Modeling the transmission dynamics of delayed pneumonia-like diseases with a sensitivity of parameters

There have been rapid and dramatic increases in the numbers of people on antiretroviral treatment in Africa [1]. These numbers should continue to grow substantially, if current treatment need is met and those progressing to later stages of infection are added into treatment programmes, especially if earlier initiation on treatment is adopted. Antiretroviral treatment itself may suppress HIV transmission rates by reducing viral load [2,3], but its preventive effect in populations will depend on its impact on sexual behaviour, with increases in risky behaviour potentially negating the benefits of reduced transmissibility and decreases in risky behaviour enhancing them. Indeed, a substantial part of the modelled impact of the ‘test and treat’ strategy promoted by the WHO was derived from an assumed change in risky behaviour [4]. The paper by Venkatesh et al.[5] published in this issue of AIDS suggests that initiation on treatment could lead to substantial reductions in risk behaviour in urban and rural populations in South Africa, which, if they can be replicated when treatment is initiated earlier and can be sustained for long periods after patients have regained their full health, would provide dramatic prevention benefits. These findings are very encouraging, especially because the comparison group comprised infected individuals who had previously received testing and counselling and could, therefore, already have adopted safer behaviour [6]. However, before we conclude that treatment will enhance HIV prevention efforts by reducing risk behaviour, we should consider, first, whether such a big effect in antiretroviral therapy (ART) patients is plausible – or is more likely to have resulted from methodological difficulties – and, second, what the influence of ART might be on the behaviour of susceptible and undiagnosed infected individuals. The study by Venkatesh et al.[5] has a prospective design (spanning periods up to and after ART initiation), a large representative sample (n = 6263 of whom 37% initiated treatment) with regular follow-ups, collected data on both sexes in both urban and rural areas, and applied rigorous statistical methods. As in some previous studies in sub-Saharan Africa [7,8], initiation on ART was associated with reductions in sexual activity, including reporting of unprotected sex and multiple sexual partnerships. However, the effect sizes in this study were particularly remarkable. The authors did not investigate the reasons for the reductions in risk behaviour in their setting and we wonder about the plausibility of such big effect sizes. Why would people who had been seriously ill (WHO stage 4 disease and/or CD4 cell count < 200 cells/μl.) and are now returned to good health become less sexually active and have fewer partners? Counselling and condoms were provided to newly diagnosed individuals as well as to those initiated on treatment; if disclosure to partners or fear of infecting others increased adoption of safer behaviours why would this be different for those who know they are infected, but are not yet treated? Methodological difficulties remain a concern: if interviewers were not blind to who was on treatment, interviewer bias could have affected the results. We are told that interviewers were trained to avoid social desirability bias, but this is very difficult to achieve especially when interviews are conducted in a clinical setting. It must be possible, therefore, that participants on ART felt a greater need to conceal risky behaviour. Concern over validity would be much assuaged by a good explanation for the behaviour change. In addition, trials showing the impact of expanded treatment on HIV incidence in populations would be welcome. In assessing the impact of ART on sexual transmission, it is important that we consider the right behaviours in the right people. The best indicator of risk is multiple partnerships, so it is notable that this was the indicator that changed the most amongst individuals initiated on ART in the South Africa study. Nevertheless, in assessing the prevention impact, it is important to consider the behaviour of those started on treatment in the context of the wider population. An increase in the fraction of the population that is HIV-positive owing to longer survival of those infected, other things being equal, might be expected to lead to an increase in exposure amongst susceptible individuals. More importantly, the effects of widespread availability of ART on behaviour will extend beyond those who are on treatment. In Western countries, studies have found evidence of behavioural disinhibition in susceptible and undiagnosed individuals in the presence of ART scale-up that resulted in the resurgence of epidemics [9]. The behaviour of these individuals will be as important, if not more so, than that of those on treatment in determining the future course of HIV epidemics within sub-Saharan Africa. The study by Venkatesh et al.[5] is a valuable contribution to our understanding of the impact of ART initiation on sexual behaviour of infected individuals and suggests that this impact could be large. However, more evidence through monitoring and evaluation is required before we can safely assume expanded treatment access is successfully reducing HIV incidence. Elucidating the reasons for changes in behaviour observed and establishing whether these changes extend to those with early treatment initiation and are sustained over time would be important contributions. Equally important, studies on the effects of increases in the availability of ART must be extended to encompass the effects on risk behaviour amongst undiagnosed and uninfected individuals. In the meantime, close integration of prevention activities within expanded care activities should continue [10].

Cancer and palliative care in Africa.

