Focus: Gujarat
Earthquake
|
|
|
|
Madhya Pradesh: Undernutrition and Starvation Deaths: An Inquiry An investigation of undernutrition
and suspected starvation deaths in a few selected villages of Barwani
district in Madhya Pradesh, a chronically drought-prone region, has thrown
up a number of issues impinging on the concepts, methods and processes
used for measuring of malnutrition and starvation. Barwani district in western Madhya Pradesh is one of the less developed districts of the state, with a large tribal population. This district with a total population of 10.8 lakhs (2001 Census) has a tribal population of about 65 per cent. The district has suffered the effects of severe drought during the last three years, which is a part of the larger scenario of failure of rains, which have affected many regions of western India. The SATHI team of CEHAT has been collaborating with a people’s organisation, Jagrit Adivasi Dalit Sangathan, a broad based health committee, Jan Swasthya Samiti and an NGO, Ashagram Trust in Barwani to develop a health initiative. Given the prevailing drought situation, these local organisations with inputs from CEHAT carried out an investigation of undernutrition in selected villages of Pati and Barwani blocks and conducted a survey of deaths in a few selected villages. The study involved a survey of undernutrition among one to five-year old childern; a study of undernutrition among adults; an assessment of suspected increase in death rates; and verbal autopsy to ascertain the causes of deaths occurring during a three-month period. Given the context of lowered food intake, the first step was a survey of nutritional status of children. The first stage of this survey was completed by ‘swasthya sathis’ (community health workers, all women), in 25 hamlets of nine villages in Pati block. The hamlets were basically those where swasthya sathis were based and hence it was a convenience sample. This survey covered all children in these hamlets who could be contacted (1,663 children) in the age range of 1 to 5 years. The swasthya sathis were given a brief training in the use of the ‘mid arm circumference tape’ and were shown how to record the findings for each child on a specially prepared record sheet. This was necessary because all of the swasthya sathis are non-literate and could not write, but could record the status of each child (red, yellow or green) on the coloured record sheet. They measured the mid arm circumference of children and all those falling in the red zone, i e, with mid arm circumference less than 12.5 cm were classified as malnourished. The results when collected and analysed, showed that 1,260 (75.7 per cent) children were malnourished. However, it was felt that this was a comparatively less sensitive method, and a more accurate estimation of the grade of undernutrition was needed. Therefore certain hamlets were selected from Pati and Barwani blocks for a weight-for age survey. These hamlets were randomly selected from clusters where there were ongoing activities of local organisations involved in the Jan Swasthya Samiti. In this way 10 hamlets from Pati and nine hamlets from Barwani block were selected. All the children in these 19 hamlets – a total of 712 children between age 1 to 5 years – were weighed, their age in months was obtained and nutritional status determined based on weight-for-age. This included recording of age in months, measurement of weight, recording of presence of edema and skin/hair changes and taking a brief dietary history. The standard IAP criterion – weight of the child being less than 80 per cent of the expected weight – was used to define undernutrition. The analysis of
this data has yielded the results in Table 1. Some 84 per cent
of the children in these villages were found to be malnourished and
nearly 22 per cent were found to be suffering from severe malnutrition.
It should be noted that these severely malnourished children are at
significant risk of succumbing to fatal infections if malnutrition is
not corrected. For comparison, Table 2 gives the nutritional
status of children in tribal areas of MP as per NNMB survey, 2000.
This data from National Nutrition Monitoring Bureau (NNMB) Diet and Nutritional Status of Tribal Population Report on First Repeat Survey is according to Gomez classification and hence has a higher cutoff point for normalcy (>90 per cent expected weight for age) which results in this extremely high prevalence of malnutrition (92 per cent). Here it may be noted that if we use the Gomez classification, about 98 per cent children in the Barwani sample would be classified as malnourished. However the figures for severe undernutrition found in Barwani district (22 per cent) seem comparable to the figures found for the general tribal population of Madhya Pradesh according to NNMB (20.4 per cent). Here the cut-off points in both classification systems are the same, i e, <60 per cent of expected body weight, and hence the proportions are directly comparable. Table 3 shows
detailed agewise analysis of nutritional status of under-five children,
however it is a sub sample of 311 children in five villages of Pati
block which have been analysed in detail. Severe undernutrition was
found to be commonest among the recently weaned (12-23 months) group
– as high as 42.1 per cent – and less common in higher age
groups.
