DETERMINANTS AND CAUSES OF INFANT MORTALITY IN

GUJARAT AND MAHARASHTRA 

 

M.M. Gandotra* & N.P. Das+

INTRODUCTION

 

          The infant mortality rate has been recognised as a summary index of the socio-economic development of a region. This recognition has spurred international organisations and national governments to intesify their efforts to reduce the level of infant mortality and promote greater child survival. The sixth and seventh Five Year Plans of India have also aimed at nation-wide programmes in this direction as the government has adopted the national goal of halving the prevailing level of infant mortality to 60 by the turn of the century. It therefore become imperative to explore the underlying causative factors that impede the reduction in infant mortality.

          An examination of the state-wise data reveals that while Gujarat and its neighbouring state of Maharashtra are more or less at par with regard to the overall level of development, differentials were evident with respect to infant mortality. Estimates of infant mortality available from the SRS for Maharashtra and Gujarat were found to be 66 and 97, per thousand live births for the year 1986-87 (Govt. of India, 1988-89). On the other hand, NFHS-1 for Gujarat and Maharashtra revealed that IMR in Gujarat (PRC Baroda & IIPS Mumbai, 1993)  and Maharashtra (PRC Pune & IIPS Mumbai, 1992-93) were 69 and 51 per 1000 live births respectively during 1988-92. Such a situation was a cause of concern not only to the planners and policymakers but also for researchers. An indepth understanding of the causes for these differentials and the underlying factors that affect infant mortality in varying culture was required. To this end the present paper is an attempt to understand the determinants and causes of infant mortality in these two states.

 

REVIEW OF LITERATURE

 

          Way back in sixties Bourgeois-Pichat (1964) indentified two types of factors viz. ‘endogenous’ and ‘exogenous’ that affect infant mortality. Exogenous factors of infant mortality are dependant on environment in which an infant is exposed and include deaths to infants due to infectious, parasitic and respiratory diseases. Such causes normally occur in the post-neonatal period (1 to 11 months of age of infant) and they are easier to control. On the other hand, endogenous causes of mortality are more biological in nature and include deaths due to congenital malformations and birth process. They occur in the neonatal period (less than 1 month of age of infant) and are rather difficult to control.

 

          Mosley and Chen (1984) provided an analytical framework for child survival in developing countries. They grouped the proximate determinants into five categories viz. maternal factors (age, parity and birth interval); environmental contamination (air, food/water etc.); nutrient deficiency; injury (accidental, intentional); and personal illness control (preventive measures and treatment).

          Gandotra and Das (1988) latter categorized the underlying factors behind the immediate causes of infant deaths into five broad groups: demographic factors; socio-economic factors; environmental, sanitation and hygienic factors; nutrient availability factors; and medical care factors.

          Recently, Pandey et. al. (1998) while analyzing the NFHS-1 data considered child’s year of birth, child’s sex, mother’s age at child birth, residence, mother’s literacy, religion – caste/tribe membership, mother’s exposure to mass media, availability of toilet facility, type of cooking fuel and ownership of goods scores as the covariates of infant mortality.

          The present study tries to see the impact of some of these covariates on infant mortality in Gujarat and Maharashtra. The independent variables considered having an impact on the level of infant mortality are discussed in the subsequent section of this paper.

 

STUDY DESIGN

                                               

          In view of the intricate and varied nature of data to be generated, multiple research tools and design mixes were employed. With the main focus of the study to examine the relationship between the antecedent and consequent variables, the first step was to identify women with 7 and more months of pregnancy during a designated one year reference period, by the selected Auxiliary Nurse Midwife (ANM)/Female Health Workers (FHWs) in their respective areas of operation. Having identified these women, the ANMs/FHWs contacted them to obtain details on height, weight, measurement of mid arm circumference, haemoglobin level and on the ultimate outcome of pregnancy. Subsequently, the ANMs followed each live birth that occurred, at regular intervals, to obtain information on infant’s birth weight and other anthropometric measurements as well as their post-natal and survival status. These periodic follow-up visits occurred till the infant completed 12 months of age or till its death (if this occurred within the first 12 months of life). In the event of an infant death, the ANMs collected data on the age at death and clinical causes of death.

 

          Another set of data was generated from the follow-up of those women whose pregnancies terminated in a live birth. These women were interviewed by trained field investigators and supervisors to obtain details on various socio-economic, demographic, cultural, environmental aspects, as well as medical and non-medical pre and post-natal care taken.

 

          The data obtained through cross-sectional and lognitudinal survey was substantiated by data obtained through qualitative approaches, particularly focus group discussions with pertinent groups. The detailed design of the study is depicted in Fig. 1.


