International Journal of Statistics and Applications

p-ISSN: 2168-5193    e-ISSN: 2168-5215

2018;  8(1): 1-8

doi:10.5923/j.statistics.20180801.01

 

Levels and Trends of Child Mortality in Zaria Local Government Area of Kaduna State, Nigeria

Tukur Dahiru1, Alhaji A. Aliyu1, Moses A. Oyefabi2, Suleiman H. Idris1, Ahmad A. Umar1

1Department of Community Medicine, Ahmadu Bello University, Zaria, Nigeria

2Department of Community Medicine, Kaduna State University, Kaduna, Nigeria

Correspondence to: Tukur Dahiru, Department of Community Medicine, Ahmadu Bello University, Zaria, Nigeria.

Email:

Copyright © 2018 Scientific & Academic Publishing. All Rights Reserved.

This work is licensed under the Creative Commons Attribution International License (CC BY).
http://creativecommons.org/licenses/by/4.0/

Abstract

Information on child mortality at sub-national levels is scanty and where available, is fragmentary. The absence of universal, reliable and accurate vital registration system in the country generating sub-national level data on child mortality presents a challenge. This study is an attempt at responding to poor data on child mortality at sub-national level. Demographic inquiry was conducted in six political wards of Zaria between March and December 2014. Relevant demographic information was collected from women of reproductive age group using a questionnaire and indirect technique of estimating child mortality was applied to the data. The Palloni-Heligman version of the original Brass method was applied to estimate the level of child mortality. The estimated under-five and infant mortality rates of 186 and 122 per 1000 live births were recorded. Generally, infant mortality rate varied between 103 and 134 per 1000 live births while under-five mortality showed more stable estimates of between 170 and 176 per 1000 live births depending on the model life table in both situations. Child mortality estimates showed a gradual downward decline over past decades. Child mortality estimates in these communities are high even though there has been steady decline over the past years. Child mortality is still a public health challenge and more policies and programmes are urgently needed to address them as we to enter sustainable development goals era.

Keywords: Child, Mortality, Residence, Zaria, North, Nigeria

Cite this paper: Tukur Dahiru, Alhaji A. Aliyu, Moses A. Oyefabi, Suleiman H. Idris, Ahmad A. Umar, Levels and Trends of Child Mortality in Zaria Local Government Area of Kaduna State, Nigeria, International Journal of Statistics and Applications, Vol. 8 No. 1, 2018, pp. 1-8. doi: 10.5923/j.statistics.20180801.01.

1. Introduction

Worldwide, child mortality has declined by more than half from 12.7 million in 1990 to 5.9 million in 2015; with an annual average of 16, 000 deaths. The under-five mortality rate declined from 91 deaths per 1000 live births to 43 deaths per 1000 live births in the same period [1]. Similar rate of decline has been observed in sub-Saharan African region with under-five mortality rate decreasing from 180 deaths per 1000 live births to 83 deaths per 1000 live births [1].
In Nigeria, where child mortality remains a major public health challenge, such reduction has not been observed. The recent 2013 Nigeria Demographic and Health Survey (NDHS) indicates that the highest reduction in child mortality was seen among under-five rate falling 78% from 201 deaths per 1000 live births in 1999-2003 period to 128 deaths per 1000 live births in 2008-2013 period while the least reduction was among neonatal mortality rate falling from 48 deaths per 1000 live births to 37 deaths per 1000 live births [2]. Furthermore, while global child deaths burdens are on the decline, Nigeria’s contribution to the global burden is on the increase. Nigeria’s contribution to the global numbers of child deaths nearly doubled from around 6.7% in 1990 to 12.6% in 2015 [1]. This reversal of trend in child mortality corresponds to the period of global and national renewed efforts to address the issues of maternal and child health such as World Summit for Children in 1990 and MDGs [3, 4].
Information about child mortality at sub-national levels in Nigeria is scanty and fragmentary; in the absence of reliable, accurate and complete vital registration system it will be difficult to close the gap. While information on child mortality rates has been generated at national and in some instances at zonal levels by the DHS programme, there is need to generate such information beyond zonal level. In the past three decades a few researchers have generated child mortality rates at sub-national levels or community levels besides those generated by the DHS programme [5-9]. However, majority of these have also emphasized heavily on the determinants of child mortality except a few [8, 9]. Review of the available documented literature revealed only three earlier reports on child mortality were from northern Nigeria [9-11]. These estimates are more than three decades old and the need exist to generate recent estimates. This paucity of data on the levels and trends on child mortality as well their fragmentary nature at community level in Northern Nigeria prompted this investigation to document levels and trend in child mortality in the region.

