Indmedica Home | About Indmedica | Medical Jobs | Advertise On Indmedica
Search Indmedica Web
Indmedica - India's premier medical portal

Indian Journal of Community Medicine

A Comparative Study on Prevelance of Anaemia in Women by Cyanmethaemoglobin A

Author(s): M. Mohanram, G.V. Ramana Rao*, J. Gowrinath Sastry**

Vol. 27, No. 2 (2002-04 - 2002-06)

p>Director, Indian Institute of Health & Family Welfare (IIHFW), Vengalrao Nagar, Hyderabad - 500038 *Deptt. of Epidemiology, IIHFW Vengalrao Nagar, Hyderabad - 500038 **National Institute of Nutrition, Adikmet, Hyderabad - 500044


Research question: What is the extent of agreement between Gold Standard Cyanmethaemoglobin (HiCN) and HemoCue (HC) methods in measuring Haemoglobin (Hb) levels in women?

Hypothesis: HemoCue method tends to overestimate the levels of Hb and as a result the prevalence rates of anaemia would be lower using this method.

Objective: Evaluate the extent of agreement between HiCN and HC methods in measuring Haemoglobin levels in women.

Study design: Cross-sectional, observational, clinic and laboratory based study.

Setting: Niloufer Hospital for women and children and Urban Health Centre, Hyderabad.

Participants: Pregnant and Non-Pregnant women.

Sample size: 207 pregnant and 212 non-pregnant women.

Study variables: Hb levels and severe anaemia in women.

Statistical analysis: Mean, standard deviation, regression analysis, percentage distribution.

Results: Mean difference in Hb levels by the two methods was found to be 2.08 with a SD of 0.08. The HC method overestimated Hb as compared to HiCN method. Correlation coefficients (r) were significantly different from zero and the regression coefficients (b) were significantly different from one as well. The data analysis clearly suggests that the correction factor would vary depending upon the level of Hb. Sixty five per cent of pregnant women were found to be "normal" by HC method in comparison to 22% by HiCN method.

Conclusions: HiCN method should continue to be the method of choice for assessment of anaemia prevalence.

Keywords : Haemoglobin, Anaemia, Pregnant and Non-pregnant women,Standard Cyanmethaemoglobin method, HaemoCue method


Iron deficiency anaemia is the most common nutritional disorder in the world, affecting over a billion people1. It is estimated that 90% of anaemic people reside in the developing countries with highest prevalence in South Asia including India. Assessment of iron status has traditionally been centered around quantitative measurement of haemoglobin (Hb) level and circulating red cell count. Cyanmethaemoglobin (HiCN) method has been the preferred choice for measuring the blood Hb level and as per the recommendations of the International Committee for Standardisation of Haematology, all other methods should be standardised and adjusted with this method2.

Recently, HemoCue (HC) method has been developed and advocated for routine measurement of Hb in nutritional surveys, because of ease and quickness in its operational features3. In the recently conducted National Family Health Survey - II (1998-99) in India, Hb levels were determined by using HC method and lower prevalence of anaemia was reported among pregnant women than in non-pregnant women4. Further, the overall prevalence of anaemia was much lower when compared with the results of other large scale community studies in the country5. Because of the policy implications involved in the interpretation of these conflicting prevalence figures, it was felt necessary to study the extent of agreement in the Hb values obtained by these two methods. A field-based study was, therefore, undertaken to examine this issue among women.

