INTRODUCTIO TO CLINICAL TRIALS

Question 1

WHAT IS A CLINICAL TRIAL?

Answer 1

• A SCIENTIFIC WAY OF TESTING A CLINICAL QUESTION
• CAN ASSESS: TRETAMENTS, SUPPORTIVE CARE, DEVICES, SCREEING PROGRAMMES, INFORMATION, DIAGNOSTIC TESTS, BIOMARKERS….. –> NEED TO COMPARE WITH CONTROL GROUP
• A FAIR, CONTROLLED (COMPARATIVE), REPRODUCIBLE EXPERIMENTAL STUDY

Question 2

(ONE OF THE) FIRST REPORTED CLINICAL TRIAL(S)

Answer 2

• CONDUCTED BY JAMES LIND
• 1747: TOOK 12 PATIENTS WITH SCURVY AND SPLIT THEM INTO 6 GROUPS OF 2 PATIENTS: CYDER, ELIXER VITRIOL, VINEGAR, SEA WATER, ORGANES & LEMONS, NUTMEG
• THE PATIENTS WERE VERY SIMILAR IN EVERY OTHER WAY BUT THE FOOD THEY’VE BEEN CONSUMING
• THE MOST SUDDEN AND VISIBLE GOOD EFFECTS WERE PERCEIVED FROM THE ORANGES AND LEMONS GROUP (SCURVY = VIT C DEFICIT)

Question 3

GOLD THERAPY FOR TUBERCULOSIS

Answer 3

• EXAMPLE OF AN EARLY CLINICAL TRIAL (LESSONS FORM HSITORY)
• POPULAR IN 1920s AND 1930s
• PETERS AND SHORT (1935) FOUND THAT CONTROLS HAD BETTER IMPROVEMENT

‘THE EXAMINATION OF OUR STATISTICS HAS BEEN A PAINFUL SHOCK, WE WERE CONVINCED TRETAMENT WAS VALUABLE; MANY OF THE CASES DID EXTREMELY WELL. BUT ONE TENDS TO FORGET MANY CASES PREVIOUSLY DID EXTREMELY WELL….’

Question 4

VACCINE THERAPY CLINICAL TRIAL (LESSONS FROM HISTORY)

Answer 4

• VACCINE THERAPY WAS A POPULAR FORM OF TREATMENT IN EARLY 1990s FOR MANY CHRONIC INFLAMMATORY DISEASES
• HOWEVER, CONTROLLED TRIALS WERE UNABLE TO DEMONSTRATE VACCINE WAS BETTER THAN PLACEBO (E.G. RHEUMATOID ARTHRITIS, SIDEL AND ADAMS, 1940)

Question 5

FUNDAMENTAL PRINCIPLES IN COMPARING TREATMENT GROUPS IN CLINICAL TRIALS

Answer 5

DIFFERENCES IN OUTCOMES BETWEEN GROUPS MAY BE DUE TO:

• TRUE TREATMENT EFFECT
• BIAS
• CHANCE
• GROUPS MUST BE ALIKE IN ALL IMPORTANT ASPECTS EXCEPT FOR THE TREATMENTS UNDER EVALUATION (POSIBEL BIAS)
• HAVE TO HAVE A LARGE ENOUGH SAMPLE (LIMIT CHANCE IMBALANCES)

Question 6

DIFFERENCES IN OUTCOMES BETWEEN GROUPS IN CLINICAL TRIALS MAY BE DUE TO:

Answer 6

• TRUE TREATMENT EFFECT
• BIAS
• CHANCE

Question 7

THE BEST WAY TO CREATE A CONTROL GROUP SIMILAR TO THE TREATMENT GROUP IN ALL RESPECTS (KNOWN AND UNKNOWN):

Answer 7

RANDOMISTION!!!!!!!!!!!!!

Question 8

DESCRIBE THE RANDOMISED CLINICAL TRIALS

Answer 8

• THERAPIES ALLOCATED BY A CHANCE MECHANISM
• NEITHER PATIENT NOR PHYSICIAN KNOW IN ADVANCE WHICH THERAPY WILL BE ASSIGNED
• ELIMINATES SELECTION BIAS
• GIVES A BETTER CHANCE OF BALANCING PROGNOSTIC FACTORS BETWEEN GROUPS
• IS A FUNDAMENTAL BASIS FOR VALIDITY OF STATISTICAL TESTS

Question 9

ADVANTAGES OF RANDOMISTION IN CLINICLA TRIALS

Answer 9

• ELIMINATES SELECTION BIAS
• GIVES A BETTER CHANCE OF BALANCING PROGNOSTIC FACTORS BETWEEN GROUPS
• IS A FUNDAMENTAL BASIS FOR VALIDITY OF STATISTICAL TESTS

