樱花视频

Skip to main content
Download PDF
  • Research
  • Published:

Predictors of pregnancy loss among urban and rural women aged 15 to 49 years in Pakistan

樱花视频 volume听25, Article听number:听950 (2025) Cite this article

Abstract

Background

The burden of pregnancy loss remains high in low- and middle-income countries like Pakistan. The Every Newborn Action Plan (ENAP) aims to decrease the stillbirth rate to 12 per 1000 total births by 2030, in every country. Current estimates indicate that Pakistan is unlikely to achieve this ENAP target, as the stillbirth rate stands at 30.6 per 1000 total births.

Methods

This study used the 2019 Pakistan Maternal Mortality Survey to identify the community-level, sociodemographic, maternal, environmental, and health services factors that are associated with pregnancy loss. Due to characteristic differences in urban and rural communities, separate analyses were carried out for ever-married women of 15 to 49 years. Mixed effects negative binomial regression was used to analyze the urban (n鈥�=鈥�5,887) and rural (n鈥�=鈥�7,136) samples of women who reported at least one pregnancy.

Results

The separate analyses found the factors associated with pregnancy loss to vary between urban and rural areas. In urban areas, pregnancy loss was associated with maternal education, maternal age, current marital status, and sanitation facility type. In rural areas, pregnancy loss was associated with region of residence, wealth index, maternal age, current marital status, drinking water source, cooking fuel type, and sanitation facility type.

Conclusion

This study carries significant implications for alleviating the burden of pregnancy loss in Pakistan, in line with ENAP objectives. The separate analyses provide a novel perspective regarding the factors influencing pregnancy loss in urban and rural areas, allowing for targeted interventions.

Peer Review reports

Background

In recent decades, strides have been made, globally, to reduce the burden of pregnancy loss. Pregnancy loss is a broad term used to describe any pregnancy that did not result in a live birth over the total length of the pregnancy, inclusive of miscarriages, stillbirths, and abortions [1,2,3]. The global stillbirth rate has reduced from 24.7 per 1000 total births in 2005 to 18.4 stillbirths per 1000 total births in 2015 [4]. The Every Newborn Action Plan (ENAP) aims to reduce this rate to 12 stillbirths per 1000 total births in every country, by 2030 [4]. Despite recent progress, the burden of pregnancy loss is still high in Low- and Middle-Income Countries (LMIC) where approximately 98% of stillbirths occur [5,6,7]. Recent estimates show that Pakistan has a stillbirth rate of 30.6 stillbirths per 1000 total births [8], indicating that the nation is not on track to meet its ENAP objectives [8, 9].

Several programs have been established by the government to improve pregnancy outcomes across the country [10]. In 2011, the constitution of Pakistan was amended to delegate healthcare administration to the provinces, allowing provincial governments to respond to healthcare demands of their communities [11,12,13]. Efforts have also been made to bolster the Lady Health Worker (LHW) Program. LHWs provide primary healthcare, obstetric care, and newborn care in their rural communities [10]. LHWs are a key pillar of healthcare delivery for many communities that otherwise would not have the resources to access such care [10, 14].

In addition to the strain on health systems, pregnancy loss also leads to social isolation as women are often blamed for the pregnancy loss, leading to dehumanization, humiliation, and even divorce [8, 15]. Therefore, understanding and reducing the burden of pregnancy loss not only eases the burden on the health system but also stands to improve the well-being of many women and their communities.

Existing literature primarily focuses on direct causes of pregnancy loss, and there is a lack of comprehensive studies examining predictors of pregnancy loss at a national scale. Moreover, the prevailing literature also presumes homogeneity between the urban and rural areas; however, recent studies have shown that estimates of pregnancy loss vary between urban and rural areas [2, 16], indicating the need for separate analyses. In addition, urban and rural communities differ in their characteristics, for example compared to urban regions, rural regions of Pakistan have poorer access to safe drinking water, lack of improved sanitation facilities, and inadequate healthcare [12]. There is also strong evidence in the literature regarding the impact of environmental factors on pregnancy loss [2, 17,18,19]. But very few studies in Pakistan have examined environmental factors at a national scale in the context of other factors.

This study aims to address these gaps by identifying the key predictors associated with pregnancy loss, among ever-married women aged 15鈥�49 years, separately for urban and rural areas of Pakistan, by using the latest national level data from 2019. This study will be one of the first to examine pregnancy loss factors, separately for urban and rural communities at a national level.

