Quantitative study.

Study design: The data set in the current study is a subsample of all women who underwent CS (n = 516) within a hospital-based prospective cross-sectional study (n = 6658) conducted to investigate maternal and newborn outcomes among pregnant women in Somaliland described elsewhere [10].

Study population: All women who underwent CS at the national referral hospital between the 15^th of April 2019 and 30^th of March 2020 and whose performance of CS was majorly influenced by the family decision making process were included in the study. Informed by our previous study [5], family consent is a “collective decision-making process in which several decision makers in the extended family structure (father/mother, father/mother in laws, and husband to the woman giving birth) are involved with deliberations on need for CS before the male family member provides formal consent for CS to be performed”. Women whose CS decision was majorly influenced by healthcare providers or financial barriers were excluded. Similarly, women with missing information related to family consent were excluded from the study. The women included were divided into two groups for comparison: women with timely family consent and those with late family consent. Timely family consent for CS to be performed was defined as CS performed in less than 3 hours following consent provided by the family, whereas late family consent for CS to be performed was defined as CS performed in more than 3 hours following delayed consent from the family members. The cut-off of 3 hours was used based on the findings of our previous study that showed that CS delayed for more than 3 hours was associated with adverse maternal outcomes [12].

International quality standards generally recommend decision-to-incision intervals of 30–75 minutes for emergency cesarean sections [16]. However, such benchmarks are often unachievable in low-resource health systems such as Somaliland, where sociocultural factors notably the need for family consent and structural limitations, including staff shortages and constrained operating theatre capacity, contribute to longer delays [5]. In recognition of these realities, a 3-hour threshold was adopted as a context-specific and evidence-based benchmark, reflecting findings from our prior study in Somaliland [12], while acknowledging that shorter intervals remain optimal for maternal and perinatal safety [17].

Sample Size and Sampling: Using Cochran’s formula for sample size calculation, an assumed proportion of 50% (which maximizes variability), a 95% confidence level, and a 5% margin of error, the initial sample size was calculated to be 385. This ensures an adequate representation of the target population while preserving statistical validity. For this study, all the 516 participants were included, which exceeds the minimum required sample size, thereby enhancing the robustness and reliability of the findings.

Data collection: A data abstraction form, based on the validated World Health Organization Maternal Near-Miss tool, was used to collect data from medical records upon admission. This standardized tool contains predefined variables assessing sociodemographic factors, obstetric history, and factors associated with family consent [18]. The tool has been widely used in similar studies focusing on emergency obstetric care interventions such as cesarean section, ensuring its validity and reliability [19,20]. Women were followed until discharge, and data were systematically recorded to maintain consistency and accuracy. Trained data collectors comprising a team of midwives, nurses and doctors extracted data from the medical records to the data abstraction form. Periodic quality assurance checks were conducted to ensure that there was no missing data in the forms. For the missing data identified, the data was retrieved from the attached medical record. If data was missing in the medical record, the person responsible with providing care to that woman was contacted to provide the information. The first and fourth author, along with a team of researchers in the maternal health research group entered the collected data into Statistical Package of Social Sciences (SPSS) version 20.

Variables: The dependent variables were timely or late performance of CS mainly due to family consent. The independent variables were sociodemographic characteristics such as age, education, residence area, obstetric characteristics such as parity, previous history of CS, underlying causes, hypertensive disorders, obstetric haemorrhage, and contributory/associated factors such as anaemia or obstructed labour/prolonged labour.