Undernutrition among Adults Adults were also
examined for undernutrition in selected villages as part of the study.
Adult undernutrition was assessed based on the Body Mass Index (BMI)
or (weight in kg/height in metres) with the following standard classification.
132 adults (above
age 18) were examined in two villages for weight and height and their
BMI was calculted. The findings were in Tables 4 and 5.
According to the findings, 63 per cent of adults were undernourished, 15 per cent adults had a BMI of less than 16 which is indicative of chronic hunger/energy deficiency and severe undernutrition, which can contribute to the development of life-threatening illnesses. Undernutrition among men seems to be more (73 per cent) than women (54 per cent). Severe undernutrition among men (18 per cent) is also somewhat greater than among women (12 per cent). According to the NNMB data collected from the tribal areas of Madhya Pradesh in general, 49 per cent of the adults are undernourished and 8 per cent are severely undernourished. As against the findings in the Barwani study, the NNMB data shows slightly greater prevalence of undernutrition among women in MP (50 per cent) as compared to men (47 per cent). Severe undernutrition too is greater among women (9 per cent) than men (6 per cent). Severe undernutrition in Barwani in adults (15 per cent) is double that of tribal areas of Madhya Pradesh in general (7.9 per cent), at the same time overall adult undernutrition is significantly greater here (63 per cent) than the NNMB data (49 per cent). Undernutrition in men at Barwani (73 per cent) seems to be much worse than MP tribals in general (47 per cent) and severe undernutrition among men in Barwani (12 per cent) is double that of NNMB data (6.3 per cent). On the other hand the prevalence of undernutrition in women seems to be comparable in Barwani and tribal areas of MP (54 per cent and 50 per cent respectively), while severe undernutrition is again somewhat higher in Barwani (12 per cent) than the MP tribal population (9.3 per cent). Overall tribal areas in MP do not score well on nutrition parameters, but the drought situation prevailing in Barwani seems to have taken an additional toll and may be responsible for the higher level of undernutrition. Suspected Increase in Death Rates Anecdotal reports were received about unusually high death rates in certain villages. Three villages in one cluster, from where there had been some reports of suspected starvation deaths, were taken up for investigation of all the deaths that had taken place in the last one year. The families of the deceased were visited, the date/month of death were recorded for all deaths in the past one year. In all, 70 deaths during the last year were documented in these three villages. To confirm the time span of these deaths and in order not to miss any deaths, an attempt was made to compare this data with the mortality records maintained by the ANM. However this additional corroborative data could not be obtained from two of the three villages as the local health authorities refused to cooperate in giving this information. Therefore the actual number of deaths is actually likely to be higher than what we could document. Local calendar, local festivals, phases of the moon and local market days were used to ascertain the date of death in case of all deaths in the last three months. The exact number of deaths in these three months was used for the calculation of death rates, as this being a short recall period, the date of deaths could be assessed fairly accurately. This data, based on the population of the village and concerning three month’s span was extrapolated for a population of 1,000 and a span of a year and the death rates were calculated. It was found that the three monthly death rate of Semli village was 3.5 deaths per 1,000 population, of Verwada it was 2.65 and of Sipahiduwali it was 5.33 deaths per 1,000 population. This brings the annual rate of death per 1,000 population in these villages to 14, 10.6 and 21.3 respectively. The total annual death rate in these three areas combined was 14.15. The average death rate in the three villages during these three months (14.15 deaths per 1,000 population per year) is somewhat higher than the crude death rate of MP (11 deaths per 1,000 population per year, SRS). However, given the small population base and time period in this sample, it was not considered possible to draw definitive conclusions from this information. Use of Verbal Autopsy It was felt that more definitive investigation of the cause of all deaths occurring in these villages in the recent period was required in order to ascertain whether these were starvation deaths. This exercise was also considered essential to develop a methodology to study and diagnose starvation deaths. It was decided to conduct a verbal autopsy (VA) on each death occurring in the selected villages during the last three months. This meant a detailed investigation of the symptoms and signs, bodily appearance, history of food intake, family food supply and other relevant factors in case of each death. Although a standardised verbal autopsy questionnaire for childhood deaths, standardised by WHO along with John Hopkins School and London School of Hygiene and Tropical Medicine is available, such was not the case with the VA questionnaire for adults. So taking a general-purpose VA questionnaire from an article in WHO Bulletin as the basis, a modified VA form was prepared which could take care of ‘starvation related deaths’. Questions relating to food intake and family food supply were added to this questionnaire. This was done based on suggestions from C S Kapse (head of department forensic medicine, D Y Patil Medical College, Pune), Veena Shatrughna (jt director, National Institute of Nutrition, Hyderabad), and H H Trivedi (ex-professor, department of medicine, M G Medical College, Bhopal) who gave their detailed and valuable inputs to modify this questionnaire and also agreed to be the expert panelists for analysing the results. The VA questionnaire had the following major sections.