BACKGROUND

Large, randomized controlled trials (RCTs) are essential in answering pivotal questions in child health.

METHODS

We created a bird’s eye view of all large, noncluster, nonvaccine pediatric RCTs with ≥1000 participants registered in ClinicalTrials.gov (last search January 9, 2020). We analyzed the funding sources, countries, outcomes, publication status, and correlation with the pediatric global burden of disease (GBD) for eligible trials.

RESULTS

We identified 247 large, nonvaccine, noncluster pediatric RCTs. Only 17 mega-trials with ≥5000 participants existed. Industry funding was involved in only 52 (21%) and exclusively funded 47 (19%) trials. Participants were from high-income countries (HICs) in 100 (40%) trials, from lower-middle-income countries (LMICs) in 122 (49%) trials, and from both HICs and LMICs in 19 (8%) trials; 6 trials did not report participants’ country location. Of trials conducted in LMIC, 43% of investigators were from HICs. Of non-LMIC participants trials (HIC or HIC and LMIC), 39% were multicountry trials versus 11% of exclusively LMIC participants trials. Few trials (18%; 44 of 247) targeted mortality as an outcome. 35% (58 of 164) of the trials completed ≥12 months were unpublished at the time of our assessment. The number of trials per disease category correlated well with pediatric GBD overall (ρ = 0.76) and in LMICs (ρ = 0.69), but not in HICs (ρ = 0.29).

CONCLUSIONS

Incentivization of investigator collaborations across diverse country settings, timely publication of results of large pediatric RCTs, and alignment with the pediatric GBD are of pivotal importance to ultimately improve child health globally.

What’s Known on This Subject:

The pediatric global burden of disease remains high, with large inequalities across the world. Large, pediatric randomized trials are pivotal in providing high quality evidence for clinicians and policy makers.

What This Study Adds:

There is a dearth of large nonvaccine, noncluster, pediatric randomized trials. Trial investigators are almost exclusively from high-income countries, and 35% of completed trials remained unpublished ≥12 months from completion.

Historically, children are an underserved population in clinical research.1  Looking at the pediatric global burden of disease (GBD) sheds light on key patterns. Mortality disproportionately affects children aged <5 years.2  Geographic variations exist as well. One-half of the diarrheal deaths among children occur in just 5 countries.3  As of today, almost 90% of the world’s pediatric population lives in nations deemed lower-middle-income countries (LMICs),2,4  where many children invariably face challenges in realizing their full developmental potential. A previous analysis of pediatric randomized controlled trials (RCTs) and nonrandomized clinical trials registered in ClinicalTrials.gov between 2006 and 2012 revealed only a moderate correlation with the pediatric GBD.5  There is a strong need for the pediatric clinical research agenda to address the diseases with the highest pediatric GBD.3,5  We set up to perform an empirical evaluation of large, noncluster, nonvaccine, pediatric RCTs registered in ClinicalTrials.gov to obtain a global health insight into the current state of pediatric research agenda involving large trials and address the following questions: How many large, pediatric RCTs exist? Who is funding these trials? Where are the investigators and the participants from? How many of these large trials are targeting mortality outcomes? How many of these large trials are published after completion? What clinical conditions are targeted? Is the pediatric research agenda of these large RCTs aligned with the pediatric GBD? Answering these questions can help guide the future of the global pediatric research agenda.

We performed a descriptive analysis of all large, noncluster, nonvaccine, pediatric RCTs registered in ClinicalTrials.gov as of January 9, 2020 with ≥1000 participants (enrolled or planned to be enrolled). ClinicalTrials.gov is a widely used trial registry platform5,6  with 45 557 registered interventional clinical trials with participants aged 0 to 17 years, as of January 9, 2020.7  We excluded cluster RCTs, vaccine trials, nonrandomized trials, trials with targeted recruitment size <1000, trials with an unreported recruitment sample size, and trials with participants from mixed age groups (adults and children) (Fig 1). We identified 508 potentially eligible trials that were further reviewed by D.C.I, S.C., and T.K. Trials that randomly assigned dyads (eg, parent and child) or households were considered eligible for inclusion if the sample size of randomized units was ≥1000. For trials with unclear unit of randomization (cluster versus individual participants), we attempted to contact the trial investigators for additional clarifications (we contacted 3 investigators of trials with an unclear unit of randomization and received 1 response confirming the trial was a cluster RCT; the remaining 2 trials without an E-mail response were excluded). Ultimately, 247 large, noncluster, nonvaccine, exclusively pediatric RCTs were included in our analysis.

FIGURE 1

Flowchart. More than 1 exclusion criteria may apply per trial.

FIGURE 1

Flowchart. More than 1 exclusion criteria may apply per trial.

Close modal

We automatically extracted information for each trial registered in ClinicalTrials.gov using the clinicaltrialr package in R (R Foundation for Statistical Computing, Vienna, Austria)8  for the following: National Clinical Trial (NCT) unique identifier number, trial registration year, trial start year, age group of enrolled children, recruitment sample size (anticipated and actually enrolled), funding source(s), country of investigators, collaborators, and/or sponsors, country of enrolled participants, targeted conditions and interventions, listed primary and secondary outcome(s), trial’s recruitment status and enrollment sample size (the anticipated sample size was used if the trial was recruiting, suspended, or active not recruiting), and trial’s publication status for completed trials.

