Geographic patterns and determinants of family planning satisfaction among women of reproductive age in Ethiopia

Abstract

Introduction

While studies in Ethiopia have explored family planning use and contraceptive determinants, limited research examines the spatial variation and predictors of satisfaction with family planning methods. This study addresses this gap using geographically weighted regression (GWR) analysis to assess spatial patterns and determinants of satisfaction with family planning methods among reproductive-age women. Utilizing nationally representative 2021 Performance Monitoring for Action Ethiopia (PMA-ET) data, the study aims to provide targeted insights for improving family planning services.

Methods

A weighted sample of 1,456 reproductive-age women was analyzed. Spatial analysis was conducted using ArcGIS 10.7 and SaTScan 9.6. Hotspot detection, ordinary least squares (OLS) regression, and geographically weighted regression (GWR) were applied. The Bernoulli model was used to identify spatial clusters of satisfaction. Associations between satisfaction and explanatory variables were assessed using OLS and GWR, with statistical significance set at P < 0.05.

Results

The overall satisfaction rate with current family planning methods was 84.1% (95% CI: 82.12, 85.97). Satisfaction levels exhibited a clustered spatial distribution. High-satisfaction hotspots were detected in Addis Ababa, most parts of Amhara, and some areas of Benishangul-Gumuz. SaTScan identified a primary spatial cluster (RR = 1.19, P < 0.001) in Addis Ababa, northern southwestern Ethiopia, and eastern Gambela, and a secondary cluster (RR = 1.18, P < 0.001) in northern Addis Ababa and southern Amhara. GWR analysis revealed that younger women (15–24 years), those with poor wealth status, and urban residents were significant spatial predictors of satisfaction.

Conclusion

The study revealed that satisfaction with the current use of family planning methods among reproductive-age women in Ethiopia exhibits regional variations. Statistically significant clusters of higher satisfaction were identified in Addis Ababa, most parts of Amhara, and some parts of Benishangul-Gumuz. Geographically Weighted Regression (GWR) analysis indicated that women aged 15–24, those with poor wealth status, and urban residents were significant spatial predictors of satisfaction at the local level. These findings suggest the need for targeted regional strategies to improve satisfaction with family planning services. Interventions should prioritize high-satisfaction areas to sustain progress while addressing disparities in underserved regions. Policymakers should focus on enhancing equitable access, strengthening awareness campaigns, and mitigating socioeconomic barriers to improve satisfaction with family planning.

Introduction

Satisfaction with family planning methods is a key indicator of service quality and influences contraceptive continuation and effectiveness [1]. Globally, an estimated 214 million women have an unmet need for family planning, and discontinuation rates remain high. In Africa, only 33% of women use modern contraceptive methods, with sub-Saharan Africa reporting a contraceptive prevalence rate of 23% and an unmet need of 24% [2,3,4]. In Ethiopia, the contraceptive prevalence rate is 36%, while 22% of women have an unmet need for family planning [5]. Despite efforts to improve access, little is known about women’s satisfaction with their current contraceptive methods. Low satisfaction may contribute to inconsistent use, leading to unintended pregnancies, unsafe abortions, maternal and child health risks, and broader socio-economic consequences such as poverty, reduced educational opportunities, and limited healthcare access [6, 7].

As indicated in previous studies, age, educational status, wealth status, media exposure, residence, marital status, and other related factors were significant factors for satisfaction with the current use of family planning methods among reproductive-age women [8,9,10].

While studies in Ethiopia have examined family planning use and contraceptive determinants, little research has explored the spatial variation and predictors of satisfaction with family planning methods. To address this gap, this study applies geographically weighted regression analysis to provide nuanced insights. Using nationally representative data from the 2021 Performance Monitoring for Action Ethiopia (PMA-ET), it examines spatial patterns in satisfaction, enabling targeted interventions in regions with lower satisfaction rates. This study aims to comprehensively analyze the geographic patterns and determinants of family planning satisfaction among reproductive-age women in Ethiopia.

