Print version ISSN 0256-9574
SAMJ, S. Afr. med. j. vol.100 no.5 Cape Town May 2010
Pathology request guidelines: haematology questioned
To the Editor: We refer to the Haematology section of the article by Stanfliet, King and Pillay.1
The assessment of iron deficiency anemia (IDA) is incomplete without demonstrating iron-deficient erythropoiesis, by looking at the red cell indices (mean corpuscular volume, mean cell haemoglobin and mean cell haemoglobin concentration (MCV, MCH and MCHC)).2,3 A serum ferritin is mandatory if the indices suggest microcytic hypochromic anaemia (MHA). A low ferritin confirms IDA, but a normal ferritin does not exclude iron deficiency. A normal or raised ferritin may also indicate another cause of MHA, such as anaemia of chronic disorder (of which HIV and TB are common causes locally).2-4 Therefore, such an over-simplified approach will misdiagnose or over-diagnose IDA.
A trial of iron without confirmation of IDA has been proposed, but only for patients who are regarded as low-risk. Included in this group would be premenopausal women and young individuals during growth spurts, especially in areas known to have a high prevalence of IDA. Even in this group, a trial of therapy with 200 mg of ferrous iron is more correctly recommended for 3 weeks prior to review for response, as opposed to the 2 months, as suggested by the authors.3 A positive response is present when the haemoglobin concentration (Hb) has increased by 2 g/dl after 3 weeks of therapy. This confirms the diagnosis of IDA, the cause of which still needs to be ascertained. Some studies have shown a statistically significant prevalence of dual pathology in this group of patients.4 In a prospective study of 187 premenopausal patients with IDA, menorrhagia was present in 67.4%. A possible gastro-intestinal (GI) cause of IDA was found in 57.2% and 43% for the upper and lower GI tract respectively (p<0.04 and 0.002). This would easily be missed if therapy is given without questioning the need to investigate for a possible cause. The main risk factor for the presence of likely GI causes in this study was the presence of upper GI symptoms (odds ratio 5.2, 95% confidence interval 1.6 -1 6.4, p=0.005).4 Moreover, in high-risk patients (such as the elderly), a trial of iron without establishing a diagnosis of the underlying IDA will result in unnecessary delays and is potentially hazardous to the patient.
Regarding haemoglobin determinations, there has been significant disparity between using the Spencer Haemoglobinometer readings and the standard laboratory readings reported in the study. Observers' Hb estimations, using the Hbmeter, ranged around 7.2 g/dl and 7.8 g/dl for a laboratory Hb of 5.9 g/dl, questioning the accuracy of an instrument on which therapeutic decisions may be made.5
A D Mahomed
Division of Medical Gastroenterology
Department of Internal Medicine
University of the Witwatersrand
Division of Haematology
Department of Internal Medicine
University of the Witwatersrand
1. Stanfliet JC, King JA, Pillay TS. Inadequacy of primary health care test request guidelines-Lack of an evidence base. S Afr Med J 2009; 99: 569-571.
2. Cook JD. Diagnosis and management of iron-deficiency anaemia. Best Prac Res Clin Haematol 2005; 18(2): 319-332.
3. Frewin R, Henson A, Provan D. ABC of clinical haematology: iron deficiency anaemia. BMJ 1997; 314: 360-363.
4. Smellie WSA, Forth J, Coleman JJ, et al. Best practice of primary care pathology: Review 3. J Clin Pathol 2006; 59: 781-789.
5. Vannella L, Aloe Spiriti MA, Cozza G. Benefit of concomitant gastrointestinal and gynaecological evaluation in pre-menopausal women with iron deficiency anaemia. Aliment Pharmacol Ther 2008; 28: 422-430.
6. Discombe G, Goddard AJ, Hill ID. The accuracy of haemoglobin determinations with the Spencer Haemoglobinometer. J Coll Gen Pract 1966; 12(2): 251-252.
Stanfliet, King and Pillay respond: Linegar et al. investigated the Buffalo haemoglobinometer.1 The guidelines recommend the use of the Spencer haemoglobinometer from American Optical. However, the design and principles of both instruments are similar, and identical considerations apply.
