\documentclass[10pt]{article}
\usepackage{fullpage}
\usepackage{setspace}
\usepackage{parskip}
\usepackage{titlesec}
\usepackage[section]{placeins}
\usepackage{xcolor}
\usepackage{breakcites}
\usepackage{lineno}
\usepackage{hyphenat}
\PassOptionsToPackage{hyphens}{url}
\usepackage[colorlinks = true,
linkcolor = blue,
urlcolor = blue,
citecolor = blue,
anchorcolor = blue]{hyperref}
\usepackage{etoolbox}
\makeatletter
\patchcmd\@combinedblfloats{\box\@outputbox}{\unvbox\@outputbox}{}{%
\errmessage{\noexpand\@combinedblfloats could not be patched}%
}%
\makeatother
\usepackage{natbib}
\renewenvironment{abstract}
{{\bfseries\noindent{\abstractname}\par\nobreak}\footnotesize}
{\bigskip}
\titlespacing{\section}{0pt}{*3}{*1}
\titlespacing{\subsection}{0pt}{*2}{*0.5}
\titlespacing{\subsubsection}{0pt}{*1.5}{0pt}
\usepackage{authblk}
\usepackage{graphicx}
\usepackage[space]{grffile}
\usepackage{latexsym}
\usepackage{textcomp}
\usepackage{longtable}
\usepackage{tabulary}
\usepackage{booktabs,array,multirow}
\usepackage{amsfonts,amsmath,amssymb}
\providecommand\citet{\cite}
\providecommand\citep{\cite}
\providecommand\citealt{\cite}
% You can conditionalize code for latexml or normal latex using this.
\newif\iflatexml\latexmlfalse
\providecommand{\tightlist}{\setlength{\itemsep}{0pt}\setlength{\parskip}{0pt}}%
\AtBeginDocument{\DeclareGraphicsExtensions{.pdf,.PDF,.eps,.EPS,.png,.PNG,.tif,.TIF,.jpg,.JPG,.jpeg,.JPEG}}
\usepackage[utf8]{inputenc}
\usepackage[ngerman,english]{babel}
\usepackage{float}
\begin{document}
\title{Subgroup analysis in Haematologic Malignancies Phase III Clinical
Trials: A systematic review}
\author[1]{Nerea Báez Gutiérrez}%
\author[1]{Héctor Rodríguez-Ramallo}%
\author[1]{Laila Abdel-kader Martín}%
\author[1]{Sandra Flores-Moreno}%
\affil[1]{Virgen del Rocio University Hospital}%
\vspace{-1em}
\date{\today}
\begingroup
\let\center\flushleft
\let\endcenter\endflushleft
\maketitle
\endgroup
\selectlanguage{english}
\begin{abstract}
Aims: To assess the appropriateness of the use and interpretation of
subgroup analysis in haematology randomized clinical trials (RCT).
Method: A systematic review of Medline including Haematology phase III
RCT published between January 2013 and October 2019 was carried to
identify subgroup analysis reported. Information related to trials
characteristics, subgroup analysis reported and claims of subgroup
difference were collected. Results: A total of 98 studies reporting
subgroup analyses were identified. Of those, 24 RCT reported 46 claims
of subgroup difference. Among them, 44 were claims for the primary
outcome, of which 25 were considered strong claims and 17 were
considered suggestions of a possible effect. Authors included subgroup
variables for the primary outcome measured at baseline for 38 claims (n
= 86.36\%), used subgroup variable as stratification factor at
randomization for 15 (34.09\%), clearly prespecify their hypothesis for
11 (25\%), the subgroup effect was one of a small number of hypothesised
effects tested for 17 (38.36\%), carried out a test of interaction that
provide statistically significant for 18 (40.91\%), documented
replication of a subgroup effect with previously related studies for 11
(25\%), identify consistency of a subgroup effect across related outcome
for 10 (22.72\%), and provided a biological rationale for the effect for
8 (18.18\%). Of the 44 claims for the primary outcome, 34 (77.27\%) met
4 or fewer of the 10 credibility criteria. Conclusion: Credibility of
subgroup claims reported in haematology RCT lack of credibility, even
when claims are strong. Information about subgroup difference should be
interpreted ca%
\end{abstract}%
\sloppy
\textbf{Introduction}
Subgroup analysis are important elements in the report of results of
randomized clinical trials (RCTs)\textsuperscript{1 -2}. Clinical
practice guidelines (GPC) recommendations in haematology are guided by
phase III RCTs results. Usually only average results are reported in
RCTs and trial participants are frequently recruited from a
heterogeneous population. Subgroup analysis are born with the aim to
detect subgroup effects, in other words they have the aim of predicting
which patients will benefit more from therapies\textsuperscript{2-4}.