Fig 1 : Framework Indicating the Basic Design of the Prospective Study in Gujarat and Maharastra

 

 

 

 

 

TWO SELECTED STATES

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

GUJARAT

(Geographic Division)

 

 

 

MAHARASHTRA

(Geographic Divisions)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

South Gujarat

 

Saurashtra

 

North Gujarat

 

 

Bombay

 

Pune

 

Aurangabad

 

Nagpur

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Selected District

 

Selected District

 

Selected District

 

 

Selected District

 

Selected District

 

Selected District

 

Selected District

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

10 Selected PHCs from each district

 

 

 

Identification of pregnant women (7-9 months) during the reference period by ANM/female health worker to elicit information on prenatal status and to know ultimate outcome of preganancy

 

10 Selected PHCs from each district

 

 

 

 

 

 

 

 

 

 

 

1. Detailed survey of each of these households where women have given a live birth by trained investigators and their follow-up visit for second time to record postnatal status of the infants

 

2. follow-up of infants for 12 months after birth by the ANM/FHWs to elicit information on natal, post-natal & survival status as well as to obtain age at death and causes of death information in the event of infant mortality

 

 


Rural areas of the two selected states (Gujarat and Maharashtra) were chosen for this study as the problem of high infant mortality is invariably being experienced in the rural rather than urban areas.

          Geographically, Gujarat has been divided into three division viz. Saurashtra, North Gujarat and South Gujarat. These divisions include 7, 5 and 7 districts respectively. One district each was selected randomly, from each of these divisions, out of the total number of districts in the respective division. Three districts selected from Gujarat state for this study were: Bhavnagar, Sabarkantha and Surat.

          Similarly, Maharashtra state has been divided into four divisions viz. Bombay, Pune, Aurangabad and Nagpur having 6, 7, 6 and 8 districts respectively. Here again one district from each division was selected randomly. Four districts, thus, selected from Maharashtra were: Nasik, Pune, Aurangabad and Akola.

          Ten PHCs were selected randomly from each of the selected district. Thus, in all, 30 PHCs were selected from Gujarat and 40 PHCs from Maharashtra for the study. It was decided to select the PHC village itself or the nearby village under the jurisdiction of only one ANM/Female Health Worker covering thereby a  population of three to five thousand per PHC depending upon whether the area was tribal or non-tribal.

LEVEL OF IMR

          Infant mortality rate in Gujarat and Maharashtra, as obtained from the sample, were 67.3 and 45.9 per thousand live births during the period 1992-95 respectively. The neonatal and post-neonatal mortality in Gujarat was 45.0 and 22.3 per thousand live births while the corresponding values in Maharashtra were 32.7 and 13.2 respectively. This, in other words, means that while neonatal mortality rate in Maharashtra is about 27 percent lower than that in Gujarat, the post-neonatal mortality rate is still lower and is of the order of 41 percent. The  study further  reveals that while the  neonatal and post-neonatal mortality  as percentage of  infant  mortality in  Gujarat was 67 to 33; it was 71 to 29 in Maharashtra. The above analysis illustrate clearly that the decline in infant mortality in Maharashtra compared to Gujarat is largely due to the decline in its post-neonatal component of mortality. 

SOCIO-DEMOGRAPHIC AND MATERNAL CARE

 FACTORS AFFECTING IMR

 

The risk of infant death at the individual level is determined by considering the survival status of 2332 live births in Gujarat and 3270 live births in Maharashtra. For the present paper following eight variables viz. maternal age, age at effective marriage of wife, number of live births, interval between last two births, mother’s literacy status, prenatal care taken by mother, type of delivery (full term or premature) and birth weight of the baby are considered which may be having some impact on infant mortality in the respective area under study.

Maternal Age

          Table 1 reveals that the infants of teen aged mothers had higher risk of dying in both Gujarat as well as Maharashtra than the infants of mothers in the age bracket 20-29. It further reveals that births to mothers in the older age group, 30 and over, also had a higher risk of death in Gujarat but this pattern was not found to be true in the case of Maharashtra. Thus, the risk of infant death seems to be higher when the mother is either very young or relatively old (30 and over)  especially in Gujarat. In other words, the pattern of infant mortality follows a U-shaped curve with the age of mother. On the other hand, the infant mortality was high in the age group 15-19 of mothers in Maharashtra and went on declining with the increase in the age of mother in this state illustrating thereby that when the overall level of infant mortality is low the U-shaped pattern of IMR changes into the reverse J-shape pattern.

