Abstract
Background
The relationship between serum anion gap (AG) and 28-day mortality in critically ill patients with infective endocarditis is currently not well established.
Objective
This study aims to investigate the impact of serum AG on 28-day mortality in critically ill patients with infective endocarditis.
Methods
A retrospective cohort study was conducted involving 449 participants diagnosed with infective endocarditis and admitted to intensive care units (ICU). Vital signs, laboratory parameters and comorbidity were collected for all participants to analyze the association between anion gap levels and 28-day mortality.
Results
A total of 449 critically ill patients with infective endocarditis (IE) were included in the study. The mean age was 57 years, and 64% were male. The overall 28-day mortality rate was 20%. A greater AG on admission were significantly associated with increased 28-day mortality in unadjusted analysis (hazard ratio [HR] 1.13; 95% confidence interval [CI] 1.09–1.18; p < 0.001). After adjusting for all confounders, the association remained significant (adjusted HR 1.07; 95% CI 1.02–1.13; p = 0.003). When AG was converted into categorial variables (quartiles), the risk of 28-day mortality in the greatest Q4 group was significantly higher compared with that in the lowest Q1 group (model 4: HR = 2.62, 95%CI: 1.17–5.83, p = 0.019). Subgroup analysis showed consistent results across different groups.
Conclusion
A greater AG on admission were independently associated with increased 28-day mortality in critically ill patients with IE. These findings suggest that the AG can serve as a prognostic marker in this population, aiding in risk stratification and guiding clinical management.
Keywords: Anion gap, Infective endocarditis, Critically ill patients, Mortality, MIMIC-IV
Introduction
Infective endocarditis (IE), typically caused by bacterial infections of the inner lining of the heart chambers and heart valves, can lead to serious complications such as valvular damage, stroke, and organ failure [1]. Symptoms may include fever, chills, fatigue, weakness, symptoms related to IE complications such as heart failure or embolic events (e.g., stroke). Treatment usually involves a combination of antibiotics and, in some cases, surgery to repair or replace damaged heart valves. The incidence of IE is 3–10 per 100,000 people [2–4]. Despite advancements in diagnostic tools and therapeutic measures, the in-hospital mortality rate due to IE remains at approximately 25% [5], leading to a high economic burden with average hospitalization costs ranging from $37,000 to $55,000 per patient [6]. The prognosis of patients with IE is often influenced by multiple factors, including the severity of the infection and the timeliness of early diagnosis and treatment. Various factors, including individual patient characteristics, cardiac and non-cardiac comorbidities, infecting microbial species, and echocardiographic disease status, contribute to the poor prognosis of IE [4, 7]. Early diagnosis is critical to improving patient outcomes and reducing mortality associated with IE [8]. However, research on preventive factors for IE is limited, presenting a challenge. In clinical practice, it is crucial to identify reliable predictive markers to assess disease progression and prognosis, which can guide treatment decisions.
Serum anion gap (AG), which is cheap and useful, a calculated parameter derived from electrolyte measurements, has been widely used in clinical practice as an indicator of acid-base imbalances and metabolic acidosis [9]. It reflects the balance between unmeasured anions and cations in the blood and is calculated using the formula [10]:
 |
The average serum AG in healthy individuals measured in mmol/L is 12 ± 4. Several earlier researches have demonstrated that an increase in the AG may be associated with a worse prognosis in patients with critically ill [11, 12]. Several researches have indicated that a greater AG is linked to a higher mortality in a number of serious conditions, including sepsis [13], heart failure [14], acute renal injury [15], and asthma [12]. Its correlation in patients with IE in critical condition, however, has not been thoroughly investigated. In order to investigate the relationship between AG and mortality in critically ill patients with IE, we are conducting the current study. We hypothesized that a greater AG would be associated with an increased risk of mortality in critically ill patients with IE.
Method
Database
The Medical Information Mart for Intensive Care IV (MIMIC IV) database v2.2 (https://mimic.mit.edu/) provided the data used in this study [16]. It includes details on 73,181 hospitalizations of critically ill patients who were admitted to Boston’s Beth Israel Deaconess Medical Center between 2008 and 2019 [17]. Numerous data elements are included in the database, including survival statue, vital signs, laboratory indicators, illness names, and treatment plans. Because of the MIMIC database’s abundance of high-quality data and its other benefits, an increasing number of academics are using it to conduct research [18–20]. Two of the authors, Yingxiu Huang, and Ting Ao were given access to the database following successful completion of an online course and test (Certificate ID: 56513391 Yingxiu Huang, Certificate ID: 58844105 Ting Ao). Since the database protects patients’ privacy, informed permission was not required. The present cohort study was in compliance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) [21].