Is Africa prepared for tackling the COVID-19 (SARS-CoV-2) epidemic. Lessons from past outbreaks, ongoing pan-African public health efforts, and implications for the future

Cardiovascular disease (CVD) remains the leading cause of mortality and a major contributor to disability globally. High dietary sodium is one of the major risk factors for CVD. It is associated with hypertension and increased risk of stroke and heart disease. According to the Global Burden of Disease study, in 2019, high dietary sodium accounted for an estimated 1.7 million cardiovascular deaths worldwide, representing ~ 10% of the overall cardiovascular mortality. In view of the huge burden of high dietary sodium, salt reduction has been identified by the World Health Organization (WHO) as one the most efficient and cost-effective measures that can improve population health outcomes. The WHO recommends that adults consume less than 2 g of sodium per day which corresponds to less than 5 g (just under a teaspoon) of salt per day. The Global Burden of Diseases Nutrition and Chronic Diseases Expert Group (NutriCoDE) estimated the global mean sodium intake at 3.95 g per day in 2010, almost twice the recommended maximum level of intake. The WHO projected that nearly 2.5 million deaths could be prevented each year if global salt consumption were reduced to the recommended level.

Although 2015 is only three years away, many countries are not on track to reach the United Nations Millennium Development Goals (MDGs), particularly goals 4 and 5, which call for reductions in child and maternal mortality rates. Targeted interventions are needed for rapid progress to take place. However, reliable estimates of the numbers, causes and determinants of death are needed to design the interventions. Such estimates are the basis of functional health information systems, which are the sources of data for the development and monitoring of evidence-based health policies and programmes.1 Countries unable to record the number of people who die or why they die cannot realize the full potential of their health systems.2 Nonetheless, many countries lack data on the social, behavioural and health systems determinants of child and maternal deaths. In some high-mortality settings, civil registration systems are so deficient that most deaths go unrecorded.3 Globally, fewer than one third of the 350 000 maternal deaths and 7 600 000 child deaths4 occurring annually are medically certified.5 The data needed to estimate cause-specific mortality rates can be obtained through alternative methods such as verbal autopsy, which consists of the use of standardized interview tools to question the caretakers of recently deceased persons about the symptoms that preceded the death.2Area-specific disease profiles generated from these data can help subnational and national health planners improve health resource allocation, as illustrated by the United Republic of Tanzania’s Essential Health Interventions Project.6 Yet despite the resources invested in developing verbal autopsy methods and tools, the non-biological factors contributing to a death cannot be determined from verbal autopsy data. This has led to the development of a complementary interview method known as “social autopsy”, which explores the social, behavioural and health systems determinants of maternal and child deaths.7 During social autopsy, the caretaker of the deceased is interviewed in detail regarding any preventive care received by the deceased, the diagnostic procedures followed, the type and timing of any treatment provided inside or outside the home, and any barriers encountered during care seeking.8 While verbal autopsy data can be used to prioritize health problems and evaluate health programme impact, social autopsy data, which focus on modifiable factors present in the home, community and health system, can inform policies and practices for increasing access to and use of preventive and curative services. The data afforded by social autopsy can create awareness that maternal and child deaths are preventable, empower communities to actively participate in interventions for reducing child and maternal deaths, and increase health programme responsiveness and accountability.9 Despite its importance, social autopsy has not been widely practiced and still lacks standard methods for data collection and analysis. Two key models, known as Pathway to Survival10 and Three Delays,11 have been used to organize the care-seeking data generated by social autopsy.8,9,12 These models should guide future social autopsy studies, whose aim should be to maximize social autopsy’s awareness-raising potential through participatory data sharing and development of interventions.7 The World Health Organization’s Child Health Epidemiology Reference Group has been further developing and testing the Pathway to Survival and Three Delays analytical models in several low-income countries. The INDEPTH Network has also been developing its own social autopsy tools. Both groups have identified similar challenges, primarily the need to reduce the number of interview questions to make data collection faster, and to develop a suitable strategy for data analysis. The standard social autopsy tool intended for use in combination with verbal autopsy is already under review to try to reduce the number of questions, but more resources should be invested in trying to simplify and standardize social autopsy methods and in encouraging the use of social autopsy, which is a new and still unfamiliar method. Resources should also be invested in bringing together interested parties with the aim of reaching consensus on standardized formats and computer-aided data analysis plans. All of these measures will enhance the visibility of social autopsy and promote routine collection and use of quality social autopsy data. Social autopsy instruments should be widely disseminated, along with tools for country adaptation, fieldworker training and community data sharing, and for integration with population-wide surveys. Tools of this type are direly needed for low-income countries to achieve the MDGs established for the reduction of child and maternal mortality.

The rising burden of non-communicable diseases in sub-Saharan Africa

Importance of tuberculosis control to address child survival

Integrated approach to child health in developing countries

In low-income and middle-income countries, such as those in sub-Saharan Africa and Latin America, the COVID-19 pandemic has had substantial implications for women's wellbeing. Policy responses to the COVID-19 pandemic have highlighted the gendered aspect of pandemics; however, addressing the gendered implications of the COVID-19 pandemic comprehensively and effectively requires a planetary health perspective that embraces systems thinking to inequalities. This Viewpoint is based on collective reflections from research done by the authors on COVID-19 responses by international and regional organisations, and national governments, in Latin America and sub-Saharan Africa between June, 2020, and June, 2021. A range of international and regional actors have made important policy recommendations to address the gendered implications of the COVID-19 pandemic on women's health and wellbeing since the start of the pandemic. However, national-level policy responses to the COVID-19 pandemic have been partial and inconsistent with regards to gender in both sub-Saharan Africa and Latin America, largely failing to recognise the multiple drivers of gendered health inequalities. This Viewpoint proposes that addressing the effects of the COVID-19 pandemic on women in low-income and middle-income countries should adopt a systems thinking approach and be informed by the question of who is affected as opposed to who is infected. In adopting the systems thinking approach, responses will be more able to recognise and address the direct gendered effects of the pandemic and those that emerge indirectly through a combination of long-standing structural inequalities and gendered responses to the pandemic.

