Each year since 2002, evidence on child and adolescent health derived from randomised controlled trials (RCTs) in low- and middle-income countries (LMICs) have been summarised in a booklet and distributed widely. The aim is to provide information to review treatment policies, clinical guidelines, public health approaches, to teach about evidence-based medicine and to learn about child health in the wider world. That world has changed in the last 20 years, and there are now many more sources of information, including an expansion of social media and the number of open access journals. Twenty years ago, very few RCTs were published as open access, now a high proportion are. But still a busy paediatrician or healthcare worker may not have the time to search and find relevant publications, and with the vast information in the social media age there is added complexity. The full version of this is in online supplemental file 1, and all editions of this publication (2002–2023) are available at: https://pngpaediatricsociety.org/research-2/

The method of identifying studies uses PubMed, a search engine that is freely available and widely used throughout the world (http://www.ncbi.nlm.nih.gov/sites/entrez). The search strategy is listed in the panel. It was chosen to capture as many relevant studies as possible. This year the search yielded 4664 results, of which 540 papers were directly relevant. As in past years, meta-analyses are also included if they include RCTs from LMICs and the authors are from those countries. For some topics, such as the timing of cord clamping and chemoprophylaxis in malaria, numerous small RCTs have been published over several years. However, it is often when a meta-analysis is published that the messages for guidelines and policy become clearer.

Randomised trials are far from the only valuable scientific evidence, and some RCTs, because of problems with design or implementation have limited value. However, the method of the randomised trial is the gold standard for determining attributable benefit or harm from clinical and public health interventions. When done properly, they eliminate bias and confounding. The results should not be accepted uncritically but they should be evaluated for quality and validity. Before the result of an RCT can be generalised to another setting, there must be consideration of wider applicability or reproducibility, feasibility and potential for sustainability.

The 540 trial publications come from LMICs across all regions of the world. Many RCTs from 2023 to 2024 are expected to contribute to significant changes in child health recommendations, and some have already had an impact.

There have been a marked evolution of the content and nature of the 4200 studies published since 2002. This reflects several factors: the changing epidemiology of child health; the amazing application of new technologies in vaccines and drugs against ever-changing pathogens; the ‘grand convergence’ of health and medicine between the ‘LMICs’ and ‘high-income’ countries of last century, partly evidenced by significant increase in trials conducted in India, China and South Africa; global targets such as the Millennium and Sustainable Development Goals (SDGs); approaches like WHO’s ‘life-course’ and ‘Survive and Thrive’ frameworks and the influences (and sometimes distortions) of global funding agencies and social and political trends on the research agendas.

Encouragingly, there are now more trials addressing the broader context of those global targets in the SDG era. These include field trials of interventions to provide better water, sanitation and hygiene, as well as reduce pollution in the poorest communities; community-based programmes for home gardening and poultry rearing; trials addressing the local impacts of climate change; increased focus on the development, psychological and mental health of children and adolescents; efforts to reduce violence against children; holistic approaches to improve the health and education of adolescent girls and boys often co-designed by them and trials aimed at improving maternal health and parent-child interactions. RCTs are being conducted not just in hospitals and healthcare settings, but in schools, villages and communities.

In urban Bangladesh, a community-based early child development (ECD) social safety-net programme using home visitation improved the development of children of young mothers, including improved cognitive and motor skills, reduced experience of violence by mothers and more engagement by fathers.1 Similar positive results for community ECD were found in rural India.2

In Rwanda, Bandebereho, a programme that worked with parents to build couple relationship and parenting skills and included reflection on gender norms had lasting effects on reduced family violence and physical punishment of children, plus multiple health and relationship outcomes.3

In India, among 27 000 households where water sanitation and hygiene (WASH) training was conducted, households that engaged in a combination of four WASH characteristics: safe source of water for daily use, safe source of drinking water, private or shared flush toilet use and always handwashing with soap after defaecation had a 30% lower risk of cholera than those that did not have these four characteristics.4

In Bangladesh, several studies explored the effects of improved WASH on childhood diarrhoea. One large cluster RCT involving 360 clusters and 4941 children showed that improved WASH substantially reduces the risk of diarrhoea, with greatest benefits observed among children from the lowest socio-economic households and during the monsoon season. The study estimated that improved WASH prevented an estimated 734 cases (95% CI 385, 1085) per 1000 children per month during the seasonal monsoon.5 A substudy showed improved WASH may improve epigenetics including reduced methylation of cortisol genes, and enhanced adaptive responses of the physiological stress system in early childhood.6