Funding Sources

Funding sources were further categorized into nonprofit, foundations, or not-for-profit, public (national or international), and/or industry. The primary difference between a not-for-profit and nonprofit organization is that donations to nonprofits are deemed tax-exempt by the Internal Revenue Service.9  Thus, not-for-profit organizations are not required to operate for the benefit of the public good and may simply serve the goals of their members, unlike nonprofits, which must provide a public benefit.10  Nonprofit sources include the Wellcome Trust and Children’s Miracle Network; examples of foundations include the Bill and Melinda Gates Foundation and the Children’s Investment Fund Foundation; examples of not-for-profit organizations include the Nutrition Third World and Medicines for Malaria Venture; examples of public funding include governmental sources, such as the National Institutes of Health, the US Agency for International Development, UK Medical Research Council or international sources, such as the World Health Organization, Joint United Nations Program on HIV/AIDS, and World Food Program.

Countries: High-Income Countries and LMICs

Trials were categorized as investigators’ multicountry trials (and participants’ multicountry trials, respectively), if there were ≥2 listed countries. Furthermore, trials were categorized as investigators' high-income country (HIC) trials, LMIC trials, or mixed HIC and LMIC trials, and, respectively, participants' HIC trials, LMIC trials, or mixed HIC/LMIC trials, according to the economic development status of the country of investigators and participants. We used the World Bank’s definition for HIC and LMIC.11  All low-income, lower-middle-income, and upper-middle-income economies were classified as LMICs, with the exception of Eastern European upper-middle-income countries, which were classified along with HICs because of similar operational research standards across all other Eastern European (EE) HICs.12,13  Of the 8 trials with participants in upper-middle income, EE countries, only 1 trial’s classification was impacted by the above recategorization; NCT 01211145 trial with participants in Serbia (among several other HICs) was classified as a HIC trial instead of mixed HIC/LMIC trial; the remaining 7 trials were already classified as mixed HIC/LMIC trials. There were 18 additional trials with participants from other EE countries which were classified according to the World Bank’s country-classification. The above reclassification did not impact the Investigators Countries’ income status for any of the study trials.

Interventions

Trial interventions were categorized into 12 intervention categories: drugs (systemic or topical), biological agents (eg, intravenous immunoglobulin, transplants), nutritional supplements (eg, zinc, iron, protein), behavioral interventions, medical algorithms, medical devices, procedures (eg, radiation), infant formulas, diagnostics, surgical interventions, and traditional medicine.

Outcomes

Primary and secondary trial outcomes were categorized into mortality, nonmortality, or composite outcomes.

Recruitment Status

Trial’s recruitment status (completed, recruiting, enrolling by invitation, active not recruiting, suspended, terminated, or unknown status) was automatically extracted from the ClinicalTrials.gov registry on June 23, 2020. The investigators of 56 trials with statuses active not recruiting, suspended, terminated, or unknown status were contacted, and trial status was updated accordingly on August 10, 2020 on the basis of their E-mail responses.

Publication Status

For the identification of the publication status, we searched PubMed for any publication(s) associated with the unique NCT trial identifier number for all 247 registered trials (last search October 5, 2020). We analyzed the number of trials with published trial results for completed and for trials completed ≥ 12 months at the time of our last assessment. PubMed citations linked to the NCT trial identification referring to only pilot studies, feasibility studies, or partial result publications were not considered full-result publications.

Disease Categories and Pediatric GBD

Targeted condition(s) were categorized according to the 22 Institute for Health Metrics and Evaluation (IHME) level 2 disease condition categorization.14  A total of 18 trials targeted multiple conditions, and 8 trials were not suitable for categorization under any of the 22 IHME conditions (eg, inpatient survey experience, pediatric warning system). These 8 trials were excluded from the GBD tree-maps and GBD-correlation plots. Data on the pediatric GBD were obtained from the University of Washington’s IHME database.14  We selected the disability-adjusted life-years (DALYs) metric, which is the most commonly used GBD-metric that combines both the mortality and the morbidity from a condition15  and accounts for years lost because of premature death, disability, and illness.16  We extracted the 2017 GBD estimates for all children aged 0–4 and 5–14 years. We combined the DALYs corresponding to the IHME’s 22 level 2 disease categories for each disease category and for the 2 age group categories.

We analyzed the number of large, nonvaccine, noncluster pediatric trials registered in ClinicalTrials.gov, study characteristics, the number of mega-trials with ≥5000 participants (and targeted diseases thereof), the relative role of nonindustry versus industry funding, the number of trials with multicountry (and multisite, respectively) patients’ enrollment, the number of trials with multicountry (and multisite respectively) investigators’ collaborations, the number of trials with investigators from HICs versus LMICs, respectively, the role of investigators from HICs in trials performed in HICs versus LMICs versus mixed HICs/LMICs, the number of trials (and targeted diseases thereof) targeting mortality as a primary or secondary outcome, the recruitment status of these trials, and the publication status of trials that were completed ≥12 months from the time of our assessment. We also analyzed the number of trials (and participants thereof) targeting each of the 22 IHME disease categories and estimated the correlation between the number of trials with the pediatric GBD by using the Spearman’s rank correlation coefficient.

Our primary unit of analysis for the above analyses was the number of large trials.

The data are openly available under a CC-BY-NC 4.0 license, and the code is available under a GPL 3.0 license on GitHub at https://github.com/serghiou/pediatric-clinical-trials. A new R package called clinicaltrialr was developed for the purposes of easily downloading salient data about registered trials from ClinicalTrials.gov: https://github.com/serghiou/clinicaltrialr.

We identified 247 large, nonvaccine, noncluster pediatric RCTs registered in ClinicalTrials.gov, with a total of 651 429 children (enrolled or planned to be enrolled). (Table 1, Supplemental Table 3, Fig 1). These large, pediatric RCTs represented 0.5% (247 of 45557) of all interventional, randomized or nonrandomized, pediatric trials registered in ClinicalTrials.gov. There were only 17 pediatric mega-trials with ≥5000 participants. The start date of clinical trials spanned a period from January 1988 to June 2021, and there were 127 trials registered retrospectively. The majority of these trials (67%) pertained to children <5 years of age (Table 1, Supplemental Fig 4).

