX, XL, and KW conducted the data analysis. PG and YM drafted the manuscript. PW and WS critically reviewed, edited, and approved the manuscript. PG and YM finalized the manuscript based on all the authors’ comments. All authors provided comments and approved the final manuscript.

Corresponding author

Correspondence to Peixi Wang.

Ethics declarations

Ethics approval and consent to participate

The database was approved for research use by the Institutional Review Boards of the Massachusetts Institute of Technology and Beth Israel Deaconess Medical Center. Studies using the database were granted a waiver of informed consent, and all methods were performed in accordance with the relevant guidelines and regulations.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Electronic supplementary material

Below is the link to the electronic supplementary material.

12871_2024_2738_MOESM1_ESM.docx

Supplementary Material 1: Supplementary File: Supplementary Fig. 1. Graph of the percentage of missing variables for patients undergoing acute cardiac surgery in the MIMIC database. Supplementary Table 1. The ICD-9 and ICD-10 codes for identifying cardiac surgeries. Supplementary Table 2. The variance inflation factor for all covariates of the fully adjusted model. Supplementary Table 3. Univariate logistic regression analysis of postoperative delirium in cardiac surgery patients. Supplementary Table 4. The association between various bicarbonate groups and the risk of postoperative delirium in patients undergoing cardiac surgery after excluding patients with renal disease (n=4, 585). Supplementary Table 5. The association between various bicarbonate groups and the risk of postoperative delirium in patients undergoing cardiac surgery after excluding patients with cerebrovascular disease (n=4, 881). Supplementary Table 6. Relationship between different bicarbonate groups and risk of postoperative delirium in patients undergoing cardiac surgery in the original data before multiple interpolation (n=4, 810).

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Guo, P., Ma, Y., Su, W. et al. Association between baseline serum bicarbonate and the risk of postoperative delirium in patients undergoing cardiac surgery in the ICU: a retrospective study from the MIMIC-IV database. BMC Anesthesiol 24, 347 (2024). https://doi.org/10.1186/s12871-024-02738-9

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12871-024-02738-9

Keywords

BMC Anesthesiology

ISSN: 1471-2253

Contact us