Abstract
Background
Respiratory rate (RR) is one important vital sign that is often neglected in patients with acute myocardial infarction (AMI). The present study aimed to evaluate the impact of admission RR on the prognosis in patients with AMI.
Methods
This study included 5631 AMI patients from Medical Information Mart for Intensive Care IV (MIMIC-IV) and 1323 patients from The First Affiliated Hospital of Chongqing Medical University (validation cohort). The primary endpoint was 30-day all-cause mortality. Patients were divided into increased RR group (RR > 20 breaths/min) and normal RR group (RR ≤ 20 breaths/min) based on admission RR.
Results
In the MIMIC-IV cohort, the median age was 70 years and 3503 (62.2%) patients were male. Compared to patients with normal RR, patients with admission RR > 20 breaths/min had significantly increased risk of 30-day all-cause mortality (14.2% vs. 27.5%, p < 0.001). After multivariate adjustment, admission RR > 20 breaths/min was independently associated with increased risk of 30-day all-cause mortality (HR 1.715; 95%CI 1.507–1.952, p < 0.001). The findings from MIMIC-IV were validated in the data from The First Affiliated Hospital of Chongqing Medical University and the results were consistent.
Conclusion
RR is not only a vital sign but also a simple and practical indicator for predicting the prognosis of patients with AMI. Increased RR upon admission independently predicted a higher risk of poor outcomes, making it a valuable tool for early risk stratification and guiding timely interventions.
Supplementary Information
The online version contains supplementary material available at 10.1186/s12872-025-04729-1.
Keywords: Respiratory rate, Acute myocardial infarction, Vital signs, 30-day all-cause mortality
Introduction
Acute Myocardial Infarction (AMI) also known as “heart attack”, is a condition characterized by a decreased or complete cessation of blood flow to a portion of the myocardium, leading to myocardial cell death and necrosis [1]. Despite substantial advances in medical and reperfusion therapy, AMI remains one of the leading causes of death representing 13% of the total deaths in 2021 [2], and more than four out of five cardiovascular deaths are caused by AMI [3].
Currently, there are several tools available in clinical practice for risk stratification for AMI [4–8]. However, due to the significant heterogeneity in the clinical characteristics of myocardial infarction patients, these risk stratification tools have their limitations in the daily practice. Additionally, the complex algorithms required by some of these tools restrict their clinical application. Therefore, identification of novel markers for risk stratification would be helpful in better guiding clinical treatment and prognosis assessment.
Blood pressure (BP), heart rate (HR), and respiratory rate (RR) are important parameters of vital signs and previous studies have demonstrated that HR [9] and BP [10, 11] are significantly associated with the prognosis in patients with AMI, while the impact of RR on the outcome in patients with AMI is not well understood. In fact, RR often changes in patients with cardiovascular disease and has been linked with the outcome of acute ischemic stroke [12, 13], heart failure [14, 15], and cardiac arrest [16, 17]. In the case of AMI, particularly when complicated by heart failure or cardiogenic shock (CS), systemic hypoxia and pulmonary congestion typically lead to an increased RR. Additionally, the activation of the sympathetic nervous system following AMI can contribute to an elevated RR. Despite these associations, there is limited research exploring the role of RR changes in predicting the prognosis of AMI patients. Considering the close relationship between RR regulation and AMI, we hypothesized that an increased RR in AMI patients would be associated with a higher risk of poor prognosis. We tested this hypothesis using data from The Medical Information Mart for Intensive Care IV Database (MIMIC-IV version 2.2) and validated the findings in an institutional cohort.
Methods
Data source and ethics
The present study utilized a publicly and freely accessible clinical database that contains critical care data of 73,181 patients of Beth Israel Deaconess Medical Centre in Boston from 2008 to 2019 [18]. The MIMIC-IV version 2.2 contained charted events, such as demographic characteristics, vital signs, laboratory data, discharge summaries, procedures, and drug data. Diagnostic information was obtained based on the International Classification of Diseases, Ninth and Tenth Revision (ICD- 9 and ICD- 10).
The validation cohort was the patients diagnosed with AMI from December 2015 to December 2018 in The First Affiliated Hospital of Chongqing Medical University (Chongqing, China). Data including demographic characteristics, vital signs, laboratory parameters, treatments, and procedures were obtained from the hospital’s medical record system.
The establishment of the database was approved by the institutional review boards of the Massachusetts Institute of Technology in Cambridge, Massachusetts, the Beth Israel Deaconess Medical Centre in Boston, Massachusetts, and The First Affiliated Hospital of Chongqing Medical University (Chongqing, China). The Collaborative Institutional Training Initiative Examination (Certification number: 57772730) has been completed by author B.R.S., who has been granted permission to extract data from the database. In addition, we are aware of and abide by all applicable rules and guidelines related to research ethics and the use of data in our study (Declaration of Helsinki).
Study population
Adults with AMI as defined by the ICD- 9 codes (code = 410) and ICD- 10 codes (code = I21 and I22) were extracted from the MIMIC-IV database. The inclusion criteria of the study were: 1) Patients aged between 18 and 91 years old diagnosed with AMI; 2) Patients who were admitted to the intensive care unit (ICU) for more than 24 h; 3) First hospital and first admission. Patients younger than 18 years old, who had more than one admission, had incomplete information regarding the study outcomes, had missing data on admission RR, and stayed less than 24 h in the ICU were excluded from this study.
Patients in the validation cohort were diagnosed with AMI and included in the study if presented with the following features: chest pain or equivalent symptoms, dynamic electrocardiographic changes, elevated cardiac serum biomarkers, and first coronary care unit (CCU) admission. Patients without RR data (n = 12) were excluded. Follow-ups were conducted through the electronic medical record system and/or telephone interviews (Fig. 1).
