The use of a mobile obstetric emergency system to improve obstetric referrals in Bong County, Liberia: A pre‐post study

Liberia experiences an unmet need for cesarean section with about 5% population coverage, lower than 9%–19% coverage associated with improved maternal and newborn outcomes. Delays in the referral process for comprehensive emergency obstetric and newborn care (CEmONC) services due to ineffective communication between a rural health facility (RHF) and a district hospital contribute to the low CS rate. This study examined the association between mobile obstetric emergency system (MORES) implementation and referral time for obstetric emergencies as well as maternal/newborn outcomes.


| INTRODUC TI ON
Countries in sub-Saharan Africa (SSA) account for 2/3 of the world's maternal deaths. 1 Delays that contribute to these maternal death and injuries are often categorized using the "three delays" framework. 2The first delay refers to the delay in deciding to seek care, the second in presenting to the health center and the third in receiving care once the woman arrives at the health facility. 2However, additional delays can further occur when a woman needs to be referred from a rural health facility (RHF) to a district hospital for emergency obstetric care services such as cesarean section (CS). 3,4ral health facilities (RHFs) in African countries often have the capacity to perform basic emergency obstetric and newborn care (BEmONC), which includes parenteral antibiotics, parenteral uterotonic, and parenteral anticonvulsants administration, manual removal of retained placenta, vacuum aspiration, assisted vaginal delivery, and basic neonatal resuscitation. 5Alternatively, district hospitals offer comprehensive emergency obstetric and newborn care (CEmONC), which includes blood transfusion and CS in addition to the seven BEmONC functions. 5Hence, if a woman shows indications for a blood transfusion or a CS, healthcare providers at the RHF should refer the woman to a district hospital for additional care. 4,5In the process of a referral, the same three delays can occur.
According to a verbal autopsy conducted in Bong County Liberia, ineffective communication between RHFs and hospitals is a contextual cause contributing to preventable maternal deaths. 6Having no standardized referral process for communication of important information as well as the lack of feedback once the patient is referred to the hospital were identified as ineffective communication. 6A study conducted in Nigeria found that women who were referred from a RHF were three times more likely to travel longer than 60 min to get to a hospital compared to women who went directly to a hospital. 7Furthermore, a study conducted in Rwanda found that longer travel time from RHF to a hospital was significantly associated with adverse neonatal outcome, emphasizing the need for strategies to reduce the transfer delay from health centers to district hospitals. 4spite this need, only 40% of RHFs in Liberia were ready to make an emergency referral, defined as having access to a functional ambulance or other vehicle stationed at the facility or access to an ambulance and a functioning telephone, either a landline or a mobile phone. 8Hence, there is a need for an efficient and effective communication mechanism between RHFs and hospitals.
0][11] Building upon these studies, this study piloted a mobile obstetric emergency system (MORES) using the free WhatsApp platform as an obstetric referral intervention examining the association between the implementation of the MORES intervention and transfer times, maternal outcome, and newborn outcomes.

| Design
This study was a pre-post descriptive study.We collected data from 20 RHFs and two district hospitals.We then merged the two data sets to examine the number of obstetric referrals from RHF to hospitals to identify common reasons for referrals, maternal and newborn

| Study setting
Liberia is a West African country with a maternal mortality ratio of 662 deaths per 100 000 live births, ninth highest globally. 2Bong County is the third most populous county in Liberia, with a population of approximately 329 000. 12The most recent Demographic and Health Survey of Liberia revealed 84% of women delivered with a skilled provider, up from 61% in 2013. 13Bong County reports 88% of pregnant women receiving four or more antenatal care visits and 86.8% delivering with a skilled provider.In 2019, there were 15231 facility births attended by a skilled provider in Bong County. 13Of these, 12,362 (81%) were delivered at a RHF with a midwife or nurse.Bong County was selected as the study setting based on the long-lasting relationship between and established work by the Bong County health team and the academic research team.

