Anemia is a common disease in pediatrics, presenting in 15–74% of this population (1–6). A reduction in oxygen-carrying capacity can lead to inadequate tissue oxygenation and organ function (7). As a result, anemia may impair the body's resistance to surgical stress. Recent studies indicate that preoperative anemia is associated with higher in-hospital mortality in children undergoing noncardiac surgery (2, 3), but little is known about this association in children undergoing cardiac surgery. Limited evidence has shown that preoperative hemoglobin (Hb) <11.0 g/dl is associated with a higher risk of postoperative acute kidney injury in children with congenital heart disease (CHD) (8).

CHD can be divided into two categories according to the presence or absence of cyanosis (9–11). Cyanosis and hypoxemia may occur if a significant right-to-left shunt exists in a heart defect, and generally, these are defects with significant complexity (10). Cyanosis can increase early and late postoperative mortality in children undergoing cardiac surgery (9) and enhance the long-term risks of noncardiac surgery and type 2 diabetes mellitus in adulthood (10, 11). Although cyanotic CHD is more complex and associated with poor clinical outcomes, cyanotic patients were excluded in some anemia-related studies (6, 7). The main obstacle is that the optimal Hb level is difficult in cyanotic patients when hypoxemia leads to secondary erythrocytosis and an increase in the Hb concentration (12, 13).

The main function of hemoglobin is to transport oxygen. Under normal nutritional and hematopoietic conditions, there is an inverse relationship between the severity of compensatory erythrocytosis and resting oxygen saturation in cyanotic patients (14). Therefore, this study integrated preoperative Hb and oxygen saturation to predict postoperative outcomes in children with cyanotic CHD.

Optimal preoperative Hb level and anemia criteria are difficult to define in cyanotic children (12, 13). The 15 g/dl cutoff value was adopted in the study published by Okoromah et al. (5). This value was significantly higher than the lower limits of normal Hb ranges of all age groups of healthy children (1, 19). However, a fixed Hb cutoff value may not be appropriate for cyanotic children because the severity of compensatory erythrocytosis is not fixed but inversely correlated with resting SpO₂ (14). In cyanotic adults with compensatory erythrocytosis, a strong linear correlation was found between Hb and SpO₂ (Hb = 61 – SpO₂/2) (20). In addition to oxygen saturation, age may also influence the Hb level of children (1, 19). In children with cyanotic heart disease, the regression equation was found as follows: Hb concentration = 34.4 – 0.22 × (aortic oxygen saturation) + 0.14 × age (21). However, this regression equation was restricted to children with sufficient iron stores and oxygen saturation >75% (21). Based on the theory of an inverse relationship between the severity of compensatory erythrocytosis and resting oxygen saturation (14) and the estimation formula of CaO₂ (18), we presumed that the value of preoperative Hb × SpO₂ can evaluate whether cyanotic children achieved adequate Hb compensation. If the value of preoperative Hb × SpO₂ of cyanotic children was below the lower limit of normal Hb of each age group, this indicated that the preoperative Hb concentration was not sufficient to obtain similar CaO₂ to healthy children, and preoperative anemia would be considered.

Preoperative anemia is common in neonates and children undergoing noncardiac operations, with an estimated incidence of 24–32% (2, 3). Similarly, the incidence was 23% in acyanotic children with a ventricular septal defect or an atrioventricular canal (4). However, the rate of preoperative anemia was only 4.5% in our cyanotic children if the diagnosis of anemia was made based on the presence of actual Hb concentration < aaHb. Hypoxemia may induce secondary erythrocytosis and then increase the hemoglobin concentration of cyanotic patients (12, 13). Obviously, the incidence of preoperative anemia in cyanotic children would be seriously underestimated by the actual Hb concentration, but the anemia rate could increase to 21.6% according to the Hb × SpO₂ value, which may improve the detection rate of preoperative anemia in cyanotic children. Moreover, univariate analyses indicated that the presence of preoperative Hb × SpO₂ < aaHb was associated with more types of severe postoperative events, including in-hospital mortality (Table 1). Both multivariable logistic regression analysis and propensity-score matching analysis showed that preoperative Hb × SpO₂ < aaHb was significantly associated with the composite postoperative outcome (Table 3). This was consistent with the findings that preoperative anemia was associated with poor postoperative outcomes in adults undergoing cardiac surgery (7, 22, 23) and in neonates and children undergoing noncardiac surgery (2, 3). Therefore, compared with the actual Hb concentration, the value of Hb × SpO₂ below the lower limit of normal Hb of each age group has stronger prognostic power and may be more suitable to evaluate preoperative anemia in children with cyanotic CHD.

