Urban school neighbourhoods dominated by unhealthy food retailers and advertisements in Greater Tunis: a geospatial study in the midst of the nutrition transition.

OBJECTIVE: Food environments are a major determinant of children's nutritional status. Scarce evidence on food environments exists in low- and middle-income countries (LMIC). This study aims to fill this gap by documenting the obesogenicity of food environments around schools in Greater Tunis, Tunisia - an LMIC of the Middle East and North Africa region with an ongoing nutrition transition and increasing rates of childhood obesity. DESIGN: In this cross-sectional study, we assessed built food environments around fifty primary schools. Ground-truthing was performed to collect geographic coordinates and pictures of food retailers and food advertisement sets within an 800-m road network buffer of each school. Retailers and advertisement sets were categorised as healthy or unhealthy according to a NOVA-based classification. Associations between school characteristics and retailers or advertisement sets were explored using multinomial regression models. SETTING: Greater Tunis, Tunisia. PARTICIPANTS: Random sample of fifty (thirty-five public and fifteen private) primary schools. RESULTS: Overall, 3621 food retailers and 2098 advertisement sets were mapped. About two-thirds of retailers and advertisement sets were labelled as unhealthy. Most retailers were traditional corner stores (22 %) and only 6 % were fruit and vegetable markets. The prevailing food group promoted was carbonated and sugar-sweetened beverages (22 %). The proportion of unhealthy retailers was significantly higher in the richest v. poorest areas. CONCLUSIONS: School neighbourhood food environments included predominantly unhealthy retailers and advertisements. Mapping of LMIC food environments is crucial to document the impact of the nutrition transition on children's nutritional status. This will inform policies and interventions to curb the emergent childhood obesity epidemic.

Perspectives and experiences of healthcare providers on the response to the COVID-19 pandemic in three maternal and neonatal referral hospitals in Guinea in 2020: a qualitative study.

BACKGROUND: The COVID-19 pandemic has adversely affected access to essential healthcare services. This study aimed to explore healthcare providers' perceptions and experiences of the response to the COVID-19 pandemic in three referral maternal and neonatal hospitals in Guinea. METHODS: We conducted a longitudinal qualitative study between June and December 2020 in two maternities and one neonatology referral ward in Conakry and Mamou. Participants were purposively recruited to capture diversity of professional cadres, seniority, and gender. Four rounds of in-depth interviews (46 in-depth interviews with 18 respondents) were conducted in each study site, using a semi-structured interview guide that was iteratively adapted. We used both deductive and inductive approaches and an iterative process for content analysis. RESULTS: We identified four themes and related sub-themes presented according to whether they were common or specific to the study sites, namely: 1) coping strategies & care reorganization, which include reducing staffing levels, maintaining essential healthcare services, suspension of staff daily meetings, insertion of a new information system for providers, and co-management with COVID-19 treatment center for caesarean section cases among women who tested positive for COVID-19; 2) healthcare providers' behavior adaptations during the response, including infection prevention and control measures on the wards and how COVID-19-related information influenced providers' daily work; 3) difficulties encountered by providers, in particular unavailability of personal protective equipment (PPE), lack of financial motivation, and difficulties reducing crowding in the wards; 4) providers perceptions of healthcare service use, for instance their fear during COVID-19 response and perceived increase in severity of complications received and COVID-19 cases among providers and parents of newborns. CONCLUSION: This study provides insights needed to be considered to improve the preparedness and response of healthcare facilities and care providers to future health emergencies in similar contexts.

Wealth-based inequality in the continuum of maternal health service utilisation in 16 sub-Saharan African countries.

