Are Neonatal Trials Better Conducted and Reported over the Last 6 Decades? An Analysis on Their Risk-of-Bias Status in Cochrane Reviews

Nai Ming Lai, Joni M.J. Ong, Kee Hsin Chen, Nathorn Chaiyakunapruk, Colleen Ovelman, Roger Soll

Research output: Contribution to journalReview article

3 Citations (Scopus)

Abstract

Background: The introduction of Neonatology as a subspecialty in 1960 has stimulated an enormous amount of neonatal research. A large proportion of neonatal randomizedcontrolled trials (RCTs) have been included in the Cochrane reviews, within which methodological quality or risk-of-bias (ROB) assessment is an integral feature. Objectives: We described the ROB profile of neonatal RCTs published since the 1950s. Methods: We analyzed individual studies within the Cochrane Neonatal reviews published up to December 2016. We extracted the reviewers' judgments on the ROB domains including random sequence generation, allocation concealment, blinding, incomplete outcome data, and selective reporting. We evaluated blinding of personnel in trials in which blinding was considered feasible. Results: We assessed 1980 RCTs published between 1952 and 2016 from 294 Cochrane Neonatal systematic reviews, with full ROB assessments performed in 848 trials (42.8%). Among the ROB domains, the highest proportion of trials (73%) were judged as satisfactory ("low risk") in handling incomplete outcome data, while fewest trials achieved blinding of outcome assessor (38.4%). In the last 6 decades, a progressive increase has been observed in the proportion of trials that were rated as low risk in random sequence generation, allocation concealment, and selective reporting. However, blinding was achieved in less than half of the trials with no clear improvement across decades (23-44% since the 1980s). Conclusions: Despite steady improvement in the overall quality of neonatal RCTs over the last 6 decades, blinding remained unsatisfactory in the majority of the trials.

Original languageEnglish
Pages (from-to)123-131
Number of pages9
JournalNeonatology
Volume116
Issue number2
DOIs
Publication statusPublished - Aug 1 2019

Fingerprint

Neonatology
Research Design
Research

Keywords

  • Neonatology
  • Randomized controlled trials
  • Risk of bias

ASJC Scopus subject areas

  • Pediatrics, Perinatology, and Child Health
  • Developmental Biology

Cite this

Are Neonatal Trials Better Conducted and Reported over the Last 6 Decades? An Analysis on Their Risk-of-Bias Status in Cochrane Reviews. / Lai, Nai Ming; Ong, Joni M.J.; Chen, Kee Hsin; Chaiyakunapruk, Nathorn; Ovelman, Colleen; Soll, Roger.

In: Neonatology, Vol. 116, No. 2, 01.08.2019, p. 123-131.

Research output: Contribution to journalReview article

Lai, Nai Ming ; Ong, Joni M.J. ; Chen, Kee Hsin ; Chaiyakunapruk, Nathorn ; Ovelman, Colleen ; Soll, Roger. / Are Neonatal Trials Better Conducted and Reported over the Last 6 Decades? An Analysis on Their Risk-of-Bias Status in Cochrane Reviews. In: Neonatology. 2019 ; Vol. 116, No. 2. pp. 123-131.
@article{f563ae2b5b7e4e669b383d58b1f038f4,
title = "Are Neonatal Trials Better Conducted and Reported over the Last 6 Decades? An Analysis on Their Risk-of-Bias Status in Cochrane Reviews",
abstract = "Background: The introduction of Neonatology as a subspecialty in 1960 has stimulated an enormous amount of neonatal research. A large proportion of neonatal randomizedcontrolled trials (RCTs) have been included in the Cochrane reviews, within which methodological quality or risk-of-bias (ROB) assessment is an integral feature. Objectives: We described the ROB profile of neonatal RCTs published since the 1950s. Methods: We analyzed individual studies within the Cochrane Neonatal reviews published up to December 2016. We extracted the reviewers' judgments on the ROB domains including random sequence generation, allocation concealment, blinding, incomplete outcome data, and selective reporting. We evaluated blinding of personnel in trials in which blinding was considered feasible. Results: We assessed 1980 RCTs published between 1952 and 2016 from 294 Cochrane Neonatal systematic reviews, with full ROB assessments performed in 848 trials (42.8{\%}). Among the ROB domains, the highest proportion of trials (73{\%}) were judged as satisfactory ({"}low risk{"}) in handling incomplete outcome data, while fewest trials achieved blinding of outcome assessor (38.4{\%}). In the last 6 decades, a progressive increase has been observed in the proportion of trials that were rated as low risk in random sequence generation, allocation concealment, and selective reporting. However, blinding was achieved in less than half of the trials with no clear improvement across decades (23-44{\%} since the 1980s). Conclusions: Despite steady improvement in the overall quality of neonatal RCTs over the last 6 decades, blinding remained unsatisfactory in the majority of the trials.",
keywords = "Neonatology, Randomized controlled trials, Risk of bias",
author = "Lai, {Nai Ming} and Ong, {Joni M.J.} and Chen, {Kee Hsin} and Nathorn Chaiyakunapruk and Colleen Ovelman and Roger Soll",
note = "{\circledC} 2019 S. Karger AG, Basel.",
year = "2019",
month = "8",
day = "1",
doi = "10.1159/000497423",
language = "English",
volume = "116",
pages = "123--131",
journal = "Neonatology",
issn = "1661-7800",
publisher = "S. Karger AG",
number = "2",