Tuberculosis (TB) kills ∼1.5 million people worldwide each year.1 Among children and adolescents infected with TB, those <5 years old and those with HIV coinfection have a higher risk of death.2–4 Not only have few researchers quantified the impact of antiretroviral therapy (ART) regarding those with coinfected with HIV,4–6 there are important gaps in our overall knowledge for adolescents treated for TB. Numerous observational studies conducted before the availability of anti-TB antibiotics revealed that TB disease and death peaked in the first 2 years of life and then decreased until adolescence, after which it rose steeply.7 Some reports revealed sex-based differences in incidence and mortality during adolescence.8–11 In recent decades, however, adolescents have received little attention in TB research, and their vulnerability to the disease has been largely forgotten, mostly because TB surveillance systems group individuals <15 years with children and those ≥15 years with adults.1,12 Disaggregated TB data have not been reported routinely for adolescents, although this practice will change in 2021 after the World Health Organization’s request for countries to disaggregate TB data by 5-year age bands for people <25 years old.In this issue of Pediatrics, Osman et al13 address these research gaps by analyzing the South African Electronic Treatment Register, a national surveillance database of individuals treated for drug-susceptible TB. The study includes ∼85 000 HIV coinfected children and adolescents, of whom nearly one-half were not receiving ART at TB diagnosis. The authors found that, compared with HIV-uninfected children and adolescents, those with HIV on ART were >5 times as likely to die within 6 months of TB diagnosis, whereas those with HIV not on ART were >8 times more likely to die. These findings highlight the importance of TB preventive therapy and timely ART initiation in HIV-infected children and adolescents.4,5Using finely age-disaggregated case fatality ratios (CFRs), Osman et al13 found that the risk of death was high in the first year of life for both sexes (4.1%), early adolescence for boys (4.3%), and late adolescence for girls (4.2%). Estimating standardized mortality ratios (SMRs), the authors found that, in 2016, observed TB deaths occurred at 4 to 5 times the rate of expected deaths in children <5 years of age. In contrast, in 2016, TB deaths occurred at 77 to 92 times the rate of expected deaths in adolescents 10 to 14 years old. Notably, SMRs were much higher for adolescent girls than for adolescent boys. As the authors hypothesize, lower SMRs in young children may reflect their susceptibility to other diseases, and the discrepancy between adolescent boys and girls may reflect the high risk of death from injury and violence among boys in South Africa. Nonetheless, SMRs reveal the impact of TB on specific groups and should be incorporated into future research to inform public health decisions.Another strength of this study is its examination of time trends between 2004 and 2016. The authors observed similar CFRs in all age groups in 2004. Thereafter, CFRs declined in children <10 years old (most steeply in those <5 years old) but remained stable in adolescents. This discrepancy is consistent with a global trend of sharp mortality declines in young children, compared with adolescents.14 Taken together, the study observations underscore the need to devote more resources to adolescents: identifying and addressing their risk factors for TB mortality, prioritizing them in TB-contact investigations and screening activities, and increasing the prescription of and adherence to TB preventive treatment in this age group.A key consideration is that the data available for this study do not include people who are diagnosed with TB but do not initiate treatment, are never diagnosed with TB at all, or have drug-resistant TB. Because an estimated 20% of children <15 years old with untreated TB disease die within 1 year and treatment outcomes for those with drug-resistant TB are likely worse,4 the findings represent the “best case scenario” for child and adolescent TB mortality in South Africa.In the study, Osman et al13 clarify the impact of HIV and ART on TB mortality in children and adolescents and expose adolescents’ overall high risk of poor TB outcomes. In future research efforts, researchers should incorporate the study’s granular age categories and innovative use of both CFRs and SMRs to examine TB deaths in children and adolescents in diverse settings. For example, in former Soviet republics, where >15% of adolescents have rifampin-resistant TB, characterizing the impact of drug-resistance on TB mortality is particularly important.15 The World Health Organization’s recent request for data disaggregation by 5-year age bands will facilitate more detailed evaluations of TB deaths among children and adolescents around the world. By analyzing these disaggregated data, researchers can help improve outcomes for the most vulnerable children and adolescents.

