发表文献中的统计分析与方法学指南(Statistical Analyses and Methods in the Published Literature, the SAMPL guideline)——概述

发布于 2023年8月16日 星期三 20:53:00 浏览:734
原创不易,转载请注明来源,感谢!

合理的统计设计、正确的统计分析与解读、规范的统计报告是保证研究质量和数据结果真实性的前提,更是提高研究透明度和可重复性的重要保障。本文主要介绍发表文献中的统计分析与方法学指南(Statistical Analyses and Methods in the Published Literature, the SAMPL guideline)的背景、内容及注意事项。

关键词:医学期刊; 医学论文; 报告规范; 统计分析报告规范; 统计结果报告规范; SAMPL

一、SAMPL指南背景

生物医学文献中统计学方法误用、结果解读错误以及统计报告不足等问题较为普遍,不仅存在于一些高级统计学方法,更多地是一些基础统计方法的误用、误解。目前,不少生物医学相关期刊并未对投稿文章的统计学提出明确或详细要求。鉴于此,发布一套全面、详细且易于理解的统计学方法结果相关报告指南十分必要。

2013年,Lang和Altman整理并发布了发表文献中的统计分析与方法学指南(the Statistical Analyses and Methods in the Published Literature guideline, the SAMPL guideline),该指南与国际医学期刊编辑委员会(International Committee of Medical Journal Editors,ICMJE)发布的“Recommendations for the conduct, reporting, editing, and publication of scholarly work in medical journals(医学期刊学术著作实施、报告、编辑和发表建议)”一起为众多期刊明确自身的统计学报告要求提供了重要参考。

SAMPL指南包括报告统计方法和结果的指导原则(guiding principles for reporting statistical methods and results)、统计方法报告的一般原则(general principles for reporting statistical methods)、统计结果报告的一般原则(general principles for reporting statistical results)三部分内容。

二、SAMPL指南内容

(一) 统计方法和结果报告的指导原则

SAMPL指南对统计方法和结果报告的第一个指导原则来自ICMJE对提交到生物学期刊的稿件提出的统一要求。具体如下:描述统计方法时须提供足够的细节,从而使一个受过专业教育的读者获得原始数据时可以验证报告的结果。如果可能的话,作者应量化结果并用合适的误差指标或不确定性(如置信区间)表示;同时应避免完全依赖于统计学假设检验,如P值,这样会导致无法传达效应大小等重要信息。报告时应明确定义统计术语、缩写和大部分符号,并说明使用的分析软件。

SAMPL指南对统计方法和结果报告的第二指导原则是报告时应提供足够的细节使结果可以合并到其他分析。这一原则一般要求报告可推导出其他统计量的估计结果,如百分数的分子和分母,特别是风险比、比值比和危害比;还需要报告变量的描述性统计结果,包括样本量、P值相关的估计值(或“效应大小”)和对估计值的精度测量,通常是95%置信区间,注意不可仅报告P值。

(二) 统计方法报告的一般原则

统计方法报告的一般原则包括初步分析、主要分析、补充分析三个方面。具体如下:

1  Preliminary analyses (初步分析)
(1)Identify any statistical procedures used to modify raw data before analysis. Examples include mathematically transforming continuous measurements to make distributions closer to the normal distribution, creating ratios or other derived variables, and collapsing continuous data into categorical data or combining categories确定分析前原始数据清洗及转换的统计过程。例如,转换连续变量使其更服从正态分布;创建率比或其他衍生变量;将连续变量转换为分类变量;对分类变量的不同类别进行合并等
2  Primary analyses (主要分析)
(1)Describe the purpose of the analysis描述分析目的
(2)Identify the variables used in the analysis and summarize each with descriptive statistics.确定分析中使用的变量,并给出每一个变量的描述性统计结果
(3)When possible, identify the smallest difference considered to be clinically important如可能,确定具有临床意义的最小差异
(4)Describe fully the main methods for analyzing the primary objectives of the study全面描述分析主要研究目标使用的主要方法
(5)Make clear which method was used for each analysis, rather than just listing in one place all the statistical methods used.指明每个分析使用的方法,而不是只在一个地方列出所有使用的统计学方法
(6)Verify that data conformed to the assumptions of the test used to analyze them. In particular, specify that(1)skewed data were analyzed with non-parametric tests(2) paired data were analyzed with paired tests(3)the underlying relationship analyzed with linear regression models was linear.验证数据是否满足相应统计方法的假设前提和适用条件。特别要说明:(1) 偏态数据采用了非参数检验配对数据使用了配对假设检验(3) 使用线性回归模型分析的潜在变量间关系本质是线性的
(7)Indicate whether and how any allowance or adjustments were made for multiple comparisons (performing multiple hypothesis tests on the same data)指出是否对多重比较(对同一数据执行多次假设检验)进行校正,以及校正的方法
(8)If relevant, report how any outlying data were treated in the analysis如涉及,报告分析时如何处理异常值
(9)Say whether tests were one- or two-tailed and justify the use of one-tailed tests说明检验是单侧还是双侧,如使用的是单侧检验,解释其理由
(10)Report the alpha level (e.g., 0.05) that defines statistical significance报告具有统计学意义的α水平(如0.05)
(11)Name the statistical package or program used in the analysis报告分析中使用的统计软件或程序包的名称
3  Supplementary analyses (补充分析)
(1)Describe methods used for any ancillary analyses, such as sensitivity analyses, imputation of missing values, or testing of assumptions underlying methods of analysis描述辅助分析的统计方法,如敏感性分析、缺失值填补、或统计方法所基于的假设检验
(2)Identify post-hoc analyses, including unplanned subgroup analyses, as exploratory指出探索性的事后分析,包括未事先计划的亚组分析

