Brænd Anja Maria, Straand Jørund, Klovning Atle
Department of General Practice, Institute of Health and Society, Faculty of Medicine, University of Oslo, Postbox 1130 Blindern, N-0318, Oslo, Norway.
BMC Fam Pract. 2017 Dec 29;18(1):113. doi: 10.1186/s12875-017-0680-7.
When reading a report of a clinical trial, it should be possible to judge whether the results are relevant for your patients. Issues affecting the external validity or generalizability of a trial should therefore be reported. Our aim was to determine whether articles with published results from a complete cohort of drug trials conducted entirely or partly in general practice reported sufficient information about the trials to consider the external validity.
A cohort of 196 drug trials in Norwegian general practice was previously identified from the Norwegian Medicines Agency archive with year of application for approval 1998-2007. After comprehensive literature searches, 134 journal articles reporting results published from 2000 to 2015 were identified. In these articles, we considered the reporting of the following issues relevant for external validity: reporting of the clinical setting; selection of patients before inclusion in a trial; reporting of patients' co-morbidity, co-medication or ethnicity; choice of primary outcome; and reporting of adverse events.
Of these 134 articles, only 30 (22%) reported the clinical setting of the trial. The number of patients screened before enrolment was reported in 61 articles (46%). The primary outcome of the trial was a surrogate outcome for 60 trials (45%), a clinical outcome for 39 (29%) and a patient-reported outcome for 25 (19%). Clinical details of adverse events were reported in 124 (93%) articles. Co-morbidity of included participants was reported in 54 trials (40%), co-medication in 27 (20%) and race/ethnicity in 78 (58%).
The clinical setting of the trials, the selection of patients before enrolment, and co-morbidity or co-medication of participants was most commonly not reported, limiting the possibility to consider the generalizability of a trial. It may therefore be difficult for readers to judge whether drug trial results are applicable to clinical decision-making in general practice or when developing clinical guidelines.
阅读一项临床试验报告时，应该能够判断其结果是否与你的患者相关。因此，影响试验外部有效性或可推广性的问题应该被报告。我们的目的是确定那些全部或部分在全科医疗中进行的完整药物试验队列的已发表结果的文章，是否报告了足够的试验信息以考量其外部有效性。
先前从挪威药品管理局档案中识别出一组196项挪威全科医疗中的药物试验，其批准申请年份为1998 - 2007年。经过全面的文献检索，确定了134篇报告2000年至2015年发表结果的期刊文章。在这些文章中，我们考量了以下与外部有效性相关问题的报告情况：临床背景的报告；纳入试验前患者的选择；患者合并症、联合用药或种族的报告；主要结局的选择；以及不良事件的报告。
在这134篇文章中，只有30篇（22%）报告了试验的临床背景。61篇文章（46%）报告了入组前筛查患者数量。60项试验（45%）的主要结局为替代结局，39项（29%）为临床结局，25项（19%）为患者报告结局。124篇文章（93%）报告了不良事件的临床细节。54项试验（40%）报告了纳入参与者的合并症，27项（20%）报告了联合用药情况，78项（58%）报告了种族/民族情况。
试验的临床背景、纳入试验前患者的选择以及参与者的合并症或联合用药情况最常未被报告，这限制了考量试验可推广性的可能性。因此，读者可能难以判断药物试验结果是否适用于全科医疗中的临床决策或制定临床指南时的情况。
