Introduction

The extensive application of standardized anticoagulation and antiplatelet therapy has significantly improved the clinical prognosis of acute myocardial infarction (AMI), albeit with an increased risk of bleeding.1,2,3 Periprocedural bleeding, a common complication in AMI patients undergoing percutaneous coronary intervention (PCI), has been found to be significantly associated with poor prognosis, including increased mortality.3,4,5,6,7,8,9,10

At present, a hemoglobin drop threshold of 3 to 5 g/dL is mostly used to grade the severity of bleeding. Specifically, a hemoglobin drop of 5 g/dL without overt bleeding, or a drop of 3 g/dL with overt bleeding, is commonly defined as major bleeding and used as endpoint events in most clinical trials and registries.5,11,12,13 Nevertheless, the incidence of major bleeding events was relatively low in the clinical practice, with most patients suffering from a hemoglobin decline in the absence of overt bleeding.14,15 The data regarding the association between in-hospital hemoglobin drop without overt bleeding and clinical outcomes in AMI patients remain limited.

In this retrospective study, we aimed to investigate the impacts of in-hospital hemoglobin drop without overt bleeding on 1-year clinical outcomes of AMI patients after PCI.

Results

Demographic and baseline characteristics

A total of 5036 AMI enrolled in final analysis with a 99.3% follow-up rate at 1 year. The flowchart is presented in Supplemental Fig. 1. The demographic and baseline characteristics are shown in Table 1. The level of hemoglobin at admission and nadir during hospitalization are shown in Supplementary Materials Table S1. Baseline characteristics varied considerably among the three groups. In general, with incremental hemoglobin drop stratification, patients were more likely to present with STEMI, as well as lower eGFR (91.51 ± 24.00 vs. 88.73 ± 24.10 vs. 83.54 ± 25.44 ml/min/1.73 m2, P < 0.0001) and lower LVEF values (55.29 ± 8.73 vs. 55.21 ± 8.88 vs. 53.48 ± 9.39%, P < 0.0001). Besides, lower proportion of PCI performed with trans-radial access, and an increase in the proportion of right coronary artery as a target lesion were found in patients with incremental hemoglobin drop (Table 2). Regarding discharge medications, the use of aspirin, statin and ACEI/ARB decreased with the degree of hemoglobin drop increased (Table 3).

Table 1 Demographic and baseline characteristics.
Full size table
Table 2 Procedure information.
Full size table
Table 3 Medical treatment.
Full size table

Stratified clinical outcomes

1-year Clinical outcomes of patients stratified by the degree of hemoglobin drop are reported in Table 4. There were significant differences in 1-year incidence of ischemic events (1.88% vs. 3.27% vs. 3.46%; P = 0.0114), all-cause death (1.45% vs. 2.18% vs. 2.94%; P = 0.0128), as well as cardiac death (1.15% vs. 1.82% vs. 2.37%; P = 0.0282) and MI (0.30% vs. 1.15% vs. 0.92%; P = 0.0175) across tertiles of hemoglobin drop values. However, no significant differences were observed in bleeding events. Time-to-event curves are shown in Fig. 1.

Fig. 1
figure 1

Kaplan–Meier survival analysis. Time-to-event curves for the incidences of (a) the primary endpoint – ischemic events, (b) all-cause death.

Full size image
Table 4 Clinical outcomes at 1-year follow-up.
Full size table

Considering hemoglobin drop as a continuous variable, the adjusted risk of 1-year ischemic events was increased by 2% for each 1 g/dL hemoglobin drop (HR: 1.02, 95%CI 1.001–1.04, P = 0.01). An independent association between in-hospital hemoglobin drop and 1-year all-cause death was found in AMI patients and showed that for each 1 g/dL hemoglobin drop, the adjusted risk for 1-year all-cause death was increased by 3% (HR: 1.03, 95%CI 1.01–1.04, P = 0.008) (Table 5).

