Clinical Trials

(testing of hypothesis) Study designs Study types Descriptive (formulation of hypothesis) Individual based Case studies Case series Cross-sectional surveys Population based co-relational Analytical (testing of hypothesis) Observational Case-control cohort interventional RCT’s (III) Quazi experimental

Randomized Control Trials In public health and clinical practice our objective is to modify natural history of disease i.e. to delay death or disability and to promote health Challenge is to select best available preventive or therapeutic measure. RCT is the ideal design to evaluate the effectiveness and efficiency of these measures

History of RCT Unintentional trial be Ambroise Pare (1537) Treatment of battle wounds with Hot oil ointment made of rose oil, turpentine oil and egg yolk Planned trial by James Lind (1747) Treatment of scurvy on board “salisbury” 12 patients 2 had apple drink, 2 had elixir vitriol, 2 had vinegar, 2 had lemon, 2 had nutmeg and 2 had to drink sea water

Learning Objectives Students should learn at the end of this lecture What is a clinical trial What are the phases of clinical trials How many types of clinical trials are available Design of a clinical trial Protocol Informed consent Sample size

What is clinical trial? It is an experimental epidemiological method. Subjects are randomly allocated into groups (usually study and control group) to receive or not to receive an experimental, preventive or therapeutic procedure, maneuver, or intervention.

Why they are done? As they provide better evidence of the effect or the outcome that cannot be obtained with any observational method. The variation is minimized and bias is controlled, hence more valid and their results speak truth.

Why they are done? 3. Enjoying maximum confidence just like any other scientific laboratory experiment 4. Scientifically most rigorous, Fastest and safest method of testing hypothesis to find new treatments that work in patients and ways to improve quality of health

When They Are Done When the margin of the expected benefit or the outcome of an intervention is doubtful. After the laboratory and animal studies yield the most promising results of the intervention, these results are tested by clinical trials

Phases Of Clinical Trials Phase -I trials Phase -II trials Phase -III trials Phase -IV trials

Phase-l Trials After laboratory and animal testing, first time human testing In a small group of 2 – 10 usually terminally ill patients Purpose is To evaluate Maximum Tolerated Dose (MTD) To determine a safe dosage range Rectify side effects

Phase-ll Trials Tested in large group of 10-30 people Purpose is to further evaluate Effectiveness with optimal dose Safety

Phase – lll Trials (RCT) Still large group 100-3000 people are tested Purpose is to confirm Its effectiveness Monitor side effects Compare with commonly used treatments To gather information regarding safe use

Phase-lV Trials Post marketing studies To know about Drug risks after prolong use Benefits Optimal use

What are its types? Treatment trials Prevention trials Diagnostic trials Screening trials Quality of life trials

1) Treatment Trials They test New treatments New combination of drugs New surgical techniques New Radiation therapies

2) Prevention Trials They try to find better ways to prevent disease in people and to prevent disease recurrence using Vaccines Vitamins Minerals Life style changes Preventive Medicines Preventive med. Going to area which is endemic with malaria, and using antimalarials for prevention.

3) Diagnostic Trials To find better tests for diagnosis of a disease

4) Screening Trials To find out the best way to detect certain diseases or conditions

5) Life Style Trials Also called Supportive care trials Often employed for the chronically ill patients or old homes. They explore the ways to improve comfort To improve the quality of life

How is it done? Design of Clinical Trial Selecting the reference population Selecting the experimental population . Selecting the study population (Participants) Random allocation into Intervention group Comparison group Apply intervention No intervention (placebo/standard treatment) Uniform assessment of outcomes Improved Not improved Improved Not improved

Randomized control study Design Study Population Randomization Study Group (Experimental) Control Group (Comparison) Base line Data Collection Base line Data Collection Intervention Compare Data Collection Data Collection 22

Quazi-Experimental study Design Study Population Study Group (Experimental) Control Group (Comparison) Base line Data Collection Base line Data Collection Intervention Compare Data Collection Data Collection 23