(三)  统计结果报告的一般原则

根据需要报告的统计结果类型不同,统计结果报告的一般原则分为数值和描述性统计(reporting numbers and descriptive statistics)报告,风险、率和比值报告(reporting risk, rates, and ratios),假设检验报告(reporting hypothesis tests),关联分析报告 (reporting association analyses),相关分析报告(reporting correlation analyses), 回归分析报告(reporting regression analyses),方差或协方差分析报告[reporting analyses of variance (ANOVA) or of covariance (ANCOVA)],生存(时间-事件)分析报告[reporting survival (time-to-event) analyses],贝叶斯分析报告 (reporting Bayesian analyses)9个部分。

1. 数值和描述性统计报告

(1)Report numbers—especially measurements— with an appropriate degree of precision. For ease of comprehension and simplicity, round to a reasonable extent. For example, mean age can often be rounded to the nearest year without compromising either the clinical or the statistical analysis. If the smallest meaningful difference on a scale is 5 points, scores can be reported as whole numbers; decimals are not necessary报告数值(特别是测量值)及其相应的精确度。为了便于理解和简化,可将数值保留至合理的精确度。例如,通常可将平均年龄四舍五入到年,且不会影响临床或统计分析。如果量表里有意义的最小差值是5分,则可按整数报告分值,没有必要报告小数
(2)Report total sample and group sizes for each analysis报告每个分析的总样本量和每组样本量
(3)Report numerators and denominators for all percentages报告所有百分比的分子和分母
(4)Summarize data that are approximately normally distributed with means and standard deviations (SD). Use the form: mean (SD), not mean±SD用均数和标准差描述近似正态分布的数据。格式为均数(标准差),而不是均数±标准差
(5)Summarize data that are not normally distributed with medians and interpercentile ranges, ranges, or both. Report the upper and lower boundaries of interpercentile ranges and the minimum and maximum values of ranges, not just the size of the range用中位数和百分位数间距、全距或两者结合来描述非正态分布的数据。报告百分位数间距的上下边界和全距的最大、最小值,而不仅是全距的大小
(6)Do NOT use the standard error of the mean (SE) to indicate the variability of a data set. Use standard deviations, inter-percentile ranges, or ranges instead. (The SE is an inferential statistic — it is about a 68% confidence interval — not a descriptive statistic.)不要使用均值的标准误(SE)来表示数据的变异程度,而应使用标准差、百分位数间距或全距。SE是统计学推断统计量而不是描述性统计量,它相当于68%的置信区间
(7)Display data in tables or figures. Tables present exact values, and figures provide an overall assessment of the data用图或表展示数据。表展现了准确的变量及数据,图提供了对数据的整体评估