Table 5 Associations between hemoglobin drop and clinical outcomes.
Full size table

The diagnostic ability of the in-hospital hemoglobin drop to discriminate patients with respect to 1-year ischemic events and all-cause death was assessed by ROC curve analysis. Applied for the whole group of patients, the AUC for 1-year ischemic events and all-cause death was 0.58 (95% CI 0.53–0.62) and 0.60 (95% CI 0.55–0.65), respectively (Fig. 2).

Fig. 2
figure 2

Receiver operating characteristic curve. Receiver operating characteristic curve for (a) the primary endpoint – Ischemic events, (b) all-cause death.

Full size image

Discussion

The results of the study, which was conducted over a cohort of AMI patients undergoing PCI, indicated that an decline in hemoglobin during hospitalization was associated with an increased risk of 1-year ischemic events and all-cause death, even in the absence of overt bleeding. Besides, for each 1 g/dL hemoglobin drop, the adjusted risk of 1-year ischemic events was 2% higher in AMI patients, as well as 3% higher adjusted risk for all-cause death.

Patients diagnosed with AMI may present with decreased hemoglobin during hospitalization due to the application of invasive treatment and antiplatelet or anticoagulant therapy, both of which can cause hemorrhage.3,14,15,16,17,18 Numerous clinical investigations have provided substantial evidence for the association between major bleeding events and adverse outcomes after PCI.6,7,8,9,10,19 However, major bleeding events were not commonly observed in the clinical practice.14,15 Hence, the attention paid to non-major bleeding has been relatively limited, resulting in the neglect of adverse outcomes caused by potential bleeding events. Nevertheless, evidence supporting the association between hemoglobin decrease without overt bleeding and clinical outcomes is insufficient.

In our analysis, there was a significant association between the decline in hemoglobin, without overt bleeding during hospitalization and an increased risk of 1-year ischemic events and all-cause death. Our findings are consistent with those of previous studies. Among ACS patients with PCI, in-hospital hemoglobin drop ≥ 3 g/dL has been found to be independently associated with increased risk for 1-year mortality in the absence of overt bleeding.14 Moreover, Ndrepepa et al. demonstrated that the association between hemoglobin decline and 1-year mortality remained significant when patients with obvious in-hospital bleeding were excluded, and the minimal decline in hemoglobin levels associated with mortality was 1.13 g/dL in patients without overt bleeding.15 Besides, a decline in hemoglobin ≥ 0.9 g/dL in patients with ACS during intensive care unit (ICU) stay was found to be associated with adverse clinical outcomes within 6 months, which provided additional prognostic information on top of the GRACE score.20 Another study focused on non-overt bleeding ICU-admitted patients with AMI revealed that in-hospital hemoglobin drop was independently associated with higher 180-day all-cause mortality.21 Hence, the above studies support the association between in-hospital hemoglobin drop and clinical events to some extent.

The mechanism of in-hospital hemoglobin drop is probably complex and influenced by multiple factors. Previous studies suggested that therapeutic fluid use,22 suppression of erythropoiesis due to cytokines generated as a response to myocardial necrosis,23 and baseline anemia24 might be possible causes of in-hospital hemoglobin drop without overt bleeding. Therefore, adverse cardiovascular risk profile related to in-hospital hemoglobin decline, and the exacerbation of myocardial ischemia induced by anemia, may account for the poor prognosis in these patients. Honestly, prior studies14,15,21 have indicated a potential correlation between hemoglobin decline and subsequent clinical outcomes in ACS patients who do not present with overt bleeding. However, the current study has made significant improvements in two aspects. Firstly, it focuses on the specific STEMI patient with appropriate sample size, which could support to eliminate potential confounding from different subtypes of ACS and prevent insufficient statistical power. Second, unlike most prior studies, it is of utmost importance to provide evidence from the East Asian population.