Methods

Data source

This study utilized the Pakistan Maternal Mortality Survey 2019 (PMMS) administered by the National Institute of Population Studies (NIPS) and the Inner City Fund (ICF) [20]. The PMMS implemented a multistage and multiphase cluster sampling design which utilized the 6th Population and Housing Census of 2017 [20]. The survey was designed to obtain samples from the urban and rural areas of Punjab, Sindh, Khyber Pakhtunkhwa, Balochistan, Azad Jammu and Kashmir, Gilgit Baltistan, Islamabad Capital Territory (sampled with Punjab), and Federally Administered Tribal Areas (sampled with Khyber Pakhtunkhwa), resulting in 12 sampling strata for the first stage [20]. Further details of the survey can be found at Pakistan Maternal Mortality Survey 2019 [20]. The final survey sample consisted of 15,143 ever-married women between the ages of 15 to 49 years [20]. The study sample consisted of 13,075 women who had reported at least one pregnancy.

Outcome

The total number of pregnancy losses, over the respondent鈥檚 life course, was the outcome variable, a count variable that ranged from 0 to 12.

Potential explanatory variables

The explanatory variables were selected based on existing literature regarding predictors of pregnancy loss in Pakistan [7,8,9, 15, 16, 21, 22]. This study also adapted the Mosley and Chen framework which is widely used to study child survival in LMIC [17, 23,24,25,26,27,28]. The framework consists of community level factors, socioeconomic factors, maternal factors, environmental factors, and health services factors as seen in Fig.听1.

Fig. 1
figure 1

Conceptual framework for factors associated with pregnancy loss in Pakistan. 鈥� Factors derived from Household Questionnaire which measures household factors (All members of the household will be assigned the same value). * Factors derived from the Services Questionnaire which measures access to care factors at a cluster level for rural areas (All members of the cluster will be assigned the same value)

Full size image

Region of residence consisted of Punjab, Sindh, Khyber Pakhtunkhwa, Balochistan, Gilgit Baltistan, and Azad Jammu and Kashmir. The wealth index variable was created by the DHS, and it categorized households into five quintiles from poorest to richest. Maternal education was categorized as per previous work from Pakistan [8] into the highest level of education that the respondent has completed, and the categories were 鈥渘o education鈥�, 鈥渓ess than 9 years鈥�, 鈥�9 to 12 years鈥�, and 鈥�13 or more years鈥� of education. Maternal age was maintained as a continuous variable ranging from 15 to 49 years. The currently married variable was a binary variable with 鈥淵es鈥� and 鈥淣o鈥� responses.

Drinking water source was categorized into improved (any piped water, tube well or borehole, protected well or spring, rainwater, cart with small tank, tanker truck, or bottled water) and unimproved (unprotected well or spring, surface water, or other) sources as per DHS guidelines [29]. Type of cooking fuel was categorized into solid fuels (coal or lignite, charcoal, wood, straw/shrubs/grass, agricultural crop, animal dung, kerosene, or other) or clean fuels (electricity, LPG, natural gas, or biogas) as per DHS guidelines [29, 30], and kerosene was recoded as a solid fuel since it is considered a polluting fuel [30]. Sanitation facility type was also recoded as per DHS guidelines with categories of improved (flush toilet to sewer, flush toilet to septic tank, flush toilet to pit latrine, flush toilet to don鈥檛 know where, ventilated pit latrine, improved pit latrine, pit latrine with slab, and composting toilet), unimproved (flush toilet to somewhere else, pit latrine without slab/open pit, bucket toilet, hanging toilet/hanging latrine, and other), and open defecation [29]. Time to nearest health facility was categorized based on previous findings [7] as 鈥�10 minutes or less鈥�, 鈥�11鈥�30 minutes鈥�, and 鈥�31 minutes or more鈥� and 鈥渄on鈥檛 know鈥� was recoded to missing. LHW in village was a binary variable with responses of 鈥淵es鈥� and 鈥淣o鈥�, with 鈥淒on鈥檛 know鈥� being recoded to missing.

