1. Md Aminul Haque, Assistant Professor, Department of Population Sciences, University of Dhaka. Bangladesh.
Master in Governance and Development, Institute of Development Policy of Management, University of Antwerp, Belgium. Master of Social Sciences in Sociology, University of Dhaka, Dhaka-1000, Bangladesh

2. Amir Mohammad Sayem, Master of Population Sciences, University of Dhaka, Dhaka-1000, Bangladesh

3. Dr. Nilufar Yeasmin Nili, Masters in Population Sciences, University of Dhaka

Correspondence:
Md. Aminul Haque
Assistant Professor
Department of Population Sciences
University of Dhaka. Dhaka-1000
Bangladesh
Ph: 88-01712529264
E-mail: aminuldps@gmail.com , aminul1975@yahoo.com

INTRODUCTION

The International Conference on Population and Development (ICPD) held in Cairo in 1994 emphasized safe motherhood in which all pregnant women are expected to receive basic antenatal care (ANC). Such ANC package includes monitoring pregnancies for signs of complications, treating concurrent problems of pregnancy, providing iron supplements, tetanus toxoid immunization and counseling on preventive care and related issues. The purposes of providing such services was to identify signs of, or risk factors for, complicated pregnancies and identify the appropriate health provider (Vanneste and others, 2000; Magadi, Madise and Rodrigues, 2000; Matthews and others, 2001).

Such care necessity is due to the fact that complications during the antenatal period can result in various complications at the time of delivery such as hemorrhage, which increases the risk of excessive hemorrhage during delivery; the risk of obstructed labor increases significantly if abdominal pain is observed during the antenatal period; prolonged labor appears to be significantly higher for the first pregnancy; and mothers suffering from abdominal pain during pregnancy tend to have a higher risk of prolonged labor during delivery (Islam et al, 2004).

In most developing countries, women of reproductive age constitute more than one-fifth of the total population. These women are exposed repeatedly to the risk of pregnancy and childbearing and, under existing socioeconomic conditions and the inadequacy of medical and health facilities, are at greater risk of morbidity and mortality from causes related to pregnancy (Bhatia, 1993). As expected, in Bangladesh, about 12,000 women die each year from maternal causes while the estimated lifetime risk of dying from pregnancy and child birth-related causes is about 100 times higher than that in developed countries (NIPORT 2003). A tragic consequence of such deaths is that about 75% of the babies born to these women are also likely to die within the first week of their life (WHO 2004). In this regard, different studies in developing countries identified the causes of maternal deaths have repeatedly emphasized the need for antenatal care utilization (Maine, 1986; Fauveach et al, 1988; Fortney et al, 1988). Receipt of ANC services plays a pivotal role in bringing the women into delivery care-seeking. Moreover, women who receive ANC were more likely to receive Postnatal care (PNC) (Chakraborty, et al, 2003) and preventive care for children aged one to three years (Bracken et al, 2003).

Generally, the utilization of ANC service is poor in Bangladesh to say the least; moreover, in poor socio-economic areas such service uptake is at an unacceptable level, though the World Health Organization (WHO) recommended to ensure at least three visits during pregnancy. However, to the best of our knowledge, there is little or almost no extensive work on differential determinants of uptake of ANC in slum areas. Considering such matters, our study focused on socio-economic, demographic and cultural factors that influence the use of a number of ANC services among women in slum areas of Dhaka city.

Study Area, Population and Data Collection
The data were drawn from a randomly selected group of slum dwelling women of reproductive age through a semi structure based survey questionnaire that collected information on the socio-economic, demographic and cultural characteristics of respondents as well as the family along with uptake of the number of ANC visits during pregnancy. A cluster sampling technique was used in selecting the study population where at first a cluster of slum areas was randomly selected which included Khilgaon, Meradia and Rampura slums. From each slum area women who had at least one birth on or before March 1, 2008 were randomly selected and a total of 540 women were successfully interviewed. Before applying the final interview schedule a pre-test of the questionnaire was conducted among 25 women in Khilgaon and Meradia slums in order to maintain the sequence, coding and to localize the wording of the questionnaire. Informed consent was taken from each subject before interview. In this regard, the purpose and rationale of this study was fully explained to respondents.

Measuring Variables
The times of ANC uptake during the last pregnancy was used as a dependent variable while several socio-economic, demographic and cultural variables were included as the independent variables. Among the independent variables, four variables such as mass media exposure to respondents, respondent's autonomy, respondent's attitude towards maternal health care services and male participation in maternity care were used as index variables. The mass media exposure was measured with three indicators like frequency of listening to radio, watching television and reading newspaper, per month. Respondent's autonomy, positive attitude and male participation were measured by respectively drawing information on 9, 8 and 6 yes-no type questions. In this case, yes was numbered as 1 while no was 0, which later converted into scale score following the arithmetic transformation procedure.

