对与术后恶心呕吐相关的系统评价和荟萃分析摘要中“spin”的评估。

Evaluation of spin in the abstracts of systematic reviews and meta-analyses relating to postoperative nausea and vomiting.

作者信息

Bruns Matthew, Manojkumar Arvind, Ottwell Ryan, Hartwell Micah, Arthur Wade, Roberts Will, White Brad, Young Jeff, Martin Janet, Wright Drew N, Chen Suhao, Miao Zhuqi, Vassar Matt

机构信息

From the Office of Medical Student Research, Oklahoma State University Center for Health Sciences, Tulsa, Oklahoma (MB, RO, MH, WA, WR, MV), Arkansas College of Osteopathic Medicine, Fort Smith, Arkansas (AM), Department of Internal Medicine, University of Oklahoma, School of Community Medicine, Tulsa, Oklahoma (RO), Department of Dermatology, St. Joseph Mercy Hospital, Ann Arbor, Michigan (RO), Department of Anesthesiology, Oklahoma State University Medical Center, Tulsa, Oklahoma (WR, BW, JY), Samuel J. Wood Library & C. V. Starr Biomedical Information Center, Weill Cornell Medical College, New York, New York, USA (MH, DNW), Department of Anesthesia & Perioperative Medicine, MEDICI Centre, Western University, London, Ontario, Canada (JM), Department of Psychiatry and Behavioral Sciences, Oklahoma State University Center for Health Sciences, Tulsa (MV), School of Industrial Engineering and Management (SC) and Center for Health Systems Innovation, Oklahoma State University, Stillwater, Oklahoma, USA (ZM).

出版信息

Eur J Anaesthesiol. 2022 Jul 6:701-710. doi: 10.1097/EJA.0000000000001709.

DOI:10.1097/EJA.0000000000001709

PMID:35796313

Abstract

INTRODUCTION

Spin - the beautification of study results to emphasise benefits or minimise harms - is a deceptive reporting strategy with the potential to affect clinical decision-making adversely. Few studies have investigated the extent of spin in systematic reviews. Here, we sought to address this gap by evaluating the presence of the nine most severe forms of spin in the abstracts of systematic reviews on treatments for postoperative nausea and vomiting (PONV). PONV has the potential to increase hospital costs and patient burden, adversely affecting outcomes.

METHODS

We developed search strategies for MEDLINE and Embase to identify systematic reviews focused on PONV. Following title and abstract screening of the reviews identified during the initial search, those that met inclusion criteria were evaluated for the presence of spin and received a revised AMSTAR-2 (A Measurement Tool to Assess Systematic Reviews) appraisal by two investigators in a masked, duplicate manner. Study characteristics for each review were also extracted in duplicate.

RESULTS

Our systematic search returned 3513 studies, of which 130 systematic reviews and meta-analyses were eligible for data extraction. We found that 29.2% of included systematic reviews contained spin (38/130). Eight of the nine types of spin were identified, with spin type 3 ('selective reporting of or overemphasis on efficacy outcomes or analysis favouring the beneficial effect of the experimental intervention') being the most common. Associations were found between spin and funding source. Spin was more likely in the abstracts of privately funded than nonfunded studies, odds ratio (OR) 2.81 [95% confidence interval (CI), 0.66 to 11.98]. In the abstracts of studies not mentioning funding spin was also more likely than in nonfunded studies, OR 2.30 (95% CI, 0.61 to 8.70). Neither of these results were statistically significant. Significance was found in the association between the presence of spin and AMSTAR-2 ratings: 'low' quality studies were less likely to contain spin than 'high' quality, OR 0.24 (95% CI, 0.07 to 0.88): 'critically low' studies were also less likely to contain spin than 'high' quality studies, OR 0.21 (95% CI, 0.07 to 0.65). There were no other associations between spin and the remaining extracted study characteristics or AMSTAR-2 ratings.

CONCLUSION

Spin was present in greater than 29% of abstracts of systematic reviews and meta-analyses regarding PONV. Various stakeholders must take steps to improve the reporting quality of abstracts on PONV.

摘要

引言

结果粉饰——美化研究结果以强调益处或最小化危害——是一种具有欺骗性的报告策略，有可能对临床决策产生不利影响。很少有研究调查系统评价中的结果粉饰程度。在此，我们试图通过评估关于术后恶心呕吐（PONV）治疗的系统评价摘要中九种最严重的结果粉饰形式的存在情况来填补这一空白。PONV有可能增加医院成本和患者负担，对治疗结果产生不利影响。

方法

我们制定了针对MEDLINE和Embase的检索策略，以识别聚焦于PONV的系统评价。在对初始检索中识别出的系统评价进行标题和摘要筛选后，对符合纳入标准的系统评价进行结果粉饰情况评估，并由两名研究人员以盲法、重复的方式对其进行修订后的AMSTAR-2（评估系统评价的测量工具）评价。每个系统评价的研究特征也进行了重复提取。

结果

我们的系统检索返回了3513项研究，其中130项系统评价和荟萃分析符合数据提取条件。我们发现，纳入的系统评价中有29.2%包含结果粉饰（38/130）。识别出了九种结果粉饰类型中的八种，其中类型3（“选择性报告或过度强调疗效结果或分析偏向实验干预的有益效果”）最为常见。结果粉饰与资金来源之间存在关联。在摘要中，私人资助的研究比未获资助的研究更有可能出现结果粉饰，比值比（OR）为2.81[95%置信区间（CI）为0.66至11.98]。在未提及资金的研究摘要中，出现结果粉饰的可能性也高于未获资助的研究，OR为2.30（95%CI为0.61至8.70）。这两个结果均无统计学意义。结果粉饰的存在与AMSTAR-2评分之间存在显著关联：“低”质量的研究比“高”质量的研究更不可能包含结果粉饰，OR为0.24（95%CI为0.07至0.88）；“极低”质量的研究比“高”质量的研究更不可能包含结果粉饰，OR为0.21（95%CI为0.07至0.65）。结果粉饰与其余提取的研究特征或AMSTAR-2评分之间没有其他关联。