TABLE 1

Trial Characteristics of the 247 Large Pediatric RCTs Included in This Study

Trial Characteristics (N = 247)n (%), n, or Median
Trial start year, n (%)  
 Before 2000 15 (6) 
 2000–2010 105 (43) 
 2011–2020 127 (51) 
Trial registration year, n (%)  
 Before 2000 5 (2) 
 2000–2010 108 (44) 
 2011–2020 134 (54) 
No. children, n 651 429 
Target recruitment size, n (%)  
 1000–2500 194 (79) 
 2501–5000 37 (15) 
 5001–10 000 9 (4) 
 >10 000 7 (3) 
All trials estimated recruitment size, median (IQR) 1652 (1200-2400) 
Mega-trials estimated recruitment size, median (IQR) 8402 (6346-21 500) 
Top 10 diseases,an (%)  
 Maternal and neonatal disorders 49 (19) 
 Nutritional deficiencies 40 (16) 
 Respiratory infections and TB 38 (15) 
 NTDs and malaria 28 (11) 
 Other noncommunicable diseases 23 (9) 
 Enteric infections 14 (6) 
 Neoplasms 12 (5) 
 Chronic respiratory diseases 8 (3) 
 Other infectious diseases 8 (3) 
 Mental disorders 7 (3) 
Top 10 interventions,an (%)  
 Drug 99 (40) 
 Supplement 51 (21) 
 Behavioral intervention 25 (10) 
 Medical device 16 (6) 
 Medical algorithm 15 (6) 
 Topical 9 (4) 
 Procedure 8 (3) 
 Biological 7 (3) 
 Infant formula 7 (3) 
 Diagnostic 2 (1) 
Primary focus of trials, n (%)  
 Treatment 118 (48) 
 Prevention 110 (44) 
 Diagnostic 12 (5) 
 Screening 7 (3) 
Primary outcomes, n (%)  
 Mortality 16 (7) 
 Nonmortality 198 (80) 
 Composite 33 (13) 
Secondary outcomes, n (%)  
 Mortality 28 (11) 
 Nonmortality 168 (68) 
 Composite 14 (6) 
 None 37 (15) 
Trial Characteristics (N = 247)n (%), n, or Median
Trial start year, n (%)  
 Before 2000 15 (6) 
 2000–2010 105 (43) 
 2011–2020 127 (51) 
Trial registration year, n (%)  
 Before 2000 5 (2) 
 2000–2010 108 (44) 
 2011–2020 134 (54) 
No. children, n 651 429 
Target recruitment size, n (%)  
 1000–2500 194 (79) 
 2501–5000 37 (15) 
 5001–10 000 9 (4) 
 >10 000 7 (3) 
All trials estimated recruitment size, median (IQR) 1652 (1200-2400) 
Mega-trials estimated recruitment size, median (IQR) 8402 (6346-21 500) 
Top 10 diseases,an (%)  
 Maternal and neonatal disorders 49 (19) 
 Nutritional deficiencies 40 (16) 
 Respiratory infections and TB 38 (15) 
 NTDs and malaria 28 (11) 
 Other noncommunicable diseases 23 (9) 
 Enteric infections 14 (6) 
 Neoplasms 12 (5) 
 Chronic respiratory diseases 8 (3) 
 Other infectious diseases 8 (3) 
 Mental disorders 7 (3) 
Top 10 interventions,an (%)  
 Drug 99 (40) 
 Supplement 51 (21) 
 Behavioral intervention 25 (10) 
 Medical device 16 (6) 
 Medical algorithm 15 (6) 
 Topical 9 (4) 
 Procedure 8 (3) 
 Biological 7 (3) 
 Infant formula 7 (3) 
 Diagnostic 2 (1) 
Primary focus of trials, n (%)  
 Treatment 118 (48) 
 Prevention 110 (44) 
 Diagnostic 12 (5) 
 Screening 7 (3) 
Primary outcomes, n (%)  
 Mortality 16 (7) 
 Nonmortality 198 (80) 
 Composite 33 (13) 
Secondary outcomes, n (%)  
 Mortality 28 (11) 
 Nonmortality 168 (68) 
 Composite 14 (6) 
 None 37 (15) 

NTD, neglected tropical disease; IQR: interquartile range.

a

Denotes that there are trials that are represented >1 time; thus, the total may exceed 100%. Numbers were rounded to whole numbers; thus, total percentage may not equal 100%.

Almost 50% (123 of 247) of these large trials were exclusively funded by nonprofit organizations, foundations, or public organizations. Industries funded only 21% (52 of 247) of trials and exclusively funded only 19% (47 of 247) (Table 2, Supplemental Fig 5).

TABLE 2

Funding Source, Geographic Characteristics (Countries of Investigators and Countries of Participants Thereof) and Publication Status of Large Pediatric RCTs