Question 10

1ST PROPERLY DESIGNED RANDOMLY CONTROLLED TRIAL

Answer 10

• SIR AUSTIN BRADFORD HILL
• 1948
• ASSESSING STREPTOMYCIN FOR TUBERCULOSIS IN 107 PATIENTS

CONTROL: BED REST ALONE (52 PATIENTS)
TRETAMENT: STRETPTOMYCIN (55 PATIENTS)

• SHOWED THAT TRETAMENT LEADS TO X RAY IMPROVEMETN AND REDUCED NUMBER OF DEATHS

Question 11

CONTROLLED CLINICAL TRIAL AND RANDOMISED CLINICAL TRIAL DEFINITIONS

Answer 11

CCT IS A PROSPECTIVE!!! STUDY COMPARING EFFECT (S) AND VALUE OF AN INTERVENTION AGAINST A CONTROL IN HUMAN SUBJECTS

RCT IS CONTROLLED TRIAL WHERE THE THERAPIES ARE ALLOCATED BY A CHNCE MECHANISM

Question 12

UNCONTROLLED CLINICAL TRIAL?

Answer 12

INVOLVES NO CONTROL GROUP

Question 13

CLASSIFICATION OF CLINICAL TRIALS BY DESIGNS (FROM MOST RELIABLE TO LESS RELIABLE)

Answer 13

1) RANDOMISED CONTROLLED
2) CURRENT NON-RANDOMISED CONTROLLED
3) HISTORICAL CONTROLS
4) UNCONTROLLED

Question 14

STATISTICAL INFERENCE?

Answer 14

We are often interested in understanding something that is unobserved in the wider population, this could be the average blood pressure in a population of pregnant women for example, or the true effect of a drug on pregnancy rate, or whether a new treatment perform better or worse than the standard treatment. In these situations we have to recognise that almost always we observe only one sample or do one experiment. If we took another sample or did another experiment, then the result would almost certainly vary. This means that there is uncertainty in our result, if we took another sample or did another experiment and based our conclusion solely on the observed sample data, we may even end up drawing a different conclusion!
The purpose of statistical inference is to estimate this sample to sample variation or uncertainty. Understanding how much our results may differ if we did the study again, or how uncertain our findings are, allows us to take this uncertainty into account when drawing conclusions. It allows us to provide a plausible range of values for the true value of something in the population, such as the mean, or size of an effect, and it allows us to make statements about whether our study provides evidence to reject a hypothesis.

Question 15

THE PLACEBO EFFECT

Answer 15

EVEN IF THE THERAPY IS IRRELEVANT TO THE PATIENT’S CONDITION, THE PATIENT’S ATTITUDE TO HIS OR HER ILLNES, AND INDEED TH ILLNESS ITSELF, MAY BE IMPROVED BY A FEELING THAT SOMETHING IS BEING DONE ABOUT IT

Question 16

‘BLIND’ TRIALS?

Answer 16

• BLINDING ‘STRENGTHENS’ RANDOMISTAION

SINGLE BLIND; ONE OF THE PATIENT, CLINICAN, ASSESSOR DOES NOT KNOW THE TREATMENT ALLOCATION (USUALLY PATIENT)

DOUBLE BLIND; 2 OF PATIENT, CLINICAN, ASSESSOR DO NOT KNOW THE TREATMENT ALLOCATION (USUALLY PATIENT + CLINICIAN/ASSESSOR)

(IF ALL 3 ARE BLIND. IT’S A TRIPLE BLIND TRIAL)

• AIMS TO REMOVE DIFFERENTIAL PLACEBO EFFECT THAT COULD BIAS COMPARIES BETWEEN THERAPIES
• NOT ALWAYS POSSIBLE TO BLIND CERTAIN STAKEHOLDERS DUE TO THE NATURE OF THE TRETAMENT (E.G. THE DOCTOR CANNOT BE BLINDED IF THEY NEED TO PERFORM A SURGERY)

Question 17

WHAT PROPERIES DOES PLACEBO HAVE TO HAVE?

Answer 17

HAS TO BE IDENTICAL IN TASE, APPEREANCE, TEXTURE, DOSAGE REGIME ETC TO THE REAL TREATMENT

Question 18

DESCRIBE HOW CLINICAL TRIALS ARE CONDUCTED

Answer 18

1) ELIGIBLE PATIENT PRESENTS
2) CONSENT
3) RANDOMISATION INTO NEW AND CONTROL
4) GIVE TREATMENT (REGULAR FOLLOW UP TO MONITOR FOR NON COMPLIANCE)
5) ASSESSMENT (COMPARE THE OUTCOME FAIRLY TO SEE IF THERE IS DIFFERENCE AND IS IT STATISTICALLY SIGNFICANT AND CLINICALLY IMPORTANT + CAN IT REALLY BE ATTRIBUTED TO THE INTERVENTION IN THE TRIAL)

Question 19

WHAT CAN BE CONSIDERE IN CLINICAL TRIALS WHEN ASSESSING WHETHER THE NEW TRETAMENT IS BETTER THAN THE STANDARD? DIFFERENCE IN TYPES OF TRIALS BASED ON THIS?

