Abstract
Major adverse cardiovascular events (MACEs) in geriatric patients are an important cause of increased mortality and morbidity. The results of current studies regarding the predictive value of the NT-proBNP, H-FABP, and AUB-HAS2 scales for cardiovascular complications are inconsistent, and there is no relevant large sample study. Therefore, this study aimed to investigate whether preoperative NT-proBNP, H-FABP, and AUB-HAS2 alone or in combination can effectively predict postoperative cardiovascular complications in geriatric patients. A total of 1736 geriatric patients (aged ≥ 65 years) who were scheduled for elective non-cardiac surgery under general anesthesia were enrolled. AUB-HAS2 risk assessment is required for each patient, and blood was collected 1 h before surgery for the measurement of NT-proBNP and H-FABP. The primary outcomes were MACEs within 30 days after surgery. The secondary outcomes were other complications. Its predictive value was analyzed by receiver operating characteristic (ROC) curves. Of the 1736 patients, 71 (4.1%) had MACEs. NT-proBNP was a predictor of MACEs (AUC = 0.763; 95% CI 0.695–0.832; P < 0.001). When H-FABP was combined with AUB-HAS2, AUB-HAS2 increased the predictive value of H-FABP (AUC = 0.736; 95% CI 0.673–0.799; P < 0.001). Multiple logistic regression analysis revealed increased predictive value of the modified AUB-HAS2 scale for MACEs (AUC = 0.794, 95% CI = 0.737–0.851, P < 0.001). Our study revealed the predictive efficacy and prognostic value of NT-proBNP, H-FABP and the AUB-HAS2 score alone or in combination for postoperative MACE risk assessment in geriatric patients undergoing non-cardiac surgery.
This trial was registered at the Chinese Clinical Trial Registry (2019/09/27 ChiCTR1900026223, https://www.chictr.org.cn).
Supplementary Information
The online version contains supplementary material available at 10.1038/s41598-025-95987-8.
Keywords: Geriatric patients, Non-cardiac surgery, N-terminal pro-B-type natriuretic peptide, Heart type fatty acid binding protein, AUB-HAS2 cardiovascular risk index, Major adverse cardiovascular events
Subject terms: Biomarkers, Medical research
Introduction
Globally, more than 300 million adults undergo non-cardiac surgery annually, with an average overall complication rate of 7–11%1,2. The mortality rate within 30 days after surgery ranged from 0.5 to 2%, of which the most important cause of death was major adverse cardiovascular events(MACEs), accounting for up to 42% of the perioperative death risk3,4. MACEs are also considered to be significantly associated with increased morbidity, prolonged hospital stays, and increased medical costs in patients after non-cardiac surgery, especially in geriatric patients5. Although a large number of clinical trials and prospective observational studies have improved awareness of the risk prediction and management of MACEs, this remains a major public health concern. Therefore, preoperative prediction is crucial for providing optimal medical care for geriatric patients and contributes to the improvement of perioperative cardiovascular management.
Preoperative identification of populations at high risk of MACEs is essential for perioperative cardiovascular management in geriatric. Currently, the risk index and blood biomarkers are the main tools for risk stratification before surgery. The various clinical scores can also predict MACEs. The revised cardiac risk index (RCRI) and American College of Surgeons/National Surgical Quality Improvement Program (ACS/NSQIP) scale are recommended in the 2014 ACC/AHA guidelines for the risk assessment of cardiovascular complications6. In addition, the American University of Beirut (AUB)-HAS2 Cardiovascular Risk Index has also been frequently used, and studies have shown that this index is more effective than the commonly used RCRI7. However, some studies have shown differences among different surgical types and have not considered high-risk conditions with low incidence, which can easily lead to an underestimation of patient risk8.
Previous studies have shown that the preoperative N-terminal type-B natriuretic peptide precursor (NT-proBNP) is closely related to cardiovascular death and myocardial infarction within 30 days after non-cardiac surgery and improves the prediction of cardiovascular risk9,10. Studies have also shown that high-sensitivity troponin I (hs-TnI), high-sensitivity troponin T (hs-TnT), and C-reactive protein (CRP) also have predictive effects on postoperative MACEs11–13. Cardiac-type fatty acid binding protein (H-FABP) is a cardiac protein that plays an important role in the metabolism of fatty acids (FAs) in cardiomyocytes and has a high degree of cardiomyocyte specificity14. Previous studies have shown that H-FABP is related to pulmonary embolism15, myocardial infarction16, renal injury17 and postoperative cognitive dysfunction18; however, there are relatively few studies on the application of H-FABP in the perioperative period, and the relationship between it and MACEs after non-cardiac surgery is not clear.
Therefore, while the risk scale is a convenient tool that takes into account multiple factors, its effectiveness can be influenced by the characteristics of the modeling cohort. In contrast, biomarkers for predicting cardiac events/prognoses are objective and widely used. However, relying on a single indicator has limitations because its sensitivity is dependent on its source and biological metabolism. Additionally, different levels of the same biomarker in patients with varying underlying diseases may have differing clinical significance. Both types of tools have their advantages, and combining them could improve clinical accuracy and accessibility. Nevertheless, there is currently no confirmation from relevant large sample studies regarding how to apply this combination or whether critical values change after combination therapy. Therefore, this study explored the ability of the combination of NT-proBNP, H-FABP and AUB-HAS2 to predict MACEs after non-cardiac surgery in geriatric. We developed a modified AUB-HAS2 risk index based on the study results to improve its clinical application value.
Results
Patients’ baseline characteristics
From June 2020 to December 2021, 2017 patients met the study inclusion criteria. Four patients were excluded because of rapid changes of their condition and change from elective surgery to emergency surgery. And one patient was excluded due to a preoperative psychiatric disorder. After surgery, 248 patients were excluded because 239 patients had undergone surgery for less than 90 min, and 9 patients had missing data. Overall, 1764 patients were included in the study, and 1736 patients (98.4%) completed the 30-day follow-up. Patient screening and recruitment are shown in Fig. 1.
Fig. 1.
Participant screening, recruitment, and follow-up.
The mean age of the included patients was 71.5 ± 5.2 years, 30.4% were female, and the median BMI was 23.03 (21.08, 25.25). The types of surgery mainly include abdominal surgery (67.0%) and urology surgery (23.6%). The ASA classification was mostly Grade II (46.6%) or Grade III (52.9%). Among them, 71 patients (4.1%) had MACEs. At baseline, patients who developed postoperative MACEs were significantly older, had higher ASA grades, and were more likely to have comorbidities before surgery. In patients with MACEs, NT-proBNP, H-FABP, and AUB-HAS2 scores were all higher than those in the unaffected group (P<0.001). The demographic characteristics and preoperative comorbidities of these patients are shown in Table 1.
Table 1.