Materials and methods

Study design, study area, and data source

In the PMA-ET 2021 study, a community-based cross-sectional design employed a two-stage cluster approach with residential areas (urban and rural) and sub-regions as strata, ensuring representation across all 12 geographic regions in Ethiopia. The study divided each region into urban and rural zones, resulting in 321 sampling strata. Households were then systematically selected through random sampling. Eligible participants included all women of reproductive age who were regular members of the chosen households or who stayed overnight in the household before the survey. Notably, 95% of the target population, women aged 15–49 years, resides in four key regions: Addis Ababa, Amhara, Oromia, and SNNP. To address regions with less than 5% of the target population, a sixth synthetic region, denoted as “other,” was created. Due to population distribution and resource constraints, regional representative samples were taken exclusively in the four major regions. The sampling design, comprising 321 Enumeration Areas, aims to achieve a national-level margin of error below 2%, below 3% for urban and rural estimates, and below 5% at each of the four regional levels, ensuring robust and precise estimates of male involvement in family planning decision-making among reproductive-age women. The secondary data for this analysis were obtained from PMA-ET of 2021 which was found in the PMA portal (https://www.pmadata.org/ _ 2021) (Fig. 1).

Fig. 1

Map of the study area (Ethiopia) Shapefile source: CSA, 2013; URL: https://africaopendata.org/dataset/ethiopia- shapefiles

Population

The study population comprised all reproductive-age women surveyed in Ethiopia. A weighted total sample of 1,456 women of reproductive age were included in the analysis, encompassing all variables of interest. Women with missing responses on satisfaction were excluded, particularly those from enumeration areas with zero coordinates, as their geographic information was incomplete, preventing accurate spatial analysis. the Performance Monitoring for Action Ethiopia (PMA-ET) dataset follows standardized data collection protocols, including interviewer training, supervision, and real-time data validation to ensure accuracy and reliability.

Study variables

Dependent variable

The proportion of respondents at the cluster level who report satisfaction with their current family planning method. This is calculated by dividing the number of satisfied respondents by the total number of respondents in each cluster.

Independent variables

Women’s age (aggregated proportion of 15–24, aggregated proportion of 25–34, and aggregated proportion of 35–49), Women’s education status (aggregated proportion of no education, aggregated proportion of primary education, aggregated proportion of secondary education, and aggregated proportion of higher education), residence (aggregated proportion of urban or rural), marital status (aggregated proportion of single, married, divorced/separated, and widowed), and wealth index (aggregated proportion of poor, middle and rich) of current use of any type of family planning method.

Operational definition

Family planning satisfaction

“satisfaction” with family planning methods is operationally defined as women’s self-reported contentment with their current contraceptive method, based on their responses in the 2021 PMA-ET survey, where those expressing approval were classified as satisfied, while those indicating dissatisfaction or uncertainty were considered unsatisfied.

Data management and statistical analysis

Spatial autocorrelation analysis

The spatial autocorrelation analysis aims to examine the degree of similarity between nearby locations in spatial data, helping identify spatial patterns and potential clustering, thus informing targeted interventions and policy decisions. The data underwent cleaning using STATA version 17.0 software and Microsoft Excel. For data analysis, Arc GIS 10.7 and SaTScan 9.6 were utilized. To assess the spatial distribution of satisfaction towards the current use of family planning methods in Ethiopia, the Global Moran’s I statistic was employed. A Moran’s I value nearing − 1 indicates dispersed satisfaction, close to + 1 suggests clustering and a value of zero signifies a random distribution, and if Moran’s I value zero shows randomly distributed and a statistically significant Moran’s I (p < 0.05) leads to rejection of the hypothesis [11].

Incremental autocorrelation analysis

To assess spatial autocorrelation across a range of distances, a line graph depicting these distances and their corresponding Z-scores was created. The Z-scores indicate both the degree of spatial clustering and statistical significance. Peaks in Z-scores highlight distances where spatial processes fostering clustering are most prominent. These peak distances serve as valuable references for tools incorporating Distance Band or Distance Radius parameters. Utilizing this information aids in selecting an optimal distance threshold or radius for tools that employ such parameters, including those used in hotspot analysis [12].