BackgroundSickle cell disease (SCD) is common in many parts of sub-Saharan Africa (SSA), where it is associated with high early mortality. In the absence of newborn screening, most deaths among children with SCD go unrecognized and unrecorded. As a result, SCD does not receive the attention it deserves as a leading cause of death among children in SSA. In the current study, we explored the potential utility of verbal autopsy (VA) as a tool for attributing underlying cause of death (COD) in children to SCD.MethodsWe used the 2007 WHO Sample Vital Registration with Verbal Autopsy (SAVVY) VA tool to determine COD among child residents of the Kilifi Health and Demographic Surveillance System (KHDSS), Kenya, who died between January 2008 and April 2011. VAs were coded both by physician review (physician coded verbal autopsy, PCVA) using COD categories based on the WHO International Classification of Diseases 10th Edition (ICD-10) and by using the InterVA-4 probabilistic model after extracting data according to the 2012 WHO VA standard. Both of these methods were validated against one of two gold standards: hospital ICD-10 physician-assigned COD for children who died in Kilifi District Hospital (KDH) and, where available, laboratory confirmed SCD status for those who died in the community.ResultsOverall, 6% and 5% of deaths were attributed to SCD on the basis of PCVA and the InterVA-4 model, respectively. Of the total deaths, 22% occurred in hospital, where the agreement coefficient (AC1) for SCD between PCVA and hospital physician diagnosis was 95.5%, and agreement between InterVA-4 and hospital physician diagnosis was 96.9%. Confirmatory laboratory evidence of SCD status was available for 15% of deaths, in which the AC1 against PCVA was 87.5%.ConclusionsOther recent studies and provisional data from this study, outlining the importance of SCD as a cause of death in children in many parts of the developing world, contributed to the inclusion of specific SCD questions in the 2012 version of the WHO VA instruments, and a specific code for SCD has now been included in the WHO and InterVA-4 COD listings. With these modifications, VA may provide a useful approach to quantifying the contribution of SCD to childhood mortality in rural African communities. Further studies will be needed to evaluate the generalizability of our findings beyond our local context.

Innovations in pneumonia diagnosis and treatment: a call to action on World Pneumonia Day, 2013.

