Transferrin saturation for the diagnosis of iron deficiency in febrile anaemic children

C H Li 李澤荷, A C W Lee 李志偉, T W L Mak 麥永禮, S C Szeto 司徒紹昌

HK Pract 2003;25:363-366

Summary

Objective: To study the sensitivity of commonly used laboratory markers in the diagnosis of iron deficiency anaemia (IDA) in children found incidentally to be anaemic during hospitalisation.

Design: A retrospective study of all children evaluated for iron deficiency during their admission to a general hospital between July 1999 and June 2001.

Subjects: Children aged between 6 months and 15 years who were found to be anaemic and subsequently confirmed to have IDA by a positive therapeutic response were included.

Main outcome measures: The sensitivities of the following laboratory results were calculated and compared: red cell distribution width (RDW), Mentzer's index (mean corpuscular volume/red blood count), serum iron, transferrin saturation, total iron binding capacity (TIBC), and ferritin.

Results: 45 children were diagnosed with IDA. 14 (31%) were also found to have either or thalassaemia trait. The sensitivities of RDW, Mentzer's index, serum iron, TIBC, transferrin saturation, and serum ferritin in identifying IDA were 71%, 60%, 83%, 23%, 95% and 64%, respectively. Eliminating serum ferritin from the biochemical tests reduces the cost from HK$35 to $5 per patient.

Conclusion: Transferrin saturation was the most sensitive and cost-effective marker of IDA found incidentally in children evaluated for febrile illnesses.

Keywords: Anaemia, Children, Ferritin, Iron deficiency anaemia, Transferrin saturation

摘要

目的： 通過住院後意外發現患有貧血的兒童進行缺鐵性貧血（IDA）診斷，研究相應常用實驗室指標的敏感性。

設計： 對自1999年7月至2001年6月，曾在某一綜合醫院住院期間，進行缺鐵性貧血檢查的所有兒童進行的回顧性研究。

研究對象： 發現有貧血並對治療反應良好被證實患有IDA的6個月至15歲的兒童。

主要測量內容： 對下列實驗室檢查結果的敏感度進行計算比較：紅細胞分佈寬度（RDW）、Mentzer指數（紅細胞平均體積/紅細胞計數）、血清鐵、轉鐵蛋白飽和度、總鐵結合力（TIBC）和鐵蛋白。

結果： 45名兒童被診斷為IDA。其中14人（31%）還有或地中海貧血的特徵。RDW、Mentzer指數、血清鐵、TIBC、轉鐵蛋白飽和度和血清鐵蛋白 診斷IDA的敏感度分別為71%、60%、83%、23%、95%和64%。如果從這些生化檢查中去掉血清鐵蛋白檢查，則成本即可由每人35港元降至5港元。

結論： 轉鐵蛋白飽和度是在發熱兒童診斷IDA最敏感、最有成本效益性的指標。

詞彙： 貧血，兒童，鐵蛋白，缺鐵性貧血，轉鐵蛋白飽和度。

Introduction

Iron deficiency is the commonest nutritional deficiency in the world and constitutes a major health problem in children, even in developed regions.¹ The clinical manifestations of iron deficiency anaemia (IDA) can be subtle, and significant psychomotor retardation may occur in the young child if iron deficiency is severe and prolonged.² Definitive diagnosis of iron deficiency relies on the demonstration of depletion of iron reserves in the body; for example, the absence of stainable iron in the bone marrow.³ However, bone marrow aspiration is an invasive procedure and is not recommended for routine diagnosis.⁴ Instead, a number of biochemical and haematological parameters including ferritin, transferrin saturation, total iron binding capacity (TIBC), and red cell distribution width (RDW), have been used as surrogate markers for iron deficiency and hence initiation of iron supplementation.^4,5 Under such circumstances, the ultimate diagnosis of IDA rests on the correction of anaemia after adequate treatment.²

These biochemical and haematological surrogate markers of iron deficiency are often used in combination clinically, but their relative merits for the recognition of IDA are unclear. In clinical practice, where anaemia is often detected "incidentally" during the management of acute infectious disorders, the reliability of biochemical markers has been queried.⁶ This retrospective study was carried out in order to assess the sensitivity of a number of commonly used laboratory parameters as markers of IDA.

Methods

The study was carried out in the paediatric unit of an acute general hospital, where children under the age of 16 years were admitted. The management of anaemia (haemoglobin <11g/dL) found during the course of hospitalisation has been previously described.⁷ In brief, the affected children were offered the following tests: full blood counts, serum iron, TIBC, transferrin saturation (serum iron/TIBC x 100%), serum ferritin, and haemoglobin electrophoresis. The patients were also offered a 4-week course of iron therapy, after which they would return for follow-up of their initial laboratory findings and a repeat measurement of haemoglobin.

This is a retrospective hospital chart review of all children who were diagnosed with IDA between July 1999 and June 2001. The WHO criterion of a therapeutic response - defined as a rise of haemoglobin by at least 1g/dL in a 4-week interval after iron therapy (4-6mg of elemental iron/kg/day) - was used as the standard for diagnosis.² The inclusion criterion was age between 6 months and 15 years. Children were excluded if they had received blood transfusion, had prior history of iron treatment, or had known chronic inflammatory illnesses. The results of the following laboratory tests were obtained: mean corpuscular volume (MCV), red blood count (RBC), RDW, serum iron, transferrin saturation, TIBC, and ferritin. According to the hospital reference values, a positive test for iron deficiency was defined as RDW >15%, serum iron <8 mol/L, TIBC >79 mol/L, transferrin saturation <20%, or serum ferritin <67pmol/L. We also examined the usefulness of the Mentzer's index (MCV/RBC), for which a ratio of greater than 13.5 was taken as a positive test.⁸ The sensitivity of each of these haematological and biochemical markers was calculated and compared for their clinical usefulness.