Baseline characteristics of patients undergoing major non-cardiac surgery with or without postoperative major adverse cardiovascular events.
| All surgeries(n = 1736) | No postoperative MACEs(n = 1665) | Postoperative MACEs (n = 71) | P value | ||
|---|---|---|---|---|---|
| Age | 71.5 ± 5.2 | 71.4 ± 5.1 | 74.1 ± 6.2 | <0.001 | |
| Sex | Female | 528 (30.4%) | 510 (30.6%) | 18 (25.4%) | 0.344 |
| Male | 1208 (69.6%) | 1155 (69.4%) | 53 (74.6%) | ||
| BMI | 23.03 (21.08, 25.25) | 23.03 (21.08, 25.24) | 23.37 (21.01, 25.71) | 0.583 | |
| Procedure type | Abdominal surgery | 1163 (67.0%) | 1113 (66.8%) | 50 (70.4%) | 0.689 |
| Orthopaedic surgery | 46 (2.6%) | 44 (2.6%) | 2 (2.8%) | ||
| Urological surgery | 409 (23.6%) | 396 (23.8%) | 13 (18.3%) | ||
| Chest non-cardiac surgery | 118 (6.8%) | 112 (6.7%) | 6 (8.5%) | ||
| Cancer-related surgery | No | 123 (7.1%) | 116 (7.0%) | 7 (9.9%) | 0.352 |
| Yes | 1613 (92.9%) | 1549 (93.0%) | 64 (90.1%) | ||
| Type of surgery | Open surgery | 877 (50.5%) | 840 (50.5%) | 37 (52.1%) | 0.784 |
| Endoscopic surgery | 859 (49.5%) | 825 (49.5%) | 34 (47.9%) | ||
| General situation | Good | 1254 (72.2%) | 1211 (72.7%) | 43 (60.6%) | 0.003 |
| Common | 471 (27.1%) | 446 (26.8%) | 25 (35.2%) | ||
| Bad | 11 (0.6%) | 8 (0.5%) | 3 (4.2%) | ||
| Frail scale | No frailty | 118 (6.8%) | 115 (6.9%) | 3 (4.2%) | 0.098 |
| Frailty prophase | 1606 (92.5%) | 1540 (92.5%) | 66 (93.0%) | ||
| Frailty | 12 (0.7%) | 10 (0.6%) | 2 (2.8%) | ||
| Smoke | No | 1421 (81.9%) | 1360 (81.7%) | 61 (85.9%) | 0.365 |
| Yes | 315 (18.1%) | 305 (18.3%) | 10 (14.1%) | ||
| Alcohol drinking | No | 1654 (95.3%) | 1590 (95.5%) | 64 (90.1%) | 0.072 |
| Yes | 82 (4.7%) | 75 (4.5%) | 7 (9.9%) | ||
| Grade of dyspnea | No dyspnea | 1503 (86.6%) | 1440 (86.5%) | 63 (88.7%) | 0.023 |
| I | 97 (5.6%) | 95 (5.7%) | 2 (2.8%) | ||
| II | 117 (6.7%) | 114 (6.8%) | 3 (4.2%) | ||
| III | 18 (1.0%) | 16 (1.0%) | 2 (2.8%) | ||
| IV | 1 (0.1%) | 0 (0.0%) | 1 (1.4%) | ||
| NYHA cardiac functional grading | I | 995 (57.3%) | 964 (57.9%) | 31 (43.7%) | 0.010 |
| II | 691 (39.8%) | 657 (39.5%) | 34 (47.9%) | ||
| III | 48 (2.8%) | 42 (2.5%) | 6 (8.5%) | ||
| IV | 2 (0.1%) | 2 (0.1%) | 0 (0.0%) | ||
| MET | >6MET | 272 (15.7%) | 266 (16.0%) | 6 (8.5%) | <0.001 |
| 3-6MET | 1215 (70.0%) | 1173 (70.5%) | 42 (59.2%) | ||
| <3MET | 249 (14.3%) | 226 (13.6%) | 23 (32.4%) | ||
| ASA | I | 2 (0.1%) | 2 (0.1%) | 0 (0.0%) | <0.001 |
| II | 809 (46.6%) | 793 (47.6%) | 16 (22.5%) | ||
| III | 919 (52.9%) | 865 (52.0%) | 54 (76.1%) | ||
| IV | 6 (0.3%) | 5 (0.3%) | 1 (1.4%) | ||
| Grade of renal function | I | 1686 (97.1%) | 1624 (97.5%) | 62 (87.3%) | <0.001 |
| II | 38 (2.2%) | 32 (1.9%) | 6 (8.5%) | ||
| III | 10 (0.6%) | 8 (0.5%) | 2 (2.8%) | ||
| IV | 2 (0.1%) | 1 (0.1%) | 1 (1.4%) | ||
| Hypertension | No | 1012 (58.3%) | 979 (58.8%) | 33 (46.5%) | 0.039 |
| Yes | 724 (41.7%) | 686 (41.2%) | 38 (53.5%) | ||
| Ischemic heart disease | No | 1670 (96.2%) | 1607 (96.5%) | 63 (88.7%) | 0.002 |
| Yes | 66 (3.8%) | 58 (3.5%) | 8 (11.3%) | ||
| Coronary heart disease | No | 1656 (95.4%) | 1597 (95.9%) | 59 (83.1%) | <0.001 |
| Yes | 80 (4.6%) | 68 (4.1%) | 12 (16.9%) | ||
| Postoperative stent placement | No | 1706 (98.3%) | 1640 (98.5%) | 66 (93.0%) | 0.002 |
| Yes | 30 (1.7%) | 25 (1.5%) | 5 (7.0%) | ||
| Valvular heart disease | No | 1731 (99.7%) | 1660 (99.7%) | 71 (100.0%) | 1.000 |
| Yes | 5 (0.3%) | 5 (0.3%) | 0 (0.0%) | ||
| Congenital heart disease | No | 1731 (99.7%) | 1660 (99.7%) | 71 (100.0%) | 1.000 |
| Yes | 5 (0.3%) | 5 (0.3%) | 0 (0.0%) | ||
| Dilated cardiomyopathy | No | 1729 (99.6%) | 1662 (99.8%) | 67 (94.4%) | <0.001 |
| Yes | 7 (0.4%) | 3 (0.2%) | 4 (5.6%) | ||
| Congestive heart-failure | No | 1731 (99.7%) | 1662 (99.8%) | 69 (97.2%) | 0.015 |
| Yes | 5 (0.3%) | 3 (0.2%) | 2 (2.8%) | ||
| Arrhythmia | No | 1594 (91.8%) | 1544 (92.7%) | 50 (70.4%) | <0.001 |
| Yes | 142 (8.2%) | 121 (7.3%) | 21 (29.6%) | ||
| Peripheral vascular disease | No | 1713 (98.7%) | 1646 (98.9%) | 67 (94.4%) | 0.013 |
| Yes | 23 (1.3%) | 19 (1.1%) | 4 (5.6%) | ||
| COPD | No | 1667 (96.0%) | 1602 (96.2%) | 65 (91.5%) | 0.097 |
| Yes | 69 (4.0%) | 63 (3.8%) | 6 (8.5%) | ||
| Stroke | No | 1687 (97.2%) | 1621 (97.4%) | 66 (93.0%) | 0.068 |
| Yes | 49 (2.8%) | 44 (2.6%) | 5 (7.0%) | ||
| TIA | No | 1710 (98.5%) | 1641 (98.6%) | 69 (97.2%) | 0.663 |
| Yes | 26 (1.5%) | 24 (1.4%) | 2 (2.8%) | ||
| Abnormal thyroid function | No | 1712 (98.6%) | 1643 (98.7%) | 69 (97.2%) | 0.257 |
| Yes | 24 (1.4%) | 22 (1.3%) | 2 (2.8%) | ||
| Diabetes | No | 1400 (80.6%) | 1354 (81.3%) | 46 (64.8%) | 0.001 |
| Yes | 336 (19.4%) | 311 (18.7%) | 25 (35.2%) | ||
| Anemia | No | 1150 (66.2%) | 1116 (67.0%) | 34 (47.9%) | 0.001 |
| Yes | 586 (33.8%) | 549 (33.0%) | 37 (52.1%) | ||
| Albumin (g/l) | 41.35 (38.53, 43.80) | 41.1 (38.7, 43.8) | 39.2 (37.3, 43.3) | 0.005 | |
| Creatinine (umol/l) | 80.0 (68.0, 92.0) | 79.0 (67.0, 91.0) | 86.0 (74.0, 99.0) | 0.002 | |
| Blood glucose concentration (mmol/l) | 5.21 (4.75,5.97) | 5.20 (4.75, 5.97) | 5.27 (4.68, 6.11) | 0.597 | |
| Hemoglobin (g/l) | 128.00 (114.00, 140.00) | 129.00 (115.00, 140.00) | 119.00 (102.00, 138.00) | 0.029 | |
| NT-proBNP (pg/ml) | 122.24 (75.34, 201.71) | 120.05 (74.46, 194.72) | 344.40 (139.56, 867.35) | <0.001 | |
| H-FABP (ng/ml) | 2.24 (1.62, 3.04) | 2.21 (1.61, 3.01) | 2.81 (1.99, 3.54) | <0.001 | |
| AUB-HAS2 | 1.00 (0.00, 1.00) | 1.00 (0.00, 1.00) | 1.00 (1.00, 2.00) | <0.001 | |
| Operation time | 180.00 (134.25, 243.75) | 180.00 (134.00, 243.00) | 193.00 (135.00, 250.00) | 0.304 | |
| Anesthesia time | 251.00 (196.00, 320.00) | 251.00 (195.00, 320.00) | 270.00 (199.00, 330.00) | 0.275 |
MACEs, major adverse cardiovascular events; BMI, Body mass index; ASA, American Society of Anesthesiologists Physical Status; MET, Metabolic equivalent; COPD, Chronic obstructive pulmonary disease; TIA, Transient ischemic attack; NT-proBNP, N-terminal type-B natriuretic peptide precursor; H-FABP, Cardiac-type fatty acid binding protein; AUB-HAS2, American University of Beirut (AUB)-HAS2 Cardiovascular Risk Index. Grade of renal function is currently differentiated according to international standards and glomerular filtration rate.
