Antenatal corticosteroids for pregnant women at risk of preterm labor in low- and middle-income countries: utilization and facility readiness

47

48 Antenatal corticosteroids (ACS) use among pregnant women with a high likelihood of preterm 49 labor improves newborn survival. ACS adoption in low- and middle-income countries (LMICs) 50 remains limited. Giving ACS in inadequately equipped settings could be harmful to mothers and 51 newborns. Thus, health facilities have to demontrate readiness to administer ACS. However, the 52 degree to which health systems are ready is unknown. 53

54 We assessed facility readiness to administer ACS based on the 2022 WHO recommendations on 55 ACS use and ACS utilization. 56

57 The study used Service Provision Assessment surveys administered between 2013 and 2022 in 58 nine LMICs. The primary outcome was whether facilities had ever provided ACS. We also 59 assessed injectable corticosteroid (dexamethasone or betamethasone) availability and facility 60 readiness to administer ACS. We used a total of 35 indicators, grouped into four readiness 61 categories based on the WHO recommendations, to measure facility readiness. 62 Findings 63 Across eight countries with comparable sampling strategies, only 10.7% (median, range 6.7% - 64 35.2%) of facilities had ever provided ACS; one-fourth (median 25.3%, range 4.6% - 61.5%) of 65 facilities had injectable corticosteroids available at the time of the survey; overall readiness 66 indices were low ranging from 8.1% for Bangladesh to 32.9% for Senegal. Across four readiness 67 categories, the readiness index was the lowest for criterion 1 (ability to assess gestational age 68 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 4 accurately and identify a high likelihood of preterm birth) (7.3%), followed by criterion 2 (ability 69 to identify maternal infections) (24.8%), criterion 4 (ability to provide adequate preterm care) 70 (31.3%), and criterion 3 (ability to provide adequate childbirth care) (32.9%). 71

72 We proposed a strategy for measuring facility readiness to implement one of the most effective 73 interventions to improve neonatal survival. Countries should operationalize readiness 74 measurement, improve facilities readiness to deliver this life-saving intervention, and encourage 75 ACS uptake by targeting facilities that are well-equipped. 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 5

92 Antenatal corticosteroids (ACS) use among pregnant women at risk of preterm labor is one of 93 the most effective interventions to improve neonatal outcomes. By accelerating fetal lung 94 maturity, ACS can reduce the risk of respiratory distress syndrome by 30% and neonatal death 95 by 20%.(1) The World Health Organization (WHO) and other professional medical organizations 96 recommend giving ACS to pregnant women at risk of imminent preterm labor from gestational 97 age (GA) of 24 to 34 weeks.(2-5) 98 While ACS is commonly used in high-income countries, its adoption remains extremely limited 99 in low- and middle-income countries (LMICs) despite LMICs contributing to 80% of preterm 100 births worldwide.(6) An estimated 13.4 million newborns were born prematurly in 2020.(7) 101 Globally, preterm births are the leading cause of neonatal death, accounting for nearly half (46%) 102 of under-five moratlity (U5M).(8) Additionally, 75% of all neonatal deaths occur in the first 103 week of life, with 1 million neonatal deaths happening within the first 24 hours after birth (9, 10), 104 which motivates the interventions targteing this critical period.(11, 12) ACS has been considered 105 “the lowest-hanging and sweetest fruit” as an intervention to improve preterm outcomes in 106 LMICs.(13) However, ACS uptake in LMICs has been a subject of widespread debate.(13-18) 107 Two major evidence gaps remain. First, the current status of ACS use in LMICs is unknown 108 despite some outdated coverage data. Second, the structural readiness of health facilities in 109 LMICs to provide ACS based on international guidelines is unclear. In 2022, the WHO released 110 its recommendations emphasizing five conditions for safe and effective administration of ACS: 1) 111 GA can be accurately assessed, 2) there is a high likelihood of preterm birth within 7 days of 112 starting ACS therapy, 3) there is no evidence of maternal infections, 4) adequate childbirth care 113 is available, and 5) the preterm newborn can receive adequate care.(2) 114 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 6 The WHO recommendations caution that these five conditions might not be met consistently 115 across settings because of varied capabilities, highlighting the potential harms of ACS in places 116 lacking the capacity to meet the criteria.(2). As a potent anti-inflammatory drug, ACS suppresses 117 immune functions. Maternal infections remain a major concern if ACS is given to vulnerable 118 pregnant women who are ineligible for this intervention. Also, observational studies found 119 increased neurocognitive disorders among late preterm infants (born at GA 34 to 36 weeks) who 120 were exposed to ACS.(19-21) The balance between the benefits and risks emphasized the 121 importance of locations in which ACS should be given. However, knowledge about facility 122 readiness in resource-constraint settings remains scarce. In 2021, Kankaria et al. found that 123 primary and secondary facilities in northern India were not ready to administer ACS safely, 124 which is, to date and to our knowledge, the only study assessing facility readiness to give 125 ACS.(22) As facility readiness is crucial in providing healthcare of good quality (23), there is an 126 urgent need to expand evidence on ACS use in low resource countries. 127 This study aimed to address these knowledge gaps. The objectives were to assess ACS use, 128 corticosteroid availability, and facility readiness to administer ACS according to the WHO 129 criteria. 130

131 Study sample 132 This study used data from the Service Provision Assessment (SPA), a health facility survey on 133 service availability and quality of care.(24) We restricted our analysis to SPAs done in the past 134 10 years and used the latest survey available for countries with SPA data. This study included 10 135 surveys from 9 countries: Afghanistan 2018-2019, Bangladesh 2017-2018, Nepal 2021, Haiti 136 2017-2018, the Democratic Republic of Congo (DRC) 2017-2018, Ethiopia 2021-2022, Malawi 137 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 7 2013-2014, Senegal 2018 and 2019, and Tanzania 2014-2015. All surveys were completed 138 before the release of the 2022 WHO recommendations but most, except Malawi 2013-2014 and 139 Tanzania 2014-2015 surveys, were conducted after the WHO recommendations on interventions 140 to improve preterm birth outcomes in 2015, which contained guidelines on ACS use.(25) The 141 sampling strategies varied based on country needs.(26-35) The majority of surveys adopted 142 stratified random sampling strategies to obtain a national representative sample in the 143 country.(26-28, 31, 32) Malawi 2013-2014 and Haiti 2017-2018 SPAs were national census.(33, 144 35) Unlike other surveys, Afghanistan 2018-2019 mainly sampled urban hospitals.(34) Senegal 145 implemented continuous SPA over a consecutive 5 years to survey all health facilities.(29, 30, 36) 146 In order to have similar sample sizes across countries, we merged two years (2018 and 2019) of 147 data from Senegal (Supplemental Table 1). Our study used data from two core instruments (out 148 of five) of the SPA survey questionnaire: facility inventory and health worker interviews. In each 149 country, only facilities that provided antenatal care (ANC), performed normal deliveries, and/or 150 performed Cesarean deliveries were included. 151 Measures 152 Our primary outcome was ACS utilization, defined as facilities that had ever provided ACS to 153 pregnant women. We focused on two secondary outcomes: corticosteroid availability (injectable 154 dexamethasone or betamethasone) and facility structural readiness. Most countries did not survey 155 corticosteroid availability (except for Afghanistan 2018-2019) within the maternal and child 156 health care sections in SPA surveys. We reported the availability of injectable corticosteroids 157 that were surveyed in the section on medicines for non-communicable diseases. 158 We assessed facility readiness to provide ACS in accordance with the 2022 WHO 159 recommendations on ACS use. We identified 35 indicators from the SPA questionnaire, which 160 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 8 were grouped into four readiness categories based on the five WHO criteria (collapsing criterion 161 1 and criterion 2) as outlined in Supplemental Table 2. The first category focused on the 162 facility’s ability to assess GA accurately and to identify pregnant women with a high likelihood 163 of preterm labor; we included only one indicator for this category (presence of a functional 164 ultrasound). The second category included 4 indicators and covered the facility’s ability to 165 identify maternal infections. The third and fourth categories included 13 and 17 indicators and 166 assessed the facility’s readiness to provide adequate childbirth care and preterm newborn care, 167 respectively. For each category, readiness indices were calculated by dividing the number of 168 indicators available by the total number of indicators assessed, where higher percentages 169 indicating higher readiness. An overall readiness index was calculated by averaging the readiness 170 indices from four categories. This approach was chosen based on previous studies on facility 171 readiness to implement health improvement interventions.(37, 38) 172 Statistical analysis 173 First, we presented descriptive statistics of ACS utilization, corticosteroid availability, and 174 facility readiness indices, taking into account survey sampling weights. Results were stratified by 175 three facility levels. Level 1 facilities were those that only provided ANC and did not conduct 176 deliveries; ANC was defined as the care that pregnant women receive before birth, including risk 177 identification, prevention, and management of pregnancy related health conditions, education, 178 and health promotion.(39) Level 2 facilities performed normal deliveries, but not Cesarean 179 deliveries, and level 3 facilities performed Cesarean deliveries. Because Afghanistan mainly 180 sampled hospitals, which substantially differed from other countries, it was excluded from the 181 summary descriptive analysis of our outcomes of interest and included only in analyses by 182 facility levels. 183 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 9 Second, we assessed the relationships between ACS use and, respectively, ACS availability, and 184 overall readiness indices at sub-national levels (e.g. regions in Ethiopia, provinces in DRC). This 185 was driven by the hypothesis that facilities without corticosteroids or with low readiness might 186 refer patients in need of ACS to facilities in geographic proximity. We averaged each of the three 187 measures (ACS use, corticosteroid availability, and readiness index) for all facilities in each sub-188 national region. Lastly, we examined the difference in the overall readiness index between 189 facilities that had ever and never used ACS by country. All analyses were performed using R. 190