Interpretation of subgroup analysis is potentially important for
treatment decisions in medical practice. Subgroup analysis can provide
clinicians with a better perspective on the individualized treatment of
patients, which is particularly interesting in the field of haematology
due to the lower therapeutic index and higher toxicity of
used-drugs\textsuperscript{5-6}. However, subgroup analysis can
introduce analytical challenges leading to misleading and exaggerated
results, which may result in denial of a beneficial treatment or even
receiving a potentially harmful or ineffective
treatment\textsuperscript{7-9.}
Subgroups analysis have the potential to generate hypotheses for further
prospective investigation\textsuperscript{10}, their exploratory nature
requires results to be confirmed in a new study to ensure their findings
with statistical reliability. However, confirmatory studies are
generally never carried out and decisions in clinical practice are made
with this lack of information. On the other hand, the option of
completely discarding subgroups analysis finding is also a decision that
has its consequences, and is especially controversial in situations with
very high risks or costs that are difficult to assume, which are not
uncommon in clinical practice\textsuperscript{11}.
Concerns about the correct interpretation of subgroup analysis has
recently grown. With the intention of reducing the problems related to
subgroup analysis misinterpretation, several tools have been developed
to assess the credibility of the effects of subgroups reported in
RCTs\textsuperscript{12-17}.
With the results of this study we will be able to determine if subgroup
analysis claims of phase III RCTs in haematology malignancies are
carried out correctly.
The main objective of this study is to assess the appropriateness of the
use and interpretation of subgroup analysis in recently published
haematologic malignancies RCTs. To achieve our objective the following
aspects will be evaluated:
1- To describe subgroups analysis and claims of subgroup effects.
2- To assess study characteristics of subgroup analyses.
3- To examine the analysis and interpretation of subgroup effects for
primary outcomes and to assess the credibility of subgroup claims using
``the 10 criteria for assessing the credibility of a subgroup claim'' by
Sun et al 2012\textsuperscript{17}.
\textbf{Methods:}
\emph{Literature Search}
This Systematic review was designed to summarize the available data
addressing the following research question, framed in the
Population-Intervention-Comparator- Outcome-Study design (PICOS)
framework: (Population) Patients with haematological malignancies;
(Intervention) subgroup analysis; (Comparison) studies with comparator
will be considered; (Outcomes), subgroup analysis; (Study design), phase
III randomized clinical trials.
A systematic search was performed following Preferred Reporting Items
for a Systematic Review and Meta-analysis (PRISMA)
guidelines\textsuperscript{18}. The search was performed using Mesh
terms-controlled vocabulary and keywords in MEDLINE database (OVID
interfaz including In-process and Epub ahead of print) between January
2013 and October 2019, to identify publications of phase III RCT
assessing systemic therapies for haematological malignancies.
The search was performed on October 2019. The full literature search
strategy is available at supplemental material (Appendix A).
The following criteria were used for trial selection
Eligibility criteria:
We considered eligible all published Phase III randomized clinical
trials for haematological malignancies with subgroup analysis reported.
Not language restriction was applied.
Exclusion criteria:
\begin{enumerate}
\tightlist
\item
Paediatric patients (\textless{}18 years of age).
\item
Pooled data from two or more trials.
\item
Studies exploring devices, behavioural or supportive care
interventions.
\item
The report does not include the entire population enrolled in the
original article (i.e. the report focuses on a subset of the original
study population).
\end{enumerate}
In cases were multiple publications from the same trial were identified,
the initial publication was used for the analysis if it was published
during the studied period.
\emph{Study Screening and Selection}
Two investigators independently examined the titles and abstracts of the
search results using the predefined inclusion criteria. For all titles
that appear to meet the inclusion criteria or those where there was some
uncertainty, full text was accessed. The two reviewers assessed whether
the articles met the selection criteria. Any disagreements were resolved
by discussion or arbitration from a third reviewer. Reasons for
excluding studies were recorded and is available at supplemental
material.