2. Materials and Methods

Background to the study area
Zaria is a major city in Kaduna State in Northwest geopolitical zone of Nigeria. This metropolitan city consists of two local government areas (LGAs), Zaria and Sabon Gari. Popularly and traditionally called Zazzau, it was one of the original seven Hausa city-states. According to the 2006 Census its population was 408,198 [12]. Zaria LGA is the study location bounded by Sabon Gari in the west, Soba LGA in the east and Igabi LGA in the south. The administrative headquarter is Zaria city and it has thirteen political wards seven in the northern and six in the southern division of the LGA namely Kufena, Tukur Tukur, Tudun Wada, Gyellesu, Dambo, Angwan Fatika and Kwarabai B in the north, and Angwan Juma, Angwan Limancin Kona, Kaura and Kwarbai A, Dutsen Abba and Wucicciri in the southern division of the LGA. There are thirty-six PHCs (primary health care centers) and four health clinics in the LGA. This study was conducted in five out of these wards with their estimated populations given: Angwa Juma (35,915), Angwan Limanci Kona (52,441), Dambo (41021), Kwarbai A (38,021) and Angwan Fatika (3307) wards. Each of these wards has a primary health center (PHC) run by the staff of Health Department of Zaria LGA. Data was collected in the five wards in March while data was collected in the last (Dutsen Abba) ward in December 2014 as part of the community health diagnosis of final year medical students.
Methodology
Five of the six wards were randomly selected while Dutsen Abba was purposively selected as a community for community health diagnosis posting for the final year medical students of Ahmadu Bello University, Zaria. An interviewer-administered questionnaire was used to collect relevant information for the purpose of estimating child mortality rates: age of respondent, children ever born, children still alive and children dead as well as other socio-demographic information about the respondent/ household: educational attainment, employment, marital status etc. The data required for this method are: (i) the number of children ever born, classified by five-year age group of mother; (ii) the number of children surviving (or the number dead) classified by five-year age group of mother; (iii) the total number of women (irrespective of marital status) classified by five-year age group. Since the purpose of this study was not to generalize the findings to the whole of Zaria LGA, sample size was not calculated; the aim was to estimate levels of child mortality cognizance of the fact that estimates available are over three decades old.
Statistical analysis
The analysis is based primarily on one of the two versions of the original Brass’s indirect technique for indirect estimation of child mortality rates as described in the United Nations Guide [13]. The Trussell version [14] utilized the Trussell multipliers and adopted the Coale-Demeny model life tables (North, East, West and South) [15]. The second version is that developed by Palloni and Heligman [16], it uses the Palloni-Heligman multipliers but also requires mean age of child bearing among the women. Both these methods are based on the principle of converting the proportion of dead children among children ever born into probabilities of dying between birth and exact ages. The Palloni-Heligman version utilizes the UN model life tables (Latin America, Chilean, South Asia, Far East and General). The estimation procedure as proposed by Brass is of the form:
(1)
where q(x) is the probability of dying between birth and exact age x, D(i) is the proportion dead of children ever born to women in age group i and K(i) is a multiplier that converts D(i) into estimate of q(x). The proportion of children dead is calculated from the ratio of children dead (D) and children ever born (CEB):
(2)
Average parity is derived in a similar manner, which is the ratio of children ever born to that of women in a given five-year age group:
(3)
where P(i) is the average parity, and FP total number of women in a given five-year age group.
The multipliers, K(i) are calculated from the Trussell regression equation:
(4)
In this study we chose the Palloni-Heligman version and therefore the coefficients a(i), b(i) and c(i) are tabulated values from the Palloni-Heligman regression equation. In this exercise, we chose the General model life table of the United Nations since we believe it best suits Zaria mortality pattern [15]. In the case of Palloni-Heligman version and additional age at maternity or mean age at child bearing (MACB) was added to equation (4) above such that:
(5)
Where M is the mean age child bearing computed as follows:
(6)
In equation (6), B(i) is number of births to women in age group i and mp(i) is the mid-point in years of age group i. Time-dating of child mortality rates were estimated using the method developed by Brass but adopting separate coefficients for Palloni-Heligman equations [16]. The time-reference to which the estimates refer to are derived from, again Brass’s equation using multipliers for Trussell and Palloni-Heligman separately:
(7)
Where t(i) refers to time (in years) to which the estimate refer to before the survey, e(i), f(i) and g(i) are the coefficients from the tabulated values. Once the values of t(i) are obtained, they were converted into actual dates by subtracting them from the reference date of survey expressed in decimal terms. Finally, common indices of infant and child mortality and , were derived from the estimates of by first converting into a value of or the level of parameter of relational logit model life table. The calculated is then used to estimate the corresponding probability of dying between birth and exact age 5, . Each value of is converted into its logit by means of the relationship [17]:
(8)
From this relationship, is calculated for each 5q0 using the following relationship:
(9)
The are the estimated values obtained above and are the logit transformations of the standard life table selected, West Model Life Table at life expectancy of 60years. Series of are then generated and for each, are calculated based on the relationship:
(10)
In this paper common indices for both and are calculated. Finally, Mortpak suite of programme was also used to generate child mortality estimates for all the families of model life tables used here; the output is here attached as Appendix I [18].
Assumptions
This indirect method is based on the following assumptions: 1) the age reporting by mothers does not show any digit preference or heaping; 2) the risk of child’s death is independent of age of mother or any other associated factors such as socio-economic status; 3) information on children ever born and those surviving is accurate; 4) fertility and mortality patterns as well as levels remain constant; 5) the age pattern of mortality of the population is similar to the selected model life table. For the Palloni-Heligman version, MACB was assumed to be 29.80 years from the 2015 World Population Prospects [19].