Hotspot and cold spot analysis

Hotspot and cold spot analysis, commonly used in spatial epidemiology and public health, helps identify areas with statistically significant clusters of high or low values, respectively, guiding resource allocation and intervention strategies for disease prevention and control. The study employed Gettis-Ord Gi* statistics to assess spatial autocorrelation differences in the study location for each area, particularly focusing on satisfaction with the current use of the family planning method. Gettis-Ord Gi* statistics were utilized to identify hotspot areas, signifying pronounced spatial clustering. The z-score was computed to validate the statistical significance of clustering, and the p-value was determined at p-value < 0.05 with 90%, 95%, and 99% confidence intervals. A z-score less than − 1.68, -1.96, and − 2.58 designates a cold spot, while a z-score greater than + 1.68, + 1.96, and + 2.58 indicates hotspot areas [3, 13, 14].

Spatial interpolation

The spatial interpolation technique is used to predict satisfaction with the current use of the family planning method for unsampled areas based on sampled clusters [15]. Deterministic and geostatistical interpolation methods were applied in this study. To evaluate these interpolation methods, a geostatistical analysis, identifying the technique with the lowest mean predicted error (MPE) and root mean square predicted error (RMSPE) as the most fitting for predicting satisfaction towards the current use of the family planning method. Smaller MPE and RMSPE values suggest a closer alignment between predicted and observed values, indicating the precision and efficacy of the selected interpolation technique [16].

Spatial scan statistics

This study employed Bernoulli-based spatial Kuldorff’s Scan statistics within SaTScan version 9.6.1 software to identify the geographical locations with statistically significant spatial windows for satisfaction towards the current use of the family planning method [17]. The outcome variable has a Bernoulli distribution, so the Bernoulli model was employed using Kuldorff’s method for purely spatial analysis.

The scanning window, moving across the study area, identified cases with satisfaction towards the current use of family planning method as well as controls dissatisfaction towards the current use of family planning method, fitting the Bernoulli model. The default maximum spatial cluster size, set at < 50% of the population, served as an upper limit. Identification of the most likely clusters relied on p-values and likelihood ratio tests derived from 999 Monte Carlo replications. Secondary clusters were generated using non-overlapping options in SaTScan version 9.6.1, and the mapping of clusters and attributes of satisfaction towards the current use of the family planning method, produced by SaTScan, was accomplished using ArcGIS software version 10.7.

Spatial regression

Exploratory Regression was employed to identify a model adhering to the assumptions of the Ordinary Least Squares (OLS) method, focusing on models with high Adjusted R2 values. The Ordinary Least Square regression model, being global, estimates a single coefficient per explanatory variable across the entire study region. We utilized explanatory regression to verify the assumptions of spatial regression, incorporating specific tests. The Jarque-Bera test assessed normality assumptions for residuals, and the statistically significant Koenker (BP) statistic indicated inconsistencies in the modeled relationships, possibly attributable to non-stationarity or heteroscedasticity. Multicollinearity, assessed through the Variance Inflation Factor, ensured the absence of redundancy among predictor variables, with coefficients displaying the expected sign and statistical significance, along with robust Adjusted R2 values.

A geographically weighted regression model gives local parameter estimates to reflect variations over space in the association between the current use of family planning methods and predictor variables [18]. The geographically weighted regression model utilized the aggregated proportion of satisfaction towards the current use of the family planning method and all relevant predictor variables for each cluster.