Statistical analysis

All statistical procedures were undertaken with STATA 17 statistical software. The dataset was split into urban and rural samples for analyses. In the univariate analysis, mean and standard deviation was obtained for the continuous variables and the frequency of distribution was obtained for the categorical variables. For the bivariate and multivariate analyses, mixed effects negative binomial regression was used to obtain Incidence Rate Ratios (IRR) for pregnancy loss and respective explanatory variables, in the urban and rural samples. Both analyses adjusted for the clustering effect of the 鈥渃luster鈥� variable (qhclust), where households within each cluster were interviewed [20]. A 5% significance level for the variables were considered as statistically significant. Collinearity among the variables was also assessed for the urban and rural models. The unadjusted and adjusted IRR with 95% confidence intervals have been reported. All analyses were conducted without weights.

Results

The univariate characteristics of the urban and rural areas are outlined in Table听1.

Table 1 Univariate analysis of pregnancy loss predictors stratified by urban and rural areas
Full size table

Pregnancy loss incidence was higher in urban areas (0.59鈥壜扁��0.98), when compared to rural areas (0.54鈥壜扁��0.96). Over 60% of women in urban areas were found to be in the highest two wealth quintiles whereas nearly 60% of rural women were in the lowest two wealth quintiles. Similarly with education, urban women had higher education levels compared to rural women. Compared to urban women, a greater percentage of rural women reported access to unimproved sources of drinking water, 2.6% and 8.9% respectively. Regarding cooking fuels, while 74.0% of urban women reported using clean fuels, only 23.6% rural women reported using clean fuels. As per access to sanitation facility, 93.1% of urban women report using improved sanitation facility, while only 72.4% of rural women report using improved sanitation facility. Further, 21.6% of rural women report open defecation while only 2.0% of urban women report open defecation. Majority of rural women reside in a village that is 31听min or more from the nearest health facility. LHWs are present in only 56.4% of rural villages.

Table听2 describes the bivariate and multivariate results for the urban areas. The multivariate results find maternal education, maternal age, current marital status, and type of sanitation facility to be associated with pregnancy loss in urban areas. Women with less than 9 years of education had an adjusted pregnancy loss IRR that was 18% higher (IRR: 1.18; 95% CI: 1.06, 1.32) than women with no education. Each year of increase in age was associated with an adjusted pregnancy loss IRR of 1.03 (95% CI: 1.02, 1.04). Compared to currently married women, those who reported not being currently married had an adjusted pregnancy loss IRR that was 38% lower (IRR: 0.62; 95% CI: 0.49, 0.78). Women who reported accessing unimproved sanitation facilities had a 35% higher (IRR: 1.35; 95% CI: 1.11, 1.64) pregnancy loss incidence when compared to women who accessed improved sanitation facilities. The conditional alpha value is greater than 0, an indication that the mixed effects negative binomial model is better fitting due to overdispersion in the outcome variable. The cluster variable has an associated coefficient of 0.09, indicating that approximately 9% of the variation in pregnancy loss incidence can be attributed to clusters.

Table 2 Bivariate and multivariate analyses of pregnancy loss predictors among urban women
Full size table

Table听3 outlines the results from the bivariate and multivariate analyses for the rural areas. In rural areas, the multivariate analysis revealed region of residence, wealth index, maternal age, current marital status, drinking water source, cooking fuel type, and type of sanitation facility to be associated with pregnancy loss incidence. Compared to residing in rural Punjab, residing in rural Khyber Pakhtunkhwa was associated with an adjusted incidence of pregnancy loss that was 14% lower (IRR: 0.86; 95% CI: 0.74, 0.99), rural Balochistan with 24% lower (IRR: 0.76; 95% CI: 0.63, 0.91), and residing in rural Gilgit-Baltistan with a pregnancy loss incidence 19% lower (IRR: 0.81; 95% CI: 0.67, 0.97). Women from the highest wealth quintile had an adjusted pregnancy loss IRR of 0.73 (95% CI: 0.57, 0.93) when compared to women from the lowest wealth quintile. Each year of increase in age was associated with an adjusted pregnancy loss IRR of 1.03 (95% CI: 1.03, 1.04). Women who reported not being currently married had a 24% lower (IRR: 0.76; 95% CI: 0.61, 0.93) pregnancy loss incidence when compared to those who reported being currently married. Accessing unimproved source of drinking water over improved was associated with a 17% higher (IRR: 1.17; 95% CI: 1.01, 1.37) adjusted pregnancy loss incidence. Compared to cooking with clean fuels, cooking with solid fuels was associated with an adjusted pregnancy loss IRR of 0.85 (95% CI: 0.75, 0.97). Compared to women using improved sanitation facilities, women who reported open defecation had an adjusted pregnancy loss incidence, 19% higher (IRR: 1.19; 95% CI: 1.05, 1.36). The conditional alpha value being greater than 0 indicates that the mixed effects negative binomial model is better fitting due to overdispersion in the outcome variable. The cluster variable coefficient of 0.10 indicates that approximately 10% of the variation in pregnancy loss incidence can be attributed to cluster variations.