2. 风险、率和比值报告

(1)Identify the type of rate (e.g., incidence rates; survival rates), ratio (e.g., odds ratios; hazards ratios), or risk (e.g., absolute risks; relative risk differences), being reported指明率、比值及风险的具体类型,如发病率、生存率,比值比、危害比,绝对风险、相对风险差异
(2)Identify the quantities represented in the numerator and denominator (e.g., the number of men with prostate cancer divided by the number of men in whom prostate cancer can occur)报告分子和分母的数值(如患前列腺肿瘤的人数除以可能患前列腺肿瘤的人数)
(3)Identify the time period over which each rate applies报告每个率所适用的时间段
(4)Identify any unit of population (that is, the unit multiplier: e.g., ×100; ×10,000) associated with the rate报告与率相关的人群单位(即数量级,如×100、×10 000)
(5)Consider reporting a measure of precision (a confidence interval) for estimated risks, rates, and ratios报告估算的风险、率、比值等效应值的精确度(置信区间)

3. 假设检验报告

(1)State the hypothesis being tested陈述待检验的假设
(2)Identify the variables in the analysis and summarize the data for each variable with the appropriate descriptive statistics指明分析中使用的变量,并使用合理的描述性统计量概括每一个变量
(3)If possible, identify the minimum difference considered to be clinically important如可能,确定有临床意义的最小差异
(4)For equivalence and non-inferiority studies, report the largest difference between groups that will still be accepted as indicating biological equivalence (the equivalence margin)对于等效性和非劣效性研究,报告可认为生物学等效性的组间最大差异(等效性界值)
(5)Identify the name of the test used in the analysis. Report whether the test was one- or two-tailed (justify the use of one-tailed tests) and for paired or independent samples给出分析中使用的假设检验名称。报告检验是单侧还是双侧(给出选择单侧检验的理由),数据是配对还是独立样本
(6)Confirm that the assumptions of the test were met by the data确保数据符合假设检验的适用前提
(7)Report the alpha level (e.g., 0.05) that defines statistical significance报告具有统计学意义的α水平(如0.05)
(8)At least for primary outcomes, such as differences or agreement between groups, diagnostic sensitivity, and slopes of regression lines, report a measure of precision, such as the 95% confidence interval至少报告主要结局的精确度(如95%置信区间),如组间差异或一致性、诊断灵敏度、回归线的斜率
(9)Do NOT use the standard error of the mean (SE) to indicate the precision of an estimate. The SE is essentially a 68% confidence coefficient: use the 95% confidence coefficient instead不要使用均值的标准误(SE)来描述参数估计的精确度。SE本质上是68%的置信区间,应使用95%置信区间
(10)Although not preferred to confidence intervals, if desired, P values should be reported as equalities when possible and to one or two decimal places (e.g., P = 0.03 or 0.22 not as inequalities: e.g., P < 0.05). Do NOT report ‘‘NS’’,give the actual P value. The smallest P value that need be reported is P<0.001, save in studies of genetic associationsP值并不能替代置信区间,但尽可能报告P值的具体数值,并保留1~2位小数(例如P=0.03或者0.22;不应报告为不等式,如P<0.05)。不要报告“无统计学意义”,应给出确切的P值。除遗传关联性研究外,一般需要报告的最小P值是P<0.001
(11)Report whether and how any adjustments were made for multiple statistical comparisons说明是否对多重比较进行调整,如有调整说明使用的方法
(12)Name the statistical software package used in the analysis报告分析中使用的统计软件包的名称

4. 关联分析报告

(1)Describe the association of interest报告所研究的关联大小
(2)Identify the variables used and summarize each with descriptive statistics给出分析的变量,并对每一个变量进行描述性分析
(3)Identify the test of association used给出所使用的关联性检验的方法
(4)Indicate whether the test was one- or two-tailed. Justify the use of one-tailed tests说明检验是单侧还是双侧,若为单侧检验,说明其合理性
(5)For tests of association (e.g., a chi-square test), report the P value of the test (because association is defined as a statistically significant result)对于关联检验(如卡方检验),报告检验的P值(因为关联的定义是指具有统计学意义的结果)
(6)For measures of association (i.e., the phi coefficient), report the value of the coefficient and a confidence interval. Do NOT describe the association as low, moderate, or high unless the ranges for these categories have been defined. Even then, consider the wisdom of using these categories given their biological implications or realities对关联的估计(如phi相关系数),报告系数的估计值和置信区间。除非已经提前定义了关联为大、中、小的范围,否则不要使用这些词语来描述关联大小。考虑到这些分类的生物学意义或实际意义,即使使用这些词语来描述关联程度也需慎重
(7)For primary comparisons, consider including the full contingency table for the analysis对于主要比较,可考虑使用完整的列联表进行分析
(8)Name the statistical package or program used in the analysis报告分析中使用的统计软件包的名称