Among 225 Ugandan school children with hookworm, dual-dose albendazole (400 mg/day for 2 days) improved the cure rate of hookworm compared with a single-dose of 400 mg albendazole (96% vs 84%).7 And in Gabon, a childhood hookworm vaccine underwent a phase II trial that included assessing serological responses.8

In a large cluster RCT in Bangladesh, certain enteropathogens were more prevalent in the hot rainy season, including Cryptosporidium, Escherichia coli, Shigella, Campylobacter, Aeromonas and adenovirus.9

Three trials (in Niger, Mali and Kenya) showed the benefits of involving community health workers and decentralised care in the management of children with acute malnutrition, with equivalent or greater recovery rates, shorter length of hospital stay and reduced costs.10

In a large community-based RCT in the Punjab, India, involving over 70 000 households and 15 000 births, a maternal and newborn health package—along with training for community-based and facility-based healthcare workers, and community counselling and education sessions—resulted in a reduced neonatal mortality rate (39.2/1000 live births vs 52.2/1000 live births in control clusters), improved clean delivery practices and increased use of chlorhexidine for cord care.11

In Zambia and Burkina Faso, breast fed infants without HIV, whose mothers had HIV and a viral load >1000 copies/mL, lamivudine prophylaxis in the first 12 months of life in addition to maternal antiretroviral therapy reduced the risk of postnatal mother-to-child HIV transmission.12

In Lesotho, an integrated parenting and HIV testing intervention held in 34 community clusters showed improvements in child language development, and child HIV testing.13

In Africa there were many studies further exploring aspects of azithromycin mass drug administration (MDA), including effect on mortality, pneumococcal and diarrhoeal pathogen resistance and delivery issues. In Burkina Faso, twice-yearly azithromycin resulted in a non-significant reduction in mortality in children in the presence of seasonal malaria chemoprevention. And among preschool children in Niger, mass azithromycin distribution aimed at reducing mortality was associated with an increase in macrolide resistance determinants in the gut, but no study showed a clear sign that MDA substantially increased resistance.14 15

The malaria vaccine R21/Matrix-M was well tolerated and offered high efficacy against clinical malaria in over 3000 children in 4 African countries.16 And in implementation trials of the RTS,S/AS01E malaria vaccine (RTS,S), introduced by national immunisation programmes in Ghana, Kenya and Malawi, over 650 000 children had received at least one dose, and RTS,S vaccine introduction was associated with a 32% reduction (95% CI 5% to 51%) in hospital admission with severe malaria, and a 9% reduction (95% CI 0% to 18%) in all-cause (non-injury) mortality.17

In Tanzania, in a study of over 4500 people, chlorfenapyr-pyrethroid impregnated bed nets reduced the incidence of malaria infection compared with standard pyrethroid insecticide-treated bed nets over the 3-year bed-net lifespan.18 In a parallel study in Benin, the beneficial effect on malaria transmission of chlorfenapyr-pyrethroid impregnated bed nets waned by the third year, likely related to reduced bed-net usage.19

In a systematic review of trials in malaria-endemic countries in Africa, postdischarge malaria chemoprevention with sulfadoxine-pyrimethamine or artemether-lumefantrine administered monthly for 2 or 3 months, or until the end of the malaria season, was found to reduce mortality by 77% (rate ratio (RR) 0.23 (95% CI 0.08 to 0.70), p<0.001) and decrease readmissions in recently discharged children recovering from severe anaemia.20

A systematic review showed a significant impact of wMel-Wolbachia-carrying Aedes aegypti mosquitoes in preventing dengue infection in an endemic setting, mostly based on study involving over 6000 participants in Yogyakarta, Indonesia. The odds of contracting virologically confirmed dengue were reduced by 77% (OR 0.23, 95% CI 0.15 to 0.35).21

The Butantan-Dengue Vaccine, a single-dose, live, attenuated, tetravalent vaccine against dengue, provided high protective efficacy over 2 years of follow-up in a study in Brazil involving over 10 000 vaccine recipients.22

In Malawi, as shown elsewhere, a single dose of typhoid conjugate vaccine (Vi-TT) is highly efficacious for at least 4 years among children aged 9 months to 12 years, including those younger than 2 years.23

In a large RCT involving 680 patients (including adolescents) with rifampicin-resistant pulmonary tuberculosis in Uzbekistan, Belarus and South Africa, a 24-week regimen of oral bedaquiline, pretomanid, linezolid and moxifloxacin resulted in fewer adverse outcomes (a composite of treatment failure, death, treatment discontinuation, disease recurrence or loss to follow-up) compared with 36 months of standard care (12% vs 41%).24