Some further Designs Non-RCT; groups allocated without Randomization Cross over designs; same person in intervention group and then shifted to control group and vice versa Usually for analgesics For non-fatal diseases

Some further Designs Historical controls; single group intervention and it is compared with controls of the past (records/other studies) Group allocation; groups are allocated instead of individuals (school, villages) Hybrid designs; combination of 2 or more than 2 designs

PROTOCOL (STUDY PLAN) Study plan is designed to safeguard the health of the participants to answer the specific research question

PROTOCOL Written agreement between researcher, participant and scientific community; should be developed before enrolment and once implemented should not change. It has to be explained in detail, to all the participants, particularly regarding benefits and risks. It should be submitted to the ethical committee for prior approval before commencing the study.

PROTOCOL It describes what types of people can participate, the schedule of tests, procedures, medications, dosages The duration of study

EXPLAINING THE BENEFITS that the participants play active role in their own health care They gain access to the new treatments before they are widely available They obtain expert medical care Help others by contributing to research

EXPLAINING THE RISKS TO PARTICIPANTS Unpleasant, serious or even life threatening side effects Failure of treatment Time waste for the participants

Protocol Outline of the protocol A:- introduction of the study B:- Objectives Primary question and its response Secondary question and its response Adverse effects C:- Design of the study Study population (inclusion and exclusion criteria) Sample size and enrolment of the subjects Informed consent

Protocol 4:- Intervention 5:- Follow up 6:- Termination policy Assessment of eligibility Base line examination Intervention allocation (Randomization) 4:- Intervention Treatment schedule Measure of compliance 5:- Follow up 6:- Termination policy

INFORMED CONSENT Informed consent document will be obtained from the participants in the study population after explaining them The purpose, Duration of the study, Required procedures, Expectations, Risks and benefits, Adverse effects of the trial if any, Participants’ rights

INFORMED CONSENT It is a continuous process of learning of key facts by participants about a clinical trial. It also explains the rights of the participant. It is not a contract and the participant can withdraw from the trial at any time. Informed consent is only a communication document.

INFORMED CONSENT Participants are human beings with a motive to help the researcher and society. Researcher should never be over- enthusiastic in his intervention to get his results while dealing with participants.

RANDOM ALLOCATION The participants in the study population are randomly allocated into two groups (Arms) using Random Number tables, or Computers to avoid selection and confounding biases.

Block Randomization Disadvantage of Randomization is, it produces imbalance in the number of participants assigned to each group, if groups are small Then we do block randomization; it produces more balanced groups

Block Randomization The investigator wants that after every 4th randomized participant the number in each group is equal Then block of four. Four participants are ranked according to random number table; eg; 1, 8, 4, 7 then these numbers are arranged as 1,4,7,8 and then one of the following 6 combination is used decided in advance ABBA, ABAB, AABB, BBAA, BABA, BAAB Suppose we are using the first combination then patients assigned the numbers 1 and 8 will go to group A and the others (4,7) to group B

1000 Patients (600 males and 660 young) Stratified Randomization To achieve between group comparability of certain characteristics. eg. Age and sex 1000 Patients (600 males and 660 young) 600 males 400 females 360 young 240 old 300 young 100 old 180 + 120 + 150 + 50 Intervention group 180 + 120 + 150 + 50 Control group

BENEFITS OF RANDOMIZATION? Allocation or selection bias is minimized and produces groups comparable to known and unknown risk factors. This elimination of allocation bias will greatly enhance the validity of the trial. Allocation bias is minimized by selecting randomly. Chance plays the role but not the choice of the investigator during randomization. Each individual will have equal chance of to enter the trial Investigator will not have any have any choice to allocate into any group, either the intervention group or the comparison group by his choice according to his whims and fancies, if he allocates randomly into two arms.

After random allocation into groups INTERVENTION After random allocation into groups The intervention (new drug or new procedure) is applied to the one group and placebo or standard or old procedure to the other group.