Answer 19

2 DIFFERENT INTERPRETTAIONS:

IS THE PHYSIOLOGIAL ACTION OF THE NEW TREATMENT BETTER THAN STANDARD –> ‘EXPLANATORY TRIAL’/’AS-TREATED ANALYSIS’ –> ONLY INCLUDES THOSE PARTICIPANTS THAT ACCURATELY ADHERED TO TRIAL INSTRUCTIONS —> LOSES SOME EFFECTS OF RANOMISATION BECAUSE COMPLIERS ARE DIFFERENT THAN NON COMPLIERS

IS THE NEW TREATMENT BETTER THAN THE STANDARD TREATMENT IN ROUTINE CLINICAL PRACTICE –> PRAGMATIC TRIAL/’INTENTION TO TREAT’ ANALYSIS —> KEEPS EVERYONE IN THE TRIAL REGARDLESS OF COMPLIANCE (AS LONG AS NON COMPLIERS MAKE UP <10%)

Question 20

‘AS TREATED’ VS ‘INTENTION TO TREAT’ ANALYSIS

Answer 20

AS TREATED: TENDS TO GIVE LARGER SIZES OF EFFECT (KEEPS ONLY PARTICIPANTS WHO ADHERED TO TREATMENT)

INTENTION TO TREAT: TENDS TO GIVE SMALLE EFFECT SIZES AND REFLECTS EFFECT IN CLINICAL PRACTICE

• DEFINITIVE CLINICAL TRIALS SHOULD NORMALLY BE ANALYSED ON AN ‘INTENTION-TO-TREAT’ BASIS

Question 21

KEY ETHICAL DILEMMAS IN CLINICAL TRIALS ND HOW CNA THEY BE SURPASSED

Answer 21

• CLINICIANS SHOULD PROVIDE BTS TREATMENT FOR EACH PATIENT, BUT SCIENTIFIC INTEGRITY REQUIRES TRETAMENT CHOSEN RANDOMLY

CLINICAL EQUIPOISE: REASONABLE UNCERTAINTY ABOUT WHICH TREATMENT (INCLUDING NON-TREATMENT) IS BETTER –> RANDOMISATION DOES NOT DENY ANY PATIENT THE BS TREATMENT, BECAUSE THE BEST TREATMENT IS NOT KNOWN/ IT IS BEING INVESTIGATED

• INFORMED CONSENT IS KEY (EXPLAIN: THAT SOMEONE IS INVITED FOR A TRIAL, WHAT ARE THE ALTERNATIVE TRETAMENTS AND SIDE EFFECTS, RANDOM ALLOCATION, RIGHT TO WITHDRAW + GIVE INFO VERBALY AND IN WRITING WITH ‘COOLING OFF’ TIME)

Question 22

CLINICAL EQUIPOISE?

Answer 22

• ETHICAL RATIONALE FOR RNADOMISATION IN CLINICAL TRIALS

REASONABLE UNCERTAINTY ABOUT WHICH TREATMENT (INCLUDING NON-TREATMENT) IS BETTER –> RANDOMISATION DOES NOT DENY ANY PATIENT THE BS TREATMENT, BECAUSE THE BEST TREATMENT IS NOT KNOWN/ IT IS BEING INVESTIGATED

• THE ONLY QUALIFICTAION IS UNCERTAINTY

Question 23

THE ZELEN METHOD

Answer 23

• ALTERNATIVE APPROACH/STEPS TO STANDARD RANDOMISATION

STANDARD: ELIGIBILITY - CONSENT - SOME DROP FROM THE TRIAL SOME ARE RANDOMISED - GIVE TREATMENT A AND B

ZELEN METHOD: ELIGIBILITY - RANDOMISE TO ‘NO CONSENT’ AND ‘SEEK CONSENT’ - NO CONSENT RECEIVES STANDARD TREATMENT - IF ‘SEEK CONSENT’ SAYS YES, GIVE THE NEW TRETMENT, IF THEY SAY NO, GIVE THE SNADARD TREATMENT

• ALLOWS EASIER ALLOCATION AND RECRUITMENT BUT CAN’T GET BY ETHICS COMMITTEES