Table 3 Bivariate and multivariate analyses of pregnancy loss predictors among rural women
Full size table

Discussion

This study identified distant factors associated with pregnancy loss among ever-married women in urban and rural areas of Pakistan. The prevalence of pregnancy loss was higher in urban areas than rural areas. Since the PMMS dataset lacks primary sampling unit and strata variables, the prevalence estimates are not directly comparable to existing literature, which typically account for sampling design. The split analyses of urban and rural areas revealed that the factors associated with pregnancy loss differed between these contexts.

Consistent with the existing literature, rural areas of Khyber Pakhtunkhwa, Balochistan, and Gilgit-Baltistan had lower pregnancy loss incidence when compared to rural Punjab [8]. The poor pregnancy outcomes in rural Punjab have been attributed to the inadequate government support for healthcare, underscoring the critical role of healthcare infrastructure in determining pregnancy loss [8]. Moreover, the compounding effect of limited resources in these regions further accentuates the challenges faced by pregnant women in these regions, within the broader context of a resource-constrained country. Rural region-specific policies and interventions are necessary to support regions with poorer pregnancy outcomes.

In line with previous research, the current study found that rural women from the highest wealth quintile had better pregnancy outcomes than rural women from the lowest wealth quintile [7,8,9, 16]. Lower wealth status has consistently emerged as a strong predictor of increased pregnancy loss incidence, attributed in part to the financial constraints that may deter healthcare utilization among households with limited resources [16]. However, intriguingly, this protective effect of increased wealth index was not observed in urban areas. The concentration of quality healthcare services in urban settings may potentially mitigate the impact of wealth on pregnancy loss, as relative accessibility to quality healthcare may mean poorer women are still able to access services [12, 31].

Maternal education exhibited a negative association with pregnancy loss in the bivariate analyses, however after adjusting for other factors, such a protective effect was not observed. Rather, among urban women with less than 9 years of education, an increase in pregnancy loss incidence was observed when compared to women with no education. This is contradictory with existing literature where maternal education has been shown to be protective against poor pregnancy outcomes [8, 9], specifically due to better health seeking behaviour [8, 16]. But it is also reported that uneducated mothers might underreport pregnancy loss during the interview due to social taboos and fear of stigmatization, potentially misrepresenting the true impact of maternal education on pregnancy loss [8].

Consistent with literature, among urban and rural women, age was found to be positively associated with pregnancy loss as increased age is associated with health complications but also more pregnancies over the lifespan [7, 9, 16, 19]. Currently not being married was identified as protective since these women are no longer 鈥渁t risk鈥� of getting pregnant; specifically in a conservative society where marital status can be a proxy for sexual activity leading to pregnancies.

The analyses revealed that environmental factors had a significant influence, particularly in rural areas. Accessing unimproved sources of drinking water was associated with pregnancy loss among rural women and similar findings have been reported in the literature from South Africa [18]. Existing research has linked air particulate matter exposure to an increased risk of pregnancy loss [2, 19]. The current study employed 鈥渃ooking fuel type鈥� as a proxy to measure air particulate matter exposure and found a protective effect between using solid fuels and pregnancy loss in rural areas. This unexpected finding necessitates further investigation, and one potential explanation could be that the variable may not directly capture air particulate matter exposure as previous researchers have directly asked respondents about ventilated cooking facilities [19]. Exposure to poor sanitation facilities were associated with pregnancy loss in both urban and rural areas. This finding aligns with research conducted in Nepal, where utilization of poor sanitation facilities has been linked to pregnancy loss [17, 32]. Utilizing poor sanitation facilities increases the risk of potential exposure to bacterial infections, consequently contributing to pregnancy loss.

None of the health services factors were found to be statistically significant despite previous qualitative works having identified a link with pregnancy loss [15, 22]. The current study used health services factors that were captured at the cluster level and then linked to individuals, primarily capturing the availability aspect of health services in rural areas rather than individual service utilization. Further, disparities in health service utilization may be indirectly captured through the sociodemographic factors within the model, as studies have shown associations between health service utilization and socioeconomic status [8, 16].

