Global estimates on the number of people blind or visually impaired by cataract: a meta-analysis from 2000 to 2020.

BACKGROUND: To estimate global and regional trends from 2000 to 2020 of the number of persons visually impaired by cataract and their proportion of the total number of vision-impaired individuals. METHODS: A systematic review and meta-analysis of published population studies and gray literature from 2000 to 2020 was carried out to estimate global and regional trends. We developed prevalence estimates based on modeled distance visual impairment and blindness due to cataract, producing location-, year-, age-, and sex-specific estimates of moderate to severe vision impairment (MSVI presenting visual acuity <6/18, ≥3/60) and blindness (presenting visual acuity <3/60). Estimates are age-standardized using the GBD standard population. RESULTS: In 2020, among overall (all ages) 43.3 million blind and 295 million with MSVI, 17.0 million (39.6%) people were blind and 83.5 million (28.3%) had MSVI due to cataract blind 60% female, MSVI 59% female. From 1990 to 2020, the count of persons blind (MSVI) due to cataract increased by 29.7%(93.1%) whereas the age-standardized global prevalence of cataract-related blindness improved by -27.5% and MSVI increased by 7.2%. The contribution of cataract to the age-standardized prevalence of blindness exceeded the global figure only in South Asia (62.9%) and Southeast Asia and Oceania (47.9%). CONCLUSIONS: The number of people blind and with MSVI due to cataract has risen over the past 30 years, despite a decrease in the age-standardized prevalence of cataract. This indicates that cataract treatment programs have been beneficial, but population growth and aging have outpaced their impact. Growing numbers of cataract blind indicate that more, better-directed, resources are needed to increase global capacity for cataract surgery.

Application of referenced thermodynamic integration to Bayesian model selection.

Evaluating normalising constants is important across a range of topics in statistical learning, notably Bayesian model selection. However, in many realistic problems this involves the integration of analytically intractable, high-dimensional distributions, and therefore requires the use of stochastic methods such as thermodynamic integration (TI). In this paper we apply a simple but under-appreciated variation of the TI method, here referred to as referenced TI, which computes a single model's normalising constant in an efficient way by using a judiciously chosen reference density. The advantages of the approach and theoretical considerations are set out, along with pedagogical 1 and 2D examples. The approach is shown to be useful in practice when applied to a real problem -to perform model selection for a semi-mechanistic hierarchical Bayesian model of COVID-19 transmission in South Korea involving the integration of a 200D density.

Changes in life expectancy and house prices in London from 2002 to 2019: hyper-resolution spatiotemporal analysis of death registration and real estate data.

Background: London has outperformed smaller towns and rural areas in terms of life expectancy increase. Our aim was to investigate life expectancy change at very-small-area level, and its relationship with house prices and their change. Methods: We performed a hyper-resolution spatiotemporal analysis from 2002 to 2019 for 4835 London Lower-layer Super Output Areas (LSOAs). We used population and death counts in a Bayesian hierarchical model to estimate age- and sex-specific death rates for each LSOA, converted to life expectancy at birth using life table methods. We used data from the Land Registry via the real estate website Rightmove (www.rightmove.co.uk), with information on property size, type and land tenure in a hierarchical model to estimate house prices at LSOA level. We used linear regressions to summarise how much life expectancy changed in relation to the combination of house prices in 2002 and their change from 2002 to 2019. We calculated the correlation between change in price and change in sociodemographic characteristics of the resident population of LSOAs and population turnover. Findings: In 134 (2.8%) of London's LSOAs for women and 32 (0.7%) for men, life expectancy may have declined from 2002 to 2019, with a posterior probability of a decline >80% in 41 (0.8%, women) and 14 (0.3%, men) LSOAs. The life expectancy increase in other LSOAs ranged from <2 years in 537 (11.1%) LSOAs for women and 214 (4.4%) for men to >10 years in 220 (4.6%) for women and 211 (4.4%) for men. The 2.5th-97.5th-percentile life expectancy difference across LSOAs increased from 11.1 (10.7-11.5) years in 2002 to 19.1 (18.4-19.7) years for women in 2019, and from 11.6 (11.3-12.0) years to 17.2 (16.7-17.8) years for men. In the 20% (men) and 30% (women) of LSOAs where house prices had been lowest in 2002, mainly in east and outer west London, life expectancy increased only in proportion to the rise in house prices. In contrast, in the 30% (men) and 60% (women) most expensive LSOAs in 2002, life expectancy increased solely independently of price change. Except for the 20% of LSOAs that had been most expensive in 2002, LSOAs with larger house price increases experienced larger growth in their population, especially among people of working ages (30-69 years), had a larger share of households who had not lived there in 2002, and improved their rankings in education, poverty and employment. Interpretation: Large gains in area life expectancy in London occurred either where house prices were already high, or in areas where house prices grew the most. In the latter group, the increases in life expectancy may be driven, in part, by changing population demographics. Funding: Wellcome Trust; UKRI (MRC); Imperial College London; National Institutes of Health Research.