Data analysis: Univariate, bivariate, and multivariable analyses were performed using SPSS. The Chi-square test was chosen to compare differences between the two groups (timely vs. late family consent for CS) as it is appropriate for categorical data. Bivariate logistic regression was conducted to identify variables significantly associated with timely or late family consent, as this method is suitable for assessing relationships between independent variables and a binary outcome [21]. To control for potential confounders, variables with a p-value of < 0.2 in the bivariate analysis were included in the multivariable model. This threshold is commonly used in regression modeling to ensure that important predictors are not excluded prematurely [22]. The Hosmer–Lemeshow goodness-of-fit test was used to evaluate model adequacy, yielding a p-value of 0.12, indicating that the model provided a good fit to the data (p > 0.05). To control for multiple comparisons, the Bonferroni correction was applied to adjust the significance threshold for the number of tests performed. Since 6 variables were tested, the corrected p-value threshold was set to p ≤ 0.05/ 6 = 0.0083. As such, factors associated with timely or late family consent were considered statistically significant at p ≤ 0.008. This adjustment helps reduce the risk of Type I errors due to multiple testing [23].

Qualitative study.

Study design: An exploratory study design [24] using in-depth interviews was applied in this study. In-depth interviews were used because they provide rich descriptive data and insight and enabled the authors to explore deeply the experiences and perspectives of women who underwent CS following timely or late family consent [24].

Study participants: Women who experienced maternal near miss and underwent CS at the national referral hospital between the 15^th of April 2019 and 30^th of March 2020 during the data collection for the quantitative part of the study were eligible to be included.

Sampling: Purposive sampling was used to select women who had experienced maternal near miss to give deeper understanding information about the phenomenon under investigation [25,26]. This method enabled us to obtain a study sample that provided rich data related to experiences, perspectives barriers, and facilitators of family consent for CS to be performed. A local midwife trained in data collection with the fourth author recruited mothers at the obstetrics department. The fourth author and the local midwife asked the women meeting the criteria of maternal near miss if they wanted to participate in the study. Women who accepted were followed up for in-depth interviews at the hospital. Data saturation was reached after 16 interviews, meaning no new themes or insights emerged with additional interviews. None of the women approached declined to participate. Overall, 16 women who underwent CS participated in the study. The sample was drawn from the quantitative study participants encompassing CS deliveries (n = 16).

Data collection: In-depth interviews were conducted in Somali by trained data collectors at the national referral hospital in Somaliland. A semi-structured interview guide was used to collect data. The interview guide contained questions regarding circumstances related to near-miss events, such as, “Please explain your situation and how you experienced this difficulty”; “How was the decision about CS made?”; “Were there delays in the family decision for CS?”; “Who provided the signature for CS?”; and “How long did it take?”. Probing questions were included on barriers and facilitators to family decision making for CS consent. The interviews took approximately 40 minutes and were recorded, transcribed, and translated into English.

Data analysis: The analysis of the data in this study followed a deductive thematic approach [26,27]. The preliminary analysis was discussed among the first, second, third and the last author. In the first phase, the first author read the transcripts several times to be familiarized with the data set. The second phase involved generation of codes using the deductive approach of coding the data text in relation to the barriers to and facilitators of family consent for CS to be performed. Examples are shown in Table 1. In the third phase, the codes were examined and re-examined to find repeated patterns, differences, and similarities; these were then organized into themes according to two thematic areas of barriers to and facilitators of family consent for CS to be performed. In the fourth phase, a further check was carried out by mapping the themes back onto the transcripts to ensure that they accurately reflected the original meaning of the text. The fifth and final phase of the process involved further defining and refining every theme and interpreting the data set beyond its original description. Throughout the phases of data analysis other researchers in this study were involved and discussed the analysis back and forth to reach consensus.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Example of the coding matrix used to identify subthemes according to themes on barriers and facilitators of timely family decision-making for CS to be performed.

https://doi.org/10.1371/journal.pone.0342475.t001

Ethical considerations: Permission to conduct the study was obtained from the Somaliland Ministry of Health Development. Ethical clearance was provided by the research ethics committee of the University of Hargeisa (Dr: CS/41/05/18). The researchers applied the ethical principles and guidelines for research of human subjects. All participants received oral information about the study and provided verbal consent to participate. The consent process was documented in each participant’s study record. This approach was approved by the University of Hargeisa research ethics commitee as part of the ethical review for the study, and all procedures followed were in line with the ethical guidelines and regulations established by the research ethics commitee. They were informed that they had the right to withdraw from the study at any time and that their participation in the study was voluntary. They were also informed that the data would be kept confidential.