All (N = 247)aLMIC Participants (n = 122)bNon-LMIC Participants (n = 119)c
Funding sources, n (%)    
 A: foundation or nonprofit 29 (12) 26 (21) 3 (3) 
 B: industry 47 (19) 9 (7) 38 (32) 
 C: public 77 (31) 27 (22) 50 (42) 
 A + B 3 (1) 3 (2) 0 (0) 
 A + C 17 (7) 11 (9) 6 (5) 
 B + C 5 (2) 4 (3) 1 (1) 
 A + B + C 2 (1) 1 (1) 1 (1) 
 Not reported 67 (27) 41 (34) 26 (22) 
Participants’ locations, n (%)    
 Multicountry trials 61 (25) 14 (11) 47 (39) 
 Multisite trials 106 (43) 32 (26) 74 (62) 
Investigators’ locations, n (%)    
 Multicountry trials 37 (15) 24 (20) 13 (11) 
 Multisite trials 56 (23) 33 (27) 23 (19) 
Top 10 countries of study investigators, n (%)d    
 United States 108 (44) 36 (30) 72 (61) 
 United Kingdom 33 (13) 20 (16) 13 (11) 
 China 15 (6) 15 (12) 0 (0) 
 Germany 14 (6) 1 (1) 13 (11) 
 Canada 13 (5) 2 (2) 11 (9) 
 India 12 (5) 11 (9) 1 (1) 
 France 8 (3) 2 (2) 6 (5) 
 Denmark 8 (3) 4 (3) 4 (3) 
 Belgium 7 (3) 3 (2) 4 (3) 
 Finland 7 (3) 0 (0) 7 (6) 
 Not provided 2 (1)   
Top 10 countries of study participants, n (%)d    
 United States 73 (30) 73 (61) 
 Germany 28 (11) 28 (24) 
 Canada 27 (11) 27 (23) 
 United Kingdom 23 (9) 23 (19) 
 Australia 18 (7) 18 (15) 
 India 18 (7) 15 (13) 3 (3) 
 Italy 18 (7) 18 (15) 
 Poland 18 (7) 18 (15) 
 Spain 17 (7) 17 (14) 
 Belgium 15 (6) 15 (13) 
 Not provided 6 (2)   
Trial status (as of August 10, 2020), n (%)    
 Completed 172 (70) 88 (72) 84 (71) 
 Recruiting 35 (14) 11 (9) 24 (20) 
 Enrolling by invitation 2 (1) 0 (0) 2 (2) 
 Active, not recruiting 10 (4) 6 (5) 4 (3) 
 Not yet recruiting 14 (6) 11 (9) 3 (3) 
 Suspended 2 (1) 0 (0) 2 (2) 
 Terminated 3 (1) 1 (1) 2 (2) 
 Unknown status 9 (4) 5 (4) 4 (3) 
Publication status of completed trials, n (%)e    
 Published full results 106 (62) 55 (63) 51 (61) 
 Published pilot or partial results 5 (3) 1 (1) 4 (5) 
 Did not publish full, pilot, or partial results 61 (25) 32 (36) 29 (35) 
All (N = 247)aLMIC Participants (n = 122)bNon-LMIC Participants (n = 119)c
Funding sources, n (%)    
 A: foundation or nonprofit 29 (12) 26 (21) 3 (3) 
 B: industry 47 (19) 9 (7) 38 (32) 
 C: public 77 (31) 27 (22) 50 (42) 
 A + B 3 (1) 3 (2) 0 (0) 
 A + C 17 (7) 11 (9) 6 (5) 
 B + C 5 (2) 4 (3) 1 (1) 
 A + B + C 2 (1) 1 (1) 1 (1) 
 Not reported 67 (27) 41 (34) 26 (22) 
Participants’ locations, n (%)    
 Multicountry trials 61 (25) 14 (11) 47 (39) 
 Multisite trials 106 (43) 32 (26) 74 (62) 
Investigators’ locations, n (%)    
 Multicountry trials 37 (15) 24 (20) 13 (11) 
 Multisite trials 56 (23) 33 (27) 23 (19) 
Top 10 countries of study investigators, n (%)d    
 United States 108 (44) 36 (30) 72 (61) 
 United Kingdom 33 (13) 20 (16) 13 (11) 
 China 15 (6) 15 (12) 0 (0) 
 Germany 14 (6) 1 (1) 13 (11) 
 Canada 13 (5) 2 (2) 11 (9) 
 India 12 (5) 11 (9) 1 (1) 
 France 8 (3) 2 (2) 6 (5) 
 Denmark 8 (3) 4 (3) 4 (3) 
 Belgium 7 (3) 3 (2) 4 (3) 
 Finland 7 (3) 0 (0) 7 (6) 
 Not provided 2 (1)   
Top 10 countries of study participants, n (%)d    
 United States 73 (30) 73 (61) 
 Germany 28 (11) 28 (24) 
 Canada 27 (11) 27 (23) 
 United Kingdom 23 (9) 23 (19) 
 Australia 18 (7) 18 (15) 
 India 18 (7) 15 (13) 3 (3) 
 Italy 18 (7) 18 (15) 
 Poland 18 (7) 18 (15) 
 Spain 17 (7) 17 (14) 
 Belgium 15 (6) 15 (13) 
 Not provided 6 (2)   
Trial status (as of August 10, 2020), n (%)    
 Completed 172 (70) 88 (72) 84 (71) 
 Recruiting 35 (14) 11 (9) 24 (20) 
 Enrolling by invitation 2 (1) 0 (0) 2 (2) 
 Active, not recruiting 10 (4) 6 (5) 4 (3) 
 Not yet recruiting 14 (6) 11 (9) 3 (3) 
 Suspended 2 (1) 0 (0) 2 (2) 
 Terminated 3 (1) 1 (1) 2 (2) 
 Unknown status 9 (4) 5 (4) 4 (3) 
Publication status of completed trials, n (%)e    
 Published full results 106 (62) 55 (63) 51 (61) 
 Published pilot or partial results 5 (3) 1 (1) 4 (5) 
 Did not publish full, pilot, or partial results 61 (25) 32 (36) 29 (35) 

Numbers were rounded to whole numbers; thus, total percentage may not equal 100%.

a

Six trials did not have information regarding participants’ location, and 2 trials did not have information for investigator’s location. Non-LMIC participant trials include HIC and LMIC trials.

b

Of the 122 LMIC participants' trials, 52 were done by HIC investigators, 49 were done by LMIC investigators, and 21 were done by both HIC and LMIC investigators.

c

One hundred non-LMIC participants trials had participants exclusively in HICs (98 of them also with investigators in HICs, 1 with investigators in LMICs and 1 with investigators in both HICs and LMICs) and 19 trials had participants in both HICs and LMICs (18 with investigators in HICs and 1 with investigators in both HICs and LMICs).

d

Denotes that investigators of individual trials could have been from >1 country and/or that participants of individual trials could have been enrolled in >1 country.

e

One hundred seventy-two trials were published (88 trials with LMIC participants). Median time to publication of full results was 751 d (interquartile range : 567-1128 d).