| MORES intervention
Following baseline collection of referral data, a two-day interactive obstetric triage and MORES referral training was conducted with the nurses and midwives at 20 RHFs and two district hospitals in Bong County, Liberia.The training focused on recognition and management of acutely ill pregnant woman, managing obstetric emergencies, and prioritization of care.The healthcare providers were provided with an overview of MORES, assigned unique ID, and messaging templates for RHF referrals to the district hospitals (Figure 1) and templates for received hospital referrals back to the RHF provider were designed to ensure bidirectional communication (Figure 2).Templates for referrals from providers at RHFs included information such as patient initials, reasons for referral, time of referral, type of transportation, and any other information related to patient condition.Templates for providers at the district hospitals to communicate back to RHFs providers included information such as unique ID, patient initials, referral received confirmation, patient's arrival time and maternal and newborn outcomes.Completed templates were shared only within the individual WhatsApp group between each RHF and district hospitals.A study designated smart phone was provided to the district hospitals and a small fee to cover data fee for the RHF staff as they used their personal device for referrals.Prior to the implementation of the MORES intervention, referrals were handwritten at the RHF and carried by the women's family to the district hospital.Furthermore, these handwritten referrals would often get misplaced and forgotten, causing the necessary information to never reach the hospital.

| Data collection and procedures
Research assistants (RAs) from the Bong County health team obtained permission from the RHFs and district hospitals to conduct chart reviews and collect transfer data.A retrospective review of the hospital labor and delivery logs for baseline referrals sent from the 20 RHFs to the two district hospitals was conducted prior to implementation of the MORES intervention.Data collection was repeated at endline, during the final 6 months of the study.Identifying data was collected to link data from hospital labor and delivery log and patient charts to data from RHF logs.The linking document, and identifiable data, were destroyed after clinical data were collected.Deidentified data with no direct identifiers from participant data, were collected and used for analysis included date and time of departure from RHF and hospital arrival, type of delivery, and maternal and newborn outcomes.The hospital logs contained more referral records than those recorded in the RHF logs, as not all women referred arrived at the district hospital.Thus, data were sorted as complete (hospital data linked to RHF data) or incomplete (hospital records only).Descriptive analyses were conducted for all baseline and endline data and analyzed as separate cohorts (complete and incomplete).

| Data management and analysis
Frequency and percentage were tabulated for RHFs, referrals to district hospitals, reasons for referral, and outcome variables including mode of delivery, maternal outcome, newborn outcome, and transfer time from RHF to district hospital.Transfer time was further presented as mean with (SD), as well as the median time, and categorized into 2 h or less, between 2 and 12 h, and more than 12 h.The 2 h or less category was included per The Lancet's definition of geographic accessibility, with access to a health facility with the capacity to provide essential surgical and anesthesia services, including CS, within 2 h. 14Furthermore, the 12 h or less and more than 12 h were included because despite a referral from a RHF, woman often go back home before going to the referred hospital.

| Ethics
All ethical approvals were obtained from the University of Michigan (no.UM HSBS-HUM0019544) and University of Liberia (date of approval February 5, 2021).The study used retrospective medical records, the data were fully anonymized before analysis, and the ethics committee waived the requirement for the informed consent.