3. Results

Table 1 shows information on the number of women interviewed, number of children ever born by these women as well as the number of children dead from those given birth by the cohort of women. Table 2 shows the results of the estimated child mortality using General model of the United Nations Model Life Tables (Palloni-Heligman version). It shows the average parity, the proportion of children dead among those ever born and the multiplying factors. It also shows the probability of dying between birth and exact age 1, 2, 3, 5, 10, 15 and 20; as well as the reference date to which these estimates refer. Importantly are the common indices of child mortality: and that where estimated to determine their trends over the previous 20years. For the six communities studied here, the corresponding infant and under-five mortality rates are 122 and 186 per 1000 live births respectively. Figures 1 and 2 show the trends in and by age group of women and reference dates for the six communities, both sexes combined. The results indicate downward trends in the levels of both and ; these probabilities of child death refer to period between June 2013 and September 2001. A dip in both and estimates corresponding to 2007 coinciding with women in the age group 35-39years in both versions are observed. This dip could possibly be due to under-reporting error by these women as has been reported elsewhere among older women as described in Manual X [20].
Table 1. Children ever born, children dead and age of mother Zaria, December 2014
Table 2. Estimates of infant and child mortality (both sexes) using indirect technique (Palloni-Heligman version), in six communities of Zaria, December 2014
Figure 1. Trends in child mortality rates for both sexes by age group of women in six Zaria communities (Palloni-Heligman version), December 2014
Figure 2. Trends in child mortality rates for both sexes by reference date in six Zaria communities (Palloni-Heligman version), December 2014