Spatio-temporal estimates of HIV risk group proportions for adolescent girls and young women across 13 priority countries in sub-Saharan Africa.

The Global AIDS Strategy 2021-2026 identifies adolescent girls and young women (AGYW) as a priority population for HIV prevention, and recommends differentiating intervention portfolios geographically based on local HIV incidence and individual risk behaviours. We estimated prevalence of HIV risk behaviours and associated HIV incidence at health district level among AGYW living in 13 countries in sub-Saharan Africa. We analysed 46 geospatially-referenced national household surveys conducted between 1999-2018 across 13 high HIV burden countries in sub-Saharan Africa. Female survey respondents aged 15-29 years were classified into four risk groups (not sexually active, cohabiting, non-regular or multiple partner[s] and female sex workers [FSW]) based on reported sexual behaviour. We used a Bayesian spatio-temporal multinomial regression model to estimate the proportion of AGYW in each risk group stratified by district, year, and five-year age group. Using subnational estimates of HIV prevalence and incidence produced by countries with support from UNAIDS, we estimated new HIV infections in each risk group by district and age group. We then assessed the efficiency of prioritising interventions according to risk group. Data consisted of 274,970 female survey respondents aged 15-29. Among women aged 20-29, cohabiting (63.1%) was more common in eastern Africa than non-regular or multiple partner(s) (21.3%), while in southern countries non-regular or multiple partner(s) (58.9%) were more common than cohabiting (23.4%). Risk group proportions varied substantially across age groups (65.9% of total variation explained), countries (20.9%), and between districts within each country (11.3%), but changed little over time (0.9%). Prioritisation based on behavioural risk, in combination with location- and age-based prioritisation, reduced the proportion of population required to be reached in order to find half of all expected new infections from 19.4% to 10.6%. FSW were 1.3% of the population but 10.6% of all expected new infections. Our risk group estimates provide data for HIV programmes to set targets and implement differentiated prevention strategies outlined in the Global AIDS Strategy. Successfully implementing this approach would result in more efficiently reaching substantially more of those at risk for infections.

Subnational HIV incidence trends in Malawi: large, heterogeneous declines across space.

The rate of new HIV infections globally has decreased substantially from its peak in the late 1990s, but the epidemic persists and remains highest in many countries in eastern and southern Africa. Previous research hypothesised that, as the epidemic recedes, it will become increasingly concentrated among sub-populations and geographic areas where transmission is the highest and that are least effectively reached by treatment and prevention services. However, empirical data on subnational HIV incidence trends is sparse, and the local transmission rates in the context of effective treatment scale-up are unknown. In this work, we developed a novel Bayesian spatio-temporal epidemic model to estimate adult HIV prevalence, incidence and treatment coverage at the district level in Malawi from 2010 through the end of 2021. We found that HIV incidence decreased in every district of Malawi between 2010 and 2021 but the rate of decline varied by area. National-level treatment coverage more than tripled between 2010 and 2021 and more than doubled in every district. Large increases in treatment coverage were associated with declines in HIV transmission, with 12 districts having incidence-prevalence ratios of 0.03 or less (a previously suggested threshold for epidemic control). Across districts, incidence varied more than HIV prevalence and ART coverage, suggesting that the epidemic is becoming increasingly spatially concentrated. Our results highlight the success of the Malawi HIV treatment programme over the past decade, with large improvements in treatment coverage leading to commensurate declines in incidence. More broadly, we demonstrate the utility of spatially resolved HIV modelling in generalized epidemic settings. By estimating temporal changes in key epidemic indicators at a relatively fine spatial resolution, we were able to directly assess, for the first time, whether the ART scaleup in Malawi resulted in spatial gaps or hotspots. Regular use of this type of analysis will allow HIV program managers to monitor the equity of their treatment and prevention programmes and their subnational progress towards epidemic control.