We conducted a retrospective cohort study using data from MIMIC IV 2.2. We included all patients admitted to the ICU with IE for the first time. The data collected included patient demographics, comorbidities, illness severity scores, left ventricular ejection fraction (LVEF), blood culture result (Staphylococcus spp., Streptococcus spp., Enterococcus spp., HACEK organisms, culture-negative), and laboratory parameters. The illness severity was assessed using the Sequential Organ Failure Assessment (SOFA) score. The comorbidities included congestive heart failure, renal disease, chronic pulmonary disease, diabetes, renal disease, acute kidney injury within 2days, continuous renal replacement therapy (CRRT), valvular surgery, and sepsis. The calculation of sepsis was dependent on SOFA score, which was based on the scores of six different systems, including the respiratory, cardiovascular, hepatic, coagulation, renal and neurological. The endpoint of the study was 28-day mortality.
Study population
Criteria
First, we included patients who had been diagnosed with IE during their hospital admission from 2008 to 2019, and the diagnostic criteria were based on the International Classification of Diseases(ICD), ICD9 and ICD10( code = 03642, 07422, 0932X, 09884, 11281, 11504, 11514, 11594, 3911, 421X, 4249X, A3282, A3951, A5203, B3321, B376, I011, I38, I33, I330, I339 ,I39, M3211). Secondly, only the first ICU admission record was selected. Finally, we limited our analysis to individuals whose serum AG values were obtained within 24 h of their ICU admission.
Covariates
The MIMIC-IV database was queried using a structured query language to extract relevant information, which was then stored in PostgreSQL. The extracted data included patient demographics, such as age and sex, race, the SOFA score, laboratory indicators, such as the anion gap, white blood cells (WBC), hemoglobin, platelets, blood urea nitrogen (BUN), prothrombin time (PT), creatinine, lactate, and glucose, on the first day of ICU admission. The comorbidities included congestive heart failure, renal disease, chronic pulmonary disease, diabetes, renal disease, and sepsis. Sepsis is defined as a life-threatening organ dysfunction caused by a dysregulated host response to infection. Organ dysfunction is identified by an increase in the Sequential Organ Failure Assessment (SOFA) score of 2 or more points due to infection.
Congestive heart failure: also known as heart failure, the ICD code was following: 39,891, 40,201, 40,211, 40291,40401,40403, 40,411, 40,413, 40,491, 40493,4254, 4255,4257,4258,4259,428X, I099, I110, I130, I132 I255 I420 I425 I426 I427 I428 I429 I43 I501 I5020 I5021 I5022 I5023 I5030, I5031 I5032 I5033 I5040 I5041 I5042 I5043 I50810 I50811 I50812 I50813 I50814 I5084. Chronic pulmonary disease: Defined as a group of diseases that cause airflow blockage and breathing-related problems, including chronic obstructive pulmonary disease (COPD), asthma, interstitial lung disease, and chronic bronchitis, the ICD code were following: 4168, 4169, 490, 4910, 4911, 49,120, 49,121, 49,122, 4918, 4919, 4920, 4928, 49,300, 49,301, 49,302, 49,310, 49,312, 49,320, 49,321, 49,322, 49,381, 49,382 49,390, 49,391, 49,392, 4940, 4941, 4952, 4957, 4958, 4959, 496, 500, 501, 502, 5081, 5088, I2781, I2782, I2783, I279, J40, J410, J411, J42, J430, J431, J432, J438, J439 J440, J441, J449, J4520, J4521, J4522, J4530, J4531, J4540, J454, J4550, J4551, J4552, J45901, J45902, J45909, J45990, J45991, J45998, J470, J471, J479, J60, J61, J628, J632, J634, J636, J64, J662, J668, J672, J678, J679, J684, J701, J703. Renal disease: defined as chronic kidney disease (CKD), characterized by a gradual loss of kidney function over time, the ICD code was following: 40301,40311,40391,40403,40413,40492,40493,5821,5822,5824,58281,58289,5829,5830,5831,5832,5834,5836,5851,5852,5853,5854,5855,5856,5859,586,5880,I120,I1310,I1311,N032,N052,N055,N057,N181,N182,N183,N184,N185,N186,N189,N19N250,V420,V420,V451,V451,V4511,V4512,V560,V560,V561,V561,V562,V562,V5631,V568,Z4901,Z4902,Z940,Z992.