Fig. 1.
Flowchart of the study
Data extraction
Structured Query Language (SQL) was used to extract clinical data, laboratory data, demographic data, diagnostic information, procedures, and medications of the MIMIC-IV database. Meanwhile, clinical data of the validation cohort was extracted from the hospital’s electronic medical record system. Demographic data included age, weight, height, body mass index (BMI), sex, smoking status, and family history. Laboratory and clinical data included vital signs, hospital and out of hospital death, coexisting disorders, leukocyte, neutrophil, hematocrit, creatinine kinase MB (CK-MB), partial arterial pressure of oxygen (PaO2), partial arterial pressure of carbon dioxide (PaCO2), cardiac troponin T, N-terminal pro-B-type natriuretic peptide (NT-proBNP), B-type natriuretic peptide (BNP), alanine transaminase (ALT), aspartate transaminase (AST), hemoglobin, glucose, international normalized ratio (INR), triglyceride, total cholesterol, low-density lipoprotein cholesterol (LDL-C), high-density lipoprotein cholesterol (HDL-C), albumin, lactate, treatments, and procedures. The laboratory data were extracted within the first 24 h after ICU or CCU admission.
Outcomes
The primary endpoint was 30-day all-cause mortality.
Statistical analysis
Continuous variables with normal distribution were presented as mean and standard deviation (SD), while non-normally or skewed continuous variables were analyzed using the Mann-Whittney U test or the Kruskal–Wallis test as appropriate and presented as the median and interquartile range (25%—75%). Categorical variables were expressed as the number of cases and percentages and were analyzed using Pearson or Fisher’s exact test as appropriate. Patients were subsequently divided into two groups based on the normal physiological range of RR in adults: the increased RR group (RR > 20 breaths/min) and the normal RR group (RR ≤ 20 breaths/min). Thereafter, the baseline characteristics between the two groups were compared.
Kaplan–Meier (K-M) survival curves were performed to compare the 30-day survival rates between the two groups. Restricted Cubic Spline (RCS) models were performed to examine the association between admission RR and the endpoint. Cox regression analyses were conducted to analyze the independent correlation between admission RR and outcome. The multivariate Cox regression analysis was adjusted based on variables that were clinically relevant or had p-values below 0.05 in the univariate analysis. A forward stepwise approach was used to apply the Cox proportional model, with a p < 0.05 entry criteria. The computed values were the adjusted hazard ratio (HR) and its corresponding 95% confidence interval (CI). A two-sided p-value of less than 0.05 was regarded as statistical significance. An HR > 1.0 and a p-value < 0.05 suggested a deleterious association, while an HR < 1.0 and a p-value < 0.05 demonstrated a protective association. All statistical analyses were carried out using the SPSS statistical software, version 25.0 (IBM, Armonk, NY, USA), GraphPad Prism 8.4.3 (GraphPad Software, Inc., San Diego, CA, USA), and R (version 3.6.3, R foundation for statistical computing, Vienna, Austria).
Results
Baseline characteristics in MIMIC-IV
In the MIMIC-IV database, a total of 5631 patients with first admission to the ICU and diagnosed with AMI were included. The median age of this study population was 70 years and 3503 (62.2%) were male. The mean admission vital signs of this cohort were BP 116/64 mmHg, HR 84 beats per minute, and RR 18 breaths/min. Common comorbidities in this cohort included hypertension (39.9%), atrial fibrillation (34.9%), hypertensive heart failure (48.0%), and renal failure (42.1%). After admission to ICU, the most common procedure was coronary artery bypass grafting (CABG) (22.7%) and ventilator (22.7%), followed by dilatation of coronary artery (DCA) (8.1%), IABP (3.2%), cardiopulmonary resuscitation (CPR) (1.9%), and extracorporeal membrane oxygenation (ECMO) (0.4%) (Table 1).
Table 1.
A comparison of baseline characteristics of the 2 groups in MIMIC-IV Database
| Total n = 5631 | RR ≤ 20 breaths/min n = 3652 | RR > 20 breaths/min n = 1979 | P value | |
|---|---|---|---|---|
| Age (years) | 70 (61–79) | 70 (60–78) | 72 (62–82) | < 0.001 |
| Male (%) | 3503 (62.2) | 2345 (64.2) | 1158 (58.5) | < 0.001 |
| Smoking (%) | 1037 (18.4) | 737 (20.2) | 300 (15.2) | < 0.001 |