191 This study included a total of 8669 facilities from ten surveys in nine countries. All surveys had a 192 high response rate (median 94.9%) ranging from 88.8% in Afghanistan to 99.0% in Tanzania 193 (Supplemental Table 3). 194 The majority (88.9%, median; range 66.1% - 98.8%) of facilities provided maternal health care 195 services and were included in our analysis (Supplemental Table 3). The median sample size 196 was 929 facilities (median; range 108 -1500) (Table 1). Among eight countries (excluding 197 Afghanistan), 22.6% (median; range 7.1% - 88.7%) of facilities were urban. The proportion of 198 facilities of different levels varied across countries. Across eight countries, one third (median 199 32%) of the facilities were level 1 facilities that provided antenatal care, with Bangladesh having 200 the highest proportion (76.2%) of level 1 facilities and the DRC having the lowest proportion 201 (1.9%). The proportion of level 2 facilities was 59.4% (median) with Tanzania having the 202 highest proportion (83.3%) and Bangladesh with the lowest proportion (19.5%). Only 6.2% 203 (median) of facilities were level 3 facilities that performed Cesarean deliveries, with 26.5% of 204 facilities in DRC and 2.7% in Ethiopia were level 3 (Table 1). 205 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 10 ACS utilization. ACS was underutilized across all countries. Excluding Afghanistan, median 206 ACS use was 10.7%, ranging from 6.7% of facilities in Bangladesh to 35.2% of facilities in the 207 DRC having ever administered ACS (Figure 1a). In Afghanistan, 68.7% of facilities had 208 administered ACS at the time of the survey. We did not find higher ACS use in surveys that were 209 done in more recently. While none of the level 1 facilities had provided ACS, ACS use increased 210 by facility level (Figure 1b). Within facility level across 9 countries, 21.6% (median, range 5.0% 211 - 26.9%) of level 2 facilities and 76.3% (median, range 52.7% - 90.3%) of level 3 facilities had 212 provided ACS. 213 ACS availability. Corticosteroid availability was limited. Across eight countries, a fourth of the 214 facilities (median 25.3%, range 4.6% - 61.5%) had at least one valid corticosteroid (injectable 215 betamethasone or dexamethasone) available at the time of the survey, while 93.3% of facilities in 216 Afghanistan had corticosteroids available (Figure 1c). Corticosteroid availability generally 217 increased by facility level within each country (Figure 1d). Also, gaps between corticosteroid 218 availability and ACS utilization existed. For example, 48.5% of level 2 facilities in Nepal had 219 corticosteroid available at the time of survey but only 7.3% had ever used it (Supplemental 220 Figure 5). 221 Facility readiness. Readiness indices were low (Figure 2). Overall, only 22% of the facilities in 222 the sample had an overall readiness index above 50%. Other than Afghanistan, overall readiness 223 indices were low among the eight countries ranging from 8.1% in Bangladesh to 32.9% in 224 Senegal. Afghanistan had an overall readiness index of 57.7%. We did not see higher overall 225 readiness indices in surveys done in more recent years. Also, overall readiness indices increased 226 by facility level. Across four readiness categories among 8 countries, facilities performed the 227 worst in the ability to assess GA accurately, with a readiness index of 7.3% (median), followed 228 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 11 by the ability to identify maternal infection (median, 24.8%), provide adequate preterm care 229 (median 31.3%), and provide adequate childbirth care (median 32.9%). When delving into the 230 details of specific indicators among the 35 ones used, we found very limited availability for 231 respiratory support-related equipment across countries in the readiness category of adequate 232 preterm newborn care (Supplemental Table 4, Supplemental Figures 1- 4). 233 ACS use, availability, and readiness at sub-national levels. At the sub-national level, positive 234 associations were observed between corticosteroid availability and ACS use as well as between 235 average readiness and ACS use (Figures 3a and 3b). A few Ethiopian regions (Dire Dawa and 236 Harari) had an average overall readiness index above 50% but low ACS use (<25%). In contrast, 237 some regions of the DRC (Kasaï Central, Kongo Central, Kinshasa, and Haut-Katanga) had low 238 overall readiness indices (<50%), but ACS use greater than 50% (Figure 3b). 239 At the facility level, overall readiness indices differed between facilities that had ever and never 240 provided ACS within each country (Supplemental Figures 6-8). In Afghanistan, Bangladesh, 241 DRC, Senegal, and Tanzania, facilities that had utilized ACS had higher overall readiness indices 242 compared to facilities that had never used ACS. On the contrary, the results were reversed for 243 Nepal, Haiti, Ethiopia, and Malawi. 244

245 This study assessed antenatal corticosteroid use, corticosteroid availability and structural 246 readiness to administer the drug adequately based on international recommendations among 247 8,669 health facilities from nine resource-constrained countries. We had three major findings: 1) 248 antenatal corticosteroids were substantially underused, 2) corticosteroid availability was limited, 249 and 3) facilities in these countries had low levels of readiness. 250 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 12 We found limited ACS use with only one out of ten health facilities having ever administered 251 ACS. However, it is challenging to compare our results with existing literature as previous work 252 mostly measured ACS coverage among pregnant women who delivered preterm infants.(40-45) 253 A 2011 WHO maternal and newborn health survey for 29 countries found 54% of women who 254 gave birth at GA 26 to 34 weeks were given ACS, with the lowest use in Afghanistan (16%), 255 Nepal (20%), and DRC (16%).(41) Another analysis using the data from the 2015 Antenatal 256 Corticosteroids Trial (ACT trial) showed an overall low ACS use in control clusters; in Kenya, 257 only 3.8 % of pregnant women of infants born with birthweight less than the 5th percentile, a 258 proxy for preterm births, received ACS.(40) Our findings adds value in understanding the status 259 of ACS use at the facility level. Also, we found that only one third of the facilities in the sample 260 had either injectable dexamethasone or betamethasone. Dexamethasone is on the WHO List of 261 Essential Medicines, but limited availability remains a major barrier to ACS use.(46) Strategies 262 to improve ACS coverage need to ensure drug availability. 263 The location where ACS is given is crucial to ensure safe and effective use. Two landmark 264 studies on the effects of ACS in LMICs presented conflicting results, highlighting the importance 265 of settings in which ACS is given.(47, 48) The 2015 ACT trial, a cluster randomized trial of a 266 multifaceted intervention to promote ACS use in six LMICs unexpectedly found increased 267 neonatal deaths and suspected maternal infections among intervention clusters.(47) On the 268 contrary, the WHO Antenatal Corticosteroids for Improving Outcomes in Preterm Newborns 269 Trial (ACTION-I trial) in 2020, an RCT in five LMICs, showed ACS reduced neonatal deaths 270 without increasing maternal infections.(48) These contradictory findings can partially be 271 explained by the different settings for the two trials; the ACT trial was done in all levels of care 272 including clinics and primary care centers, whereas the ACTION-I trials included secondary or 273 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 13 tertiary hospitals.(48) The drastically different findings emphasized the importance of locations – 274 or more precisely, the readiness of facilities – in providing ACS. Our study offered a strategy for 275 measuring facility readiness to implement this life-saving intervention. However, 276 recommendations on what levels of facilities should give ACS need to be made carefully. 277 Assuming essential equipment, medicines and trained staff are available, some level 2 and level 278 3 facilities with high readiness that have never provided ACS should be targeted for the 279 expansion of ACS use. Also, our sub-national level analyses showed that some provinces in 280 DRC with low overall readiness (<50%) frequently administered ACS. Policymakers should 281 ensure that ACS is delivered in well-equipped settings. 