4. Discussion

This field survey was conducted to determine the levels and trends of child mortality in six communities in Zaria, northern Nigeria by utilizing the indirect method developed initially by Brass and further refined by Trussell and Palloni-Heligman. The DHS programme which started operation in Nigeria in mid-1980s has provided child mortality rate estimates at national levels but sub-national estimates has been very scanty. According to the recent report, the infant mortality and under-five mortality rates are 69 and 128 per 1000 respectively for the whole country; for the North West, North East and North Central zones are 89 and 185; 77 and 160; and 66 and 100 per 1000 live births [2]. The infant mortality rate estimated from our study is substantially higher than the regional and national average while under-five mortality rate virtually same with the regional average but much higher than the national average. Bradley has reported an infant and under-five mortality rates of 169 and 321 per 1000 from around two villages in Malumfashi area which is situated about 100km to the north of Zaria. However, Bradley’s estimates refer to period around 1980 (35years ago). Considering the ecological, geopolitical and cultural similarities between Malumfashi and Zaria area it may be safe to infer that the infant and under-five mortality rates reported in our study is plausible considering that child mortality has declined both at national and global level; and that there has been a similar decline in child mortality over the same period. This shows that the rates of child mortality have gradually declined to current level in Zaria area over the past 35years if we use the Malumfashi mortality rates as our reference point [10]. Furthermore, child mortality rates were reported during a field posting of final year medical students of Ahmadu Bello University, Zaria during the period April/May 1985 in Yakawada area of present Giwa LGA of Kaduna State; Yakawada is around 40km to the North-West of Zaria along the major road to Sokoto. The infant and under-five mortality rates are 109 and 99 per 1000 respectively [11]. However, the Yakawada estimates indicates reversal between infant mortality and under-five rate; generally infant mortality rates are lower than under-five mortality rate except there are some events leading to excessive loss of life before the age of one year, notably measles epidemic. Udjo, in 1987 reported child mortality rates in some four villages surrounding Maiduguri area including some two wards in Maiduguri town, North Eastern Nigeria. The reported infant mortality rates ranged from 71 per 1000 in Maiduguri to an average of 123 per 1000 in the four villages around Maiduguri area. The under-five mortality rate ranged also from 132 per 1000 in Maiduguri to an average of 219 per 1000 in the four villages surveyed. Using the MDG targets for Nigeria from 213 per 1000 live births in 1990 it is reasonable that the estimates derived here are plausible; the under-five mortality rate was targeted to decline to 71 per 1000 live births by end of 2015 [9].
The fragmentary nature of these estimates from different parts of Northern Nigeria makes it difficult to make any reasonable conclusion on the levels and trends. However, from available literature, especially the serial Nigeria DHS reports it is possible to make a professional guestimate on the levels and trends of child mortality. Different components of child mortality rates have declined at varying rates; from 7.3 in under-five mortality rate to 3.1 in infant mortality rate annually. It is possible that child mortality rate has also declined in the past three decades or more drawing from decline at the national level.
The inability to make reasonable conclusion emanating from lack of data at community or sub-national level calls for the need to undertake a nation-wide vital registration system and systematic and organized surveys at sub-national levels to assess levels and trends in child mortality and other health indicators. This exercise will close the gap in lack of estimates of child mortality in the country and it will help in calibrating progress in reducing child mortality especially with the new SDGs [21]. Furthermore, the generation of child mortality rates estimates will serve as benchmark for both governmental and non-governmental agencies working in the area of improving child health [22].

5. Conclusions

Using one of the several indirect techniques of estimating child mortality, we estimated child mortality in some rural settlement in Zaria, northern Nigeria. Despite progress made in reducing mortality globally, estimates of child mortality from these communities indicate that child mortality is still high. Therefore, there is urgent need to accelerate reduction in child mortality especially in the light of achieving SDG 3.

Limitations of the Study

Estimating child mortality using the method applied here depends on the extent to which those assumptions have been met. Satisfying these assumptions is difficult in these communities where accurate reporting on age of the woman and children ever born is difficult. Fertility and mortality remaining constant or at least declining at steady rate cannot also be assumed to hold. However, it appears that model life table selected is similar to mortality pattern of these communities. Even though we have not reported here (but the results are in Appendix I just for purposes of comparison), using Mortpak to simulate child mortality show that whichever model life table chosen (either Coale-Demeny using Trussell equations or original Coale-Demeny), the estimate of child mortality are very similar. The differences are apparent using Coale-Demeny and Unite Nations model life table. As the study was not aimed at identifying the causes of child mortality, information on these causes were not captured. Finally, the social desirability poses some potential limitation on the study since the respondents’ social expectation might not have aligned with that of the study.