| BMI (kg/m2) | 28.65 ± 6.48 | 28.71 ± 6.27 | 28.53 ± 6.89 | 0.065 |
| Vital Signs | ||||
| Heart Rate (b.p.m) | 84.0 (74.0–97.0) | 81.0 (72.0–91.0) | 92.0 (79.0–107.0) | < 0.001 |
| SBP (mmHg) | 116.0 (98.0–134.0) | 115.0 (96.0–132.0) | 119.0 (101.0–138.0) | < 0.001 |
| DBP (mmHg) | 64.0 (52.0–78.0) | 63.0 (50.0–76.0) | 66.0 (54.0–81.0) | < 0.001 |
| RR (breaths/min) | 18.0 (15.0–23.0) | 16.0 (14.0–18.0) | 24.0 (22.0–28.0) | < 0.001 |
| Coexisting Disorders (%) | ||||
| Atrial Fibrillation | 1963 (34.9) | 1224 (33.5) | 739 (37.3) | 0.004 |
| Primary Hypertension | 2244 (39.9) | 1573 (43.1) | 671 (33.9) | < 0.001 |
| Type 2 Diabetes Mellitus | 1534 (27.2) | 981 (26.9) | 553 (27.9) | 0.384 |
| Cardiogenic Shock | 785 (13.9) | 429 (11.7) | 356 (18.0) | < 0.001 |
| Chronic Kidney Disease | 1418 (25.2) | 848 (23.2) | 570 (28.8) | < 0.001 |
| COPD | 461 (8.2) | 247 (6.8) | 214 (10.8) | < 0.001 |
| Dilated Cardiomyopathy | 27 (0.5) | 15 (0.4) | 12 (0.6) | 0.310 |
| Cerebral Hemorrhage | 170 (3.0) | 106 (2.9) | 64 (3.2) | 0.488 |
| Dyslipidemia | 2977 (52.9) | 2032 (55.6) | 945 (47.8) | < 0.001 |
| Old Myocardial Infarction | 749 (13.3) | 495 (13.6) | 254 (12.8) | 0.448 |
| Cerebral Infarction | 34 (0.6) | 20 (0.5) | 14 (0.7) | 0.460 |
| Acute Respiratory Failure | 1316 (23.4) | 669 (18.3) | 647 (32.7) | < 0.001 |
| Ischemic Heart Disease | 479 (8.5) | 268 (7.3) | 211 (10.7) | < 0.001 |
| Renal Failure | 2368 (42.1) | 1319 (36.1) | 1049 (53.0) | < 0.001 |
| Cardiac Arrest | 339 (6.0) | 207 (5.7) | 132 (6.7) | 0.131 |
| Hypertensive Heart Failure | 2704 (48.0) | 1565 (42.9) | 1139 (57.6) | < 0.001 |
| Pneumonia | 804 (14.3) | 415 (11.4) | 389 (19.7) | < 0.001 |
| Anemia | 1289 (22.9) | 773 (21.2) | 516 (26.1) | < 0.001 |
BMI Body mass index, SBP Systolic blood pressure, DBP Diastolic blood pressure, MI Myocardial infarction, COPD Chronic obstructive pulmonary disease
The comparison of baseline characteristics between patients with increased RR and normal RR was presented in Table 1. Compared to patients with normal RR, patients with RR > 20 breaths/min were older, had fewer males, and presented with higher admission blood pressure and heart rate (p < 0.001). Moreover, they tended to have a higher percentage of concomitant CS, CKD, COPD, acute respiratory failure (ARF), ischemic heart disease, renal failure, hypertensive heart failure, pneumonia, and anemia (all p < 0.001). In addition, patients with admission RR > 20 breaths/min also had a relatively higher admission lactate, urea, creatinine, leukocytes, and glucose, but significantly lower PaO2 (all p < 0.001). As for medical therapy and procedures, dual antiplatelet therapy, statins, beta-blockers, angiotensin-converting enzyme inhibitors, and angiotensin receptor blockers were less likely to be administered to patients with increased RR, while torsemide, vasoactive drugs, and morphine had a higher prescription rate in patients with increased RR (all p < 0.001). Additionally, patients in the admission RR > 20 breaths/min were less likely to undergo invasive procedures, such as CABG, but had a higher reliance on minimally invasive procedures and mechanical support, such as DCA, ventilator, and ECMO (all p < 0.05) (Table 2).
Table 2.
A comparison of laboratory parameters and treatment between the two groups in MIMIC-IV Database
| Total n = 5631 | RR ≤ 20 breaths/min n = 3652 | RR > 20 breaths/min n = 1979 | P value | |
|---|---|---|---|---|
| Laboratory Parameters | ||||
| Troponin T (ng/mL) | 0.33 (0.1–1.1) | 0.34 (0.1–1.05) | 0.32 (0.11–1.18) | 0.345 |
| CK-MB (ng/mL) | 12.0 (5.0–37.0) | 12.0 (5.0–40.0) | 11.0 (5.0–35.0) | 0.281 |
| NT-proBNP (pg/mL) | 6188.0 (2326.0–16,291.0) | 5885.0 (2143.8–15,117.0) | 6482.0 (2518.0–17,083.0) | 0.166 |
| Lactate (mmol/L) | 1.8 (1.2–2.9) | 1.6 (1.2–2.5) | 2.2 (1.4–3.7) | < 0.001 |
| Albumin (g/dL) | 3.4 (2.9—3.8) | 3.4 (2.9–3.8) | 3.3 (2.9–3.7) | 0.002 |
| ALT (u/L) | 32.0 (18.0–69.0) | 31.0 (19.0–65.5) | 33.0 (18.0–75.0) | 0.194 |
| AST (u/L) | 55.0 (30.0—139.0) | 54.0 (29.0–132.0) | 56.0 (31.0–153.0) | 0.104 |
| Total Cholesterol (mg/dL) | 161.0 (128.0–194.0) | 162.0 (128.0–197.5) | 157.0 (127.3–185.8) | 0.205 |
| Triglyceride (mg/dL) | 112.0 (81.3–161.8) | 110.5 (80.8–161.0) | 112.5 (85.0–162.3) | 0.345 |
| HDL-C (mg/dL) | 43.0 (35.0–54.0) | 43.0 (35.0–53.0) | 43.0 (34.0–54.0) | 0.687 |
