DOI: https://doi.org/10.21203/rs.3.rs-1231501/v1
Utilization of maternal health services is a direct and indirect indicator of perinatal death, and socio-economic development. Evidence on individual and community level determinants of maternal health services in Ethiopia, particularly in the study region was not found. Hence, this study fills this gap.
Among 2,198 study subjects, proportion of pregnant women who visited antenatal care (ANC4+), received skilled delivery care, and postnatal care (PNC) were 66.1%, 58.3%, and 58.6% respectively. Besides these, community and different individual-level factors for the three indicators were detected. For ANC 4th visits: place of resident (AOR=3.82), information on MHS (AOR=2.13), history of pregnancy-related problems (AOR=1.83) and women's decision-making (AOR=1.74). For skilled delivery were belonged in 3rd quintile household wealth (AOR=2.23), women's education (AOR=1.71), attended ANC visit 4th (AOR=2.29), delay initiation of ANC visit between 4–6 months (AOR=0.66). Finally, for PNC services: partner education (AOR=3.67), attending recommended ANC visits (AOR=10.8), iron folic acid supplementation (AOR=1.96) and skilled delivery (AOR=1.63). Hence, community based interventions are strongly recommended to improve utilization of maternal health services.
Regardless of significant reduction of maternal and neonatal mortality globally and developed countries, still in Ethiopia, it was the highest and no significant reduction(1, 2). Maternal health services are essential for women and baby health(1). Even though, utilization of maternal health services are an indirect indicator of maternal and perinatal death, fewer women were using ANC, facility delivery and PNC that is an alarming challenge to reduce maternal mortality(3, 4). This underutilization of maternal health services were due to different factors: socio-demographic and economic(1, 4–12), obstetric(4–6, 9, 10, 13), availability of HFs(10), transportation services and quality of services(9, 11), household index(1, 5–8), women empowerment(6), health insurance(6), information on maternal health services(5, 8, 9).
However, prior studies were determining magnitude and individual level factors using traditional logistic regression which excluded community level factors and advance statistical modeling. This might lead to underestimate or overestimate the magnitude and their predictors which are crucial for the establishment of community based interventions for maternal health services. Multilevel mode is an appropriate method for controlling the nesting effect of clusters at different levels, which is not addressed in previous studies. Therefore, by overcoming the limitation of previous study, the aim of this study was to determine individual (level-1) and community level (level-2) determinants of maternal health services.
Study settings
This study was conducted in Benishangul Gumuz Regional State. It is one of the eleven regions constituting in Ethiopia, located in the Northwest Ethiopia.
Study design and period
A community and health facility linked prospective follow up study design has been conducted from March 2020 to January 2021.
Source population and study participants
All pregnant women within the study area during the time of baseline survey were source population. A randomly selected pregnant women using sampling technique were study participants.
Sample size determination and sampling procedure
Sample size was computed using both single and double population proportion formulas. Then, the calculated maximum sample size using two methods were 812 and 874 respectively. However, this study was part of larger research work, and the sample size determined for another objective was 2,402, which was used as the final sample size for this study. Multistage sampling technique was employed to reach the study participants. Finally, 51 kebeles were randomly selected from the selected districts. Similarly, all eligible public HFs which served the selected kebeles were recruited and make candidate for survey. Thus, 46 health facilities were included for health facility based survey.
Data collection and quality control
Questionnaire was prepared in English, adopted from different relevant sources (2, 3, 14–16). Then, training, pretest, supervision and use of local languages were made to ensure the quality of data. The trained data collectors gather information through face to face interviews. After all, completed questionnaires were reviewed by supervisors on monthly base for accuracy and consistency.
Variables: This study have three outcomes: receiving recommended ANC visits, skilled delivery and PNC 4th visits. Independent variables were categorized in to two levels. Individual level variables (level-1) included individual and household related factors: socio-demographic and economic, obstetric and decision making power. Higher level variables (cluster-2) included community and health facility related factors: place of resident and access to HF.