BACKGROUND: Persistent inequalities in coverage of maternal health services in sub-Saharan Africa (SSA), a region home to two-thirds of global maternal deaths in 2017, poses a challenge for countries to achieve the Sustainable Development Goal (SDG) targets. This study assesses wealth-based inequalities in coverage of maternal continuum of care in 16 SSA countries with the objective of informing targeted policies to ensure maternal health equity in the region. METHODS: We conducted a secondary analysis of Demographic and Health Survey (DHS) data from 16 SSA countries (Angola, Benin, Burundi, Cameroon, Ethiopia, Gambia, Guinea, Liberia, Malawi, Mali, Nigeria, Sierra Leone, South Africa, Tanzania, Uganda, and Zambia). A total of 133,709 women aged 15-49 years who reported a live birth in the five years preceding the survey were included. We defined and measured completion of maternal continuum of care as having had at least one antenatal care (ANC) visit, birth in a health facility, and postnatal care (PNC) by a skilled provider within two days of birth. We used concentration index analysis to measure wealth-based inequality in maternal continuum of care and conducted decomposition analysis to estimate the contributions of sociodemographic and obstetric factors to the observed inequality. RESULTS: The percentage of women who had 1) at least one ANC visit was lowest in Ethiopia (62.3%) and highest in Burundi (99.2%), 2) birth in a health facility was less than 50% in Ethiopia and Nigeria, and 3) PNC within two days was less than 50% in eight countries (Angola, Burundi, Ethiopia, Gambia, Guinea, Malawi, Nigeria, and Tanzania). Completion of maternal continuum of care was highest in South Africa (81.4%) and below 50% in nine of the 16 countries (Angola, Burundi, Ethiopia, Guinea, Malawi, Mali, Nigeria, Tanzania, and Uganda), the lowest being in Ethiopia (12.5%). There was pro-rich wealth-based inequality in maternal continuum of care in all 16 countries, the lowest in South Africa and Liberia (concentration index = 0.04) and the highest in Nigeria (concentration index = 0.34). Our decomposition analysis showed that in 15 of the 16 countries, wealth index was the largest contributor to inequality in primary maternal continuum of care. In Malawi, geographical region was the largest contributor. CONCLUSIONS: Addressing the coverage gap in maternal continuum of care in SSA using multidimensional and people-centred approaches remains a key strategy needed to realise the SDG3. The pro-rich wealth-based inequalities observed show that bespoke pro-poor or population-wide approaches are needed.

Length of stay and determinants of early discharge after facility-based childbirth in Cameroon: analysis of the 2018 Demographic and Health Survey.

BACKGROUND: A minimum length of stay following facility birth is a prerequisite for women and newborns to receive the recommended monitoring and package of postnatal care. The first postnatal care guidelines in Cameroon were issued in 1998 but adherence to minimum length of stay has not been assessed thus far. The objective of this study was to estimate the average length of stay and identify determinants of early discharge after facility birth. METHODS: We analyzed the Cameroon 2018 Demographic and Health Survey. We included 4,567 women who had a live birth in a heath facility between 2013 and 2018. We calculated their median length of stay in hours by mode of birth and the proportion discharged early (length of stay < 24 h after vaginal birth or < 5 days after caesarean section). We assessed the association between sociodemographic, context-related, facility-related, obstetric and need-related factors and early discharge using bivariate and multivariable logistic regression. RESULTS: The median length of stay (inter quartile range) was 36 (9-84) hours after vaginal birth (n = 4,290) and 252 (132-300) hours after caesarean section (n = 277). We found that 28.8% of all women who gave birth in health facilities were discharged too early (29.7% of women with vaginal birth and 15.1% after a caesarean section). Factors which significantly predicted early discharge in multivariable regression were: maternal age < 20 years (compared to 20-29 years, aOR: 1.44; 95%CI 1.13-1.82), unemployment (aOR: 0.78; 95%CI: 0.63-0.96), non-Christian religions (aOR: 1.65; 95CI: 1.21-2.24), and region of residence-Northern zone aOR:9.95 (95%CI:6.53-15.17) and Forest zone aOR:2.51 (95%CI:1.79-3.53) compared to the country's capital cities (Douala or Yaounde). None of the obstetric characteristics was associated with early discharge. CONCLUSIONS: More than 1 in 4 women who gave birth in facilities in Cameroon were discharged too early; this mostly affected women following vaginal birth. The reasons leading to lack of adherence to postnatal care guidelines should be better understood and addressed to reduce preventable complications and provide better support to women and newborns during this critical period.

Insights into barriers and facilitators in PrEP uptake and use among migrant men and transwomen who have sex with men in Belgium.

