BACKGROUND: Clinical trial registries can be used to monitor the production of trial evidence and signal when systematic reviews become out of date. However, this use has been limited to date due to the extensive manual review required to search for and screen relevant trial registrations. Our aim was to evaluate a new method that could partially automate the identification of trial registrations that may be relevant for systematic review updates. MATERIALS AND METHODS: We identified 179 systematic reviews of drug interventions for type 2 diabetes, which included 537 clinical trials that had registrations in ClinicalTrials.gov. Text from the trial registrations were used as features directly, or transformed using Latent Dirichlet Allocation (LDA) or Principal Component Analysis (PCA). We tested a novel matrix factorisation approach that uses a shared latent space to learn how to rank relevant trial registrations for each systematic review, comparing the performance to document similarity to rank relevant trial registrations. The two approaches were tested on a holdout set of the newest trials from the set of type 2 diabetes systematic reviews and an unseen set of 141 clinical trial registrations from 17 updated systematic reviews published in the Cochrane Database of Systematic Reviews. The performance was measured by the number of relevant registrations found after examining 100 candidates (recall@100) and the median rank of relevant registrations in the ranked candidate lists. RESULTS: The matrix factorisation approach outperformed the document similarity approach with a median rank of 59 (of 128,392 candidate registrations in ClinicalTrials.gov) and recall@100 of 60.9% using LDA feature representation, compared to a median rank of 138 and recall@100 of 42.8% in the document similarity baseline. In the second set of systematic reviews and their updates, the highest performing approach used document similarity and gave a median rank of 67 (recall@100 of 62.9%). CONCLUSIONS: A shared latent space matrix factorisation method was useful for ranking trial registrations to reduce the manual workload associated with finding relevant trials for systematic review updates. The results suggest that the approach could be used as part of a semi-automated pipeline for monitoring potentially new evidence for inclusion in a review update.

OBJECTIVES: Trial registries can be used to measure reporting biases and support systematic reviews, but 45% of registrations do not provide a link to the article reporting on the trial. We evaluated the use of document similarity methods to identify unreported links between ClinicalTrials.gov and PubMed. STUDY DESIGN AND SETTING: We extracted terms and concepts from a data set of 72,469 ClinicalTrials.gov registrations and 276,307 PubMed articles and tested methods for ranking articles across 16,005 reported links and 90 manually identified unreported links. Performance was measured by the median rank of matching articles and the proportion of unreported links that could be found by screening ranked candidate articles in order. RESULTS: The best-performing concept-based representation produced a median rank of 3 (interquartile range [IQR] 1-21) for reported links and 3 (IQR 1-19) for the manually identified unreported links, and term-based representations produced a median rank of 2 (1-20) for reported links and 2 (IQR 1-12) in unreported links. The matching article was ranked first for 40% of registrations, and screening 50 candidate articles per registration identified 86% of the unreported links. CONCLUSION: Leveraging the growth in the corpus of reported links between ClinicalTrials.gov and PubMed, we found that document similarity methods can assist in the identification of unreported links between trial registrations and corresponding articles.

PURPOSE: Greater than 50% of drugs lack pediatric labeling information, resulting in widespread "off-label" use in children. To increase pediatric prescribing information, the Pediatric Research Equity Act (PREA) was passed in 2003, requiring new drug applications to include pediatric assessments. We evaluated the study of new drugs in children since PREA was implemented. METHODS: We performed a cross-sectional analysis of new drug applications submitted to the FDA from December 2003 to July 2012, using publicly available documents at Drugs@FDA. We calculated the proportion of new drugs that included a pediatric assessment at the time of approval and at a final review performed in July 2016. RESULTS: We identified 92 new drugs requiring pediatric assessments. Only 20 (21.7%) had been fully studied in children at the time of approval, while 9 (9.8%) had been partially assessed, and 63 (68.5%) did not have pediatric data. Among drugs approved without pediatric assessments, 4.2% met FDA deferral deadlines and 34.9% eventually submitted pediatric data. At the time of our final review, allowing for a mean of 8.6 years since drug approval, 57.6% of new drugs had not been fully assessed in children. The mean time between approval in adults and availability of pediatric data for drugs approved without pediatric assessments was 6.5 years. CONCLUSIONS: These results represent a comprehensive evaluation of the study of new drugs in children and demonstrate that many drugs continue to be approved without pediatric data. Our findings serve to inform approaches to strengthen adherence to PREA requirements.