Limitation

There are several limitations that should be considered in this study. Firstly, patients without baseline (on admission) or nadir (lowest hemoglobin value measured during hospitalization) inpatient hemoglobin values were excluded, which may introduce selection bias and limit the generalizability of the findings. Secondly, the possible reasons for the hemoglobin drop in AMI patients were not analyzed, as hematological disorders, malignant tumors or other anemia-related diseases conditions could also cause a drop in hemoglobin levels, and these factors may potentially influence the prognosis of the patients. Consequently, the disregard of these potential confounding factors may limit the validity of the results. Further studies with larger sample sizes, prospective designs, and comprehensive analyses of potential confounding factors are needed for a better understanding of our findings.

Conclusion

In AMI patients who have undergone PCI, an in-hospital decline in hemoglobin levels is associated with an increased risk of 1-year ischemic events and all-cause death, even in the absence of overt bleeding. This finding provides important evidence for the clinical management and prognostic assessment of AMI patients, suggesting that clinicians should closely monitor the changes in patients’ hemoglobin levels during the treatment process to enable timely intervention measures and improve prognoses.

Methods

We retrospectively analyzed patients who underwent PCI at the General Hospital of Northern Theater Command from March 2016 to March 2019. A standard web-based data collection system (CVNET, Crealife Technology) was used to collect patients’ demographic and clinical characteristics, including age, sex, medical history, PCI indication, laboratory findings, angiographic and procedural characteristics, and medication treatment. These patients had both baseline (on admission) and nadir (lowest hemoglobin value measured during hospitalization) inpatient hemoglobin values. Of 5,036 included patients reported an in-hospital hemoglobin drop without overt bleeding. These patients were divided into 3 groups based on tertiles of the degree of hemoglobin drop: hemoglobin drop < 0.8 g/dL (n = 1652), 0.8–1.49 g/dL hemoglobin drop (n = 1651) and ≥ 1.5 g/dL hemoglobin drop (n = 1733). Follow-up information was obtained by telephone at 3 months, 6 months, and 1 year after the discharged by staff members who were unaware of the clinical process. This study was approved by the institutional ethical committee of the General Hospital of Northern Theater Command [No. Y(2023)198] with a waiver of the requirement to obtain informed consent to conduct this analysis. And the study complied with the provisions of the Declaration of Helsinki.

Definition

The definition of hemoglobin drop was baseline hemoglobin (at admission) minus nadir (lowest hemoglobin value measured during hospitalization). Besides, a standardized definitions of endpoints was used in the present study.25

The primary outcome was ischemic events at 1 year, defined as a composite of cardiac death, myocardial infarction (MI), or stroke. Secondary endpoints were defined as 1-year all-cause death, Bleeding Academic Research Consortium (BARC) types 2, 3, or 5, and 3 or 5 bleeding.

Statistical analysis

Continuous variables are expressed as mean ± standard deviation (SD) and compared using analysis of variance. Categorical variables are presented as number (percentage) and were compared using the χ2 test or Fisher’s exact test. Time-to-event outcomes were analyzed by the Kaplan–Meier method, and compared by the log-rank test. The Cox proportional hazard regressions were performed to calculate the hazard ratio (HR) and 95% confidence interval (CI) and evaluate the associations between hemoglobin drop and clinical outcomes. Variables selected for adjusted in multivariable model were patient characteristics including age, gender, hypertension, diabetes, a history of MI, previous stroke, previous PCI, smoking status, types of AMI, estimated glomerular filtration rate, procedure information (transradial access, coronary arteries treated), and medical treatment (aspirin, P2Y12 inhibitors, statin, ACEI/ARB, β-blockers). Receiver operating characteristic curve (ROC) for hemoglobin drop was constructed to assess the predictive accuracy for 1-year ischemic events with the area under the curve (AUC). Unless otherwise specified, a 2-sided P value less than 0.05 was considered to indicate statistical significance. The statistical analysis was conducted using SAS software version 9.4 (SAS Institute, Cary, NC, USA).