Assessment of COVID-19 as the Underlying Cause of Death Among Children and Young People Aged 0 to 19 Years in the US.

Importance: COVID-19 was the underlying cause of death for more than 940 000 individuals in the US, including at least 1289 children and young people (CYP) aged 0 to 19 years, with at least 821 CYP deaths occurring in the 1-year period from August 1, 2021, to July 31, 2022. Because deaths among US CYP are rare, the mortality burden of COVID-19 in CYP is best understood in the context of all other causes of CYP death. Objective: To determine whether COVID-19 is a leading (top 10) cause of death in CYP in the US. Design, Setting, and Participants: This national population-level cross-sectional epidemiological analysis for the years 2019 to 2022 used data from the US Centers for Disease Control and Prevention Wide-Ranging Online Data for Epidemiologic Research (WONDER) database on underlying cause of death in the US to identify the ranking of COVID-19 relative to other causes of death among individuals aged 0 to 19 years. COVID-19 deaths were considered in 12-month periods between April 1, 2020, and August 31, 2022, compared with deaths from leading non-COVID-19 causes in 2019, 2020, and 2021. Main Outcomes and Measures: Cause of death rankings by total number of deaths, crude rates per 100 000 population, and percentage of all causes of death, using the National Center for Health Statistics 113 Selected Causes of Death, for ages 0 to 19 and by age groupings (<1 year, 1-4 years, 5-9 years, 10-14 years, 15-19 years). Results: There were 821 COVID-19 deaths among individuals aged 0 to 19 years during the study period, resulting in a crude death rate of 1.0 per 100 000 population overall; 4.3 per 100 000 for those younger than 1 year; 0.6 per 100 000 for those aged 1 to 4 years; 0.4 per 100 000 for those aged 5 to 9 years; 0.5 per 100 000 for those aged 10 to 14 years; and 1.8 per 100 000 for those aged 15 to 19 years. COVID-19 mortality in the time period of August 1, 2021, to July 31, 2022, was among the 10 leading causes of death in CYP aged 0 to 19 years in the US, ranking eighth among all causes of deaths, fifth in disease-related causes of deaths (excluding unintentional injuries, assault, and suicide), and first in deaths caused by infectious or respiratory diseases when compared with 2019. COVID-19 deaths constituted 2% of all causes of death in this age group. Conclusions and Relevance: The findings of this study suggest that COVID-19 was a leading cause of death in CYP. It caused substantially more deaths in CYP annually than any vaccine-preventable disease historically in the recent period before vaccines became available. Various factors, including underreporting and not accounting for COVID-19's role as a contributing cause of death from other diseases, mean that these estimates may understate the true mortality burden of COVID-19. The findings of this study underscore the public health relevance of COVID-19 to CYP. In the likely future context of sustained SARS-CoV-2 circulation, appropriate pharmaceutical and nonpharmaceutical interventions (eg, vaccines, ventilation, air cleaning) will continue to play an important role in limiting transmission of the virus and mitigating severe disease in CYP.

Global disparities in SARS-CoV-2 genomic surveillance.

Genomic sequencing is essential to track the evolution and spread of SARS-CoV-2, optimize molecular tests, treatments, vaccines, and guide public health responses. To investigate the global SARS-CoV-2 genomic surveillance, we used sequences shared via GISAID to estimate the impact of sequencing intensity and turnaround times on variant detection in 189 countries. In the first two years of the pandemic, 78% of high-income countries sequenced >0.5% of their COVID-19 cases, while 42% of low- and middle-income countries reached that mark. Around 25% of the genomes from high income countries were submitted within 21 days, a pattern observed in 5% of the genomes from low- and middle-income countries. We found that sequencing around 0.5% of the cases, with a turnaround time <21 days, could provide a benchmark for SARS-CoV-2 genomic surveillance. Socioeconomic inequalities undermine the global pandemic preparedness, and efforts must be made to support low- and middle-income countries improve their local sequencing capacity.

Effective refractive error coverage in adults aged 50 years and older: estimates from population-based surveys in 61 countries.