The findings of this study carry significant implications for reducing the burden of pregnancy loss in Pakistan, aligning with the ENAP objective of achieving 12 stillbirths per 1000 births by 2030 [4]. Our findings can enable policymakers to adopt a targeted strategy to address the unique challenges faced by urban and rural communities. For example, rural Punjab was found to be underperforming in pregnancy outcomes compared to other rural regions, but no significant differences were seen between the urban parts of the regions. The exemplary rural regions with lower pregnancy loss incidence rates can serve as potential models, prompting further research to understand the differences in healthcare administration leading to the protective effects.

In rural areas, it appears that financial hardships may be contributing to the burden of pregnancy loss. Policies targeted towards reducing the financial barriers to accessing healthcare, specifically in rural areas, would prove beneficial. Policies should be targeted towards improving or at least offsetting the wealth disparities within communities. This could be carried out by bolstering and increasing support for the LHW Program which is responsible for administering primary healthcare services to rural areas. Previous researchers have shown that the LHW Program can be leveraged to increase health service utilization and accessibility in rural areas [14]. Thus, it stands to reason that investments in expanding the LHW Program would help reduce many barriers currently faced by the rural communities in Pakistan.

Limitations

When interpreting the findings from this study, some limitations should be considered. Due to the use of cross-sectional survey data, causality cannot be established between pregnancy loss and the explanatory variables. As alluded to earlier, social desirability bias is another concern with sensitive topics such as pregnancy loss. Due to fears of stigmatization, women may opt to misrepresent their pregnancy loss. In addition, the survey question regarding pregnancy loss did not differentiate between intentional and unintentional pregnancy loss. So, some pregnancies could have been intentional. Also, the survey was only administered to ever-married women; thus, these findings may not be generalizable to never-married women. Further, due to unavailability of sampling design variables, unweighted analyses were conducted which limit comparability of prevalence rates with other datasets. Lastly, health services factors were only available at a cluster level within the dataset, and individual level health service utilization could not be measured.

Conclusion

The findings from this study provides policymakers with a unique perspective regarding the factors impacting pregnancy loss in Pakistan. By conducting a separate analysis for urban and rural areas, it revealed unique factors that impact specific communities. Given this nuanced understanding, targeted policies and interventions can be implemented to aid in achieving the ENAP goal of 12 stillbirths per 1000 total births by 2030.

Data availability

The PMMS data is publicly available, by request, from the online DHS Program website (). The first author requested access and was granted permission to use the PMMS data. The authors are not permitted to share the dataset directly.

Abbreviations

DHS:

Demographic and Health Survey

ENAP:

Every Newborn Action Plan

ICF:

Inner City Fund

IRR:

Incidence Rate Ratio

LHW:

Lady Health Worker

LMIC:

Low- and Middle-income Country

NIPS:

National Institute of Population Studies

PMMS:

Pakistan Maternal Mortality Survey

References

  1. Lakew D, Tesfaye D, Mekonnen H. Determinants of stillbirth among women deliveries at Amhara region, Ethiopia. 樱花视频 Pregnancy Childbirth. 2017;17(1):375.

    听 听 听 听

  2. Xue T, Guan T, Geng G, Zhang Q, Zhao Y, Zhu T. Estimation of pregnancy losses attributable to exposure to ambient fine particles in South Asia: an epidemiological case-control study. Lancet Planet Health. 2021;5(1):e15鈥�24.

    听 听 听

  3. Zakar R, Nasrullah M, Zakar MZ, Ali H. The association of intimate partner violence with unintended pregnancy and pregnancy loss in Pakistan. Int J Gynecol Obstet. 2016;133(1):26鈥�31.

    听 听

  4. Lawn JE, Blencowe H, Waiswa P, Amouzou A, Mathers C, Hogan D, et al. Stillbirths: rates, risk factors, and acceleration towards 2030. Lancet. 2016;387(10018):587鈥�603.

    听 听 听

  5. Christou A, Dibley MJ, Raynes-Greenow C. Beyond counting stillbirths to Understanding their determinants in low- and middle-income countries: a systematic assessment of stillbirth data availability in household surveys. Trop Med Int Health. 2017;22(3):294鈥�311.

    听 听 听

  6. McClure EM, Saleem S, Goudar SS, Garces A, Whitworth R, Esamai F, et al. Stillbirth 2010鈥�2018: a prospective, population-based, multi-country study from the global network. Reprod Health. 2020;17(2):146.