 

Age at Effective Marriage

 

          Age at effective marriage of the mother seems to be associated with the risk of infant mortality inversely i.e. lower the age at effective marriage, higher would be the risk of infant mortality. This relationship is found to be true in the case of both Gujarat and Maharashtra (Table 1). It is, however, quite likely that when multivariate analysis will be conducted subsequently, this variable may loose importance in favour of maternal age.

 

Birth Order

          The Table 1 clearly delineates that the first birth had a much higher risk of dying than 2-3 births especially in Gujarat. The risk of infant death among the births of order 2-3 was low both in Gujarat and Maharashtra. However, the risk of infant death again rises with the increase in parity. And this has been found to be true for 4 or higher order births in both Gujarat and Maharashtra. This variable, thus, seems to have an independent affect on the risk of infant mortality.  The pattern of infant mortality by order of birth seems to follow a U-shaped curve in the case of both Gujarat and Maharashtra. However, the percentage decline in the level of infant mortality from birth order 1 to 2-3 as well percentage increase in the level of infant mortality from birth order 2-3 to 4 and over was much sharper in the case of Gujarat than Maharashtra.

 

Birth Interval

          Besides birth order, shorter birth interval (below 18 months) between the two live births was also found to have been playing a significant role in the higher risk of dying among the infants as compared to the infants where the interval was higher than 18 months. This factor was found to be an important determinant of infant mortality in both Gujarat and Maharashtra as seen from the results presented in Table 1. Thus, unless a spacing of more than 18 months is maintained between the two pregnancies the  risk of infant death becomes high.

 

 


Table 1 :

Infant Mortality Rates Per Thousand Births by Socio-Demographic Factors in Gujarat and Maharashtra, 1992-95

 

Characteri-stics

Infant mortality rate

Number of births

Gujarat

Mahara-shtra

Total

Gujarat

Mahara-shtra

Total

Mother’s age

 

 

 

 

 

 

15-19

98.2

72.1

79.1

112

305

417

20-29

60.7

44.1

50.9

  1828

    2628

4456

30 & over

82.6

35.6

60.8

387

337

724

Age at Effective

marriage of mother

 (in years)

 

 

 

 

 

£ 17

72.2

52.7

57.6

693

2086

2779

18 & over

63.6

33.8

51.1

1634

1184

2818

Birth order

 

 

 

 

 

 

1

96.8

46.7

68.6

651

836

1487

2-3

40.5

37.7

38.8

1112

1566

2678

4 & above

86.2

62.2

71.8

545

820

1365

Interval between last two live births

(in months)

 

 

 

 

 

1-17

61.0

83.1

72.7

213

227

440

18 & over

54.2

41.4

46.5

1439

2176

3615

Literacy status of mother

 

 

 

 

 

Illiterate

71.4

54.5

62.7

1233

1303

2536

Literate

66.3

40.2

47.4

1094

1967

3061

 

 

Mother’s Education

 

          The Table 1 reveals that babies whose mothers were illiterate faced relatively higher risk of mortality than those whose mother were literate in Gujarat as well as Maharashtra. Literacy level of mother, thus, plays an important role in the survival status of infants. Impact of education on infant mortality, however, was found to be more marked in Maharashtra than Gujarat.

 

Prenatal Care

 

          It was observed in this study that babies whose mothers had taken prenatal care during pregnancy in terms of having 2 doses of Tetanus Toxoid had a relatively lower risk of death than those whose mothers either had 1 or no dose of TT (Table 2). And this relationship was found to hold good in both Gujarat as well as Maharashtra illustrating thereby that the prenatal care in terms of administration of two doses of TT is an important determinant of infant mortality.

 


Table 2:

Infant Mortality Rates Per Thousand Live Births by Maternal Care Factors in Gujarat and Maharashtra, 1992-95

 

Characteri-stics

Infant mortality rate

Number of births

Gujarat

Mahara-shtra

Total

Gujarat

Mahara-shtra

Total

Doses of Tetanus

toxoid taken

during pregnancy

 

 

 

 

 

0 or 1 dose

109.6

82.6

102.4

301

     109

410

2 doses

61.1

44.6

51.1

2028

3161

5189

Type of Delivery

 

 

 

 

 

Full term

53.0

32.9

41.3

2281

3187

5468

Premature

705.9

542.2

604.5

51

       83

134

Birth Weight of child

(in gms)

 

 

 

 

 

£ 2400

166.7

180.7

174.9

   168

238

406

> 2400

37.5

23.2

27.6

   666

1509

2175

 

Type of Delivery

          It is obvious from the Table 2 that premature deliveries in Gujarat and Maharashtra had a very high risk of infant death as compared to the full-term deliveries. This, however, is a biological factor and may be difficult to control unless proper medical care is sought by the pregnant mothers during the course of their pregnancies.