The evaluation of geographical heterogeneity for each coefficient involved comparing the AICc between the GWR model and the global OLS regression model. Model comparison, utilizing the corrected Akaike Information Criteria (AIC) and Adjusted R-squared, was performed for both the OLS (global model) and GWR (local) model. The determination of the best-fit model for local parameter estimates hinged on selecting the model with the lowest AICc value and a higher adjusted R-squared. The comparison between the OLS and GWR models was based on robust statistical metrics, including the corrected Akaike Information Criteria (AICc) and Adjusted R-squared. By evaluating geographical heterogeneity through these measures, we aimed to demonstrate how the GWR model captures spatial variability more effectively than the OLS model. The GWR model’s superiority lies in its ability to account for spatial heterogeneity by allowing coefficients to vary across different locations, unlike the OLS model, which assumes spatial homogeneity. By comparing AICc values and adjusted R-squared between the two models, we identified the GWR model as providing a better fit for capturing the spatial nuances inherent in our dataset.

Findings

Socio-demographic characteristics and proportion of satisfaction towards current use of family planning method

A weighted total of 1,456 reproductive age group women were included in the study. Approximately half of the participants, constituting 648 (44.50%) women, fell within the age range of 25–34 years. The majority of the participants resided in rural areas, accounting for 861 (59.15%). Regarding educational background, 610 (41.86%) had primary education, and rich (54.18%) belonged to the rich wealth status category. The majority of the study participants, 1,375 (94.46%) were married women. The overall proportion of satisfaction towards the current use of family planning methods among reproductive-age women in Ethiopia was 84.10% (82.12, 85.97) (Table 1).

Table 1 Socio-demographic characteristics and proportion of satisfaction towards current use of family planning method among reproductive-age women in Ethiopia using 2021 PMA-ET

Spatial analysis result

Spatial autocorrelation (Global Moran’s I) and incremental Spatial autocorrelation analysis

At a distance threshold of 229,712 m, statistically significant z-scores suggest that spatial factors contribute to clustering patterns. The incremental spatial autocorrelation analysis, beginning at 18,385.7 m, identified ten distance bands with significant clustering (Fig. 2). The spatial distribution of satisfaction with current family planning use among reproductive-age women demonstrated a non-random clustering pattern at the regional level. The Global Moran’s I index was 0.308358 (p-value < 0.001), indicating a moderate but statistically significant level of spatial clustering. The Z-score of 7.1 suggests that there is less than a 1% probability that the observed clustering occurred by random chance (Fig. 3).

Fig. 2

Incremental spatial autocorrelation of satisfaction towards current use of family planning method among reproductive-age women in Ethiopia using 2021 PMA-ET

Fig. 3

Spatial autocorrelation of satisfaction towards current use of family planning method among reproductive-age women in Ethiopia using 2021 PMA-ET

Hotspot and cold spot regions for satisfaction with current use of family planning methods

Hotspot analysis identifies statistically significant regions with either high (hot spots) or low (cold spots) levels of satisfaction with current family planning use. The results indicate that regions with higher satisfaction levels (hot spots) were detected in Addis Ababa, most parts of Amhara, and some areas of Benishangul-Gumuz. Conversely, lower satisfaction levels (cold spots) were found in Harari, Dire Dawa, parts of Oromia, sections of Gambela, and some areas within Benishangul-Gumuz and Somali regions (Fig. 4).

Fig. 4

Hot spot analysis of satisfaction towards current use of family planning method among reproductive-age women in Ethiopia using 2021 PMA-ET

Spatial interpolation

The Ordinary Kriging spatial interpolation method was used to predict satisfaction with the current use of family planning methods in unobserved areas, with mean predicted error (MPE) and Root Mean Square Predicted Error (RMSP) used for comparison. Among the interpolation methods evaluated, Disjunctive Kriging was identified as the most suitable, yielding the lowest MPE (-0.02669) and RMSP (0.56540). The analysis, based on the 2021 PMA-ET dataset, indicated regional variations in predicted satisfaction, with a transition from red (lower predicted satisfaction) to green (higher predicted satisfaction) areas (Table 2). Illustrate that Addis Ababa and some parts of the Amhara region were predicted as areas with higher satisfaction towards the current use of family planning methods compared to other regions (Fig. 5).