    听 听 听 听

  7. Poudel S, Ghimire PR, Upadhaya N, Rawal L. Factors associated with stillbirth in selected countries of South Asia: A systematic review of observational studies. PLoS ONE. 2020;15(9):e0238938.

    听 听 听 CAS听 听

  8. Shakeel A, Kamal A, Ijaz M, Siddiqa M, Tesema GA, Abushal T. Trends and risk factors of stillbirth among women of reproductive age in Pakistan: A multivariate decomposition analysis. Front Public Health. 2023 [cited 2024 Apr 19];11. Available from:

  9. Afshan K, Narjis G, Qayyum M. Risk factors and causes of stillbirths among pregnant women in Pakistan. Afr Health Sci. 2019;19(1):1507鈥�16.

    听 听 听 听

  10. Aziz A, Saleem S, Nolen TL, Pradhan NA, McClure EM, Jessani S, et al. Why are the Pakistani maternal, fetal and newborn outcomes so poor compared to other low and middle-income countries? Reprod Health. 2020;17(3):190.

    听 听 听 听

  11. Hassan A, Mahmood K, Bukhsh HA. Healthcare system of Pakistan. Int J Adv Res Publ. 2017;1(4):170鈥�3.

    听

  12. Nishtar S, Boerma T, Amjad S, Alam AY, Khalid F, Haq Iul, et al. Pakistan鈥檚 health system: performance and prospects after the 18th constitutional amendment. Lancet. 2013;381(9884):2193鈥�206.

    听 听 听

  13. Pakistan Bureau of Statistics. National Health Accounts-Pakistan 2019-20. Government of Pakistan. 2022. Available from:

  14. Bhutta ZA, Soofi S, Cousens S, Mohammad S, Memon ZA, Ali I, et al. Improvement of perinatal and newborn care in rural Pakistan through community-based strategies: a cluster-randomised effectiveness trial. Lancet. 2011;377(9763):403鈥�12.

    听 听 听

  15. Zakar MZ, Zakar R, Mustafa M, Jalil A, Fischer F. Underreporting of stillbirths in Pakistan: perspectives of the parents, community and healthcare providers. 樱花视频 Pregnancy Childbirth. 2018;18(1):302.

    听 听 听 听

  16. Shakeel A, Kamal A, Tesema GA, Siddiqa M. Analysis of Spatial patterns and associated factors of stillbirth in Pakistan, PDHS (2017鈥�18): A Spatial and multilevel analysis. J Stat Theory Pract. 2023;17(1):19.

    听 听

  17. Ghimire PR, Agho KE, Renzaho AMN, Nisha MK, Dibley M, Raynes-Greenow C. Factors associated with perinatal mortality in Nepal: evidence from Nepal demographic and health survey 2001鈥�2016. 樱花视频 Pregnancy Childbirth. 2019;19(1):88.

    听 听 听 听

  18. Akombi BJ, Ghimire PR, Agho KE, Renzaho AM. Stillbirth in the African great lakes region: A pooled analysis of demographic and health surveys. PLoS ONE. 2018;13(8):e0202603.

    听 听 听 听

  19. Rozi S, Butt ZA, Zahid N, Wasim S, Shafique K. Association of tobacco use and other determinants with pregnancy outcomes: a multicentre hospital-based case鈥揷ontrol study in Karachi, Pakistan. BMJ Open. 2016;6(9):e012045.

    听 听 听 听

  20. National Institute of Population Studies (NIPS) [Pakistan], ICF. Pakistan Maternal Mortality Survey 2019. Islamabad, Pakistan, and Rockville, Maryland, USA: NIPS and ICF. 2020. Available from:

  21. Ahmed J, Alam A, Khokhar S, Khowaja S, Kumar R, Greenow CR. Exploring women鈥檚 experience of healthcare use during pregnancy and childbirth to understand factors contributing to perinatal deaths in Pakistan: A qualitative study. PLoS ONE. 2020;15(5):e0232823.

    听 听 听 CAS听 听

  22. Mcnojia SZ, Saleem S, Feroz A, Khan KS, Naqvi F, Tikmani SS, et al. Exploring women and traditional birth attendants鈥� perceptions and experiences of stillbirths in district Thatta, Sindh, Pakistan: a qualitative study. Reprod Health. 2020;17(1):3.

    听 听 听 听

  23. Ezeh OK, Agho KE, Dibley MJ, Hall JJ, Page AN. Risk factors for postneonatal, infant, child and under-5 mortality in Nigeria: a pooled cross-sectional analysis. BMJ Open. 2015;5(3):e006779.

