Arch Pediatr. 2021 May 28. pii: S0929-693X(21)00094-4. [Epub ahead of print]
INTRODUCTION: At the end of April 2020, three European pediatric societies published an alert on a new hyperinflammatory disorder linked to SARS-CoV-2. This disease has alternatively been called Kawasaki-like disease, pediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 infection (PIMS-TS), and multisystem inflammatory syndrome in children (MIS-C). These alerts provide a clear starting point from which to study the early response of the medical and scientific community to a new disease in terms of scientific publications, and to compare the timeline of this response with levels of general public interest. To this aim, we conducted a bibliometric analysis of articles on this disease published between 1 April and 5 July 2020.METHOD: A literature search was performed using PubMed and in three preprint repositories. For each article, the name used for the disease in the title, the number of authors, the number of patients, the citations according to Google Scholar, the journal impact factor, and the Altmetric score were retrieved. Google search trends for the terms "Kawasaki" and "COVID," "COVID-19," and "coronavirus" were also retrieved, as was the number of Reuters news articles published on the topic. These data were compared longitudinally on a weekly basis. The quality of the reporting of the study was evaluated using the STROBE guidelines for observational studies with more than three patients and using the CARE guidelines for case reports of three or fewer patients.
RESULTS: Eighty-six articles were included, among which ten were preprints (three of which were subsequently published) and 49 were clinical articles (57%). A total of 857 patients were described. The median number of authors per article was five (range, 1-45), the median number of patients was four (1-186), the median number of citations was one (0-170), the median Altmetric score was 12 (0-7242), and the median journal impact factor was 3.7 (1-74.7). For the clinical articles, the median percentage of STROBE or CARE checklist items satisfied was 70% (IQR, 56.75-79.25; range, 40-90). Guideline adherence was significantly higher for observational studies than for case reports (median percentage of checklist items satisfied, 78.5% vs 61.5%; P<0.001); however, guideline adherence did not differ significantly between peer-reviewed and preprint articles (median percentage of checklist items satisfied, 57% vs 72%; P=0.205). The only statistically significant difference between clinical articles and other types of articles was the number of authors (median, 7 vs 2; P=2.53E-9). Fifty-seven of the 86 articles were authored by researchers from just three countries (the USA, 31; France, 14; and the UK, 12). The names most frequently used in the title were Kawasaki-like disease (n=37), followed by MIS-C (n=27), PIM-TS (n=14), and other names involving the term "inflammatory" (n=12). Google searches for related terms peaked between weeks 18 and 21, following the initial alerts and decreased rapidly thereafter. The number of Reuters articles on the subject was correlated with Google search trends (ρ: 0.86, 95% CI [0.59; 0.96]; P=0.00016), but the number of medical articles published was not (ρ: -0.54, 95% CI [-0.87; 0.14]; P=0.11). The first small case series was published less than 2 weeks after the initial alert; however, if all articles had been deposited as preprints when they were submitted to journals, the cumulative number of reported cases would have been 300% higher in week 18 (3 vs 1), 400% higher in week 19 (44 vs 11), 70% higher in week 20 (124 vs 73), and 54% higher in week 21 (129 vs 84).
CONCLUSION: In a period of 9 weeks after the initial alerts from European pediatric societies, 85 medical articles were published, involving 856 patients (one case report was published before the alerts), allowing rapid dissemination of research information. However, general public interest followed the news cycle rather than scientific releases. The quality of the reporting, as assessed by adherence to STROBE or CARE guidelines, was adequate with more than two-thirds of checklist items satisfied. Learned societies play an important role in the early dissemination of up-to-date peer-reviewed information. Preprint deposition should be encouraged to accelerate the dissemination of research information.
Keywords: Bibliometric; COVID-19; Kawasaki-like; Multisystem inflammatory syndrome in children (MIS-C); Pediatric inflammatory multisystem syndrome temporally associated with SARS-CoV-2 infection (PIMS-TS); Preprint; SARS-CoV-2