Severe iron deficiency anemia in the paediatric emergency department: A retrospective study

Abstract

Background

Transfusion is discouraged in hemodynamically stable children with severe iron deficiency anemia (IDA). Intravenous (IV) iron sucrose (IS) could be an alternative for some patients; however, there is a paucity of data on its use in the paediatric emergency department (ED).

Methods

We analyzed patients presenting with severe IDA at the Children’s Hospital of Eastern Ontario (CHEO) ED between September 1, 2017, and June 1, 2021. We defined severe IDA as microcytic anemia <70 g/L and either a ferritin <12 ng/mL or a documented clinical diagnosis.

Results

Of 57 patients, 34 (59%) presented with nutritional IDA and 16 (28%) presented with IDA secondary to menstrual bleeding. Fifty-five (95%) patients received oral iron. Thirteen (23%) patients additionally received IS and after 2 weeks, the average Hgb was similar to transfused patients. The median time for patients receiving IS without PRBC transfusion to increase their Hgb by at least 20 g/L was 7 days (95%CI 0.7 to 10.5 days). Of 16 (28%) children who were transfused with PRBC, there were three mild reactions, and one patient who developed transfusion associated circulatory overload (TACO). There were two mild and no severe reactions to IV iron. There were no return visits to the ED due to anemia in the following 30 days.

Conclusions

Management of severe IDA with IS was associated with a rapid rise in Hgb without severe reactions or returns to ED. This study highlights a strategy for management of severe IDA in hemodynamically stable children that spares them the risks associated with PRBC transfusion. Paediatric specific guidelines and prospective studies are needed to guide the use of IV iron in this population.

Iron deficiency anemia (IDA) is the most common cause of anemia worldwide (1). Iron deficiency anemia occurs much more frequently in some northern Indigenous and Inuit communities but is reported less often in urban Canadian cities (2–4). Severe IDA represents an uncommon extreme presentation of this disorder that is likely under-reported with only 195 cases reported in Canada between 2009 and 2011 (5). Choosing Wisely guidelines discourage the use of packed red blood cell (PRBC) transfusion in asymptomatic, hemodynamically stable children with IDA (6), but acute management of children with severe IDA often involves PRBC transfusion which is reported in 14% to 56% of cases (7–11). Tachycardia and delayed capillary refill are predictors of hemodynamic instability; however, both findings are confounded in children with anemia making it unclear which patients with severe IDA are at risk of hemodynamic instability (12–15). The overuse of transfusion in this could be explained by a perceived urgency of correcting a low Hgb, that obscures careful consideration of the risks of serious adverse events including transfusion associated cardiac overload (TACO) and alloimmunization.

Protocols directing the use of intravenous (IV) iron developed in the adult emergency department (ED) setting have been successful in promoting rational use of PRBC products in adult patients with severe IDA (16–18). A legacy of adverse reactions from high molecular weight iron dextran products have delayed the widespread use of IV iron products, despite thorough systematic reviews revealing no increased risk of severe adverse events with IV iron products compared with intramuscular iron, oral iron, and placebo controls (19,20). Newer, non-high weight dextran IV iron preparations such as iron sucrose (IS) have an estimated incidence of severe adverse drug reactions of < 1:200,000 doses (21). There are a growing number of single institution studies reporting the safety of IV iron in children (22–35). These studies include mainly patients with mild to moderate anemia, and there are no studies examining the use of IV iron for management of patients with severe anemia or those presenting to the paediatric ED. We hypothesized that patients who receive IV iron in the ED respond with a rapid increase in Hgb that exceeds the rate expected with oral (PO) iron, and that patients given IV iron can be managed safely without PRBC transfusion. The purpose of this study was to describe the presentation, management, and response to therapy of paediatric patients with severe IDA presenting to a paediatric ED.

METHODS

This retrospective, observational study was conducted at a paediatric tertiary care hospital in an urban setting. The primary outcome was frequency of IV iron therapy and PRBC transfusion. Secondary outcomes were Hgb response following IV iron therapy and the rate of adverse infusion and transfusion reactions. Medical records were reviewed for all patients presenting with severe IDA at the Children’s Hospital of Eastern Ontario (CHEO) ED between September 1, 2017 and June 1, 2021. The study period was chosen to optimize data collection as it coincides with the introduction of a new electronic medical record at our institution. Severe anemia was defined as Hgb <70 g/L because it is a common transfusion threshold extrapolated from adult transfusion guidelines (36). Iron deficiency was identified by microcytosis and low ferritin according to the age specified reference ranges at our institution (Supplementary Table 1), iron studies (serum iron, total iron binding capacity, transferrin saturation) when available, or a clinical diagnosis of iron deficiency documented in the patient record (ED, inpatient or outpatient) when iron studies were not available. Patients were excluded if a diagnosis was present such as malignancy, hemolytic anemia, or hemoglobinopathy because we believed anemia would be managed differently in the context of these conditions. Complete medical records were reviewed for all patients with severe IDA. Data abstracted included demographic information, underlying diagnosis, treatment administered, laboratory values before and after therapy, and adverse reactions. This retrospective study was approved by the Research Ethics Board at CHEO.