Postoperative complications and ROC analysis
The incidence of complications within 30 days after surgery is shown in Table A.1 in the Supplemental Digital Content (SDC)-A. 71 (4.1%) patients developed MACEs. Among these patients, 56 (78.9%) had arrhythmia present or aggravated, and 39 had atrial fibrillation. The all-cause mortality in the entire cohort was 0.3%, and four patients died due to multiple organ failure and two patients died from respiratory and circulatory failure. And cardiac arrest, myocardial infarction or ischemic stroke occurred in 0.9% of patients who underwent postoperative surgery. After stratification by the normal values of NT-proBNP, significant differences were found in the incidence of complications (except ischemic stroke) at different NT-proBNP levels. There were statistically significant differences in the incidence of MACEs, arrhythmia, and renal dysfunction at different H-FABP levels (P < 0.001).
The ROC analysis demonstrated the predictive value of the NT-proBNP, H-FABP, and AUB-HAS2 scores for MACEs and other postoperative complications (Table 2; SDC-Figure A.1). For MACEs, the AUC for NT-proBNP was 0.763 (P < 0.001), while the AUC was 0.628 (P < 0.001) for H-FABP, and the AUC for AUB-HAS2 was 0.712 (P < 0.001). However, when H-FABP was combined with NT-proBNP, the AUC was 0.780 (P < 0.001), and when H-FABP was combined with AUB-HAS2, the AUC was 0.736 (P < 0.001), and the predictive value of H-FABP increased. The AUC of the combination of NT-proBNP, H-FABP and AUB-HAS2 was 0.788 (P < 0.001), indicating that this combination had the highest predictive value.
Table 2.
Predictive performance of NT-proBNP、H-FABP and AUB-HAS2 for 30-day postoperative major adverse cardiovascular events and other 30-day postoperative complication.
| Variables | NT-proBNP | H-FABP | AUB-HAS2 | NT-proBNP *H-FABP | NT-proBNP *AUB-HAS2 | H-FABP *AUB-HAS2 | NT-proBNP *H-FABP *AUB-HAS2 | modified AUB-HAS2 | modified AUB-HAS2*H-FABP | |
|---|---|---|---|---|---|---|---|---|---|---|
| MACEs | AUC (95%CI) | 0.763(0.695, 0.832) | 0.628(0.558, 0.698) | 0.712(0.651, 0.774) | 0.780(0.713, 0.846) | 0.781(0.720, 0.842) | 0.736(0.673, 0.799) | 0.788(0.728, 0.848) | 0.794(0.737, 0.851) | 0.800(0.744, 0.856) |
| P value | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| Arrhythmia | AUC (95%CI) | 0.783(0.708, 0.858) | 0.607(0.530, 0.685) | 0.695(0.623, 0.766) | 0.797(0.727, 0.867) | 0.774(0.706, 0.843) | 0.711(0.638, 0.785) | 0.777(0.710, 0.845) | 0.771(0.703, 0.838) | 0.765(0.695, 0.834) |
| P value | <0.001 | 0.006 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| Atrial fibrillation | AUC (95%CI) | 0.785(0.696, 0.874) | 0.537(0.444, 0.629) | 0.693(0.611, 0.775) | 0.779(0.691, 0.868) | 0.790(0.706, 0.873) | 0.704(0.628, 0.781) | 0.783(0.707, 0.859) | 0.753(0.669, 0.837) | 0.763(0.683, 0.844) |
| P value | <0.001 | 0.434 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| Other types of arrhythmias | AUC (95%CI) | 0.796(0.674, 0.918) | 0.744(0.651, 0.836) | 0.700(0.576, 0.824) | 0.873(0.813, 0.932) | 0.762(0.646, 0.877) | 0.761(0.652, 0.869) | 0.793(0.695, 0.891) | 0.811(0.715, 0.907) | 0.840(0.764, 0.915) |
| P value | <0.001 | <0.001 | 0.002 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| Death | AUC (95%CI) | 0.720(0.527, 0.913) | 0.553(0.254, 0.852) | 0.699(0.493, 0.905) | 0.527(0.238, 0.817) | 0.777(0.626, 0.927) | 0.729(0.530, 0.927) | 0.680(0.419, 0.941) | 0.757(0.620, 0.895) | 0.734(0.579, 0.889) |
| P value | 0.063 | 0.653 | 0.092 | 0.817 | 0.019 | 0.053 | 0.127 | 0.029 | 0.047 | |
| Cardiac arrest | AUC (95%CI) | 0.833(0.746, 0.920) | 0.796(0.638, 0.954) | 0.676(0.431, 0.922) | 0.640(0.325, 0.954) | 0.759(0.591, 0.927) | 0.816(0.621, 1.000) | 0.725(0.404, 1.000) | 0.820(0.678, 0.962) | 0.839(0.706, 0.972) |
| P value | 0.010 | 0.022 | 0.173 | 0.280 | 0.045 | 0.015 | 0.082 | 0.013 | 0.009 | |
| Myocardial infarction | AUC (95%CI) | 0.661(0.371, 0.950) | 0.647(0.432, 0.861) | 0.630(0.378, 0.882) | 0.654(0.369, 0.938) | 0.731(0.500, 0.962) | 0.754(0.563, 0.944) | 0.732(0.503, 0.961) | 0.790(0.659, 0.920) | 0.789(0.654, 0.923) |
| P value | 0.174 | 0.214 | 0.271 | 0.193 | 0.050 | 0.032 | 0.050 | 0.014 | 0.015 | |
| Stroke | AUC (95%CI) | 0.688(0.493, 0.883) | 0.799(0.722, 0.877) | 0.768(0.644, 0.891) | 0.798(0.697, 0.899) | 0.829(0.719, 0.938) | 0.876(0.807, 0.944) | 0.869(0.794, 0.944) | 0.829(0.713, 0.945) | 0.864(0.778, 0.949) |
| P value | 0.111 | 0.011 | 0.023 | 0.012 | 0.005 | 0.001 | 0.002 | 0.005 | 0.002 | |
| Cardiac insufficiency | AUC (95%CI) | 0.679(0.590, 0.768) | 0.582(0.477, 0.688) | 0.704(0.623, 0.786) | 0.676(0.587, 0.765) | 0.753(0.675, 0.831) | 0.717(0.627, 0.806) | 0.753(0.675, 0.832) | 0.724(0.637, 0.811) | 0.721(0.629, 0.813) |
| P value | <0.001 | 0.071 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| Renal inadequacy | AUC (95%CI) | 0.832(0.749, 0.916) | 0.701(0.577, 0.826) | 0.763(0.682, 0.844) | 0.762(0.642, 0.881) | 0.837(0.770, 0.905) | 0.809(0.721, 0.897) | 0.822(0.736, 0.909) | 0.858(0.789, 0.926) | 0.859(0.783, 0.935) |
| P value | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | <0.001 | |
| MODS | AUC (95%CI) | 0.790(0.609, 0.970) | 0.668(0.356, 0.979) | 0.711(0.469, 0.953) | 0.633(0.321, 0.946) | 0.842(0.709, 0.975) | 0.722(0.445, 0.998) | 0.663(0.325, 1.000) | 0.819(0.678, 0.960) | 0.811(0.649, 0.974) |
| P value | 0.025 | 0.194 | 0.103 | 0.303 | 0.008 | 0.086 | 0.207 | 0.014 | 0.016 |
MACEs, major adverse cardiovascular events; MODS, multiple organ dysfunction syndrome.