}

TY - JOUR

T1 - Are Neonatal Trials Better Conducted and Reported over the Last 6 Decades? An Analysis on Their Risk-of-Bias Status in Cochrane Reviews

AU - Lai, Nai Ming

AU - Ong, Joni M.J.

AU - Chen, Kee Hsin

AU - Chaiyakunapruk, Nathorn

AU - Ovelman, Colleen

AU - Soll, Roger

N1 - © 2019 S. Karger AG, Basel.

PY - 2019/8/1

Y1 - 2019/8/1

N2 - Background: The introduction of Neonatology as a subspecialty in 1960 has stimulated an enormous amount of neonatal research. A large proportion of neonatal randomizedcontrolled trials (RCTs) have been included in the Cochrane reviews, within which methodological quality or risk-of-bias (ROB) assessment is an integral feature. Objectives: We described the ROB profile of neonatal RCTs published since the 1950s. Methods: We analyzed individual studies within the Cochrane Neonatal reviews published up to December 2016. We extracted the reviewers' judgments on the ROB domains including random sequence generation, allocation concealment, blinding, incomplete outcome data, and selective reporting. We evaluated blinding of personnel in trials in which blinding was considered feasible. Results: We assessed 1980 RCTs published between 1952 and 2016 from 294 Cochrane Neonatal systematic reviews, with full ROB assessments performed in 848 trials (42.8%). Among the ROB domains, the highest proportion of trials (73%) were judged as satisfactory ("low risk") in handling incomplete outcome data, while fewest trials achieved blinding of outcome assessor (38.4%). In the last 6 decades, a progressive increase has been observed in the proportion of trials that were rated as low risk in random sequence generation, allocation concealment, and selective reporting. However, blinding was achieved in less than half of the trials with no clear improvement across decades (23-44% since the 1980s). Conclusions: Despite steady improvement in the overall quality of neonatal RCTs over the last 6 decades, blinding remained unsatisfactory in the majority of the trials.

AB - Background: The introduction of Neonatology as a subspecialty in 1960 has stimulated an enormous amount of neonatal research. A large proportion of neonatal randomizedcontrolled trials (RCTs) have been included in the Cochrane reviews, within which methodological quality or risk-of-bias (ROB) assessment is an integral feature. Objectives: We described the ROB profile of neonatal RCTs published since the 1950s. Methods: We analyzed individual studies within the Cochrane Neonatal reviews published up to December 2016. We extracted the reviewers' judgments on the ROB domains including random sequence generation, allocation concealment, blinding, incomplete outcome data, and selective reporting. We evaluated blinding of personnel in trials in which blinding was considered feasible. Results: We assessed 1980 RCTs published between 1952 and 2016 from 294 Cochrane Neonatal systematic reviews, with full ROB assessments performed in 848 trials (42.8%). Among the ROB domains, the highest proportion of trials (73%) were judged as satisfactory ("low risk") in handling incomplete outcome data, while fewest trials achieved blinding of outcome assessor (38.4%). In the last 6 decades, a progressive increase has been observed in the proportion of trials that were rated as low risk in random sequence generation, allocation concealment, and selective reporting. However, blinding was achieved in less than half of the trials with no clear improvement across decades (23-44% since the 1980s). Conclusions: Despite steady improvement in the overall quality of neonatal RCTs over the last 6 decades, blinding remained unsatisfactory in the majority of the trials.

KW - Neonatology

KW - Randomized controlled trials

KW - Risk of bias

UR - http://www.scopus.com/inward/record.url?scp=85066912320&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85066912320&partnerID=8YFLogxK

U2 - 10.1159/000497423

DO - 10.1159/000497423

M3 - Review article

C2 - 31108494

AN - SCOPUS:85066912320

VL - 116

SP - 123

EP - 131

JO - Neonatology

JF - Neonatology

SN - 1661-7800

IS - 2

ER -