| LDL-C (mg/dL) | 90.0 (62.0–116.0) | 91.0 (62.0–117.0) | 88.0 (62.0–113.0) | 0.405 |
| Urea (mg/dL) | 23.0 (16.0–40.0) | 22.0 (15.0–35.0) | 27.0 (18.0–46.0) | < 0.001 |
| Creatinine (mg/dL) | 1.2 (0.9–1.8) | 1.1 (0.8–1.7) | 1.3 (0.9–2.1) | < 0.001 |
| Uric Acid (mg/dL) | 7.8 (5.3–11.4) | 7.9 (5.4–12.1) | 7.8 (5.2–10.7) | 0.374 |
| Leukocytes (× 109/L) | 11.2 (8.3–15.3) | 10.9 (8.1–14.6) | 11.9 (8.8–16.5) | < 0.001 |
| Neutrophil (%) | 80.7 (72.8–76.6) | 79.9 (72.0–85.8) | 82.4 (74.0–87.7) | < 0.001 |
| Hemoglobin (g/L) | 11.5 (9.7–13.2) | 11.7 (9.9–13.5) | 11.1 (9.2–12.8) | < 0.001 |
| Glucose (mg/dL) | 139.0 (110.0–191.0) | 134.0 (107.0–182.0) | 147.0 (115.0–205.0) | < 0.001 |
| INR | 1.2 (1.1–1.5) | 1.2 (1.1–1.5) | 1.3 (1.1–1.5) | < 0.001 |
| PaO2 (mmHg) | 106.0 (56.0–289.0) | 164.0 (66.0–353.0) | 75.0 (47.0–130.0) | < 0.001 |
| PaCO2 (mmHg) | 40.0 (35.0–46.0) | 40.0 (36.0–45.0) | 40.0 (34.0–47.0) | 0.084 |
| Treatments (%) | ||||
| Aspirin | 5052 (89.7) | 3360 (92.0) | 1692 (85.5) | < 0.001 |
| DAPT | 2336 (41.5) | 1573 (43.1) | 763 (38.6) | 0.001 |
| Statins | 4889 (86.80 | 3279 (89.8) | 1610 (81.4) | < 0.001 |
| ACEI | 2305 (40.9) | 1589 (43.5) | 716 (36.2) | < 0.001 |
| ARB | 206 (3.7) | 160 (4.4) | 46 (2.3) | < 0.001 |
| Digoxin | 249 (4.4) | 137 (3.8) | 112 (5.7) | < 0.001 |
| Beta-blockers | 4628 (82.2) | 3144 (86.1) | 1484 (75.0) | < 0.001 |
| Warfarin | 1233 (21.9) | 815 (22.3) | 418 (21.1) | 0.301 |
| Diuretics | ||||
| Furosemide | 3772 (67.0) | 2416 (66.2) | 1356 (68.5) | 0.072 |
| Torsemide | 434 (7.7) | 240 (6.6) | 194 (9.8) | < 0.001 |
| Bumetanide | 101 (1.8) | 59 (1.6) | 42 (2.1) | 0.171 |
| Vasopressors and Inotropes | ||||
| Dopamine | 438 (7.8) | 252 (6.9) | 186 (9.4) | < 0.001 |
| Dobutamine | 286 (5.1) | 149 (4.1) | 137 (6.9) | < 0.001 |
| Epinephrine | 584 (10.4) | 379 (10.4) | 205 (10.4) | 0.982 |
| Norepinephrine | 1489 (26.4) | 834 (22.8) | 655 (33.1) | < 0.001 |
| Milrinone | 257 (4.6) | 180 (4.9) | 77 (3.9) | 0.075 |
| Morphine | 613 (10.9) | 335 (9.2) | 278 (14.0) | < 0.001 |
| Procedures (%) | ||||
| CPR | 106 (1.9) | 68 (1.9) | 38 (1.9) | 0.878 |
| IABP | 182 (3.2) | 110 (3.0) | 72 (3.6) | 0.205 |
| CABG | 1281 (22.7) | 1096 (30.0) | 185 (9.3) | < 0.001 |
| Ventilator | 1276 (22.7) | 708 (19.4) | 568 (26.7) | < 0.001 |
| Dilation of Coronary Artery | 455 (8.1) | 258 (7.1) | 197 (10.0) | < 0.001 |
| ECMO | 21 (0.4) | 7 (0.2) | 14 (0.7) | 0.002 |
ALT Alanine transaminase, AST Aspartate transaminase, CK-MB Creatinine kinase-MB, NT-proBNP N-terminal pro-B-type natriuretic peptide, HDL-C High-density lipoprotein cholesterol, LDL-C Low-density lipoprotein cholesterol, DAPT Dual antiplatelet therapy, ACEI Angiotensin converting enzyme inhibitors, ARB Angiotensin receptor blocker, CPR Cardiopulmonary resuscitation, IABP Intra-aortic balloon pump, CABG Coronary artery bypass grafting, ECMO Extracorporeal membrane oxygenation, INR International normalized ratio, PaO2 arterial partial pressure of oxygen, PaCO2 arterial partial pressure of carbon dioxide
Outcome in MIMIC-IV
During the 30-day follow-up, a total of 1061 patients (18.8%) died. Patients in the elevated RR group had a significantly higher 30-day all-cause mortality compared to those with normal RR (27.5% vs. 14.2%, p < 0.001).
Association of RR with the outcome in MIMIC-IV
The K-M curves for admission RR and the primary endpoint were shown in Fig. 2A. Patients with an admission RR > 20 breaths/min exhibited a significantly higher cumulative incidence of events compared to those with a normal RR (log rank p < 0.001).
Fig. 2.
Kaplan–Meier survival analysis (A) and multivariate restricted cubic spline (RCS) (B) of admission RR with 30-day all-cause mortality in the MIMIC-IV database
The RCS analysis reveals a non-linear relationship between admission RR and the primary outcome (p < 0.05 for non-linearity, as shown in Fig. 2B), with a significant increase in mortality risk observed at higher RR levels (RR > 20 breaths/min).
Univariate and multivariate Cox regression models were constructed to evaluate the association between admission RR and the outcome (Table 3). After adjusted for age, sex, smoking, pneumonia, chronic obstructive pulmonary disease (COPD), CKD, cardiac troponin T, CK-MB, SBP, dyslipidemia, T2DM, HTN, HR, DCA, and old MI, admission RR > 20 breaths/min was independently associated with a higher risk of 30-day all-cause mortality (HR 1.715; 95%CI 1.507–1.952, p < 0.001).