Poor communication and understanding delays informed choice for CS.

In most of the interviews, women brought to the health facility for obstetric care and their family members initially refused CS. When healthcare providers informed them of the need for CS, it was revealed that they had misconceptions about and low understanding of CS. This was because they preferred a vaginal delivery. They pointed out that lack of adequate education about the indication for CS deterred them from accepting the intervention. The women interviewed expressed the need for effective communication between healthcare providers, family members, and the woman in need of a CS. In this regard, the participants’ accounts of delayed family consent related to inadequate information from healthcare providers and, consequently, a lack of understanding of the need for CS.

I believed if I underwent CS, my baby would die, I would have constant pain at the surgical site, back pain, and I might easily get an infection; then pain would follow me for the rest of my life, I would not be able to take care of my baby, my milk supply will be affected, and CS would limit my future pregnancies...at that moment, if I got before someone who educated me about CS, I could have accepted early because I realized that I harmed my life due to lack of information. (Interview 15, woman with eclampsia)

When asked about how consent was sought, many of the women expressed their frustration with not being adequately informed about how healthcare providers arrived at the decision about CS. They mentioned that doctors performed procedures such as obstetric ultrasound and made the decision about CS without informing the woman and her family members on the severe condition of the woman or the foetus. One woman said the following:

The doctors did not talk to me in a more respectful way, and when I repeated questions or when the family members wanted to make sure the intervention was necessary, the doctors did not answer the questions or re-explain these difficult things they were talking about… my family refused CS, so it took more than three hours, and lastly, they accepted when they realized that I had a very serious situation. (Interview 12, woman with ruptured uterus). Another woman said,

The doctor decided to perform CS, and actually I refused. I was angry about the ultrasound that found there was something wrong with my baby, but no one really explained to me what this meant, and I was worried the baby was already dead, or would it die during delivery? I had many unanswered questions, so that is why I refused the care. (Interview 8, woman with eclampsia)

In the quantitative results, the unadjusted odds ratio showed that women who had no formal education had 1.67 times higher odds of late family consent (crude odds ratio 1.67; 95% CI 1.17–2.39) compared to women with formal education. The adjusted odds ratio showed no statistical significance, indicating that no formal education was not an independent predictor of late family consent. However, the qualitative finding, suggests that a lack of education among women contributes to delayed family consent due to poor understanding and communication about the necessity of CS. The qualitative findings gave a deeper understanding of the complex situation of family consent and that other factors may cause the delay because provision of consent is in the hands of the family.

Differing understanding between family members on need for CS.

The quantitative results showed that women with obstetric haemorrhage had higher odds of a late family consent (aOR 3.21; 95% CI 1.16–8.89) compared to women without obstetric haemorrhage. Moreover, women with hypertensive disorders had higher odds of late family consent (aOR 8.49; 95% CI 1.08–66.99) compared to women without hypertensive disorders. Consent for CS was given late despite the maternal complications. This quantitative finding was consistent with qualitative results.

The participants who were interviewed pointed out that it was time consuming for families to consult on the care indicated by healthcare providers. Moreover, families argued with each other before agreeing for CS to be performed due to a lack of trust in the healthcare and the doctors’ judgement. Family disagreements were said to result due to differences in opinions between family members on what constituted the best care after doctors proposed that CS should be performed. Some of the family members preferred other modes of delivery rather than CS, even though their significant other had a severe maternal complication. One of the women said, “My father believed it was very important to give birth at home with a normal delivery, and hence, he didn’t want any more hospital involvement and treatment” (Interview 7, eclampsia). Some of the women interviewed mentioned that some of their family members held the view that CS was unnecessary and that the woman would be able to give birth normally. Other family members had accepted CS as a mode of birth that could resolve the woman’s complication. It took time for them to have discussions with the family members to agree on giving consent. One woman said,

