IAP vs ZScore Classification for Growth Charts

1
IAP vs Z-Score Classification for Growth Charts

• A.K.Nigam
• Director
• Institute of Applied Statistics and Development
  Studies

2

• Prevention and management of severely
  malnourished children is an important component
  of the ICDS and IMCI (Integrated Management of
  Children Illness) strategy. WHO also emphasizes
  on the management of severely malnourished
  children.
• Prevalence of malnutrition in children is
  described in terms of the percentage of
  individuals below a specific cut-off, such as
  certain per cent of the median or standard
  deviation (sd) in terms of z scores of the
  reference population. Classifications like
  Gomezs or Indian Association of Pediatricians
  (IAP) are on the basis of cut - off points as
  percentage of median (reference) weight-for-age.

3
(No Transcript)
4

• NNMB/NIN reports (UNICEF also in some studies)
  nutritional status using each of these
  classifications with no inter-comparisons and
  guidelines about which one should be used for
  programme interventions.
• SD classifications are now widely used by major
  stake holders like WHO, USAID among others. NFHS
  I, II, and the latest III, the largest national
  surveys, which form the basis for planning by
  policy makers and programme implementers, present
  nutritional status data in terms of sd
  classifications only.
• One is therefore in a dilemma which
  classification to use.

5

• Although sd classification has distinct
  statistical advantages over others, its use is
  not widespread, mainly because it is felt (though
  erroneously) that it involves relatively
  cumbersome calculations. Even research workers
  find them difficult, as it needs development of
  appropriate software for different indicators of
  nutritional status.
• Because of this, ICDS uses IAP classification. At
  each AWC, the Anganwadi Worker has to prepare and
  monitor growth chart for each child. These charts
  are based upon IAP classification and use 60 of
  median (standard) weight-for-age for identifying
  severely malnourished children.
• It is known that the prevalence of severe
  malnutrition as derived by ICDS functionaries is
  far below in comparison to that reported by NFHS
  and other nutrition surveys. Even for the same
  survey (say, NNMB/NIN) the prevalence of severe
  malnutrition using below 3sd cut-off is much
  higher in comparison to that obtained using IAP
  cut-off of 60 of median weight.

6
Gap between NCHS reference and IAP classification
in assessing severe undernutrition in boys
7

• Nigam (2003) showed that 3sd cut-off matches with
  67 of median (standard) weight-for-age. One
  readily notices the rationale behind the gap in
  reporting by realizing that 3sd cut-off
  corresponds to 67 (instead of 60 as in IAP) of
  median (standard) weight-for-age.
• The equivalence relation facilitates the use of
  sd classification under field conditions and also
  by research workers. The growth charts, in terms
  of both IAP and sd classifications are given in
  another paper by Nigam (2005) separately for boys
  and girls. These charts can replace the existing
  growth charts at AWC.

8
(No Transcript)
9
(No Transcript)
10

• A comparison was made by Nigam (2005) between
  the two approaches, 60 and 67 of median, to
  evaluate the percentage of severely malnourished
  children being left out by IAP classification
  which is being used by agencies like ICDS. For
  this, district level results from the NIN/IASDS
  district level Reports from the study
  Nutrition Profile of Community in Uttar Pradesh,
  were utilized for Uttar Pradesh and Uttaranchal
  states.

11
(No Transcript)
12
At the State level, 61.1 of severely
malnourished children were left out in UP. At the
regional level, in Uttar Pradesh, these left out
percentages were 58.1 in Western region, 59.9
in Central, 67 in Eastern and 53.4 in
Bundelkhand. In numbers, in UP alone, out of
about 6 million estimated severely malnourished
children over 3.5 million such children were
likely to be left out.
13

• The results reveal and support earlier findings
  that the percent of median cut-off points under
  IAP do not capture a substantial number of
  children identified as malnourished through sd /
  z- score classifications.
• As in most programme intervention projects,
  only severely malnourished are targeted and
  monitored, the huge gap in the two assessments
  also raises ethical considerations.
• As ICDS uses IAP classifications for growth
  monitoring and identifying severely malnourished
  children, it is not difficult to realize the
  gravity and magnitude of the problem with regard
  to left out severely malnourished children at
  the national level. Corrective measure in this
  direction would prove to be very effective in
  tackling malnutrition deaths.

14
New WHO growth standards for assessing severe
undernutrition in boys
15
New WHO growth standards for assessing severe
undernutrition in boys
16
Growth Curves (Severe Undernutrition) For Boys
From WHO And NCHS Populations (0-59 months)
17
Growth Curves (Severe Undernutrition) For Boys
From WHO And NCHS Populations (0-24 months)
18
Growth Curves (Severe Undernutrition) For Boys
From WHO And NCHS Populations (24-59 months)
19
WHO Standards Some Observations/Concerns

• It is seen that in 24-59 month period -3sd
  cut-offs vary considerably with average cut-off
  70 of median.
• Because for weight-based measures outer tails are
  highly affected by even few extreme data points.
• This could also be because of lower emphasis
  given to uniformity on different aspects of
  complementary feeding initiation, quantity,
  quality and frequency among children of
  cross-sectional group. This perhaps explains
  lower dropouts in cross-sectional design
• A limitation of WHO standards is non inclusion of
  some important ethnic groups (South East Asia,
  Australia, NZ etc.), and even in India children
  from cities from South and East.
• The overall design, a mix of longitudinal and
  cross-sectional generates confounding
  overlapping between 18-24 months, uneven visits
  in cross-sectional design.
• WHO standards are limited to 0-59 months
  children what about older children?

20
Overall Conclusion

• No single cut-off is entirely satisfactory
• More stratification is required within each
  country
• For some time both NCHS and WHO should be
  concurrently used on pilot basis.

21

• THANK YOU

22
Weight For Age Tables For BOYS and GIRLS In NCHS
Reference Population
23
BOYS GIRLS
24
BOYS GIRLS
25
Percentage of severely malnourished children left
out by IAP classification
26
(No Transcript)
27
(No Transcript)