Investigators and participants were from 53 and 93 countries, respectively (Table 2) (Supplemental Figs 610). The top 3 listed countries of investigators were the United States, United Kingdom, and China (Table 2, Supplemental Fig 13); the top 3 listed countries where children were enrolled or planned to be enrolled were the United States, Germany, and Canada (Table 2, Supplemental Fig 14). Approximately 25% (61 of 247) of the large trials enrolled or planned to enroll children in multiple countries, and 43% (106 of 247) in multiple sites. Additionally, 15% (37 of 247) of large RCT investigators were from multiple countries, and 23% (56 of 247) were from multiple sites (Table 2). Although 39% (47 of 119) of non-LMIC participants trials were multicountry and 62% (74 of 119) were multisite, only 11% (14 of 122) and 26% (32 of 122) of exclusively LMIC participants trials were multicountry and multisite trials, respectively. For the majority of the trials, investigators were exclusively from HICs (70%, 172 of 247), whereas 20% (50 of 247) of the trial’s investigators were exclusively from LMICs; investigators from only 9% (23 of 247) of trials were from both HICs and LMICs (Table 2, Supplemental Fig 18). The countries of collaborators and sponsors are revealed in Supplemental Figs 11, 12, 15, and 16.

Approximately 40% (100 of 247) of pediatric participants were from HICs, 49% (122 of 247) from LMICs, and 8% (19 of 247) from HICs and LMICs (Table 2); 6 trials did not report the participants’ country. Trial investigators of patients enrolled in HICs or HICs and LMICs were almost always from HICs (Supplemental Fig 19). Approximately 97% (97 of 100) of investigators of trials enrolling patients in HICs were from HICs. Similarly, 95% (18 of 19) of investigators of trials enrolling patients in both HICs and LMICs were from HICs. Trial investigators of patients enrolled in LMICs were more diversified; 43% (52 of 122) were from HICs, 40% (49 of 122) from LMICs, and 17% (21 of 122) from both HICs and LMICs.

Only a minority of trials (18%; 44 of 247) targeted mortality as an outcome; 7% (16 of 247) as a primary outcome and 11% (28 of 247) as a secondary outcome (Table 2, Supplemental Table 4, Supplemental Fig 21).

Overall, 70% (172 of 247) of the registered trials were completed and 66% (164 of 247) had been completed for ≥12 months at the time of our assessment (as of October 5, 2020); another 35 trials were still recruiting (Table 2). Premature termination of participant enrollment occurred in one trial because of the COVID-19 pandemic, and the trial’s status was changed to completed status; 2 trials suspended recruitment because of the COVID-19 pandemic.

We identified 106 publications with full trial results, corresponding to a 62% publication rate (106 of 172) among completed trials and 65% (106 of 164) among those completed ≥12 months from the time of our assessment (Table 2). Among trials with published results, the majority (65%; 69 of 106) reached the anticipated target recruitment; and, among trials not reaching the exact target sample size, only 7 trials did not reach 1000 participants sample size. For trials that did not fulfill recruitment, median percentage difference in sample size from the expected enrollment is 9% (median absolute difference is 146 participants). Publication rates among LMICs and non-LMIC participants trials were similar.

The top 3 targeted conditions, according to the 22 IHME disease categorization, were maternal and neonatal disorders, nutritional deficiencies, and respiratory infections and tuberculosis (TB) (Table 1, Supplemental Fig 20). The top 3 targeted conditions with the highest DALYs were maternal and neonatal disorders, respiratory infections and TB, and enteric infections with respective DALYs ranging from 50 million DALYs to >100 million DALYs (Supplementary Table 5). The targeted conditions by the pediatric mega-trials are shown in Supplementary Table 6.

The 4 IHME-targeted conditions with the highest GBD (DALYs ranging from 50 million to >100 million) were the maternal and neonatal disorders (n = 49 trials; 167 268 496 DALYs), respiratory infections and TB (n = 38 trials; 83 648 680 DALYs), enteric infections (n = 14 trials; 63 732 876 DALYs), and other noncommunicable diseases, which included congenital anomalies, gynecologic diseases, urinary diseases, oral disorders, hemoglobinopathies and hemolytic anemia, and sudden infant death syndrome (n = 23 trials; 61 494 156 DALYs) (Fig 2). The 4 IHME-targeted conditions with the highest number of large trials were maternal and neonatal disorders (n = 49 trials), nutritional deficiencies (n = 40 trials), respiratory infections and TB (n = 38 trials) and neglected tropical diseases and malaria (n = 28 trials) (Fig 2). Among trials conducted in LMICs, we did not identify large trials targeting several conditions, including mental disorders (12 540 977 DALYs), neoplasms (9 414 291 DALYs), other noncommunicable diseases (61 494 156 DALYs), and self-harm and interpersonal violence (5 474 592 DALYs) as noted by the gray dots in Fig 3. Among all conditions with ≥2 million DALYs, the least studied conditions in this research agenda were the musculoskeletal disorders (1 trial) and self-harm and interpersonal violence (1 trial).

