Indian Journal of Pharmaceutical Education and Research Association of Pharmaceutical Teachers of India

Clinical Research Methodology

Purvi Gandhi*
Department of Pharmacology, B. S. Patel Pharmacy College, Mehsana, India.

ABSTRACT Submitted: 29/5/2010 Revised: 29/8/2010 Accepted: 04/10/2010

Medical trials occupy a greater part of clinical trials and exceed in number over other types of clinical trials. Although clinical researches or
studies other than those falling under the category of trial, lack the commercial aspect, they may contribute to the betterment of clinical practice
and may constitute a base for large-scale commercial trials. Explanatory clinical studies are the most widely accepted and executed type of
clinical researches/ trials. They explain not only the therapeutic aspects of a drug but also provide with a detailed description on adverse events
and pharmacovigilance. Any new drug to be marketed first requires proving its safety, efficacy and the need over existing product range by
passing through various phases of clinical trial that follows animal studies. For this reason, randomized controlled trials (RCTs) with placebo-
control and blinding fashion is the only accepted form of clinical trials with few exceptions. Besides explanatory, descriptive clinical research is
also useful for certain types of studies such as epidemiological research. While conducting RCTs on humans (either healthy volunteers or
patients), various components of the trials viz. study design, patient population, control group, randomization, blinding or non-blinding,
treatment considerations and outcome measures are important to strengthen the outcomes of the trial. Similarly, proper utilization of statistics,
for sample-size calculation, data collection, compilation and analysis by applying proper statistical tests, signifies the outcomes of the clinical
research. The presented article includes a brief, however, an informative review of literature on methods in clinical research including clinical
trial.
KEY WORDS: Clinical research methods, Clinical trial methods, Statistics in clinical research

INTRODUCTION
either on healthy volunteers or on volunteer patients. Healthy
Clinical research is a branch of medical science dealing with volunteers are generally included in such trial that determines
any research or study in living humans. 'Clinical trials' is the pharmacokinetics, tolerability, safety and even efficacy of
term interchangeably used with the terms 'clinical research' or certain types of drugs (e.g. hypoglycemic, hypnotic, diuretic
'clinical study'. Although there are many definitions of etc.). Otherwise, for majority of drugs (e.g. antiepileptic,
clinical trials, they are generally considered to be biomedical antipsychotic, anti-inflammatory, antitubercular etc.),
or health-related research studies in human beings that follow efficacy can only be assessed in patients.2
a pre-defined and –designed protocol. Clinical trial is defined
as “a systematic study of new drug(s) in human subject(s) to The research entailing the use of human participants is
generate data for discovering and/ or verifying the clinical, considered to be absolutely essential after a due consideration
pharmacological (including pharmacodynamic and of all alternatives in the light of the existing knowledge in the
pharmacokinetic) and/ or adverse effects with the objective of proposed area of research. In other words, when a new drug is
determining safety and/ or efficacy of the new drug”.1 Clinical with clear significant benefit at human side, human as
trial is company-sponsored, meant for a new drug or device participants for trial experimentation becomes justified.
and carried out for a specific new use of an intervention; while International Conference on Harmonization has provided a
clinical research is meant for academic and guideline on Good Clinical Practice (ICH GCP) as an
pharmacovigilance. Large number of literature describes on international ethical and scientific quality standard for
clinical trial and its phases. However, it has been now started designing, conducting, recording and reporting trials that
to include a detailed description on clinical research or study. involve the participation of human subjects.3 World Health
Organization Guidelines for good clinical practice for trials
Ethics in clinical trial
on pharmaceutical products also describe provisions and
Depending upon the objective, clinical trial is conducted prerequisites for a clinical trial, protocol and protection of
trial subjects, responsibilities of the investigator,
*Address for Correspondence:
responsibilities of the sponsor, responsibilities of the monitor,
Purvi S. Gandhi, Department of Pharmacology, B. S. Patel Pharmacy College , Saffrony
Institute of Technology Campus, Nr. Saffrony Holiday Resort, Ahmedabad –Mehsana monitoring of safety, record-keeping and handling of data,
Highway, Linch, Dist. Mehsana,Gujarat, India statistics and calculations, handling of and accountability for
E-mail: [email protected] pharmaceutical products, role of the drug regulatory