Both my sister and my husband refused to give consent for Caesarean section since I had prolonged labour within two days. They believed that I could give birth naturally, so that he went to town and switched off his phone, and my mother called him many times, but he could not respond. Although he accepted finally, but it took to provide consent after 2 days to beg him, and later he consented, and the doctor performed CS. (Interview 14, woman with severe postpartum haemorrhage)

Nonetheless, some of the women stated that their family members understood the severity of their condition and agreed to provide consent. However, other family members refused to provide consent. They were not convinced that CS was the appropriate care. In this regard, differences in understanding between what families found to be the appropriate care resulted in delayed family decision-making.

My husband was ready to provide the signature, although we were shocked about this. However, my family refused to let my husband to sign because my family was not confident about the Caesarean section, and it took me to wait within two additional hours; after more consultation, my family allowed, and lastly, my husband signed. (Interview 9, woman with severe antepartum haemorrhage and postpartum sepsis)

One woman who experienced obstetric haemorrhage due to ruptured uterus said, “Midday of the 11^th day, the doctor decided on caesarean section, and around 6-7 hours, my family and relatives argued about the decision of my care, whether to agree to caesarean section or not, but lastly, they accepted when they realized my severity” (Interview 5, woman with ruptured uterus).

Another woman said,

Although my husband was ready to sign, my mother was very upset and refused him to sign for my CS; after two hours, my father convinced her that she should get out of this and accept his decision to sign as maybe otherwise I would die too like the baby. Then she accepted. (Interview 3, woman with severe antepartum haemorrhage and systemic infection)

The qualitative findings help explain the quantitative results, particularly the unexpected association between hypertensive disorders and late family consent. Although women with hypertensive complications require urgent care, the interviews revealed that family members often misunderstood the severity of such conditions and spent considerable time debating whether CS was necessary. These differing perceptions, coupled with mistrust in healthcare providers, contributed to delayed consent even in clinically urgent situations.

Absence of the person providing formal consent.

The qualitative finding on absence of the person giving consent was an additional finding to the quantitative results. This finding gave a deeper understanding of the barriers to timely family consent. The participants who were interviewed mentioned that when healthcare providers informed them that CS should be performed, the person with the authority to give consent was not available at the hospital to provide consent. For this reason, consent could not be given on time because it was time consuming for the person to come to the hospital to give consent.

The women interviewed stated that when the person who was supposed to provide consent was not available, it further delayed performance of CS because delays had already been experienced due to the initial refusal by the woman indicated for CS. In addition, when another family member was called in to provide consent because the husband was not available, the family member declined to provide consent. One woman said the following:

Two days later, I agreed, but my husband was still not found, and my brother-in-law was brought to the hospital in hopes that he could take responsibility and sign on his brother’s behalf, and also, he refused to sign the consent as he was worried his brother would get angry with him. My husband came and consented on the 5^th day after admission, and the Caesarean section was performed. (Interview 10, woman with eclampsia)

Really there was lot of delay because first i refused the c/section and other care, and it took to accept around ten hours; then, when i accepted, my husband was not available, so i called her father to sign, and when, he arrived he signed for my care. (Interview 8, woman with eclampsia)

Some of the other women interviewed explained that they had to contact their fathers, to provide consent for CS to be performed. However, when the family members were contacted, they initially refused to sign the consent for the operation to be performed. This contributed to further delays in timely provision of consent for CS. One of the women said, “My father gave signature to my operation because my husband was not available, although my father refused first, but later he obeyed...when the condition became serious and my baby died” (Interview 12, woman with ruptured uterus).

Husband’s autonomous consent about CS.