NT-proBNP had good predictive value for arrhythmias, cardiac arrest, renal insufficiency, and MODS, with AUCs of 0.783 (P < 0.001), 0.833 (P = 0.010), 0.832 (P < 0.001), and 0.790 (P = 0.025), respectively. H-FABP had a high predictive value for cardiac arrest and ischemic stroke, with AUCs of 0.796 (P = 0.022) and 0.799 (P = 0.011), respectively. For ischemic stroke and renal insufficiency patients, the AUCs of AUB-HAS2 were 0.768 (P = 0.023) and 0.763 (P < 0.001), respectively.
When NT-proBNP and H-FABP were combined, the AUCs for arrhythmia, ischemic stroke, and renal insufficiency were 0.797 (P < 0.001), 0.789 (P = 0.012), and 0.762 (P < 0.001), respectively, and the combined prediction efficacy was greater than that of the prediction alone. When NT-proBNP and AUB-HAS2 were combined, they had good predictive value for arrhythmia, all-cause death, ischemic stroke, cardiac dysfunction, renal dysfunction, and multiple organ dysfunction syndrome (AUC > 0.75). When H-FABP and AUB-HAS2 were combined, the AUC for other types of arrhythmias in addition to atrial fibrillation was 0.761 (P < 0.001), the AUC for cardiac arrest was 0.816 (P = 0.015), the AUC for ischemic stroke was 0.876 (P = 0.001), and the AUC for renal insufficiency was 0.809 (P < 0.001). The combined predictive value of NT-proBNP, H-FABP or AUB-HAS2 was greater than that of either parameter alone. The results of the ROC analysis are shown in Table 2 and SDC-Figure A.1.
Changes in the cutoff values after AUB-HAS2 stratification
Patients were classified as low-risk (AUB-HAS2 < 2) or medium- to high-risk (AUB-HAS2 ≥ 2) based on the AUB-HAS2 scores. The cutoff concentration for NT-proBNP for predicting MACEs was 301.79 pg/ml. After stratification by the AUB-HAS2, the AUC of NT-proBNP for predicting MACEs in the middle- and high-risk groups was 0.781 (P < 0.001), the cutoff concentration was 300.40 pg/ml, and it exhibited the best combined sensitivity (0.862) and specificity (0.761). The cutoff concentration of NT-proBNP for predicting MACEs in the low-risk group was 166.98 pg/ml (AUC = 0.700, P < 0.001). For arrhythmias, in the middle- and high-risk groups, the cutoff concentration of NT-proBNP was 314.980 pg/ml (AUC = 0.791, P < 0.001).
The cutoff concentration for H-FABP in predicting MACEs in the middle- and high-risk groups was 3.07 ng/ml (AUC = 0.682, P < 0.001), which was less than the normal value given by the manufacturer, and H-FABP exhibited the best combined sensitivity (0.690) and specificity (0.710). The AUC of H-FABP for predicting cardiac arrest was 0.987 (P < 0.001), with a cutoff concentration of 7.370 ng/ml. The cutoff concentrations for both NT-proBNP and H-FABP for predicting cardiac insufficiency were reduced in the intermediate-to-high-risk group. The results of the ROC analysis after AUB-HAS2 stratification are shown in Table 3 and SDC-Figure A.2.
Table 3.
Changes in cutoff values for NT-proBNP and H-FABP predicting 30-day postoperative maces and complications after risk stratification according to the AUB-HAS2 score.
| Outcomes | All Surgeries | AUB-HAS2<2 | AUB-HAS2 ≥ 2 | ||||
|---|---|---|---|---|---|---|---|
| NT-proBNP | H-FABP | NT-proBNP | H-FABP | NT-proBNP | H-FABP | ||
| MACEs | AUC (95%CI) | 0.763(0.695, 0.832) | 0.628(0.558, 0.698) | 0.700(0.676, 0.724) | 0.554(0.528, 0.580) | 0.781(0.730, 0.826) | 0.682(0.626, 0.734) |
| P value | <0.001 | <0.001 | <0.001 | 0.269 | <0.001 | <0.001 | |
| Cut-off | 301.790 | 2.260 | 166.980 | 2.260 | 300.400 | 3.070 | |
| Arrhythmia | AUC (95%CI) | 0.783(0.708, 0.858) | 0.607(0.530, 0.685) | 0.731(0.707, 0.753) | 0.541(0.515, 0.567) | 0.791(0.741, 0.835) | 0.649(0.593, 0.703) |
| P value | <0.001 | 0.006 | <0.001 | 0.459 | <0.001 | 0.007 | |
| Cut-off | 319.630 | 2.280 | 254.780 | 2.290 | 314.980 | 3.070 | |
| Atrial fibrillation | AUC (95%CI) | 0.785(0.696, 0.874) | 0.537(0.444, 0.629) | 0.752(0.729, 0.774) | 0.535(0.509, 0.561) | 0.781(0.730, 0.826) | 0.599(0.542, 0.655) |
| P value | <0.001 | 0.434 | <0.001 | 0.570 | <0.001 | 0.153 | |
| Cut-off | 326.310 | 2.290 | 280.320 | 1.420 | 386.300 | 2.290 | |
| Other types of arrhythmias | AUC (95%CI) | 0.796(0.674, 0.918) | 0.744(0.651, 0.836) | 0.724(0.700, 0.747) | 0.745(0.721, 0.767) | 0.784(0.734, 0.829) | 0.696(0.641, 0.747) |
| P value | <0.001 | <0.001 | 0.044 | <0.001 | <0.001 | 0.015 | |
| Cut-off | 319.630 | 2.280 | 191.940 | 2.360 | 314.980 | 3.110 | |
| Death | AUC (95%CI) | 0.720(0.527, 0.913) | 0.553(0.254, 0.852) | 0.729(0.705, 0.752) | 0.582(0.556, 0.608) | 0.617(0.560, 0.672) | 0.675(0.620, 0.728) |
| P value | 0.063 | 0.653 | 0.115 | 0.690 | 0.599 | 0.548 | |
| Cut-off | 100.230 | 5.480 | 109.330 | 2.680 | 300.400 | 5.260 | |
| Cardiac arrest | AUC (95%CI) | 0.833(0.746, 0.920) | 0.796(0.638, 0.954) | 0.842(0.822, 0.861) | 0.685(0.660, 0.709) | 0.706(0.652, 0.757) | 0.987(0.967, 0.996) |
| P value | 0.010 | 0.022 | <0.001 | 0.012 | <0.001 | <0.001 | |
| Cut-off | 174.270 | 2.320 | 174.270 | 2.320 | 300.400 | 7.370 | |
| Myocardial infarction | AUC (95%CI) | 0.661(0.371, 0.950) | 0.647(0.432, 0.861) | 0.567(0.540, 0.592) | 0.655(0.630, 0.680) | 0.673(0.618, 0.726) | 0.616(0.559, 0.671) |
| P value | 0.174 | 0.214 | 0.818 | 0.012 | 0.478 | 0.666 | |
| Cut-off | 522.410 | 2.260 | 166.980 | 2.260 | 518.890 | 3.520 | |
| Stroke | AUC (95%CI) | 0.688(0.493, 0.883) | 0.799(0.722, 0.877) | 0.705(0.681, 0.728) | 0.855(0.836, 0.873) | 0.538(0.480, 0.595) | 0.667(0.611, 0.720) |
| P value | 0.111 | 0.011 | 0.269 | <0.001 | 0.854 | 0.031 | |
| Cut-off | 176.470 | 2.570 | 195.810 | 2.970 | 982.180 | 2.570 | |
| Cardiac insufficiency | AUC (95%CI) | 0.679(0.590, 0.768) | 0.582(0.477, 0.688) | 0.629(0.603, 0.654) | 0.501(0.475, 0.527) | 0.649(0.593, 0.703) | 0.624(0.567, 0.678) |
| P value | <0.001 | 0.071 | 0.039 | 0.993 | 0.056 | 0.079 | |