BACKGROUND: In 2021, WHO Member States endorsed a global target of a 40-percentage-point increase in effective refractive error coverage (eREC; with a 6/12 visual acuity threshold) by 2030. This study models global and regional estimates of eREC as a baseline for the WHO initiative. METHODS: The Vision Loss Expert Group analysed data from 565 448 participants of 169 population-based eye surveys conducted since 2000 to calculate eREC (met need/[met need + undermet need + unmet need]). A binary logistic regression model was used to estimate eREC by Global Burden of Disease (GBD) Study super region among adults aged 50 years and older. FINDINGS: In 2021, distance eREC was 79·1% (95% CI 72·4-85·0) in the high-income super region; 62·1% (54·7-68·8) in north Africa and Middle East; 49·5% (45·0-54·0) in central Europe, eastern Europe, and central Asia; 40·0% (31·7-48·2) in southeast Asia, east Asia, and Oceania; 34·5% (29·4-40·0) in Latin America and the Caribbean; 9·0% (6·5-12·0) in south Asia; and 5·7% (3·1-9·0) in sub-Saharan Africa. eREC was higher in men and reduced with increasing age. Global distance eREC increased from 2000 to 2021 by 19·0%. Global near vision eREC for 2021 was 20·5% (95% CI 17·8-24·4). INTERPRETATION: Over the past 20 years, distance eREC has increased in each super region yet the WHO target will require substantial improvements in quantity and quality of refractive services in particular for near vision impairment. FUNDING: WHO, Sightsavers, The Fred Hollows Foundation, Fondation Thea, Brien Holden Vision Institute, Lions Clubs International Foundation.

π VAE: a stochastic process prior for Bayesian deep learning with MCMC.

Stochastic processes provide a mathematically elegant way to model complex data. In theory, they provide flexible priors over function classes that can encode a wide range of interesting assumptions. However, in practice efficient inference by optimisation or marginalisation is difficult, a problem further exacerbated with big data and high dimensional input spaces. We propose a novel variational autoencoder (VAE) called the prior encoding variational autoencoder ( π VAE). π VAE is a new continuous stochastic process. We use π VAE to learn low dimensional embeddings of function classes by combining a trainable feature mapping with generative model using a VAE. We show that our framework can accurately learn expressive function classes such as Gaussian processes, but also properties of functions such as their integrals. For popular tasks, such as spatial interpolation, π VAE achieves state-of-the-art performance both in terms of accuracy and computational efficiency. Perhaps most usefully, we demonstrate an elegant and scalable means of performing fully Bayesian inference for stochastic processes within probabilistic programming languages such as Stan.

Orphanhood and Caregiver Loss Among Children Based on New Global Excess COVID-19 Death Estimates.

This study assesses estimates of new orphanhood based on excess deaths to provide a comprehensive measure of the COVID-19 pandemic's long-term impact on orphanhood and caregiver loss.

PriorVAE: encoding spatial priors with variational autoencoders for small-area estimation.

Gaussian processes (GPs), implemented through multivariate Gaussian distributions for a finite collection of data, are the most popular approach in small-area spatial statistical modelling. In this context, they are used to encode correlation structures over space and can generalize well in interpolation tasks. Despite their flexibility, off-the-shelf GPs present serious computational challenges which limit their scalability and practical usefulness in applied settings. Here, we propose a novel, deep generative modelling approach to tackle this challenge, termed PriorVAE: for a particular spatial setting, we approximate a class of GP priors through prior sampling and subsequent fitting of a variational autoencoder (VAE). Given a trained VAE, the resultant decoder allows spatial inference to become incredibly efficient due to the low dimensional, independently distributed latent Gaussian space representation of the VAE. Once trained, inference using the VAE decoder replaces the GP within a Bayesian sampling framework. This approach provides tractable and easy-to-implement means of approximately encoding spatial priors and facilitates efficient statistical inference. We demonstrate the utility of our VAE two-stage approach on Bayesian, small-area estimation tasks.

Spatial and temporal fluctuations in COVID-19 fatality rates in Brazilian hospitals.

The severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) Gamma variant of concern has spread rapidly across Brazil since late 2020, causing substantial infection and death waves. Here we used individual-level patient records after hospitalization with suspected or confirmed coronavirus disease 2019 (COVID-19) between 20 January 2020 and 26 July 2021 to document temporary, sweeping shocks in hospital fatality rates that followed the spread of Gamma across 14 state capitals, during which typically more than half of hospitalized patients aged 70 years and older died. We show that such extensive shocks in COVID-19 in-hospital fatality rates also existed before the detection of Gamma. Using a Bayesian fatality rate model, we found that the geographic and temporal fluctuations in Brazil's COVID-19 in-hospital fatality rates were primarily associated with geographic inequities and shortages in healthcare capacity. We estimate that approximately half of the COVID-19 deaths in hospitals in the 14 cities could have been avoided without pre-pandemic geographic inequities and without pandemic healthcare pressure. Our results suggest that investments in healthcare resources, healthcare optimization and pandemic preparedness are critical to minimize population-wide mortality and morbidity caused by highly transmissible and deadly pathogens such as SARS-CoV-2, especially in low- and middle-income countries.