Birth Weight

          Birth weight of the babies is determined to a large extent on the nutritional intake of the mothers. It was observed from the Table 2 that among those babies who had a birth weight of less than 2400 grams in Gujarat as well as Maharashtra,  had a much higher risk of death than the babies having birth weight to be more than 2400 grams . Birth weight of the babies at birth is, thus, an excellent indicator of infant death.

          Based on the preliminary analyses of the data conducted in this study, it appears that all the above factors play an important role in determining the level of infant mortality.  The demographic factors like early age at marriage, age at first pregnancy can influence the risk of death in early infancy through their effects on the prenatal and natal causes such as prematurity, birth injuries, genetic disorders, and congenital malformations. On the other hand, environmental conditions and personal hygiene may be associated with infective and respiratory diseases. Socio-economic factors such as mother’s education, living habits and economic conditions might be influencing the survival status of the infant through medical and non-medical care during the neonatal and post-neonatal period. Similarly, nutrient availability factors such as the food habits of the mother and her nutrition level during the prenatal period might be responsible for premature births resulting into infant death. Further, malnutrition of mothers and infant feeding practices, for example breastfeeding for at least the first few months and the time of introduction of supplementary food, affect the chances of infant survival by reducing endocrine, nutritional and metabolic diseases. Finally, medical care (administration of anti-tetanus vaccine) starting from conception to the first birthday is an important determinant of infant mortality.

LOGISTIC MODEL TO DETERMINE FACTORS

ASSOCIATED WITH INFANT MORTALITY

          In the earlier section importance of each category of individual variable (covariate) on survival status of an infant was ascertained. Now, an effort is being made to see the impact of each category of a covariate in the presence of other covariates, on the survival status of the infant in Gujarat and Maharashtra separately as well as collectively with the help of Logistic Model. The Model used to test the covariates is :

                                   n

          log (P/1-P) = a +  S bixi

                                                i=1

 

            Where xi = (1, 2, .... n) is the covariate of the dependent variable

 

          P = Probability of a birth ending into an infant death

 

Best models in the case of Gujarat and Maharashtra were obtained separately as well as collectively  for the two states together first by excluding the dependent variable birth weight of the  infant and then by including this variable in the model. The impact of each dependent variable, when controlled by other dependent variables, on infant mortality was ascertained and the results are presented in the Tables 3 and 4 respectively.

          Table 3 presents the coefficients and odds ratios of the covariates, excluding the covariate birth weight, which were found to be significant or otherwise for infant mortality. The table clearly indicates that the type of delivery (full-term or premature) has been the single most important covariate of infant mortality in both Gujarat as well as Maharashtra followed by birth order. Literacy was another significant variable affecting infant mortality in Gujarat but this variable was not found to be important in Maharashtra. On the other hand, the variables ‘age’ and the ‘doses of tetanus toxoid taken during pregnancy’ were also found to be important predictor of infant mortality in Maharashtra but not in Gujarat.

 

 

 

 

 

Table 3 :

Estimated Logit Coefficients and the Odds Ratios for the Infant Mortality in Gujarat & Maharashtra, 1992-95

 

Covariate

Logit coefficient

Odds ratio

 

Gujarat

Mahara-shtra

Combined

Gujarat

Mahara- shtra

Combined

 

 

 

 

 

 

 

Intercept

 1.51

 1.27

 1.46

 

 

 

 

 

 

 

 

 

 

Mother’s Education

 

 

 

 

 

 

·      Illiterate

 0.00

 0.00

 0.00

1.00

1.00

1.00

·      Literate

 0.43*

-0.17

-0.34**

0.65

0.84

0.71

 

 

 

 

 

 

 

Mother’s age

(in yrs.)

 

 

 

 

 

 

·      15-19

 0.00

 0.00

 0.00

1.00

1.00

1.00

·      20-29

 0.26

-0.47

-0.14

1.30

0.62

0.87

·      30 & over

 0.52

-1.14*

-0.22

1.68

0.32

0.80

 

 

 

 

 

 

 

Age at effective marriage (in yrs.)