Fig. 5

Interpolation of satisfaction towards current use of family planning method use among Reproductive-age Women PMA-ET 2021

Table 2 Interpolation method comparison for satisfaction towards current use of family planning method among reproductive-age women PMA-ET 2021

Spatial scan statistics

The Kulldorff’s Spatial Scan analysis identified seven spatial clusters, of which two were statistically significant at p < 0.05. The primary cluster, represented by a green-colored ring, was predominantly located in central Ethiopia, covering Addis Ababa, the northern part of southwestern Ethiopia, and the eastern part of Gambela (Fig. 6). This spatial window, centered at 9.954596°N, 38.209724°E, had a 73.07 km radius, a Log-Likelihood Ratio (LLR) of 13.21, and a relative risk (RR) of 1.19 at p < 0.001, indicating that women within this area were 1.19 times more likely to report satisfaction with the current use of family planning methods compared to those outside the cluster. The secondary cluster, marked by a tourmaline blue ring, covered northern Addis Ababa and the southern part of the Amhara region, with a center at 8.011829°N, 35.638621°E, a 113.07 km radius, an LLR of 8.4, and an RR of 1.18 at p < 0.001, suggesting that women within this cluster had 1.18 times higher satisfaction than those outside it (Table 3).

Fig. 6

Spatial SaTScan analysis of satisfaction towards current use of family planning method among reproductive-age women, PMA-ET 2021

Table 3 Significant clusters of satisfaction towards current use of family planning method among reproductive-age women, PMA-ET 2021

Spatial regression analysis

The spatial regression analysis assessed factors influencing spatial variations in satisfaction with the current use of family planning methods using the Ordinary Least Squares (OLS) model, which explained 71.0% of the variation (Adjusted R² = 0.71), meeting all model assumptions. The coefficients indicated the strength and direction of each explanatory variable’s effect, with statistical significance confirmed using robust probability estimates. The Koenker (BP) statistic was significant (Koenker BP = 24.65, p < 0.001), suggesting that the relationships between explanatory variables and satisfaction with family planning use varied across locations. The Joint Wald statistic confirmed the overall model significance (p < 0.01), while Variance Inflation Factor (VIF) values < 7.5 indicated no multicollinearity issues. Additionally, the Jarque-Bera test was non-significant, confirming that model residuals followed a normal distribution. Given the significant Koenker test result, the analysis suggests that relationships between some explanatory variables and satisfaction with family planning use are non-stationary across different geographic areas, warranting the application of Geographically Weighted Regression (GWR) for a more localized assessment (Table 4).

Table 4 Global beta coefficients of the ordinary least square model summary and diagnostics for satisfaction towards current use of family planning method among reproductive-age women, PMA-ET 2021

Geographically weighted regression (GWR) analysis

The Geographically Weighted Regression (GWR) analysis identified women aged 15–24, poor wealth status, and urban residence as significant spatial predictors of satisfaction with the current use of family planning methods among reproductive-age women in Ethiopia. The effect of age (15–24 years) on satisfaction exhibited spatial variation, with coefficients ranging from 1.062880 to 1.062969, indicating a consistent positive influence across different geographic areas. Notably, regions such as Addis Ababa, Amhara, Oromia, SNNP, northwest Ethiopia, Gambela, and Afar demonstrated a particularly strong and positive association between younger women (15–24 years old) and higher satisfaction with family planning use (Fig. 7).

Fig. 7

Geographically varying values of coefficients per cluster for predictor women who had age 15–24, PMA-ET 2021

The Geographically Weighted Regression (GWR) analysis also revealed that the statistical significance of poor wealth status varied across different regions of Ethiopia. The coefficients associated with women with media exposure exhibited spatial variation, ranging from 0.733614 to 0.733722, indicating a positive spatial effect of poor wealth status on satisfaction with the current use of family planning methods. This relationship was particularly strong and positive in Oromia and some parts of the Afar region (Fig. 8).