BACKGROUND: PrEP uptake is low among non-Belgian men and transwomen who have sex with men, although the HIV epidemic among men who have sex with men in Belgium is diversifying in terms of nationalities and ethnicity. We lack an in-depth understanding of this gap. METHODS: We conducted a qualitative study using a grounded theory approach. The data consists of key informants interviews and in-depth interviews with migrant men or transwomen who have sex with men. RESULTS: We identified four underlying determinants which shape our participants' experiences and contextualize the barriers to PrEP use. These include (1) the intersectional identities of being migrant and men and transwomen who have sex with men, (2) migration related stressors, (3) mental health and (4) socio-economic vulnerability. Identified barriers include: the accessibility of services; availability of information, social resources and providers' attitudes. These barriers influence PrEP acceptance and mediated by individual agency this influences their PrEP uptake. CONCLUSION: An interplay of several underlying determinants and barriers impacts on PrEP uptake among migrant men and transwomen who have sex with men, illustrating a social gradient in access to PrEP. We need equitable access to the full spectrum of HIV prevention and care for all priority populations, including undocumented migrants. We recommend social and structural conditions that foster exercising these rights, including adapting PrEP service delivery, mental health and social support.

Effect of the COVID-19 pandemic on maternal and neonatal health services in three referral hospitals in Guinea: an interrupted time-series analysis.

INTRODUCTION: In sub-Saharan Africa, there is limited evidence on the COVID-19 health-related effect from front-line health provision settings. Therefore, this study aimed to analyse the effect of the COVID-19 pandemic on routine maternal and neonatal health services in three referral hospitals. MATERIALS AND METHODS: We conducted an observational study using aggregate monthly maternal and neonatal health services routine data for two years (March 2019-February 2021) in three referral hospitals including two maternities: Hôpital National Ignace Deen (HNID) in Conakry and Hôpital Regional de Mamou (HRM) in Mamou and one neonatology ward: Institut de Nutrition et de Santé de l'Enfant (INSE) in Conakry. We compared indicators of health service utilisation, provision and health outcomes before and during the COVID-19 pandemic periods. An interrupted time-series analysis (ITSA) was performed to assess the relationship between changes in maternal and neonatal health indicators and COVID-19 through cross-correlation. RESULTS: During COVID-19, the mean monthly number (MMN) of deliveries decreased significantly in HNID (p = 0.039) and slightly increased in HRM. In the two maternities, the change in the MMN of deliveries were significantly associated with COVID-19. The ITSA confirmed the association between the increase in the MMN of deliveries and COVID-19 in HRM (bootstrapped F-value = 1.46, 95%CI [0.036-8.047], p < 0.01). We observed an increasing trend in obstetric complications in HNID, while the trend declined in HRM. The MMN of maternal deaths increased significantly (p = 0.011) in HNID, while it slightly increased in HRM. In INSE, the MMN of neonatal admissions significantly declined (p < 0.001) and this decline was associated with COVID-19. The MMN of neonatal deaths significantly decreased (p = 0.009) in INSE and this decrease was related to COVID-19. CONCLUSION: The pandemic negatively affected the maternal and neonatal care provision, health service utilisation and health outcomes in two referral hospitals located in Conakry, the COVID-19 most-affected region.

Can machine learning models predict maternal and newborn healthcare providers' perception of safety during the COVID-19 pandemic? A cross-sectional study of a global online survey.

BACKGROUND: Maternal and newborn healthcare providers are essential professional groups vulnerable to physical and psychological risks associated with the COVID-19 pandemic. This study uses machine learning algorithms to create a predictive tool for maternal and newborn healthcare providers' perception of being safe in the workplace globally during the pandemic. METHODS: We used data collected between 24 March and 5 July 2020 through a global online survey of maternal and newborn healthcare providers. The questionnaire was available in 12 languages. To predict healthcare providers' perception of safety in the workplace, we used features collected in the questionnaire, in addition to publicly available national economic and COVID-19-related factors. We built, trained and tested five machine learning models: Support Vector Machine (SVM), Random Forest (RF), XGBoost, CatBoost and Artificial Neural Network (ANN) for classification and regression. We extracted from RF models the relative contribution of features in output prediction. RESULTS: Models included data from 941 maternal and newborn healthcare providers from 89 countries. ML models performed well in classification and regression tasks, whereby RF had 82% cross-validated accuracy for classification, and CatBoost with 0.46 cross-validated root mean square error for regression. In both classification and regression, the most important features contributing to output prediction were classified as three themes: (1) information accessibility, clarity and quality; (2) availability of support and means of protection; and (3) COVID-19 epidemiology. CONCLUSION: This study identified salient features contributing to maternal and newborn healthcare providers perception of safety in the workplace. The developed tool can be used by health systems globally to allow real-time learning from data collected during a health system shock. By responding in real-time to the needs of healthcare providers, health systems could prevent potential negative consequences on the quality of care offered to women and newborns.