A dataset of non-pharmaceutical interventions on SARS-CoV-2 in Europe.

During the second half of 2020, many European governments responded to the resurging transmission of SARS-CoV-2 with wide-ranging non-pharmaceutical interventions (NPIs). These efforts were often highly targeted at the regional level and included fine-grained NPIs. This paper describes a new dataset designed for the accurate recording of NPIs in Europe's second wave to allow precise modelling of NPI effectiveness. The dataset includes interventions from 114 regions in 7 European countries during the period from the 1st August 2020 to the 9th January 2021. The paper includes NPI definitions tailored to the second wave following an exploratory data collection. Each entry has been extensively validated by semi-independent double entry, comparison with existing datasets, and, when necessary, discussion with local epidemiologists. The dataset has considerable potential for use in disentangling the effectiveness of NPIs and comparing the impact of interventions across different phases of the pandemic.

National, regional, and global estimates of anaemia by severity in women and children for 2000-19: a pooled analysis of population-representative data.

BACKGROUND: Anaemia causes health and economic harms. The prevalence of anaemia in women aged 15-49 years, by pregnancy status, is indicator 2.2.3 of the UN Sustainable Development Goals, and the aim of halving the anaemia prevalence in women of reproductive age by 2030 is an extension of the 2025 global nutrition targets endorsed by the World Health Assembly (WHA). We aimed to estimate the prevalence of anaemia by severity for children aged 6-59 months, non-pregnant women aged 15-49 years, and pregnant women aged 15-49 years in 197 countries and territories and globally for the period 2000-19. METHODS: For this pooled analysis of population-representative data, we collated 489 data sources on haemoglobin distribution in children and women from 133 countries, including 4·5 million haemoglobin measurements. Our data sources comprised health examination, nutrition, and household surveys, accessed as anonymised individual records or as summary statistics such as mean haemoglobin and anaemia prevalence. We used a Bayesian hierarchical mixture model to estimate haemoglobin distributions in each population and country-year. This model allowed for coherent estimation of mean haemoglobin and prevalence of anaemia by severity. FINDINGS: Globally, in 2019, 40% (95% uncertainty interval [UI] 36-44) of children aged 6-59 months were anaemic, compared to 48% (45-51) in 2000. Globally, the prevalence of anaemia in non-pregnant women aged 15-49 years changed little between 2000 and 2019, from 31% (95% UI 28-34) to 30% (27-33), while in pregnant women aged 15-49 years it decreased from 41% (39-43) to 36% (34-39). In 2019, the prevalence of anaemia in children aged 6-59 months exceeded 70% in 11 countries and exceeded 50% in all women aged 15-49 years in ten countries. Globally in all populations and in most countries and regions, the prevalence of mild anaemia changed little, while moderate and severe anaemia declined in most populations and geographical locations, indicating a shift towards mild anaemia. INTERPRETATION: Globally, regionally, and in nearly all countries, progress on anaemia in women aged 15-49 years is insufficient to meet the WHA global nutrition target to halve anaemia prevalence by 2030, and the prevalence of anaemia in children also remains high. A better understanding of the context-specific causes of anaemia and quality implementation of effective multisectoral actions to address these causes are needed. FUNDING: USAID, US Centers for Disease Control and Prevention, and Bill & Melinda Gates Foundation.

Comparative analysis of the risks of hospitalisation and death associated with SARS-CoV-2 omicron (B.1.1.529) and delta (B.1.617.2) variants in England: a cohort study.

