ノート
Trends in Study Design Assessment for Anti-Influenza Agents
2019 年 45 巻 8 号 p. 451-459
詳細
2019 年 45 巻 8 号 p. 451-459
Utilization of a clinical article requires consideration of the quality of the study. Greater numbers of cases make predicted values closer to true values. To conduct meta-analysis and to create guidelines, a sample size must be taken into account to identify the timing of analysis, as well as being aware of language bias. We investigated the quality and sample size of clinical articles over time to examine the timing of analysis and language bias.
Using PubMed and ICHUSHI Web, we searched for clinical articles on anti-influenza agents, followed by a study on the changes in quality and sample size in the selected articles, sorted by language. We assessed the quality as it related to the Jadad scale and the processing of the analysis sets.
Our assessment of the articles revealed that the quality of the English articles was higher than that of the Japanese articles. For the English articles, observation of the changes in quality and sample size confirmed a positive correlation in Phase I and a negative correlation in Phase II, whereas for the Japanese articles, a negative correlation was confirmed for both phases.
We found a linguistic bias between English and Japanese articles, with the temporal changes in the articles different for such languages. When analyzing clinical studies for treatment plans or meta-analysis, the internal validity should be assessed in light of the results described above. Should the bias not be completely eliminated, the reason should be clearly stated under the Limitations section before conducting the assessment.