Data Processing and Analysis
The collected data was edited two times i.e., once by supervisor during the data collection and another by the researcher just before entering the data into computer in order to maintain the quality of the data. Moreover, data were doubly entered in the computer for maintaining accuracy of the data. At the analysis stage, univariate, bivariate and multivariate analyses were conducted. Mainly frequency distribution was done in univariate analysis, while at bivariate level, mean comparison and Pearson correlation coefficient were used respectively for observing differences between and/or among different levels of independent variables and correlation between each independent and dependent variable. As the dependent and independent variables were interval level of measurement, simple linear regression analysis was considered in order to identify the factors determining the times of ANC uptake. In this regard the following regression equation was used to estimate the regression coefficients:

Y= a + b1*X1…………+ bk* Xk + e

Where, Y = dependent variable, a = constant, b = the regression coefficient, X = independent variables of the model, k = end number of the series and e= error term

Level of ANC Uptake
A significant portion of women received ANC services; however, the number of such visits was less in slum areas of Dhaka city. On average, around 2 ANC visits were reported among women aged 15-49 years with standard deviation 1.849 (Table 1). In total, 38.7% women did not receive any ANC service while 61.3% women reported to receive at least one ANC service. Around 15% of women were found to receive ANC for 1-2 times, while around half of the women reported they received ANC at least three times. Though the number of women was decreasing with the increasing number of visits, a small percentage of women reported to receive it for 6, 7 and 8 times.

Ways of Receiving ANC Services
Women were more likely to receive ANC service from government hospital than that of private clinic (Figure 1). Among 331 women, who received ANC, 46.22% reported receiving it by visiting a government hospital while 36.25% received it by visiting a private clinic. Slightly more than one-tenth of women were also likely to receive it from both government and private health facilities. Moreover, 5.14% of women reported receiving an ANC service from a service provider who visited the respondents.

Figure 1. Ways of receiving ANC services among women in slum areas

Correlates of ANC Uptake
The differential pattern of number of ANC uptakes was observed by several socio-economic, demographic and cultural factors. The results are displayed in Table 2 which made it clear that the use of a number of ANC services varied with different socio-economic characteristics of Table 2. The distribution of mean comparison and correlation coefficients of uptake of number of ANC services by different socio-economic, demographic and cultural factors of respondents are taken into consideration. Women with primary and >primary education were more likely to use a higher number of ANC services compared to women with no education. On average, women without schooling used ANC services 1.41 times while these were used 2.14 and 3.25 times respectively by women with primary and >primary education. Between number of ANC services and respondent's education a moderate correlation was found with P<0.01. Similarly, a positively significant relationship was observed with varying increasing mean number of ANC services received by women with their increasing monthly family income, husband's monthly income, respondent's monthly income, mass media exposure, male participation, respondent's autonomy and respondent's positive attitude.

Women's age at present, had a negative significant correlation with number of ANC services received by women. Women aged <20, on average, received ANC 1.94 times while women aged 20-29 and 30-39 were found to receive on average slightly more, respectively, 2.17 and 2.11 times. However, women aged 40-49 reported they received only 1.16 number of ANC services. Distance between home and clinic was found to have significant correlation with the number of ANC services received by women. Women whose household was within 1 and 2-3 kilometers from the clinic, on average, received 1.63 and 2.24 ANC services while women with increased distance from home to clinic (>3 KM) women reported to receive only 1.35 ANC services. Similarly a significant relationship with the same direction was found in regard to the order of the respondent's order of last birth. Women with first parity received ANC, on average, 1.67 times; as expected the mean number of ANC services received was reduced to 1.02 for women with parity >3.

Determinants of ANC Uptake
The main objective of this study was to examine the determinants of the number of ANC services received by women of reproductive age group. To do so, simple linear multiple regression technique was applied after meeting the criteria such as homoscedasticity of variance, multi-collinearity, interval level of measurement of dependent and independent variables. All the variables used at bivariate analysis were included in the regression model; however, as stepwise method was applied during multivariate analysis, only significant variables were displayed in the output (Table 3).

Overall, the regression model explained 31.1% of variance in the number of ANC services received by women (P<0.001). All the variables found significant at bivariate analysis, were also found significant at multivariate analysis. The most explanatory variable was respondents' education, which explained 11.6% variance (p<0.001) indicating the higher the education of respondents, the higher the number of ANC services received by women. Among other significant variables respondents' autonomy, mass media exposure, male participation, monthly family income, monthly income of husband, and respondents' monthly income positively explained 3.4%, 2.9%, 2.6%, 2.0%, 2.0% and 1.1% of variance respectively. Results of such positively significant variables suggest that women with higher autonomy, mass media exposure, male participation, monthly family income and their own monthly income were higher receivers of ANC.