Severe meningococcal disease in childhood

Fertility decline in most countries of sub-Saharan Africa has thus far started later and proceeded more slowly than in countries in Asia and Latin America and the Caribbean undergoing the transition in fertility since the 1950s from high levels to near-replacement or even below-replacement levels (Bongaarts and Casterline 2013). Yet there is considerable variation among countries in sub-Saharan Africa: in the duration and magnitude of fertility decline, whether stalls in fertility decline have occurred, shifts in the timing of births, and even the economic and population subgroups that have led declines in family size (Caldwell, Orubuloye, and Caldwell 1992; Bongaarts and Casterline 2013; Cleland, Onuoha, and Timæus 1994; Cohen 1998; Ezeh, Mberu, and Emina 2009; Garenne 2008; Kirk and Pillet 1998; Rossier, Corker, and Schoumaker 2015; Timæus and Moultrie 2008). With current fertility estimated at 5.1 births per woman in the region and 19 countries in sub-Saharan Africa at or above that level and another 21 countries with at least four births per woman on average (United Nations 2015a), the pathways that future fertility takes will significantly determine population growth and age structure shifts not only in the region, but increasingly for the world. Sub-Saharan Africa is projected to grow from 840 million people in 2010 to nearly 1.4 billion in 2030 (United Nations 2015a). Above-replacement fertility is projected to account for 61 percent of this population increase from 2010 to 2030 compared to 4 percent from mortality reduction, 37 percent from a young age structure in 2010 (population momentum), and a small negative contribution from migration (United Nations 2015b). These projections draw on the United Nations medium variant and do not take into account the uncertainty around current and future fertility levels, uncertainty that only increases the farther the projection period extends. For high-fertility countries in sub-Saharan Africa, the wide uncertainty around where fertility is headed can result in substantial differences in population projections (Ezeh, Mberu, and Emina 2009; Fuchs and Goujon 2014; Gerland et al. 2014). Beyond population numbers alone, the uncertainty about fertility decline also bears on policy-relevant questions such as the degree to which the region may realize a demographic dividend (e.g., how fast the shift will occur toward a higher ratio of working-age population to non-working-age population and the consequent effects on economic growth) (Bloom et al. 2013) or the extent to which greenhouse gas emissions might be reduced by slowing population growth (O'Neill et al. 2010). Earlier reviews of the fertility transitions in sub-Saharan Africa through the 1980s and 1990s showed fertility declines underway in most countries and particularly rapid declines in several countries in Eastern Africa (Kenya, Rwanda, and Zimbabwe) and Southern Africa (Botswana and South Africa) (Cleland, Onuoha, and Timæus 1994; Cohen 1998; Kirk and Pillet 1998). Subsequent survey data suggested an apparent slowdown in the pace of fertility decline in more than ten sub-Saharan African countries (Bongaarts 2008). Further analyses of the data indicated far fewer stalls in fertility decline had in fact occurred in the region, with evidence strongly supportive of stalls in Kenya and Rwanda, and the stalls that do occur have been of relatively short duration (Garenne 2011; Machiyama 2010; Schoumaker 2009, 2014). Most countries in sub-Saharan Africa still lack complete and accurate vital registration data on births, so these and other analyses of fertility trends will continue to rely heavily on survey data and require reconciling estimates from different sources (Alkema et al. 2012; United Nations 2015c). Our aim in this chapter is to provide an updated and concise description of the diversity of fertility decline patterns among countries1 in sub-Saharan Africa, drawing on the latest series of fertility estimates that take into account many different data sources and that are harmonized with other demographic components (United Nations 2015d). We focus on the level of fertility prior to the start of fertility decline, the time period of the fertility transition, and the estimated pace of decline. We also explore the implications of different fertility decline patterns for future fertility and population projections in the region. We draw on the distinct patterns of fertility decline among countries worldwide that are advanced in (or have completed) their first fertility transition to construct probabilistic fertility and population projections for sub-Saharan African countries. The illustrative comparisons of projections highlight the demographic impact if future fertility decline in sub-Saharan countries were to accelerate and follow the rapid pace of decline already experienced by a diverse group of countries. The UN Population Division publishes estimates and projections of period total fertility rates in World Population Prospects (WPP) every two years. The estimates of total fertility presented in this chapter are from the 2015 Revision and are for five-year time periods for countries or areas with 90,000 persons or more in 2015 (United Nations 2015a). The most recent data underlying the total fertility estimates from the 2015 Revision for 50 sub-Saharan African countries2 (United Nations 2015c) are from the period 2013–2014 for 14 countries, 2010–2012 for 30 countries, and 2005–2009 for six countries. A common challenge in estimating total fertility over time, especially for countries without accurate or complete vital registration data,3 is that estimates vary across data sources and by the methodology used to derive those estimates. Schoumaker (2014) showed that the underlying data from standardized, high-quality surveys such as the Demographic and Health Surveys vary considerably within and across countries, yielding total fertility estimates from recent fertility data of good quality (e.g., Gabon, Lesotho, Namibia, and Zimbabwe) and of poor quality (e.g., Benin, Burkina Faso, Cameroon, Chad, Ethiopia, Guinea, Madagascar, Mali, Mozambique, Niger, Nigeria, and Uganda). Total fertility estimates based on births in the last three years tend to be under-estimated by 10 percent or more in most of the surveys with poor quality fertility data from retrospective birth histories. Figure 1 illustrates the variation in total fertility estimates based on survey data and estimation methods (direct methods (D) and cohort-completed (C) fertility) for Nigeria for the period 1985 to 2015. The thick lines show the total fertility estimates from the 2010, 2012, and 2015 Revisions of WPP. Given new data from the 2008 DHS and other surveys, total fertility in the 2012 Revision was re-estimated at a higher level than the 2010 Revision beginning in the mid-1980s, resulting in almost half a birth per woman difference in the 2005–2010 period. The 2015 Revision used new data from the 2013 Demographic and Health Survey. The 2013 survey estimates highlight a recurring pattern in which fertility estimates based on a recent reference period are consistently