Data analysis
Data were coded and entered into Epi. Info software, then it was exported, cleaned, edited, and analyzed using STATA software. Descriptive statistics and crude odds ratio at 95%CI were computed for all variables to select candidate variables (p<0.25). Composite indicator of household wealth index was computed and categorized into three categories using Principal Component Analysis. Before running the full model, coefficient of the interaction term at p ≥0.1 and variance inflation factors >10% were determined. Thus, all included variables had no multi-collinearity and interaction effect. Goodness of fit for multilevel model was tested by the log likelihood ratio (LR) test found to be statistical significant, data fit the model. Therefore, multilevel regression model was applied.
Pattern of ANC visit were 1st ANC 1919(87.3%), 2nd ANC 1815(82.6%), 3rd ANC 1674(76.2%), and 4th ANC 1453(66.1%). Key interventions received during ANC contact were informed on danger signs of pregnancy 1740(79.2%) and blood pressure measured 1701(77.4%). Two thirds (65.5%) were initiated their first ANC contact within 4 – 6 months of pregnancy.
The prevalence of skilled delivery service was 58.3%. The reasons for home delivery: labour was going well 424(53.4%) and feeling more comfortable 392(49.4%). Beside, 295(14.3%) of women were suffered from pregnant-related complications. Among them, obstructed/prolonged labour 187(63.4%) and excessive bleeding 102(34.6%) were common problems.
Pattern of PNC visits: 1783(86.3%), 1545(74.8%), 1373(66.5%) and 1210(58.6%) of women were attended 1st PNC, 2nd PNC, 3rd PNC and 4th PNC services respectively. Key services received during PNC visits: immunization services 1692(81.9%) and physical examination 1248(60.4%). The main reasons explored for not utilize PNC were they didn’t teach them well 181(64.2%) and ignorance of her privacy 137(48.6%). Around, 249(12.1%) of women had postpartum complications, of them, the common problems were headache with visual disturbances 170(68.3%) and convulsions/rigidity 118(47.4%)(Table 1).
|
Variables |
Frequency |
Percent |
|---|---|---|
|
Visit of ANC received during last pregnancy 1st ANC contact 2nd ANC contact 3rd ANC contact 4th ANC contact |
1919 1815 1674 1453 |
87.3 82.6 76.2 66.1 |
|
Key interventions received during ANC contact (n = 1919, multiple response) Informed on danger sign of pregnant Blood pressure measured Iron foliate supplementation Nutritional counseling Urine sample taken Blood sample taken Protection of birth from tetanus Other |
1740 1701 1677 1623 1607 1578 1562 22 |
79.2 77.4 76.3 73.8 73.1 71.8 71.1 1.0 |
|
Reason for dropout from ANC follow up (n = 466) Didn’t know about importance of ANC No problems encountered Fear of lack of privacy Influence of other peoples Couldn’t pay for transportation Health institution was too far No transportation services Male health professionals No money to pay for services |
324 298 139 137 114 111 56 45 38 |
69.5 63.9 29.8 29.4 24.5 23.8 12.0 9.7 8.2 |
|
Time of first ANC services initiation (n = 2032) 1-3 months of pregnancy 4-6 months of pregnancy After 6 months of pregnancy |
570 1330 132 |
28.1 65.5 6.5 |
|
Delivery services for last delivery Skilled care Unskilled care |
1281 917 |
58.3 41.7 |
|
Reason for health facility delivery (n = 1404) She was informed to deliver in HFs Previous bad experience from home delivery She faced problems: forced to deliver in HF Others |
1277 263 227 53 |
91.0 18.7 16.2 3.8 |
|
Reason for Home delivery (n = 794) The labour was going well She feel more comfortable at home Close attention from relatives/family It is usual practice No transportation services Cannot pay for transportation services Previous bad experience from ID Cannot afford to pay for health services Culture doesn’t allow to give birth at HF Others |