In the qualitative interviews, participants explained that in cases whereby the husband was alone when healthcare providers requested consent, he provided consent without delays. In these instances, the in-laws were not present at the healthcare facility. This enabled the husband to have full independence in making a timely decision for CS to be performed. This finding meant that when there was limited involvement of other family members in the decision-making process when CS was indicated, the husband provided consent early, without delays, and had full authority.

One woman said, “My husband signed because my family lived in Ethiopia and no one of my family could stay, so he had full authority to give the signature for my operation” (Interview 2, woman with severe antepartum haemorrhage).

In addition, some of the women stated that their husbands proceeded with giving consent immediately after informing the family that their wives were undergoing CS. In these cases, the process of seeking consent was not time consuming as the husband did not have extensive discussions with family members prior to giving consent. Husbands, in this regard, had more control over the decision-making about CS.

One woman said, “When it comes to the signature, my husband was with me and accounted for my care, and he suddenly told my family and continued to sign rapidly because I was severely sick and I had massive bleeding” (Interview 2, woman with severe antepartum haemorrhage).

The quantitative results also sought to investigate the role of place of residence in timely family consent about CS. The unadjusted odds ratio showed that women from rural areas had 0.47 lower odds of late family consent for CS (cOR 0.47; 95% CI 0.26–0.85) compared to women from urban areas. In the qualitative findings, some of the women who were brought to the health facilities from the rural areas came with their husbands. The presence of the husband only or fewer family members at the healthcare facility when consent was requested by healthcare providers contributed to timely family consent. One woman said, “I lived in Daami Village...but it took me to travel to the hospital around one hour. My husband and my friends stayed with me so that as soon as the doctor decided on an operation, he signed, and there was no delay” (Interview 6, eclampsia). Another woman said,

It was very long journey from Moolli village in Gashaamo near the border of Ethiopia, because I left home at 3am and came to Hargeisa at 4 pm since the car went straight because I was severely sick, and my husband encouraged the driver to speed up...my family and my husband provided the consent since my baby was dying; my family agreed, and consent was provided. (Interview 13, woman with severe antepartum haemorrhage)

However, the adjusted odds ratio showed no statistical significance between place of residence and late family consent for CS. The qualitative finding on fewer family members from rural areas during decision-making about CS gave a deeper understanding of the quantitative results on rural residence as a facilitator of timely family consent.

Adequate disclosure of all relevant information about CS.

When the family received more information regarding the possible risks and bad outcomes of not giving consent early, consent was given on time without delays. Participants also pointed out that when the responsible doctor explained and engaged in a discussion related to the care provided, there was timely family consent. One woman said, “Although my father first disagreed [about] my operation, after more consultation, he agreed” (Interview 7, woman with eclampsia).

The health care providers told him to sign consent for my care in order to save my life because my situation was very serious, and he did not have time to take other options since I had more bleeding, and they told him the only way to save my life was caesarean section. My baby died early as vaginal bleeding started, and also my situation was severe, and it was mandatory for my husband to consent to my operation. (Interview 2, woman with severe antepartum haemorrhage)

Some of the women who survived their complications and gave birth to a healthy baby following CS emphasized the importance of explaining adequately to family members about the complications that the women were having. One woman said, “Although I and my baby survived the problem, when women experience any difficulties, they might suffer many times because the relative did not know what happened or any complication that arose, so that it is better to tell the families what is the cause of complication” (Interview 14, woman with severe postpartum haemorrhage).

The qualitative finding on adequate disclosure of all relevant information about CS was not in the quantitative results. However, this finding provided a deeper understanding about what facilitates a timely family consent.

Previous CS.

The quantitative results showed that women who had had a previous CS had higher odds of a timely family consent CS (cOR 0.59; 95% CI 0.38–0.87) compared to women who had not had a previous CS. However, there was no statistical significance in the adjusted odds ratio. There was no qualitative finding that integrated with the finding about previous CS.