Selecting a study population - PowerPoint Presentation

Slide1

Selecting a study population for clinical trials

Dr Greg FoxUniversity of Sydney, Australia

Slide2

OverviewSelecting a suitable study population to answer our research question: where? who’s in? who’s out?Minimising biases in randomized trialsCase study: selecting a study population for a randomized study of LTBI treatment

Slide3

Part I:Selecting a suitable study population

Slide4

Specifying the study populationTotal Population

Target population

Accessible population

Population (of the world

)

Population

which

the

target

population

is hoped to representTarget populationGroup from whom study population is drawnAccessible population / Reference populationUsually defined by time & placePopulation sample / Study populationSelected subset of the study population

Study population

Slide5

5Who’s in and who’s out? Selecting a study population

Why select ? Rarely

possible to study all target

population

Criteria

for

selection

:

relevant to the

study

objectives; practicality (accessible); usually defined by time & placeSources of study population: Community, workplace, school, hospital;

Slide6

6Choosing the accessible populationCommon options are:Clinic basedPopulation basedHospital basedEach setting has

its problemsHow might biases affect your results?

Slide7

7Study population: select how? Requires enumeration of population

Random sampling:

Random: Each

person

(unit) has an

equal

chance

Stratified

random

: Random samples from specified sub groupsSystematic sampling: Use regular interval to sample (every 5th person)Cluster: random sample of groups (households) Convenience (‘grab’): easily accessed but not random

Slide8

How can you be sure your study sample is representative of the target population?8

Slide9

Selecting the study setting and study populationSelecting a suitable study settingSetting is often determined by available clinical links or existing collaborationsSome study questions may not be answerable in some settings (e.g. treatment for DR-LTBI in a low-prevalence setting, complex interventions in weak health care setting

)Single site or multi-site recruitment ?Defining a suitable study population to answer the research questionDefine study subjects precisely and unambiguously

Consider whether to choose broad vs narrow selection criteriaDesired target population (e.g. household vs all ‘close’ contacts)

Intended

generalisability

(e.g. adults

vs

all ages)

High risk groups of

particular clinical importance (

e.g. solid organ transplant recipients, PLHIV, children)Consider efficiency of recruitment (e.g. TST positive only vs all contacts)

Slide10

Case study 1 : MDR-TB prevention among household contactsWe will illustrate the issues relating to selecting study populations through the design of a clinical trial of levofloxacin as treatment for latent TB infection among contacts of MDR-TB patients

Slide11

Background to V-QUIN MDR trialTreating active MDR-TB is complex, toxic and costlyClose contacts of people with MDR-TB have a high risk of developing TB1Preventive therapy is routinely offered to infected contacts of people with drug-susceptible TB (isoniazid, rifampicin, isoniazid+rifapentine…)There is not yet evidence from RCTs to determine whether preventive therapy may be effective in MDR-TB contacts“There is an urgent need for a multicenter, randomized

, controlled trial”… of preventive therapy21 –

Kritski 1996; 2 - Schaaf et al, Paediatrics, 2002

Case

study

1: MDR-TB

prevention

among

household

contacts

Slide12

Research question“What is the effectiveness of levofloxacin given for 6 months, compared to placebo, in the prevention of active TB among close contacts of patients with MDR-TB who have latent tuberculosis infection?” Which settings and study populations could best answer this question?

Case study 1: MDR-TB prevention among household

contacts

Slide13

Study setting: requirementsAvailability of a sufficiently large target population for recruitment Capacity of health system to implement the studyEffective local engagement in research

Sufficient infrastructure to support technical aspects of the trial

Case study 1: MDR-TB prevention

among

household

contacts

Slide14

Choosing study eligibility criteriaIf criteria are too narrow:Unable to reach recruitment targetsResults not generalizable to other important patient populationsRecruitment process too complexIf criteria are too broad:May reduce average effect size (choosing some who may not actually benefit)May include individuals less likely to comply (reducing follow-up)Proportion of eligible subjects recruited may be lower, with potential for selection bias

Choosing a balance betweenInternal validity (ability to identify what is ‘true’ in the study population) and

Generalizability (external validity – an extension of the observed results to a larger population)

Slide15

Choosing inclusion criteriaInclusion criteria: exampleAny age [?]

Living in the same household as the index patient within the previous 3 months [

? why not ‘close’ contacts]TST result:

Tuberculin

skin test positive (a size of 10mm or greater at first reading

); OR

Any

TST size if known to be HIV positive or severely malnourished

; OR

New

TST conversion on the second

reading**defined as: (a) If the first test was <5mm: a size of 10mm or greater at second reading; OR (b) If the first test was 5-9mm: An increase of 6mm or greater at the second reading Case study 1: MDR-TB prevention among household contactsMinimize risk & enhance participant safety Select subjects likely to benefit from the interventionInclude subjects for whom the intervention may be considered in future policy and praxisUse standard definitions

Slide16

Choosing exclusion criteriaExclusion criteria: exampleA diagnosis of current active TB disease made during initial assessment [?how]

Known to be pregnantUnable to take oral medicationDocumented previous treatment for MDR-

TBDialysis-dependent chronic kidney diseaseetc

etc

.