OBJECTIVES: To explore the variation in diagnostic testing and management for males diagnosed with three testicular conditions (testicular torsion, appendix testis torsion, epididymitis/orchitis) using a large pediatric health care database. Diagnostic testing is frequently used in evaluation of the acute scrotum; however, there is likely variability in the use of these tests in the emergency department setting. METHODS: We conducted a cross-sectional study of males with the diagnoses of testicular torsion, appendix testis torsion, and epididymitis/orchitis. We identified emergency department patients in the Pediatric Health Information Systems (PHIS) database from 2010 to 2015 using diagnostic and procedure codes from the International Classification of Diseases Codes 9 and 10. Frequencies of diagnoses by demographic characteristics and of procedures and diagnostic testing (ultrasound, urinalysis, urine culture and sexually transmitted infection testing) by age group were calculated. We analyzed testing trends over time. RESULTS: We identified 17,000 males with the diagnoses of testicular torsion (21.7%), appendix testis torsion (17.9%), and epididymitis/orchitis (60.3%) from 2010 to 2015. There was substantial variation among hospitals in all categories of testing for each of the diagnoses. Overall, ultrasound utilization ranged from 33.1-100% and urinalysis testing ranged from 17.0-84.9% for all conditions. Only urine culture testing decreased over time for all three diagnoses (40.6% in 2010 to 31.5 in 2015). CONCLUSIONS: There was wide variation in the use of diagnostic testing across pediatric hospitals for males with common testicular conditions. Development of evaluation guidelines for the acute scrotum could decrease variation in testing.

BACKGROUND: Few medicines have been approved for children, leading to rates of off-label prescribing reported to be as high as 90%. In 2007, the European Union adopted the Paediatric Regulation, which mandates that pharmaceutical companies conduct paediatric studies for all new medicines, unless granted a waiver. We aimed to evaluate the availability of paediatric trial results from studies required under the Paediatric Regulation for new medicines authorised in the EU. METHODS AND FINDINGS: The European Medicines Agency (EMA) public database of paediatric investigation plans was searched for new medicines centrally authorised in the EU between 1 January 2010 and 31 December 2014 with at least 1 required paediatric study. For our study cohort of paediatric clinical trials required for these medicines, we used internal EMA databases and publicly available trial registries to determine changes to the planned completion date or study design, rates of trial completion, time to trial completion, and results reporting (peer-reviewed publication or posting on trial registry). Cox proportional hazards regression models were constructed to examine factors associated with study completion. A total of 326 paediatric clinical trials were required for 122 novel medicines authorised by the EMA between 2010 and 2014. In all, 76% (247/326) of paediatric studies were not planned to be completed until after the initial marketing authorisation. The planned completion dates for 50% (162/326) were further postponed by a median of 2.2 years. Overall, 38% (124/326) of paediatric studies were completed as of 30 November 2017. The rate of trial completion for paediatric studies planned to be completed after initial marketing authorisation was 23% (56/247), versus 86% (68/79) for trials planned to be completed before authorisation (adjusted hazard ratio 0.11; 95% CI 0.06-0.19). Among completed studies, the results were reported in a public registry or in the peer-reviewed literature for 85% (105/124) at a median of 1.1 years after study completion, and 60% (74/124) were published in a peer-reviewed journal. Limitations of this study include the potential lack of generalisability to medicines not authorised by the EMA and the possibility for more of these trials to be completed or published in the future. CONCLUSIONS: The completion of many paediatric studies required under the Paediatric Regulation has been delayed. Paediatric studies planned to be completed after marketing authorisation were associated with a lower likelihood of eventual completion, highlighting the need to examine the implementation of current policies in ensuring timely availability of important paediatric information.