BACKGROUND: The omicron variant (B.1.1.529) of SARS-CoV-2 has demonstrated partial vaccine escape and high transmissibility, with early studies indicating lower severity of infection than that of the delta variant (B.1.617.2). We aimed to better characterise omicron severity relative to delta by assessing the relative risk of hospital attendance, hospital admission, or death in a large national cohort. METHODS: Individual-level data on laboratory-confirmed COVID-19 cases resident in England between Nov 29, 2021, and Jan 9, 2022, were linked to routine datasets on vaccination status, hospital attendance and admission, and mortality. The relative risk of hospital attendance or admission within 14 days, or death within 28 days after confirmed infection, was estimated using proportional hazards regression. Analyses were stratified by test date, 10-year age band, ethnicity, residential region, and vaccination status, and were further adjusted for sex, index of multiple deprivation decile, evidence of a previous infection, and year of age within each age band. A secondary analysis estimated variant-specific and vaccine-specific vaccine effectiveness and the intrinsic relative severity of omicron infection compared with delta (ie, the relative risk in unvaccinated cases). FINDINGS: The adjusted hazard ratio (HR) of hospital attendance (not necessarily resulting in admission) with omicron compared with delta was 0·56 (95% CI 0·54-0·58); for hospital admission and death, HR estimates were 0·41 (0·39-0·43) and 0·31 (0·26-0·37), respectively. Omicron versus delta HR estimates varied with age for all endpoints examined. The adjusted HR for hospital admission was 1·10 (0·85-1·42) in those younger than 10 years, decreasing to 0·25 (0·21-0·30) in 60-69-year-olds, and then increasing to 0·47 (0·40-0·56) in those aged at least 80 years. For both variants, past infection gave some protection against death both in vaccinated (HR 0·47 [0·32-0·68]) and unvaccinated (0·18 [0·06-0·57]) cases. In vaccinated cases, past infection offered no additional protection against hospital admission beyond that provided by vaccination (HR 0·96 [0·88-1·04]); however, for unvaccinated cases, past infection gave moderate protection (HR 0·55 [0·48-0·63]). Omicron versus delta HR estimates were lower for hospital admission (0·30 [0·28-0·32]) in unvaccinated cases than the corresponding HR estimated for all cases in the primary analysis. Booster vaccination with an mRNA vaccine was highly protective against hospitalisation and death in omicron cases (HR for hospital admission 8-11 weeks post-booster vs unvaccinated: 0·22 [0·20-0·24]), with the protection afforded after a booster not being affected by the vaccine used for doses 1 and 2. INTERPRETATION: The risk of severe outcomes following SARS-CoV-2 infection is substantially lower for omicron than for delta, with higher reductions for more severe endpoints and significant variation with age. Underlying the observed risks is a larger reduction in intrinsic severity (in unvaccinated individuals) counterbalanced by a reduction in vaccine effectiveness. Documented previous SARS-CoV-2 infection offered some protection against hospitalisation and high protection against death in unvaccinated individuals, but only offered additional protection in vaccinated individuals for the death endpoint. Booster vaccination with mRNA vaccines maintains over 70% protection against hospitalisation and death in breakthrough confirmed omicron infections. FUNDING: Medical Research Council, UK Research and Innovation, Department of Health and Social Care, National Institute for Health Research, Community Jameel, and Engineering and Physical Sciences Research Council.

Global, regional, and national minimum estimates of children affected by COVID-19-associated orphanhood and caregiver death, by age and family circumstance up to Oct 31, 2021: an updated modelling study.