    听 听 听 听

  24. Lu SSM, Stewart Williams J, Sommar JN. Inequalities in early childhood mortality in Myanmar - Association between parents鈥� socioeconomic status and early childhood mortality. Glob Health Action. 2019;12(1):1603516.

    听 听 听 听

  25. Kibret GD, Demant D, Hayen A. Bayesian Spatial analysis of factors influencing neonatal mortality and its geographic variation in Ethiopia. PLoS ONE. 2022;17(7):e0270879.

    听 听 听 CAS听 听

  26. Kibria GMA, Burrowes V, Choudhury A, Sharmeen A, Ghosh S, Mahmud A, et al. Determinants of early neonatal mortality in Afghanistan: an analysis of the demographic and health survey 2015. Glob Health. 2018;14(1):47.

    听 听

  27. Nasejje JB, Mbuvha R, Mwambi H. Use of a deep learning and random forest approach to track changes in the predictive nature of socioeconomic drivers of under-5 mortality rates in sub-Saharan Africa. BMJ Open. 2022;12(2):e049786.

    听 听 听 听

  28. Mosley WH, Chen LC. An analytical framework for the study of child survival in developing countries. Popul Dev Rev. 1984;10:25鈥�45.

    听 听

  29. Croft TN, Marshall AMJ, Allen CK, et al. Guide to DHS statistics. Rockville, Maryland, USA: ICF; 2018.

    听

  30. Arku RE, Brauer M, Duong M, Wei L, Hu B, Ah Tse L, et al. Adverse health impacts of cooking with kerosene: A multi-country analysis within the prospective urban and rural epidemiology study. Environ Res. 2020;188:109851.

    听 听 听 CAS听 听

  31. Ansari RM, Ansari Y, Ansari SY. Status of general practice and challenges to healthcare system of Pakistan. Open J Prev Med. 2015;5(12):463鈥�7.

    听 听

  32. Ghimire PR, Agho KE, Renzaho A, Christou A, Nisha MK, Dibley M, et al. Socio-economic predictors of stillbirths in Nepal (2001鈥�2011). PLoS ONE. 2017;12(7):e0181332.

    听 听 听 听

Acknowledgements

Not applicable.

Funding

Not applicable.

Author information

Authors and Affiliations

  1. Department of Epidemiology and Biostatistics, Western University, London, ON, Canada

    Kevin George Samuel,听Ngianga-Bakwin Kandala,听Bridget L. Ryan听&听Amardeep Thind

  2. Departments of Family Medicine and Medicine, Schulich School of Medicine and Dentistry, Western University, London, ON, Canada

    Bridget L. Ryan听&听Amardeep Thind

  3. Interfaculty Program in Public Health, Western University, London, ON, Canada

    Amardeep Thind

Authors
  1. Kevin George Samuel

    You can also search for this author in听

  2. Ngianga-Bakwin Kandala

    You can also search for this author in听

  3. Bridget L. Ryan

    You can also search for this author in听

  4. Amardeep Thind

    You can also search for this author in听

Contributions

KGS: Conceptualization, methodology, data analysis, interpretation, prepared the main manuscript, and editing. AT: Conceptualization, methodology, data analysis, interpretation, revisions, editing, and supervision. N-BK: Conceptualization, methodology, data analysis, interpretation, revisions, and editing. BLR: Conceptualization, methodology, data analysis, interpretation, revisions, and editing. All authors have read and approved the manuscript.

Corresponding author

Correspondence to Kevin George Samuel.

Ethics declarations

Ethics approval and consent to participate

Respondent consent was obtained for all data collected by the PMMS. Permission to use this data was obtained from the DHS Program website. Since this study uses publicly available data with anonymized responses, it was exempt from the Research Ethics Board review at Western University. Further details regarding DHS privacy procedures can be found at .

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Additional information

Publisher鈥檚 note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Rights and permissions

Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article鈥檚 Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article鈥檚 Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit .

About this article

Cite this article

Samuel, K.G., Kandala, NB., Ryan, B.L. et al. Predictors of pregnancy loss among urban and rural women aged 15 to 49 years in Pakistan. 樱花视频 25, 950 (2025). https://doi.org/10.1186/s12889-025-22165-w

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s12889-025-22165-w

Keywords

樱花视频

ISSN: 1471-2458

Contact us