OBJECTIVES: Describe the trends in pediatric sedation use over time and determine variation in use of procedural sedation across children's hospital emergency departments (EDs). METHODS: We analyzed ED data from 35 hospitals within the Pediatric Health Information System for patients <19 years old who received sedation medications and were discharged from 2009 to 2014. Patients with chronic comorbidities or undergoing intubation were excluded. We determined frequency and trends in use of sedation and compared these between EDs. Descriptive statistics with appropriate weighting were used. RESULTS: Of the 1 448 011 patients potentially requiring sedation who presented to the ED, 99 951 (7.9%) underwent procedural sedation. Medication usage in 2014 included ketamine (73.7%), fentanyl and midazolam (15.9%), ketofol (7.3%), and propofol (2.7%). Use of fentanyl and midazolam increased, whereas use of ketamine, pentobarbital, etomidate, chloral hydrate, and methohexital decreased over time. Significant variation exists in the use of sedation across hospitals; in 2014, the sedation rate ranged 0.2% to 32.0%, with a median of 8.0%. The diagnosis with the largest variation in procedural sedation use was dislocation, with sedation rates ranging from 2% to 35%. CONCLUSIONS: There is significant variability across pediatric EDs in the use of procedural sedation, suggesting sedations may be performed too often or too little in some hospitals.

OBJECTIVES: To measure exclusion of elderly adults from randomized trials studying drug interventions for ischemic heart disease (IHD) and describe the characteristics of these trials. DESIGN: Cross-sectional analysis. SETTING: Interventional clinical trials studying a drug intervention for IHD that started in 2006 and after were identified in ClinicalTrials.gov. Data were extracted on study features, including age-based inclusion criteria. Data on participants and their age distribution were collected from trial publications, investigator inquiry, and result data in ClinicalTrials.gov. PARTICIPANTS: Individuals aged 65 and older. MEASUREMENTS: Proportion of trials excluding individuals based on age, mean age of trial participants, and proportion of enrolled participants aged 65 and older and 75 and older. RESULTS: Of 839 identified trials, 446 (53%) explicitly excluded elderly adults. The most-frequent upper age limits were 80 (n = 164) and 75 (n = 114), with a median upper age limit of 80 (interquartile range 75-80). Trials with upper age limit exclusions tended to be smaller (median number of participants 100 vs 201, P < .001) and were more likely to be funded primarily by nonindustry sources (78.3% vs 70.0%, P = .006). The overall mean age of trial participants was 62.7 (mean maximum age 74). The estimated proportion of participants aged 65 and older was 42.5% and the estimated proportion aged 75 and older was 12.3%. CONCLUSION: Despite the high burden of IHD in elderly adults, the majority of drug trials do not enroll participants reflective of age-related prevalence of the disease.

OBJECTIVES: Trial registries can be used to measure reporting biases and support systematic reviews but 45% of registrations do not provide a link to the article reporting on the trial. We evaluated the use of document similarity methods to identify unreported links between ClinicalTrials.gov and PubMed. STUDY DESIGN AND SETTING: We extracted terms and concepts from a dataset of 72,469 ClinicalTrials.gov registrations and 276,307 PubMed articles, and tested methods for ranking articles across 16,005 reported links and 90 manually-identified unreported links. Performance was measured by the median rank of matching articles, and the proportion of unreported links that could be found by screening ranked candidate articles in order. RESULTS: The best performing concept-based representation produced a median rank of 3 (IQR 1-21) for reported links and 3 (IQR 1-19) for the manually-identified unreported links, and term-based representations produced a median rank of 2 (1-20) for reported links and 2 (IQR 1-12) in unreported links. The matching article was ranked first for 40% of registrations, and screening 50 candidate articles per registration identified 86% of the unreported links. CONCLUSIONS: Leveraging the growth in the corpus of reported links between ClinicalTrials.gov and PubMed, we found that document similarity methods can assist in the identification of unreported links between trial registrations and corresponding articles.