Table 3.
Univariate and Multivariate analysis COX regression analysis of MIMIC-IV and Validation Cohort
| 30-day All-cause Mortality | |||||
|---|---|---|---|---|---|
| Univariate (HR) | p value | Multivariate (HR) | p value | Akaike Information Criterion | |
| MIMIC-IV Cohort | 17,561.235 | ||||
| RR > 20 breaths/min | 2.122 (1.881–2.394) | < 0.001 | 1.715 (1.507–1.952) | < 0.001 | |
| Age | 1.033 (1.028–1.038) | < 0.001 | 1.033 (1.028–1.039) | < 0.001 | |
| Male | 0.733 (0.649–0.827) | < 0.001 | 0.822 (0.726–0.929) | 0.002 | |
| Old MI | 0.988 (0.827–1.180) | 0.893 | |||
| T2DM | 1.023 (0.894–1.170) | 0.743 | |||
| Primary HTN | 0.602 (0.528–0.688) | < 0.001 | 0.635 (0.555–0.726) | < 0.001 | |
| CKD | 1.562 (1.375–1.774) | < 0.001 | |||
| Dyslipidemia | 0.608 (0.538–0.687) | < 0.001 | 0.658 (0.581–0.744) | < 0.001 | |
| Troponin T | 1.042 (1.022–1.062) | < 0.001 | 1.063 (1.043–1.084) | < 0.001 | |
| CK-MB | 1.000 (0.999–1.001) | 0.854 | 1.001 (1.000–1.002) | 0.018 | |
| SBP | 0.998 (0.996–0.999) | < 0.001 | 0.996 (0.995–0.997) | < 0.001 | |
| HR | 1.013 (1.010–1.016) | < 0.001 | 1.008 (1.005–1.011) | < 0.001 | |
| DCA | 0.675 (0.522–0.875) | 0.003 | 0.636 (0.489–0.827) | < 0.001 | |
| Smoking | 0.759 (0.641–0.899) | 0.001 | |||
| COPD | 1.419 (1.170–1.721) | < 0.001 | |||
| Pneumonia | 1.828 (1.582–2.111) | < 0.001 | 1.457 (1.258–1.657) | < 0.001 | |
| Validation Cohort | 789.388 | ||||
| RR > 20 breaths/min | 2.436 (1.520–3.904) | < 0.001 | 1.868 (1.102–3.168) | 0.020 | |
| Age | 1.056 (1.039–1.074) | < 0.001 | 1.053 (1.028–1.079) | < 0.001 | |
| Male | 0.425 (0.263–0.688) | < 0.001 | |||
| Old MI | 1.075 (0.338–3.419) | 0.902 | |||
| T2DM | 1.063 (0.664–1.702) | 0.798 | |||
| Primary HTN | 1.588 (0.978–2.578) | 0.062 | |||
| CKD | 2.494 (1.004–6.195) | 0.049 | |||
| Dyslipidemia | 0.944 (0.570–1.564) | 0.824 | |||
| Troponin T | 0.986 (0.964–1.008) | 0.215 | |||
| CK-MB | 0.999 (0.994–1.004) | 0.623 | |||
| SBP | 0.981 (0.974–0.989) | < 0.001 | 0.978 (0.970–0.986) | < 0.001 | |
| HR | 1.020 (1.009–1.031) | < 0.001 | 1.013 (1.003–1.023) | 0.011 | |
| DCA | 1.205 (0.617–2.353) | 0.586 | |||
| Smoking | 0.332 (0.206–0.535) | < 0.001 | 0.401 (0.230–0.700) | 0.001 | |
| COPD | 1.431 (0.521–3.925) | 0.487 | |||
| Pneumonia | 2.977 (1.657–5.347) | < 0.001 | |||
Old MI Old myocardial infarction, T2DM type 2 diabetes mellitus, HTN Hypertension, CKD Chronic kidney disease, CK-MB Creatinine kinase MB, SBP Systolic blood pressure, HR Heart rate, DCA dilatation of coronary artery, COPD Chronic obstructive pulmonary disease
A subgroup analysis was conducted to investigate the association between admission RR and the outcome in different subgroup patients (Fig. 3). There was interaction in patients age > 75 years old (HR 1.979; 95%CI 1.684–2.326) vs. ≤ 75 years old (HR 0.913; 95%CI 0.774–1.078) (p for interaction < 0.001), CS (HR 2.619; 95%CI 2.168–3.163) vs. without CS (HR 0.952; 95%CI 0.817–1.109) (p for interaction < 0.001), with ARF (HR 1.876; 95%CI 1.584–2.222) vs. without ARF (HR 0.979; 95%CI 0.836–1.147) (p for interaction < 0.001). In the subgroups of sex, smoking status, presence of atrial fibrillation/flutter, and presence of CKD, although there was an interaction, the effect of RR remained consistent, with only numerical differences.
Fig. 3.
Subgroup analysis of the primary endpoint according to demographic characteristics in the MIMIC-IV database
Validation of the findings
An institutional cohort was used to validate the findings from the MIMIC-IV. The baseline characteristics of the patients were shown in supplement Table 1 and supplement Table 2. The K-M curves of for 30-day all-cause mortality stratified by RR > 20 and ≤ 20 breaths/min were shown in Fig. 4. Consistent with the findings in MIMIC-IV cohort, patients with an RR > 20 breaths/min had significantly higher cumulative mortality compared to those with an RR ≤ 20 breaths/min (log rank p < 0.001).
Fig. 4.
