2 e-Letters

  • External valdiation of QFracture-2012

    As authors of the QFracture papers1-3, we read this article by Livingstone et al with interest. They stated that the had externally validated the QFracture-2016 algorithm using CPRD4. The authors report that whilst there was very good to excellent discrimination, calibration was poor. The authors attributed an apparent under-prediction to their outcome definition using the CPRD validation dataset since this included GP data linked to hospital data. However, we think this under-prediction is due to the authors using the wrong algorithm – the authors have confirmed that they had used a previous version (QFracture-2012) which is based on unlinked data. The QFracture-2016 algorithm is the version which is currently recommended and used in the NHS and is derived from the QResearch database including GP data linked to hospital and mortality data3. Therefore, the authors need to correct their paper and update their conclusions accordingly. We would also like to highlight that the code groups for QFracture are available here https://www.qresearch.org/data/qcode-group-library/
    1. Hippisley-Cox J, Coupland C. Predicting risk of osteoporotic fracture in men and women in England and Wales: Prospective derivation and validation of QFractureScores. BMJ (Online) 2009;339(7733):1291-95. doi: 10.1136/bmj.b4229
    2. Hippisley-Cox J, Coupland C. Derivation and validation of updated QFracture algorith...

    Show More
  • Are not umbrealla reviews the same as overviews of reviews?

    We read with great interest the article on ‘Conducting umbrella reviews’. In our view, it is important to highlight that what is described as ‘umbrella reviews’ in this article is often referred to as ‘overviews (of reviews)’, and shares similar definitions and goals, namely to synthesise evidence at the systematic review-level. The term ‘overviews of reviews’ is employed by Cochrane, a leading international organization for evidence synthesis. The corresponding chapter in the Cochrane handbook (1) was revised a few years ago and provides a summary of methods research for this type of evidence synthesis along with recommendations for conducting overviews of reviews. Although this has been prepared for Cochrane, we think that the vast majority of the content can also be used outside Cochrane and for a range of research questions. It provides guidance for topics not mentioned in your article such as dealing with overlapping primary studies across reviews on the same topic, decision tools supporting the inclusion of reviews, and updating reviews by conducting supplemental searches for primary studies. Most importantly, in the article by Belbasis et al. there is no explicit mention of assessing the quality or risk of bias of the included reviews and there is no mention of the recently published reporting guideline for overviews of reviews of healthcare interventions (2).

    A large body of research by many evidence synthesis groups over the past 10+ years exists to advan...

    Show More