Financial relationships between patient and consumer representatives and the health industry: A systematic review.

BACKGROUND: Patients and consumers are increasingly engaged in health policymaking, research and drug regulation. Having financial relationships with the health industry creates situations of conflicts of interest (COI) and might compromise their meaningful and unbiased participation. OBJECTIVE: To synthesize available evidence on the financial relationships between the health industry and patient and consumer representatives and their organizations. METHODS: We systematically searched MEDLINE and EMBASE. We selected studies and abstracted data in duplicate and independently. We reported on outcomes related to financial relationships of individuals with, and/or funding of organizations by the health industry. RESULTS: We identified a total of 14 510 unique citations, of which 24 reports of 23 studies were eligible. Three studies (13%) addressed the financial relationship of patient and consumer representatives with the health industry. Of these, two examined the proportion of public speakers in drug regulatory processes who have financial relationships; the proportions in the two studies were 25% and 19% respectively. Twenty studies (87%) addressed funding of patient and consumer organizations. The median proportion of organizations that reported funding from the health industry was 62% (IQR: 34%-69%) in questionnaire surveys, and 75% (IQR: 58%-85%) in surveys of their websites. Among organizations for which there was evidence of industry funding, a median proportion of 29% (IQR: 27%-44%) acknowledged on their websites receiving that funding. CONCLUSION: Financial relationships between the health industry and patient and consumer representatives and their organizations are common and may not be disclosed. Stricter regulation on disclosure and management is needed.

Authorship in reports of clinical practice guidelines: A systematic cross-sectional analysis.

BACKGROUND: A transparent and explicit reporting on authors' contributions to the development of clinical practice guidelines and on panelists' characteristics is essential for their credibility and trustworthiness. We did not find published studies on authorship or panel involvement in clinical practice guidelines. OBJECTIVE: To describe the approach to authorship in reports of clinical practice guidelines, and the characteristics of individual authors. METHODS: We conducted a cross-sectional survey of guidelines listed in the National Guideline Clearing House (NGC) in 2016. We abstracted data on the general characteristics of the guidelines, report approach to authorship, and individual authors characteristics. Data abstraction was in duplicate and independent manner using standardised form. Data analyses were both descriptive and regression analyses. RESULTS: Overall, 139 eligible guidelines with published papers were identified. Of these, 48 (35%) included a group authorship statement in the author byline. A third of these guidelines (n = 45; 32%) reported on authors' contributions, while about half of the guidelines (n = 74; 53%) reported who of the authors served as panel members. Around one-fifth of the guidelines (n = 30; 22%) reported group membership (eg, content expert, patient representative) for at least 1 author. Less than one-seventh of the eligible guidelines indicated who selected the panel members (n = 18; 13%), reported the types of panel members (n = 18; 13%) or the selection criteria (n = 12; 9%). Higher journal impact factor was associated with both "reporting of the author contributions" (OR = 1.07) and "the inclusion of a panel membership section in the guideline report" (OR = 1.21). CONCLUSION: Low percentages of clinical practice guidelines report information on important aspects of authorship and characteristics of individual authors. Better reporting of some of these criteria was associated with journal impact factor.

Public health journals' requirements for authors to disclose funding and conflicts of interest: a cross-sectional study.

BACKGROUND: Public health journals need to have clear policies for reporting the funding of studies and authors' personal financial and non-financial conflicts of interest (COI) disclosures. This study aims to assess the policies of public health journals on reporting of study funding and the disclosure of authors' COIs. METHODS: This is a cross-sectional study of "Public, Environmental & Occupational Health" journals. Teams of two researchers abstracted data in duplicate and independently using REDCap software. RESULTS: Of 173 public health journals, 155 (90%) had a policy for reporting study funding information. Out of these, a majority did not require reporting of the phase of the study for which funding was received (88%), nor the types of funding sources (87%). Of the 173 journals, 163 (94%) had a policy requiring disclosure of authors' COI. However, the majority of these journals did not require financial conflicts of interest disclosures relating to institutions (75%) nor to the author's family members (90%) while 56% required the disclosure of at least one form of non-financial COI. CONCLUSIONS: The policies of the majority of public health journals do not require the reporting of important details such as the role of the funder, and non-financial COI. Journals and publishers should consider revising their editorial policies to ensure complete and transparent reporting of funding and COI.