FIGURE 2

Tree map showing the relationship between the number of large pediatric RCTs for the 22 IHME disease categories and the pediatric GBD (expressed in DALYs for children aged 0–15 years). Box size corresponds to the pediatric global burden for each of the 22 IHME disease categories in terms of DALYs. Box color intensity corresponds to the number of trials. Font color corresponds to the percentage of trials completed or actively recruiting within each disease category. Not visible in this tree map is the disease category substance abuse disorder because of its low DALYs in comparison with other disease categories (substance use disorders are responsible for 71 964 DALYs).

FIGURE 2

Tree map showing the relationship between the number of large pediatric RCTs for the 22 IHME disease categories and the pediatric GBD (expressed in DALYs for children aged 0–15 years). Box size corresponds to the pediatric global burden for each of the 22 IHME disease categories in terms of DALYs. Box color intensity corresponds to the number of trials. Font color corresponds to the percentage of trials completed or actively recruiting within each disease category. Not visible in this tree map is the disease category substance abuse disorder because of its low DALYs in comparison with other disease categories (substance use disorders are responsible for 71 964 DALYs).

Close modal
FIGURE 3

Scatter plot depicting the association between pediatric GBD and number of large pediatric trials. Panel A, all trials; Panel B, trials with participants exclusively in HICs; Panel C, trials with participants exclusively in LMICs. In Panel A, all trials include LMIC, HIC, and LMIC/HIC trials. 18 trials with multiple conditions were counted for as many IHME conditions as they addressed. Noncommunicable diseases are “noninfectious and nontransmissible diseases that may be caused by genetic or behavioral factors and generally have a slow progression and long duration.”31 IHME categorizations of noncommunicable diseases include oral disorders, gynecologic disorders, urinary diseases and male infertility, congenital disorders, hemoglobinopathies and hemolytic anemias, endocrine, metabolic, blood, and immune disorders, and sudden infant death syndrome.

FIGURE 3

Scatter plot depicting the association between pediatric GBD and number of large pediatric trials. Panel A, all trials; Panel B, trials with participants exclusively in HICs; Panel C, trials with participants exclusively in LMICs. In Panel A, all trials include LMIC, HIC, and LMIC/HIC trials. 18 trials with multiple conditions were counted for as many IHME conditions as they addressed. Noncommunicable diseases are “noninfectious and nontransmissible diseases that may be caused by genetic or behavioral factors and generally have a slow progression and long duration.”31 IHME categorizations of noncommunicable diseases include oral disorders, gynecologic disorders, urinary diseases and male infertility, congenital disorders, hemoglobinopathies and hemolytic anemias, endocrine, metabolic, blood, and immune disorders, and sudden infant death syndrome.

Close modal

The number of large trials targeting each of the 22 IHME disease categories moderately correlated with the pediatric GBD (Fig 3). The Spearman’s correlation coefficient for all trials across all country settings was ρ = 0.76 (P = .00004); for HICs, ρ = 0.29 (P = .20) (96 trials and 192 642 participants); and for LMICs, ρ = 0.69 (P = .00036) (139 trials and 364 909 participants respectively).

There was a dearth of large, nonvaccine, noncluster pediatric RCTs with ≥1000 participants, representing only 0.5% (247 of 45 000) of all registered trials for participants aged 0 to 17 years in ClinicalTrials.gov. Only 17 pediatric mega-trials with ≥5000 participants exist. Approximately one-half of the large, pediatric RCTs were funded exclusively by foundations, nonprofit organizations, and public organizations, whereas only 19% of these trials were exclusively funded by the industry. The top 3 countries of investigators are the United States, the United Kingdom, and China, whereas the top 3 countries of enrolled children are the United States, Germany, and Canada. Only 15% of large, pediatric RCTs were international, multicountry investigator collaborations, and one-quarter of the trials enrolled (or planned to enroll) children in >1 country. The majority (70%) of investigators of large pediatric RCTs were exclusively from HICs, whereas only 9% of trials involved investigators from both HICs and LMICs. Only 20% of trial investigators were exclusively from LMICs. Only a minority of the large pediatric RCTs targeted mortality as a primary or secondary outcome. Although most of the identified large pediatric trials were completed at the time of our assessment, less than two-thirds of those completed >12 months at the time of our assessment were published. The total number of trials per disease category (across all trials targeting the same disease category) correlated well with the pediatric GBD overall and in LMICs, but not in HICs.

There are many challenges to conducting a clinical trial in a resource-poor environment.17,18  Obstacles identified include financial barriers (although trials may be more cost-effective in LMICs), lack of research environment, competing demands, health system maturity, and ethical and regulatory demands.17  A key solution proposed is the implementation of local investigator-initiated trials that could capitalize on local knowledge of the health care infrastructure and population. Only a minority (17%) of large trials performed exclusively in LMICs had investigators from both HICs and LMICs. Incentivization of collaborations between investigators from LMICs and HICs, particularly for trials performed exclusively in LMICs, should be a priority.