| RE SULTS
Descriptive statistics for the baseline sample of 225 women referred from 20 RHFs to the two districts hospitals, CB Dunbar (75.1%) and Phebe (24.8%) are presented in Table 1.Of the 225 women, 93 (41.3%) had complete records and 132 (58.6%) had records only from the hospital.The most common reasons for referrals were obstructed labor (33.3%), other (12%) which included premature rupture of membranes, anemia, and previous CS (8%).Most women delivered via spontaneous vaginal delivery (56.8%), with 18 women (8%) and 92 babies (40.8%) experiencing complications.After the referral was initiated at the RHF, the median time for women to reach the hospital was 5.8 h.Only 12.9% of the women arrived at the hospital within 2 h or less.
Endline data included 287 women with 159 (55.4%) complete records (Table 2).Of the total sample, 155 (54%) women were referred to CB Dunbar and 132 (46%) referred to Phebe.Similar to the baseline sample, the most common reasons for referral among the endline sample were obstructed labor (22.6%), previous CS (15.6%), and other (12.8%).Less than half of the women delivered via spontaneous vaginal delivery (41.4%), with 13 women (4.5%) and 67 babies (23.3%) experiencing complications.The median time it took for women to arrive at the district hospital following referral from the RHF was 7.2 h and 18 (11.3%)women arrived at the hospital in 2 h or less from the time of referral.
Table 3 shows the association between timepoints, mode of delivery, maternal outcome, newborn outcome, and transfer time.
The unadjusted model showed that women at endline were more likely to undergo a CS (OR: 1.86; 95% CI: 0.99-3.46)compared to women at baseline.There was no statistically significant association between timepoints and maternal outcomes.Newborns at endline were less likely to be non-vigorous (OR: 0.31; 95% CI: 0.14-0.68),defined as an infant with presence of poor respiratory effort, poor muscle tone, or heart rate <100 beats per minute during the delivery room provider's initial assessment, compared to newborns at baseline.Lastly, there was no statistical significance in transfer time from RHF to hospital between baseline and endline.

| DISCUSS ION
This study examined the association between a MORES intervention implementation and mode of delivery, maternal outcome, newborn outcome, and transfer time from 20 RHFs to two district hospitals in rural Liberia.Following implementation of MORES, women had higher odds of receiving a CS and newborns had lower odds of being non-vigorous.No statistically significant association was observed between the intervention and maternal outcomes and transfer time.
Prompt and high-quality CS can significantly improve maternal and newborn outcomes. 15A study of all 194 WHO member states examining the relationship between CS rate and maternal and F I G U R E 2 Message template for received hospital referrals.
newborn mortality found a CS rate of 9%-19% of the population coverage was associated with decreased maternal and neonatal mortality. 8,16,17While the global CS rate continues to rise with current estimates at 21.1%, the CS rate in sub-Saharan Africa is only 5%. 8,18beria's facility-based deliveries significantly increased between 2004 and 2017, from 37% to 80%, most of the increased from rural areas. 8However, the CS rate did not reflect that improvement, with 6.1% born via CS in urban areas compared to 3.7% in rural areas, far below the 9%-19% of the ideal population coverage. 8,16,17Furthermore, in 2019, about 84% of deliveries in Bong County occurred in a health facility, with 81% taking place within RHFs. 13 This study found that the MORES intervention was associated with approximately 1.8 times higher CS rate, indicating that the intervention may be improving the unmet need of CS.
Relatedly, this study found that newborns were significantly less likely to exhibit poor respiratory effort and muscle tone after the intervention implementation.This may further indicate improved preparations to receive women being referred from RHFs due to better communication with the MORES intervention, leading to more timely care with improved newborn outcome.A study TA B L E 1 Descriptive statistical for overall baseline sample and stratified sample of women with both rural health facility (RHF) and hospital data and women with only hospital data.Abbreviations: IUFD, intrauterine fetal death; RHF, rural health facility; SD, standard deviation.a More than one reasons for referrals could have been recorded so the total percentage may not add up to 100.
b Observations that had no significant maternal and newborn outcomes were combined with missing observations.c Percentage calculated only among those that had date or departure, date of arrival, time of departure, and time of arrival.