Material and Methods:

After obtaining prior informed consent, 207 pregnant women attending Antenatal Clinic at the government maternity ward of Niloufer Hospital and 212 non-pregnant women from Urban Health Centre in Hyderabad were included in the study. A purposive sample of about 400 was selected based on the resources available. HemoCue apparatus was supplied by International Institute of Population Sciences (IIPS), Mumbai to IIHFW, with these particulars - B-haemoglobin photometer, HemoCue AB, Angelhelm, Sweden (Equipment number 9824002379). Hb estimations were carried out simultaneously by both HiCN and HC methods in all the 419 samples. Information on age and gestation period in the case of pregnant women was also recorded. Gestation period was recorded because pregnancy would lead to hemodilution which may differentially affect the relationship between these two methods. The mean age of pregnant women was 22.9 years with range of 15-39 years; while for non-pregnant women mean was 27.6 years and range was 15-49 years. For estimation of Hb by HiCN method, 20 ul of capillary blood samples were collected by finger prick and transferred to Whatman filter paper No.1 and labelled appropriately. Hb levels were measured by HiCN technique within 4-5 days after collection of samples2 at the National Institute of Nutrition, Hyderabad. For HC method used in the field, finger prick blood was drawn into self-filling control micro-cuvettes with reagents in dry form. Immediately, the cuvettes were placed in the HemoCue equipment, which directly displayed the Hb levels within 45-60 seconds6.

Data on haemoglobin levels obtained through HiCN and HC methods were entered in Statistical Package for Social Science (SPSS) and analysis was carried out. Levels of anaemia among pregnant, non-pregnant and combined groups were determined adopting the WHO criteria7. Regression analysis for these groups was carried out and levels of significance calculated.*


Table I: Haemoglobin values (g/dL) among women (Mean S.D.).

Group Number HiCN method HC method
Non-pregnant 212 10.46 ±1.75 12.46 ±1.90
Ist trimester 9 9.93 ±1.44 12.29± 1.04
2nd trimester 73 9.83±1.84 11.84 ±2.19
3rd trimester 125 9.30 ±1.71 11.22±2.09
Pooled 207 9.52 ±1.76 11.49 ±2.11
Grand total 419 9.99±1.81 11.98± 2.06

The data pertaining to the differences in Hb values is presented in Table I. It was observed that HC method overestimated Hb levels, on an average, by about 2 g/dL, which was significantly different from zero (p<0.01). Only 18.7% of the values differed within 1 g/dL. In 51.3% of the women studied, the HC method overestimated the Hb levels by more than 2 g/dL.

Table II: Regression analysis of the data.

Group Corr. Coeff (r) Regression equation SE of the estimate
Non- pregnant 0.762 HiCN = 1.700 + 0.703 (HC) *±0.519)±0.041) 1.136
p<.001 p<.001, p<.001
Pregnant 0.680 HiCN = 2.995 + 0.568 (HC) *±0.499)±0.043) 1.293
p<.001 p<.001 p<.001
Pooled 0.736 HiCN = 2.232 + 0.648 (HC) *±0.355)±0.029) 1.231
p<.001 p<0.001 p<.001

*Figures in the parentheses indicate the Standard Errors (SEs) of the coefficients.*

Regression analysis was undertaken to see if a constant correction factor could be identified to convert data of HC method to that of the HiCN method. The results are given in Table II. The analysis showed that:

  1. The correlation coefficients were significantly different from Zero (p<.001), but are low in magnitude. This is reflected in the large SE of the estimate (mean), which is more than 1.0 g/dL;
  2. The intercepts as well as the regression coefficients were significantly different from zero; and
  3. The regression coefficients were, also, significantly different from one. This means that a constant correction factor cannot be applied to convert data of HC method to that of HiCN method.

When the study subjects were categorised into different grades of anaemia by Hb levels using WHO criteria, it was observed that a large proportion of pregnant women were normal (65%) by HC method as against (22%) by HiCN method. While the HiCN method revealed that 55% of pregnant women were suffering from "moderate/severe" anaemia, the values obtained through HC method placed only 21% of women in this category.


It was earlier reported that the HC method provides consistently higher values of Hb than Blood Cell Counter (BCC) method and a constant correction factor 0.5 g/dL was suggested to be subtracted from the Hb estimates obtained by this method8. However, their results clearly indicate that the regression coefficient (b) is significantly different from 1(p<0.05) which does not support the use of constant correction factor, even on averages. The regression coefficient arrived at in the present study (Table II) is significantly different from not only zero but also from one indicating that a single factor cannot be applied to correct the overestimated values obtained by HC method. The data analysis clearly suggested that the correction factor would vary depending upon the level of Hb.