282 Another critical issue centers around the different aspects of readiness to ensure safe and 283 effective ACS use. Our readiness index was developed based on the WHO criteria. However, it 284 is debatable, first, whether some criteria could be met across facility levels, and second, whether 285 facilities need to meet all criteria to safely and effectively administer ACS. For example, 286 ultrasound examination in early pregnancy is the gold standard for GA assessment. However, we 287 found that only 7.3% (median across eight countries, range 0.2% to 12.3%; 82.4% for 288 Afghanistan) of maternal care facilities had a functional ultrasound. Access to ultrasound for GA 289 dating remains a major barrier to proper ACS use despite its increasing use in obstetric care in 290 LMICs.(17, 49) In this case, GA dating should occur early in pregnancy, while ACS use occurs 291 later in pregnancy. Thus, the decision to administer ACS should not be based on the ultrasound 292 availability in place, but the availability of accurate GA assessment (which may come from care 293 obtained at another facility). In addition, regarding the WHO recommendations for adequate 294 preterm newborn care, one important element is non-invasive respiratory support, such as 295 continuous positive airway pressure (CPAP). However, this recommendation might exclude most 296 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 14 preterm infants who might benefit from ACS use because they usually do not have access to 297 respiratory support in LMICs.(50) A similar concern applies to adequate childbirth care, 298 referring to nine signal functions of Comprehensive Emergency Obstetric and Newborn Care 299 (CEmONC), including blood transfusions and Cesarean sections. Again, it is debatable whether a 300 facility can give ACS only when it can do blood transfusions. 301 Our study had several strengths. To our knowledge, this is the first study to assess ACS use, 302 availability and structural readiness to give ACS at the facility level. Also, we included data from 303 multiple countries to understand the landscape of ACS use and identify policy directions in areas 304 with a high burden of preterm births. Nonetheless, our study has a few limitations. First, 305 sampling strategies and survey years varied across countries spanning 9 years. All surveys were 306 done prior to the release of the 2022 WHO recommendations while most of them were 307 completed after the release of 2015 WHO recommendations on interventions to improve preterm 308 birth outcomes that contained guidelines on ACS utilization.(2, 25) Also, countries have 309 different national guidelines regarding ACS use. One policy analysis on ACS use in Africa 310 found ACS could be given in lower levels of care before referral in Ethiopia and Tanzania, but 311 DRC and Malawi only allowed ACS use in hospitals.(51) Hence, we caution careful 312 interpretations when comparing across countries. Secondly, some readiness indicators served as 313 proxies. For instance, SPA surveys do not assess the availability of corticosteroids and 314 ultrasound in the maternal care section. Instead, corticosteroid availability is surveyed in the 315 section for medicines for non-communicable diseases, and ultrasound availability is assessed 316 within the facility in general. We also used the availability of rapid diagnostic tests for HIV and 317 syphilis as a proxy to estimate facilities’ ability to detect maternal infections. This approach 318 might overestimate or underestimate true availability or readiness. Thirdly, we were unable to 319 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 15 draw correlations between ACS use or facility readiness with preterm health outcomes because 320 SPAs do not have patient-level data for newborns. 321 Antenatal corticosteroid use has gained tremendous international attention as its population-level 322 health benefits could be profound.(12, 52-55) ACS should not be used as a “just-in-case” 323 intervention.(56) It should be given to the right people (pregnant women at risk of imminent 324 preterm labor and without maternal infections), at the right time (the specified GA window), and 325 in the right place (settings that are properly equipped and ready to provide quality maternal and 326 newborn care). Future research and programs should operationalize readiness measurement and 327 assess facility readiness for safe and effective ACS use to enhance readiness and encourage ACS 328 uptake among well-prepared facilities. 329 Disclosure of relationships and activities: All authors have completed the ICMJE uniform 330 disclosure form at www.icmje.org/coi_disclosure.pdf and declare no support from any 331 organization for submitted work; no financial relationships with any organization that might have 332 an interest in the submitted work in the previous three years; no other relationships or activities 333 that could appear to have influenced the submitted work. 334 Supporting information 335 S1 Supplementary file 336 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 16

1. McGoldrick E, Stewart F , Park er R, Dalziel SR. Antenatal corticosteroids for accelerating f etal lung maturation for women at risk of preterm birth. Cochrane Database Syst Rev . 2020;12(12):Cd004454. Epub 2020/12/29. doi: 10.1002/14651858.CD004454.pub4. PubMed PMID: 33368142; PubMed Central PMCID: PMCPMC8094626 Stuart Dalziel reports receiving research funding, not associated with the review topic, from Cure Kids, Health Research Council, and Starship Foundation, New Zealand. Stuart Dalziel is employed by The University of Auckland and Auckland District Health Board. He is also on the board of the Advanced Paediatric Lif e Support New Zealand. 2. The W orld Health Organization. 2022 WHO recommendations on ACS to improve preterm birth outcomes. 2022. Epub 2022/09/29. 3. Reddy UM, Deshmukh U, Dude A, Harper L, Osmundson SS. Society for Maternal-Fetal Medicine Consult Series #58: Use of antenatal corticosteroids for individuals at risk for late preterm delivery: Replaces SMFM Statement #4, Implementation of the use of antenatal corticosteroids in the late preterm birth period in women at risk for preterm delivery , August 2016. American journal of obstetrics and gynecology . 2021;225(5):B36-b42. Epub 2021/08/08. doi: 10.1016/j.ajog.2021.07.023. PubMed PMID: 34363784. 4. Norman J, Shennan A, Jacobsson B, Stock SJ, Birth FWGfP . FIGO good practice recommendations on the use of prenatal corticosteroids to improve outcomes and minimize harm in babies born preterm. Int J Gynaecol Obstet. 2021;155(1):26-30. Epub 2021/09/15. doi: 10.1002/ijgo.13836. PubMed PMID: 34520057. 5. El-Sayed YY , Borders AEB, Gyamfi-Bannerma C. ACOG committee opinion, Antenatal Corticosteroid Therapy for Fetal Maturation The American College of Obstetricians and Gynecologists. 2017;130(2 ):e102-e9. 6. Chawanpaiboon S, V ogel JP , Moller AB, Lumbiganon P , Petzold M, Hogan D, et al. Global, regional, and national estimates of levels of preterm birth in 2014: a systematic review and modelling analysis. The Lancet Global health. 2019;7(1):e37-e46. Epub 2018/11/06. doi: 10.1016/s2214-109x(18)30451-0. PubMed PMID: 30389451; PubMed Central PMCID: PMCPMC6293055. 7. Ohuma EO, Moller AB, Bradley E, Chakwera S, Hussain-Alkhateeb L, Lewin A, et al. National, regional, and global estimates of preterm birth in 2020, with trends from 2010: a systematic analysis. Lancet (London, England). 2023;402(10409):1261-71. Epub 2023/10/08. doi: 10.1016/s0140- 6736(23)00878-4. PubMed PMID: 37805217. 8. Perin J, Mulick A, Y eung D, Villavicencio F , Lopez G, Strong KL, et al. Global, regional, and national causes of under-5 mortality in 2000-19: an updated systematic analysis with implications for the Sustainable Development Goals. The Lancet Child & adolescent health. 2022;6(2):106-15. Epub 2021/11/21. doi: 10.1016/s2352-4642(21)00311-4. PubMed PMID: 34800370; PubMed Central PMCID: PMCPMC8786667. 9. The W orld Health Organization. Newborn mortality 2024 [cited 2024 May 13]. Available from: https://www . who.int/westernpacific/news-room/fact-sheets/detail/newborn-mortality . 10. Sankar MJ, Natarajan CK, Das RR, Agarwal R, Chandrasekaran A, Paul VK. When do newborns die? A systematic review of timing of overall and cause-specific neonatal deaths in developing countries. Journal of perinatology : official journal of the California Perinatal Association. 2016;36 Suppl 1(Suppl 1):S1-s11. Epub 2016/04/26. doi: 10.1038/jp.2016.27. PubMed PMID: 27109087; PubMed Central PMCID: PMCPMC4848744. 11. Hofmeyr GJ, Black RE, Rogozińska E, Heuer A, W alk er N, Ashorn P , et al. Evidence-based antenatal interventions to reduce the incidence of small vulnerable newborns and their associated poor outcomes. Lancet (London, England). 2023;401(10389):1733-44. Epub 2023/05/12. doi: 10.1016/s0140- 6736(23)00355-0. PubMed PMID: 37167988. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 17 12. Bhutta ZA, Das JK, Bahl R, Lawn JE, Salam RA, Paul VK, et al. Can available interventions end preventable deaths in mothers, newborn babies, and stillbirths, and at what cost? Lancet (London, England). 2014;384(9940):347-70. Epub 2014/05/24. doi: 10.1016/s0140-6736(14)60792-3. PubMed PMID: 24853604. 13. Jobe AH, Kemp MW , Kamath-Rayne B, Schmidt AF . Antenatal corticosteroids for low and middle income countries. Semin Perinatol. 2019;43(5):241-6. Epub 2019/04/14. doi: 10.1053/j.semperi.2019.03.012. PubMed PMID: 30979597. 14. Sultana S, V ogel JP , Oladapo O T . The efficacy of antenatal corticosteroids to improve preterm newborn outcomes in low-resource countries: Are we there yet? BJOG : an international journal of obstetrics and gynaecology . 2023. Epub 2023/08/02. doi: 10.1111/1471-0528.17611. PubMed PMID: 37530472. 15. V ogel JP , Oladapo O T , Pileggi-Castro C, Adejuyigbe EA, Althabe F , Ariff S, et al. Antenatal corticosteroids for women at risk of imminent preterm birth in low-resource countries: the case for equipoise and the need for efficacy trials. BMJ Glob Health. 