Uncovering community needs regarding violence against women and girls in southern Ethiopia: An explorative study.

BACKGROUND: Violence against women and girls (VAWG) is a significant global public health problem and a violation of human rights experienced by one in three women worldwide. This study explores community perceptions of and responses to VAWG and challenges in accessing support services among female violence survivors in Arbaminch City. METHODS: We adopted a phenomenological explorative qualitative study design. A total of 62 participants including female violence survivors, religious leaders, service providers, police, women, and men in participated in interviews, focus group discussions, and observations in August 2022. Participants were selected purposively, and the findings were analyzed thematically. We applied data source and respondent triangulation to increase the findings' trustworthiness. RESULTS: Community perceptions of VAWG, specifically of intimate partner violence (IPV) and non-partner sexual violence (NPSV), varied depending on gender, age, and social position. IPV and NPSV were normalized through tolerance and denial by young and married men, while resistance to all forms of violence was common among women. Survivors of violence responded to the act of violence by leaving their homes, separating from their husbands, or taking harsh actions against their husbands, such as murder. Support for VAWG survivors was available through health care, free legal services, and a temporary shelter. Yet factors ranging from individual to societal levels, such as fear, lack of knowledge, lack of family and community support, and social and legal injustice, were barriers to accessing existing services. Nonetheless, violence survivors desired to speak about their experiences and seek psychosocial support. CONCLUSIONS: Our qualitative evidence gathered here can inform tailored VAWG prevention and response services such as interventions to shift social norms and the perception towards VAWG among different population group through raising awareness in schools, health care settings, faith-based venues, and using social media.

Stress and safety of maternal and newborn healthcare workers early in the COVID-19 pandemic: a repeat cross-sectional analysis from a global online survey from March 2020 to March 2021.

OBJECTIVES: This study aims to characterise the physical and psychological well-being of maternal and newborn healthcare workers (MNHCWs) during the COVID-19 pandemic. DESIGN: Observational repeated cross-sectional study. SETTING: An online questionnaire was distributed to MNHCWs around the globe in three separate rounds from March 2020 to March 2021. PARTICIPANTS: Total samples of N=1357 (round 1) and N=420 (round 3) primarily consisted of doctors, midwives and nurses in maternal and newborn specialties. Samples represented all WHO regions, with 33% (round 1) and 42% (round 3) from low- or middle-income countries (LMICs). PRIMARY AND SECONDARY OUTCOME MEASURES: Responses from rounds 1 (March-June 2020) and 3 (December 2020-March 2021) were analysed to measure self-reported levels of relative stress and workplace protection from COVID-19, while associated factors were determined through multivariable ordinal logistic regression. RESULTS: In round 1, 90% of MNHCWs reported increased stress levels and 45% reported insufficient personal protective equipment (PPE) access. Nurses and physicians were less likely to report increased stress than midwives at the pandemic onset. Factors associated with increased stress included being female, being from an LMIC and insufficient PPE. In round 3, 75% reported similar or increased stress while 10% reported insufficient PPE. In both rounds, over 50% of MNHCWs felt relatively or completely unprotected from COVID-19 in the workplace. Those from LMICs were more likely to report feeling unprotected, while receiving organisational information that valued safety was associated with better feelings of protection in the workplace. CONCLUSIONS: Among our international sample of MNHCWs, we observed high rates of self-reported stress increase at the start of the pandemic with persistence or increase up to a year later. High rates of feeling unprotected persisted even as PPE became more available. These results may inform interventions needed to support and protect MNHCWs during this and future pandemics.

Neonatal and perinatal mortality in the urban continuum: a geospatial analysis of the household survey, satellite imagery and travel time data in Tanzania.

