SIGN

Methodology Checklist 2: Controlled Trials

Study identification (Include author, title, year of publication, journal title, pages) Guideline topic: Before completing this checklist consider: !" #s the paper a randomised controlled trial or a controlled clinical trial$ #f in dou%t check the study design algorithm availa%le from S#GN and make sure you have the correct checklist" #f it is a controlled clinical trial &uestions !"' !"( and !") are not relevant and the study cannot %e rated higher than !* '" #s the paper relevant to key &uestion$ +nalyse using ,#-. (,atient or ,opulation #ntervention -omparison .utcome)" #/ N. R010-2 (give reason %elow)" #/ 30S complete the checklist" Reason for re4ection: !" ,aper not relevant to key &uestion '" .ther reason (please specify): SECTION 1: INTERNAL VALIDITY In a well conducted RCT study !"! !"' !"( !") !"7 !"9 !": !"; 2he study addresses an appropriate and clearly focused &uestion"i 2he assignment of su%4ects to treatment groups is randomised"ii +n ade&uate concealment method is used"iii Su%4ects and investigators are kept 6%lind5 a%out treatment allocation"iv 2he treatment and control groups are similar at the start of the trial"v 2he only difference %etween groups is the treatment under investigation"vi +ll relevant outcomes are measured in a standard valid and relia%le way"vii <hat percentage of the individuals or clusters recruited into each treatment arm of the study dropped out %efore the study was completed$viii +ll the su%4ects are analysed in the groups to which they were randomly allocated (often referred to as intention to treat analysis)"ix <here the study is carried out at more than one site results are compara%le for all sites"@ 3es -an5t say 3es -an5t say No >oes not apply No >oes not apply Does this study do it? 3es -an5t say 3es -an5t say 3es -an5t say 3es -an5t say 3es -an5t say 8 3es -an5t say 3es -an5t say No No No No No No No Key Question No: Reviewer:

!"=

!"!?

/ile name : -hecklist ' A -ontrolled 2rials ,roduced %y: -arolyn Sleith

Bersion '"? ,age ! of (

';C?7C'?!' Review date: None

SECTION : OVERALL ASSESS!ENT O" T#E ST$DY '"! Dow well was the study done to minimise %ias$ Code as follows:xi Digh &uality (**) +ccepta%le (*) Enaccepta%le A re4ect ? '"' 2aking into account clinical considerations your evaluation of the methodology used and the statistical power of the study are you certain that the overall effect is due to the study intervention$ +re the results of this study directly applica%le to the patient group targeted %y this guideline$ Notes% Summarise the authors5 conclusions" +dd any comments on your own assessment of the study and the e@tent to which it answers your &uestion and mention any areas of uncertainty raised a%ove"

'"( '")

/ile name : -hecklist ' A -ontrolled 2rials ,roduced %y: -arolyn Sleith

Bersion '"? ,age ' of (

';C?7C'?!' Review date: None

Enless a clear and well defined &uestion is specified it will %e difficult to assess how well the study has met its o%4ectives or how relevant it is to the &uestion you are trying to answer on the %asis of its conclusions"
ii

Random allocation of patients to receive one or other of the treatments under investigation or to receive either treatment or place%o is fundamental to this type of study"
iii

+llocation concealment refers to the process used to ensure that researchers are unaware which group patients are %eing allocated to at the time they enter the study" Research has shown that where allocation concealment is inade&uate investigators can overestimate the effect of interventions %y up to )?F"
iv

Glinding refers to the process where%y people are kept unaware of which treatment an individual patient has %een receiving when they are assessing the outcome for that patient" #t can %e carried out up to three levels" Single %linding is where patients are unaware of which treatment they are receiving" #n dou%le %lind studies neither the clinician nor the patient knows which treatment is %eing given" #n very rare cases studies may %e triple %linded where neither patients clinicians nor those conducting the analysis are aware of which patients received which treatment" 2he higher the level of %linding the lower the risk of %ias in the study"
v

,atients selected for inclusion in a trial must %e as similar as possi%le" 2he study should report any significant differences in the composition of the study groups in relation to gender mi@ age stage of disease (if appropriate) social %ackground ethnic origin or coHmor%id conditions" 2hese factors may %e covered %y inclusion and e@clusion criteria rather than %eing reported directly" /ailure to address this &uestion or the use of inappropriate groups should lead to the study %eing downgraded"
vi

#f some patients received additional treatment even if of a minor nature or consisting of advice and counselling rather than a physical intervention this treatment is a potential confounding factor that may invalidate the results" If &ro'(s )ere not treated e*'all+, t-e st'd+ s-o'ld .e re/ected 'nless no ot-er e0idence is a0aila.le% #f the study is used as evidence it should %e treated with caution"
vii

2he primary outcome measures used should %e clearly stated in the study" If t-e o'tcome meas'res are not stated, or t-e st'd+ .ases its main concl'sions on secondar+ o'tcomes, t-e st'd+ s-o'ld .e re/ected% <here outcome measures re&uire any degree of su%4ectivity some evidence should %e provided that the measures used are relia%le and have %een validated prior to their use in the study"
viii

2he num%er of patients that drop out of a study should give concern if the num%er is very high" -onventionally a '?F drop out rate is regarded as accepta%le %ut this may vary" Some regard should %e paid to why patients dropped out as well as how many" #t should %e noted that the drop out rate may %e e@pected to %e higher in studies conducted over a long period of time" + higher drop out rate will normally lead to downgrading rather than re4ection of a study"
i@

#n practice it is rarely the case that all patients allocated to the intervention group receive the intervention throughout the trial or that all those in the comparison group do not" ,atients may refuse treatment or contraHindications arise that lead them to %e switched to the other group" #f the compara%ility of groups through randomisation is to %e maintained however patient outcomes must %e analysed according to the group to which they were originally allocated irrespective of the treatment they actually received" (2his is known as intention to treat analysis") #f it is clear that analysis was not on an intention to treat %asis the study may %e re4ected" #f there is little other evidence availa%le the study may %e included %ut should %e evaluated as if it were a nonHrandomised cohort study"
@

#n multiHsite studies confidence in the results should %e increased if it can %e shown that similar results were o%tained at the different participating centres"
@i

Rate the overall methodological &uality of the study using the following as a guide: #i&- *'alit+ (**): Ia4ority of criteria met" Jittle or no risk of %ias" Results unlikely to %e changed %y further research" Acce(ta.le (*): Iost criteria met" Some flaws in the study with an associated risk of %ias -onclusions may change in the light of further studies" Lo) *'alit+ (?): 0ither most criteria not met or significant flaws relating to key aspects of study design" -onclusions likely to change in the light of further studies"