2017;2(3):e000398. Epub 2017/10/31. doi: 10.1136/bmjgh-2017-000398. PubMed PMID: 29082019; PubMed Central PMCID: PMCPMC5656119. 16. McClure EM, Goldenberg RL, Jobe AH, Miodovnik M, Koso-Thomas M, Buek ens P , et al. Reducing neonatal mortality associated with preterm birth: gaps in knowledge of the impact of antenatal corticosteroids on preterm birth outcomes in low-middle income countries. Reprod Health. 2016;13(1):61. Epub 2016/05/26. doi: 10.1186/s12978-016-0180-6. PubMed PMID: 27221397; PubMed Central PMCID: PMCPMC4877818. 17. T ownsend R, Calvert C. Antenatal corticosteroids in adequately equipped facilities in low- resource settings. The Lancet Global health. 2022;10(10):e1379-e80. Epub 2022/09/17. doi: 10.1016/s2214-109x(22)00387-4. PubMed PMID: 36113517. 18. Azad K, Costello A. Extreme caution is needed before scale-up of antenatal corticosteroids to reduce preterm deaths in low-income settings. The Lancet Global health. 2014;2(4):e191-2. Epub 2014/08/12. doi: 10.1016/s2214-109x(14)70020-8. PubMed PMID: 25103050. 19. Ninan K, Liyanage SK, Murphy KE, Asztalos EV , McDonald SD. Evaluation of Long-term Outcomes Associated With Preterm Exposure to Antenatal Corticosteroids: A Systematic Review and Meta-analysis. JAMA pediatrics. 2022;176(6):e220483. Epub 2022/04/12. doi: 10.1001/jamapediatrics.2022.0483. PubMed PMID: 35404395; PubMed Central PMCID: PMCPMC9002717. 20. Lin YH, Lin CH, Lin MC, Hsu Y C, Hsu CT . Antenatal Corticosteroid Exposure Is Associated with Childhood Mental Disorders in Late Preterm and T erm Infants. J Pediatr . 2022. Epub 2022/10/07. doi: 10.1016/j.jpeds.2022.09.050. PubMed PMID: 36202238. 21. Raikk onen K, Gissler M, Kajantie E. Associations Between Maternal Antenatal Corticosteroid T reatment and Mental and Behavioral Disorders in Children. JAMA. 2020;323(19):1924-33. Epub 2020/05/20. doi: 10.1001/jama.2020.3937. PubMed PMID: 32427304; PubMed Central PMCID: PMCPMC7237984. 22. Kankaria A, Duggal M, Chauhan A, Sarkar D , Dalpath S, Kumar A, et al. Readiness to Provide Antenatal Corticosteroids for Threatened Preterm Birth in Public Health F acilities in Northern India. Global health, science and practice. 2021;9(3):575-89. Epub 2021/10/02. doi: 10.9745/ghsp-d-20-00716. PubMed PMID: 34593583; PubMed Central PMCID: PMCPMC8514043. 23. Leslie HH, Spiegelman D, Zhou X, Kruk ME. Service readiness of health facilities in Bangladesh, Haiti, Kenya, Malawi, Namibia, Nepal, Rwanda, Senegal, Uganda and the United Republic of T anzania. Bulletin of the W orld Health Organization. 2017;95(11):738-48. Epub 2017/11/18. doi: 10.2471/blt.17.191916. PubMed PMID: 29147054; PubMed Central PMCID: PMCPMC5677617. 24. The DHS program. The Demographic and Health Surveys, The Service Provision Assessment USAID; [cited 2023 July 20]. Available from: https://dhsprogram.com/methodology/Survey- T ypes/SP A.cfm . . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 18 25. The W orld Health Organization. WHO recommendations on interventions to improve preterm birth outcomes. 2015 26. Ministry of Health, ICF . Nepal Health Facility Survey 2021. Kathmandu, Nepal, Rockville, Maryland, USA: Ministry of Health and Population, New ERA and ICF , 2022. 27. Ethiopian Public Health Institute, Ministry of Health E, ICF . Ethiopia Service Provision Assessment 2021-22. 2021. 28. National Institute of Population Research, Ministry of Health, ICF . Bangladesh Health F acility Survey 2017. Dhaka, Bangladesh: NIPORT - ICF , 2020. 29. Agence Nationale de la Statistique et de la Démographie/ ANSD , ICF . Senegal: Enquête Continue sur la Prestation des Services de Soins de Santé (ECPSS) 2018. Dakar , Sénégal: ANSD/ICF , 2020. 30. Agence Nationale de la Statistique et de la Démographie , ICF . Senegal: Enquête Continue sur la Prestation des Services de Soins de Santé (ECPSS) 2019. Dakar , Sénégal: ANSD/ICF , 2020. 31. Ecole de Santé Publique de Kishasa ESPK/Kinshasa RDdC, ICF . Republique Democratique du Congo É valuation des Presentations des Services de souns de Santé EPSS 2017-2018. Kinshasa, République Démocratique du Congo: ESPK and ICF , 2019. 32. Ministry of Health T anzania, National Bureau of Statistics T anzania, Office of Chief Government Statistician/T anzania, ICF International. T anzania Service Provision Assessment Survey 2014-2015. Dar es Salaam, T anzania: MoHSW/T anzania, MoH/T anzania, NBS/T anzania, OCGS/T anzania, and ICF International, 2016. 33. Ministry of Health - MoH/Malawi, ICF International. Malawi Service Provision Assessment 2013- 14. Lilongwe, Malawi: MoH/Malawi and ICF International, 2014. 34. The W orld Bank. The W orld Bank Microdata Library . Service Provision Assessment Survey Af ghanistan 2018-2019 Available from: https://microdata. worldbank.org /index.php/catalog /3645/study- description . 35. Institut Haïtien de l'Enfance , ICF . Haiti Évaluation de la Prestation des Services de Soins de Santé (EPSSS 2017-2018). 2019. 36. The DHS program. Continuous Service Provision Assessment. 37. Mallick L, T emsah G, W ang W . Comparing summary measures of quality of care for family planning in Haiti, Malawi, and T anzania. PLOS ONE. 2019;14(6):e0217547. doi: 10.1371/journal.pone.0217547. 38. Sheffel A, Zeger S, Heidkamp R, Munos MK. Development of summary indices of antenatal care service quality in Haiti, Malawi and T anzania. BMJ Open. 2019;9(12):e032558. doi: 10.1136/bmjopen- 2019-032558. 39. The W orld Health Organization. WHO recommendations on antenatal care for a positive pregnancy experience. 2016. 40. Berrueta M, Hemingway-Foday J, Thorsten VR, Goldenberg RL, Carlo W A, Garces A, et al. Use of antenatal corticosteroids at health facilities and communities in low-and-middle income countries. Reprod Health. 2016;13(1):66. Epub 2016/05/28. doi: 10.1186/s12978-016-0176-2. PubMed PMID: 27228986; PubMed Central PMCID: PMCPMC4882797. 41. V ogel JP , Souza JP , Gülmezoglu AM, Mori R, Lumbiganon P , Qureshi Z, et al. Use of antenatal corticosteroids and tocolytic drugs in preterm births in 29 countries: an analysis of the WHO Multicountry Survey on Maternal and Newborn Health. Lancet (London, England). 2014;384(9957):1869- 77. Epub 2014/08/17. doi: 10.1016/s0140-6736(14)60580-8. PubMed PMID: 25128271. 42. Riganti AA, Caff erata ML, Althabe F , Gibbons L, Segarra JO, Sandoval X, et al. Use of prenatal corticosteroids for preterm birth in three Latin American countries. Int J Gynaecol Obstet. 2010;108(1):52-7. Epub 2009/11/07. doi: 10.1016/j.ijgo.2009.08.022. PubMed PMID: 19892349; PubMed Central PMCID: PMCPMC3401047. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 19 43. Pattanittum P , Ewens MR, Laopaiboon M, Lumbiganon P , McDonald SJ, Crowther CA. Use of antenatal corticosteroids prior to preterm birth in four South East Asian countries within the SEA- ORCHID project. BMC pregnancy and childbirth. 2008;8:47. Epub 2008/10/18. doi: 10.1186/1471-2393- 8-47. PubMed PMID: 18925968; PubMed Central PMCID: PMCPMC2596081. 44. Krauss Silva L PT , Franklin R, Oliveira N. Assessment of quality of obstetric care and corticoid use in preterm labor . Cadernos de Salude Publica. 1999;15(4):1-23. 45. Tita A T , Selwyn BJ, W aller DK, Kapadia AS, Dongmo S. Evidence-based reproductive health care in Cameroon: population-based study of awareness, use and barriers. Bulletin of the W orld Health Organization. 2005;83(12):895-903. Epub 2006/02/08. PubMed PMID: 16462981; PubMed Central PMCID: PMCPMC2626486. 46. The W orld Health Organization. The WHO Model List of Essential Medicines. 2023. 47. Althabe F , Belizán JM, McClure EM, Hemingway-Foday J, Berrueta M, Mazzoni A, et al. A population-based, multifaceted strategy to implement antenatal corticosteroid treatment versus standard care for the reduction of neonatal mortality due to preterm birth in low-income and middle- income countries: the ACT cluster-randomised trial. The Lancet. 2015;385(9968):629-39. doi: 10.1016/s0140-6736(14)61651-2. 48. Collaborators W A T , Oladapo O T , V ogel JP , Piaggio G, Nguyen MH, Althabe F , et al. Antenatal Dexamethasone for Early Preterm Birth in Low-Resource Countries. N Engl J Med. 2020;383(26):2514-25. Epub 2020/10/24. doi: 10.1056/NEJMoa2022398. PubMed PMID: 33095526; PubMed Central PMCID: PMCPMC7660991. 49. Zahroh RI, Hazfiarini A, Eddy KE, V ogel JP , T unçalp ;iZH, Minckas N, et al. Factors influencing appropriate use of interventions for management of women experiencing preterm birth: A mix ed-

systematic review and narrative synthesis. PLoS medicine. 2022;19(8):e1004074. doi: 10.1371/journal.pmed.1004074. 50. Hall M, V alencia CM, Soma-Pillay P , Luyt K, Jacobsson B, Shennan A. E ffective and simple interventions to improve outcomes for preterm infants worldwide: The FIGO PremPrep-5 initiative. Int J Gynaecol Obstet. 2024;165(3):929-35. Epub 2024/01/24. doi: 10.1002/ijgo.15269. PubMed PMID: 38264849. 51. Greensides D, Robb-McCord J, Noriega A, Litch JA. Antenatal Corticosteroids for W omen at Risk of Imminent Preterm Birth in 7 sub-Saharan African Countries: A Policy and Implementation Landscape Analysis. Global health, science and practice. 2018;6(4):644-56. Epub 2018/12/24. doi: 10.9745/ghsp-d- 18-00171. PubMed PMID: 30573455; PubMed Central PMCID: PMCPMC6370350. 52. Sebastian E, Byk ersma C, Eggleston A, Eddy KE, Chim ST , Zahroh RI, et al. Cost-eff ectiveness of antenatal corticosteroids and tocolytic agents in the management of preterm birth: A systematic review . EClinicalMedicine. 2022;49:101496. Epub 2022/06/25. doi: 10.1016/j.eclinm.2022.101496. PubMed PMID: 35747187; PubMed Central PMCID: PMCPMC9167884. 53. Eddy KE, V ogel JP , Scott N, Fetene D, Tidhar T , Oladapo O T , et al. Antenatal dexamethasone for improving preterm newborn outcomes in low-resource countries: a cost-effectiveness analysis of the WHO ACTION-I trial. The Lancet Global Health. 2022;10(10):e1523-e33. doi: 10.1016/s2214- 109x(22)00340-0. 54. Bill and Melinda Gates Foundation. 2023 Goalk eepers Report. 2023. 55. The United Nations Population Fund. UN Commission on Lif e-Saving Commodities for W omen and Children Commissioners' Report. 2012. 