Statistical analysis
Patients were classified into four groups based on the quartile distribution of serum AG. The characteristics were summed together using descriptive statistics. Continuous variables were expressed as mean and standard deviation (SD) for normal distributions or median and interquartile range for skewed distributions. While categorical variables were presented as proportions (percentages). For the analysis of baseline characteristics, data were compared using Student-t test or the Mann–Whitney test for continuous variables and the chi-square test for categorical variables. For variables with missing data rates below 50%, we employed the K-Nearest Neighbors (KNN) imputation method [22].
In order to evaluate the independent relationship between serum AG and 28-day mortality, multivariable Cox hazards regression models were used to determine the hazard ratios (HRs) and 95% confidence intervals (CIs) for the association between serum AG /quartiles and the risk of mortality.
To create various variables adjusted models, an extended Cox model was used. The lowest quartile of serum AG was used as the reference group. Model 1: unadjusted. Model 2: Adjust for age, sex, race. Model 3: Adjust for age, sex, race, diabetes, congestive heart failure, chronic pulmonary disease, renal disease, and sepsis, acute kidney injury within 2 day, CRRT. Model 4: Adjust further for LVEF, blood culture result, surgery. The tests for trend were carried out by incorporating the median value of each quartile as a continuous variable in the models.
A subgroup analysis was conducted to ascertain whether the impact of serum AG levels on mortality differed across distinct subgroups. Furthermore, Kaplan–Meier analysis was employed to generate survival curves, which were subsequently subjected to log-rank analysis for comparison.
A receiver operating characteristic (ROC) curve was constructed and the area under the curve (AUC) was calculated in order to compare the ability of the AG and SOFA score in predicting 28-day mortality.
The statistical software packages R 3.3.2 (http://www.R-project.org, The R Foundation) and Free Statistics software versions 1.9.2 were used for all of the studies [23]. A two-tailed test with a significance level of p < 0.05 deemed statistically significant.
Results
Participants characteristics
After a rigorous screening process based on the inclusion and exclusion criteria, a total of 449 patients were selected from the initial 453 critically ill patients with IE in the MIMIC IV database. The 4 patients were excluded due to missing data (refer to Fig. 1 for specifics). The final analysis included 449 patients, with 64% being male and a median age of 57 years. Among these patients, the 28-day mortality rate was 20%. Detailed baseline characteristics of all participants are presented in Table 1.
Fig. 1.
Flowchart of the study
Table 1.
Baseline characteristics of subjects
| Variables | Total (n = 449) |
Anion gap(mmol/L) | p | |||
|---|---|---|---|---|---|---|
| Q1(≤12) (n = 99) |
Q2(13–14) (n = 109) |
Q3(15–17) (n = 115) |
Q4( ≥18) (n = 126) |
|||
| Sex (Male), n (%) | 288 (64.1) | 67 (67.7) | 76 (69.7) | 69 (60) | 76 (60.3) | 0.299 |
| Age, years | 57.4 ± 18.1 | 54.6 ± 18.5 | 56.0 ± 18.3 | 58.7 ± 18.0 | 59.4 ± 17.4 | 0.163 |
| Race (White), n (%) | 307 (68.4) | 75 (75.8) | 82 (75.2) | 75 (65.2) | 75 (59.5) | 0.019 |
| Heart rate (bpm) | 108.5 ± 21.7 | 101.7 ± 19.5 | 106.1 ± 19.6 | 111.0 ± 23.0 | 113.4 ± 22.4 | < 0.001 |
| Mean blood pressure (mmhg) | 75.5 ± 10.4 | 74.7 ± 8.9 | 77.6 ± 10.3 | 76.8 ± 10.3 | 73.1 ± 11.2 | 0.003 |