Kaplan–Meier survival analysis and cumulative incidence of 30-day all-cause mortality according to admission RR in the Institutional Cohort
The results of the Cox regression analysis were presented in Table 3. Consistent with the findings from the MIMIC-IV database, an admission RR > 20 breaths/min was independently associated with an increased risk of 30-day all-cause mortality (HR 1.868, 95% CI 1.102–3.168, p = 0.020). Furthermore, the Akaike Information Criterion (AIC) comparison indicated that our institutional model (AIC 789.388) provided a better fit than the MIMIC-IV model (17,561.235). Additionally, the Schoenfeld residual test was performed to assess the proportional hazards assumption, and the results indicated the association between RR and 30-day all-cause mortality was independent of time. This finding suggests that RR consistently predicted mortality risk throughout the follow-up period without significant time-dependent effects (Supplementary Fig. 1).
Discussion
In this study, we demonstrated that RR, the often-overlooked vital sign in cardiovascular diseases plays an important prognostic role in patients with AMI. A significantly increased RR had higher mortality risk and was independently associated with poor prognosis. This study highlights the potential and importance of RR as a novel marker for risk stratification and prognosis assessment of AMI patients. To the best of our knowledge, this is the first study to evaluate the impact of admission RR on the outcome in patients with AMI patients.
RR is a physiological variable that is simple, easy to measure, practical, and can be evaluated at the bedside. In fact, the heart and the lung make up an inseparable anatomic and functional unit, where changes in one affect the other and vice versa. Previous studies have shown that RR was associated with the prognosis in patients with cardiovascular diseases, such as heart failure [14]; however, the prognostic value of RR in patients with AMI is limited. There were only two studies focusing on this issue. An early study in 1975 with 475 AMI patients treated in the coronary care unit found that maximum RR during the coronary care unit stay was one of the prognostically most important factors [19]. The other contemporary study was conducted in 2013, which enrolled 941 AMI patients and found RR was a significant predictor of death (HR 1.14, 95%CI 1.07–1.22, p < 0.0001); however, this study measured the RR within 2 weeks after AMI rather than on admission [20]. In contrast, our study aimed to evaluate the predictive value of admission RR in patients with AMI and found that an admission RR > 20 breaths/min was independently associated with an increased risk of poor outcomes. This finding underscores the potential of RR as a reliable, practical, cost-effective, and easily accessible prognostic marker that can help clinicians quickly identify high-risk patients and facilitate more timely and targeted interventions to improve patients’ outcomes. It is worth noting that in the subgroup analysis, there were interactions based on age (> 75 years and ≤ 75 years), the presence of CS, and the presence of ARF. This suggests that in older patients and those with CS or ARF, the adverse effects of elevated RR should be given more attention.
To date, the mechanisms why increased RR is associated with higher mortality remain unclear, but several findings may shed light on such a complex and multifactorial relationship. First, AMI, especially in the setting of anterior wall myocardial infarction, is prone to complications such as heart failure and subsequent pulmonary congestion, which can reflexively lead to an increased RR [21]. Second, AMI is often accompanied by increased sympathetic nervous system activity and there is sympathetic nerve distribution on the bronchial smooth muscle, the excitation of the sympathetic nervous system after AMI can also cause an increase in RR [22]. Third, the decrease in cardiac output after AMI leads to systemic hypoxia [23], further causing increased RR. In addition, some patients may have concurrent chronic pulmonary diseases, anemia, infection, or other conditions that could synergistically cause increased RR. Consistently, similar findings were found in our study where patients with increased RR had a lower oxygen partial pressure, higher percentage of COPD, anemia, and complicated with ARF compared to those with a normal RR. This finding suggests that an elevated RR in AMI patients may serve as an early indicator of a more severe clinical condition, such as a large infarct size, the presence of complications like heart failure, or the coexistence of multiple comorbidities. Therefore, closer monitoring and prompt identification of the underlying cause of increased RR are crucial for initiating timely and appropriate interventions (e.g. closer monitoring, oxygen therapy, and escalation of care), which may ultimately improve patients’ outcome.
The significance of this study lies in the inclusion of only patients who were initially admitted to the ICU or CCU, ensuring that the study population comprised critically ill individuals. Furthermore, we validated the findings from the MIMIC-IV database using our own dataset, and the consistency of results between the two cohorts strengthens the reliability of our conclusions. Notably, in the multivariate analysis, we accounted for covariates that may directly or indirectly influence RR in AMI patients, including age [24], sex [25], COPD [26], smoking [27], HR [28], CKD [29], and T2DM [30]. Even after such adjustment, admission RR remained an independent predictor for 30-day all-cause mortality. In addition, it was found in the subgroup analysis, elderly (age > 75 years), males, and those with complicating conditions such as CS and ARF had a significantly increased risk of mortality. Several possible explanations for these findings include that older patients often have poorer cardiopulmonary compensatory function and a higher burden of comorbidities, which can contribute to worse outcomes. Additionally, in patients with CS, hyperactivation of the sympathetic nervous system and pulmonary congestion can lead to an increased RR as a compensatory mechanism. In cases of ARF, the body may increase the RR to improve oxygenation and remove excess carbon dioxide. Male patients, especially those with a history of smoking, may experience a decline in lung function, which can lead to an elevated RR as a compensatory response, along with the potential development of COPD. These possible factors contribute to poor outcomes in these subgroup patients. However, further studies are needed to confirm these hypotheses.