lower than fertility estimates from reconstructed birth histories for the same time point. Looking only at fertility estimates from a three-year reference period, the 2013 DHS shows a decline in total fertility to 5.5 births per woman from a stalling pattern of 5.7 births per woman in the 2003 and 2008 DHS. Yet the absolute differences are large between these three-year reference period estimates and those for the same time point from the reconstructed birth histories: about half a birth difference in the mid-2000s (comparing the 2008 and 2013 survey estimates) and about one birth difference in the early 2000s (comparing the 2003 survey estimate to those from the 2008 and 2013 surveys). Estimates of the total fertility rate for Nigeria 1985–2015 based on various data sources and estimation methods, and WPP estimates from the 2010, 2012, and 2015 Revisions NOTES: DHS = Demographic and Health Survey; MICS = Multiple Indicator Cluster Survey; MIS = Malaria Indicator Survey. (C) refers to cohort completed fertility (i.e., average number of children ever born) for women aged 40–44 and 45–49 at the date of the survey and backdated using their mean age of childbearing. (D) refers to direct fertility estimates based on maternity histories or recent births in the 12 or 24 months preceding the survey. SOURCES: Federal Office of Statistics of Nigeria (1992); National Bureau of Statistics (2008, 2012); National Population Commission (2000, 2002, 2004, 2009, 2012, 2014). WPP fertility estimates consider as many types and sources of empirical estimates as possible, including retrospective birth histories and direct and indirect fertility estimates (Gerland 2014). The 2015 Revision updated all total fertility estimates taking into account new data and the inconsistencies among estimates. Moreover, total fertility estimates are derived to ensure as much internal consistency as possible with all other demographic components and intercensal cohorts enumerated in successive censuses (United Nations 2015d). The advantages of this approach are that the estimates are internally consistent within a country over time and with respect to other related demographic information. A disadvantage is that the estimates can depart from a country's official estimates of fertility. Figure 2 shows the estimated trends in period total fertility for sub-regions of sub-Saharan Africa from 1950 to 2015. It was high (above six births per woman) in all sub-regions in 1950–1955. Fertility remained high in Eastern and Western Africa until the 1980s, after which it began a slow decline to 4.9 births per woman in Eastern Africa and 5.5 in Western Africa in 2010–2015. Fertility in Middle Africa began to decline a decade later and more slowly, reaching 5.8 births per woman in 2010–2015. Southern Africa departed from the overall trends with a decline beginning in the 1950s and dropping below three births per woman in the 2000s. The 2010–2015 estimate of 2.5 births per woman in Southern Africa is about half the total fertility level in Eastern, Middle, and Western Africa. Sub-regional trends in total fertility, sub-Saharan Africa, 1950–2015 SOURCE: United Nations 2015a The sub-regional fertility levels mask diverse levels among countries. Figure 3 shows country-specific total fertility levels in 2010–2015. Among the 16 countries in Western Africa, total fertility ranged from 2.4 in Cabo Verde to 7.6 in Niger. One additional country had a fertility level of more than six births per woman (Mali), six countries had fertility between five and six births per woman (Burkina Faso, Côte d'Ivoire, Gambia, Guinea, Nigeria, and Senegal), and seven had fertility between four and five births per woman. Total fertility levels in countries in Africa, 2010–2015 SOURCE: United Nations 2015a. Total fertility levels ranged even more widely among the 20 countries in Eastern Africa, from 1.5 in Mauritius to 6.6 in Somalia. One additional country in Eastern Africa still had a fertility level above six in 2010–2015 (Burundi) and six countries had fertility between five and six births per woman (Malawi, Mozambique, South Sudan, Uganda, Tanzania, and Zambia). Eight countries had fertility levels between four and five births per woman. The lowest levels of fertility were in Djibouti (3.3 births per woman) and in the small island countries of Mauritius, Réunion, and Seychelles (less than three births per woman). The nine countries of Middle Africa all had fertility levels at or above four children per woman. In Angola, Chad, and DR Congo, fertility levels were six or more births per woman, and in the remaining six countries fertility levels were between four and five births per woman. While fertility in Southern Africa is largely dominated by South Africa's pattern, the range in fertility among the five countries in the sub-region is narrow, from 2.4 births per woman in South Africa to 3.6 in Namibia. Both Botswana and South Africa now have fertility levels below three births per woman. The start of the fertility transition also varies widely across sub-Saharan Africa. We analyze when and at what level of fertility a country experienced a maximum total fertility level before the onset of fertility decline. This maximum is defined as the most recent five-year time period where total fertility is within half a child of the overall maximum fertility in the country over the 1950–2015 estimation period, thus excluding random fluctuations in pre-transition fertility (Alkema et al. 2011). Figure 4 shows the diversity across countries and within sub-regions in the level and time period of the maximum fertility level before the onset of fertility decline, as assessed in the 2015 Revision. By the late 1970s, 29 sub-Saharan countries were on the verge of a fertility decline, increasing to 40 countries by the early 1980s. While the maximum fertility before the onset of fertility decline was reached in all countries in Southern Africa by the late 1970s, the range of experiences was much wider among countries in Eastern Africa (from the early 1950s in Réunion to the late 1990s in Somalia), Middle Africa (from the late 1960s in Angola to the late 1990s in Chad and DR Congo), and Western Africa (from the early 1960s in Cabo Verde to the late 1990s in Niger). Maximum total fertility in the time period before the onset of fertility transition, sub-Saharan African countries by sub-region SOURCE: United Nations 2015a. The maximum fertility before the onset of fertility transition ranged from less than six births per woman in five countries (Central African Republic, Equatorial Guinea, Gabon, Lesotho, and Seychelles) to more than eight births per woman in two countries (Kenya and Rwanda). There was a positive but weak relationship between the maximum fertility level and the timing of when fertility transition commenced (R2 = .04). The transition from the maximum fertility to the current estimated fertility level in 2010–2015 has been slow for most countries in sub-Saharan Africa. We examine both the maximum decline in fertility in a five-year period and the duration of time between the maximum fertility level and when a country achieved a 10 percent decline in total fertility. A 