(1) Personal identification details. (2) Family food supply related information including irrigated and non-irrigated land owned, state of the harvests, wages earned and ongoing government relief work, categorywise status of food items being consumed by the family. (3) Individual dietary history during the week and the month before death. Calorific value of each food item consumed was calculated in order to analyse this data. (4) Unnatural food consumption patterns such as begging or borrowing food, consumption of unusual foods such as leaves of plants, forest tubers, etc. (5) Signs and symptoms during the last illness, as well as any medical records and prescriptions. (6) Physical appearance at the time of death. Establishing the case of death – at the end of the questionnaire there were sections to record (a) immediate cause of death; (b) underlying cause of death; and (c) contributory cause of death. This diagnosis was to be made by each panelist based on the findings of the Verbal Autopsy. For this the complete sets of copies of all the filled questionnaires were sent to each of the three panelists for analysis. Their opinion about immediate, underlying and contributory causes of death were taken and collated. A drawback noticed in the process was that the currently available verbal autopsy questionnaires are quite medicalised. Although good quality training was imparted regarding it’s administration, and the help of locally available doctors was taken wherever necessary, it was felt that a qualified doctor would have been better suited for the job. This again brings us into the realm of over-medicalising an essentially social problem. Not only that ‘death related to starvation’ almost never appears in the death certificate of a doctor certifying a death, but even undernutrition seldom appears as an underlying cause. To add to this we could not obtain any clear guidelines about how to define a starvation death, or the parameters to certify such a death (see discussion). A tool less medical in nature, and one that can be administered with minimal training is definitely needed. Our modified verbal autopsy form might be seen as one step in the direction of developing such a tool. All the deaths (19 deaths), which occurred during March 2001 to May 2001, in the villages of Semli, Sipahiduwali and Verwada were investigated. These villages were purposively selected, where local activists suspected an unusually large number of deaths. Six deaths during the six months prior to this period (September 2000 to February 2001), which were strongly suspected to be starvation deaths were also analysed. Thus a total of 25 deaths were investigated, seven of which were children, and 18 were adults. The questionnaire was translated into Hindi. Bhausaheb Aher who administered the questionnaire and Amulya Nidhi who assisted him (both MSWs) were both trained and acquainted with the medical phrases and clinical conditions that appear in the questionnaire. This was pre-tested, by investigating the deaths of five children and four adults in these villages, and then finalised for use. Both the questionnaires were used to investigate all the deaths in the selected three villages – Semli, Sipahiduwali and Verwada. So far the verbal autopsy forms for all the adult deaths have been analysed (18 deaths). The completed forms were sent to all three panelists who gave their independent opinions, which were then compiled in a table. If at least two of the three panelists stated starvation or malnutrition as the underlying cause of death, this was taken as the probable underlying cause of death. A similar definition was used for a probable contributory cause of death. Among the 18 adult deaths, in case of three adult deaths, starvation was identified as the probable underlying cause of death. Apart from these, in an additional three deaths, starvation or undernutrition has been identified as the probable contributory cause of death. Some of the many issues, which emerged in the course of this study are: (A) Though the phenomenon of starvation is widely discussed and reported, we could not find a very clear definition of starvation and especially no clear definition of a ‘starvation death’. However some indicators which were pointed out by Veena Shatrughna and which helped to initially guide us when we were grouping for some working definitions, were as follows: a dietary intake in adults of less than 500 kcal per day is starvation (NIN report on drought in Gujarat); doubling of the proportion of adults with a BMI of less than 16, compared to the baseline, is indicative of starvation; and consumption of abnormal or unusual foods (forest leaves/tubers/wild fruits not usually eaten, etc) is indicative of starvation. However, the first