| Cut-off | 474.290 | 3.180 | 136.480 | 2.900 | 461.650 | 3.170 | |
| Renal inadequacy | AUC (95%CI) | 0.832(0.749, 0.916) | 0.701(0.577, 0.826) | 0.831(0.811, 0.850) | 0.613(0.587, 0.638) | 0.788(0.737, 0.832) | 0.751(0.698, 0.798) |
| P value | <0.001 | <0.001 | <0.001 | 0.269 | <0.001 | 0.002 | |
| Cut-off | 278.060 | 2.940 | 168.000 | 2.610 | 300.400 | 3.350 | |
| MODS | AUC (95%CI) | 0.790(0.609, 0.970) | 0.668(0.356, 0.979) | 0.729(0.705, 0.752) | 0.653(0.628, 0.678) | 0.706(0.652, 0.757) | 0.843(0.798, 0.882) |
| P value | 0.025 | 0.194 | 0.362 | 0.649 | 0.145 | 0.006 | |
| Cut-off | 180.990 | 2.670 | 699.620 | 0.500 | 180.930 | 5.260 | |
Multivariate logistic regression and model development
Univariate analysis of each component was first performed to investigate the risk factors associated with MACEs. NT-proBNP, H-FABP, AUB-HAS2, age, general condition, NYHA cardiac functional grade, exercise equivalent, renal function grade, ASA grade, hypertension, history of ischemic heart disease, and albumin and creatinine were all predictors of MACEs (SDC-Table A.2).
To adjust for the effects of confounding factors, forward stepwise regression was used to include covariates to construct a multivariate logistic regression model. The results revealed that a history of dilated cardiomyopathy (OR = 15.174, 95% CI 2.922–78.813, P = 0.001), preoperative arrhythmia (OR = 2.685, 95% CI 1.391–5.183, P = 0.003), diabetes (OR = 1.929, 95% CI 1.105–3.368, P = 0.021) and peripheral vascular disease (OR = 4.940, 95% CI 1.518–16.083, P = 0.008) showed statistically significant differences in the model and were risk factors for MACEs (SDC-Table A.3). And the subgroup analysis of operative type showed that the predictive value of NT-proBNP and H-FABP was not significantly changed in abdominal surgery and urinary surgery. The modified AUB-HAS2 combines the above risk factors, NT-proBNP stratification and the AUB-HAS. NT-proBNP was stratified by 166 pg/ml and 300 pg/ml based on the cutoff concentrations of NT-BNP in the low-risk and mid-to-high-risk groups, as shown in Table 3. The modified AUB-HAS2 scale is shown in Table 4. The ability of the modified AUB-HAS2 to predict MACEs increased (AUC = 0.794, 95% CI = 0.737–0.851, P < 0.001), as shown in Fig. 2; Table 2. And the difference from the traditional model score was statistically significant (P<0.001)(SDC-table A.4).
Table 4.
Elements and risk stratification of the AUB-HAS2 and modified AUB-HAS2.
| AUB-HAS2 cardiovascular risk index | Modified AUB-HAS2 cardiovascular risk index |
|---|---|
|
• History of heart disease (1 points) • Symptoms of heart disease (angina or dyspnea) (1 points) • Age ≥ 75 years (1 points) • Anemia (hemoglobin < 12 mg/dL) (1 points) • Vascular surgery (1 points) • Emergency surgery (1 points) |
• Dilated cardiomyopathy (1 points) • Arrhythmia (1 points) • History of heart disease except dilated cardiomyopathy and arrhythmia (1 points) • Symptoms of heart disease (angina or dyspnea) (1 points) • Age ≥ 75 years (1 points) • Anemia (hemoglobin < 12 mg/dL) (1 points) • Peripheral vascular disease (1 points) • Diabetes (1 points) • Vascular surgery (1 points) • Emergency surgery (1 points) • NT-proBNP: <166pg/ml (1 points) 166-300pg/ml (2 points) >300 pg/ml (3 points) |
|
• Low risk (score 0–1) • Intermediate risk (score 2–3) • High risk (score > 3) |
• Low risk (score 0–3) • Intermediate risk (score 4–6) • High risk (score > 6) |
Fig. 2.
Results of ROC analysis for modified AUB-HAS2 scale predicting MACEs within 30 days after surgery.
Discussion
This study demonstrated that the preoperative NT-proBNP concentration was highly valuable for predicting postoperative MACEs, but the H-FABP and AUB-HAS2 levels were poorly able to predict postoperative MACEs. However, the predictive value of NT-proBNP, H-FABP and AUB-HAS2 was greater than that of any single index. When H-FABP was combined with the preoperative AUB-HAS2, its ability to predict MACEs significantly improved. Our study extends the findings of other cohort studies in this area, most of which are focused on preoperative NT-proBNP versus MACE studies19–21 and the use of H-FABP in acute pulmonary embolism22,23. According to our results, the modified AUB-HAS2 scale was developed with the joint application of each index, and the predictive value of the improved scale for MACEs was significantly improved. Therefore, this study not only revealed the predictive and prognostic value of the combination of H-FABP, NT-proBNP and preoperative AUB-HAS2 scores but also established an optimal strategy for the preoperative prediction of MACEs.
NT-proBNP is a polypeptide hormone synthesized and secreted by cardiomyocytes, whose main function is to regulate the hemodynamics of the heart. When the heart is subjected to increased pressure load, myocardial ischemia, or heart failure, cardiomyocytes synthesize and release more NT-proBNP. Therefore, increased preoperative NT-proBNP levels will increase the risk of postoperative cardiovascular complications9. In this study, the AUCs of NT-proBNP and NT-proBNP AUB-HAS2 for predicting MACEs were 0.763 and 0.781, respectively. Nielsen H et al. demonstrated that the AUC of NT-proBNP for predicting MACE in geriatric patients aged 60 to 74 years was 0.8219. Another study showed that the AUC of long-term cardiac events by BNP was 0.71 in 270 patients24. The predictive value of NT-proBNP varies greatly among different studies, and the results of this study provide new evidence supporting the use of NT-proBNP for the prediction of MACEs. In this study, we found that the cutoff value for NT-proBNP lies within the reference range. From the cut-off value, we identified a proportion of high-risk patients whose NT-proBNP examination was normal. For such patients, it helps to identify the occurrence of MACEs.