BACKGROUND: In the 6 months following our estimates from March 1, 2020, to April 30, 2021, the proliferation of new coronavirus variants, updated mortality data, and disparities in vaccine access increased the amount of children experiencing COVID-19-associated orphanhood. To inform responses, we aimed to model the increases in numbers of children affected by COVID-19-associated orphanhood and caregiver death, as well as the cumulative orphanhood age-group distribution and circumstance (maternal or paternal orphanhood). METHODS: We used updated excess mortality and fertility data to model increases in minimum estimates of COVID-19-associated orphanhood and caregiver deaths from our original study period of March 1, 2020-April 30, 2021, to include the new period of May 1-Oct 31, 2021, for 21 countries. Orphanhood was defined as the death of one or both parents; primary caregiver loss included parental death or the death of one or both custodial grandparents; and secondary caregiver loss included co-residing grandparents or kin. We used logistic regression and further incorporated a fixed effect for western European countries into our previous model to avoid over-predicting caregiver loss in that region. For the entire 20-month period, we grouped children by age (0-4 years, 5-9 years, and 10-17 years) and maternal or paternal orphanhood, using fertility contributions, and we modelled global and regional extrapolations of numbers of orphans. 95% credible intervals (CrIs) are given for all estimates. FINDINGS: The number of children affected by COVID-19-associated orphanhood and caregiver death is estimated to have increased by 90·0% (95% CrI 89·7-90·4) from April 30 to Oct 31, 2021, from 2 737 300 (95% CrI 1 976 100-2 987 000) to 5 200 300 (3 619 400-5 731 400). Between March 1, 2020, and Oct 31, 2021, 491 300 (95% CrI 485 100-497 900) children aged 0-4 years, 736 800 (726 900-746 500) children aged 5-9 years, and 2 146 700 (2 120 900-2 174 200) children aged 10-17 years are estimated to have experienced COVID-19-associated orphanhood. Globally, 76·5% (95% CrI 76·3-76·7) of children were paternal orphans, whereas 23·5% (23·3-23·7) were maternal orphans. In each age group and region, the prevalence of paternal orphanhood exceeded that of maternal orphanhood. INTERPRETATION: Our findings show that numbers of children affected by COVID-19-associated orphanhood and caregiver death almost doubled in 6 months compared with the amount after the first 14 months of the pandemic. Over the entire 20-month period, 5·0 million COVID-19 deaths meant that 5·2 million children lost a parent or caregiver. Our data on children's ages and circumstances should support pandemic response planning for children globally. FUNDING: UK Research and Innovation (Global Challenges Research Fund, Engineering and Physical Sciences Research Council, and Medical Research Council), Oak Foundation, UK National Institute for Health Research, US National Institutes of Health, and Imperial College London.

Robust deep learning optical autofocus system applied to automated multiwell plate single molecule localization microscopy.

We presenta robust, long-range optical autofocus system for microscopy utilizing machine learning. This can be useful for experiments with long image data acquisition times that may be impacted by defocusing resulting from drift of components, for example due to changes in temperature or mechanical drift. It is also useful for automated slide scanning or multiwell plate imaging where the sample(s) to be imaged may not be in the same horizontal plane throughout the image data acquisition. To address the impact of (thermal or mechanical) fluctuations over time in the optical autofocus system itself, we utilize a convolutional neural network (CNN) that is trained over multiple days to account for such fluctuations. To address the trade-off between axial precision and range of the autofocus, we implement orthogonal optical readouts with separate CNN training data, thereby achieving an accuracy well within the 600 nm depth of field of our 1.3 numerical aperture objective lens over a defocus range of up to approximately +/-100 µm. We characterize the performance of this autofocus system and demonstrate its application to automated multiwell plate single molecule localization microscopy.

Many microscopy experiments involve extended imaging of samples over timescales from minutes to days, during which the microscope can 'drift' out of focus. When imaging at high magnification, the depth of field is of the order of one micron and so the imaging system should keep the sample in the focal plane of the microscope objective lens to this precision. Unfortunately, temperature changes in the laboratory can cause thermal expansion of microscope components that can move the focal plane by more than a micron and such changes can occur on a timescale of minutes. This is a particular issue for super-resolved microscopy experiments using single molecule localization microscopy (SMLM) techniques, for which 1000s of images are acquired, and for automated imaging of multiple samples in multiwell plates. It is possible to maintain the sample in the focal plane focus position by either automatically moving the sample or adjusting the imaging system, for example by moving the objective lens. This is called 'autofocus' and is frequently achieved by reflecting a light beam from the microscope coverslip and measuring its position of beam profile as a function of defocus of the microscope. The correcting adjustment is then usually calculated analytically but there is recent interest in using machine learning techniques to determine the required focussing adjustment. Here, we present a system that uses a neural network to determine the required defocus correcting adjustment from camera images of a laser beam that is reflected from the coverslip. Unfortunately, this approach will only work when the microscope is in the same condition as it was when the neural network was trained - and this can be compromised by the same drift of the optical system that causes the defocus needing to be corrected. We show, however, that by training a neural network over an extended period, for example 10 days, this approach can 'learn' about the optical system drifts and provide the required autofocus function. We also show that an optical system utilizing a rectangular slit can make two measurements of the defocus simultaneously, with one measurement being optimized for high accuracy over a limited range (±10 µm) near focus and the other providing lower accuracy but over a much longer range (±100 µm). This robust autofocus system is suitable for automated super-resolved microscopy of arrays of samples in a multiwell plate using SMLM, for which an experiment routinely lasts more than 5 h.