Fig. 8

Geographically varying values of coefficients per cluster for predictor women had poor wealth status, PMA-ET 2021

The Geographically Weighted Regression (GWR) analysis further indicated that the statistical significance of women’s residence in urban areas varied across different regions of Ethiopia. The coefficients associated with urban residence exhibited spatial variation, ranging from 0.418157 to 0.418243, suggesting a positive spatial effect on satisfaction with the current use of family planning methods. This relationship was particularly strong and positive in Harari, Dire Dawa, and some parts of the Amhara and Oromia regions (Fig. 9).

In comparing the Ordinary Least Squares (OLS) and Geographically Weighted Regression (GWR) models, the corrected Akaike Information Criterion (AICc) and adjusted R² were used as evaluation metrics. Both models exhibited a similar AICc value of 953.11, indicating comparable model fit. However, the GWR model demonstrated a superior explanatory power, achieving an adjusted R² of 73.0%, compared to 71.0% in the OLS model (Table 4 and Table 5).

Fig. 9

Geographically varying values of coefficients per cluster for predictor women residence in urban, PMA-ET 2021

Table 5 Geographic weighted regression (GWR) model for satisfaction towards current use of family planning method among reproductive-age women, PMA-ET 2021

Discussion

This study aimed to explore the spatial variation and predictors of satisfaction with the current use of family planning methods among reproductive-age women in Ethiopia using data from the 2021 Performance Monitoring for Action Ethiopia (PMA-ET). The findings revealed that 84.10% (82.12–85.97) of reproductive-age women in Ethiopia were satisfied with their current family planning methods.). his result is consistent with a study conducted in Nigeria, which reported a similar satisfaction rate of 83.5% among women of reproductive age [19]. However, findings from Ghana indicated a lower satisfaction level of 71% [20]. while a study in the Philippines reported a significantly higher satisfaction rate of 92% [21]. These variations across countries may be attributed to differences in healthcare accessibility, quality of family planning services, cultural perceptions, and socioeconomic factors. For instance, Ghana’s lower satisfaction rate may reflect barriers to service access, limited contraceptive choices, or provider biases, whereas the Philippines’ higher satisfaction rate could result from well-established family planning programs, extensive community outreach, and improved service delivery. Unexpectedly, despite Ethiopia’s relatively lower contraceptive prevalence rate (36%) and persistent unmet need for family planning (22%), satisfaction levels were relatively high. This may indicate that women who do access family planning services generally find them acceptable. Additionally, variations in spatial clusters suggest that local healthcare infrastructure, provider-client interactions, and regional socio-cultural factors play a role in shaping women’s satisfaction levels, warranting further qualitative exploration.

The Getis-Ord Gi* hotspot analysis identified significant hotspots, indicating hotspot (high satisfaction) regions for satisfaction towards the current use of family planning method were detected in Addis Ababa, most parts of Amhara, and some parts of Benshangul-Gumz. The spatial Kuldorff’s Scan analysis is statistically significant at a P-value < 0.05 in Addis Ababa, the north part of southwestern Ethiopia, the south part of Amhara, and the east part of Gambela. A study conducted in Nigeria used spatial analysis to identify hotspots of contraceptive use and satisfaction with family planning services [22]. The study found significant hotspots of high satisfaction in urban areas, while rural areas showed lower satisfaction levels. This contrasts with the findings in Ethiopia, where hotspots were identified in both urban (Addis Ababa) and rural (Amhara and Benshangul-Gumz) areas. In another study conducted in Kenya, spatial analysis was used to assess the geographical disparities in satisfaction with family planning services. The results revealed hotspots of high satisfaction in the coastal region and parts of the Rift Valley [23]. This differs from the Ethiopian findings, where hotspots were identified in Addis Ababa and various regions across the country. The discrepancy in contraceptive use satisfaction between women of reproductive age may be due to variations in access to a wide range of contraceptive methods, quality of family planning services, cultural attitudes towards contraception, and the availability of information and education on contraceptive options.

