Reduction in prevalence of anaemia in pregnant women

Sir,

We read the article by Kalaivani and Ramachandran1 with great interest. The study analyzed the datasets of National Family Health Survey (NFHS)-II, III, IV, District Level Household Survey (DLHS) II, IV and Annual Health Survey (AHS)-Clinical and Anthropometric and Biochemical (AHS CAB). Authors reported that there has been a reduction in the prevalence of anaemia among pregnant women in the past 15 years.

We would like to discuss a few concerns about the interpretation of the predicted trend of reduction in the prevalence of anaemia using the National Survey data:

1. The sampling procedure and inclusion criteria of Pregnant women differed greatly in NFHS II (1998-1999)2, NFHS-III (2005-2006)3 and NFHS-IV (2015-2016)4 as compared to DLHS-II (2002-2004)5, DLHS-IV (2012-2013)6 and AHS CAB surveys (2014)7 presented in the study1.

2. NFHS II and III were designed to provide State-level estimates of anaemia. However, DLHS II, IV, AHS and AHS CAB and NFHS IV were designed to provide district level estimates of anaemia. There were large variations in the total number of pregnant mothers included for estimation of prevalence of anaemia in NFHS-II (n=2796), III (n=3788) and IV (n=30,320), DLHS II (n=38,710), IV (n=12,306) and AHS CAB (n=20,832).

3. The methods used for haemoglobin estimation were different in NFHS-II, NFHS-III and NFHS-IV as compared to DLHS-II, IV and AHS surveys. The NFHS used HemoCue analyzer for estimation of haemoglobin while DLHS-II, IV utilized cyanmethaemoglobin method. In addition, different models of HemoCue analyzers were used during NFHS-II, III and IV as consistent results of Hb estimations were not produced by the earlier models of the machine8.

4. The classification for grading of anaemia used in NFHS-II, NFHS-III and NFHS-IV as compared to DLHS-II, IV and AHS surveys also differed across the surveys. The NHFS II, III and IV graded anaemia according to the WHO grading of anaemia9; pregnant women with Hb levels ≥11 g/dl were graded as non-anaemic; those with Hb levels between 10.0 and 10.9 g/dl as mildly anaemic, those with Hb levels between 7.0 and 9.9 g/dl as moderately anaemic and those with Hb levels below 7.0 g/dl as severely anaemic. Whereas, DLHS 2 used the grading of anaemia as per the earlier published Indian data based on functional decompensation101112 which has been associated with a fall in Hb levels. Pregnant women with Hb ≥11 g/dl were graded as not anaemic; while those with Hb levels between 8.0 and 10.9 g/dl as mildly anaemic, those with Hb levels between 5.0 and 7.9 g/dl as moderately anaemic and those with Hb levels below 5.0 g/dl as severely anaemic.

5. All the NFHS surveys (II, III and IV)234 have documented lower prevalence of anaemia in pregnant women as compared to non-pregnant women. This is in contradiction to the existing knowledge according to which the prevalence of anaemia among pregnant women is higher due to haemo dilution during pregnancy9. The WHO also has recommended lower “cut-off” for Hb by 0.5 g/dl for defining anaemia among pregnant mothers9.

6. High reduction in the prevalence of anaemia was recorded between NFHS-III and IV in Chhattisgarh (63-41%), Assam (72-44%), Haryana (71-51%), Odisha (68-47%) and Kerala (62-45%). However, the coverage of iron folic acid (IFA) supplementation (major intervention to reduce anaemia) among pregnant women was only 30.3 per cent in Chhattisgarh, 32.0 per cent in Assam, 32.5 per cent in Haryana, 36.5 per cent in Odisha and 67.1 per cent in Kerala in NFHS-IV4. The distribution of IFA tablets and monitoring of their consumption were poorly undertaken due to various logistic reasons. Furthermore, anaemic pregnant women, possibly received only prophylactic dose of iron (instead of therapeutic dose) while the majority of them were suffering from anaemia234. It has been suggested that only up to 50 per cent of women with anaemia in countries of South East Asia region are amenable to iron supplementation13.

7. The drastic reduction in the prevalence of anaemia mentioned in this study1 between DLHS II to AHS conducted in Odisha (97-62%), Chhattisgarh (96-63%), Jharkhand (97-80%) and Madhya Pradesh (97-71%) could possibly be due to limitations in the process of estimation of haemoglobin rather than health interventions for reduction in anaemia.

In view of the above, combining the haemoglobin data of NFHS, DLHS and AHS surveys and concluding reduction in the prevalence of anaemia possibly does not provide true scenario. True trends in the prevalence of anaemia could be provided by utilizing data from similar sampling framework surveys with the same method for haemoglobin estimation.

Conflicts of Interest: None.