\emph{Data extraction}
For data extraction additional sources referenced in the included study
(i.e., trial register, published protocol and online supplements) were
used. Data were extracted and entered in a structured Microsoft Excel
(Redmond, WA, USA) database.
Eligible RCTs were evaluated to determine whether a subgroup analysis
was reported. A subgroup analysis was defined as a statistical analysis
that explores whether effects of the intervention differ according to
status of a subgroup variable. A subgroup effect was defined as a
difference in the magnitude of a treatment effect across a group of a
study population\textsuperscript{16}. For each RCT reporting subgroup
analysis and subgroup claims the following information was collected:
\begin{enumerate}
\tightlist
\item
\emph{Trial characteristics:} information on funding source, year and
journal of publication, journal impact factor (\textless{}10 o
\textgreater{}10), haematological malignancy type, disease status
(naive/untreated or refractory/relapse), type of intervention
(chemotherapy, immunotherapy or haematopoietic transplant),centre
(multicentric or unicentric), trial design (parallel, cross-over or
factorial), trial type (superiority, non-inferiority or equivalence),
allocation concealment, blinding of patients, number of patients
recruited and randomized for the trial and number of treatment arms.
The primary endpoint was categorized according to whether results were
statistically significant and the type of outcome variable
(time-to-event, binary, continuous or count).
\item
\emph{Reporting of subgroup analysis} : number of subgroup factors,
type of subgroup factors (clinical factors or biomarkers), number of
subgroup analysis and outcomes for subgroup analysis reported, forest
plots used, prespecify or post hoc subgroup, statistical method used
to assess heterogeneity of the treatment effect (descriptive only,
subgroup P values and confidence interval or interaction test).
\end{enumerate}
A subgroup factor was defined as each of the subgroup analysed in the
RCT (i.e. sex, age, presence of a mutation).
\emph{Claims of subgroup effects:} Subgroup claims mode of presentation
(abstract or text only), number of subgroup claims, subgroup variable
(primary or secondary outcome) and number of outcomes for subgroup
claims were recorded. A subgroup effect was considered claimed when the
authors states in the abstract or discussion that the effect of
intervention differs between the categories of the subgroup variable.
Claims of subgroup effect were classified according to the strength of
the claim into 3 categories: Strong claim, claim of a likely effect or
suggestion of a possible effect based on Sun et al 2009
clasification\textsuperscript{16}(Appendix B). To evaluate the
credibility of subgroup claims for primary outcomes ``the 10 criteria
for assessing the credibility of a subgroup claim'' by Sun et al
2012\textsuperscript{17} were applied (Appendix C). These criteria have
been widely used\textsuperscript{13-15,17} and are recommended for
assessing how much confidence to place in subgroup
analyses\textsuperscript{19}. If the subgroup claim met less than half
of criteria, the credibility of this claim was considered low.
\emph{Assessment of risk of bias}
Risk of bias was assessed using the Cochrane Collaboration\selectlanguage{ngerman}´s tool for
assessing risk of bias in randomized trials\textsuperscript{20}. This
tool is composed by 5 domains: bias arising from the randomization
process; bias due to deviations from intended interventions; bias due to
missing outcome data; bias in measurement of the outcome; and bias in
selection of the reported result. For each domain, the tool comprises:
a series of `signalling questions'; a judgement about risk of bias for
the domain, which is facilitated by an algorithm that maps responses to
the signalling questions to a proposed judgement; free text boxes to
justify responses to the signalling questions and risk-of-bias
judgements; and an option to predict (and explain) the likely direction
of bias. Risk of bias was assessed by two independent reviewers.
Possible disagreements between reviewers were resolved by discussion or
arbitration by a third reviewer when consensus could not be reached.
\emph{Data analysis}
A descriptive analysis was developed. Continuous and categorical
variables were presented as mean (range) and n (\%), respectively.
For those RCTs that stated a subgroup effect without providing an
interaction test, p interaction was calculated using the Joaquin Primo
calculator\textsuperscript{21}\textbf{,} to verify that there was indeed
statistical significance.
\textbf{Results}
The literature search identified 1622 studies. After a first review by
title or abstract and removing duplicates, 321 articles were selected
for a full text review. Finally, 98 articles were included. (Figure 1).
Articles excluded and the reason of exclusion are available at
supplemental material (Appendix D).