| Respirotary rate (bpm) | 31.3 ± 7.7 | 30.6 ± 7.6 | 30.5 ± 6.8 | 32.9 ± 8.7 | 30.9 ± 7.2 | 0.058 |
| Temperature (°C) | 37.7 ± 0.9 | 37.6 ± 0.9 | 37.7 ± 0.9 | 37.7 ± 0.9 | 37.8 ± 1.0 | 0.532 |
| SPO2 (%) | 91.5 ± 6.8 | 92.2 ± 4.2 | 92.5 ± 3.7 | 92.0 ± 4.4 | 89.7 ± 11.0 | 0.006 |
| Hemoglobin (g/L) | 9.2 ± 1.9 | 9.0 ± 1.8 | 9.3 ± 1.8 | 9.2 ± 2.0 | 9.1 ± 2.0 | 0.682 |
| WBC (109/L) | 14.9 (10.6, 21.9) | 13.7 (9.2, 18.9) | 14.4 (10.0, 20.2) | 15.5 (10.7, 21.5) | 16.1 (12.5, 24.3) | 0.05 |
| Anion gap (mmol/L) | 15.7 ± 4.5 | 10.9 ± 1.4 | 13.5 ± 0.5 | 15.9 ± 0.9 | 21.4 ± 4.0 | < 0.001 |
| Platelets (109/L) | 157.0 (100.0, 236.0) | 171.0 (119.0, 232.0) | 175.0 (111.0, 244.0) | 162.0 (99.5, 267.5) | 137.0 (69.5, 201.8) | 0.01 |
| Creatinine (mg/dL) | 1.1 (0.8, 1.8) | 0.9 (0.7, 1.2) | 0.9 (0.8, 1.2) | 1.2 (0.8, 1.7) | 2.0 (1.2, 4.7) | < 0.001 |
| PT (s) | 18.7 ± 9.6 | 17.5 ± 6.1 | 18.3 ± 7.9 | 18.7 ± 11.0 | 20.0 ± 11.6 | 0.254 |
| Lactate (mmol/L) | 2.1 (1.4, 3.0) | 2.0 (1.3, 2.5) | 2.0 (1.4, 2.8) | 1.9 (1.4, 2.9) | 2.4 (1.5, 3.9) | 0.06 |
| SOFA score | 5.8 ± 3.9 | 4.5 ± 2.9 | 4.8 ± 3.2 | 5.6 ± 3.9 | 7.7 ± 4.5 | < 0.001 |
| Sodium (mEq/L) | 136.8 ± 5.3 | 137.9 ± 4.7 | 136.8 ± 5.2 | 137.5 ± 5.0 | 135.3 ± 5.9 | 0.001 |
| Potassium (mEq/L) | 4.2 ± 0.8 | 4.2 ± 0.6 | 4.1 ± 0.6 | 4.2 ± 0.7 | 4.3 ± 1.0 | 0.108 |
| Congestive heart failure, n (%) | 192 (42.8) | 32 (32.3) | 48 (44) | 53 (46.1) | 59 (46.8) | 0.119 |
| Chronic pulmonary disease, n (%) | 103 (22.9) | 22 (22.2) | 24 (22) | 31 (27) | 26 (20.6) | 0.68 |
| Diabetes, n (%) | 121 (26.9) | 27 (27.3) | 24 (22) | 28 (24.3) | 42 (33.3) | 0.226 |
| Renal disease, n (%) | 89 (19.8) | 11 (11.1) | 15 (13.8) | 22 (19.1) | 41 (32.5) | < 0.001 |
| Sepsis, n (%) | 318 (70.8) | 59 (59.6) | 74 (67.9) | 81 (70.4) | 104 (82.5) | 0.002 |
| LVEF (%) | 52.1 ± 9.2 | 51.7 ± 8.3 | 54.9 ± 7.5 | 51.3 ± 11.4 | 50.7 ± 8.8 | 0.043 |
| Acute kidney injury 2 day, n (%) | 273 (60.8) | 48 (48.5) | 60 (55) | 68 (59.1) | 97 (77) | < 0.001 |
| CRRT | 35 ( 7.8) | 2 (2) | 2 (1.8) | 7 (6.1) | 24 (19) | < 0.001 |
| Surgery, n (%) | 24 ( 5.3) | 3 (3) | 9 (8.3) | 9 (7.8) | 3 (2.4) | 0.091 |
| Blood culture, n (%) | 0.063 | |||||
| Negative | 209 (47.6) | 50 (51) | 48 (44.4) | 61 (54) | 50 (41.7) | |
| Staphylococcus spp | 151 (34.4) | 36 (36.7) | 41 (38) | 30 (26.5) | 44 (36.7) | |
| Streptococcus spp | 23 ( 5.2) | 4 (4.1) | 10 (9.3) | 5 (4.4) | 4 (3.3) | |
| Enterococcus spp | 25 ( 5.7) | 5 (5.1) | 3 (2.8) | 6 (5.3) | 11 (9.2) | |
| HACEK | 2 ( 0.5) | 1 (1) | 1 (0.9) | 0 (0) | 0 (0) | |
| Double infection | 29 ( 6.6) | 2 (2) | 5 (4.6) | 11 (9.7) | 11 (9.2) | |
Abbreviations: PT, prothrombin time; WBC, white blood cell; LVEF, left ventricular ejection fraction; CRRT, continuous renal replacement therapy; HACEK, Haemophilus, Aggregatibacter, Cardiobacterium, Eikenella, Kingella; spp, species
Relationship between serum anion gap and mortality
We performed four models (Model 1: unadjusted, Model 2: Adjust for age, sex, race, Model 3: Adjust for age, sex, race, diabetes, congestive heart failure, chronic pulmonary disease, renal disease and sepsis, Model 4: Adjust for all potential confounders) for investigating the association of serum AG with 28-day mortality in patients with infective endocarditis (Table 2). In model 4, there was still a significant association between AG and 28-day mortality (HR = 1.07, 95%CI: 1.10–1.13, p = 0.003). Then, when AG was converted into categorial variables (quartiles), the risk of 28-day mortality in Q4 group was significantly higher compared with that in Q1 group (model 4: HR = 2.62, 95%CI: 1.17–5.83, p = 0.019). The values of P for trend in all four models were both < 0.05.