 

 

 

 

 

 

·      £ 17

 0.24

 0.27

 0.06

1.27

1.31

1.06

·      18 & over

 0.00

 0.00

 0.00

1.00

1.00

1.00

 

 

 

 

 

 

 

Birth order

 

 

 

 

 

 

·      1

 0.00

 0.00

 0.00

1.00

1.00

1.00

·      2-3

-0.94**

-0.06

-0.53**

0.39

0.94

0.59

·      4 & above

-0.22

 0.64*

 0.17

0.80

1.90

1.19

 

 

 

 

 

 

 

Type of delivery

 

 

 

 

 

 

·      Full term

 0.00

 0.00

 0.00

1.00

1.00

1.00

·      Premature

-3.91**

-3.64**

-3.64**

0.20

0.03

0.26

 

 

 

 

 

 

 

Doses of Tetanus Toxoid taken during pregnancy

 

 

 

 

 

 

·      0 or 1 dose

-0.31

-0.86*

-0.59**

0.73

0.42

0.55

·      2 doses

0.00

0.00

 0.00

1.00

1.00

1.00

 

*   Significant at 0.05 level

** Highly significant at 0.01 level

 

         

Among the variables which were significantly associated with infant mortality, odds ratios were also calculated and are depicted in this table. The Gujarat data reveals that the infants of literate mothers had about 35 percent less chance of death compared to the infants of illiterate mothers. Similarly, the chance of infant deaths to the births that occurred to the mothers of first birth order was estimated to be about 61 percent higher compared to the chance of infant deaths to the births that occurred to the mothers of second or third births order. Further, premature births in Gujarat had 80 percent higher chance of infant death compared to full-term babies. However, this chance of infant death in Maharashtra was even higher (97 percent) when the birth was premature as compared to full-term babies. 

          Maharashtra data also indicates that the chances of infant death among the births to the mothers aged 30 and over was about 68 percent lower than that of the infants born to the mothers of the younger age (15-19). It also reveals that the births to the mothers, who took 0 or 1 dose of tetanus toxoid during pregnancy, had about 58 percent higher chance of death during infancy compared to the births to the mothers who had two doses of tetanus toxoid during their pregnancy. Further, the risk of death among 4 and higher order births was about 90 percent higher than that of first order birth in Maharashtra.

 

After having studied the impact of selected dependent variables, in the absence of birth weight of the infant, on infant mortality an attempt is now made to see the behaviour of these variables on infant mortality in the presence of the variable ‘weight of the child at birth’ in the logistic model and the results are presented in the Table 4.

 

          It is very obvious from this table that ‘Weight of the child at birth’ and ‘Type of Delivery’ are the two most important variables associated with infant mortality in both Gujarat as well as Maharashtra. The chances of a birth ending into an infant death becomes 67 and 80 percent higher in Gujarat and Maharashtra respectively when the birth weight of the babies is less than 2400 grams compared to those where the birth weight of the children is more than 2400 grams. Similarly, the chances of premature births ending into infant deaths are 94 and 95 percent higher in comparison to the full term births in Gujarat and Maharashtra respectively. The variable mother’s education, which was found to be an important determinant of infant mortality in Gujarat when the variable ‘weight of the child at birth’ was excluded lost its importance when the later was included in the model illustrating thereby that mother’s education was weakly associated with infant mortality in Gujarat. Further, this variable was not found to be an important determinant of infant mortality in Maharashtra. It thus, appears that the variable “mother’s education” remain an important determinant of infant mortality only upto a certain threshold level of infant mortality but once the infant mortality in a certain area goes down the threshold level the importance of this variable also vanishes.

 

          The birth order continues to be an important determinant of infant mortality for both Gujarat as well as Maharashtra even after the introduction of the dependent variable ‘weight of child at both’ in the model. The table reveals that second and third order births are exposed to 77 percent less infant mortality than first order births in Gujarat. On the other hand, 4 and higher order births in Maharashtra had 3.7 times higher chance of infant death compared to first order births. This means that first order and/or higher order births (4 and above) have higher chance of ending into infant deaths compared to intervening second or third order births.

 

Table 4:

Estimated Logit Coefficients and the Odds Ratios for the Infant Mortality,   when the Dependent Variable Weight was Introduced, in Gujarat & Maharashtra, 1992-95

 

Covariate

Logit coefficient

Odds ratio

 

Gujarat

Mahara-shtra

Combined

Gujarat

Mahara- shtra

Combined

 

 

 

 

 

 

 

Intercept

 1.11

 1.11

 1.27

 

 

 

 

 

 

 

 

 

 

Mother’s Education

 

 

 

 

 

 

·      Illiterate

 0.00

 0.00

 0.00

1.00

1.00

1.00

·      Literate

-0.44

 0.01

-0.20

0.64

1.01

0.82

Mother’s age (in yrs.)

 

 

 

 

 

 

·      15-19

 0.00

 0.00

 0.00

1.00

1.00

1.00

·      20-29

 0.13

-0.86++

-0.47

1.14

0.42

0.62

·      30 & over

 0.35

-1.45*

-0.74

1.42

0.23

0.48

Age at effective marriage (in yrs.)