Most pediatric trials in the scientific literature are underpowered to detect differences for hard clinical outcomes, such as mortality.19  For example, a trial targeting all-cause mortality as an outcome with an assumed relative risk reduction by the targeted intervention of 20% and a 30% expected death rate in the control group would have required a sample size of 2298 to detect a statistically significant difference in mortality (α = .05; β = .10). The noninclusion of mortality outcome as a primary or secondary end point in the majority of these large, pediatric trials represents a missed unique opportunity to address survival benefit from the targeted interventions in these pivotal large pediatric trials. Lastly, the nonpublication rate of 38% for the completed large pediatric RCTs, and 35% for those completed ≥12 months, was significant. This accounts for clinical data from 135 722 children whose participation in these clinical trials was not effectively used. Participants may be exposed to risk in a trial, but there is no ultimate societal benefit if results are nonpublished. Existing regulations to ensure full-result publications must be reinforced to maximize use of funds and resources and decrease research waste.20  Our findings are similar to that of a retrospective, cross-sectional study from 2008 to 2010 analyzing pediatric RCTs in ClinicalTrials.gov, in which ∼30% of pediatric trials were not published.21 

Some study limitations should be acknowledged. Our empirical analysis captured only large pediatric RCTs registered in ClinicalTrials.gov. Although not all large pediatric RCTs were captured, nevertheless, we believe we captured a representative sample. A cross validation of trial-queries based on the ClinicalTrials.gov registry against the World Health Organization International Clinical Trial Research Platform (an international clinical trials registry platform that combines data from 14 primary registries, including the ClinicalTrial.gov,22  revealed that the ClinicalTrials.gov registry captures ∼83% of pediatric trials performed worldwide (84% of trials from HICs, 84% of trials from MICs, and 81% of trials from lower-income countries). The ClinicalTrials.gov registry was previously used by investigators in our team to assess the association between pediatric clinical trials (of any study design and any sample size) and the GBD.5  This current work expands on this previous work.

We identified a dearth of large, pediatric trials, with such trials representing only 0.5% of the total number of pediatric trials registered in ClinicalTrials.gov. However, we were not able to say whether this finding was unique for pediatrics because we did not comparatively assess the respective proportion of large adult trials.

To address clinically important questions with larger certainty around the reported estimates (narrow confidence intervals), mega-trials (trials with ≥5000 participants) may also be needed. Smaller trials can often have large residual uncertainty around the reported effect-size point estimates and often cannot exclude clinically important benefits or harms.

We excluded cluster-randomized trials from our analysis because the effective sample size in cluster RCTs is not the actual number of individual participants in the clusters.23  The responses of participants in the same cluster tend to be correlated; thus, the effective sample size is somewhere in between the number of randomized clusters and the number of participants within these clusters, depending on the average cluster size and the degree of correlation of responses of individuals within the same cluster.23,24  Although several informative pragmatic cluster RCTs are successfully performed in LMICs, nevertheless, their characteristics are not directly comparable to those of noncluster RCTs. Furthermore, we excluded phase 3 vaccine trials from our analysis because these trials have different trial logistics compared with nonvaccine therapeutic or preventive pediatric RCTs.25 

Although 25% of large trials were single country trials (for participants’ country), a substantial portion of those trials were multisite trials. For the same sample size, single site or single country trials may have advantages or disadvantages versus multisite or multicountry trials. In multisite and multicountry trials, there may be greater variation across sites or countries on the complementary interventions applied (backbone interventions). However, given the randomized study design, these interventions are expected to be similarly implemented in the experimental and control groups and, thus, unlikely to differentially affect the results for the compared interventions across multicountry and multisite study settings.

The use of DALYs may underrepresent the burden of disease, particularly in LMICs where disease surveillance is lacking,26,27  and some public health researchers suggest that the use of DALYs does not accurately capture the societal aspects of disability or felt-disability, which is predicated on the affected person’s socioeconomic capital and access to health care, among many other factors.28  However, DALYs is one of the most widely used metric for global health analyses.16,29 

In summary, incentivization for international collaborations between investigators from HICs and LMICs in the pediatric research agenda of large pediatric RCTs should be a priority. There is a strong need to build clinical research capacity in LMICs (ie, mandatory to have LMIC investigators if a study is based in such a country). Timely publication of the results of large pediatric RCTs is also of paramount importance to gain maximal benefit for the broader community from the allocated clinical and funding resources. A greater number of children could benefit if the research agenda of these large pediatric RCTs is closely aligned with the GBD in children.

FUNDING: Dr Klassen is supported by a Tier 1 Canada Research Chair in Clinical Trials. The Tier 1 Canada Research Chair in Clinical Trials is a federal government national research award available to awardees for 7 years. Dr Terry Klassen received funding since April 1, 2020.

COMPANION PAPER: A companion to this article can be found online at www.pediatrics.org/cgi/doi/10.1542/peds.2021-051589.

Ms Cho acquired the data from ClinicalTrials.gov, contributed to the study design, performed the data analyses and interpretation of data, and prepared the first manuscript draft; Dr Serghiou acquired the data from ClinicalTrials.gov, performed the data analyses, contributed to the interpretation of data, and critically reviewed the final manuscript draft; Dr Klassen conceptualized the project, contributed to the study design, data analysis, and interpretation of data, and critically reviewed the final manuscript draft; Dr Ioannidis conceptualized the project, designed the study protocol, contributed to the data analysis and interpretation of data, and critically reviewed the final manuscript draft; Dr Contopoulos-Ioannidis conceptualized the project, designed the protocol, acquired the data from ClinicalTrials.gov, performed the data analyses and interpretation of data, contributed to the preparation of the first manuscript draft, and critically reviewed and revised the final manuscript draft; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.