Table 2.
Association between anion gap (AG) and 28-day mortality risk in different models
| Model1 HR(95%CIs) |
P | Model2 HR(95%CIs) |
P | Model3 HR(95%CIs) |
P | Model4 HR(95%CIs) |
P | |
|---|---|---|---|---|---|---|---|---|
| AG (mmol/L) | 1.13 (1.09 ~ 1.18) | < 0.001 | 1.13 (1.09 ~ 1.18) | < 0.001 | 1.12 (1.07 ~ 1.17) | < 0.001 | 1.08 (1.04 ~ 1.13) | < 0.001 |
| AG quartile group(mmol/L) | ||||||||
| Q1(≤ 12) | 1(Ref) | 1(Ref) | 1(Ref) | 1(Ref) | ||||
| Q2(13–14) | 2.11 (0.92 ~ 4.86) | 0.079 | 2.06 (0.89 ~ 4.73) | 0.09 | 1.73 (0.73 ~ 4.11) | 0.215 | 1.46 (0.86 ~ 2.48) | 0.165 |
| Q3(15–17) | 2.65 (1.19 ~ 5.93) | 0.017 | 2.46 (1.1 ~ 5.51) | 0.029 | 2.25 (0.95 ~ 5.34) | 0.066 | 1.75 (1.07 ~ 2.84) | 0.025 |
| Q4(≥ 18) | 4.78 (2.24 ~ 10.2) | < 0.001 | 4.28 (2 ~ 9.17) | < 0.001 | 3.60 (1.60 ~ 8.07) | 0.002 | 2.59 (1.69 ~ 3.99) | < 0.001 |
| P for trend | < 0.001 | < 0.001 | 0.001 | 0.004 |
Model 1: unadjusted
Model 2: Adjust for age, sex, race
Model 3: Adjust for age, sex, race, diabetes, sepsis, congestive heart failure, cerebrovascular disease, chronic pulmonary disease, rheumatic disease, acute kidney injury, CRRT
Model 4: Adjust for age, sex, race, diabetes, sepsis, congestive heart failure, cerebrovascular disease, chronic pulmonary disease, rheumatic disease, acute kidney injury, CRRT, LVEF, surgery, blood culture, SOFA score
Abbreviations: AG, anion gap; Q, quartile; HR, hazard ratio; CI, confidence interval; Ref, reference; LVEF, left ventricular ejection fraction
Subgroup analyses
To further examine the robustness of the relationship between serum AG levels and mortality, we conducted a subgroup analysis based on gender, age, race, SOFA score, diabetes, sepsis and congestive heart failure, which is shown in the forest plot in Fig. 2. The results indicated that the association between serum AG and mortality were consistent across patients with different genders, age and comorbidities. The interaction analysis revealed that age had interaction with AG in predicting mortality (P < 0.05).
Fig. 2.
Subgroup analyses for the association of serum AG with 28-day mortality in the critically ill patients with infective endocarditis
Kaplan-Meier survival curve analysis
The Kaplan-Meier curve showed that the 28-day cumulative survival rates were lower in Q4 group than other groups (p < 0.0001) (Fig. 3).
Fig. 3.