Case

study

1: MDR-TB

prevention

among

household contactsConsider excluding:Those with clear, recognised contraindications to the interventionThose highly unlikely to comply with trial protocolThose in whom the intervention may not be effective, and/or ethically justifiableHowever, avoid unnecessary complexity and narrow criteria

Slide17

An aside: Including children in TB trialsBarriers to including

children in clinical trials for TB

includeLack

of

pharmacokinetic

and

pharmacodynamic

data

Lack

of

appropriate

drug formulationsConcerns by IRBs and cliniciansLack of funding (2% of total TB drug research funding, 25% of need)1Parental concernsConsensus statement on child TB trials:« Children

should

be

included

in

studies

at

the

early

phases of

drug

development

and

be

an

integral

part of the

clinical

development

plan,

rather

than

after

approval

 »

1

1

Nachman et al, Towards early inclusion of children in tuberculosis drugs trials: a consensus statement. Lancet ID 2015

Slide18

Case study 2: MDR prevention among liver transplant candidates Torre-Cisneros, CID 2015

Slide19

Study designMulti-centre, prospective, non-inferiority RCT comparing isonazid with levofloxacin in treatment of LTBI in patients eligible for liver transplantation500mg daily levofloxacin for 9 months vs 300mg isoniazid for 9 monthsTarget sample size 870 subjects to be randomized

Torre-Cisneros, CID 2015Case study 2:

MDR-TB prevention among liver

transplant candidates

Slide20

Torre-Cisneros, CID 2015Case study 2: MDR-TB prevention among liver transplant candidates

Inclusion criteria

On the waiting list for solid organ transplant within a network of Spanish hospitals

Aged ≥18 years

No evidence of active TB

One of:

Latent TB infection (TST ≥ 5mm or positive IGRA); or

History of ‘improperly treated TB’, or

Recent TB contact, or

Xray

changes consistent with old TB (apical nodules, calcified lymph nodes, pleural thickening)

Slide21

ResultsTorre-Cisneros, CID 2015

Slide22

Results33/33 LEV and 27/31 INH patients took steroids2/33 LEV patients (6%) and 7/18 INH patients (38.9%) developed severe hepatotoxicity6/33 LEV patients (18.2%) developed tenosynovitis, affecting knee in 5 and achilles tendon in 1, permanently discontinued in 5Study terminated early: “Due to high frequency and intensity of this unexpected side effect the trial was definitively stopped” (?pre-defined stopping rules)Torre-Cisneros, CID 2015

Slide23

Considerations in selecting study populationsStudy population affects interpretation of findingsGeneralizability of findings from Torre-Cisneros ? – determined by participant characteristicsEnsure the research question can be addressed within the intended study population Clearly specify inclusion and exclusion criteria in detailConsider generalizability of findingsConsider advantages vs disadvantages of multiple sites

Slide24

Part II:Minimising bias inclinical trials

Slide25

Minimising bias in clinical trials

Slide26

Bias in clinical trialsOur goal in conducting clinical trials is to obtain valid (‘truthful’) and precise (‘accurate’) estimates of the relationship between an intervention and outcomeThe main threats to validity are caused by bias: a tendency of an estimate to deviate in one direction from a true valueLeading to underestimation or overestimation of the effect of the interventionIt is impossible to know for sure whether a clinical study is biased, as we cannot know ‘the truth’

Slide27

Important forms of biasKey forms of bias in clinical trials include:Confounding (an ‘imbalance’ between groups that may be systematic, or by chance)*Selection bias (selection for an intervention is based upon the outcome)Information bias (measurement error in the exposure, outcome or covariates = ‘misclassification’**)How can we minimise

biases in clinical trials?*Confounders can be describe as variables that are: (a) Independently predictive of disease, within strata of exposure, (b) Associated with the exposure, (c) Not an intermediate in the causal pathway between exposure and outcome **

Misclassification bias for categorical variables

Slide28

RandomizationRandomization is the random allocation of an individual or group to an interventionEach individual theoretically has the same opportunity to be assigned to each of the study groupsIf done properly, randomization can ensure study groups are balanced - for both measured and unmeasured factors (confounders)

Randomization can satisfy assumptions required by statistical methods (e.g. independence between observations, no unmeasured confounding)

Slide29

Viera, Fam Med 2007

Slide30

Key components of adequate randomization:1. Truly random sequence generation

✓

✗Computer generated

Random numbers tables

Draw numbers from a hat

Toss a coin

Recruiting on alternate days to each group

Assigning random letter by last name

Hospital chart numbers

Day of the week

Randomization is good at achieving balance in measured and unmeasured covariates

Slide31

e.g.