In children in Bangladesh with persistent diarrhoea, green banana mixed with rice suji (semolina) was found to be more effective than rice alone for managing persistent diarrhoea in young children.25

Among children with acute diarrhoea and severe dehydration accompanied by severe non-anion gap metabolic acidosis, rehydration with Hartmann’s solution plus additional bicarbonate to correct the deficit led to earlier resolution of metabolic acidosis, less utilisation of critical care facilities and fewer adverse outcomes compared with rehydration with Hartmann’s solution alone.26

In 100 mechanically ventilated children in a paediatric intensive care unit in India, restricting fluids to 40% of maintenance compared with 70%–80% resulted in lower fluid overload at 7 days, more ventilator-free days and a lower mortality rate, all of which were non-significant trends.27

In a stepped-wedge cluster RCT conducted across 20 hospitals in Uganda involving over 2000 children with pneumonia, improving access to oxygen with solar-powered systems resulted in a mortality benefit, with a relative risk reduction of 48.7% (95% CI 8.5 to 71.5).28

In Ethiopian general hospitals, the introduction of locally made bubble-continuous positive airway pressure (CPAP), supervised by general practitioners and paediatricians, was associated with a reduced risk of treatment failure in children with severe pneumonia and hypoxaemia compared with standard low-flow oxygen therapy. Additionally, bubble-CPAP was associated with a sustained lower mortality over 9 years in a tertiary hospital in Dhaka, Bangladesh.29

In a meta-analysis, children with asthma treated with fluticasone and salmeterol, compared with fluticasone alone, experienced a greater proportion of time asthma-symptom free and had a lower need for short-acting beta-2 agonists after 12 weeks of treatment.30

In China, a meta-analysis of 24 RCTs involving over 2000 patients with mycoplasma pneumonia found that the combination of budesonide with azithromycin improved lung function, reduced inflammatory markers and reduced the duration of symptoms.31

Among 60 Indian children with new-onset type 1 diabetes, a blinded RCT showed that 6 months of supplementation with probiotics containing lactobacillus and bifidobacterium improved glycaemic control, resulting in a greater reduction in haemoglobin A1C levels, and improved immunoregulatory markers of islet cell function, compared with placebo.32 This year, other studies on type 1 diabetes in India explored the role of milk and calcium supplementation to improve bone health among children in underprivileged communities, as well as the adjunctive use of metformin to improve glycaemic control in adolescents.

Among 767 children in Indonesia with congenital heart disease undergoing cardiac surgery, tri-iodothyronine supplementation was associated with reduced duration of ventilation and shorter hospital stays. However, this effect was not observed in a parallel study conducted in the USA, where malnutrition rates were lower and baseline TSH and T3 were normal, unlike the lower levels seen in Indonesia.33

In 160 Indian children with frequently relapsing nephrotic syndrome, levamisole (2–2.5 mg/kg on alternate days) reduced the risk of relapse from 40% to 23%, was steroid-sparing and lowered the risk of steroid toxicity.34 And daily dosing of levamisole may reduce use of steroids more than second daily dosing, in frequently relapsing nephrotic syndrome.35

In Uganda, among children with Nodding syndrome, a neurological disorder in Africa that carries a high mortality rate, treatment with doxycycline reduced acute seizure-related hospitalisations (RR 0.43 (95% CI 0.20 to 0.94), p=0.028) and deaths (RR 0.46 (0.24 to 0.89), p=0·028). This was based on the hypothesis that nodding syndrome is a neuroinflammatory disorder, induced by antibodies to Onchocerca volvulus.36

In two RCTs in India and China, among mothers and their newborns, increasing the duration of early skin-to-skin contact from 60 to 90 min increased the likelihood of exclusive breastfeeding rates in the first 14 weeks of life.37

In a systematic review of 20 studies involving 3260 infants, deferred cord clamping, compared with immediate cord clamping, reduced death before discharge (OR 0.68 (95% CI 0.51 to 0.91).38

In a systematic review of 24 studies involving 1100 newborns, the use of common salt for treating umbilical granulomas had a treatment success rate of nearly 94%.39

In 286 preterm infants (gestational age 23–30 weeks) in China, breast milk enema reduced the time to achieve meconium evacuation and full enteral feeding compared with saline enema.40