Descriptive statistics with median and interquartile range (IQR) for continuous variables, and frequencies and percentages for categorical variables were used to analyze population demographics, and frequency of treatments prescribed (PO iron, IV IS, and PRBC transfusion). Variations of hemoglobin levels in the two iron treatment groups (IV iron with PO iron and PO iron only) over time were modelled using local regression, specifically LOESS smooth (37). Results were illustrated with LOESS curves using ggplot2 package from R statistical software (38,39). The distribution of time from ED encounter to 20 g/L increase of Hgb in each iron treatment group was modeled using the Kaplan–Meier estimator. The exact time of Hgb increase by 20 g/L was approximated as the midpoint between the date the increase was observed and the date of the most recent previous Hgb measurement. Results were illustrated with Kaplan–Meier curves using survminer package from R statistical software (40).

RESULTS

Patient characteristics

There were 57 patients presenting with severe IDA during the study period (Figure 1). Fifty-eight ED encounters were included in the analysis because one patient presented again, one year later (Table 1). Median age at presentation was 4.2 years (IQR 1.7 to 13.2) and the youngest patient was 11 months. The most common underlying diagnosis was nutritional IDA (n = 34, 59%) followed by heavy menstrual bleeding (HMB) (n = 16, 28%). Most patients with nutritional IDA were patients under 3 years of age (26, 76%). Median Hgb at presentation was 51 g/L (IQR 41 to 61.5). There were 26 (45%) patients that presented with a Hgb less than 50 g/L, and of these, 15 (58%) were toddlers less than 3 years old with nutritional IDA. Patients presenting with Hgb less than 40 g/L (11, 19%) were almost exclusively children (median 2 y, IQR 1.5 to 2.3) with nutritional IDA (n = 9, 82%). There were 21 (36%) patients admitted to an inpatient ward from the ED, 14 (67%) with nutritional IDA, and 3 (14%) with HMB. No patient was admitted to the PICU from the ED but one patient was transferred to the PICU on Day 3 of admission.

Figure 1.

Flowchart of cohort identification. Three hundred and fifty-one hospital encounters were identified based on a Hgb < 70 g/L. Fifty-eight encounters met the inclusion criteria of Hgb <70 and low MCV or IDA diagnosis at time of presentation to the ED. Reasons for exclusion are listed.

Table 1.

Characteristics of patients with severe IDA at ED presentation (N = 58)

Characteristics
Age at ED presentation, N (%), years
0–0.99	1 (2)
1–5.99	29 (51)
6–10.99	4 (7)
11–17.99	23 (40)
Female sex, N (%)	42 (74)
Most responsible diagnosis, N (%)
Nutritional	34 (59)
HMB	16 (28)
IBD	2 (3)
Other*	6 (10)
Laboratory Investigations at ED presentation
Hgb, median (IQR), g/L	52 (41–60)
Platelet count, median (IQR), ×109/L	363 (276–450)
MCV, median (IQR), fL	56 (50–66)
Ferritin, median (IQR), µg/L	2 (2–6)
Admitted, N (%)	21 (36)

ED Emergency department; Hgb Hemoglobin; HMB Heavy menstrual bleeding; IBD Inflammatory bowel disease; IQR Interquartile range.

*Other diagnoses: GI bleed (3), chronic kidney disease (1), unknown (2).

Treatment administered

Oral iron was prescribed in nearly all encounters (n = 55, 95%), most often as monotherapy for patients presenting with Hgb >40 g/L (n = 29, 66%) compared to <40 g/L (n = 3, 27%). Ferrous sulfate was the most common iron supplement (n = 52, 94%). Prescribed doses varied, the most common dosing was 6 mg/kg/day (n = 25, 45%) followed by 60 mg/dose (n = 16, 29%), lower dosing of 3 mg/kg/day were uncommon (n = 4, 7%). Dose frequency included daily (n = 17, 32%), twice daily (n = 21, 38%), and three times daily (n = 17, 31%).

Iron sucrose was administered during 13 (22%) encounters (Table 2, Supplementary Table 2). Ten (76%) of these patients were admitted. Iron sucrose dose was 7 mg/kg (maximum 300 mg) for 11 (85%) patients. Two transfused patients received IS as well, one received a dose in the ED and a subsequent dose on day six of admission, the other received IS on day 14 of admission. Eleven (20%) patients were managed without transfusion and most of these patients had nutritional IDA (n = 7, 64%) followed by HMB (n = 4, 36%). Nine (82%) patients treated with IS alone received a single dose, and two (18%) received additional doses.