DALY

disability-adjusted life-year

GBD

global burden of disease

HIC

high-income country

IHME

Institute for Health Metrics and Evaluation

LMIC

lower-middle-income country

RCT

randomized controlled trial

TB

tuberculosis

1
Klassen
TP
,
Hartling
L
,
Craig
JC
,
Offringa
M
.
Children are not just small adults: the urgent need for high-quality trial evidence in children
.
PLoS Med
.
2008
;
5
(
8
):
e172
2
Suchdev
PS
,
Howard
CR
;
Section On International Child Health
.
The role of pediatricians in global health
.
Pediatrics
.
2018
;
142
(
6
):
e20182997
3
Kyu
HH
,
Pinho
C
,
Wagner
JA
, et al;
Global Burden of Disease Pediatrics Collaboration
.
Global and national burden of diseases and Injuries Among Children and Adolescents Between 1990 and 2013: Findings From the Global Burden of Disease 2013 Study
.
JAMA Pediatr
.
2016
;
170
(
3
):
267
287
4
Denno
D
.
Global child health
.
Pediatr Rev
.
2011
;
32
(
2
):
e25
e38
5
Bourgeois
FT
,
Olson
KL
,
Ioannidis
JP
,
Mandl
KD
.
Association between pediatric clinical trials and global burden of disease
.
Pediatrics
.
2014
;
133
(
1
):
78
87
6
Tse
T
,
Fain
KM
,
Zarin
DA
.
How to avoid common problems when using ClinicalTrials.gov in research: 10 issues to consider
.
BMJ
.
2018
;
361
:
k1452
7
US National Library of Medicine
.
Advanced search
.
2020
.
8
R Core Team
.
R: A Language and Environment for Statistical Computing
.
Vienna, Austria
:
R Foundation for Statistical Computing
;
2018
9
Internal Revenue Service
.
Exemption requirements – 501(c)(3) organizations
.
10
Investopedia
.
Not for profit
.
Available at: https://www.investopedia.com/terms/n/not-for-profit.asp. Accessed December 7, 2020
11
The World Bank
.
World Bank country and lending groups
.
12
Caldron
PH
,
Gavrilova
SI
,
Kropf
S
.
Why (not) go east? Comparison of findings from FDA Investigational New Drug study site inspections performed in Central and Eastern Europe with results from the USA, Western Europe, and other parts of the world
.
Drug Des Devel Ther
.
2012
;
6
:
53
60
13
Murthy
S
,
Mandl
KD
,
Bourgeois
FT
.
Industry-sponsored clinical research outside high-income countries: an empirical analysis of registered clinical trials from 2006 to 2013
.
Health Res Policy Syst
.
2015
;
13
:
28
14
Institute for Health Metrics and Evaluation
.
GBD compare
.
2020
.
Available at: https://vizhub.healthdata.org/gbd-compare/. Accessed June 12, 2020
15
Voigt
K
,
King
NB
.
Out of alignment? Limitations of the global burden of disease in assessing the allocation of global health aid
.
Public Health Ethics
.
2017
;
10
(
3
):
244
256
16
Troeger
C
,
Colombara
DV
,
Rao
PC
, et al
.
Global disability-adjusted life-year estimates of long-term health burden and undernutrition attributable to diarrhoeal diseases in children younger than 5 years
.
Lancet Glob Health
.
2018
;
6
(
3
):
e255
e269
17
Alemayehu
C
,
Mitchell
G
,
Nikles
J
.
Barriers for conducting clinical trials in developing countries- a systematic review
.
Int J Equity Health
.
2018
;
17
(
1
):
37
18
von Saint André-von Arnim
AO
,
Attebery
J
,
Kortz
TB
, et al;
Pediatric Acute Lung Injury and Sepsis Investigators (PALISI) Network
.
Challenges and priorities for pediatric critical care clinician-researchers in low- and middle-income countries
.
Front Pediatr
.
2017
;
5
:
277
19
Jakobsen
JC
,
Wetterslev
J
,
Gluud
C
.
Considerations on the strengths and limitations of using disease-related mortality as an outcome in clinical research
.
BMJ Evid Based Med
.
2019
;
26
(
3
):
127
130
20
Manzoli
L
,
Flacco
ME
,
D’Addario
M
, et al
.
Non-publication and delayed publication of randomized trials on vaccines: survey
.
BMJ
.
2014
;
348
:
g3058
21
Pica
N
,
Bourgeois
F
.
Discontinuation and nonpublication of randomized clinical trials conducted in children
.
Pediatrics
.
2016
;
138
(
3
):
e20160223
22
Hemming
K
,
Eldridge
S
,
Forbes
G
,
Weijer
C
,
Taljaard
M
.
How to design efficient cluster randomised trials
.
BMJ
.
2017
;
358
:
j3064
23
World Health Organization
.
International clinical trials registry platform
.
2021
.
Available at: https://www.who.int/clinical-trials-registry-platform. Accessed March 1, 2021
24
Brown
AW
,
Li
P
,
Bohan Brown
MM
, et al
.
Best (but oft-forgotten) practices: designing, analyzing, and reporting cluster randomized controlled trials
.
Am J Clin Nutr
.
2015
;
102
(
2
):
241
248
25
Kahn
R
,
Rid
A
,
Smith
PG
,
Eyal
N
,
Lipsitch
M
.
Choices in vaccine trial design in epidemics of emerging infections
.
PLoS Med
.
2018
;
15
(
8
):
e1002632
26
Zheleva
B
,
Atwood
JB
.
The invisible child: childhood heart disease in global health
.
Lancet
.
2017
;
389
(
10064
):
16
18
27
GBD 2016 DALYs and HALE Collaborators
.
Global, regional, and national disability-adjusted life-years (DALYs) for 333 diseases and injuries and healthy life expectancy (HALE) for 195 countries and territories, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016 [published correction appears in Lancet. 2017;390(10106):e38]
.
Lancet
.
2017
;
390
(
10100
):
1260
1344
28
Grosse
SD
,
Lollar
DJ
,
Campbell
VA
,
Chamie
M
.
Disability and disability-adjusted life years: not the same
.
Public Health Rep
.
2009
;
124
(
2
):
197
202
29
World Health Organization
.
International classification of functioning, disability and health (ICF)
.
2001
.
30
World Health Organization
.
Noncommunicable diseases
.
2015
.
Available at: www.who.int/topics/noncommunicable_diseases/en/. Accessed December 7, 2020

Competing Interests

POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.

FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.

Supplementary data