 

 

 

 

 

 

·      £ 17

 0.11

 0.25

 0.03

1.12

1.28

1.03

·      18 & over

 0.00

 0.00

 0.00

1.00

1.00

1.00

Birth order

 

 

 

 

 

 

·      1

 0.00

 0.00

 0.00

1.00

1.00

1.00

·      2-3

-1.47**

 0.36

-0.47+

0.23

1.43

0.62

·      4 & above

-0.21

 1.32**

 0.57+

0.81

3.74

0.57

Type of delivery

 

 

 

 

 

 

·      Full term

 0.00

 0.00

 0.00

1.00

1.00

1.00

·      Premature

-2.76**

-3.02**

-2.87**

0.06

0.05

0.06

Doses of Tetanus Toxoid taken during pregnancy

 

 

 

 

 

 

·      0 or 1 dose

 0.05

-0.28

-0.06

1.05

0.76

0.94

·      2 doses

 0.00

 0.00

 0.00

1.00

1.00

1.00

Weight of child at birth

 

 

 

 

 

 

·      £ 2400 gms

-1.10**

-1.63**

-1.41**

0.33

0.20

0.24

·      More than 2400 gms

 0.00

 0.00

 0.00

1.00

1.00

1.00

*   Significant at 0.05 level                                   + Significant at 0.08 level

** Highly significant at 0.01 level                       ++ Significant at 0.06 level

 

 

 

          Another important covariate of infant mortality in Maharashtra was found to be the age of mother. Births to the younger mothers (15-19) had relatively higher risk of infant deaths compared to the births to older mothers (20 & over). The risks of infant death to the births to mothers in the age group 20-29 and 30 & over was found to be 58 and 77 percent lower respectively compared to the corresponding risk to the births of 15-19 age mothers.

 

          The above analyses clearly illustrate that birth weight of the child, type of delivery and the demographic factors birth order and age of mother at delivery are some of the important determinants of infant mortality.

 

CAUSES OF INFANT DEATHS - PROSPECTIVE DATA

 

          As mentioned earlier, cause of death data was collected on a prospective basis by paramedical personnel - Auxiliary Nurse Midwife (ANM) and  Lady Health Visitor (LHV) - stationed at the respective Primary Health Centres (PHCs) in seven selected districts (3 in Gujarat and 4 in Maharashtra). The selected ANMs were either residents of the PHC area or were familiar with the local community. The paramedical personnel were trained in the methodology of identification of causes of infant death. The causes of infant deaths were determined from the information about the symptoms, conditions, anatomical site and duration of diseases. These causes, obtained by paramedical personnel, were later confirmed by the medical officers of each of the Primary Health Centres and the requisite data were classified as per the WHO classification (1978).

          The percentage distribution of infant deaths by immediate cause of death is shown in Table 5. The data reveal that about 42 percent of deaths in Gujarat and 49 percent of deaths in Maharashtra were due to causes classified as ‘Certain conditions originating the perinatal period’. Prematurity under this group was the major underlying cause both in Gujarat (25 percent) and Maharashtra (33 percent) followed by Hypoxia and Asphyxia (10 percent each in Gujarat and Maharashtra). Respiratory infection was found to be another major underlying cause in Maharashtra (4.0 percent)  but not that much in Gujarat (1.3 percent). However, Birth Trauma was more evident in Gujarat (3.2 percent) but almost negligible (0.7 percent) in Maharashtra.

          Prematurity is a function of malnutrition, prenatal care, age, parity and interval between two births. Malnutrition of mothers, especially during the gestation period, may be one of the major direct causes of low birth weight babies, who may either develop rickets and marasmus or else may die in infancy. It has been seen in the earlier discussion of determinants of infant mortality that the risk of infant mortality is especially high among babies with a birth weight of less than 2400 grams and that the babies with more than 2400 grams of weight have higher chances of survival.

          Diseases of Respiratory System were the next major cause of infant deaths with 23 and 22 percent of the total infant deaths in Gujarat and Maharashtra respectively occurring due to this main cause. The major underlying causes under this main cause has been Broncho-Pneumonia/ Pneumonities (12 percent in Maharashtra and 4 percent in Gujarat), Pneumonia unspecified (10 percent in Gujarat and 5 percent in Maharashtra), and upper respiratory infection (4 percent each in Gujarat and Maharashtra).