Two systematic reviews investigated whether a shorter duration of antibiotic therapy (7–10 days) is equivalent to the traditional longer duration of therapy (10–14 days) in terms of treatment failure or mortality for both culture-proven and culture-negative neonatal sepsis. For culture-proven sepsis, the numbers were too small to detect a difference in mortality, and no apparent difference in treatment failure was observed. For culture-negative sepsis, short-course antibiotic therapy was equally effective in all measures of morbidity and significantly shortened hospitalisation.41 42

In a meta-analysis of RCTs involving 26 studies and 132 000 very low birthweight babies, the prevalence of chronic lung disease as defined by bronchopulmonary dysplasia or oxygen requirement at 1 month of age was 35% (95% CI 28% to 42%).43

In a meta-analysis of 71 RCTs on cooling for neonatal hypoxic ischaemic encephalopathy, involving 5821 surviving infants assessed for hearing impairment, the prevalence rate of hearing loss in LMICs was 7%–8%.44

In a meta-analysis of 106 RCTs involving nearly 24 000 participants, vitamin A supplementation markedly reduced the incidence of retinopathy of prematurity (ROP) in preterm infants compared with placebo (OR 0.59, 95% credible interval (CrI) 0.33, 0.85). Also effective were probiotics (OR 0.48, 95% CrI 0.32, 0.97), human milk (OR 0.50, 95% CrI 0.21, 0.78), fish oil-based lipid emulsion (OR 0.57, 95% CrI 0.24, 0.90) and early erythropoietin (OR 0.51, 95% CrI 0.34, 0.98).45

In exclusively breastfed Gambian infants, iron supplementation starting at 6 weeks of age was associated with a significant improvement in markers of iron status by around 6 months of age.46

Among Kenyan infants, adding prebiotics to iron-fortified infant cereal increases iron absorption and reduces the adverse effects of iron on the gut microbiome and inflammation.47

In an RCT involving year 8 students in 12 public schools in north India, a health-promotion intervention aimed at reducing the behavioural risk factors of chronic diseases led to a 5% reduction in salt intake and the proportion of current alcohol users, an 18 g/day increase in fruit consumption and increased physical activity. However, there was no effect on smoking rates.48

Among 232 children living in rural India, a programme of daily protein supplementation, yoga and physical exercise improved muscle function.49

Among 750 Ugandan children aged 1–5 years with stunting, 12 weeks of lipid-based nutritional supplementation containing milk or soy protein and whey or maltodextrin substantially increased haemoglobin levels and improved iron, cobalamin and folate status, but had no effect on vitamin A compared with children who received no supplementation.50

Positive results were observed from peer support groups for adolescents with chronic illnesses, including HIV in South Africa,51 and from school health programmes in Tanzania that included school meals, nutrition education, school gardens and community workshops.52

Among 224 sexually active adolescent girls in South Ethiopia, an RCT of a school-based peer education intervention effectively improved contraceptive use and reduced unmet needs.53

In rural India, in a controlled trial involving 235 adolescent school students, high-dose vitamin D supplementation (2250 IU/day for 9 weeks) reduced depression scores compared with a group who received a lower dose of vitamin D (250 IU/day for 9 weeks) along with calcium supplementation.54

In 2024, four studies involving refugee populations were published: addressing interventions to improve the psychosocial health of mothers and infants among Syrian refugees in Egypt55 and adolescent refugees in Lebanon56; to improve hand-washing practices in a refugee settlement in Sudan57 and to increase the very low rates of vaccine coverage in Lebanon.58

This represents an extraordinary output of studies from many countries with limited resources for research. The implications for child health are many, including the need for time, capacity and resources to assess the generalisability and applicability of findings in different contexts, as well as to translate research into policy and practice where appropriate. Some research will simply reinforce existing approaches, some will encourage wider implementation and some may eventually change both global and local practices. Conversely, the efficacy of certain approaches demonstrated in small trials may not be confirmed by further research. National systems and more local translational research capacity and resources are needed in LMICs, and paediatricians everywhere have an important role to play.

Rather than being satisfied with the current research, we should ask ourselves which children in the world still do not benefit from these advancements. In industry and business, research and development go hand in hand; in medicine and health, this is often not the case. And research priorities are not always proportional to the burden of the problem: there are, for example, 47 million refugees under the age of 18 years in the world. Not all settings or problems are suitable for RCTs, and other methods of implementation research will be more appropriate in some settings and addressing certain problems. Much more support is needed for developing clinical and public health research capacity in the low-income countries, with a focus on children living in the poorest and most volatile conditions. The SDGs urge us to reduce inequities to improve child health, and clinical and public health researchers have a crucial role to play in this effort.