5. 相关分析报告

(1)Describe the purpose of the analysis描述分析目的
(2)Summarize each variable with the appropriate descriptive statistics用合适的描述性统计量描述每一个分析变量
(3)Identify the correlation coefficient used in the analysis (e.g., Pearson, Spearman)指明分析中所使用的相关系数(例如Pearson、Spearman相关系数)
(4)Confirm that the assumptions of the analysis were met确保数据符合检验的假设前提
(5)Report the alpha level (e.g., 0.05) that indicates whether the correlation coefficient is statistically significant报告相关系数具有统计学意义的α水平 (例如0.05)
(6)Report the value of the correlation coefficient. Do NOT describe correlation as low, moderate, or high unless the ranges for these categories have been defined. Even then, consider the wisdom of using these categories given their biological implications or realities报告相关系数的大小。除非已经提前定义了相关系数为大、中、小的范围,否则不要使用这些词语来描述关联大小。考虑到这些分类的生物学意义或实际意义,即使使用这些词语来描述关联程度也需慎重
(7)For primary comparisons, report the (95%) confidence interval for the correlation coefficient, whether or not it is statistically significant对于主要比较,无论是否具有统计学意义,均应报告相关系数的95%置信区间
(8)For primary comparisons, consider reporting the results as a scatter plot. The sample size, correlation coefficient (with its confidence interval), and P value can be included in the data field对于主要比较,建议使用散点图报告结果,图中应包括样本量、相关系数(包含置信区间)和P值等信息
(9)Name the statistical package or program used in the analysis报告分析中使用的统计软件或程序包的名称

6. 回归分析报告

(1)Describe the purpose of the analysis描述分析目的
(2)Identify the variables used in the analysis and summarize each with descriptive statistics给出分析的变量,并对每一个变量进行描述性分析
(3)Confirm that the assumptions of the analysis were met. For example, in linear regression indicate whether an analysis of residuals confirmed the assumptions of linearity确保数据符合假设检验的适用前提。例如,在线性回归中应说明残差分析结果是否满足线性回归的适用前提
(4)If relevant, report how any outlying values were treated in the analysis如涉及,报告分析时如何处理异常值
(5)Report how any missing data were treated in the analyses报告分析时如何处理缺失数据
(6)For either simple or multiple (multivariable) regression analyses, report the regression equation对于单因素或多因素回归分析,应报告回归方程
(7)For multiple regression analyses: (1) report the alpha level used in the univariate analysis; (2) report whether the variables were assessed for (a) colinearity and (b) interaction; and (3) describe the variable selection process by which the final model was developed (e.g., forward-stepwise; best subset)对多因素回归分析:(1) 报告单因素分析时使用的α水平;(2) 报告是否对变量进行(a)共线性和(b)交互作用的评估;(3) 描述进入最终模型的变量选择过程(例如前进法、逐步法、最优子集法)
(8)Report the regression coefficients (beta weights) of each explanatory variable and the associated confidence intervals and P values, preferably in a table最好在表格里报告每个解释变量的回归系数(β系数)、置信区间以及P
(9)Provide a measure of the model’s ‘‘goodness-of-fit’’ to the data (the coefficient of determination, r2, for simple regression and the coefficient of multiple determination, R2, for multiple regression)提供数据拟合优度的结果(简单回归中报告决定系数r2,多因素回归中报告决定系数R2)
(10)Specify whether and how the model was validated说明是否对模型进行了验证,以及如何进行验证
(11)For primary comparisons analyzed with simple linear regression analysis, consider reporting the results graphically, in a scatter plot showing the regression line and its confidence bounds. Do not extend the regression line (or the interpretation of the analysis) beyond the minimum and maximum values of the data对于用简单线性回归分析的主要比较,考虑用图展示结果,在散点图中展示回归直线及其置信区间。不要将回归线(或进行分析解释)延伸到数据的最小值和最大值之外
(12)Name the statistical package or program used in the analysis报告分析中使用的统计软件或程序包的名称