424 392 332 331 248 202 65 22 15 5 |
53.4 49.4 41.8 41.8 31.2 25.4 8.2 2.8 1.9 0.6 |
|
Pregnant related complications at labour (n = 2065) No Yes |
1769 295 |
85.7 14.3 |
|
Pregnant related complications at childbirth (n = 295) Obstructed/prolonged labour Excessive bleeding during labour Elevated blood pressure Premature rupture of membranes Intrauterine fetal death Preterm labour Others |
187 102 76 69 47 12 5 |
63.4 34.6 25.8 23.4 15.9 4.1 1.7 |
|
Component of PNC contact, she received (n = 2065) 1st contact of PNC services 2nd contact of PNC services 3rd contact of PNC services 4th contact of PNC services |
1783 1545 1373 1210 |
86.3 74.8 66.5 58.6 |
|
The key interventions offered during PNC (n = 206) Immunization of baby Counseling on proper nutrition Breast feeding education Physical examination Family planning services Other |
1692 1516 1436 1248 1074 30 |
81.9 73.4 69.5 60.4 52.0 1.5 |
|
Reason for seeking PNC services (n = 1783) The baby needed it’s immunization The midwife had told me to seek the services She wanted to start family planning She wanted to make sure she is back to normal Because of illness |
1598 1305 975 834 272 |
89.6 73.2 54.7 46.8 15.3 |
|
Reason for not seeking PNC services (n = 282) They did not teach properly Fear of privacy Waiting more time at HF They examined roughly Religious forbidden Health professional shouted at me Other |
181 137 95 77 64 37 14 |
64.2 48.6 33.7 27.3 22.7 13.1 5.0 |
|
Postpartum complications (n = 2065) No Yes |
1816 249 |
87.9 12.1 |
|
Type of postpartum complications (n = 249) Headache, visual disturbances Convulsions/rigidity Fever with or without chills Heavy bleeding Loss of consciousness Foul smelling discharge Severe abdominal pain Other* |
170 118 85 66 66 52 5 3 |
68.3 47.4 34.1 26.5 26.5 20.9 2.0 1.2 |
| * Leg edema, nausea, vomiting and sever malaria during pregnancy | ||
Before running multilevel model, ICC (ρ) was calculated in the empty model for ANC 4th, skill delivery and PNC 4th. Meantime, ICC/rho (ρ) was calculated as a full model for each outcome. Rho (ρ)/ICC and test preference of log-likelihood was determine in the empty and full model for ANC 4th visit, skilled delivery care and PNC 4th visits and indicating that statistical significant association. Hence, multilevel model is the preference model for each outcome (Table 2).
|
Models |
Fixed intercept -cons(95%CI) |
Random effect as Level-2 variance var(-cons (95%CI)) |
Intra-class Correlation Coefficient: ICC(ρ) |
Log likelihood (LR)-deviance |
Significance of LR test Vs Logistic regression (P-value) |
|---|---|---|---|---|---|
|
ANC 4th and more Empty model Full model |
2.54(1.7, 3.78) 0.26(0.04,1.67) |
1.92(1.23, 3.04) 2.76 (1.54, 4.96) |
0.37 = 37% 0.46 = 46% |
-1161.48 -633.06 |
P < 0.0001 P < 0.00001 |
|
Skilled delivery Empty model Full model |
1.48(1.17, 1.87) 1.18(0.3,4.66) |
0.59(0.36, 0.96) 1.1(0.63, 1.92) |
0.15 = 15% 0.25 = 25% |
-1402.31 -706.65 |
P < 0.0001 P < 0.00001 |
|
PNC 4th Empty model Full model |
1.59(0.99, 2.54) 0.08(0.01,0.71) |
2.72(1.72, 4.33) 2.25 (1.27, 3.98) |
0.45 = 45% 0.40 = 40% |
-1099.36 -523.25 |
P < 0.0001 P < 0.00001 |
| P value less than 0.05 is statistically significant and the data fit for the multilevel model | |||||
| * Multilevel regression model applied to measure the effect of factors on outcome | |||||
After controlling confounders, among community level variables: place of resident was statistically significant association with recommended ANC visits but it was not statistical significant association with the rest outcomes. The odds of receiving the recommended ANC visits among women who reside in urban (AOR=3.82; 95%CI: 1.35-10.78) were four times higher than among women who reside in rural area.