The H-FABP concentration was better at predicting cardiac arrest (AUC = 0.796) and ischemic stroke (AUC = 0.799) and was much better for other cardiovascular events in combination with the AUB-HAS2, with the AUC of MACEs increasing from 0.628 to 0.736. The combined AUB-HAS2 score significantly improved the predictive efficacy of H-FABP, which was similar to the results of NT-proBNP and AUB-HAS2. The prediction efficacy did not improve further when the three indicators were used. Therefore, clinically, we selected only a single biomarker and the AUB-HAS2 score to achieve a better prediction effect. In this study, when H-FABP was calculated according to the normal reference value of 6.8 ng/ml given by the manufacturer, there was no good discrimination for predicting the risk of postoperative complications, and its prediction effect was inferior to that of NT-proBNP. On the one hand, this prompted us to set the cutoff concentration of H-FABP according to the different complications.
On the other hand, mechanistically, the lower predictive value of H-FABP than of NT-proBNP may be related to the mechanism of action and release of H-FABP. Approximately 50–80% of the energy of the heart is provided by lipid oxidation, and cardiac FA-binding protein (H-FABP) ensures the intracellular transport of insoluble fatty acids; therefore, H-FABP is particularly important for myocardial homeostasis25. Its ability to be released from the injured myocardium was very similar to that of myoglobin, a sensitive marker of myocardial injury, and its cardiac specificity was greater than that of myoglobin26. As a cytoplasmic protein with a molecular weight of only 1,215 kDa, its molecular size determines that it can be released very early after myocardial injury27. It appears in the circulation as early as 90 min after symptoms and peaks within 6 h25. These features make H-FABP an excellent candidate marker for myocardial injury. However, most patients in this study may not have experienced significant myocardial injury before surgery, and the incidence of postoperative myocardial infarction was also low; therefore, the predictive value of the preoperative H-FABP concentration for MACEs may be inferior to that of NT-proBNP. However, our findings also suggest that we can detect the H-FABP concentration at 6 h after surgery in the future and that the predictive value of H-FABP for myocardial injury may be better than that of myoglobin. Some data suggest that the applicability of H-FABP testing in patients with skeletal muscle injury, as well as in patients with renal failure, may be limited28,29. Most of the patients included in this study underwent orthopedic surgery, which may have caused skeletal muscle damage, and the patients showed significant differences in the baseline level of renal function, which may also have led to the reduced predictive effect of H-FABP. However, when H-FABP was combined with AUB-HAS2, the results in combination with NT-proBNP and AUB-HAS2 were similar, indicating that the AUB-HAS2 scoring system effectively compensated for the defects in H-FABP prediction.
The AUB-HAS2 score is an assessment tool that takes into account multiple risk factors that may have adverse effects on the cardiovascular system and increase the risk of postoperative cardiovascular complications. The overall cardiovascular risk can be quantified by AUB-HAS2 score6. When we used the AUB-HAS2 score to classify patients into low-risk and intermediate-high-risk groups, the predictive value of NT-proBNP and H-FABP for MACEs increased for medium- and high-risk patients. However, the cutoff values of NT-proBNP and H-FABP were greater in the high-risk group than in the low-risk group, probably because of the overall greater NT-proBNP and H-FABP values in high-risk patients. However, it is worth noting that for atrial fibrillation, cardiac arrest and renal dysfunction, NT-proBNP also had a good predictive effect in the low-risk and low-risk groups, and the cutoff value was lower than the overall cutoff value. This suggests that in clinical application, the cutoff of biomarkers should be set separately according to low-risk and medium-high-risk populations. Multiple logistic regression analysis revealed that patients with dilated cardiomyopathy had a 15.174-fold greater risk of MACEs, patients with a history of arrhythmias had a 2.685-fold greater risk of MACEs, and patients with peripheral vascular disease or diabetes mellitus had a 4.940- and 1.929-fold greater risk of postoperative MACEs. The modified AUB-HAS2 combines each risk factor and NT-proBNP, which significantly increases its predictive effect or prognostic value for postoperative MACEs.
Compared with other studies, this study had a larger sample size and more reliable results. There are relatively few previous studies on the relationship between H-FABP and cardio-cerebrovascular complications after non-cardiac surgery, and there are no reports on the ability of the AUB-HAS2 score or the combination of NT-proBNP or H-FABP to predict MACEs. Therefore, this study clarified the important value of the preoperative combination of NT-proBNP, H-FABP or AUB-HAS2 for predicting postoperative MACEs and revealed that AUB-HAS2 combined with H-FABP improves the predictive efficacy of H-FABP. The results of this study can be translated into a clinically meaningful modified AUB-HAS2 scale for clinical application in the future.
There are several limitations in the current study. First, this is a single-center prospective cohort study that can only identify associations, not chance associations. The generalizability of the current observations may need to be confirmed in multicenter studies. Second, the presence of mismatched baseline data in this study makes the results susceptible to confounding factors. However, we excluded the influence of confounding factors as much as possible by statistical methods. Third, excluding patients with surgery time less than 90 min may affect the applicability of conclusions in patients with short surgery. Surgical time is a risk factor affecting the occurrence of MACEs, and our findings may only be apply to patients in major non-cardiac surgery. In addition, cancer-related surgery was associated with MACEs, but there was no statistically significant difference between the two groups in this study. As most elderly patients underwent cancer-related surgery, the sample size was not large enough to perform further stratified analysis. Moreover, the type of surgery may potentially impact on our outcome. Through subgroup analysis, we found no statistically significant effect of the type of surgery on the outcome in this study. However, the subgroup analysis may not yield reliable conclusions due to the small sample size of each subgroup in this study. Further studies are needed to investigate the impact.
Conclusion
Our study revealed the predictive efficacy and prognostic value of NT-proBNP, H-FABP and the AUB-HAS2 score alone or in combination for postoperative MACE risk assessment in geriatric patients undergoing non-cardiac surgery. NT-proBNP should be measured before surgery in geriatric patients undergoing non-cardiac surgery, and the modified AUB-HAS2 should be routinely evaluated. This study also suggested that further studies are needed to evaluate whether H-FABP concentrations within 6 h can guide intraoperative and early postoperative management in geriatric patients and whether reducing NT-proBNP and H-FABP concentrations by treatment can improve postoperative clinical outcomes.
Methods
Study design and patients
This prospective single-center cohort study was conducted between June 2020 and December 2021 in a comprehensive Grade A tertiary hospital. Patients who were scheduled for elective non-cardiac surgery under general anesthesia were continuously included. The inclusion criteria were patients who underwent non-cardiac surgery between June 2020 and Dec 2021, who were aged ≥ 65 years, who were ASA I-IV, who were male or female, who underwent elective surgery, and who were mentally competent enough to provide informed written consent. The exclusion criteria were surgery < 90 min, emergency surgery, sedative or psychotropic drug dependence, mental disorders, and a history of allergy to narcotics. Consent was provided by all participating patients or their guardians.
Ethics approval
Ethics approval for this study was obtained from the Ethics Committee on Biomedical Research, West China Hospital of Sichuan University, Sichuan, China (No. 2019 − 624). This trial was registered at the Chinese Clinical Trial Registry (ChiCTR1900026223, https://www.chictr.org.cn). Written informed consent was obtained from all study volunteers. And we confirmed that all experiments were performed in accordance with relevant guidelines and regulations.
Anesthesia and surgical procedure
Every patient underwent a routine preoperative assessment, which included a 12-lead ECG, laboratory blood testing, a physical examination, and a medical history. If additional testing was indicated, the surgical team carried it out. Some responsible anesthesiologist was asked to subjectively assess the patient the day before surgery. Comorbidities were determined by self-reports and medical history.
