Aim This study aimed to investigate the association between Twitter exposure and the number of citations for coloproctology articles. Method Original articles from journals using Twitter between June 2015 and May 2016 were evaluated for the following characteristics: publishing journal; article subject; study design; nationality, speciality and affiliation of the author(s); and reference on Twitter. Citation data for these articles were retrieved from Google Scholar (https://scholar.google.com) in January 2018. We performed a univariate analysis using these data followed by a multivariate, logistic regression analysis to search for factors associated with a high citation level, which was defined as accrual of more than five citations. Results Out of six coloproctology journals listed on the InCites JCR database, three (Diseases of the Colon & Rectum, Colorectal Disease and Techniques in Coloproctology) used Twitter, where 200 (49.5%) out of a total of 404 articles had been featured. Citation rates of articles that featured on Twitter were significantly higher than those that did not (11.4 ± 9.2 vs 4.1 ± 3.1, P < 0.001). In multivariate analysis, Twitter exposure (OR 8.6, P = 0.001), European Union nationality (OR 2.4, P = 0.004), Colorectal Disease journal (OR 3.3, P = 0.005) and systematic review articles (OR 3.4, P = 0.009) were associated with higher citation levels. Conclusion Article exposure on Twitter was strongly associated with a high citation level. Medical communities should encourage journals as well as physicians to actively utilize social media to expedite the spread of new ideas and ultimately benefit medical society as a whole.

The accuracy of three automated haematology analysers [Sysmex XE-2100 (both optical and impedance mode), Bayer Advia 120, and Beckman Coulter LH-750] was compared with the immunoplatelet reference method for platelet measurement. A total of 165 blood specimens were obtained from patients and platelet counts were determined using the four-automated haematology analyser methods and the immunoplatelet reference method. The coefficients of determination (R(2)) between the automated haematology analyser methods and the immunoplatelet reference method for the overall platelet range were >0.98. A bias study, however, showed some disagreement. The use of a coincidence correction calculation for the immunoplatelet method did not improve the correlation between the immunoplatelet method and the automated haematology analyser methods. To estimate the possibility of inadequate platelet transfusion, the number of prophylactic platelet transfusion indications determined by the automated haematology analyser platelet counts were compared with the number of transfusion indications according to the platelet counts determined by the immunoplatelet method. An additional 48 blood specimens were included in this analysis. All of the automated haematology analysers showed some disagreement in the transfusion indications when compared with the immunoplatelet method, suggesting the possibility of inadequate platelet transfusion.