\emph{Characteristics of trials included in the analysis}
The characteristics of the trials included in this study are listed in
table 1. These 98 publications reported data on 48,245 randomized
patients (Median: 402; range: 82-1623).
A 77.25\% (n = 76) studies were funded by industry. Most of the trials
were published during 2015 (18.36\%; n = 18) and 2016 (20.41\%; n = 20).
The New England Journal of Medicine (26.53\%; n = 26) and Lancet
Oncology (20.41\%; n = 20) were the most selected journals for
publication of these trials. An 85.7\% (n = 84) of the studies were
published in high impact journals (impact factor \textgreater{}10).
The most common malignancies explored were Non-Hodgkin lymphoma
(25.51\%; n = 25), multiple myeloma (20.41\% n= 20), acute myeloid
leukaemia (20.41\%; n = 20), and chronic lymphocytic leukaemia (20.41\%;
n = 20). The most common intervention was chemotherapy (50\%; n = 49).
Stated primary endpoint was statistically significant in 65.31\% (n =
64) of trials.
\emph{Subgroup analysis}
Characteristics of reported subgroup analysis are listed in table 2.
Subgroup analysis were mentioned in the method section for 46.94\% (n =
46) trials, 89.90\% (n = 88) in the results sections, 56.12\% (n=55) in
the discussion section and 26.53\% (n = 26) in supplemental appendix.
At least 6 subgroup factors were reported in 63.26 \% (n = 62) of
trials. Related the type of subgroup factors 30.61\% (n = 30) were
clinical factor and 66.33\% (n = 65) were clinical factor plus
biomarkers. More than 6 subgroup analysis were reported in 71.43\% (n =
70) of the trials. More than one outcome was reported in 25.51\% (n =
25) of trials (mean:1; range:1-3). To show the results of subgroup
analysis forest plots were used in 77.55\% (n = 76) of the trials.
For 11.22\% (n = 11) of trials, it was unclear whether subgroup analysis
was prespecify or post hoc, in 50\% (n = 49) of trials were prespecify
and 31.63\% (n = 31) were post hoc.
Only 18.37\% (n = 18) use an interaction test to assess heterogeneity of
the treatment effect; a 17.35\% (n =17) reported subgroup analysis
without any statistical analysis.
\emph{Claims of subgroup effects}
Characteristics of subgroup claims are listed in table 3. In 24 RCTs
authors claim heterogeneity of treatment effect of at least one subject
subgroup, 13 made a claim for a primary outcome, 2 for secondary
outcomes and 9 for both primary and secondary outcomes. Six (25.00\%) of
these RCTS presented subgroup claims in the articles abstract and five
(20.83\%) were based on significant interaction tests, whereas the
claims were based only on within-subgroup comparisons for most of trials
(54.17\%; n = 13). More than one subgroup claim was made in 54.17\% (n =
13) of trials.
A total of 46 subgroup difference were claimed in these 24 trials (44
for primary outcomes and 2 for secondary outcomes). These claims were
classified as 26 (59.10\%) strong claims, two (4.54\%) as claims of a
likely effect and 18 (40.91\%) as suggestion of a possible effect.
Respect to the 10 criteria to assess credibility of subgroups claims
(table 4): Authors included subgroup variables for the primary outcome
measured at baseline for 38 claims (86.36\%), used subgroup variable as
stratification factor at randomization for 14 (34.09\%) claims, clearly
prespecify their hypothesis for 11 (25.00\%) claims, correctly
prespecify direction for 5 (11.36\%) claims, tested a small number of
hypothesis for 17 (38.63\%) claims, carried out a test of interaction
that provide statistically significant for 18 (40.91\%) claims,
documented replication of a subgroup effect with previously related
studies for 11 (25.00\%) claims, identify consistency of a subgroup
effect across related outcome for 10 (22.72\%) claims, and provided a
biological rationale for the effect for 8 (18.18\%) claims. Of the 44
claims for the primary outcomes, 34 (77.27\%) met 4 or fewer of the 10
criteria. For strong claims, 15 (60.00 \%) met three or less criteria
and only 6 (24.00\%) met more than 5 criteria.
Risk of Bias Graphs Within Studies and across studies is available at
supplemental material (Appendix D).