Kaplan–Meier survival curves for critically ill patients with infective endocarditis based on serum AG quartile
ROC curve analysis
A comparison of the ROC curves between AG and SOFA score revealed a similar result. The ability of AG to predict 28-day ICU mortality was similar to SOFA score (AUCs of 0.662 [95% CI, 0.599–0.724] and 0.699 [95% CI: 0.640–0.759], respectively) (Fig. 4). The cut-off value for AG was 16.5, whereas the cut-off value for the SOFA score was 6.5 (Table 3).
Fig. 4.
Comparison of receiver operating characteristic curves between AG and SOFA score for predicting 28-day mortality
Table 3.
ROC analysis of AG and SOFA
| AG | SOFA score | |
|---|---|---|
| Cut-off | 16.5 | 6.5 |
| sensitivity | 0.589 | 0.622 |
| specificity | 0.694 | 0.691 |
| positive predictive value | 0.326 | 0.335 |
| negative predictive value | 0.871 | 0.879 |
| positive likelihood ratio | 1.930 | 2.025 |
| negative likelihood ratio | 0.592 | 0.545 |
Abbreviations: AG, anion gap
Discussion
The retrospective cohort study has revealed a significant association between elevated serum AG levels and increased mortality risk among critically ill patients with IE admitted to the ICU. The results of the multivariate Cox hazard regression analysis indicate that, even after adjusting for confounding variables, greater serum AG levels were independently linked to a higher 28-day mortality rate in critically ill patients with IE. Furthermore, the greatest AG group had a 2.62-fold increased risk of all-cause mortality compared to the lowest AG group. It’s worth noting that this study is the first of its kind to investigate the relationship between serum AG levels and infective endocarditis.
Serum AG is a cheap and useful method for differentiating between distinct acid-base imbalances and metabolic acidosis. Both serum and plasma electrolytes can be used to compute it; serum values are more commonly employed and show the variations in the amounts of unmeasured anions and cations. In healthy individuals, the average serum AG (measured in mmol/L) is 12 ± 4. Serum anion gap can now be obtained more easily thanks to the development of automated analyzers that make electrolyte testing possible in large cohorts of patients, especially those who were critically ill [24]. Thus, an increase in the anion gap may have prognostic importance in a number of diseases in critically ill patients. According to several prior research, serum AG was positively associated with outcomes of other critical illness like COPD [25], covid-19 [26], asthma [27], acute kidney injury [28], sepsis [13], and cerebral infarction [29], acute pancreatitis [30]. The evidence including the current study indicates that serum AG is a more general and reliable risk factor for critically ill diseases [24], suggesting the clinical implications of this easily obtained predictor for prognosis. A prospective longitudinal cohort study of 500 critically ill patients in ICU showed that a significant association between AG with higher mortality rates and longer hospital stays [31]. Li et al. [11] also reported that initial serum AG ≥ 16 mmol/L after ICU admission was associated with increased mortality in critically ill patients from a large multicenter Cohort Study. Our study demonstrated that a greater AG was associated with an increased risk of 28-day mortality in critically ill patients with IE, which was consistent with the above studies, emphasizing the importance of clinical attention and intervention to reduce mortality in this population. In our analysis, we observed a significant difference of mortality between different AG group. And this difference is statistically significant, underscoring the robustness of our findings.
Patients with severe IE frequently have elevated AG levels, which can happen for a number of reasons. First, metabolic acidosis may lead to an increase in serum anion gap. In metabolic acidosis, an excess of lactate is produced in the body, leading to an increase in anions in the blood, thereby increasing the anion gap [32]. Second, the accumulation of other acids in the extracellular fluid, such as beta-hydroxybutyric acid and acetoacetic acid, can also lead to increased AG levels. Third, patients with severe IE are at risk of acute kidney injury. Impaired renal function can result in decreased excretion of unmeasured anions, which may also contribute to the elevation of AG [33]. Therefore, renal dysfunction might also underlie the association of AG with mortality [28].
There are some strengths of our study. First, the study has a relatively large sample size, and the data is from the MIMIC-IV database, which is a large-scale, real-world study with high-quality data. Second, the study conducted subgroup analysis and sensitivity analysis, providing further evidence of the results’ robustness. However, our study has certain limitations. First of all, selection bias exists in the analysis because it was a retrospective analysis. Second, we only extracted the anion gap data for patients based on admission to the intensive care unit. As a result, we were unable to identify the trends in the anion gap’s alterations, which could have affected the accuracy of the results. Third, we were unable to calibrate all of the relevant confounders because data on prognostic factors, like lactate, β-hydroxybutyrate, and acetoacetate, was missing for some individuals in the database. Lastly, because this was an observational study, we were unable to verify the hypothesized mechanism of action linking a greater AG to IE severity and prognosis.