All patients were routinely monitored by pulse oximetry, ECG, noninvasive arterial blood pressure measurements, and bispectral indices (BIS, Covidien LLC, MA, US) in the operating room, and invasive arteriovenous monitoring was used if necessary. General anesthesia was administered by intravenous injection of 1.5 ~ 2.5 mg/kg propofol followed by 0.4 µg/kg sufentanil, 0.2 mg/kg cisatracurium, or 0.1 mg/kg vecuronium. After tracheal intubation, the patient was ventilated using 0.5% of the inspired oxygen fraction and fresh gas flows of 2 L/min of oxygen and air. Anesthesia was maintained with intravenous infusion of remifentanil and sevoflurane or desflurane inhalation or propofol target controlled infusion (TCI). A target BIS value of 40–60 was maintained by adjusting the dose of general anesthetics. After postoperative skin closure, the treatment with desflurane and remifentanil was terminated. A treatment of 20 µg/kg neostigmine combined with 10 µg/kg atropine was given to antagonize the residual neuromuscular blockade unless contraindicated. The patient was transferred to the post-anesthesia care unit (PACU) or ward after tracheal extubation. Some patients were referred directly to the intensive care unit if intensive monitoring and life support were needed.
AUB-HAS2, NT-proBNP and H-FABP tests
All patients underwent preoperative risk assessment for AUB-HAS2. All patients underwent venous blood measurements for NT-proBNP and H-FABP concentrations within 1 h before surgery. Blood samples were collected in an EDTA vacuum tube, immediately centrifuged, and then analyzed in the laboratory using a Finecare3 + FS-205 immunofluorescent autoanalyzer (Guangzhou Wanfu Biotechnology Co., Ltd., Guangzhou, China). The normal reference ranges for NT-proBNP and H-FABP were 0–300 pg/ml and 0–6.8 ng/ml, respectively.
Patient follow-up and outcomes
The patient follow-up study investigators followed patients once daily during their hospital stay to determine the presence of postoperative complications. After discharge, the patient was contacted by telephone at 30 days after surgery to confirm the incidence of complications after discharge. The major postoperative complications were based on follow-up during hospitalization and telephone follow-up after discharge.
The primary outcomes were MACEs within 30 days after surgery, defined as the onset or progression of arrhythmia, inpatient all-cause mortality, cardiac arrest, acute myocardial infarction (AMI), and ischemic stroke30. Postoperative arrhythmia refers to the progression and aggravation of newly developed arrhythmia or electrocardiogram abnormalities and preoperatively diagnosed arrhythmia and is mainly divided into atrial fibrillation and other types of arrhythmias. Cardiac arrest refers to cardiac arrest of various causes with loss of consciousness and/or advanced cardiac life support31. Myocardial infarction was defined as a cardiac biomarker exceeding the upper 99th percentile plus at least one evidence of myocardial ischemia (including symptoms, ECG ischemic changes, pathological Q wave or imaging evidence)32. The diagnostic criteria for ischemic stroke were the latest definition of stroke in the 21st century published by the American Heart Association/American Stroke Association33,34.
The secondary outcomes were other complications within 30 days after surgery, including cardiac insufficiency, renal insufficiency, and multiple organ dysfunction syndrome (MODS). Cardiac insufficiency occurs due to various reasons, such as a decrease in the systolic function of the heart muscle and a decrease in blood discharge from the heart, resulting in blood stasis in the systemic circulation or pulmonary circulation symptoms. Postoperative renal dysfunction mainly includes acute kidney injury and the exacerbation of chronic renal insufficiency. Multiple organ dysfunction syndrome (MODS) refers to the acute dysfunction or failure of more than one system or/or organ in the course of acute diseases such as severe infection, trauma or major surgery35.
Statistical analysis
The sample size was calculated based on the area under the curve (AUC) comparing the receiver operating characteristic (ROC) curve using MedCalc (version 20). Based on literature reports of non-cardiac surgery mortality rates ranging from 0.8-1.5%1, the incidence of mortality events in this study was assumed to be 0.8%. The moderately good AUC of blood NT-proBNP or H-FABP concentrations was 0.75, and a sample size of 1750 patients with 90% power was used to detect clinically relevant differences in AUC values (bilateral alpha of 0.05). To account for 10% of patients potentially lost to follow-up, we aimed to recruit a total of 1925 patients into the study. The analyses described in the manuscript were performed using the complete-case method.
Statistical analysis was performed using SPSS 29.0 (IBM SPSS Statistics Corporation, Armonk, NY). Statistical tests were two-sided, and a P value < 0.05 was considered to indicate statistical significance. Continuous normally distributed variables are reported as the mean ± standard deviation (SD) and were tested using Student’s t test and ANOVA for multiple comparisons. Continuously skewed variables are presented as medians (interquartile spacings) or medians (P25, P75) and were compared using the rank sum test. Categorical variables are reported as percentages and were compared using a chi-square test. ROC curve analysis was used to determine the predictive value of NT-proBNP and H-FABP concentrations and the AUB-HAS2 score for postoperative complications. The performance of continuous variables was analyzed using receiver operating characteristic (ROC) curves, and the area under the curve (AUC) was calculated. Considering the reference range of the indicators, and the sensitivity and specificity of MACEs diagnosis are equally important, we used the value corresponding to the maximum Youden index (Sensitivity + specificity-1) in the ROC analysis as the best cutoff value. The relative risk (RR) and 95% confidence interval (CI) were determined using the chi-square test.
In order to modify the scale, the univariate logistic regression was first performed. The selection of confounding variables was based on previous evidence, which reflects the clinically relevant factors usually considered in preoperative evaluation. Next, we adjusted the confounding factors by forward stepwise multivariate logistics regression. The AUB-HAS2 scale was modified according to the analysis results. ROC analysis and paired-sample T-test were used to analyze the significance of the difference in prediction performance between modified and conventional models. Subgroup analysis was further applied to explore impact of surgery type on the model performance and cutoff value.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Author contributions
Data curation: [Xialian Hu, Yi Zhao]; Formal analysis: [Xialian Hu]; Methodology: [Xialian Hu, Yi Zhao], Writing original draft: [Xialian Hu]; Investigation: [Yi Zhao]; Study conception and design: [Mengchan Ou]; Visualization and nvestigation: [Xuechao Hao]; Supervision: [Tao Zhu]. All the authors contributed to the refinement of the study protocol and approved the final manuscript.
Funding
This work was supported by the by Sichuan Natural Science Foundation project (2023NSFSC1647); the National Natural Science Foundation of China (Beijing, China) (grant number 82371280); and 135 Project for Disciplines of Excellence, West China Hospital, Sichuan University (grant number ZYJC21008); the Postdoctoral Researcher Fund of West China Hospital (grant number 2023HXBH131); and the Science and Technology Department of Sichuan Province, China (grant number 2023NSFSC1565).
Data availability
The data underlying this article are available in the article and in its online supplementary material. Request for additional information can be made to the corresponding author.
Declarations
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.
Xialian Hu, Yi Zhao and Mengchan Ou have contributed equally to the research and share first authorship.
Contributor Information
Tao Zhu, Email: 739501155@qq.com.
Xuechao Hao, Email: aneshxc@163.com.