\textbf{Discussion}
Limitations of reporting subgroup analysis in RCT have been widely
reported on the literature. Inflated false positives due to multiple
testing, high false negatives due to inadequate statistical power and
inappropriate a priori specification are well-known limitations of
subgroup analysis\textsuperscript{2,7-8,22-24}. A prespecified subgroup
analysis is one that is planned and documented before any examination of
the data. They are more reliable than those no prespecified because
their hypotheses are based on biological rationale or data obtained on
previous studies. In this review only half of trials conducted
prespecified subgroup analysis. When analysis of a large number of
subgroups are made, even if a hypothesis has been clearly specified,
their results should be considered cautiously, since the strength of
inference associated with the apparent confirmation of any single
hypothesis will decrease if it is one of a large number that have been
tested\textsuperscript{25}. In this systematic review, multiple subgroup
analyses were performed, around three quarters of trials reported at
least 6 subgroups. Statistical analysis of interaction establishes the
difference in benefit between subgroups by calculating interaction
probability (p), which suggests that chance is an unlikely explanation
for apparent differences, therefore the interaction test is the
appropriate method to analyse subgroups. In this review only a few
trials (18.37\%) used an interaction test to assess heterogeneity of the
treatment effect.
Due to important methodological problems bias, subgroup interpretation
can lead to erroneous conclusions, producing wrongful clinical decision
making. Several tools have been developed to assess the credibility of
the effects of subgroups reported in clinical
trials\textsuperscript{12-17}. In our study we have based ourselves on
the ``10 criteria to assess credibility of subgroup claims'' by Sun et
al 2012\textsuperscript{17}. The credibility of subgroup claims in phase
III haematology RCT was low. Of the 44 claims of a subgroup effect for
the primary outcome identified, 26 were strong claims and only 24\% (n =
6) of these claims were able to satisfy at least half of the credibility
criteria and none satisfied all criteria. Multiple significant
interactions were the only criteria satisfied by more than 50\% of the
claims. All 24 assessed studies failed to prespecify the correct
direction of the subgroup hypotheses, and the hypothesis was
prespecified for only 11 (25\%) claims.
Sun et al 2012\textsuperscript{17} considered three out of their 10
criteria as critical: the use of subgroup variables measured at
baseline, prespecification of subgroup hypothesis and statistical
significance of interaction test. In our study the first of these
criteria was met for most of trials (86.36\%), however the other two
criteria were only met by 25.2\% and 40.91\% respectively. As stated
before, interaction test is the appropriate method to analyse subgroups,
but only a 40\% of strong claims of this review were made base on this
test. This finding indicates that most authors are unaware of how to
interpret a subgroup analysis correctly and make statements based on
intragroup comparisons, instead of intergroup comparisons. The latter
determines evidence of differences in the results for different
subgroups, this comparison is made by the interaction test. The lack of
compliance of previously cited criteria in the claims of the haematology
RCTS demonstrates their limited credibility.
Similar results have been reported in other studies areas. Zhang et al
2015\textsuperscript{26}, reported low credibility of subgroup claims in
phase III RCT solid tumours using The CONSORT statements to evaluate
subgroup claims\textsuperscript{27}. They found as most common problems
for reporting subgroup analysis the great number of subgroups reported,
although frequently not prespecified and the underused of interaction
test. Sun et al. 2012\textsuperscript{17} reported low credibility of
subgroup claims in pharmacological RCT published in 2007. Most of these
trials failed to prespecify the hypotheses or present significant
interaction tests. Two recent reviews investigated subgroup analysis
quality in low back pain management trials\textsuperscript{28-29} and
reported the failure to specify the subgroup hypotheses a prior as a
common problem in trials, which is also consistent with our findings.
Vidic et al 2016\textsuperscript{10} reviewed phase III cardiovascular
RCTs with subgroup analysis, concluding that subgroup analysis were
reported with several shortcomings, including lack of prespecification
and testing of a large number of subgroups without the use of the
statistically appropriate test for interaction. All these studies
reported the failure to specify the subgroup hypotheses, many subgroup
analyses conducted and underuse of interaction test as common problems
in trials, which is consistent with our findings.
By contrast in other studies the number of claims of subgroup effect in
this review was low. Zhang et al 2015\textsuperscript{26}, Sun et al
2012\textsuperscript{17}, Saragiotto et al\textsuperscript{29} and Vidic
et al 2016\textsuperscript{10} reported that a 54.26\%, 40.10\%,
57.57\%, 53.84\% of trials assessed made claims of subgroup effect,
respectively. The number of subgroup claims identify in haematological
trials was half of those reported in other areas.