Several limitations in this study should be addressed. First, RR is influenced by many factors, such as oxygen therapy, invasive mechanical ventilation, heart failure, undiagnosed pulmonary diseases, etc., which directly influence RR. Due to the retrospective nature of the study, some of these variables were not available in the database. Although we have adjusted for some relevant factors, there may still be some unidentified residual factors. Second, we only assessed the RR at the time of admission and did not monitor its changes over time. Monitoring the dynamic changes would better reflect the progression of the patient's condition and its correlation with prognosis. Future studies investigating the prognostic significance of dynamic RR changes, particularly within the first 24 to 48 h after admission, may offer valuable insights into optimizing risk stratification and improving outcome in AMI patients. Third, The population included in this study consists of AMI patients hospitalized in the intensive care unit. Whether findings from the present study can be generalized to non-ICU hospitalized patients requires further research for confirmation. Fourth, considering that an elevated RR is an independent risk factor for mortality in AMI, RR could be incorporated into current risk stratification tools for AMI, such as the GRACE or TIMI scores, to assess whether adding RR improves the predictive performance of these risk stratification models. However, some variables in the GRACE or TIMI scores are not available in the MIMIC database, thus further research is needed to evaluate whether adding RR to traditional risk stratification tools can improve predictive performance. In addition, the retrospective nature of this study may have introduced selection bias, proportional hazard bias, and inaccuracies in RR measurement. Since RR was primarily obtained from routine clinical records rather than standardized protocols, variations in measurement methods, observer bias, and differences in clinical settings could have affected the accuracy and consistency of the data. Additionally, the present study used the upper limit of normal RR as the cutoff value. However, whether a more appropriate cutoff value exists requires further research to confirm. Last but not least, although we utilized the MIMIC-IV database and validated the findings with our institutional cohort, the generalization and interpretation of this study should be approached cautiously, and more prospective studies with larger sample sizes are needed to confirm our findings.
Conclusions
RR is not only a vital sign but also a simple and practical indicator for predicting the prognosis of patients with AMI. Increased RR upon admission independently predicted a higher risk of poor outcomes, making it a valuable tool for early risk stratification and guiding timely interventions.
Supplementary Information
Supplementary Material 1: Figure 1. Univariate and multivariate Schoenfeld residuals plot for the primary outcome
Acknowledgements
None
Authors’ contributions
Author Contributions: B.R.S and S.Y.X contributed equally. B.R.S and S.Y.X designed and performed the experiments, derived the models, and analyzed the data. J.S, X.L, Y.Y, J.C, G.L contributed to the design and conceptualization of the study. Y.T.C, L.F.X, and S.Y.X contributed for data extraction. B.R.S and S.Y.X wrote the manuscript with support from B.H and S.X.L All authors read and approved the final manuscript.
Funding
This study is supported by The Innovation Program for Doctoral Student of The First Affiliated Hospital of Chongqing Medical University (CYB22199) No. CYYY-BSYJSCXXM- 202325.
Data availability
The data presented in this study are available on reasonable request from the corresponding author.
Declarations
Ethics approval and consent to participate
This study was conducted in accordance with the Declaration of Helsinki. The MIMIC-IV database has received ethical approval from the Institutional Review Boards (IRB) of Massachusetts Institute of Technology (Cambridge, MA) (No. 0403000206) and IRB of Beth Israel Deaconess Medical Centre (Boston, MA) (2001-P- 001699/14). Additionally, this study informed consent was waived and approved by the ethics committee of The First Affiliated Hospital of Chongqing Medical University. All methods in this study were carried out in accordance with the relevant guidelines and regulations in the Declaration of Helsinki.
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.
Bryan Richard Sasmita and Siyuan Xie contributed equally to this work.
Contributor Information
Suxin Luo, Email: luosuxin0204@163.com.
Bi Huang, Email: huangbi120@163.com.
References
- 1.Ojha N, Dhamoon AS. Myocardial Infarction. [Updated 2023 Aug 8]. In: StatPearls. Treasure Island (FL): StatPearls Publishing; 2024. Available from: https://www.ncbi.nlm.nih.gov/books/NBK537076/.
- 2.The top 10 causes of death. Who.int. Available from: https://www.who.int/news-room/fact-sheets/detail/the-top-10-causes-of-death.
- 3.World Health Organization. Cardiovascular diseases. [ Feb; 2024 ]. 2022. https://www.who.int/health-topics/cardiovascular-diseases/.