As in bivariate analysis, distance from home to clinic has a negative influence on the number of ANC services received by women. Individually it explained 1.0% variance indicating that the higher the distance the lower the number of ANC services received. Similar to distance, respondent's order of last birth negatively explained 4.5% of variance also suggesting the higher the order the lower the number of ANC received.

In order to reduce the risk of mother and child which ultimately reduce the maternal mortality ratio (MMR), proper antenatal care (ANC) is very crucial. In this study 61.3% of women received ANC during their last pregnancy which is lower than urban utilization (75.7%), and slightly higher than the national statistics (60.3) (NIPORT et al., 2007). The percentage of women who received ANC for at least three times was 45.9%. Overall, the mean number of visits in slum areas was only less satisfactory as only 1.95 ANC was received by study women which is notably less than the minimum required 3 visits. Among the women who received ANC service, women were comparatively more likely to receive it from government hospital than that of private clinic.

Variation was observed in receiving the number of ANC by women with differing socio-economic, demographic and cultural characteristics of respondents as well as the family. Attempting to identify the determinants of receive of ANC, this study found some significant explanatory variables influencing the number of ANC received by. As expected, women's education was found to have significant positive impact on use of ANC service which is similar to other findings (Elo, 1992; Prasad, 2000; Matsumura and Gubhaju, 2001). This is because better educated women are more aware of health problems, know more about the availability of health care services, and use this information more effectively to maintain or achieve good health status. Moreover, mother's education may enable her to enhance female autonomy so that women develop greater confidence and capability to make decisions about their own health (Caldwell, 1981; Raghupathy, 1996).

Respondent, family and husbands' monthly income had positive impact on ANC receive in study areas. This may be due to that respondents', husband of respondent and family income is usually considered as social status. Differences in attitudes to modern health care services by income of husband and family were depicted as a predisposing factor as income enables acquisition of more and better health care (Fiedler, 1981). As a result, the higher use of ANC services is expected with higher income. In a study conducted by Hadi et al (2007) it was clearly found that the health status of the population could not be improved further without fundamental changes brought about to education and income (Hadi et al, 2007).

Autonomy of women regarding maternal health care services had significant positive impact on number of ANC received by women. Autonomy is not a singular matter, rather it comes with many things like education, income, employment of women, etc. General argument suggests that autonomous women are more likely to decide about the family planning themselves or take part in such decision making process with their partner along with other household decisions. As a result, they become more conscious about their lives including heath matters.

Mass media are effective in information dissemination, which increases awareness about innovations, and fosters inter-personnel communication, which could facilitate behavioural changes allowing for the adoption of new/different behaviours (Valente et al., 1996). With regard to maternal health care uptake, it changes to beliefs and behaviour of women to receive health service. As a result, it is expected that women with higher mass media exposure are more likely to receive higher number of ANC. Similar to this concept, mass media had significant positive impact the number of ANC uptake.

Women who experienced higher male participation in their last pregnancy were significantly more likely to receive higher number of ANC compared to women with lower male participation. This is most probably due to the fact that in a traditional society like Bangladesh, the most of the decisions including family planning and health seeking decisions mainly comes from male partners. Besides, men are less likely to accompany women to hospital in health seeking where women alone are often are not permitted to visit the health facility alone. Moreover, women are generally ignored in regard to health care especially when lower income in the household limits the expenditure on the family health budget. In this regard, supporting financially in health care uptake as well as accompanying women to clinic by male partners increase the uptake of ANC among women.

The order of the last birth had been found to have significant negative impact on number of ANC received by women which is similar to other findings (Elo, 1992; Bhatia and Cleland, 1995). Because the lower parity woman tend to give careful attention to seeking antenatal care due to their inexperience in pregnancy while lower number of ANC receive, among higher parity women could be due to time and resource constraints faced by those with larger families (Wong et al., 1987; Elo, 1992; Bhatia and Cleland, 1995).

As expectedly, distance from respondent's home to clinic had negative effect on number of ANC received by women. Generally it is believed that women near to clinic are more interested to receive ANC service due to less time needs, less cost needs and women also can visit clinic alone. The lower uptake among women whose home had at more distance from clinic may be due to the poor road construction, time consuming distance, more financial necessity to receive care as well as the need of accompanying women to visit to the clinic.

The receive of ANC services is very essential in order to make the delivery safer as well as protecting the lives of both mother and child. In response to it, at least 3 ANC visits for each pregnant women is strongly recommended by World Health Organization (WHO). However, this study found that women on average were likely to use less than 2 number of ANC during their last pregnancy. The different socio-economic and cultural factors were examined and found to be associated with such use of ANC. Therefore, it may be suggested that in order to further increase the use of number of ANC in slum areas, the determinants found in this study may provide a priority based focus for the program planners and policy makers.

