Μελέτες διασταυρούμενης μετάβασης (cross over studies) (Doctoral thesis)

Λαθύρης, Δημήτριος Ν.

Randomized parallel arm design trials constitute the majority of the trials analyzed in meta-analyses. Το a lesser extent randomized crossover trials are also used. Cross over trials don’t have an accepted and well-established method for their meta-analysis. We tried to evaluate empirically how cross over trial results were analyzed in meta-analyses and if their results agreed with those of parallel arm trials in the same question. We studied 26 systematic reviews with 28 meta-analyses. Each meta-analysis had at least one cross over study. The initial sample had 334 systematic reviews retrieved by Cochrane Library Issue 2. 2003. 12 out of the 28 meta-analyses did not mention at all their approach towards cross over trials’ results, 9 used only the first period results, 3 combined results from the first and second period, 1 used only data from the second period and 3 did not have consistent approach for the analysis of cross over data. The 28 meta-analyses had 137 crossover trials with 7162 patients and 132 parallel arm trials with 11398 patients. Effect sizes correlated with the two types of design (ρ =0.72). The summary relative odd ratio for parallel arms versus cross over studies for favorable outcomes was 0.87 (95% CI, 0.74-1.02). Cross over designs contribute evidence in a fifth of systematic reviews, but few meta-analyses made use of their data. The results of cross over trials tended to agree with those of parallel arms trials, although there was a trend for more conservative treatment effect estimates in parallel arm trials. Cross over trials are not as popular as parallel arm trials. Nevertheless, one of their major indications is the trials of clinical pharmacology. Pharmaceutical industries sponsor the majority of randomized clinical trials. One can easily question if pharmaceutical industries can influence the choice of comparators of the trials that are sponsoring.85 We tried to assess how often major pharmaceutical companies sponsor trials that evaluate their products and how often they sponsor trials where their products are directly compared to licensed products of other major companies. We used randomized trials and we focused on cross over trials specifically. We studied 577 randomized trials that were sponsored by 15 major companies and were registered on ClinicalTrials.gov on 2006. 82% of them had a single industry sponsor. The compared intervention belonged to a single company in 67% of the trials. All 15 companies preferred to be unique industry sponsor and unique owner of the evaluated interventions. Co-sponsoring usually reflected co-ownership or common financial interests. Head to head comparisons of different active interventions that belonged to different companies were found only in 18 studies with more than two industry sponsors. Our findings did not change when we focused on cross- over trials. Cross over trials were 12.6% (73/577) of the randomized trials sample and were sponsored by 12 companies. 90% of cross over trials was sponsored by only one pharmaceutical industry. The interventions belonged to a single company in 86% of these trials. More than two sponsors were observed in 8 trials, but head to head comparisons among intervention owned by the sponsoring companies were not found. We can easily conclude that head to head comparisons were rarely observed in trials sponsored by pharmaceutical industries and were not observed at all in cross over trials. It seems that each company has a clinical research agenda not only focused on its own products but also aiming to avoid comparisons with products of other companies. This diminishes the ability to understand the merits of different interventions for the same condition.
Alternative title / Subtitle: Εμπειρική μετααναλυτική προσέγγιση
Institution and School/Department of submitter: Πανεπιστήμιο Ιωαννίνων Σχολή Ιατρικής Τμήμα Ιατρικής Τομέας Κοινωνικής Ιατρικής και Ψυχικής Υγείας Εργαστήριο Υγιεινής και Επιδημιολογίας
Keywords: Μετά - ανάλυση,Παράλληλες ομάδες,Μελέτες,Εμπειρική αξιολόγηση,Χορηγία τυχαιοποιημένων μελετών,Συγκρινόμενες παρεμβάσεις
URI: https://olympias.lib.uoi.gr/jspui/handle/123456789/756
http://dx.doi.org/10.26268/heal.uoi.678
Link: http://thesis.ekt.gr/thesisBookReader/id/18596#page/1/mode/2u
Appears in Collections:Διδακτορικές Διατριβές

Files in This Item:
There are no files associated with this item.


 Please use this identifier to cite or link to this item:
https://olympias.lib.uoi.gr/jspui/handle/123456789/756
  This item is a favorite for 0 people.

Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.