loading page

Value of clinical research: Usefulness tool development and systematic review of 350 randomised controlled trials in preterm birth
  • +6
  • Janneke van t Hooft,
  • Charlotte van Dijk,
  • Cathrine Axfors,
  • Noah Haber,
  • Zarko Alfirevic,
  • Martijn Oudijk,
  • Ben Mol,
  • Patrick Bossuyt,
  • John P. A. Ioannidis
Janneke van t Hooft
Amsterdam UMC Locatie AMC
Author Profile
Charlotte van Dijk
Amsterdam UMC Locatie AMC
Author Profile
Cathrine Axfors
Stanford University
Author Profile
Noah Haber
Stanford University
Author Profile
Zarko Alfirevic
University of Liverpool
Author Profile
Martijn Oudijk
Amsterdam UMC Locatie AMC
Author Profile
Ben Mol
Monash University Central Medical School
Author Profile
Patrick Bossuyt
Amsterdam UMC Locatie AMC
Author Profile
John P. A. Ioannidis
Stanford University
Author Profile

Abstract

Objective: We developed a research usefulness tool collating published criteria and examined if randomised controlled trials (RCTs) addressing preterm birth were useful. Search Strategy: Cochrane library. Selection Criteria: Published RCTs within 56 preterm birth Cochrane reviews. Data Collection and Analysis: A usefulness tool was developed with eight criteria combining 13 items identified through literature searches and consensus. RCTs were evaluated for compliance with each item by multiple assessors (reviewer agreement 95-98%). Proportions with 95% confidence interval (CI) were calculated and compared for change over time using ≧ 2010 as a cut-off, with relative risks (RR). Main Results: Among 350 selected RCTs, only 38 (11%, 95% CI 8-15%) met half of the usefulness criteria. Compared to trials before 2010, recent trials used composite or surrogate (less informative) outcomes more often (13% vs 25%, RR 1.87, 95% CI 1.19-2.93). Only 17 trials reflected real life (pragmatism) in design (5%, 95% CI 3-8%), with no improvements over time. No trials reported involvement of mothers to reflect patients’ top priorities in question definition or outcomes selection. Recent trials were more transparent with prospective registration (0.5% vs 28%, RR 58, 95% CI 8-420%), availability of protocol (0.5% vs 15%, RR 32, 95% CI 4-237%) and data sharing statements (2% vs 8%, RR 3, 95% CI 1-10%). Conclusion: Clinical trials in preterm birth lacked many usefulness features, with one tenth of trials meeting half of the items evaluated. Use of informative outcomes, patient centeredness, pragmatism and transparency should be key targets for future research planning.