10 percent decline in total fertility is one of several empirical rules that researchers have used to identify when fertility has begun to decline in a sustained manner from a pre-transitional maximum (Bongaarts and Casterline 2013; Coale and Treadway 1986; United Nations 2014), including differences in the time period selected to identify the onset of fertility decline (Casterline 2001) and the magnitude of decline (e.g., 5 percent decline with further conditions for subsequent changes applied, see Bryant 2007). While the different rules all strive to distinguish sustained decline from random fluctuations in fertility levels, each rule will have different repercussions for interpretations about the timing of onset of fertility decline, duration and pace of decline, and correlations with levels of development and other indicators. In seven countries the first steps of fertility decline took place over a long time period. In Angola, Gambia, and Uganda, a 10 percent decline from the maximum fertility level took 40 years; and in Lesotho, Mozambique, Niger, and Tanzania, it took 30 to 35 years (Appendix Table 1).4 In other countries the fertility decline never gained speed. Overall declines from the country-specific maximum fertility level to the level in 2010–2015 were very slow in nine countries (Angola, Congo, Gambia, Mali, Mozambique, Niger, Nigeria, Uganda, and Tanzania), where average fertility declines were 0.2 children per woman or less, a pace at which it would take at least 25 years to realize a decline of one birth per woman (Appendix Table 1). During the fertility transition, some countries experience an acceleration of fertility decline that, based on historical experience of past transitions in Latin America and the Caribbean and Asia, might reach a decline of more than one birth per woman in a five-year period (United Nations 2015a). In most countries in sub-Saharan Africa, however, the maximum fertility decline is much smaller. The largest declines of more than one birth per five-year period were registered in the early-transition countries in Eastern Africa before 2000 (Djibouti, Mauritius, Mayotte, Réunion, Rwanda, Seychelles, and Zimbabwe). For some countries the maximum fertility decline is projected to be reached in the future: one country in Eastern Africa (Mozambique), four in Middle Africa (Angola, DR Congo, Equatorial Guinea, and Sao Tome and Principe), and five in Western Africa (Gambia, Guinea, Mali, Niger, and Nigeria). Figure 5 shows that countries reaching the maximum fertility decline later tend to reach it at lower levels (the correlation across the 50 countries is R2 = .45). Maximum five-year decline in total fertility since the onset of fertility transition, actual and projected, sub-Saharan African countries by sub-region SOURCE: United Nations 2015a. The pace of fertility decline is illustrated in Figure 6 for Ethiopia and Nigeria, the two most populous countries in sub-Saharan Africa. Estimated levels of total fertility from 1950–1955 to 2010–2015 are shown with a fitted model of the fertility transition. Both countries reached maximum fertility levels before the onset of fertility decline between the late 1970s and early 1980s, and both reached a 10 percent decline from this maximum in the early 2000s. While Ethiopia reached the peak pace of fertility decline in 2005–2010 (a decline of nearly one birth in the five-year period), Nigeria's fertility decline has been consistently slow, with a low peak pace of decline. The maximum fertility decline per five-year period in Nigeria is projected to be reached in the future (a decline of 0.3 births in 2020–2025). The difference in the pace results in current total fertility of 4.6 births per woman in Ethiopia and 5.7 in Nigeria. The 80 percent prediction intervals (dashed lines in Figure 6) around the probabilistic projections of total fertility indicate the magnitude of the different fertility changes that could reasonably happen. For example, by 2045–2050 there is a one in ten chance that total fertility in Nigeria could be as low as 2.6 or as high as 4.4 (Figure 6 and Appendix Table 2). Observed decline in total fertility from 1950–1955 to 2010–2015, predicted decline, and time period of maximum TFR and maximum decline, Ethiopia and Nigeria SOURCES: United Nations 2015a and computations by authors. Given the slow pace of fertility decline in most sub-Saharan African countries thus far and the uncertainty around the pace of future declines, what might be the impact on future fertility levels and the growth in total population if fertility declines in the region followed an accelerated pace that has already been experienced by countries that have completed or are nearing completion of the fertility transition? To answer this question, we generated probabilistic fertility scenarios for 2015–2100 using a modeling approach described in Alkema et al. (2011) and implemented in a publicly accessible software package BayesTFR (Ševčíková, Alkema, and Raftery 2011). The standard United Nations 2015 probabilistic results pool the experience of all countries having similar fertility levels and trends (the baseline scenario). An alternative probabilistic scenario was created using only the fertility transition experiences of 21 countries5 that shared a similar pattern of accelerated decline in fertility (the accelerated scenario): specifically, a slow pace at the start of the transition that sharply rises to a peak pace of decline before steadily tapering off after total fertility has reached about four births per woman. These 21 countries, which include—among countries with more than 10 million inhabitants in 2015—Bangladesh, China, El Salvador, Morocco, Peru, South Africa, Sri Lanka, Thailand, Turkey and Uzbekistan, represent disparate institutional, social, economic, and cultural contexts and yet experienced a similar pattern of a relatively abrupt acceleration of fertility decline. The objective of this exercise is to examine the impact on population size of fertility decline that is more rapid than currently projected, and to illustrate the implications of such a hypothetical scenario. While we do not theorize or analyze the factors that produced these rapid fertility declines, the fact that a similar pattern of fertility decline took place within such a diverse group of countries raises the possibility that a similar decline could occur within a region that also reflects quite distinctive and diverse contexts. Each fertility scenario is used to simulate 10,000 probabilistic population projections for 2015–2100 under identical conditions (i.e., using the same mortality and migration assumptions) that show the population growth trajectories if sub-Saharan African countries were to follow a specific fertility decline pattern.6 The implications of this specific fertility decline pattern for future fertility trends are shown in Figure 7 for Ethiopia and Nigeria. The accelerated fertility decline scenario leads to much more rapid declines for these countries. By 2045–2050, median projected total fertility in Ethiopia would decline from 2.3 (baseline) to 2.0 (accelerated decline) and in Nigeria from 3.6 (baseline) to 