Recently, Sari et al9 have estimated Hb concentrations in 121 mothers using HiCN and HC methods and reported that the prevalence of anaemia was significantly higher when Hb levels were analysed by the indirect HiCN method involving the transfer of blood samples to Whatman filter paper, which was dried and later eluted for Haemoglobin estimates. The higher estimates of anaemia prevalence by the indirect HiCN method was attributed to possibly incomplete dissolution of blood into Drabkin's solution. The investigators suggested that the methodological differences between direct and indirect methods of HiCN should be examined critically for assessing the prevalence of anemia. The analysis carried out in the present study, however, does not support the contention of Sari et al9 as the blood samples collected on Whatman filter were found to be completely eluted as against the assumption made by them. The Hb values simultaneously estimated by the direct and indirect methods of HiCN were found to be in close agreement (Sivakumar & Madhavan Nair, NIN, personal communication) which rules out such a possibility. Also, one should use high quality Whatman #1 filter paper and ensure that the analysis be undertaken by indirect method within one week of collection of the blood samples.

Morris et al10 have used an emperical data-set to measure reliability, precision and accuracy of the portable haemoglobinometer; and observed that errors found in this method are due to unreliability of the values leading to miscalculation of anaemia status.

The data obtained in this study thus clearly suggests that until such time an alternative and reliable method is identified, HiCN method should continue to be the method of choice for assessment of anaemia prevalence.


We thank the Director, National Institute of Nutrition, Hyderabad, for extending the co-operation in measurement of the Haemoglobin levels in the study samples by HiCN method.


  1. Latham MC. Human nutrition in the developing world, FAO Food and Nutrition Series, FAO of UN, Rome. 1997: 147-55.
  2. Measurement of Iron status: A report of the International Nutritional Anaemia Consultative Group (INACG), The Nutrition Foundation, Inc. 1126 Sixteen St. NW, Washington, DC. 20036 USA, 1985: 4-8.
  3. WHO/UNICEF/UNU. Report on contribution on indication and strategies for iron deficiency and anaemia programme. World Health Organisation, Geneva, 1994: 15-26.
  4. International Institute for Population Sciences (IIPS) and ORC Macro 2000. National Family Health Survey, India (NFHS-2), 1998-99, Andhra Pradesh: Mumbai, IIPS. 152-6.
  5. Indian Council of Medical Research, Micronutrient Deficiency Disorders in 16 districts of India, Part 1, Report of an ICMR task force study, District Nutrition Project. Indian Council of Medical Research, New Delhi, 2001.
  6. Hudson-Thomas M, Bingham KC, Simmons WK. An evaluation of HemoCue for measuring haemoglobin in field studies in Jamaica; Bull World Health Organ, 1994; 72(3): 423-6.
  7. WHO, Geneva, Nutritional Anaemia, Report of WHO Group of Experts; Technical Report Series 503, 1972, Annex-2, 29.
  8. Prakash S, Kapil U, Singh G, Dwiwedi SN, Tandon M. Utility of HemoCue in estimation of haemoglobin against Standard Blood Cell counter method. J Assn Physicians India 1999; 47(10): 995-7.
  9. Sari M, Saskia de P, Martini E, Herman S, Sugiatmi, Bloem MW et al. Estimating the prevalence of anaemia; A comparison of three methods WHO 2001; Bull World Health Organ 2001; 79(6): 506-11.
  10. Morris SS, Ruel MT, Cohen RJ, Dewey KG, dela Briere B, Hassan MN. Precision, accuracy and reliability of haemoglobin assessment with use of capillary blood. Am J Clin Nutr 1999; 69(6): 1243-8.
Access free medical resources from Wiley-Blackwell now!

About Indmedica - Conditions of Usage - Advertise On Indmedica - Contact Us

Copyright © 2005 Indmedica