          The third main cause of infant deaths was reported to be Infective and Parasitic Diseases, which killed 10 percent infants in Gujarat and 6 percent babies in Maharashtra. Here, the main underlying cause was Septicemia which took a toll of 4 percent of deaths in Gujarat and 2 percent in Maharashtra. Besides the above sub-cause, the diarrheal diseases like Gastroenteritis (2 percent in Maharashtra and 1 percent in Gujarat) and Broncho-pneumonia were

 

 

Table 5:  Percent Distribution of Infant Deaths by Specific Cause of Death, 1992-95

 

 

Specific Causes of Infant Death

Percentage Distribution of Infant Death

 

 

 

Gujarat

Maharashtra

 

1

2

3

4

 

 

 

 

 

 

 

1.

Infective & Paraistic Diseases

9.6

 

6.0

 

 

Typhoid fever

0.6

 

 -

 

 

Gastroenteritis

0.6

 

2.0

 

 

Diptheria

0.6

 

 -

 

 

Erysipelas

0.6

 

 -

 

 

Tetanus

2.0

 

 -

 

 

Septicemia

4.0

 

2.0

 

 

Broncho-pneumonia

0.6

 

2.0

 

 

Malaria

0.6

 

 -

 

2.

Endocrinal, Nutritional and Metabolic diseases and immunity disorders

8.9

 

5.3

 

 

Malnutrition

5.1

 

2.0

 

 

Dehydration

3.8

 

3.3

 

3.

Diseases of Blood and Blood forming organs

3.8

 

 -

 

 

Anaemia

1.9

 

 -

 

 

Blood Dyscrasia

1.3

 

 -

 

 

Thallesemig

0.6

 

 -

 

4.

Diseases of Nervous System and Sense Organs

0.6

 

4.7

 

 

Encephalitis

0.6

 

 -

 

 

Meningitis

 -

 

3.3

 

 

Meningocele

 -

 

0.7

 

 

Nervous Degenerative disorder

 -

 

0.7

 

5.

Diseases of Respiratory System

22.9

 

22.0

 

 

Upper respiratory infection

3.8

 

4.0

 

 

Broncho-pheumonia/ Pneumonitis

3.8

 

12.0

 

 

Pneumonia unspecified

12.1

 

5.3

 

 

Bronchitis

1.3

 

 -

 

 

Bronchiectasis

1.3

 

 -

 

 

Other diseases of lungs

0.6

 

0.7

 

6.

Diseases of Digestive System

0.6

 

0.7

 

 

Intenstinal obstruction

0.6

 

 -

 

 

Pylonic sterosis

 -

 

0.7

 

7.

Diseases of the Skin & Subcutaneous Tissues

0.6

 

 -

 

 

Diseases of skin

0.6

 

 -

 

 

 

 

 

1

2

3

4

 

8.

Congenital Anomaties

10.9

 

6.0

 

 

Spina Bifida

1.3

 

0.7

 

 

Hydrocephalus

1.3

 

 -

 

 

Congenital Anomalies of Heart

5.1

 

2.7

 

 

Congenital Anomalies of Genital organ

1.3

 

0.7

 

 

Other unspecified Congenital Anomalies (Anencephalus)

1.9

 

0.7

 

 

Meningogitis

 -

 

1.3

 

9.

Certain Conditions Originating in the Perinatal Period

42.0

 

48.7

 

 

Twin Delivery

 -

 

0.7

 

 

Prematurity

24.8

 

32.6

 

 

Birth Trauma

3.2

 

0.7

 

 

Hypoxia & Asphyxia

9.6

 

6.6

 

 

Respiratory Infection

1.3

 

4.0

 

 

Neonatal Tetanus

0.6

 

 -

 

 

Hemolytic Disease (RH Incompatibility)

1.3

 

0.7

 

 

Convulsion

0.6

 

0.7

 

 

Other & ill defined conditions originating in the perinatal period

0.6

 

1.3

 

 

Breach Presentation

 -

 

0.7

 

 

Neonatal Haemorrhage

 -

 

0.7

 

10.

Symptons, Signs & Ill defined conditions

 -

 

4.7

 

 

Jaundice

 -

 

2.0

 

 

Other unknown & unspecified causes

 -

 

2.7

 

11.

Injury & Poisoning

 -

 

2.0

 

 

Cot death

 -

 

0.7

 

 

Injury to chest

 -

 

0.7

 

 

Injury to head

 -

 

0.7

 

 

 

 

 

 

 

 

Total cases

157

 

150

 

 

also the killer of 2 percent of infants in Maharashtra and 1 percent in Gujarat. Interestingly enough only 2 percent of infants died due to Tetanus in Gujarat and none in Maharashtra. Deaths due to Tetanus are thus getting controlled in Maharashtra but there is still some problem of controlling the same in Gujarat.

          The fourth major cause of death in both the states was Congenital Anomalies which was responsible for 11 percent of deaths in Gujarat and 6 percent deaths in Maharashtra. Major underlying cause of death under this broad category was found to be Congenital Anomalies of Heart (5 percent deaths in Gujarat and 3 percent in Maharashtra).