The performance of several portable haemoglobinometers is not without problems.2 The study cited by Moeng et al. quoting a CV of 3% obtained for the Spencer haemoglobinometer, was not conducted as would be the case today. No correlation co-efficient is stated, and it is not possible to comment on the appropriateness of the use of the haemoglobinometer, although they recommend it for general practice.3
There is no consensus on the investigation and management of iron deficiency anaemia (IDA).2 The gold standard for the diagnosis of iron deficiency is bone marrow aspirate stained for iron, and no single biomarker can serve as a proxy.2 We agree that a blind trial of iron supplementation is inappropriate purely on the basis of a portable haemoglobinometer, given the poor correlation of some devices in the primary health care setting, with some workers reporting a correlation factor of 0.61 against a formal haemoglobinometer, but that it serves as a screening test for anaemia in a primary health care setting.4 Nevertheless, Adamson et al. reported on children in the primary health care setting, where they reserved the measurement of serum ferritin for those with severe anaemia (Hb <9.5 g/dl). Those with milder anaemia purely diagnosed by a handheld haemoglobinometer, and no other indices, were treated with dietary advice and community follow-up.5
We suggest the use of further markers in addition to the hand-held haemoglobinometer to confirm the anaemia and to identify its likely cause.2,6 Ferritin and soluble transferrin receptors are probably the best biomarkers for the assessment of iron status in the absence of inflammation, but interpretation in the presence of inflammation and malignancy is difficult.2,6-8 Some authors only measure serum ferritin if the markers of microcytic anaemia are not consistent with IDA.6 Spuriously elevated serum ferritin levels can also be seen in hepatocellular damage and in situations such as non-steroidal anti-inflammatory agent-induced upper gastrointestinal bleed.6 Work in the interpretation of ferritin in the setting of inflammation and to improve guidelines is urgently needed.7,8
A further consideration is whether a serum ferritin is logistically possible or economically feasible in the primary health care setting in all patients in our resource-constrained settings. Further work is required to expand the current evidence base.
Vannella et al. reported that 93.6% of enrolled patients were refractory to oral iron supplementation and were referred for investigation.9 Their study population had a high incidence of coeliac disease and Helicobacter infection, and they suggested non-invasive serological tests to better decide who might benefit from gastroscopy.9 It therefore appears that referral for further investigation and delineation of the cause of IDA in 'low-risk' patients should be reserved for those who prove refractory to a trial of oral iron treatment.
With regard to assessing the response to oral iron therapy, an increase in Hb of 2 g/dl in 3 weeks is historically considered as an adequate response, but there may be better biomarkers, e.g. soluble transferrin receptors.2
Consequently, there is a need for guidelines that are clearly evidence-based, depend on multiple biomarkers, and take cognisance of varied conditions, and we call for further work and collaboration across disciplines.2
1. Linegar AG, Knottenbelt JD, Wormald PJ. Accuracy of a portable haemoglobinometer in clinical practice. S Afr Med J 1991;79: 547-548. [ Links ]
2. Clark SF. Iron deficiency anemia: diagnosis and management. Curr Opin Gastroenterol 2009; 25: 122-128. [ Links ]
3. Discombe G, Goddard AJ, Hill ID. The accuracy of haemoglobin determinations with the Spencer haemoglobinometer. J Coll Gen Pract 1966; 12: 251-252. [ Links ]
4. Neville RG. Evaluation of portable haemoglobinometer in general practice. Br Med J (Clin Res Ed) 1987; 294: 1263-1265. [ Links ]
5. Adamson EA, Bailey GR, Richards N, Wilson H. Prevalence of anaemia in an inner city primary school population. Arch Dis Child 2008; 93: 453. [ Links ]
6. Frewin R, Henson A, Provan D. ABC of clinical haematology. Iron deficiency anaemia. BMJ 1997; 314: 360-363. [ Links ]
7. Northrop-Clewes CA. Interpreting indicators of iron status during an acute phase response - lessons from malaria and human immunodeficiency virus. Ann Clin Biochem 2008; 45: 18-32. [ Links ]
8. Smellie WS, Forth J, Bareford D, et al. Best practice in primary care pathology: review 3. J Clin Pathol 2006; 59: 781-789. [ Links ]
9. Vannella L, Aloe Spiriti MA, Cozza G, et al. Benefit of concomitant gastrointestinal and gynaecological evaluation in premenopausal women with iron deficiency anaemia. Aliment Pharmacol Ther 2008; 28: 422-430. [ Links ]