Conclusion
This study demonstrates that greater serum AG on admission is independently associated with increased 28-day mortality in critically ill patients with IE. These findings highlight the potential utility of the AG as a prognostic marker in patients with AE, aiding in risk stratification and guiding clinical management. Further research is warranted to validate these findings and explore the underlying mechanisms.
Acknowledgements
The authors thanks Dr. Qilin Yang of Department of Critical Care, The Second Affiliated Hospital of Guangzhou Medical University and the Physician Scientist Team for guidance on data extraction and analysis.
Author contributions
Yingxiu Huang: Designed the protocol, write the manuscript. Ting Ao: Extracted, collected and analyzed data. Peng Zhen: Prepared tables and figures. Ming Hu: Designed the protocol, write the manuscript, reviewed and edit the manuscript.
Funding
Not applicable.
Data availability
Data in the article can be obtained from mimic-IV database (https://mimic.physionet.org/).
Declarations
Ethics approval and consent to participate
Two authors (Yingxiu Huang and Ting Ao) have passed the “Protecting Human Research Participants” examination, could access the database and was responsible for data extraction (Certificate ID: 56513391 Yingxiu Huang, Certificate ID: 58844105 Ting Ao). MIMIC-IV database used in the present study was approved by the Institutional Review Boards (IRB) of Institutional Review Boards of Beth Israel Deaconess Medical Center(2001-P-001699/14) and the Massachusetts Institute of Technology (No. 0403000206) both approved the use of the database for research. The individual information of the patients included in this database was anonymous, and ethical review and informed consent were waived. We have also complied with all relevant ethical regulations regarding the use of the data for our study.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Cahill TJ, Prendergast BD. Infective endocarditis. Lancet. 2016;387:882–93. [DOI] [PubMed] [Google Scholar]
- 2.Malakan Rad, E. et al. Infective endocarditis in north africa and the middle east, 1990‒2019: Updates from the global burden of disease study 2019. Arch. Iran. Med. 2024;27:229–238. [DOI] [PMC free article] [PubMed]
- 3.Ambrosioni J, et al. The changing epidemiology of infective endocarditis in the twenty-first century. Curr Infect Dis Rep. 2017;19:21. [DOI] [PubMed] [Google Scholar]
- 4.Che D, Hu J, Zhu J, Lyu J, Zhang X. Development and validation of a nomogram for predicting in-hospital mortality in ICU patients with infective endocarditis. BMC Med Inf Decis Mak. 2024;24:84. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Chen H, et al. The global, regional, and national burden and trends of infective endocarditis from 1990 to 2019: results from the global burden of disease study 2019. Front Med. 2022;9:774224. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Alkhouli M, Alqahtani F, Alhajji M, O Berzingi C, R Sohail M. Clinical and economic burden of hospitalizations for infective endocarditis in the United States. Mayo Clin Proc. 2020;95:858–66. [DOI] [PubMed] [Google Scholar]
- 7.Habib, G. et al. [2015 ESC guidelines for the management of infective endocarditis]. Kardiol. Pol. 2015;73:963–1027. [DOI] [PubMed]
- 8.Baumgartner H, et al. 2017 ESC/EACTS guidelines for the management of valvular heart disease. Eur Heart J. 2017;38:2739–91. [DOI] [PubMed] [Google Scholar]
- 9.Kraut JA, Madias NE. Serum anion gap: its uses and limitations in clinical medicine. Clin J Am Soc Nephrol. 2007;2:162–74. [DOI] [PubMed] [Google Scholar]
- 10.Jung B, et al. Diagnosis and management of metabolic acidosis: guidelines from a French expert panel. Ann Intensive Care. 2019;9:92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Li R et al. Relationship of admission serum anion gap and prognosis of critically ill patients: A large multicenter cohort study. Dis. Markers 2022, 5926049 (2022). [DOI] [PMC free article] [PubMed]
- 12.Xu C, et al. Serum anion gap is associated with risk of all-cause mortality in critically ill patients with acute myocardial infarction. Int J Gen Med. 2022;15:223–31. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Zhu Y, et al. Serum anion gap level predicts all-cause mortality in septic patients: a retrospective study based on the MIMIC III database. J Intensive Care Med. 2023;38:349–57. [DOI] [PubMed] [Google Scholar]