Acknowledgements
This study was conducted as a requirement of the degree of Masters of Population Sciences at the Dept. of Population Sciences, University of Dhaka. Thus, the authors are likely to acknowledge the gratitude of this Dept. for allowing us to conduct the study. The grateful appreciation also goes to the United Nation Population Fund (UNFPA) for its financial assistance during the study.

1. United Nations (1995). Report of the Fourth World Congress on Women Document A.Conf. 177/20 (New York, United Nations).
   
2. Vanneste, A.M. and others (2000). "Prenatal screening in rural Bangladesh: from prediction to care", Health Policy and Planning, vol. 15, No. 1, pp. 1-10.
   
3. Magadi, M.A, N.J. Madise and R.N. Rodrigues (2000). "Frequency and timing of antenatal care in Kenya: Explaining variations between women of different communities", Social Scienceand Medicine, vol. 51, No. 4, pp. 551-561.
   
4. Matthews, Z. and others (2001). "Antenatal care, care-seeking and morbidity in rural Karnataka, India: Results of a prospective study" Asia-Pacific Population Journal, vol. 16, No. 2, pp. 11-28.
   
5. NIPORT, ORC Macro, John Hopkins University, ICDDR,B (2003). Bangladesh maternal health services and mortality survey 2001. Dhaka and Claverton: National Institute of population Research and Training (NIPORT), ORC Macro, John Hopkins University and ICDDR,B.
   
6. World health Organization (2004). Skilled birth attendance: review of evidences in Bangladesh. Dhaka: WHO Country Office Bangladesh. http://www.whoban.org/
   pdf/Skill%20Birth%20Book.pdf accessed on 16 March 2005.
   
7. Chakraborty, N.,Islam, M.A., Chowdhury, R.I. and Bari, W. (2003) Analysis of ante-partum maternal morbidity in rural Bangladesh. Aust J Rural Health. 11 (1), Jan: 22-7.
   
8. Islam, M.A., Chowdhury, R.I., Chakraborty, N., Bari, W. and Akhter, H.H. (2004) Factors associated with delivery complications in rural Bangladesh. Eur J Contracept Reprod Health Care. 9 (4), Dec: 203-13.
   
9. Bracken, A. and Goodman, D. The Impact of Prenatal and Postnatal Home Visiting on Utilization of Care. Abstr Academy Health Meet. 2003; 20: abstract no. 665.
   
10. Elo, T. I. (1992) Utilization of maternal health-care services in Peru: the role of women's education. Health Transition Review 2, 49-69.
    
11. Caldwell, J.C. (1981). Maternal education as a factor in child mortality. World Health Forum 2:75-78.
    
12. Raghupathy, S. (1996). Education and the use of maternal health care in Thailand. Social Science & Medicine 43(4):459-471.
    
13. Fiedler, J. L. (1981) A review of the literature on access and utilization of medical care with special emphasis on rural primary care. Social Science and Medicine 15, 129-142.
    
14. Bhatia JC. Levels and causes of maternity and mortality in Southern India. Studies in Family planning. 1993; 24(5): 310-318.
    
15. Maine D. Maternal Mortality: helping women off the road to death in rural Bangladesh: an empirical study. Social Science and Medicine. 1986; 32 (1): 43- 49.
    
16. Fauveach V, Koeing M, Chakrabort I, et al. Causes or maternal mortality in rural Bangladesh: 1976-1985. Bulletin of the World Health Organization. 1988; 66 (5): 643-651.
    
17. Fortney J, Susanti S, et al. Maternal Mortality in Indonesian and Egypt. British Journal of Gynecology and obstetrics. 1988; 26: 21-32.
    
18. Matsumura,M. and Bina Gubhaju (2001). "Women's status, household structure and the utilization of maternal health services in Nepal" Asia-Pacific Population Journal, vol. 16, No. 1, pp. 23-44.
    
19. Prasad, P. (2000). "Health care access and marginalized social spaces. Leptospirosis in south Gujarat" Economic and Political Weekly, vol. 35, No. 41, pp. 3688-3694.
    
20. Hadi A, M. Naeem Mujaddidi, Taufiqur Rahman and Jalaluddin Ahmed. The Inaccessibility and Utilization of Antenatal Health-Care Services in Balkh Province of Afghanistan. Asia Pacific Popualation Journal, Vol. 22, No. 1, 2007.
    
21. Valente, T.W., Poppe, P.R., & Merritt, A.P. (1996). Mass-media-generated inter-personnel communication as sources of information about family planning. Journal of Health Communication 1(3), 247-265.
    
22. Wong, E. L., Popkin, B. M., Gullkey, D. K. and Akin, J. S. (1987) Accessibility, quality of care and prenatal care use in the Philippines. Social Science and Medicine, 24, 927-944.