This study had several strengths: It is the first systematic review of
the credibility of subgroup analysis reported on haematological
malignancies RCTs. A rigorous systematic review method was employed, and
standardized criteria were used for assessing credibility of subgroup
claims\textsuperscript{17}.
This study had several limitations: This study is based on authors'
reported trial information in published articles, which may be
vulnerable to selective reporting or underreporting. Our study was
limited to phase III RCT, although Sun et al
2012\textsuperscript{17}criteria could be applied to all phase clinical
trials. The low number of subgroup claims identified is also a
limitation of this study.
\textbf{Conclusions}
In summary, subgroup analysis in phase III haematology malignancies RCTs
are of poor quality, identifying flaws already described in other areas
of study, such as the great number of subgroups reported, inappropriate
a priori specification and the underused of interaction test.
Although not as frequent as in other areas, subgroup claims credibility
was low. Most claims do not meet critical criteria; therefore,
clinicians should interpret these results with caution. Subgroup
analysis should be carried out due to the potential information they can
provide, however researchers should be more cautious before claiming the
existence of a subgroup effect.
\textbf{Bibliography}
\par\null
\begin{enumerate}
\tightlist
\item
Guyatt G, Rennie D, Meade MO, Cook, D. Users' Guides to the Medical
Literature: A Manual for Evidence-Based Clinical Practice. Chapter 1.
Third Edition. 2015.
\item
Rothwell PM. Treating individuals 2. Subgroup analysis in randomised
controlled trials: importance, indications, and interpretation. Lancet
2005; 365 (9454): 176-86.
\item
VanderWeele TJ, Knol MJ. Interpretation of subgroup analyses in
randomized trials: heterogeneity versus secondary interventions. Ann
Intern Med 2011;154:680e3.
\item
Sun X, Ioannidis JP, Agoritsas T, Alba AC, Guyatt G.. How to use a
subgroup analysis: users' guide to the medical literature. JAMA
2014;311:405e11.
\item
Thanarajasingam G, Minasian LM, Baron F, et al. Beyond maximum grade:
modernising the assessment and reporting of adverse events in
haematological malignancies {[}published correction appears in Lancet
Haematol. 2019 Mar;6(3):e121{]}. Lancet Haematol.
2018;5(11):e563\selectlanguage{english}-e598.
\item
Bennett CL, Tigue CC, Angelotta C, McKoy JM, Edwards BJ Adverse
effects of drugs used to treat hematologic malignancies: surveillance
efforts from the research on adverse drug events and reports project.
Semin Thromb Hemost. 2007;33(4):365\selectlanguage{english}-372.
\item
Wittes J. On looking at subgroups. \emph{Circulation} .
2009;119(7):912\selectlanguage{english}-915.
\item
Wang R, Lagakos SW, Ware JH, Hunter DJ, Drazen JM. Statistics in
medicine---reporting of subgroup analyses in clinical trials. \emph{N
Engl J Med} . 2007;357(21):2189\selectlanguage{english}-2194.
\item
Koch A, Framke T. Reliably basing conclusions on subgroups of
randomized clinical trials. \emph{J Biopharm Stat} . 2014;24(1):42\selectlanguage{english}-57.
\item
Vidic A, Chibnall JT, Goparaju N, et al. Subgroup analyses of
randomized clinical trials in heart failure: facts and numbers. ESC
Heart Fail. 2016 Sep; 3(3): 152--157.
\item
Wijn SRW, Rovers MM, Le LH, et al. Guidance from key organisations on
exploring, confirming and interpreting subgroup effects of medical
treatments: a scoping review. BMJ Open. 2019;9(8):e028751.
\item
Gil-Sierra MD, Fenix-Caballero S, Abdel-Kader Martin L, et al.
Checklist for Clinical Applicability of Subgroup Analysis. J Clin
Pharm Ther. 2020 Jun;45(3):530-538
\item
Oxman AD, Guyatt GH. A consumer's guide to subgroup analyses. Ann
Intern Med 1992;116:78-84.
\item
Sun X, Briel M, Walter SD, et al. Is a subgroup effect believable?
Updating criteria to evaluate the credibility of subgroup analyses.
BMJ 2010;340:c117.
