CLINICAL TRIALSStudi Intervensi/

Experimental

Perlakuan/intervention is assigned by the investigator

OUTLINE• Definition, terminology, type, and phase of

Trial• Phase III of Clinical Trial

– Sample– Randomization– Blinding– Analysis - Comparability, Confounding,

Outcome– Challenge - Attrition– Ethical Aspect

DEFINITION, TERMINOLOGY TYPE, AND

PHASE OF TRIAL

Did investigatorassign exposures?

Experimental study

Random allocation?

Randomisedcontrolled

trial

Non-randomisedcontrolled

trial

Yes

Yes No

1. True (pure) exp.

2. Quasi (semu) exp.

1. Clinical Trial

2. Community Trial

1. Parallel

2. Cross Over

Perlakuan (intervention)

Outcome(efficacy)

Prospective

Study group

Control group

X

Y

• Cure rate

• Disease IR

• Mortality rate• Length of treatment

• Accuracy of diagnostic tools• Etc, etc, etc

• Medication

• Vaccine• New technique of surgery• New treatment protocol• New diagnostic tools• Patient nursing method

• Public health program

Study group

(X)

Controlgroup

(Y)

• Placebo (substance with out effect)• Standard medication• Standard protocol

• With out intervention

• Etc

• Etc

EXPERIMENTAL STUDY

1. True (murni)

2. Quasi (semu)

• All variables (independent and confounding) can be manipulated by the investigator• Only for animal or plant

• Not all confounding variable can be manipulated by the investigator• NOT to study disease risk factors. •To study SOMETHING BENEFITS TO HUMAN (treatment efficacy, prevention and diagnostic tools

CLINICAL TRIALS (Individual Intervention)

1000 under five children

Vaccine IR pneumonia

1000 under five children

Placebo IR pneumonia

Village AEducation

IR pneumonia

Village BClean Water Provision

IR pneumonia

COMMUNITY TRIALS (Group Intervention)

PARALLEL & CROSS OVER DESIGN

IR Pneumonia

PARALLELIR Pneumonia

IR Hypertension

IR Hypertension

CROSS OVER

• Purpose: to increase comparability between study & control group

• Only for chronic diseases (hypertension, diabetes mellitus, etc)

Aims: to know adverse effects of drugs/vaccine/others

Phase of clinical trialsA. Pre-clinical study animal study (laboratory experiment)

B. Clinical trials:human

a) Phase I : among 20-50 volunteers Aims: to know adverse effects among human

b) Phase II : among 100 - 200 volunteers (even more ~ 2000) Aims: to know efficacy of drugs/vaccine/others

c) Phase III : among bigger sample (1000 – 2000) (even more ~ 10.000) using comparison groups, randomization

d) Phase IV: post market study. Aims: to know adverse effects and efficacy of drugs/others following long period of use by large number of people

Clinical trials Phase III

REFERENCE POPULATION

SAMPLE POPULATION

EXPERIMENTAL POP. (SAMPLE)

WILLING TO PARTICIPATE

NOT WILLING TO PARTICIPATE

INFORMEDCONCENT

INTERVENTIONGROUP

COMPARISONGROUP (RECEIVE PLACEBO)

INCLUSION(EXCLUSIO

N)

RANDOMIZATION

OUTCOMEKNOWN

OUTCOMEUNKNOWN

OUTCOMEKNOWN

OUTCOMEUNKNOWN

ATTRITION ATTRITION

Flow Chart Phase III

Reference Population Target-population, study population

A population (healthy or sick people) where the study finding will be generalized

It may general or less general

All rheumatoid arthritis patients (Indonesia)

Broad

Rheumatoid arthritis patients > 40 years of age (Indonesia)

Narrow

Rheumatoid arthritis patients > 40 years who have sick >= 1 years (Indonesia)

Narrower

SAMPLE - CRITERIAS OF SAMPLE SELECTION1. Similar demographic characteristics with reference

population2. Feasibility of the trials (e.g: distance of subject

location and site study, geographical distribution of population at the study site, access of patients medical records, etc)

3. Incidence rate of the disease being studied must high enough to get sufficient number of sample.

4. Number of population at the study site must high enough to get sufficient number of sample

Multi Center study: 1. Required sample will be achieved faster shorter study period

2. Demographic characteristics of the sample will be more representative to the reference population

HOW TO DETERMINE NUMBER OF SAMPLE (SUBJECTS) IN A CLINICAL TRIALS?

1. At the planning stage of the study using formula

If dependent variable is: dichotomy (ex: cured vs. not cured)n = 1/1-f [ 2 (Z + Z) x p (100 – p)] ( pc – pt)

If the dependent variable is: interval: mean difference

n = 1/1-f [ 2 (Z + Z) x sdc2]

(Xc – Xt)

2. Sequential trials: Number of subjects (sample) is determined during the study. The data is analyzed regularly and stopped when significant different is found. If there is no significant different, the study will also be stopped when number of sample using formula is achieved.