2.6 (accelerated decline) (see Appendix Table 2). Probabilistic fertility projections (median and 80 percent prediction intervals) for two scenarios: baseline and accelerated fertility decline, Ethiopia and Nigeria SOURCES: United Nations 2015a and computations by authors. The accelerated fertility scenario has greater implications for projected fertility in Nigeria, where the recent pace of fertility decline has been much slower than in Ethiopia. The projected medians of the accelerated scenario are similar to the lower bound of the baseline scenario for parts of the projections, suggesting that Ethiopia and Nigeria would need to experience a rapid fertility decline, similar to the rapid declines experienced by the group of 21 countries, in order to realize the lower-bound fertility level of the baseline scenario. The assumption that Ethiopia and Nigeria would follow the fast fertility decline experiences of these 21 countries leads to substantially lower population projections compared with the baseline scenario (Figure 8 and Appendix Table 3). If Ethiopia experienced an accelerated fertility decline, the projected population grows from 99 million in 2015 to 182 million by 2100 (80 percent prediction interval of 93 million to 324 million) compared with the higher median projection of 234 million under the baseline scenario. If Nigeria adopted the accelerated fertility decline pattern, the projected population grows from 182 million in 2015 to a median of 466 million by 2100 (80 percent prediction interval of 282 million to 777 million) rather than the faster projected growth to 737 million under the baseline scenario. Probabilistic population projections (median and 80 percent prediction intervals) for two scenarios: baseline and accelerated fertility decline, Ethiopia, Nigeria, and sub-Saharan Africa SOURCES: United Nations 2015a and computations by authors. The counterfactual of sub-Saharan African countries following an accelerated fertility decline, as experienced by the group of 21 countries, results in a projected increase in total population in the region from 962 million in 2015 to 3.2 billion by the end of the century (with 80 percent probability of being between 2.8 and 3.7 billion). The total population projection for the baseline UN 2015 scenario would lead to a median projection of about 4 billion, or a difference of 770 million people with an 80 percent prediction interval mostly above the upper bound of the accelerated fertility decline scenario. We presented here an updated description of fertility decline in sub-Saharan Africa and have explored the implications for fertility and population projections if sub-Saharan African countries follow a pattern of accelerated fertility decline. Because estimates of fertility in the region rely almost entirely on survey and census data, there is still uncertainty about fertility change over time. Our analyses of trends over time were based on a new historical time series of total fertility estimates for the period 1950 to 2015 from the 2015 Revision of World Population Prospects. Fertility has remained persistently high in Eastern, Middle, and Western Africa (4.9 births per woman or higher as of 2010–2015) and declined rapidly in Southern Africa, which currently has about half the total fertility level of other sub-regions births per woman). Southern Africa is dominated by South Africa, TFR of 2.4 is the lowest in the Yet there is wide variation in total fertility across countries, even within from below fertility births per woman) in Mauritius to 7.6 births per woman in Niger. The fertility level before the onset of decline was high in all countries in sub-Saharan Africa, from less than six births per woman in five countries to more than eight births per woman in two countries. By the late 1970s, 29 countries were on the verge of fertility decline, increasing to 40 countries by the early 1980s. The transition from these maximum levels of fertility prior to the onset of fertility decline to current levels has been slow for most countries, including seven in which it took 30 years or more to a 10 percent decline in total fertility. While countries in Asia and in Latin America and the Caribbean experienced an acceleration in fertility decline of more than one birth per five-year time period, this level of acceleration has been experienced thus far by only seven countries in Eastern Africa before the The pace of fertility decline in sub-Saharan Africa will a large in the magnitude of future growth in We showed through an illustrative scenario that an acceleration in the pace of fertility decline to that already experienced by 21 countries South Africa) would the growth in future population in the region from a projected billion by to billion and total population size from a projected 4 billion people by the end of the century to 3.2 While these 21 countries represent a wide range of institutional, social, economic, and cultural a similar fast decline from fertility levels of six or more children per woman in the past to less than two or three in recent years. this accelerated pace of fertility decline to countries at of transition illustrates the substantial impact on future population growth of other possible patterns of fertility decline in sub-Saharan Africa. of this are assumption that the fertility estimates from the World Population Prospects are and the fact that uncertainty around the estimates of period total fertility in recent is not into By scenarios on the distinct fertility decline patterns that have been experienced thus far by countries that have completed or are advanced in their fertility we new patterns of fertility decline in the Because most countries in Middle Africa and Western Africa are still in the early of fertility transition, an sub-Saharan African pattern of fertility decline may still and Timæus The is not for the or of by the authors. than be to the for the

Since 1998, a large amount of new evidence has become available from specialist community-based studies, national surveys and censuses. Various analytical techniques have been developed that allow us to use these data in new ways. In addition to demographers and epidemiologists, social scientists and economists are now starting to look at empirical evidence on the effects of HIV, at the family, community and national level. Although modelling was the primary basis for an assessment of the size and future of the epidemic in the past, increasingly use is made of real data to inform policy and to evaluate national responses to the epidemic. In this context it is timely and appropriate to re-visit this problem area, to summarize our wider knowledge of the demographic and social impacts of the epidemic, and to broaden the range of effects that we scrutinize critically to include macro and microeconomic impacts. These impacts of AIDS are having an increased impact on policy making: the health services are stretched; the education system is losing teachers; and numbers of orphans are growing. Policy needs to be advised by good data and rigorous analysis. A scientific meeting on the Demographic and Socioeconomic Impact of AIDS was held in Durban, South