Regarding individual level factors, the odds of receiving the recommended ANC visits among women who had any information on MHS (AOR=2.13; 95%CI: 1.12-3.75), history of pregnant related problems (AOR=1.83; 95%CI: 1.15-2.2), stillbirth (AOR=1.67; 95%CI: 1.02-2.73) and decision making power (AOR=1.74; 95%CI: 1.14-2.68) were two times higher than among women belonged with their counterpart.
Similarly, the odds of utilizing skilled delivery among women who completed recommended ANC visits (AOR=2.29; 95%CI: 1.59-3.32), belonged in 3rd quintile wealth index (AOR=2.23; 95%CI: 1.27-3.89), attended primary school (AOR=1.71; 95%CI: 1.04-2.81) were higher than among women residing within their counterpart. However, women delayed 1st ANC visit initiation between 4 – 6 months of GA (AOR=0.66; 95%CI: 0.45-0.96) were lower in the odds of receiving skilled delivery.
The odds of utilizing PNC services among women who received ANC 4th visits (AOR=10.8; 95%CI: 6.79-17.2), partner attended tertiary education (AOR=3.67; 95%CI: 1.40-9.58), decision making power (AOR=1.8; 95%CI: 1.09-2.97), iron folic acid supplementation during pregnancy (AOR=1.96; 95%CI: 1.11-3.49) and skilled delivery (AOR=1.63; 95%CI: 1.11-2.42) were higher than among women who belonged within their counterpart(Table 3).
|
Variables |
ANC 4th + visits |
Skilled delivery care |
PNC 4th |
|---|---|---|---|
|
AOR (95%CI) |
AOR (95%CI) |
AOR (95%CI) |
|
|
Level – 2 (Community level) variables |
|||
|
Place of resident Rural Urban |
1 3.82(1.35, 10.78) |
1 1.22(0.55, 2.73) |
1 1.14(0.44, 2.91) |
|
Distance to Health Post < 2 Hours >= 2 Hours |
1 0.25(0.02,2.73) |
1 0.49(0.08, 2.89) |
1 0.88(0.1, 10.26) |
|
Leve-1 (individual level) variables |
|||
|
Household Wealth Index 1st Quintile (Poor) 2nd Quintile (Middle) 3rd Quintile (Rich) |
1 1.06(0.69, 1.61) 1.46(0.81, 2.62) |
1 1.13(0.78, 1.63) 2.23(1.27, 3.89) |
1 0.74(0.46, 1.21) 0.43(0.22, 1.06) |
|
Age (Years) < 20 20 – 29 >= 30 |
1 0.95(0.27, 3.31) 0.86 (0.24, 3.02) |
- - - |
1 1.53(0.42,5.62) 1.33(0.36,4.96) |
|
Women educational level No formal education Primary school High school Tertiary education |
1 0.92(0.54, 1.57) 1.14(0.57, 2.28) 1.43(0.60, 3.37) |
1 1.71(1.04, 2.81) 1.49(0.77, 2.89) 4.12(1.49, 11.33) |
1 0.69(0.38, 1.27) 0.68(0.3, 1.52) 0.52(0.2, 1.37) |
|
Partner educational level No formal education Primary school High school Tertiary education |
1 1.22(0.69, 2.15) 0.87(0.48, 1.59) 1.98(0.88, 4.45) |
1 0.66(0.39, 1.1) 0.76(0.44, 1.32) 0.56(0.25, 1.25) |
1 1.07(0.57, 1.98) 1.49(0.75, 2.95) 3.67(1.40, 9.58) |
|
Partner occupational status Governmental employee Others |
1 1.23(0.64, 2.36) |
1 0.61(0.31, 1.21) |
1 1.48(0.69, 3.18) |
|
Information on MHS No Yes |
1 2.13 (1.21, 3.75) |
1 1.23(0.7, 2.17) |
1 1.04(0.49, 2.18) |
|
Age at first marriage (year) < 18 >=19 |