The significant hotspots identified by the Getis-Ord Gi* and spatial Kuldorff’s Scan analyses offer insights into regional disparities in family planning satisfaction. Targeted interventions can prioritize areas with high satisfaction, like Addis Ababa and parts of Amhara, while addressing lower satisfaction levels in other regions, aiding policy development for equitable service delivery.

Furthermore, a global study employed spatial analysis to examine regional disparities in family planning satisfaction across multiple countries. The study identified hotspots of high satisfaction in urban areas of low-income countries, while rural areas showed lower satisfaction levels [24]. This aligns with the Ethiopian findings, where hotspots were detected in both urban and rural areas.

The geographically weighted regression model revealed that predictor variables women aged of 15–24, poor wealth status, and women who reside in urban areas were statistically significant predictors spatially for satisfaction towards the current use of family planning methods among reproductive-age women in Ethiopia. Previous studies conducted in various African countries have also explored the factors influencing satisfaction towards the current use of family planning methods among reproductive-age women. For instance, a study conducted in Nigeria found that age, education level, and socioeconomic status were significant predictors of satisfaction with family planning methods [25]. Similarly, a study in Kenya revealed that urban residence and higher education levels were associated with higher levels of satisfaction [26]. This consistency across different studies suggests that these factors play a crucial role in determining women’s satisfaction with family planning services. In Ethiopia’s socio-cultural context, women aged 15–24 may show higher satisfaction due to better exposure to education, health campaigns, and proactive use of modern family planning methods. Urban residents benefit from easier access to healthcare facilities, diverse contraceptive options, and community-level awareness initiatives. Conversely, women with poor wealth status may face challenges such as limited access to quality services or financial barriers, impacting their satisfaction. These findings reflect how socio-economic disparities and urban-rural divides influence family planning satisfaction in Ethiopia.

Strengths and limitations of study

This study’s strength lies in its use of geographically weighted regression (GWR), which provides detailed spatial insights into factors influencing family planning satisfaction across Ethiopia. The large sample size and nationwide scope enhance the generalizability of findings, while the identification of specific predictors offers valuable data for region-specific interventions. However, limitations include the cross-sectional nature of the study, which precludes causal inferences, and the exclusive focus on family planning users, excluding non-users who could provide insights into barriers to access and dissatisfaction. Additionally, reliance on secondary data limits the scope of examined variables, potentially omitting important cultural, provider-related, or behavioral factors. Spatial analysis assumptions and potential model limitations may also affect interpretation, requiring caution when generalizing results. Future research should consider longitudinal designs and a broader population to provide a more comprehensive understanding of family planning satisfaction dynamics in Ethiopia.

Practical policy implications

The findings highlight the need for region-specific policy interventions to improve family planning satisfaction in Ethiopia. Policymakers should focus on expanding access to quality family planning services in underserved areas, particularly for younger women, those with poor wealth status, and urban residents. Tailored outreach programs addressing financial and social barriers can enhance satisfaction and service utilization. Additionally, strengthening infrastructure and service delivery in identified hot spots, such as Addis Ababa, Amhara, and Benishangul-Gumuz, can sustain high satisfaction levels and guide similar efforts in other regions.

Conclusion

The study revealed that satisfaction with the current use of family planning methods among reproductive-age women in Ethiopia exhibits regional variations. Statistically significant clusters of higher satisfaction were identified in Addis Ababa, most parts of Amhara, and some parts of Benishangul-Gumuz. Geographically Weighted Regression (GWR) analysis indicated that women aged 15–24, those with poor wealth status, and urban residents were significant spatial predictors of satisfaction at the local level. These findings suggest the need for targeted regional strategies to improve satisfaction with family planning services. Interventions should prioritize high-satisfaction areas to sustain progress while addressing disparities in underserved regions. Policymakers should focus on enhancing equitable access, strengthening awareness campaigns, and mitigating socioeconomic barriers to improve satisfaction with family planning.