INTRODUCTION: Recent studies suggest that the urban advantage of lower neonatal mortality in urban compared with rural areas may be reversing, but methodological challenges include misclassification of neonatal deaths and stillbirths, and oversimplification of the variation in urban environments. We address these challenges and assess the association between urban residence and neonatal/perinatal mortality in Tanzania. METHODS: The Tanzania Demographic and Health Survey (DHS) 2015-2016 was used to assess birth outcomes for 8915 pregnancies among 6156 women of reproductive age, by urban or rural categorisation in the DHS and based on satellite imagery. The coordinates of 527 DHS clusters were spatially overlaid with the 2015 Global Human Settlement Layer, showing the degree of urbanisation based on built environment and population density. A three-category urbanicity measure (core urban, semi-urban and rural) was defined and compared with the binary DHS measure. Travel time to the nearest hospital was modelled using least-cost path algorithm for each cluster. Bivariate and multilevel multivariable logistic regression models were constructed to explore associations between urbanicity and neonatal/perinatal deaths. RESULTS: Both neonatal and perinatal mortality rates were highest in core urban and lowest in rural clusters. Bivariate models showed higher odds of neonatal death (OR=1.85; 95% CI 1.12 to 3.08) and perinatal death (OR=1.60; 95% CI 1.12 to 2.30) in core urban compared with rural clusters. In multivariable models, these associations had the same direction and size, but were no longer statistically significant. Travel time to the nearest hospital was not associated with neonatal or perinatal mortality. CONCLUSION: Addressing high rates of neonatal and perinatal mortality in densely populated urban areas is critical for Tanzania to meet national and global reduction targets. Urban populations are diverse, and certain neighbourhoods or subgroups may be disproportionately affected by poor birth outcomes. Research must capture, understand and minimise risks specific to urban settings.

Utilisation of services along the continuum of maternal healthcare during the COVID-19 pandemic in Lubumbashi, DRC: findings from a cross-sectional household survey of women.

OBJECTIVES: The continuum of maternal care along antenatal (ANC), intrapartum and postnatal care (PNC) is fundamental for protecting women's and newborns' health. The COVID-19 pandemic interrupted the provision and use of these essential services globally. This study examines maternal healthcare utilisation along the continuum during the COVID-19 pandemic in the Democratic Republic of the Congo (DRC). DESIGN: This is a cross-sectional study using data collected on a survey of 599 households in Lubumbashi, DRC, using stratified random sampling. PARTICIPANTS: We included 604 women (15-49 years) who were pregnant between March 2020 and May 2021. OUTCOME MEASURES: A structured interview involved questions on sociodemographic characteristics, attitudes regarding COVID-19 and maternal service use and cost. Complete continuum of care was defined as receiving ANC 4+ consultations, skilled birth attendance and at least one PNC check for both mother and newborn. Data were analysed in SPSS using descriptive statistics and multivariable logistic regression. RESULTS: One-third (36%) of women who gave birth during the COVID-19 pandemic completed the continuum of maternal healthcare. Factors significantly associated with completing the continuum included higher education (aOR=2.6; p<0.001) and positive attitude towards the COVID-19 vaccination (aOR=1.9; p=0.04). Reasons for not seeking maternal care included lack of money and avoiding COVID-19 vaccination. CONCLUSION: During the COVID-19 pandemic, maternal healthcare seeking behaviours were shaped by vaccine hesitancy and care unaffordability in Lubumbashi. Addressing the high cost of maternal healthcare and vaccine hesitancy appear essential to improve access to maternal healthcare.

Capturing children food exposure using wearable cameras and deep learning.

Children's dietary habits are influenced by complex factors within their home, school and neighborhood environments. Identifying such influencers and assessing their effects is traditionally based on self-reported data which can be prone to recall bias. We developed a culturally acceptable machine-learning-based data-collection system to objectively capture school-children's exposure to food (including food items, food advertisements, and food outlets) in two urban Arab centers: Greater Beirut, in Lebanon, and Greater Tunis, in Tunisia. Our machine-learning-based system consists of 1) a wearable camera that captures continuous footage of children's environment during a typical school day, 2) a machine learning model that automatically identifies images related to food from the collected data and discards any other footage, 3) a second machine learning model that classifies food-related images into images that contain actual food items, images that contain food advertisements, and images that contain food outlets, and 4) a third machine learning model that classifies images that contain food items into two classes, corresponding to whether the food items are being consumed by the child wearing the camera or whether they are consumed by others. This manuscript reports on a user-centered design study to assess the acceptability of using wearable cameras to capture food exposure among school children in Greater Beirut and Greater Tunis. We then describe how we trained our first machine learning model to detect food exposure images using data collected from the Web and utilizing the latest trends in deep learning for computer vision. Next, we describe how we trained our other machine learning models to classify food-related images into their respective categories using a combination of public data and data acquired via crowdsourcing. Finally, we describe how the different components of our system were packed together and deployed in a real-world case study and we report on its performance.