\item
Sun X, Ioannidis JPA, Agoritsas T, et al. How to use a subgroup
analysis. JAMA 2014;311:405
\item
Sun X, Briel M, Busse JW, et al. Subgroup Analysis of Trials Is Rarely
Easy (SATIRE): a study protocol for a systematic review to
characterize the analysis, reporting, and claim of subgroup effects in
randomized trials. Trials. 2009;10:101.
\item
Sun X, Briel M, Busse JW, et al. Credibility of claims of subgroup
effects in randomised controlled trials: systematic review. BMJ.
2012;344:e1553.
\item
Moher D., Liberati A., Tetzlaff J, et al. Group P. Preferred reporting
items for systematic reviews and meta-analyses: The PRISMA statement.
Int. J. Surg. 2010;8:336--341.
\item
Oxman AD. Subgroup analyses. BMJ 2012;344:e2022.
\item
Higgins JPT, Thomas J, Chandler J, Cumpston M, Li T, Page MJ, Welch VA
(editors). Cochrane Handbook for Systematic Reviews of Interventions
version 6.0 (updated July 2019). Cochrane, 2019. Available from
www.training.cochrane.org/handbook.
\item
Primo J, Escrig J. MetaSurv: Excel calculator for survival
meta-analyzes. 2008. Avaialble in:
http://www.redcaspe.org/herramientas/descargas/MetaSurv.xls
\item
Brookes ST, Whitely E, Egger M, Smith GD, Mulheran PA, Peters TJ.
Subgroup analyses in randomized trials: risks of subgroup-specific
analyses; power and sample size for the interaction test. J Clin
Epidemiol 2004; 57: 229--236.
\item
Assmann SF, Pocock SJ, Enos LE, Kasten LE.. Subgroup analysis and
other (mis) uses of baseline data in clinical trials. Lancet 2000;
355: 1064--1069.
\item
Feinstein AR. The problem of cogent subgroups: a clinicostatistical
tragedy. J Clin Epidemiol 1998; 51: 297--299.
\item
Moreira ED Jr, Stein Z, Susser E. Reporting on methods of subgroup
analysis in clinical trials: a survey of four scientific journals.
Braz J Med Biol Res. 2001;34(11):1441\selectlanguage{english}-1446.
\item
Zhang S, Liang F, Li W, Hu X. Subgroup Analyses in Reporting of Phase
III Clinical Trials in Solid Tumors.~\emph{J Clin Oncol} .
2015;33(15):1697\selectlanguage{english}-1702.
\item
Schulz KF, Altman DG, Moher D; CONSORT Group. CONSORT 2010 statement:
updated guidelines for reporting parallel group randomised trials. Int
J Surg. 2011;9(8):672\selectlanguage{english}-677.
\item
Mistry D, Patel S, Hee SW, Stallard N, Underwood M. Evaluating the
quality of subgroup analyses in randomized controlled trials of
therapist-delivered interventions for nonspecific low back pain: a
systematic review. Spine (Phila Pa 1976) 2014;39:618e29.
\item
Saragiotto BT, Maher CG, Moseley AM, et al. A systematic review
reveals that the credibility of subgroup claims in low back pain
trials was low. J Clin Epidemiol. 2016;79:3\selectlanguage{english}-9.
\end{enumerate}
\textbf{Hosted file}
\verb`Table 1.docx` available at \url{https://authorea.com/users/329737/articles/456662-subgroup-analysis-in-haematologic-malignancies-phase-iii-clinical-trials-a-systematic-review}
\textbf{Hosted file}
\verb`Table 2.docx` available at \url{https://authorea.com/users/329737/articles/456662-subgroup-analysis-in-haematologic-malignancies-phase-iii-clinical-trials-a-systematic-review}
\textbf{Hosted file}
\verb`Table 3.docx` available at \url{https://authorea.com/users/329737/articles/456662-subgroup-analysis-in-haematologic-malignancies-phase-iii-clinical-trials-a-systematic-review}
\textbf{Hosted file}
\verb`Table 4.docx` available at \url{https://authorea.com/users/329737/articles/456662-subgroup-analysis-in-haematologic-malignancies-phase-iii-clinical-trials-a-systematic-review}
\textbf{Hosted file}
\verb`Figure 1. Flow chart.pptx` available at \url{https://authorea.com/users/329737/articles/456662-subgroup-analysis-in-haematologic-malignancies-phase-iii-clinical-trials-a-systematic-review}
\selectlanguage{english}
\FloatBarrier
\end{document}