1 0.92(0.54, 1.56) |
1 1.13(0.7, 1.82) |
1 0.98(0.54, 1.73) |
|
Age at first pregnancy (year) < 19 >= 19 |
1 1.13(0.67, 1.91) |
1 1.24(0.77, 2.01) |
1 0.98(0.55, 1.74) |
|
History of pregnant related problem during labour for previous birth No Yes |
1 1.83 (1.15, 2.92) |
1 0.57 (0.37, 0.86) |
1 1.63(0.98, 2.7) |
|
History of stillbirth No No |
1 1.67(1.02, 2.73) |
1 0.43 (0.28, 0.66) |
1 0.51(0.29, 0.87) |
|
Women decision making power Didn’t make decision Make decision |
1 1.74(1.14, 2.68) |
1 1.22 (0.81, 1.85) |
1 1.8(1.09, 2.97) |
|
ANC 4th visit completed No Yes |
- |
1 2.29(1.59, 3.32) |
1 10.8(6.79, 17.2) |
|
Offered information of danger sign during ANC visit No Yes |
- |
1 0.7 (0.44, 1.12) |
1 0.71(0.38, 1.33) |
|
Time of 1st ANC visit start Within 3 months of GA 4 – 6 months of GA After 6 months of GA |
- - - |
1 0.66(0.45, 0.96) 0.33(0.16, 0.68) |
1 0.39(0.24, 0.61) 0.1(0.03, 0.24) |
|
Provision of information on health facility delivery No Yes |
- |
1 0.9(0.49, 1.68) |
1 1.05(0.51, 2.17) |
|
IFA supplementation during pregn. No Yes |
- |
- |
1 1.96(1.11, 3.49) |
|
Provision of TT during pregnant No Yes |
- |
- |
1 1.58(0.93, 2.69) |
|
Skilled delivery care No Yes |
- |
- |
1 1.63(1.11, 2.42) |
|
Pregnant related problems immediately after labour No Yes |
- |
- |
1 1.1(0.55, 2.21) |
Magnitude of maternal health services
In this study, 66.1% pregnant women were attended the recommended ANC visits which was consistent with different studies(17–20). However, it was lower than studies in South Asia and Sub-Saharan Africa (18), Debre Berhan(21) and Northern Bangladesh(22). But, it was higher than evidence in Pakistan(23), Ratanakiri(24), Khammouane(25), Sub–Saharan Africa(26), EDHS-2014(27), Arbaminch(28), West Gojjam(29) and Tigrary(30). This is due to variability of socio-demographic, wealth status of the nations, availability and accessibility HFs and medical equipment’s in the health system.
This study revealed that 58.3% of birth attended by skilled providers which was consistent with other study(23). This finding was lower than study in Cambodia(19); Tanzania(31); Bihar(20). Whereas, it was higher than evidence in Ratanakiri(24), Khammouane(25), West Gojjam(29) and South Ethiopia(28). This discrepancy might be due to variation of culture, belief, time of study and design. However, reason mention for facility delivery, the main reasons for home delivery: labour was going well and being feel more comfortable at home delivery which were consistent with study done in West Gojjam(29).
This study found that 58.6% of women received the recommended PNC visits, which was lower than study in Pakistan(23), Sub – Saharan Africa(26) and Ghana(32). Whereas, it was higher than study in West Gojjam(29), Ratanakiri(24), Arbaminch(28) and Khammouane(25).