FOLLOW-UP Better compliance will lead to better validity which in turn enables for better generalizability Both the groups are similarly followed in all aspects like duration, type of follow-up

MAINTENANCE OF COMPLIANCE Selecting high risk people as participants in study population (high risk are more motivated) Frequent contacts with the participants through phone calls, home visits, clinic/hospital visits Providing calendar packs to the participants and asking them to stick on to calendar packs without fail. Giving incentives like free medical aid in future, giving some gifts. *If the study population selected is at high risk for the disease for which the drug/intervention is tested, the participants will be strongly motivated and more cooperative compared to those at usual or normal risk. Significant difference between the outcomes of the two groups depends not only on sample size and elimination of biases but also on the compliance.

NON-COMPLIANCE Extent of non-compliance is directly proportional to the duration and complexity of the trial. Compliance is difficult when the end –points are time taking like incidence of cancers or death Non-compliance decreases the statistical power of the trial which speaks about the validity (truth of the results)

ASSESSMENT CRITERIA Depends on whether the outcomes or end-points are single or multiple, subjective or objective uniform & similar type of evaluation of end-points for both the groups is to be carried out.

ASSESSMENT CRITERIA Subjectivity of the outcome e.g. reduction of pain, may lead to observer error and poor assessment. Double blinding eliminates observer bias to a large extent.

HAWTHORNE EFFECT Sometimes the participants in comparison group may exaggerate the effects/outcomes to please the investigator or when they like the study or for some other reasons. This will affect the assessment unless controlled.

BLINDING PURPOSE Blinding or Masking is done to eliminate Investigator bias Evaluation bias Investigator may influence the trial or manipulate evaluation of the results (investigator bias) or the participant may exaggerate the effects (Hawthorne effect), Investigator may influence the trial or manipulate evaluation of the results (investigator bias) or the participant may exaggerate the effects (Hawthorne effect),

BLINDING (masking) The *participant, (study subject) the*investigator, and sometimes even the *evaluator are all kept unaware (blinded) of the outcomes of the trial and secrecy is maintained to improve the validity.

Types of Blinding x ---- X -------- Single Double Triple Subject Researcher -------- Data Analyst = Blind with respect to subject’s allocation = May be aware of subject’s allocation x ---- 50

UNBLINDING In emergencies and life threatening situations for participants, unblinding can be done.

Kaplan Mayer method is usually used for analysis ASSESSMENT Kaplan Mayer method is usually used for analysis In Veteran administration study for Hypertensive efficacy, impotence was observed 29% in intervention group & 28% in comparison group. If there is no comparison group, it may be erroneously considered that hypertensives will cause impotence In Aspirin Myocardial Infarction Study (AMIS 1980) complaints of gastritis and peptic ulcer were reported 23.7% and 14.9% in intervention and comparison groups respectively, thereby lessening the blame on Aspirin. Group A Frequency of deaths Group B time

Sample size calculation for RCT What should we know in order to estimate the sample size The estimated difference in response rate of the two arms of RCT to be detected An estimate of the response rate in one of the groups (either control or interventional) Level of significance (alpha) The value of power of the study desired (1 - beta) One sided test or 2 sided test

Differences in cure rates between 2 groups Sample size: Number of patients needed in each group to detect various differences in cure rate. (alpha=0.05, 1-beta= 0.8, 2-sided) Lower cure rate Differences in cure rates between 2 groups 5% 10% 15% 20% 25% 30% 40% 50% 420 130 69 44 36 31 20 14 680 195 96 59 41 35 23 17 910 250 120 71 48 39 25 1090 290 135 80 53 42 26 18 1250 330 150 88 57 28 1380 360 160 93 60 29 35% 1470 370 170 61 1530 390 175 97 45% 1560 391 176 -

Alpha, Beta, Power of Study Reality Your decision Treatments not differ (H0: true) Treatments differ (Ho: false) Treatments do not differ (fail to reject H0) Correct decision Type 2 error (beta) Treatments differ (reject H0) Type 1 error (alpha) Correct decision (1 – beta)

Home Work Prepare the proposal of a randomized control trial RCT on the topic of your own choice.