Table 2.

Characteristics of patients receiving IV iron and transfusion

Characteristics	IV iron only (N = 11)	Transfusion* (N = 16)
Age, median (IQR), years	2.5 (2–13)	12 (2–14)
Female sex, N (%)	10 (90)	11 (69)
Most responsible diagnosis, (N = 58)
Nutritional, N (%)	7 (64)	7 (44)
HMB, N (%)	4 (36)	6 (38)
IBD, N (%)	0	2 (12)
UGIB, N (%)	0	1 (6)
Presenting Hgb, median (IQR), g/L	47 (30–52)	48 (36–54)
Admitted, N (%)	8 (73)	11 (69)
Minor reaction, N (%)	2 (18)	3 (19)†
Major reaction, N (%)	0	1 (6)
Return to ED for anemia, N (%)	0	0

ED Emergency department; Hgb Hemoglobin; HMB Heavy menstrual bleeding; IBD Inflammatory bowel disease; IQR Interquartile range; UGIB Upper gastrointestinal bleeding.

*Two patients also received IV iron.

†One patient received IV iron in addition to PRBC transfusion, Minor Reaction: Nausea (Grade 1).

PRBC transfusion was administered to 16 (28%) patients (Table 2, Supplementary Table 3), including 7 (44%) with nutritional IDA, 6 (38%) with HMB. Eleven (68%) patients were admitted. Nine (56%) patients received a single PRBC transfusion and 6 (38%) received two or more PRBC transfusions. Total transfused PRBC volume ranged from 3 to 25 mL/kg, and three (43%) of the infants with nutritional IDA were transfused in aliquots of 10 mL/kg/dose or more. Symptomatic anemia was the most common rationale for transfusion, but there was a wide range of symptom severity. The following representative progress note extractions from the ED or inpatient record demonstrate this variability:

“lethargic, tachycardic, and febrile, appearing unwell. He was therefore transfused 10 cc/kg of PRBC”

“Gave transfusion of 50mls of RBCs for symptomatic relief […] with the hope that this would improve her motivation and ability to start feeding”

“ongoing systolic murmur, not in overt heart failure at present but significant concern for potential to decompensate”

Response to therapy

Follow-up data were available following 44 (75%) ED encounters. The median time for patients receiving IS without PRBC transfusion to increase their Hgb by at least 20 g/L was 7 days (95%CI 0.7 to 10.5 days), compared to 44 days (95%CI 20.4 to 96.8) for individuals receiving PO iron (Figure 2). Four patients treated with IS presented with an Hgb <40 g/L and were not transfused. All four patients increased their Hgb to >100 g/L by 6 weeks after their ED encounter. The most extreme anemia that was managed with IS alone was a toddler with nutritional IDA presenting with Hgb 24 g/L. This patient received three daily doses of IS and increased their Hgb to 70 g/L after 12 days. There were 11 patients that presented with bleeding, eight (73%) with HMB. Eight (73%) recipients of IS only were admitted for a mean of 6 days. Nine (64%) transfused patients were admitted for a mean of 7 days. No patients receiving IV iron or transfusion who were discharged from the ED returned to the ED within 30 days for anemia. One transfused patient received an additional transfusion at outpatient follow-up 4 weeks later. Detailed information on patients managed with IV iron and PRBC transfusion is included in the supplemental tables (Supplementary Tables 2 and 3).

Figure 2.

Time to event analysis for 20 g/L increase of Hgb from baseline for patients treated with IV and PO iron and patients receiving PO iron only. The distributions of event times were shown for the first 90 days.

Adverse reactions

Among recipients of IS there were three minor infusion reactions (CTCAE 1; fever, and nausea) and no severe reactions. Three recipients of PRBC experienced minor transfusion reactions (CTCAE 1; fever, and nausea). One infant with nutritional IDA developed transfusion-associated circulatory overload (CTCAE 4; TACO) and was admitted to the intensive care unit on day three of admission after receiving 20 mL/kg PRBC within 24 hours (10 mL/kg aliquots, Hgb rise from 17 to 74 g/L), and 10 mL/kg of 25% albumin.

DISCUSSION

The use of IV iron in the management of paediatric IDA has been limited to children who fail oral iron therapy, or those in whom oral iron is contraindicated (41). There is no approved indication for IV iron in the urgent management of severe IDA and clinicians who are unfamiliar or uncomfortable with IV iron must choose between oral iron and blood transfusion for patients with severe IDA presenting to the ED. Almost one in three of the patients in our study received PRBC transfusion. Excerpts from some patient records highlighted poorly rationalized use of PRBC transfusion for such reasons as to improve appetite or motivation. Intravenous iron products are safer than blood with significantly lower risks of serious adverse events, avoiding transfusion reactions such as TACO and alloimmunization (42). Existing guidelines provide no or unclear criteria for the use of IV iron products, and broad indications for PRBC transfusion (41,43).