- 14.Huang Z, Wang S, Yang S. Association between serum anion gap and risk of in-hospital mortality in patients with acute heart failure. Sci Rep. 2024;14:4858. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Cheng B, Li D, Gong Y, Ying B, Wang B. Serum Anion Gap Predicts All-Cause Mortality in Critically Ill Patients with Acute Kidney Injury: Analysis of the MIMIC-III Database. Dis. Markers 2020, 6501272 (2020). [DOI] [PMC free article] [PubMed]
- 16.Aew J et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data 10, (2023). [DOI] [PMC free article] [PubMed]
- 17.Johnson AEW, et al. MIMIC-III, a freely accessible critical care database. Sci Data. 2016;3:160035. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hu Y, Zhao Y, Zhang J, Li C. The association between triglyceride glucose-body mass index and all-cause mortality in critically ill patients with atrial fibrillation: A retrospective study from MIMIC-IV database. [DOI] [PMC free article] [PubMed]
- 19.Wang H, Gao C, Dantona C, Hull B, Sun J. DRG-LLaMA: tuning LLaMA model to predict diagnosis-related group for hospitalized patients. NPJ Digit Med. 2024;7:16. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Liu K, et al. Association between the ROX index and mortality in patients with acute hypoxemic respiratory failure: a retrospective cohort study. Respir Res. 2024;25:143. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Field N, et al. Strengthening the reporting of molecular epidemiology for infectious diseases (STROME-ID): an extension of the STROBE statement. Lancet Infect Dis. 2014;14:341–52. [DOI] [PubMed] [Google Scholar]
- 22.El-Menyar A, Naduvilekandy M, Asim M, Rizoli S, Al-Thani H. Machine learning models predict triage levels, massive transfusion protocol activation, and mortality in trauma utilizing patients hemodynamics on admission. Comput Biol Med. 2024;179:108880. [DOI] [PubMed] [Google Scholar]
- 23.Yang Q, et al. Association between wait time of central venous pressure measurement and outcomes in critical patients with acute kidney injury: a retrospective cohort study. Front Public Health. 2022;10:893683. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Glasmacher SA, Stones W. Anion gap as a prognostic tool for risk stratification in critically ill patients - a systematic review and meta-analysis. BMC Anesthesiol. 2016;16:68. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Chen, X. et al. Association between serum anion gap and mortality in critically ill patients with COPD in ICU: Data from the MIMIC IV database. Int. J. Chron. Obstruct. Pulmon. Dis. 2024;19:579–587. [DOI] [PMC free article] [PubMed]
- 26.Zemlin AE, et al. The association between acid-base status and clinical outcome in critically ill COVID-19 patients admitted to intensive care unit with an emphasis on high anion gap metabolic acidosis. Ann Clin Biochem. 2023;60:86–91. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Wang S, Teng H, Han H, Xu Y. The relationship between albumin corrected anion gap levels and mortality in patients with asthma in the ICU. Sci Rep. 2023;13:16903. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Zhong L, Xie B, Ji X-W, Yang X-H. The association between albumin corrected anion gap and ICU mortality in acute kidney injury patients requiring continuous renal replacement therapy. Intern Emerg Med. 2022;17:2315–22. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Liu X, et al. Serum anion gap at admission predicts all-cause mortality in critically ill patients with cerebral infarction: evidence from the MIMIC-III database. Biomark Biochem Indic Expo Response Susceptibility Chem. 2020;25:725–32. [DOI] [PubMed] [Google Scholar]
- 30.Gong F, Zhou Q, Gui C, Huang S, Qin Z. The relationship between the serum anion gap and all-cause mortality in acute pancreatitis: an analysis of the MIMIC-III database. Int J Gen Med. 2021;14:531–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Dafal A, Kumar S, Agrawal S, Acharya S, Nirmal A. Admission anion gap metabolic acidosis and its impact on patients in medical intensive care unit. J Lab Physicians. 2021;13:107–11. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Cavallazzi R. Ramirez JA. Influenza and viral pneumonia. Clin. Chest Med. 2018;39:703–21. [DOI] [PMC free article] [PubMed]
- 33.Gao P, Min J, Zhong L, Shao M. Association between albumin corrected anion gap and all-cause mortality in critically ill patients with acute kidney injury: a retrospective study based on MIMIC-IV database. Ren Fail. 2023;45:2282708. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
Data in the article can be obtained from mimic-IV database (https://mimic.physionet.org/).