TA B L E 1 (Continued)
examining the association between decision delivery interval, the time between a decision to conduct an emergency cesarean section and delivery of the baby, and maternal and fetal outcome found that there was no significant association between decision delivery interval and Apgar score, maternal blood loss, and hospital stay. 19While this study's stillbirth rates were not shown to be statistically significant, a downward trend from 20.4% to 13.8% was noted.
The transfer time from RHF to district hospitals was not significantly associated with the intervention.In fact, the median time increased from 5.87 to 7.2 h and no significant percentage changes were observed when time was examined as 2 h or less versus longer than 2 h.However, after the intervention, the percentage of women arriving to the hospital within 12 h more than doubled, compared to before the intervention.
Although all RHFs in this study were located less than 2 h from a CEmONC facility, it is worth noting that even post intervention close to 30% of the women took more than 12 h to reach district hospital following referral.This may due to the seasonal differences in road conditions affecting travel time (rainy dry season). 4Furthermore, the gap between communication, availability, and coordination of transportation services could

| Limitations
This study had several limitations.First, because we extracted referral data from routinely collected data at the RHFs and district hospitals through admission logs, additional demographic data such as women's age, marital status, education level, and reproductive history that could have functioned as a confounding factor are lacking.
Furthermore, reasons for referral could not be adjusted for because we could not collapse variables in a scientifically meaningful way to include in the model.Second, additional time points were not captured to examine where specifically the delays in travel time occurred.For example, capturing dispatch time, response time, travel time, and prehospital time could have been additional time related variables that could have enhanced the results. 22Due to these limitations, the result of this study needs to be interpreted carefully.Abbreviations: b, beta; CI, confidence interval; IUFD, intrauterine fetal death; N/A, not applicable due to small sample size; OR, odds ratio; RHFs, rural health facilities; SE, standard error.a Linear regression conducted using the mean of travel time.
outcomes, and time from RHF referral to hospital arrival.Inclusion criteria for RHFs included: staffed by providers who owned a smartphone capable of downloading WhatsApp and facility located 2 h or less from the district hospital.The two district hospitals included in the study provide the vast majority of CEmONC services to residents of the county.A designated smartphone for the study was provided to the two hospitals due to the large number of staff working at the hospitals and concerns regarding mixed communication from the receiving end of the referral.Baseline data were collected 6 months prior to implementation of the MORES intervention (November 1, 2020 to April 30, 2021) and endline data were collected during the last 6 months of the intervention (August 1, 2022 to January 31, 2023).
Referral data were collected via paper and pen, entered into Excel and exported into Stata 17 (StataCorp, College Station, TX, USA) for analysis.All paper copies were stored in a locked cabinet at the county health office in Bong County.All digital data were stored in an encrypted DropBox folder only accessible to the research team.
were fit to assess the relationship between the timepoint of the data (baseline, endline), mode of delivery, maternal outcome, newborn outcome, and transfer time.All models accounted for the clustering of individuals within hospital and RHF.All logistic regression models provided odds ratios (ORs) and 95% confidence intervals (95% CIs) F I G U R E 1 Message template for RHF to hospital referrals.and linear regression models provided coefficient and standard error (SE).Statistical significance was set at *P ≤ 0.05 and **P < 0.005.
Liberia has experienced increased facility-based deliveries in the past decade, it has not been accompanied by timely referral to hospitals capable of performing emergency obstetric services such as CS.Prior to implementation of the MORES intervention, unidirectional communication from RHFs to district hospitals prevented feedback to providers making the referral.Given that the MORES intervention was associated with a significant increase in CS and a reduction in non-vigorous newborns at birth, it may be an innovative solution to communicate emergency referral information, improve treatment upon arrival at the hospital, and improve bidirectional communication to ultimately reduce maternal and newborn morbidity and mortality.Future studies should examine the reasons for delays after referral has been made as well as interventions to overcome these challenges.TA B L E 3 Association between timepoints, mode of delivery, maternal outcome, newborn outcome, and travel time.
More than one reasons for referrals could have been recorded so the total percentage may not add up to 100.Percentage calculated only among those that had date or departure, date of arrival, time of departure, and time of arrival.Descriptive statistical for overall endline sample and stratified sample of women with both rural health facility (RHF) and hospital data and women with only hospital data.
a b Observations that had no significant maternal and newborn outcomes were combined with missing observations.c All models were accounted for the clustering of individuals within RHFs and hospitals.