56. Norman J, Shennan A, Jacobsson B, Stock SJ. FIGO good practice recommendations on the use of prenatal corticosteroids to improve outcomes and minimize harm in babies born preterm. Int J Gynaecol Obstet. 2021;155(1):26-30. Epub 2021/09/15. doi: 10.1002/ijgo.13836. PubMed PMID: 34520057. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 20 Figure legends Figure 1a. Utilization of antenatal corticosteroids by country Figure 1b. Utilization of antenatal corticosteroids by country and facility level Figure 1c. Availability of at least one valid corticosteroid by country Figure 1d. Availability of at least one valid corticosteroid by country and facility level Figure 2. Facility readiness by country, readiness category, and facility level Figure 3a. Corticosteroid availability versus ACS utilization by region Figure 3b. Overall readiness indices versus ACS utilization by region . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 21 Table 1. Characteristics of surveys and facilities included in the study Country Survey year Number of facilities surveyed1 Number of facilities included in the analysis2 Urban location Facility type Level 1: Facilities that provide antenatal care Level 2: Facilities that perform normal delivery Level 3: Facilities that perform Cesarean section N = 9793 Median 1158 (range 142 - 1576) N = 8669 Median 929 (range 108 - 1500) Median 22.6%3 (range 7.1% - 88.7%)3 Median 32.0%3 (range 1.9% - 76.2%)3 Median 59.4%3 (range 19.5% - 83.3%)3 Median 6.2%3 (range 2.7% - 26.5%)3 South Asia Afghanistan 2018-2019 142 108 107 (99.4%) 7 (4.1%) 10 (10.1%) 91 (85.9%) Bangladesh 2017-2018 1524 1498 379 (7.1%) 676 (76.2%) 551 (19.5%) 271 (4.4%) Nepal 2021 1576 1500 961 (53.2%) 693 (47.5%) 565 (47.1%) 242 (5.3%) Caribbean Haiti 2017-2018 1007 929 340 (36.5%) 567 (61.0%) 255 (27.5%) 107 (11.5%) Sub-Saharan Africa DRC 2017-2018 1380 1364 297 (22.0%) 10 (1.9%) 511 (71.6%) 843 (26.5%) Ethiopia 2021-2022 1158 911 455 (18.2%) 261 (75.0%) 309 (22.3%) 341 (2.7%) Malawi 2013-2014 977 645 119 (18.6%) 103 (16.5%) 471 (72.6%) 71 (10.9%) Senegal 2018 and 2019 841 658 594 (88.7%) 65 (14.3%) 531 (78.6%) 62 (7.1%) Tanzania 2014-2015 1188 1056 325 (19.1%) 104 (11.5%) 681 (83.3%) 271 (5.2%) 1 A total of 625 facilities that were sampled had a sampling weight of zero, indicating non-response, refusal, or closure. Those facilities were not surveyed and were excluded. 2 Only facilities that provided either antenatal care, performed normal delivery or Cesarean section were included in the analysis. 3 The median and range were obtained across eight countries excluding Afghanistan because its survey used a remarkably different sampling strategy. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 22 Figure 1a. Utilization of antenatal corticosteroids by country Figure 1b. Utilization of antenatal corticosteroids by country and facility level Figure 1c. Availability of at least one valid corticosteroid by country Figure 1d. Availability of at least one valid corticosteroid by country and facility level . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 23 Figure 2. Facility readiness by country, readiness category, and facility level . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 24 Figure 3a. Corticosteroid availability versus ACS utilization by region Figure 3b. Overall readiness indices versus ACS utilization by region . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 1 Antenatal corticosteroids for pregnant women at risk of preterm labor in low- and middle-income countries: utilization and facility readiness Supplementary file . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 2 Contents Page Supplemental Table 1 Sampling strategies of included SPA surveys 3 Supplemental Table 2 The indicators used to assess facility readiness based on the 2022 WHO recommendations on ACS 4 Supplemental Table 3 Response rates and number of facilities included in the analysis for each survey 5 Supplemental Table 4 Facility structural readiness for 35 indicators by country and facility level 6 Supplemental Figure 1 Heatmap of facility readiness for all facilities by country 11 Supplemental Figure 2 Heatmap of facility readiness for level 1 facilities by country 12 Supplemental Figure 3 Heatmap of facility readiness for level 2 facilities by country 13 Supplemental Figure 4 Heatmap of facility readiness for level 3 facilities by country 14 Supplemental Figure 5 ACS utilization and corticosteroid availability facility level 15 Supplemental Figure 6 Differences in overall readiness indexes by antenatal corticosteroids utilization for all facilities 16 Supplemental Figure 7 Differences in overall readiness indexes by antenatal corticosteroids utilization for level 2 facilities 17 Supplemental Figure 8 Differences in overall readiness indexes by antenatal corticosteroids utilization for level 3 facilities 18 . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 3 Supplemental Table 1. Sampling strategies of included SPA surveys Survey Sampling strategy Afghanistan 2018-2019 The survey focused on major public and private hospitals in the main seven provinces (including Nangarhar, Paktya, Kunduz, Balkh, Kandahar, and Herat, and Kabul) in the country. Among six provinces excluding Kabul, all 12 public hospitals, 37 private hospitals, and 52 private clinics were included. In Kabul, all 26 public and 20 private hospitals were surveyed, but a randomly selected sample of 13 private clinics (out of 84) were included. Bangladesh 2017-2018 The survey adopted a stratified random sampling strategy for 1,600 facilities to obtain nationally representative data for different facility types, different facility management authorities, and each of the eight divisions (Barisal, Chittagong, Dhaka, Khulna, Mymensingh, Rajshahi, Rangpur, and Sylhet) of the country. Nepal 2021 The survey adopted a stratified random sampling strategy for 1,633 health facilities through an equal probability systematic sampling. All government hospitals, all non-government hospitals with at least one bed, and all non-government hospitals in the provinces of Karnali and Sudurpashchim were included because of their small numbers. Also, all primary health care centers (PHCCs) and stand-alone HIV testing and counseling centers (stand-alone HTCs) were included. Haiti 2017-2018 The survey was a national census of all operational health facilities in the country during the survey period of 2017 to 2018. Demographic Republic of Congo 2017-2018 The survey adopted a probability sampling strategy from 12,059 health facilities in the country, excluding health posts. On average, approximately 50 facilities were selected for each of the 26 provinces, reaching a sample of 1,380 facilities. Ethiopia 2021-2022 The survey adopted a stratified random sampling strategy reaching a sample of 1,407 health facilities via an unequal probability systematic sampling. Within each region, stratification was achieved by facility type. All public and private hospitals were included because of their small numbers and crucial roles in the country’s health system. Health centers were sampled but all health centers in Dire Dawa and Harari were included. Clinics were sampled with all higher clinics included and all clinics in Harari included. Health posts were sampled. Malawi 2013-2014 The survey was a national census of all formal-sector health facilities, including public, private and other managing authority types, in the country during the survey period between 2013 to 2014. Senegal 2018 and 2019 Senegal implemented a continuous SPA, which was designed to continuously collect data on health facilities over a period of 5 years from 2012 to 2017, aiming to survey all health facilities in the country. Tanzania 2014-2015 The survey was a sample of formal-sector health facilities, including public and private, during the survey period between 2014 to 2015. The sample was selected to obtain national representative data by facility type, managing authority type, and regions. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 4 Supplemental Table 2. The indicators used to assess facility readiness based on the 2022 WHO recommendations on ACS use Antenatal corticosteroid therapy is recommended for women with a high likelihood of preterm birth from 24 weeks to 34 weeks of gestation when the following conditions (1 to 5) are met: WHO conditions/criteria Readiness categories Indicators extracted from SPA survey SPA dataset 1. Gestational age assessment can be accurately undertaken. Readiness to accurately assess GA and identify high likelihoods of preterm labor Equipment ultrasound Inventory dataset 2. There is a high likelihood of preterm birth within 7 days of starting therapy. 3. There is no clinical evidence of maternal infection. Readiness to identify maternal infections Equipment thermometer Diagnostics hematology analyzer HIV rapid diagnostic test syphilis rapid diagnostic test 4. Adequate childbirth care is available (including capacity to recognize and safely manage preterm labor and birth). Readiness to provide adequate childbirth care Equipment delivery pack cord clamp manual vacuum extractor vacuum aspiration kit or D&C kit forceps (large) forceps (medium) Medicines and commodities parenteral antibiotics parenteral anticonvulsants (diazepam) parenteral oxytocin Staff and guidelines a health worker who can perform C/S an anesthetist national guidelines for BEmONC national guidelines for CEmONC 5. The preterm newborn can receive adequate care (including resuscitation, kangaroo mother care, thermal care, feeding support, infection treatment and respiratory support including continuous positive airway pressure as needed). Readiness to provide preterm newborn care Equipment suction bulb or penguin sucker stethoscope newborn masks (0, 1), neonatal size self-inflating bag incubator other external heat source pulse oximeter oxygen concentrator filled oxygen cylinder oxygen distribution system Medicines and commodities glucometer glucometer strips hand-washing soap disposable latex gloves Staff and guidelines staff trained on integrated Management of Pregnancy and Childbirth (IMPAC) Health worker interview dataset 1 staff trained on Comprehensive Emergency Obstetric and Newborn Care staff trained on routine care for labor and normal vaginal delivery guidelines on management of preterm labor Inventory dataset *GA: gestational age; C/S Cesarean section; BEmONC: Basic Emergency Obstetrics and Newborn Care; CEmONC: Comprehensive Emergen cy Obstetrics and Newborn Care 1 To report readiness at the facility level when using data from health worker interview, we converted data from health worker l evel to facility level by measuring the facility with at least one health worker who had received relevant training. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 5 Supplemental Table 3. Survey response rates and number of facilities sampled, surveyed, and included in the analysis Country Survey year Response rate1 Number of facilities sampled Number of facilities surveyed4 Number (%) of facilities included in the analysis5 median 94.9% (range 88.8%, 99.0%) N =10418 median 1130 (range 160, 1626) N = 9793 median 1158 (range 142, 1576) N = 8669 median 929 (range 108, 1500) South Asia Afghanistan 2018-19 88.8%2 160 142 108 (76.1%) Bangladesh 2017-18 95.3% 1600 1524 1498 (98.3%) Nepal 2021 97.0% 1626 1576 1500 (95.2%) Caribbean Haiti 2017-18 97.5% 1033 1007 929 (92.3%) Sub-Saharan Africa DRC 2017-18 97.7% 1412 1380 1364 (98.8%) Ethiopia 2021-22 82.3% 1407 1158 911 (78.7%) Malawi 2013-14 92.2% 1060 977 645 (66.1%) Senegal 2018 89.3% 466 841 (2018: 416; 2019: 425) 601 (71.5%) Senegal 2019 94.5%3 454 Tanzania 2014-15 99.0% 1200 1188 1056 (88.9%) 1 Response rates were obtained from SPA final reports. 2 Afghanistan 2018-2019 survey does not have a publicly available SPA final report. Its response rate was manually calculated. 3 The response rate for Senegal 2019 from its final report is 94.5%, calculated by excluding health huts. The rate is 93.6% (425/454) if including health huts. 4 The numbers of facilities surveyed are smaller than the numbers of facilities sampled because of non-response, refusal, or closure. 5 Facilities included in the analysis are facilities that either provided antenatal care, performed normal delivery, or performed Cesarean sections. The proportion was calculated by dividing the number of facilities included in the analysis by the number of facilities surveyed. . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 6 Supplemental Table 4. Facility structural readiness for 35 indicators by country and facility level Four readiness categories based on five WHO criteria The proportion of facilities with the indicators available South Asia Afghanistan 2018-19 Bangladesh 2017-18 Median (range)1 Overall Level 1 Level 2 Level 3 Overall Level 1 Level 2 Level 3 1. Assess GA accurately and identify imminent preterm labor Equipment The proportion of facilities with a functional ultrasound machine in use 7.3% (0.2%, 12.3%) 92(82.4%) 3(41.2%) 5(67%) 84(86.2%) 178(3.6%) 13(0.7%) 19(1.7%) 92(82.4%) 2. Identify maternal infections Equipment The proportion of facilities with this equipment thermometer 52.3% (20.7%, 76.2%) 65(58.6%) 0(0%) 6(75%) 59(59.4%) 734(20.7%) 0(0%) 481(85.2%) 65(58.6%) Diagnostics The proportion of facilities with the following diagnostics (1) hematology analyzer 7.6% (2.7%, 30.5%) 94(92%) 6(86.3%) 5(70.1%) 83(94.8%) 131(2.7%) 11(0.5%) 16(1.7%) 94(92%) (2) HIV rapid diagnostic test 52.4% (21.2%, 82.1%) 37(25.8%) 2(27.5%) 1(7.2%) 34(27.9%) 37(25.8%) (3) syphilis rapid diagnostic test 13.8% (3.0%, 68.8%) 30(21.9%) 2(27.5%) 1(7.2%) 27(23.3%) 128(3%) 18(1.1%) 22(2.7%) 30(21.9%) 3. Provide adequate childbirth care Equipment The proportion of facilities with the following equipment to perform CEmONC (obstetric related) (1) delivery pack 62.5% (19.0%, 87.5%) 80(74.5%) 0(0%) 5(28.6%) 75(83.4%) 692(19%) 0(0%) 449(77.5%) 80(74.5%) (2) cord clamp 55.8% (15.8%, 82.1%) 78(70.3%) 0(0%) 6(37.3%) 72(77.5%) 566(15.8%) 0(0%) 358(63.6%) 78(70.3%) (3) manual vacuum extractor 9.1% (4.7%, 33.9%) 76(69.1%) 0(0%) 3(17.4%) 73(78.4%) 214(4.8%) 0(0%) 90(14.7%) 76(69.1%) (4) vacuum aspiration kit or D&C kit 10.9% (5.8%, 49.4%) 74(60.5%) 0(0%) 6(37%) 68(66.1%) 303(5.8%) 0(0%) 125(14.5%) 74(60.5%) (5) forceps (large) 32.2% (5.4%, 73.5%) 85(74.9%) 0(0%) 10(100%) 75(75.6%) 718(19.7%) 0(0%) 473(80.4%) 85(74.9%) (6) forceps (medium) 34.8% (5.2%, 82.2%) 84(74.3%) 0(0%) 10(100%) 74(74.8%) 697(19.4%) 0(0%) 451(79.4%) 84(74.3%) Medicines and commodities The proportion of facilities with the following medicines (1) parenteral antibiotics 31.4% (4.6%, 51.3%) 45(42%) 0(0%) 2(50%) 43(43%) 281(4.6%) 0(0%) 82(7.6%) 45(42%) (2) parenteral anticonvulsants (diazepam) 31.8% (4.7%, 66.6%) 49(45.1%) 0(0%) 2(50%) 47(46.7%) 264(4.7%) 0(0%) 73(8.6%) 49(45.1%) (3) parenteral oxytocin 56.4% (7.5%, 79.2%) 78(74.4%) 0(0%) 5(32.2%) 73(82.8%) 401(7.5%) 0(0%) 178(21.6%) 78(74.4%) Staff and guidelines The proportion of facilities that has the following staff (1) a health worker who can perform C/S 4.0% (2.2%, 20.5%) 46(32.5%) 0(0%) 0(0%) 46(37.8%) 158(2.5%) 0(0%) 0(0%) 46(32.5%) (2) an anesthetist 3.8% (1.9%, 12.3%) 36(26.6%) 0(0%) 0(0%) 36(31%) 118(1.9%) 0(0%) 0(0%) 36(26.6%) The proportion of facilities with the following guidelines (1) national guidelines for BEmONC 18.2% (2.8%, 40.9%) 9(5.1%) 0(0%) 0(0%) 9(5.9%) 132(2.8%) 0(0%) 80(10.9%) 9(5.1%) (2) national guidelines for CEmONC 13.4% (2.2%, 41.9%) 9(5.1%) 0(0%) 0(0%) 9(5.9%) 114(2.2%) 0(0%) 65(8.7%) 9(5.1%) 4. Provide adequate preterm newborn care Equipment The proportion of facilities with the following equipment for neonatal resuscitation (1) suction bulb or penguin sucker 46.2% (15.8%, 80.4%) 78(81.6%) 0(0%) 7(79.4%) 71(85.7%) 652(15.8%) 0(0%) 414(62.6%) 78(81.6%) (2) stethoscope (in general) 51.1% (23.1%, 81.3%) 87(87.4%) 0(0%) 5(69.7%) 82(93.7%) 804(23.2%) 0(0%) 540(97.6%) 87(87.4%) (3) newborn masks, neonatal size self- inflating bag 37.3% (13.1%, 74.4%) 83(77.3%) 0(0%) 6(37%) 77(85.7%) 601(13.1%) 0(0%) 358(48%) 83(77.3%) The proportion of facilities with equipment for thermal care (1) incubator 4.5% (2.2%, 6.6%) 54(55.4%) 0(0%) 2(11.1%) 52(63.3%) 127(2.2%) 0(0%) 29(2.7%) 54(55.4%) (2) other external heat source 14.6% (4.2%, 58.4%) 68(64.1%) 0(0%) 3(16.7%) 65(72.7%) 176(4.2%) 0(0%) 61(10.7%) 68(64.1%) The proportion of facilities with equipment for respiratory care, including safe oxygen use (1) pulse oximeter 8.8% (2.8%, 42.0%) 49(53.7%) 1(13.7%) 1(5.6%) 47(61.2%) 142(3%) 7(0.4%) 18(1.1%) 49(53.7%) (2) oxygen concentrator 9.2% (2.4%, 15.2%) 40(48.2%) 0(0%) 0(0%) 40(56.1%) 192(3.2%) 6(0.3%) 69(5.8%) 40(48.2%) (3) filled oxygen cylinder 9.8% (1.8%, 22.5%) 82(84.9%) 3(45.1%) 7(82.4%) 72(87.1%) 334(6.6%) 23(1%) 122(13.2%) 82(84.9%) (4) oxygen distribution system 3.5% (1.6%, 8.5%) 16(13.3%) 2(27.5%) 1(5.6%) 13(13.5%) 84(1.6%) 2(0.1%) 26(2%) 16(13.3%) Medicines and commodities The proportion of facilities with commodities for monitoring blood glucose (1) glucometer 22.2% (19.3%, 79.7%) 42(37%) 0(0%) 3(55.5%) 39(36.6%) 310(21.1%) 124(21.7%) 64(10.6%) 42(37%) (2) glucometer strips 20.6% (17.4%, 82.0%) 37(34%) 0(0%) 3(55.5%) 34(33.1%) 303(21.3%) 126(22.2%) 57(9.9%) 37(34%) The proportion of facilities with medicines and commodities for infection management (1) hand-washing soap 50.8% (45.3%, 88.5%) 71(62.6%) 7(100%) 4(22.7%) 60(65.5%) 1075(72.8%) 473(72.1%) 406(77.3%) 71(62.6%) (2) disposable latex gloves 89.8% (69.4%, 96.4%) 102(88.9%) 7(100%) 8(50.5%) 87(92.9%) 1158(73.7%) 458(70.3%) 472(84.7%) 102(88.9%) Staff and guidelines The proportion of facilities with at least one health worker who has received training about the following in the past 24 months (1) Integrated Management of Pregnancy and Childbirth (IMPAC) 18.6% (2.4%, 37.8%) 15(12.9%) 1(13.7%) 0(0%) 14(14.3%) 116(15%) 32(16.6%) 45(11.7%) 15(12.9%) (2) Comprehensive Emergency Obstetric and Newborn Care (CEmONC) 17.6% (2.0%, 29.6%) 7(4.2%) 0(0%) 0(0%) 7(4.9%) 84(10.6%) 19(12.3%) 32(6.1%) 7(4.2%) (3) Routine care for labor and normal vaginal delivery 24.0% (2.8%, 43.5%) 16(13.5%) 1(13.7%) 0(0%) 15(15.1%) 153(18.2%) 32(20.5%) 78(14.8%) 16(13.5%) The proportion of facilities with the following guidelines guidelines on management of preterm labor 10.2% (3.2%, 34.7%) 9(5.1%) 0(0%) 0(0%) 9(5.9%) 140(3.2%) 0(0%) 85(13.6%) 9(5.1%) 1 Median and range were obtained from data across 8 countries (excluding Afghanistan). * : data unavailable . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 7 Supplemental Table 4. Facility structural readiness for 35 indicators by country and facility level (continued) Four readiness categories based on five WHO criteria The proportion of facilities with the indicators available South Asia Caribbean Nepal 2021 Haiti 2017-2018 Overall Level 1 Level 2 Level 3 Overall Level 1 Level 2 Level 3 1. Assess GA accurately and identify imminent preterm labor Equipment The proportion of facilities with a functional ultrasound machine in use 422(12.3%) 61(5.3%) 135(10.3%) 226(93.3%) 100(10.8%) 18(3.2%) 14(5.5%) 68(63.5%) 2. Identify maternal infections Equipment The proportion of facilities with this equipment thermometer 731(48%) 0(0%) 515(91.7%) 216(90.1%) 310(33.4%) 0(0%) 227(89%) 83(77.5%) Diagnostics The proportion of facilities with the following diagnostics (1) hematology analyzer 309(8.6%) 37(3.8%) 70(5.4%) 202(79.3%) 283(30.5%) 123(21.7%) 88(34.5%) 72(67.3%) (2) HIV rapid diagnostic test 197(21.2%) 97(17.1%) 68(26.6%) 32(29.9%) (3) syphilis rapid diagnostic test 304(12.8%) 69(7.2%) 149(16.2%) 86(32.4%) 194(20.9%) 106(18.7%) 60(23.5%) 28(26.2%) 3. Provide adequate childbirth care Equipment The proportion of facilities with the following equipment to perform CEmONC (obstetric related) (1) delivery pack 767(50.6%) 0(0%) 549(97.2%) 218(88.6%) 283(30.5%) 0(0%) 198(77.6%) 85(79.4%) (2) cord clamp 714(46.8%) 0(0%) 505(89.4%) 209(87.1%) 314(33.8%) 0(0%) 218(85.5%) 96(89.6%) (3) manual vacuum extractor 313(12.1%) 0(0%) 137(17.6%) 176(70.8%) 51(5.5%) 0(0%) 20(7.8%) 