two of these criteria seemed problematic when we started actually applying them to the situation. The third seems useful but may not be a sufficient criterion in itself. The NIN criteria of starvation (mentioned in passing in its report on drought and malnutrition in Gujarat) of ‘less than 500 kcal intake per day’ seems inadequate. Those having such low consumption levels would definitely be starving. But what about those adults consuming between 500 and 1,400 kcal per day? They are consuming below the amount required for basal metabolic functions, and considering the additional fact that they would be doing some physical activities to obtain food, they would be continuously and perhaps rapidly losing body weight (probably muscle mass since fat reserves would be long gone). This is clearly an unsustainable situation. Given the fact that continuing such a low level of dietary intake will in the course of time inevitably lead to fatal results, what distinguishes this from starvation? There seems little
sense in talking of the ‘baseline’ for a community that
is already chronically undernourished. If statewise 9 per cent tribal
adults have a BMI <16, then why should ‘doubling’ of
this to 18 per cent be a criteria for starvation? By this criterion,
worse-off states would have a higher cut-off point for starvation! If
at all we have to make a comparison, should it not be with a general
standard rather than with a ‘baseline’ of an already unacceptably
poor nutritional status? In this context we can refer to the criteria
laid down in the WHO expert committee report on Anthropometry (WHO TRS
854, 1995) related to classifying low BMI as a public health problem:
By this classification, the situation in MP tribal areas in general is already ‘critical’ (49 per cent population with BMI <18.5) and the situation in Barwani can only be called ‘supercritical’ (63 per cent population with BMI <18.5)! (B) Starvation emerges
more and more as a public health problem requiring community diagnosis.
In this sense starvation deaths differ from classical ‘disease
related mortality’. The diagnosis of a death due to tuberculosis
might be considered an individual diagnosis. But the diagnosis of a
‘starvation death’ cannot be just an individual diagnosis;
we have to document the circumstances prevailing in the family and community
along with the individual to reach such a conclusion, for the simple
reason that starvation is a deeply social phenomenon. In fact, looking
at the scale and depth of malnutrition in tribal and rural areas of
our country, making individual diagnosis of ‘starvation deaths’
may seem almost incidental to the main issue. These deaths, though tragic
and extremely unfortunate especially since they could have been so easily
prevented, are just the tip of the iceberg of a situation of near universal
undernutrition in rural and tribal areas. However, the paradox is that
the government can ignore or downplay the fact that hundreds of millions
of children and adults lead lives of severe, lifelong undernutrition
since it does not provoke any public outcry. But a few starvation deaths
reported in the press make the entire government machinery go into overdrive
to ‘deny’ such an event and take some emergency measures.
Even civil society and middle class opinion which starts wringing hands
at the mention of starvation deaths, remains impervious to the implications
of findings such as NIN data according to which around 90 per cent of
children in rural areas are undernourished! So what do we do –
focus on the near universal community undernutrition/starvation or on
the few starvation deaths? One emerges as the main problem from
a public health perspective while the other is an emer-gency with
the advocacy impact of moving public opinion and the government system.
Can we develop an approach to adequately understand and document both?
Or do we interpret this as a silent yet alarmingly widespread situation of chronic severe undernutrition, with drought as an additional aggravating factor, a last straw which breaks the camel’s back and leads to obvious deaths? In summary, there seems to be an urgent need to be able to utilise the wealth of existing data on widespread malnutrition for effective advocacy, to enable people to access the right to food security. At the same time, there is a specific value of documenting starvation, or a state of extremely reduced food intake which in due course is incompatible with life, both because it is a humanitarian emergency and because it may help to shake an otherwise complacent state and civil society into some action. The challenge is to develop an approach, which has both breadth and depth, which views malnutrition and starvation as public health problems with a social dimension, and to wield such a tool effectively to establish the entitlement of all for basic food security. [The study was undertaken by the SATHI team at the Centre for Health and Allied Themes (CEHAT). Bhausaheb Aher has been involved in collecting data and conducting interviews for this study as part of his block placement with CEHAT. Abhay Shukla, Amulya Nidhi and Amita Pitre of the SATHI team have participated in various aspects of design and analysis. Activists of Jagrit Adivasi Dalit Sangathan and Ashagram Trust, Barwani contributed in various ways during all stages of the study.] Economic and Political Weekly, May 3,2003 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||