Facilitators and barriers of implementation of routine postnatal care guidelines for women: A systematic scoping review using critical interpretive synthesis.

Background: Postnatal care (PNC) has the potential to prevent a substantial burden of maternal and newborn morbidity and mortality. This scoping review aimed to identify and synthesise themes related to facilitators and barriers of implementation of guidelines on routine PNC for women (postpartum care) in all settings. Methods: This is a scoping review guided by the standard principles of Arksey & O'Malley's framework. We used the critical interpretive synthesis method to synthesise the whole body of evidence. We searched four databases (Medline, Embase, Global Health, CINAHL Plus) using a combination of search terms comprising four key concepts: postnatal care, routine care, guidelines and implementation. No restrictions on country or language of publication were applied. We excluded studies not presenting findings about PNC for women. We thematically charted the themes of studies included based on title and abstract screening. All studies included after full text screening were described and their results synthesised using the socio-ecological model framework. We did not conduct a risk of bias analysis or quality assessment of included studies. Results: We identified a total of 8692 unique records and included 43 studies which identified facilitators and barriers to implementing routine guidelines in provision of PNC to women. Three quarters of studies pertained to PNC provision in high-income countries. Specific facilitators and barriers were identified and thematically presented based on whether they affect the provision of PNC or the intersection between provision of PNC and its use by women and families. We applied a critical global health lens to synthesise three constructs in the literature: finding a balance between standardisation and individualisation of PNC, the fragmented PNC provision landscape complicating the experiences of women with intersecting vulnerabilities, and the heavy reliance on the short postpartum period as an opportunity to educate and retain women and newborns in the health system. Conclusions: This interpretive synthesis of evidence shows that the fragmented and narrow nature of PNC provision presents specific challenges to developing, adapting and implementing routine PNC guidelines. This results in a lack of linkages to social support and services, fails to address intersecting vulnerabilities and inequities among women, and negatively influences care seeking. There is a lack of evidence on how processes of individualising PNC provision can be applied in practice to support health workers in providing woman-centered PNC in various global settings. Registration: https://www.protocols.io/private/C99DA688881F11EBB4690A58A9FEAC02.

Provision and utilization of maternal health services during the COVID-19 pandemic in 16 hospitals in sub-Saharan Africa.

Objective: Maintaining provision and utilization of maternal healthcare services is susceptible to external influences. This study describes how maternity care was provided during the COVID-19 pandemic and assesses patterns of service utilization and perinatal health outcomes in 16 referral hospitals (four each) in Benin, Malawi, Tanzania and Uganda. Methods: We used an embedded case-study design and two data sources. Responses to open-ended questions in a health-facility assessment survey were analyzed with content analysis. We described categories of adaptations and care provision modalities during the pandemic at the hospital and maternity ward levels. Aggregate monthly service statistics on antenatal care, delivery, caesarean section, maternal deaths, and stillbirths covering 24 months (2019 and 2020; pre-COVID-19 and COVID-19) were examined. Results: Declines in the number of antenatal care consultations were documented in Tanzania, Malawi, and Uganda in 2020 compared to 2019. Deliveries declined in 2020 compared to 2019 in Tanzania and Uganda. Caesarean section rates decreased in Benin and increased in Tanzania in 2020 compared to 2019. Increases in maternal mortality ratio and stillbirth rate were noted in some months of 2020 in Benin and Uganda, with variability noted between hospitals. At the hospital level, teams were assigned to respond to the COVID-19 pandemic, routine meetings were cancelled, and maternal death reviews and quality improvement initiatives were interrupted. In maternity wards, staff shortages were reported during lockdowns in Uganda. Clinical guidelines and protocols were not updated formally; the number of allowed companions and visitors was reduced. Conclusion: Varying approaches within and between countries demonstrate the importance of a contextualized response to the COVID-19 pandemic. Maternal care utilization and the ability to provide quality care fluctuated with lockdowns and travel bans. Women's and maternal health workers' needs should be prioritized to avoid interruptions in the continuum of care and prevent the deterioration of perinatal health outcomes.