RANDOMIZATIONRandom allocation of subjects to either intervention arm (group) or control arm (group)

PURPOSES

1. To allows ‘comparable characteristics’ between a study and control group -- To eliminate/minimize bias (BY DESIGN)

2. To allows ‘fair’ allocation between subject in to the study or control group -- Ethics and generalisation

HOW TO RANDOM?

Random the Subjects or the Intervention

Simple, Blocked, Stratified Randomization

Example of randomization

BLIND(ING)To blind all parties involved with the study regarding the allocation of intervention and the measurement of outcome

PURPOSE : To minimize the bias

INDEPENDENT MONITOR

Process, Data, Ethics, Result, etc

BLIND(ING)

3. Triple Blind:If first, second and third parties are blind on whether the subjects receiving intervention or placebo/conventional drug.

1. Single Blind:If only subject (second party) do not know whether they got intervention or placebo/conventional drug, but investigator (first party) know

2. Double Blind:Both, first party and second party do not know which subjects got drug and which subjects got placebo. All identity of drugs and placebo were taken out, all has similar color, etc. It only has code (that can be known by PI when needed)

DATA (OUTCOME) ANALYSIS

COMPARABILITY

CONFOUNDING

OUTCOME

Why researcher do the RANDOMIZATION?

Are there other variables affected the

outcome?

Is the intervention effective?

TRIALS TO ASSESS EFFICACY OF CEFAZOLIN PERIOPERATIVE IN PREVENTING POST OP. INFECTION

Patients

Perlakuan (intervention

)

Cefazolin

Placebo

Outcome

IR post op infection

IR post op infection

Prospective

Patients

Blood loss Operation

Type

Age, Sex,Body weight,Etc

Menopause status

Confounding factors

Comparable between

Cefazolin and placebo patients?

For analytical study (using control or comparison group)

1. Data presented the comparability/similarity of characteristics of the study and control group, particularly for confounding variables

How? statistical test must be done (X2, t-test, ANOVA, etc)If the difference is statistically significant (P value <0.05), it means not similar (not comparable) there is an effect of confounding variable (bias of the study)

Intervention Control P Value

Sex (Female) 30% 34% 0,04

Age (Mean) 28,3 29,0 0,21

Education (High) 45% 56% 0,00

For analytical study (using control or comparison group)

2. Data presented the Outcome (always considering confounding variables)

How?

Calculate specific/adjusted ratio (BY ANALYSIS) if variable is categorical

Calculate difference if variable is interval

Intervention

Outcome Statistic Test Outcome Analysis

Fe tablet Anemia/Not Anemia ………………. Adjusted

Ratio

Food supplement

Increasing Weight ……………….. …………………

Education support

Passed/Not passed Chi-square …………………

TRIALS TO ASSESS EFFICACY OF CEFAZOLIN PERIOPERATIVE IN PREVENTING POST OP. INFECTION

Patients

Perlakuan (intervention

)

Cefazolin

Placebo

Outcome

IR post op infection

IR post op infection

Prospective

Patients

Blood loss Operation

Type

Age, Sex,Body weight,Etc

Menopause status

Confounding factors

Comparable between

Cefazolin and placebo patients?

MULTIFACTORIAL!!!

Vaginal Abdominal

Placebo Cefazolin Placebo Cefazolin

Total subjects 42 44 223 206

Mean age 42 thn 41 thn 41 thn 41 thn

Mean blood loss during surgery

463 cc 429 cc 395 cc 350 cc

Mean duration of surgery

1,1 jam 1,1 jam 1,2 jam 1,2 jam

Proportion of pre menopause

79% 86% 91% 90%

Proportion obese 43% 45% 49% 44%

Number of subjects who get infection after hysterectomy

9 1 47 29

TRIALS TO ASSESS EFFICACY OF CEFAZOLIN PERIOPERATIVE IN PREVENTING POST OP. INFECTION

Home care

Post heart attack patients

(325)

Prospective(3 years)

Hospital care

Post heart attack

patients(340)

Mortality rate(27%)

Mortality rate(18%)

Stat test

Study to Compare Hospital and Home Care to Decrease Mortality among Infark Myocard Patients

Study to Compare Hospital and Home Care to Decrease Mortality among Infark Myocard Patients

Hospitalcare (27%)

Homecare (18%)

Male 65% 55%

Female 35% 45%

Mean age 65,6 yrs 63,2 yrs

Mean body weight 61,2 kg 58,5 kg

Confounding variables

Mean heart attack 1,7 x 2,1 x

IF SIGNIFICANT DIFFERENT IS FOUND critical interpretation must be done:

1. Is there sampling error? (sample is not representative, external validity is not good)

2. Is there bias (systematic error)? internal validity is not good•Study and control group is not similar (not comparable), not

blind design• Disease measurement is incorrect, etc, etc

3. If sample is representative and no possible bias the different may due to the intervention

ATTRITION (LOST TO FOLLOW-UP)

• It must be anticipated at the planning stage of the study number of sample?

• If high generalization of study finding will be limited (external validity will be low)

•Retention Strategy? Maximum efforts if subject do not come home visit or telephone call must be done

ETHICAL ASPECT

• Side Effect (harm) of Intervention?

•Placebo?

• Honesty? Confidentiality?

• Etc…….