Appendix I

Appendix I

References

[1]  You D, Hug L, Ejdemyr S, Jan B. Levels and Trends in Child Mortality. Estimates Developed by the UN Inter-Agency Group for Child Mortality Estimation (UN IGME).UNICEF September 2015.
[2]  National Population Commission (NPC) [Nigeria] and ICF International. 2014. Nigeria Demographic and Health Survey 2013. Abuja, Nigeria, and Rockville, Maryland, USA: NPC and ICF International.
[3]  United Nations. Programme of Action of the International Conference on Population and Development. Adopted at the International Conference on Population and Development Cairo 5–13 September 1994.
[4]  United Nations General Assembly. United Nations Millennium Declaration. A/RES/55/2 edn. New York, NY: United Nations, September 2000.
[5]  Ogunlade S; Mezue C. In: Infant and childhood mortality and socio-economic factors in Africa. (Analysis of national World Fertility Survey data) / Mortalite infantile et juvenile et facteurs socio-economiques en Afrique. (Analyse des donnees nationales de l'Enquete Mondiale sur la Fecondite), [compiled by] United Nations. Economic Commission for Africa [ECA]. Addis Ababa, Ethiopia, United Nations, ECA, 1987. 205-26. (RAF/84/P07).
[6]  Adetunji JA. Infant and child mortality trends and differentials in Nigeria. A paper presented at the 130th Meeting of Population Association of America. Abstract #41692; Monday, November 11, 2002.
[7]  Lawoyin TO. Risk factors for infant mortality in a rural community in Nigeria. Journal of Royal Society for the Promotion of Health 2001; 121 (2): 114-18.
[8]  Oni GA. Child mortality in a Nigerian City: Its levels and socioeconomic differentials. Soc Sc & Med 1988; 27: 607–14.
[9]  Udjo EO. Levels and Trends in Infant and Child Mortality among some Kanuri of North-east Nigeria. Journal of Tropical Pediatrics 1987; 33: 43-7.
[10]  Bradley AK, MacFarlane SB, Moody JB, Gilles HM, Blacker JGC and BD Musa. Malumfashi endemic diseases research project, XX, Demographic findings: mortality. Ann Trop Med Parasitol 1982; 76: 400.
[11]  P. Singha. Demographic Estimation by Indirect Techniques. In: Introductory Text in Demography. Gaskiya Corporation Limited, Zaria, 1996. Pp116-119.
[12]  National Population Commission. Official Gazette of 2006 Census. 2007.
[13]  United Nations Population Studies No. 107. Step-by-Step Guide to the Estimation of Child Mortality. United Nations 1989.
[14]  Trussell TJ. A re-estimation of the multiplying factors for the Brass technique for determining childhood survivorship rates. Population Studies 1975; 29(1): 97-108.
[15]  UN Population Division. 1982. Model Life Tables for Developing Countries. New York: United Nations, Department of Economic and Social Affairs, ST/ESA/SER.A/77.
[16]  Moultrie TA, Dorrington RE, Hill AG, Hill K, Timaeus IM, Zaba B (eds). Tools for Demographic Estimation. Paris: International Union for the Scientific Study of Population, 2013.
[17]  MORTPAK for Windows Version 4.3. United Nations, New York; 2013.
[18]  United Nations, Department of Economic and Social Affairs, Population Division (2015). World Population Prospects: The 2015 Revision, DVD Edition.
[19]  United Nations. Manual X. Indirect Technique for Demographic Estimation. 1983.
[20]  Full report of the Open Working Group of the General Assembly on Sustainable Development Goals is issued as document A/68/970, available at http://undocs.org/A/68/970 (accessed 5th April 2015).
[21]  Countdown to 2015: A Decade of Tracking Progress for Maternal, Newborn and Child Survival. The 2015 Report.