Obstetric referrals, complications and health outcomes in maternity wards of large hospitals during the COVID-19 pandemic: a mixed methods study of six hospitals in Guinea, Nigeria, Uganda and Tanzania.

OBJECTIVES: The COVID-19 pandemic affected provision and use of maternal health services. This study describes changes in obstetric complications, referrals, stillbirths and maternal deaths during the first year of the pandemic and elucidates pathways to these changes. DESIGN: Prospective observational mixed-methods study, combining monthly routine data (March 2019-February 2021) and qualitative data from prospective semi-structured interviews. Data were analysed separately, triangulated during synthesis and presented along three country-specific pandemic periods: first wave, slow period and second wave. SETTING: Six referral maternities in four sub-Saharan African countries: Guinea, Nigeria, Tanzania and Uganda. PARTICIPANTS: 22 skilled health personnel (SHP) working in the maternity wards of various cadres and seniority levels. RESULTS: Percentages of obstetric complications were constant in four of the six hospitals. The percentage of obstetric referrals received was stable in Guinea and increased at various times in other hospitals. SHP reported unpredictability in the number of referrals due to changing referral networks. All six hospitals registered a slight increase in stillbirths during the study period, the highest increase (by 30%-40%) was observed in Uganda. Four hospitals registered increases in facility maternal mortality ratio; the highest increase was in Guinea (by 158%), which had a relatively mild COVID-19 epidemic. These increases were not due to mortality among women with COVID-19. The main pathways leading to these trends were delayed care utilisation and disruptions in accessing care, including sub-optimal referral linkages and health service closures. CONCLUSIONS: Maternal and perinatal survival was negatively affected in referral hospitals in sub-Saharan Africa during COVID-19. Routine data systems in referral hospitals must be fully used as they hold potential in informing adaptations of maternal care services. If combined with information on women's and care providers' needs, this can contribute to ensuring continuation of essential care provision during emergency.

Maternal and newborn healthcare providers' work-related experiences during the COVID-19 pandemic, and their physical, psychological, and economic impacts: Findings from a global online survey.

The COVID-19 pandemic continues to have substantial impacts on health systems globally. This study describes experiences during the COVID-19 pandemic, and physical, psychological and economic impacts among maternal and newborn healthcare providers. We conducted a global online cross-sectional survey of maternal and newborn healthcare providers. Data collected between July and December 2020 included demographic characteristics, work-related experiences, and physical, psychological, and economic impacts of COVID-19. Descriptive statistics of quantitative data and content analysis of qualitative data were conducted. Findings were disaggregated by country income-level. We analysed responses from 1,191 maternal and newborn healthcare providers from 77 countries: middle-income 66%, high-income 27%, and low-income 7%. Most common cadres were nurses (31%), midwives/nurse-midwives (25%), and obstetricians/gynaecologists (21%). Quantitative and qualitative findings showed that 28% of respondents reported decreased workplace staffing levels following changes in staff-rotation (53%) and staff self-isolating after exposure to SARS-CoV-2 (35%); this led to spending less time with patients, possibly compromising care quality. Reported insufficient access to personal protective equipment (PPE) ranged from 12% for gloves to 32% for N-95 masks. Nonetheless, wearing PPE was tiresome, time-consuming, and presented potential communication barriers with patients. 58% of respondents reported higher stress levels, mainly related to lack of access to information or to rapidly changing guidelines. Respondents noted a negative financial impact-a decrease in income (70% among respondents from low-income countries) concurrently with increased personal expenditures (medical supplies, transportation, and PPE). Negative physical, psychological and economic impacts of COVID-19 on maternal and newborn healthcare providers were ongoing throughout 2020, especially in low-income countries. This can have severe consequences for provision and quality of essential care. There is need to increase focus on the implementation of interventions aiming to support healthcare providers, particularly those in low- and middle-income countries to protect essential health services from disruption.