As there is no simple test to exclude IDA, and patients who failed iron therapy might have poor compliance or persistent blood loss, the true negative rate for IDA could not be ascertained. Hence, the specificity of the tests was not measured.

Results

Forty-five children, including 21 girls and 24 boys, were diagnosed with IDA by therapeutic response and satisfied both the inclusion and exclusion criteria. Their mean age was 6.4 years and the mean haemoglobin at presentation was 9.3g/dL. and thalassaemia traits were also found in eight (18%) and 6 (13%) of the children, respectively. Dietary insufficiency was the most frequent predisposing factor, occurring in 24 (53%) patients. Fourteen (31%) children were found to have blood loss from the gastrointestinal tract. Menstrual loss accounted for the remaining 7 (16%) cases. The sensitivity of the haematological and biochemical markers for IDA is tabulated in Table 1. Transferrin saturation, derived from the ratio of serum iron/TIBC x 100%, had the highest sensitivity at 94.8%. Serum ferritin, a classical marker of body iron stores, had a sensitivity of only 63.8%.

Discussion

Iron deficiency anaemia (IDA) remains a global childhood problem.¹ Its prevalence in Hong Kong has not been studied in full. An estimated 0.3-0.6% of middle school children may be affected.^9,10 IDA is characterised by the presence of microcytic, hypochromic anaemia, a feature that is also shared by a number of the thalassaemic syndromes. IDA and thalassaemia may co-exist in the same patient. In areas where thalassaemic syndromes are common, the diagnosis of IDA using simple haematological parameters is difficult.⁶ The absence of stainable iron in a bone marrow aspirate provides a definitive diagnosis of IDA, but the procedure is too invasive for routine clinical use. Hence, a number of biochemical tests and haematological parameters have been recommended for the indirect diagnosis of IDA. The diagnosis can be confirmed by the correction of anaemia after adequate iron treatment.²

The interpretation of these biochemical markers may be difficult in clinical practice where anaemia is often evaluated amidst acute febrile illnesses. Serum iron represents <0.1% of the total body iron. Besides IDA, a low serum iron level may also be seen during acute or chronic infections, whereas haemolysis or liver disease may increase its measurement.¹¹ Serum ferritin has been traditionally taken as a reliable marker of body iron stores, and hypoferritinaemia is commonly used as a diagnostic marker for iron deficiency.³ Because it is an acute phase reactant, however, a normal or raised serum ferritin level may be seen during acute infections or liver disease even in the presence of iron deficiency.^12,13 An increase in TIBC is also indicative of iron deficiency, but it suffers the same drawbacks as serum iron measurements.¹¹ It is therefore not surprising that both TIBC and serum ferritin were found to be relatively insensitive (23% and 63.8%, respectively) in detecting IDA in the present study.

Transferrin saturation, an indicator of iron transport that derives from the ratio of serum iron and TIBC, appeared to be a sensitive marker of iron deficiency in our study. As the conditions affecting serum iron and TIBC levels tend to increase or decrease both measurements in the same direction,¹¹ transferrin saturation may be more applicable in our clinical practice.

Others have tried to circumvent the problem by looking at red cell indices generated by automated cell counters and derived functions from these indices for the diagnosis of iron deficiency. An RDW >15%¹⁴ or a Mentzer's index (MCV/RBC) >13.5⁷ have been found to be useful indicators, especially in the differentiation of IDA from thalassaemic disorders. However, both parameters were not found to be sensitive indicators (sensitivities 71% and 60%, respectively) in our study. The high rate of co-existing IDA and thalassaemic traits in our study might have undermined the usefulness of these indices.

It would be ideal if the specificity of these tests could be evaluated; however, in order to do this, we would need to know the true prevalence of anaemic children without iron deficiency. Unfortunately, we do not have a non-invasive and routinely available test to exclude iron deficiency. Failure to respond to iron therapy cannot be taken as exclusion for IDA because poor drug compliance or ongoing blood loss might be the underlying cause. From a clinical point of view, however, the specificity of the test is not important as long as the test is sensitive enough to select the appropriate patients for iron treatment.

Our findings also allow us to assess the feasibility of cost saving in the evaluation of IDA in children. The combined tests of serum iron, TIBC, and serum ferritin cost HK$35 of which serum ferritin alone costs $30. Eliminating the routine use of serum ferritin measurement means that the cost can be reduced by 86% per patient tested.

Conclusion

The study confirms that the use and interpretation of biochemical markers and red cell indices in the evaluation of suspected IDA in children presenting with febrile illnesses should be viewed with caution. For the investigation of the commonly encountered microcytic, hypochromic anaemias, the use of transferrin saturation alone in addition to haemoglobin electrophoresis (c.f. thalassaemic disorders) may represent the most sensitive and cost-effective combination of tests.

Key messages

1. The vocational trainees on the whole are satisfied with the content, but not necessarily the process, of their training. They acquire some knowledge, skill and experience in Family Medicine.
2. The Basic Hospital Trainees are frustrated with uncertainty about their future career, unclear objectives of their training, workload more on service than training, and low self (possibly also peer) esteem.
3. The quality of training for Basic Community Trainees is probably variable, largely depending on the individual trainers.
4. The Higher Vocational Trainees are independent learners but would like to have structured programme and more guidance.
5. The trainees of different categories would like to have more communication among different parties involved in their vocational training.