31(28.9%) (4) vacuum aspiration kit or D&C kit 326(10.9%) 0(0%) 138(14.5%) 188(76.6%) 100(10.8%) 0(0%) 55(21.6%) 45(42%) (5) forceps (large) 640(40.6%) 0(0%) 441(76.7%) 199(83.9%) 50(5.4%) 0(0%) 17(6.7%) 33(30.8%) (6) forceps (medium) 707(46.1%) 0(0%) 499(88%) 208(87.1%) 49(5.3%) 0(0%) 20(7.9%) 29(27.1%) Medicines and commodities The proportion of facilities with the following medicines (1) parenteral antibiotics 575(34.4%) 0(0%) 388(64%) 187(79%) 179(19.3%) 0(0%) 109(42.8%) 70(65.3%) (2) parenteral anticonvulsants (diazepam) 358(14.7%) 0(0%) 175(22.4%) 183(76.8%) 112(12.1%) 0(0%) 67(26.3%) 45(41.9%) (3) parenteral oxytocin 760(50.5%) 0(0%) 547(97.4%) 213(86.7%) 257(27.7%) 0(0%) 173(67.9%) 84(78.4%) Staff and guidelines The proportion of facilities that has the following staff (1) a health worker who can perform C/S 166(3.5%) 0(0%) 0(0%) 166(65.6%) 54(5.8%) 0(0%) 0(0%) 54(50.4%) (2) an anesthetist 152(3.2%) 0(0%) 0(0%) 152(60.4%) 43(4.6%) 0(0%) 0(0%) 43(40.1%) The proportion of facilities with the following guidelines (1) national guidelines for BEmONC (2) national guidelines for CEmONC 124(13.4%) 0(0%) 88(34.5%) 36(33.6%) 4. Provide adequate preterm newborn care Equipment The proportion of facilities with the following equipment for neonatal resuscitation (1) suction bulb or penguin sucker 393(22.3%) 0(0%) 252(40.4%) 141(61.2%) 325(35%) 0(0%) 226(88.6%) 99(92.5%) (2) stethoscope (in general) 771(51.1%) 0(0%) 551(98%) 220(91.9%) 335(36.1%) 0(0%) 242(94.9%) 93(86.8%) (3) newborn masks, neonatal size self-inflating bag 718(47.7%) 0(0%) 511(91.5%) 207(86.2%) 207(22.3%) 0(0%) 122(47.8%) 85(79.5%) The proportion of facilities with equipment for thermal care (1) incubator 141(4.5%) 0(0%) 42(4.7%) 99(43.6%) 59(6.4%) 0(0%) 24(9.5%) 35(32.7%) (2) other external heat source 530(33.6%) 0(0%) 354(63.2%) 176(72.8%) 160(17.2%) 0(0%) 90(35.3%) 70(65.4%) The proportion of facilities with equipment for respiratory care, including safe oxygen use (1) pulse oximeter 770(42%) 211(28.1%) 340(50.4%) 219(91.8%) 155(16.7%) 58(10.2%) 54(21.2%) 43(40.1%) (2) oxygen concentrator 307(11.7%) 37(4.1%) 145(14.7%) 125(52.3%) 100(10.8%) 23(4.1%) 40(15.7%) 37(34.6%) (3) filled oxygen cylinder 507(22.5%) 85(8.5%) 237(30.3%) 185(78.7%) 149(16.1%) 36(6.3%) 55(21.7%) 58(54.2%) (4) oxygen distribution system 128(3.2%) 11(1%) 7(0.6%) 110(45.4%) 32(3.5%) 5(0.9%) 11(4.4%) 16(14.9%) Medicines and commodities The proportion of facilities with commodities for monitoring blood glucose (1) glucometer 348(19.4%) 94(11.4%) 150(24.9%) 104(42.6%) 463(49.8%) 227(40%) 159(62.3%) 77(71.9%) (2) glucometer strips 325(17.4%) 82(10%) 138(21.9%) 105(42.9%) 424(45.6%) 201(35.4%) 146(57.2%) 77(71.9%) The proportion of facilities with medicines and commodities for infection management (1) hand-washing soap 1297(88.5%) 595(86.2%) 507(91.7%) 195(81.8%) 492(52.9%) 305(53.8%) 133(52.1%) 54(50.5%) (2) disposable latex gloves 1433(95.8%) 651(93.8%) 554(97.9%) 228(94.2%) 831(89.5%) 497(87.7%) 235(92.2%) 99(92.5%) Staff and guidelines The proportion of facilities with at least one health worker who has received training about the following in the past 24 months (1) Integrated Management of Pregnancy and Childbirth (IMPAC) 87(4.5%) 4(0.7%) 54(7.7%) 29(10.2%) 257(27.8%) 76(13.5%) 109(42.8%) 72(67.3%) (2) Comprehensive Emergency Obstetric and Newborn Care (CEmONC) 98(4.4%) 5(1.2%) 57(6.8%) 36(12.5%) 218(23.5%) 59(10.4%) 91(35.7%) 68(63.5%) (3) Routine care for labor and normal vaginal delivery 180(11.1%) 7(1%) 119(20.4%) 54(18.9%) 258(27.9%) 75(13.3%) 111(43.6%) 72(67.3%) The proportion of facilities with the following guidelines guidelines on management of preterm labor 60(6.4%) 0(0%) 40(15.6%) 20(18.6%) * : data unavailable . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 8 Supplemental Table 4. Facility structural readiness for 35 indicators by country and facility level (continued) Four readiness categories based on five WHO criteria The proportion of facilities with the indicators available Sub-Saharan Africa DRC 2017-2018 Ethiopia 2021-2022 Overall Level 1 Level 2 Level 3 Overall Level 1 Level 2 Level 3 1. Assess GA accurately and identify imminent preterm labor Equipment The proportion of facilities with a functional ultrasound machine in use 450(10.7%) 1(12.5%) 8(1.2%) 441(36.2%) 344(4.5%) 16(1.2%) 38(6.1%) 290(84.9%) 2. Identify maternal infections Equipment The proportion of facilities with this equipment thermometer 1102(76.2%) 0(0%) 377(74.6%) 725(85.9%) 568(21%) 0(0%) 262(83.5%) 306(91%) Diagnostics The proportion of facilities with the following diagnostics (1) hematology analyzer 179(5.4%) 1(12.5%) 8(2%) 170(14.1%) 364(5.7%) 17(1%) 68(12.4%) 279(82.2%) (2) HIV rapid diagnostic test 654(44.2%) 4(28.1%) 189(39.3%) 461(58.5%) 616(29.1%) 62(10.5%) 258(85.6%) 296(80.1%) (3) syphilis rapid diagnostic test 218(11.8%) 1(12.5%) 53(9.9%) 164(16.9%) 238(11.1%) 24(5.2%) 107(29.1%) 107(28.2%) 3. Provide adequate childbirth care Equipment The proportion of facilities with the following equipment to perform CEmONC (obstetric related) (1) delivery pack 1240(87.5%) 0(0%) 437(87.2%) 803(94.5%) 605(22.5%) 0(0%) 276(89.5%) 329(96.9%) (2) cord clamp 1160(79.1%) 0(0%) 409(80.1%) 751(81.9%) 625(23.4%) 0(0%) 292(93.4%) 333(97.9%) (3) manual vacuum extractor 273(7.2%) 0(0%) 28(5.1%) 245(13.4%) 518(16.5%) 0(0%) 205(62.8%) 313(92.8%) (4) vacuum aspiration kit or D&C kit 729(36.4%) 0(0%) 134(26.4%) 595(65.9%) 385(10%) 0(0%) 124(36.2%) 261(70.4%) (5) forceps (large) 127(5.9%) 0(0%) 31(5.4%) 96(7.6%) 617(23.9%) 0(0%) 286(95.4%) 331(97.4%) (6) forceps (medium) 112(5.2%) 0(0%) 21(4.6%) 91(7.1%) 611(23.4%) 0(0%) 281(93.4%) 330(96.4%) Medicines and commodities The proportion of facilities with the following medicines (1) parenteral antibiotics 549(36.4%) 0(0%) 174(34.3%) 375(44.4%) 489(16.4%) 0(0%) 204(64.6%) 285(75.7%) (2) parenteral anticonvulsants (diazepam) 785(55.4%) 0(0%) 253(52.7%) 532(66.6%) 351(10.4%) 0(0%) 121(39.3%) 230(61.6%) (3) parenteral oxytocin 1110(77.3%) 0(0%) 394(77.5%) 716(82.4%) 611(22.8%) 0(0%) 283(90.7%) 328(96.7%) Staff and guidelines The proportion of facilities that has the following staff (1) a health worker who can perform C/S 725(20.5%) 0(0%) 0(0%) 725(77.3%) 312(2.2%) 0(0%) 0(0%) 312(83.5%) (2) an anesthetist 486(12.3%) 0(0%) 0(0%) 486(46.4%) 311(2.2%) 0(0%) 0(0%) 311(82.8%) The proportion of facilities with the following guidelines (1) national guidelines for BEmONC 350(11.5%) 0(0%) 147(45%) 203(55%) (2) national guidelines for CEmONC 411(24.7%) 0(0%) 118(23.2%) 293(30.5%) 162(4.3%) 0(0%) 45(15.5%) 117(33%) 4. Provide adequate preterm newborn care Equipment The proportion of facilities with the following equipment for neonatal resuscitation (1) suction bulb or penguin sucker 1099(80.4%) 0(0%) 388(80%) 711(87.1%) 577(21.3%) 0(0%) 261(84.3%) 316(93.6%) (2) stethoscope (in general) 1144(81.3%) 0(0%) 412(81.5%) 732(86.3%) 612(23.1%) 0(0%) 290(92.1%) 322(95.1%) (3) newborn masks, neonatal size self-inflating bag 561(26.9%) 0(0%) 102(21%) 459(44.6%) 593(21.6%) 0(0%) 262(85.2%) 331(97.4%) The proportion of facilities with equipment for thermal care (1) incubator 114(3%) 0(0%) 3(0.6%) 111(9.7%) 232(4.5%) 0(0%) 60(14.4%) 172(47.1%) (2) other external heat source 411(18.2%) 0(0%) 62(12.4%) 349(35.3%) 382(9.4%) 0(0%) 118(32.7%) 264(79.5%) The proportion of facilities with equipment for respiratory care, including safe oxygen use (1) pulse oximeter 78(2.8%) 1(12.5%) 2(0.4%) 75(8.6%) 318(26.8%) 23(31.1%) 85(21%) 210(64.9%) (2) oxygen concentrator 105(2.4%) 2(14.4%) 2(0.5%) 101(6.7%) 184(11.2%) 11(11.3%) 38(6.8%) 135(46.5%) (3) filled oxygen cylinder 73(1.8%) 2(14.4%) 1(0.3%) 70(4.8%) 200(13%) 12(15.7%) 43(7.9%) 145(49%) (4) oxygen distribution system 80(4.8%) 3(7.2%) 12(2.6%) 65(17.7%) Medicines and commodities The proportion of facilities with commodities for monitoring blood glucose (1) glucometer 761(34%) 3(27.7%) 76(18.2%) 682(77%) 564(21%) 38(4.1%) 202(68.8%) 324(95.6%) (2) glucometer strips 674(29.6%) 3(27.7%) 63(15.6%) 608(67.4%) 542(19.3%) 35(3.8%) 188(62.5%) 319(94.4%) The proportion of facilities with medicines and commodities for infection management (1) hand-washing soap 686(48.4%) 7(57.8%) 228(45.6%) 451(55.6%) 487(45.3%) 116(42.7%) 160(51.8%) 211(63.3%) (2) disposable latex gloves 1256(90.2%) 10(100%) 453(88.5%) 793(93.9%) 761(69.4%) 160(61.5%) 283(93%) 318(94.1%) Staff and guidelines The proportion of facilities with at least one health worker who has received training about the following in the past 24 months (1) Integrated Management of Pregnancy and Childbirth (IMPAC) 423(34.5%) 1(41.7%) 107(33.8%) 315(35.6%) 108(2.4%) 5(0.4%) 41(6.6%) 62(24.7%) (2) Comprehensive Emergency Obstetric and Newborn Care (CEmONC) 401(29.6%) 1(41.7%) 93(28.6%) 307(31.2%) 112(2%) 4(0.3%) 38(5.6%) 70(19.3%) (3) Routine care for labor and normal vaginal delivery 415(33.6%) 1(41.7%) 107(33.5%) 307(33.8%) 126(2.8%) 7(1.1%) 44(6.5%) 75(20.6%) The proportion of facilities with the following guidelines guidelines on management of preterm labor 232(11.4%) 0(0%) 46(8.8%) 186(19%) 248(7.2%) 0(0%) 86(27.1%) 162(43.2%) * : data unavailable . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 9 Supplemental Table 4. Facility structural readiness for 35 indicators by country and facility level (continued) Four readiness categories based on five WHO criteria The proportion of facilities with the indicators available Sub-Saharan Africa Malawi 2013-2014 Senegal 2018 and 2019 Overall Level 1 Level 2 Level 3 Overall Level 1 Level 2 Level 3 1. Assess GA accurately and identify imminent preterm labor Equipment The proportion of facilities with a functional ultrasound machine in use 65(10%) 4(3.8%) 7(1.5%) 54(76%) 1(0.2%) 0(0%) 1(0.2%) 0(0%) 2. Identify maternal infections Equipment The proportion of facilities with this equipment thermometer 453(69.6%) 0(0%) 390(82.6%) 63(88.6%) 495(69.9%) 0(0%) 436(80.2%) 59(96.4%) Diagnostics The proportion of facilities with the following diagnostics (1) hematology analyzer 57(8.8%) 8(7.7%) 6(1.3%) 43(60.8%) 139(16.1%) 7(28.9%) 80(10.5%) 52(52.2%) (2) HIV rapid diagnostic test 391(60.6%) 57(55.4%) 288(61.2%) 46(64.6%) 548(78.7%) 40(40.5%) 461(86.2%) 47(72.7%) (3) syphilis rapid diagnostic test 96(14.8%) 17(16.4%) 54(11.4%) 25(35.2%) 487(68.8%) 37(39.5%) 410(77.6%) 40(30.2%) 3. Provide adequate childbirth care Equipment The proportion of facilities with the following equipment to perform CEmONC (obstetric related) (1) delivery pack 486(74.8%) 0(0%) 418(88.7%) 68(95.7%) 579(83.7%) 0(0%) 521(97.8%) 58(96%) (2) cord clamp 482(74.1%) 0(0%) 416(88.2%) 66(92.9%) 568(82.1%) 0(0%) 509(95.8%) 59(96.4%) (3) manual vacuum extractor 221(33.9%) 0(0%) 165(34.8%) 56(78.8%) 96(11%) 0(0%) 57(7.6%) 39(71.5%) (4) vacuum aspiration kit or D&C kit 131(20.1%) 0(0%) 97(20.4%) 34(48%) 343(49.4%) 0(0%) 303(55.7%) 40(79.1%) (5) forceps (large) 448(69%) 0(0%) 390(82.8%) 58(81.7%) 521(73.5%) 0(0%) 464(85%) 57(94.9%) (6) forceps (medium) 434(66.9%) 0(0%) 379(80.5%) 55(77.5%) 533(75.6%) 0(0%) 474(87.5%) 59(96.4%) Medicines and commodities The proportion of facilities with the following medicines (1) parenteral antibiotics 299(46.1%) 0(0%) 244(51.9%) 55(77.4%) 362(51.3%) 0(0%) 316(60.7%) 46(51.1%) (2) parenteral anticonvulsants (diazepam) 433(66.6%) 0(0%) 371(78.6%) 62(87.2%) 363(51.6%) 0(0%) 317(58.1%) 46(83.6%) (3) parenteral oxytocin 514(79.2%) 0(0%) 449(95.3%) 65(91.4%) 451(62.3%) 0(0%) 402(74.5%) 49(52.2%) Staff and guidelines The proportion of facilities that has the following staff (1) a health worker who can perform C/S 57(8.7%) 0(0%) 0(0%) 57(80.1%) 49(3.5%) 0(0%) 0(0%) 49(48.7%) (2) an anesthetist 47(7.2%) 0(0%) 0(0%) 47(66%) 51(3.6%) 0(0%) 0(0%) 51(50.4%) The proportion of facilities with the following guidelines (1) national guidelines for BEmONC 266(40.9%) 0(0%) 231(48.9%) 35(49.1%) (2) national guidelines for CEmONC 148(22.8%) 0(0%) 117(24.8%) 31(43.6%) 304(41.9%) 0(0%) 269(47.5%) 35(64.5%) 4. Provide adequate preterm newborn care Equipment The proportion of facilities with the following equipment for neonatal resuscitation (1) suction bulb or penguin sucker 399(61.5%) 0(0%) 346(73.6%) 53(74.5%) 480(66.4%) 0(0%) 434(76.9%) 46(84.9%) (2) stethoscope (in general) 423(65.2%) 0(0%) 369(78.4%) 54(76.1%) 369(51.1%) 0(0%) 314(56.7%) 55(92.2%) (3) newborn masks, neonatal size self-inflating bag 483(74.4%) 0(0%) 418(88.7%) 65(91.4%) 477(67.4%) 0(0%) 419(77.1%) 58(94.9%) The proportion of facilities with equipment for thermal care (1) incubator 37(5.7%) 0(0%) 8(1.7%) 29(41.1%) 41(6.6%) 0(0%) 18(3%) 23(59.4%) (2) other external heat source 78(12%) 0(0%) 31(6.5%) 47(66.3%) 407(58.4%) 0(0%) 350(65.9%) 57(94.2%) The proportion of facilities with equipment for respiratory care, including safe oxygen use (1) pulse oximeter 60(9.3%) 9(8.8%) 25(5.3%) 26(36.9%) 68(8.2%) 2(1.2%) 22(2.5%) 44(85.1%) (2) oxygen concentrator 99(15.2%) 9(8.5%) 54(11.4%) 36(50.7%) 61(7.5%) 1(0.6%) 23(2.9%) 37(73.4%) (3) filled oxygen cylinder 47(7.3%) 7(6.8%) 16(3.4%) 24(34%) 99(12.4%) 5(10.9%) 44(6%) 50(85.7%) (4) oxygen distribution system 27(4.2%) 2(1.9%) 15(3.2%) 10(14.2%) 60(8.5%) 1(9.1%) 16(2.1%) 43(78.4%) Medicines and commodities The proportion of facilities with commodities for monitoring blood glucose (1) glucometer 151(23.4%) 32(31.2%) 58(12.3%) 61(85.8%) 532(79.7%) 48(72.9%) 446(81.1%) 38(77.7%) (2) glucometer strips 129(20%) 24(23.4%) 48(10.1%) 57(80.2%) 537(82%) 46(71%) 446(84%) 45(82.5%) The proportion of facilities with medicines and commodities for infection management (1) hand-washing soap 313(48.7%) 57(55.4%) 211(44.9%) 45(63.5%) 334(48.2%) 36(44.9%) 263(50.1%) 35(34.6%) (2) disposable latex gloves 622(96.4%) 96(93%) 458(97.2%) 68(95.8%) 584(86.7%) 55(81.7%) 480(91.5%) 49(44.4%) Staff and guidelines The proportion of facilities with at least one health worker who has received training about the following in the past 24 months (1) Integrated Management of Pregnancy and Childbirth (IMPAC) 137(21.2%) 3(2.9%) 103(21.9%) 31(44.2%) 223(37.8%) 10(19%) 190(41.8%) 23(21.3%) (2) Comprehensive Emergency Obstetric and Newborn Care (CEmONC) 133(20.6%) 5(4.8%) 97(20.6%) 31(44.3%) 170(27.5%) 9(17.5%) 141(29.6%) 20(18.6%) (3) Routine care for labor and normal vaginal delivery 214(33.2%) 3(2.9%) 166(35.3%) 45(64.1%) 236(43.5%) 10(46.1%) 205(45.4%) 21(18.8%) The proportion of facilities with the following guidelines guidelines on management of preterm labor 226(34.7%) 0(0%) 190(40.2%) 36(50.6%) 130(18%) 0(0%) 107(17.7%) 23(57.6%) * : data unavailable . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 10 Supplemental Table 4. Facility structural readiness for 35 indicators by country and facility level (continued) Four readiness categories based on five WHO criteria The proportion of facilities with the indicators available Sub-Saharan Africa Tanzania 2014-2015 Overall Level 1 Level 2 Level 3 1. Assess GA accurately and identify imminent preterm labor Equipment The proportion of facilities with a functional ultrasound machine in use 223(4.6%) 6(2.9%) 23(1%) 194(65.1%) 2. Identify maternal infections Equipment The proportion of facilities with this equipment thermometer 654(56.6%) 0(0%) 417(62.7%) 237(84.6%) Diagnostics The proportion of facilities with the following diagnostics (1) hematology analyzer 233(6.7%) 12(10.1%) 52(3%) 169(58.2%) (2) HIV rapid diagnostic test 879(82.1%) 76(77.4%) 573(82.6%) 230(83%) (3) syphilis rapid diagnostic test 455(40%) 50(41.1%) 275(39.5%) 130(45.4%) 3. Provide adequate childbirth care Equipment The proportion of facilities with the following equipment to perform CEmONC (obstetric related) (1) delivery pack 838(74.4%) 0(0%) 575(83.2%) 263(97.1%) (2) cord clamp 711(64.7%) 0(0%) 487(72.4%) 224(83.3%) (3) manual vacuum extractor 176(4.7%) 0(0%) 26(2.2%) 150(54.5%) (4) vacuum aspiration kit or D&C kit 197(6.6%) 0(0%) 77(5.1%) 120(45.7%) (5) forceps (large) 821(72.1%) 0(0%) 568(80.7%) 253(93.3%) (6) forceps (medium) 896(82.2%) 0(0%) 634(92.5%) 262(96.9%) Medicines and commodities The proportion of facilities with the following medicines (1) parenteral antibiotics 363(28.4%) 0(0%) 211(30.7%) 152(54.4%) (2) parenteral anticonvulsants (diazepam) 580(48.9%) 0(0%) 369(53.9%) 211(76.7%) (3) parenteral oxytocin 818(69.8%) 0(0%) 562(77.9%) 256(93.2%) Staff and guidelines The proportion of facilities that has the following staff (1) a health worker who can perform C/S 238(4.4%) 0(0%) 0(0%) 238(83.6%) (2) an anesthetist 215(3.9%) 0(0%) 0(0%) 215(74.3%) The proportion of facilities with the following guidelines (1) national guidelines for BEmONC 368(25%) 0(0%) 249(27.3%) 119(43.2%) (2) national guidelines for CEmONC 134(7.9%) 0(0%) 79(8.1%) 55(20.7%) 4. Provide adequate preterm newborn care Equipment The proportion of facilities with the following equipment for neonatal resuscitation (1) suction bulb or penguin sucker 652(57.4%) 0(0%) 427(63.9%) 225(80.7%) (2) stethoscope (in general) 732(64%) 0(0%) 490(71.4%) 242(86.4%) (3) newborn masks, neonatal size self-inflating bag 772(67.4%) 0(0%) 512(75%) 260(94.8%) The proportion of facilities with equipment for thermal care (1) incubator 97(2.4%) 0(0%) 17(1.1%) 80(28%) (2) other external heat source 188(5.6%) 0(0%) 41(3.4%) 147(52.5%) The proportion of facilities with equipment for respiratory care, including safe oxygen use (1) pulse oximeter 84(2.8%) 8(10.1%) 9(0.4%) 67(25.3%) (2) oxygen concentrator 159(5.2%) 11(12.2%) 45(2.1%) 103(38.7%) (3) filled oxygen cylinder 105(3.7%) 8(11.1%) 20(1%) 77(29.7%) (4) oxygen distribution system 51(2.1%) 5(5.5%) 12(0.9%) 34(12.9%) Medicines and commodities The proportion of facilities with commodities for monitoring blood glucose (1) glucometer 453(19.3%) 40(42.2%) 181(12%) 232(86%) (2) glucometer strips 395(17.8%) 35(39.3%) 147(11%) 213(78.4%) The proportion of facilities with medicines and commodities for infection management (1) hand-washing soap 697(63.5%) 67(64.1%) 434(63.1%) 196(70%) (2) disposable latex gloves 996(94%) 98(94.1%) 640(94%) 258(94.9%) Staff and guidelines The proportion of facilities with at least one health worker who has received training about the following in the past 24 months (1) Integrated Management of Pregnancy and Childbirth (IMPAC) 228(16%) 3(5.1%) 138(16.6%) 87(31.5%) (2) Comprehensive Emergency Obstetric and Newborn Care (CEmONC) 252(14.7%) 2(5%) 146(14.6%) 104(36.8%) (3) Routine care for labor and normal vaginal delivery 308(20.1%) 6(9.2%) 185(20.2%) 117(42.6%) The proportion of facilities with the following guidelines guidelines on management of preterm labor 153(10.2%) 0(0%) 83(10.7%) 70(24.9%) * : data unavailable . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 11 Supplemental Figure 1. Heatmap of facility readiness for all facilities by country . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 12 Supplemental Figure 2. Heatmap of facility readiness for level 1 facilities by country . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 13 Supplemental Figure 3. Heatmap of facility readiness for level 2 facilities by country . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 14 Supplemental Figure 4. Heatmap of facility readiness for level 3 facilities by country . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 15 Supplemental Figure 5. ACS utilization and corticosteroid availability by facility level . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 16 Supplemental Figure 6. Differences in overall readiness indexes by antenatal corticosteroids utilization for all facilities . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 17 Supplemental Figure 7. Differences in overall readiness indexes by antenatal corticosteroids utilization for level 2 facilities . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint 18 Supplemental Figure 8. Differences in overall readiness indexes by antenatal corticosteroids utilization for level 3 facilities . CC-BY-NC-ND 4.0 International licenseIt is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) The copyright holder for this preprint this version posted August 2, 2024. ; https://doi.org/10.1101/2024.07.31.24310863doi: medRxiv preprint