Determinant of maternal health services
In this study, women who had information on MHS were two times higher to receive recommended ANC visits which was supported by studies conducted in different setting (5, 8). Women who had history of pregnant related problems and stillbirth were two times higher to completed whole visits of ANC which was supported by other study(9). This is because prior bad experiences give good lesson for women that encourage them to consult health professional during pregnancy.
Similarly, women who had decision making power were almost two times higher to receive the recommended ANC visits. This finding is consistent with study done in Nigeria(6). This might be because if the household resources are controlled by others and no power to decide on their resources, women do not have the freedom to access health services whenever they need care. Education of women was positively significant effect on the utilization of institutional delivery services which was supported by other studies(1, 5, 6, 9, 10). This positive correlation can be explained by the fact that educated women are more aware on the importance of medical services to their mothers. Household wealth index is strongly linked to place of delivery: as household wealth index increases, the uptake of institutional delivery service is also increase. This evidence is supported by different studies(1, 5–7, 11).
Uptake of PNC service is directly related with partner education which showed a positive association which was consistent with other studies(5, 6, 12). This is because educated husbands may have a better understanding on the benefit of maternal health services. Similarly, receiving the recommended, timely initiation of ANC visits and skilled delivery are strongly significant association with utilization of PNC. This evidence is strongly supported by SRMA pooled result(33) and other studies(4–6).
Generally, the coverage of ANC 4th visits, skilled delivery and PNC services were low as compared with national target. This study explored different individual and community level factors that influenced utilization of the services which had important programmatic implication. Hence, we strongly recommended that reinforcing women’s autonomy and community based interventions to enhance utilization of maternal health services.
Even though health facility based data were collected by health workers, social desirability bias was expected which compromised the finding.
Around 8.5% of the study participates were lost to follow up that might have some deviation on the result.
Antenatal Care
Adjusted Odds Ratio
Breast Feeding
Benishangul Gumuz Region
Confidence Interval
Ethiopia Demographic and Health Survey
Gestational Age
Health Facility
Intra class Correlation
Iron Folic Acid
Income Generating Activities
Institutional Review Board
Log Likelihood
Maternal and Child Health
Maternal Health Service
Maternal and Perinatal Death Surveillance Response
Non-Governmental Organization
Principal Component Analysis
Postnatal Care
Safe Motherhood Initiative
School of Public Health
Systematic Review and Meta-analysis
Tetanus Toxoid
Variance Inflation Factors
Ethical approval was endorsed from Research Review and Ethics Committee (REC) of School of Public Health, Addis Ababa University's with protocol number SPH/3089/011 and Institutional Review Board (IRB) of College of Health Sciences of Addis Ababa University with protocol number 048/19/SPH. Necessary permission letters were obtained from Regional Health Bureau and respective all local districts. Confidentiality was maintained by avoiding any identities from the questionnaire. Before starting actual data collection, written and verbal consent was obtained from each study subjects.
Not Applicable
The datasets used and analyzed during the current study are available from the corresponding author on reasonable request.
The authors declare that they have no competing interests.
This study funded by Addis Ababa University but there is no other funding agencies for this works
MA conceived and designed the study. Then after, data was collected, analyzed, interpreted and wrote the whole document. AW and GT were critically commenting the whole document and genuinely guide the whole work. All authors read and approved the final manuscripts.
We would like to thank Addis Ababa University, College of Health for providing ethical approval and partially sponsored this project work. The authors would also want to express their gratitude to all of the pregnant women, data collectors, supervisors, and other individuals who contributed directly or indirectly to the study's success. Last but not least, the author wishes to express his gratitude to Miss Abebech Tefera, Miss Chaltu Argeta, Dr. Nigatu Disassa, and Assistant professor Atnafu Morka for their unwavering assistance in the preparation, magement and duplication of research questionnaires.