The reasons for poor prognosis among anemic patients undergoing surgery are not well known. One explanation is that anemia may be the representation of other confounding factors related to poor prognosis (7, 13). Erythrocytosis does not tend to stabilize until hypoxemia has been present for quite some time, so preoperative Hb × SpO₂ < aaHb was more likely to occur in children without adequate time for the response to hypoxemia, demonstrated by our results that the rates of preoperative Hb × SpO₂ < aaHb in children younger than 6 months (33.7%) and aged 6 months to 2 years (27.1%) were higher than those in other age groups (14.2–17.2%). Meanwhile, younger age was an independent risk factor for the postoperative composite outcome (Table 3), which was consistent with the finding in acute kidney injury after congenital cardiac surgery (8). We also found that the incidences of Hb × SpO₂ < aaHb were higher in children with patent ductus arteriosus (26.2%) or transposition of great arteries (46.6%) and were lower in children with aortopulmonary collateral arteries (17.5%) or clubbed fingers or toes (13.9%). Compared to the entire population with a median age of 24.7 months, the median age was younger in children with PDA (13.5 months) or transposition of great arteries (10.3 months) and was older in children with aortopulmonary collateral arteries (31.8 months) or clubbed fingers or toes (48.9 months). Children with transposition of great arteries or duct-dependent CHD usually require early diagnosis and intervention (24, 25) and may not have enough time to compensate for hemoglobin. In contrast, the presence of aortopulmonary collateral arteries and clubbed fingers and toes generally indicates that hypoxemia has existed for a long time. Therefore, anatomic and pathophysiologic factors may partially explain the prognostic power of Hb × SpO₂ < aaHb. A previous study revealed that aortic oxygen saturation below 75–80% was associated with an increase in erythropoietin titer, suggesting that adequate and stable erythrocytic response is not easy for patients with deep hypoxemia (21, 26). Similarly, our findings showed that low SpO₂ (<74.1%) was associated with insufficient compensation for preoperative Hb concentration (Hb × SpO₂ < aaHb) (Table 2). In addition to cyanotic CHD, living at high altitudes (low oxygen environment) can also lead to elevated hemoglobin levels in patients. In this study, eighty (10.3%) patients lived at an altitude of >1,500 m. However, the altitude of long-term residence was only slightly related to the presence of Hb × SpO₂ < aaHb (P = 0.055), probably because children moving from high altitude to the urban area of Chengdu with a 500-m altitude had adapted to the normal oxygen environment before surgery. Therefore, we conducted a propensity score matching analysis to balance these confounding factors (Table 2). Multivariable logistic regression analysis showed that preoperative Hb × SpO₂ < aaHb was still significantly associated with the composite outcome in the matched cohorts (Table 3). It requires going back to the physiological function of hemoglobin. Preoperative Hb × SpO₂ < aaHb indicates a decrease in blood oxygen content in cyanotic children. This may lead to inadequate tissue oxygen delivery for metabolic needs, consequently increasing the risk of postoperative organ dysfunction (7).

From the perspective of tissue oxygen delivery, preoperative Hb × SpO₂ < aaHb may be a modifiable risk factor for cardiac surgery. It is estimated that the incidence of iron deficiency is up to 47.1% in children with cyanotic CHD (27). Despite the lack of high-level evidence, the consensus for patient blood management recommends early management anemia before cardiac surgery, including preoperative iron supplementation for iron-deficiency anemia and consideration of erythropoietin in patients with specific conditions (12, 28). According to the formula for calculating oxygen content (18), improvement of oxygen saturation may also be a strategy for selected children with cyanotic CHD, such as oxygen administration in concomitant parenchymal lung disease or deep cyanosis (29) and prostanoid pulmonary vasodilator use in severe pulmonary arterial hypertension (29, 30) and duct-dependent CHD (25, 31). However, further studies are required to determine whether these strategies can improve postoperative outcomes in cyanotic children with preoperative Hb × SpO₂ < aaHb.

The authors acknowledge several limitations. First, there may have been selection bias in this single-institution retrospective study. Second, pulse oximetry performs poorly when SaO₂ is <80% (32). The mean SpO₂ was 4.6% higher than that of SaO₂ in cyanotic children (32). In this study, we found that 100% was the optimal cutoff value of normal SpO₂ (95–100%). Therefore, it was increased by approximately 5% on both sides of the equation (preoperative Hb × preoperative SpO₂ = normal Hb × normal SpO₂). To some extent, this counteracted the overestimation of oxygen saturation by SpO₂ in cyanotic children. Meanwhile, multiple measurements of SpO₂ are noninvasive, convenient and cost-effective. Moreover, an association, rather than causation, was identified between preoperative Hb × SpO₂ < aaHb and postoperative outcomes. We could not eliminate the influence of other confounding factors such as socioeconomic status (33, 34) and chromosome abnormality (35), which were reported as prognostic factors for children with CHD but were not recorded for all children in our study.

Our findings suggest that a preoperative Hb × SpO₂ value below the lower limit of normal Hb is significantly associated with higher postoperative mortality and morbidity and is a potential criterion to evaluate preoperative anemia in children with cyanotic CHD. Prospective multicenter studies are required to confirm these findings.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The studies involving human participants were reviewed and approved by the Ethics Committee on Biomedical Research, West China Hospital of Sichuan University (Reference Number, 2019-438). Written informed consent from the participants' legal guardian/next of kin was not required to participate in this study in accordance with the national legislation and the institutional requirements.

DZ, L-JD and Y-FL: contributed to the data curation, methodology, formal analysis, original draft, and final revision. M-LT: contributed to the conceptualization, methodology, interpretation, project administration, and final revision. All authors have read and approved the final manuscript.

This work was supported by the West China Nursing Discipline Development Special Fund Project of Sichuan University (Grant Number, HXHL19061).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.