References
- 1.Haynes, A. B. et al. A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl. J. Med.360(5), 491–499 (2009). [DOI] [PubMed] [Google Scholar]
- 2.Weiser, T. G. et al. Estimate of the global volume of surgery in 2012: an assessment supporting improved health outcomes. Lancet385(Suppl 2), S11 (2015). [DOI] [PubMed] [Google Scholar]
- 3.Garutti, I. et al. Analytic comments of the initial results of the VISION (Vascular events in noncardiac surgery patients cOhort evaluatioN): Association between troponin T and mortality at 30 days in noncardiac surgery patients. Rev. Esp. Anestesiol. Reanim. 60(6), 304–307 (2013). [DOI] [PubMed] [Google Scholar]
- 4.Devereaux, P. J. & Sessler, D. I. Cardiac complications in patients undergoing major noncardiac surgery. N. Engl. J. Med.373(23), 2258–2269 (2015). [DOI] [PubMed] [Google Scholar]
- 5.Fleischmann, K. E., Goldman, L., Young, B. & Lee, T. H. Association between cardiac and noncardiac complications in patients undergoing noncardiac surgery: outcomes and effects on length of stay. Am. J. Med.115(7), 515–520 (2003). [DOI] [PubMed] [Google Scholar]
- 6.Fleisher, L. A. et al. 2014 ACC/AHA guideline on perioperative cardiovascular evaluation and management of patients undergoing noncardiac surgery: a report of the American college of cardiology/american heart association task force on practice guidelines. Circulation130(24), e278–333 (2014). [DOI] [PubMed] [Google Scholar]
- 7.Dakik, H. A. et al. AUB-HAS2 cardiovascular risk index: performance in surgical subpopulations and comparison to the revised cardiac risk index. J. Am. Heart Assoc.9(10), e016228 (2020). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Santangelo, G. et al. Risk of cardiovascular complications during non-cardiac surgery and preoperative cardiac evaluation. Trends Cardiovasc. Med.32(5), 271–284 (2022). [DOI] [PubMed] [Google Scholar]
- 9.Duceppe, E. et al. Preoperative N-Terminal Pro-B-Type natriuretic peptide and cardiovascular events after noncardiac surgery: A cohort study. Ann. Intern. Med.172(2), 96–104 (2020). [DOI] [PubMed] [Google Scholar]
- 10.Zile, M. R. et al. Prognostic implications of changes in N-Terminal Pro-B-Type natriuretic peptide in patients with heart failure. J. Am. Coll. Cardiol.68(22), 2425–2436 (2016). [DOI] [PubMed] [Google Scholar]
- 11.Cullen, L. et al. Validation of high-sensitivity troponin I in a 2-hour diagnostic strategy to assess 30-day outcomes in emergency department patients with possible acute coronary syndrome. J. Am. Coll. Cardiol.62(14), 1242–1249 (2013). [DOI] [PubMed] [Google Scholar]
- 12.Yiu, K. H. et al. Predictive value of high-sensitivity troponin-I for future adverse cardiovascular outcome in stable patients with type 2 diabetes mellitus. Cardiovasc. Diabetol.13, 63 (2014). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Choi, J. H. et al. Preoperative NT-proBNP and CRP predict perioperative major cardiovascular events in non-cardiac surgery. Heart96(1), 56–62 (2010). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Chmurzynska, A. The multigene family of fatty acid-binding proteins (FABPs): function, structure and polymorphism. J. Appl. Genet.47(1), 39–48 (2006). [DOI] [PubMed] [Google Scholar]
- 15.Ruan, L. B., He, L., Zhao, S., Zhu, P. & Li, W. Y. Prognostic value of plasma heart-type fatty acid-binding protein in patients with acute pulmonary embolism: a meta-analysis. Chest146(6), 1462–1467 (2014). [DOI] [PubMed] [Google Scholar]
- 16.Ye, X. D. et al. Heart-type fatty acid binding protein (H-FABP) as a biomarker for acute myocardial injury and long-term post-ischemic prognosis. Acta Pharmacol. Sin. 39(7), 1155–1163 (2018). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Jiang, D. et al. Serum heart-type fatty acid-binding protein as a predictor for the development of sepsis-associated acute kidney injury. Expert Rev. Mol. Diagn.19(8), 757–765 (2019). [DOI] [PubMed] [Google Scholar]
- 18.Jiang, M., Li, Y., Cao, L., Tian, J. & Wang, D. Relations of heart-type and brain-type fatty acid-binding proteins with postoperative cognitive dysfunction in elderly patients undergoing spinal surgery. Rev. Assoc. Med. Bras. (1992). 67(3), 390–394 (2021). [DOI] [PubMed] [Google Scholar]
- 19.Wendelboe Nielsen, O., Kirk, V., Bay, M., Boesgaard, S. & Nielsen, H. Value of N-terminal pro brain natriuretic peptide in the elderly: data from the prospective Copenhagen hospital heart failure study (CHHF). Eur. J. Heart Fail.6(3), 275–279 (2004). [DOI] [PubMed] [Google Scholar]
- 20.Karthikeyan, G. et al. Is a pre-operative brain natriuretic peptide or N-terminal pro-B-type natriuretic peptide measurement an independent predictor of adverse cardiovascular outcomes within 30 days of noncardiac surgery? A systematic review and meta-analysis of observational studies. J. Am. Coll. Cardiol.54(17), 1599–1606 (2009). [DOI] [PubMed] [Google Scholar]
- 21.Farzi, S. et al. Role of N-terminal pro B-type natriuretic peptide in identifying patients at high risk for adverse outcome after emergent non-cardiac surgery. Br. J. Anaesth.110(4), 554–560 (2013). [DOI] [PubMed] [Google Scholar]
- 22.Puls, M. et al. Heart-type fatty acid-binding protein permits early risk stratification of pulmonary embolism. Eur. Heart J.28(2), 224–229 (2007). [DOI] [PubMed] [Google Scholar]
- 23.Dellas, C. et al. A novel H-FABP assay and a fast prognostic score for risk assessment of normotensive pulmonary embolism. Thromb. Haemost. 111(5), 996–1003 (2014). [DOI] [PubMed] [Google Scholar]
- 24.Breidthardt, T. et al. B-type natriuretic peptide in patients undergoing orthopaedic surgery: a prospective cohort study. Eur. J. Anaesthesiol.27(8), 690–695 (2010). [DOI] [PubMed] [Google Scholar]
- 25.Alhadi, H. A. & Fox, K. A. Do we need additional markers of myocyte necrosis: the potential value of heart fatty-acid-binding protein. QJM97(4), 187–198 (2004). [DOI] [PubMed] [Google Scholar]
- 26.Pelsers, M. M., Hermens, W. T. & Glatz, J. F. Fatty acid-binding proteins as plasma markers of tissue injury. Clin. Chim. Acta. 352(1–2), 15–35 (2005). [DOI] [PubMed] [Google Scholar]
- 27.Storch, J. & Thumser, A. E. The fatty acid transport function of fatty acid-binding proteins. Biochim. Biophys. Acta. 1486(1), 28–44 (2000). [DOI] [PubMed] [Google Scholar]
- 28.Van Nieuwenhoven, F. A. et al. Discrimination between myocardial and skeletal muscle injury by assessment of the plasma ratio of myoglobin over fatty acid-binding protein. Circulation92(10), 2848–2854 (1995). [DOI] [PubMed] [Google Scholar]
- 29.Gorski, J., Hermens, W. T., Borawski, J., Mysliwiec, M. & Glatz, J. F. Increased fatty acid-binding protein concentration in plasma of patients with chronic renal failure. Clin. Chem.43(1), 193–195 (1997). [PubMed] [Google Scholar]
- 30.Smilowitz, N. R. et al. Perioperative major adverse cardiovascular and cerebrovascular events associated with noncardiac surgery. JAMA Cardiol.2(2), 181–187 (2017). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Tsutsui, H. et al. JCS/JHFS 2021 guideline focused update on diagnosis and treatment of acute and chronic heart failure. Circ. J.85(12), 2252–2291 (2021). [DOI] [PubMed] [Google Scholar]
- 32.Hicks, K. A. et al. 2014 ACC/AHA key data elements and definitions for cardiovascular endpoint events in clinical trials: A report of the American college of cardiology/american heart association task force on clinical data standards (Writing committee to develop cardiovascular endpoints data standards). Circulation132(4), 302–361 (2015). [DOI] [PubMed] [Google Scholar]
- 33.Thygesen, K. et al. Third universal definition of myocardial infarction. Glob Heart. 7(4), 275–295 (2012). [DOI] [PubMed] [Google Scholar]
- 34.Sacco, R. L. et al. An updated definition of stroke for the 21st century: a statement for healthcare professionals from the American heart association/american stroke association. Stroke44(7), 2064–2089 (2013). [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Baue, A. E. MOF, MODS, and SIRS: what is in a name or an acronym? Shock26(5), 438–449 (2006). [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The data underlying this article are available in the article and in its online supplementary material. Request for additional information can be made to the corresponding author.