الفهرس | Only 14 pages are availabe for public view |
Abstract Purpose: To identify frequent writing flaws of the internet-published articles in Journal of the Egyptian National Cancer Institute (JENCI), Cairo University during the period from 2011 to 2016 to help improve the quality of scientific research. Material and Methods: This is a retrospective documentary study on the articles published (2011-2016) in JENCI. The articles were reviewed regarding title, abstract, introduction, methods; study settings, design, scope, sample size estimation, statistical methodology, conflict of interest, and presentation of the results, discussion, conclusions and recommendations. The articles were grouped according to the year of publication into 2 time intervals; (2011-2013), and (2014-2016) to be compared to identify any possible changes in the study characteristics over time. Results: The study included 195 articles; 139 original articles, 13 narrative reviews, 32 case reports, 11 editorial/corrigendum and no systematic reviews/meta-analyses. Most articles (81.3%) were a single-center work. Half of titles represented the aim and 9.4% represented the study design. Abstracts were mostly concise with structured writing format only 35.3% were clear. Most introductions had study aim/objectives but only 41% had a rationale. Regarding PICOT reporting, all articles reported their populations, interventions, comparator(s), while 88.5% and 89.9% applicable articles reported their outcome(s) and time period, respectively. Inclusion and exclusion criteria were clearly reported in 43.1% and 40.1% of applicable articles, respectively. Study scope was prognostic in 31.7%, therapeutic in 24.5%, and diagnostic in 21.6%. Statistical methods were mentioned in 80.6% of the articles but were complete in 30.4% and appropriate in 85.7%. Only 4 studies mentioned sample size/power estimation. Only 52.5% and 58.3% of results were exhaustive and answer the research question respectively. Misused statistical words were observed in 16.5% and incorrect statistical calculations in 41.0%. Tabular and graphical data display was independently informative in 53.1% and 41.5% with accurate title and labels in only 18.8% and 19.5%, respectively. Inappropriate selection of statistical tests/descriptive parameters was observed in 26.6% while inappropriate application of tests was found in 49.1% of articles |