Slide32

Covariate balance with randomizationSterling et al, NEJM 2011

Slide33

Key components of adequate randomization2. Allocation concealmentKeeps the group to which the study subjects are assigned unknown, or easily ascertained, up to the point that study participants are given the intervention.Aims to avoid bias in treatment allocation (selection)Inadequate allocation concealment can increase effect estimates by as much as 40%1

1Schultz et al. Empirical evidence of bias. JAMA 1995

Slide34

Blinding

Slide35

Key components of adequate randomization:3. BlindingBlinding all concerned to the intervention group can reduce ascertainment bias1. The best way to reduce ascertainment bias is to keep all participants and investigators in the study blinded as long as possible.

Blinding is not always possible, by nature of the intervention (e.g. surgery for MDR-TB – although sham surgery possible)1Ascertainment bias occurs when the results or conclusions of a

trial are systematically distorted by knowledge of which intervention each participant is receiving. 2Schultz et al. Empirical evidence of bias. JAMA 1995

Slide36

Levels of blindingSingle Blind: Subject is not aware of group allocationDouble Blind: Neither the subjects nor treating staff know group allocationTriple-Blind: Neither subjects, investigators nor data analysts and monitoring committee know group allocation

Slide37

Randomization methodsRandomization unitsSimple (e.g. coin toss, simple random numbers)Block (fixed or variable block sizes)Predicts against investigators predicting sequenceif block size is 4, there are 6 combinations: AABB, ABAB, BAAB, BABA, BBAA, and ABBA.

Stratified randomization (randomize within each stratum, to reduce variability in group comparison)Cluster randomization (groups of individuals, e.g. households) – we will discuss later1:1 randomization most often, but can use other ratiosn

Each approach has advantages and disadvantages (consult with your trial statistician early)Altman DG, Bland JM. How to randomize. BMJ 1999

Efird

J. Blocked randomization.

Int

J Environ Res Pub Health 2011

Slide38

Part III:Example of choosing appropriatesample size

Slide39

Sample size calculationsThe sample size is the expected number of participants required to adequately answer the research questionSample size is clinically and ethically importantToo few subjects: may prevent valid and precise determination of the treatment effect; may incur excessive cost and timeToo many subjects: may expose more individuals to riskBefore embarking upon the sample size calculation, you need to determine the planned primary outcome measure and measures of interest.What is the clinically important difference?

Slide40

Clinically important difference and confidence intervalsNo important effectInconclusive, needsfurther study

Clinically importantSmall but unimportant effect

At least a small effect. May be Important.

Needs further study

δ

Slide41

Example: sample size for V-QUIN trialSample size calculations require explicit decisions, including:Study design (e.g. superiority / non-inferiority; individual or cluster randomization; stratified effects required)Statistical analytic method planned (usually frequentist; could use Bayesian methods)Outcome measures (relative risk, risk difference etc)Thresholds for type I (e.g. α = 0.05) and type II (1- β = 0.8) errors

Minimum clinically important difference (δ) (prior slide)Precision of the estimates (standard deviations in each group, σ)

Expected event rates (e.g. TB incidence) based on other studiesExpected recruitment and drop-out rates

Slide42

Sample size example for binary outcomesSchlesselman (1974) - Sample size requirements in cohort and case control studies of disease, American Journal of Epidemiology 99, 381-384. Can use PS Power (which applies this formula)A good illustration of using this formula is given in Moore and Joseph, Lupus (1999) 8: 612-619

Slide43

Sample size exampleParameterValueZ(1-α/

2) for alpha = 0.051.96

Z(1-

β

)

for beta = 0.2

0.84

P

1

- proportion in

control arm

0.03P2 - Proportion in active intervention arm0.009n (in each group) prior to adjustment680Additional adjustments1.106

Design effect (clustering)

1.106

% loss to follow-up

10%

Fluoroquinolone

resistance

16.7%

Number

randomized in

both

groups

2006

Number contacts assuming 60% TST+

3344

Index patients assuming 2.1 index patients / contact

1592

Slide44

Sample size calculationsOther considerations:CostEvent rateFeasibilitySample size calculations are covered well in many places:Moore AD, Joseph L. Sample size considerations for superiority trials in systemic lupus erythematosus. Lupus, 1999.Joseph L. Bayesian and mixed Bayesian likelihood criteria for sample size determination. Stat Med 1997.Zou

KH, Normand S-L T. On determination of sample size in hierarchical binomial models. Stat Med 2001.

Slide45

Slide46

AcknowledgementsVietnam National Tuberculosis ProgramA/Prof Nguyen Viet NhungA/Prof Dinh Ngoc SyPham Ngoc Thach Hospital

An Giang, Binh Dinh, Ca Mau Can Tho

, Da Nang, Ha Noi, Tien Giang, Ho Chi Minh City, Vinh

Phuc

Tuberculosis Programs

Australian National Health and Medical Research Council (NHMRC)

Woolcock Institute of Medical Research, Sydney

Dr

Carol

Armour

, Director and staff

Woolcock Institute of Medical Research, VietnamDr Nguyen Thu AnhAnd, most importantly, the people of the participating provinces