          The fifth major cause of infant death was reported to be Endocrinal, Nutritional and Metabolic Diseases and Immunity Disorders. The important underlying causes of death under this category have been Malnutrition (which took a toll of 5 percent deaths in Gujarat and 2 percent death in Maharashtra) and Dehydration (causing death to 4 percent of infant in Gujarat and 3 percent of babies in Maharashtra).

SUMMARY & CONCLUSIONS

 

                Overall infant mortality in Gujarat and Maharashtra, during 1992-95, was estimated to be 67 and 46 per 1000 live births respectively. The study reveals that the infant mortality follows a U-shaped pattern when the IMR is high, as in the case of Gujarat, but this pattern gets changed to reverse J-shape when the infant mortality is low, as in the case of Maharashtra. However, U-shape pattern of infant mortality by birth order continues irrespective of high (Gujarat) or low (Maharashtra) level of infant mortality in the two states under study.

          Infant mortality was found to be high when (i) the age at effective marriage of mother was below 18 years; (ii) interval between last two live births was below 18 months; and (iii) when the mother was illiterate. Further, infant mortality was found to be high when the mother had either not taken any dose or had taken only 1 dose of Tetanus Toxoid as compared to those who had taken two or more doses of TT. It was also high when the baby was premature or the birth weight of the baby was less than 2400 grams. Multinomial Logit model, however, illustrates that weight of the baby at birth, type of delivery (premature or full term) and birth order are the only three covariates which have significant impact on infant mortality and other variables like mother’s age, morther’s education, age at effective marriage and doses of tetanus toxoid taken during pregnancy loses their importance in the presence of three covariates mentioned above.

          The study also obtained data on causes of infant deaths by using prospective data. The major cause of infant deaths was reported to be ‘certain conditions originating the perinatal period’ (42 percent in Gujarat and 49 percent in Maharashtra). Prematurity under this group was the major underlying cause both in Gujarat (25 percent) and Maharashtra (33 percent) followed by Hypoxia and Asphyxia (10 percent each in Gujarat and Maharashtra). Diseases of respiratory system especially brocho-pneumonia or pneumonia are the other major killer of infants. The third main cause of infant deaths was reported to be infective and parasitic diseases especially septicemia and gastroenteritis. Interestingly enough only 2 percent of infant deaths in Gujarat and none in Maharashtra took place due to tetanus. Deaths due to tetanus are thus getting controlled in this part of the country.

 

 

 

ACKNOWLEDGEMENT

          This paper is derived out of the data from the project `Determinants of Infant Mortality in Gujarat & Maharashtra’ from the financial support of IDRC, Canada. We wish to put on record our deep sense of gratitude to the IDRC, Canada for their support.

 

 

REFERENCES

 

Bourgeois-Pichat, J. 1964. Evolution Re¢cente de la Mortalite¢ Infantile. Cited in United Nations, 1973. The Determinants and Consequences of Population Trends, Vol.1, New York, United Nations, Department of Economics and Social Affairs, p. 126.

Gandotra, M.M., and Das, Nayaran, 1988. Infant Mortality: An Analysis of Recent Births. In: Infant Mortality in India – Differentials and Determinants; Edited by Anrudh K. Jain and Pravin Visaria, Sage Publications, New Delhi, 1988.

Govt. of India, 1988-89. Year Book, Family Welfare Programme in India, Ministry of Health and Family Welfare, Department of Family Welfare, New Delhi.

Mosley, W.H., and Chen, L.C., 1984. An Analytical Framework for the Study of Child Survival in Developing Countries. In: W.H. Mosely and Lincoln C. Chen (eds.), Child Survival: Strategies for Research, A supplement to Population and Development Review, 10:24-45, New York, The Population Council.

Pandey, Arvind; Choe, Minja Kim; Luther Norman Y.; Sahu, Damoder; and Jagdish Chand, 1998. Infant and Child Mortality in India, National Family Health Survey Subject Reports Number 11, IIPS Mumbai and East-West Centre Program on Population, Honolulu, Hawaii, U.S.A.

PRC, Baroda and IIPS, Mumbai. 1995. National Family Health Survey, 1993, Gujarat.

PRC, Pune and IIPS, Mumbai.1994. National Family Health Survey, 1992-93, Maharashtra.

World Health Organization, 1978. International Classification of Diseases - Manual of the International Statistical Classification of Diseases, Injuries and Causes of Death, Geneva.

 



* Additional Director, Population Research Centre, Baroda, India

+ Joint Director, Population Centre, Baroda, India