- 4.Tang EW, Wong CK, Herbison P. Global Registry of Acute Coronary Events (GRACE) hospital discharge risk score accurately predicts long-term mortality post acute coronary syndrome. Am Heart J. 2007;153(1):29–35. 10.1016/j.ahj.2006.10.004. [DOI] [PubMed] [Google Scholar]
- 5.Antman EM, Cohen M, Bernink PJ, et al. The TIMI risk score for unstable angina/non-ST elevation MI: a method for prognostication and therapeutic decision making. JAMA. 2000;284(7):835–42. 10.1001/jama.284.7.835. [DOI] [PubMed] [Google Scholar]
- 6.Solomon SD, Zelenkofske S, McMurray JJ, et al. Sudden death in patients with myocardial infarction and left ventricular dysfunction, heart failure, or both. N Engl J Med. 2005;352(25):2581–8. 10.1056/NEJMoa043938. [DOI] [PubMed] [Google Scholar]
- 7.Antman EM, Tanasijevic MJ, Thompson B, et al. Cardiac-specific troponin I levels to predict the risk of mortality in patients with acute coronary syndromes. N Engl J Med. 1996;335(18):1342–9. 10.1056/NEJM199610313351802. [DOI] [PubMed] [Google Scholar]
- 8.Babuin L, Jaffe AS. Troponin: the biomarker of choice for the detection of cardiac injury. CMAJ. 2005;173(10):1191–202. 10.1503/cmaj/05129. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Abu-Assi E, Ferreira-Gonzalez I, Ribera A, et al. Do GRACE (Global Registry of Acute Coronary events) risk scores still maintain their performance for predicting mortality in the era of contemporary management of acute coronary syndromes? Am Heart J. 2010;160(5):826-834 e1-3. 10.1016/j.ahj.2010.06.053. [DOI] [PubMed] [Google Scholar]
- 10.Verdecchia P, Reboldi G, Angeli F, et al. Systolic and diastolic blood pressure changes in relation with myocardial infarction and stroke in patients with coronary artery disease. Hypertension. 2015;65(1):108–14. 10.1161/HYPERTENSIONAHA.114.04310. [DOI] [PubMed] [Google Scholar]
- 11.Messerli FH, Shalaeva EV, Rexhaj E. Optimal BP targets to prevent stroke and MI: is there a lesser of 2 evils? J Am Coll Cardiol. 2021;78(17):1679–81. 10.1016/j.jacc.2021.09.01. [DOI] [PubMed] [Google Scholar]
- 12.Shah B, Bartaula B, Adhikari J, Neupane HS, Shah BP, Poudel G. Predictors of in-hospital mortality of acute ischemic stroke in adult population. J Neurosci Rural Pract. 2017;8(4):591–4. 10.4103/jnrp.jnrp_265_17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Jin G, Hu W, Zeng L, Ma B, Zhou M. Prediction of long-term mortality in patients with ischemic stroke based on clinical characteristics on the first day of ICU admission: an easy-to-use nomogram. Front Neurol. 2023;14:1148185. 10.3389/fneur.2023.1148185. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Metra M, Cotter G, El-Khorazaty J, et al. Acute heart failure in the elderly: differences in clinical characteristics, outcomes, and prognostic factors in the VERITAS Study. J Card Fail. 2015;21(3):179–88. 10.1016/j.cardfail.2014.12.012. [DOI] [PubMed] [Google Scholar]
- 15.Han D, Xu F, Zhang L, et al. Early prediction of in-hospital mortality in patients with congestive heart failure in intensive care unit: a retrospective observational cohort study. BMJ Open. 2022;12(7):e059761. 10.1136/bmjopen-2021-059761. Published 2022 Jul 19. [Google Scholar]
- 16.Andersen LW, Kim WY, Chase M, et al. The prevalence and significance of abnormal vital signs prior to in-hospital cardiac arrest. Resuscitation. 2016;98:112–7. 10.1016/j.resuscitation.2015.08.016. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Fieselmann JF, Hendryx MS, Helms CM, Wakefield DS. Respiratory rate predicts cardiopulmonary arrest for internal medicine inpatients. J Gen Intern Med. 1993;8(7):354–60. 10.1007/BF02600071. [DOI] [PubMed] [Google Scholar]
- 18.Johnson AEW, Bulgarelli L, Shen L, et al. MIMIC-IV, a freely accessible electronic health record dataset. Sci Data. 2023;10(1):1. 10.1038/s41597-022-01899-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Helmers C. Assessment of 3-year prognosis in survivors of acute myocardial infarction. Br Heart J. 1975;37(6):593–7. 10.1136/hrt.37.6.593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Barthel P, Wensel R, Bauer A, et al. Respiratory rate predicts outcome after acute myocardial infarction: a prospective cohort study. Eur Heart J. 2013;34(22):1644–50. 10.1093/eurheartj/ehs420. [DOI] [PubMed] [Google Scholar]
- 21.Harshe GG, Jotkar SK. Prolonged respiratory failure following acute myocardial infarction: a case report. Cardiol Res. 2011;2(2):86–9. 10.4021/cr21e. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Graham LN, Smith PA, Huggett RJ, Stoker JB, Mackintosh AF, Mary DA. Sympathetic drive in anterior and inferior uncomplicated acute myocardial infarction. Circulation. 2004;109(19):2285–9. 10.1161/01.CIR.0000129252.96341.8B. [DOI] [PubMed] [Google Scholar]
- 23.Valencia A, Burgess JH. Arterial hypoxemia following acute myocardial infarction. Circulation. 1969;40(5):641–52. 10.1161/01.cir.40.5.641. [DOI] [PubMed] [Google Scholar]
- 24.Sharma G, Goodwin J. Effect of aging on respiratory system physiology and immunology. Clin Interv Aging. 2006;1(3):253–60. 10.2147/ciia.2006.1.3.253. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.LoMauro A, Aliverti A. Sex and gender in respiratory physiology. Eur Respir Rev. 2021;30(162):210038. 10.1183/16000617.0038-2021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yentes JM, Fallahtafti F, Denton W, Rennard SI. COPD patients have a restricted breathing pattern that persists with increased metabolic demands. COPD. 2020;17(3):245–52. 10.1080/15412555.2020.1750578. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Tantisuwat A, Thaveeratitham P. Effects of smoking on chest expansion, lung function, and respiratory muscle strength of youths. J Phys Ther Sci. 2014;26(2):167–70. 10.1589/jpts.26.167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Ben-Tal A, Shamailov SS, Paton JF. Evaluating the physiological significance of respiratory sinus arrhythmia: looking beyond ventilation-perfusion efficiency. J Physiol. 2012;590(8):1989–2008. 10.1113/jphysiol.2011.222422. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Gembillo G, Calimeri S, Tranchida V, et al. Lung dysfunction and chronic kidney disease: a complex network of multiple interactions. J Pers Med. 2023;13(2):286. 10.3390/jpm13020286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Kuziemski K, Slominski W, Jassem E. Impact of diabetes mellitus on functional exercise capacity and pulmonary functions in patients with diabetes and healthy persons. BMC Endocr Disord. 2019;19(1):2. 10.1186/s12902-018-0328-1. [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.
Supplementary Materials
Supplementary Material 1: Figure 1. Univariate and multivariate Schoenfeld residuals plot for the primary outcome
Data Availability Statement
The data presented in this study are available on reasonable request from the corresponding author.