7. 方差或协方差分析报告

(1)Describe the purpose of the analysis描述分析目的
(2)Identify the variables used in the analysis and summarize each with descriptive statistics给出分析的变量,并对每一个变量进行描述性分析
(3)Confirm that the assumptions of the analysis were met. For example, indicate whether an analysis of residuals confirmed the assumptions of linearity确保数据符合假设检验的适用前提。例如,说明残差分析结果是否能符合线性的假设前提
(4)If relevant, report how any outlying data were treated in the analysis如涉及,报告分析时如何处理异常值
(5)Report how any missing data were treated in the analyses报告分析时如何处理缺失数据
(6)Specify whether the explanatory variables were tested for interaction, and if so how these interactions were treated明确说明是否检测了解释变量之间的交互作用,如果有,交互作用是如何处理的
(7)If appropriate, in a table, report the P value for each explanatory variable, the test statistics and, where applicable, the degrees of freedom for the analysis如可以,在表格中报告每个解释变量的P值、检验统计量和自由度
(8)Provide an assessment of the goodness-of-fit of the model to the data, such as R2提供对模型“拟合优度”评估的结果,例如R2
(9)Specify whether and how the model was validated说明是否对模型进行验证,以及如何进行验证
(10)Name the statistical package or program used in the analysis报告分析中使用的统计软件或程序包的名称

8. 生存(时间-事件)分析报告

(1)Describe the purpose of the analysis描述分析目的
(2)Identify the dates or events that mark the beginning and the end of the time period analyzed明确分析时所定义的起止时间或标志事件
(3)Specify the circumstances under which data were censored明确截尾数据的定义
(4)Specify the statistical methods used to estimate the survival rate说明用来估计生存率的统计方法
(5)Confirm that the assumptions of survival analysis were met确认数据符合生存分析的假设前提
(6)For each group, give the estimated survival probability at appropriate follow-up times, with confidence intervals, and the number of participants at risk for death at each time. It is often more helpful to plot the cumulative probability of not surviving, especially when events are not common对每一组,报告合适随访时间下所估计的生存概率及其置信区间,及每个随访时间下对应的存在死亡风险的研究对象人数。绘制累积死亡概率曲线图更为有价值,尤其当死亡事件发生较少时
(7)Reporting median survival times, with confidence intervals, is often useful to allow the results to be compared with those of other studies报告中位生存时间及其置信区间,这些数据有利于与其他研究结果进行比较
(8)Consider presenting the full results in a graph (e.g., a Kaplan–Meier plot) or table考虑在图(例如,Kaplan-Meier生存曲线图)或表中呈现所有结果
(9)Specify the statistical methods used to compare two or more survival curves说明用来比较两条或多条生存曲线的统计方法
(10)When comparing two or more survival curves with hypothesis tests, report the P value of the comparison采用假设检验比较两条或多条生存曲线时,报告比较的P
(11)Report the regression model used to assess the associations between the explanatory variables and survival or time-to-event报告用于评估解释变量和生存(或时间-事件)之间关联的回归模型
(12)Report a measure of risk (e.g., a hazard ratio) for each explanatory variable, with a confidence interval报告每个解释变量的风险估计值(例如风险比)及其置信区间

9. 贝叶斯分析报告

(1)Specify the pre-trial probabilities (‘‘priors’’)给出验前概率(“先验概率”)
(2)Explain how the priors were selected说明先验概率是如何确定的
(3)Describe the statistical model used描述使用的统计模型
(4)Describe the techniques used in the analysis描述分析所使用的技术
(5)Identify the statistical software program used in the analysis指明分析所使用的统计软件程序
(6)Summarize the posterior distribution with a measure of central tendency and a credibility interval报告后验分布的集中趋势和可信区间
(7)Assess the sensitivity of the analysis to different priors评估分析对不同先验概率的敏感性

三、SAMPL指南使用注意事项

本文介绍的统计方法或结果报告条目是不同研究设计类型均可能涉及的共性问题,当研究者撰写论文时,应结合自身的研究设计类型、研究目的、学术期刊要求等参考适用的条目,而非逐一报告。各种具体研究的报告规范,如CONSORT声明、STROBE声明等,也会涉及统计学方法和结果报告的细则,这些细则更具针对性,可同步参考。此外,很多期刊都有自己的统计方法或结果报告的要求,研究者投稿时也需遵从这些具体指引。值得注意的是,还有很多期刊虽然对统计分析的格式和内容有具体说明,但并不充分,建议根据国际通用的规范完善现行的统计报告要求。

关于SAMPL指南的更多内容详见Thomas A. Lang a和Douglas G. Altman发表的论文Basic statistical reporting for articles published in Biomedical Journals: The ‘‘Statistical Analyses and Methods in the Published Literature’’ or the SAMPL Guidelines (https://pubmed.ncbi.nlm.nih.gov/25441757/)。

End
文章目录 沉浸式阅读