Our study reports a cohort of patients with severe anemia who responded to IS within 2 weeks, without early treatment failure or return visits to the ED. Patients presenting at our ED were predominantly toddlers with nutritional IDA, and young women with HMB, consistent with other studies of severe paediatric IDA presenting to the ED department (30,44). The most important finding from our study was that Hgb increased by at least 20 g/L within 2 weeks for nearly all patients treated with IS. Two other studies reported a similar mean Hgb increase at 2 weeks in children with IDA treated with intravenous iron (23,35). Unlike these studies, most patients in our study received only a single dose of IS in the ED, concurrent with initiation of oral iron therapy. This dose would be insufficient to fully replace the iron deficit, but likely provided enough immediately available iron for a rapid Hgb recovery sustained by ongoing replacement with oral iron. There were no significant infusion reactions in our population, which was expected because iron sucrose has been used extensively in paediatric patients without an increase in serious adverse events (30).

Third-generation IV iron products that can replace total iron deficit replacement in a single dose may eliminate the need for oral iron therapy in this patient population. Ferric carboxymaltose is a third-generation product with multiple supportive paediatric studies, however, it is not available in Canada (22–24,27–29). Low molecular weight iron dextran is another preparation used in paediatric patients, but this product has been discontinued in Canada (32,45–47). Ferric derisomaltose is the only analogous product available in Canada, but there are no paediatric studies of the safety or efficacy of this product.

Most patients with severe chronic anemia due to IDA should be managed with oral iron alone. Another Canadian academic paediatric hospital reported only 71% of children presenting to the ED with severe (Hgb <80 g/L) microcytic anemia were started on oral iron (48). Nearly all patients in our study were prescribed oral iron, but there was heterogeneity in the dose and frequency of oral iron. A third of patients received TID dosing, which is likely a poorly tolerated regimen and recent data show that low dose, daily ferrous sulfate is effective and well tolerated by toddlers with moderate anemia (49). Few patients with Hgb less than 40 g/L were treated with oral iron alone, possibly because Hgb response following oral iron therapy is not expected within the first week of therapy (49,50).

Considering the excessive burden of IDA on some northern Indigenous communities in Canada, it is important that guidelines for management of severe IDA be created with the community hospitals that serve this population in mind. Treatment algorithms guiding the use of IV iron for IDA management have been successfully used to improve quality of care in the adult ED setting, but similar tools in paediatrics are lacking (16,51). Iron sucrose is the most likely product to be widely used because it is available in most academic and community hospitals in Canada and has the most safety and efficacy data supporting its use in children (25,30,31,33–35). The strategy of single-dose IS used in our centre is a safer alternative to transfusion for hemodynamically stable patients. Severe adverse events are more common with PRBC transfusion than with any contemporary IV iron product and therefore any health care setting that administers PRBC transfusions to children should already have the resources to offer iron infusion safely (42).

Limitations

This study is small and retrospective but clearly shows the hemoglobin response following IS administration for a typical cohort of patients with severe IDA presenting to the ED. A simplified approach of midpoint approximation when estimating time to Hgb increase may result in biased estimates of Hgb response when intervals are wide and vary between groups (52). We did not stratify our analysis by bleeding at presentation which raises the possibility that the Hgb response may be underestimated in younger patients without bleeding. Due to our small sample, we were unable to determine dose dependence of Hgb response following IS. We did not collect data on symptom resolution for patients treated with IS compared to those treated with oral iron alone. A prospective study is warranted to better characterize the effectiveness of single dose IS for patients with severe IDA. A national treatment algorithm guiding appropriate use of PRBC transfusion and IV iron is urgently needed to improve quality of care and reduce the excessive use of PRBC transfusions in this population.

CONCLUSION

Children treated with IS responded with rapid rise in Hgb within 2 weeks. Iron sucrose was safely administered with no moderate or severe infusion reactions. A significant proportion of children with severe IDA presenting to our ED received PRBC transfusion, and one severe adverse event was reported.

AUTHOR CONTRIBUTIONS

MS conceptualized and designed the study, collected data, drafted the initial manuscript, and reviewed and revised the manuscript. LR designed the data collection instruments, collected data, carried out the initial analyses, and reviewed and revised the manuscript. NM and VB conceptualized and designed the study, conducted the statistical analysis, and critically reviewed the manuscript. ML and EL conceptualized and designed the study, coordinated and supervised data collection, and critically reviewed the manuscript. All authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

FUNDING

No funding was secured for this study.

POTENTIAL CONFLICTS OF INTEREST

EL is a member of the Canadian Society of Transfusion Medicine Standards Committee and reports reimbursement for travel to for Canadian Society of Transfusion Medicine annual meeting. There are no other disclosures. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.