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 199
 Purvi Gandhi- Clinical Research Methodology

authority, quality assurance for the conduct of a clinical trial phase II and III as the actual clinical trial.11,12 This may be
and considerations for multicentre trials.4 United States Food because phase I is usually carried out on healthy volunteers
and Drug Administration Guidelines for conduct of clinical except for diseases like cancer, acquired immuno-deficiency
trials provide guidance for institutional review boards and syndrome (AIDS) etc. Moreover, phase IV may not
clinical investigators, information for health professionals on necessarily follow randomized, controlled trial (RCT) with
clinical trials and human subject protection, information for blind fashion of the most commonly-employed type since the
clinical investigators on drugs, devices and biologic as well as drug is available in market with known label-indication and it
policies and regulatory matters regarding human subjects requires chronic drug administration and observation of
research.5 Several issues and principles have been discussed patients.
in various guidelines on conducting clinical trial (especially 1) Phase I: Human/ Clinical Pharmacology trial:
drug trial) which must be addressed while conducting a trial.
These include the following: This phase aims to obtain the precise information on a) initial
safety in terms of safe dosage range and biological effects
1) Ethical justification and scientific validity of biomedical including adverse effects; b) metabolism and kinetics and c)
research involving humans drug interactions.9 The trial is carried out on healthy human
2) Ethics review board volunteers (20-80 in number) except for life-saving drugs
meant for treating life-threatening diseases such cancer, AIDS
3) Informed consent process
where actual patients only are included. The volunteers or
4) Choice of control in clinical trials patients are exposed to a single dose which is usually kept as
5) Research involving special group of research 1/12th of the effective dose found from animal studies
participants.6 especially from mice study.13

Moreover, clinical trial in India on a new drug or device or any 2) Phase II: Exploratory trial:
surgical intervention shall be initiated only after the This phase is meant to find whether or not the drug possesses
permission has been granted by the Licensing Authority under the actual therapeutic potential. It identifies the therapeutic
Rule 21 (b) and the approval obtained from the respective efficacy of the drug with dose range, kinetics as well as
Ethics Committee(s). 7 Subjects who are relatively or metabolism.9 It follows RCT type with blinding of treatment.
absolutely incapable of protecting their own interests i.e. It is carried out at one or few clinical centers only on small but
“vulnerable participants” include very poor, illiterate, sufficient number of patients (100-300 in number) to reach
terminally-ill, mentally-challenged, prisoners, students or clinical significance in outcomes. It also aims to find out
employees are examples.8 Clinical trial must preserve their therapeutic index of the drug being studied. It requires to be
rights while being conducted in such humans. Moreover, carried out in patients meeting selection criteria of age group,
there must be equal distribution of burdens and benefits in sex, presence of particular disease with pre-defined and
terms of race (in case of genetic research), economic and –diagnosed severity etc. It also reports the adverse effects of
social level, mental-healthiness and reduced autonomy, to the drug.
avoid unintentional bias and to extend the benefit of trial to
3) Phase III: Confirmatory trial:
maximum of communities.
This phase applies the same study protocol designed for phase
Phases of clinical trial
II to evaluate safety and efficacy at large.9 It is to confirm the
Clinical trial of a drug is conducted through various phases. effectiveness of the drug or treatment, to monitor side effects,
The number of phases as trial varies from literature to to compare it to commonly used treatments and to collect
literature and from author to author. Most literatures describe information that will allow the drug or treatment to be used
a clinical trial/ testing of a new molecule to comprise of four safely.14 It is simultaneously performed at a large number of
phases.9 These phases are as following: clinical centers that include patients of various geographic
1) Phase I : Human/ Clinical Pharmacology trial origins with difference in responsiveness of the disease
towards the drug treatment. Further, it covers large number of
2) Phase II : Exploratory trial patients (1,000-3,000 in number) allowing the outcome to
3) Phase III : Confirmatory trial reach not only clinical significance but also statistical
significance. Once the drug passes this phase successfully, it
4) Phase IV: Post-Marketing Surveillance.10
is licensed for commercial use. Thus, phase III of trial is a key
Some literatures describe the above-mentioned first three study forming the primary basis for regulatory approval of an
phases as actual phases of clinical trial while others consider intervention and is often referred as pivotal trial. It generally

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 200
 Purvi Gandhi - Clinical Research Methodology

proceeds as randomized, placebo-controlled, double-blind molecule and are rarely with totally new basic structure, data
clinical trial.15 from extensive pre-clinical i.e. animal studies are essential to
conduct clinical trials of a new drug. When a drug molecule is
4) Phase IV: Post-Marketing Surveillance:
a new chemical entity, detailed and extensive studies in
This phase is so named because it is carried out after the drug animals become essential and more stringent. Biotechnology
is released in the market for therapeutic use. It is mainly to products, mainly all recombinant drugs are considered as new
detect uncommon but significant adverse effects.9 Once the drugs (irrespective of international availability of the same
drug enters the market, it will be utilized by many more formulation) and hence treated accordingly. Pre-clinical
patients having other co-morbidity and co-existing diseases studies include those carried out in lower animals as well as
in addition to the disease for which the drug is indicated and higher primates with special emphasis on therapeutic and
licensed. This is also conducted to provide critical adverse effects. Further, it is also required to establish
information on drug-drug interactions or iatrogenic diseases. minimum dose producing the desired effect (i.e. therapeutic
The best example is of thiazolidinedione series of drugs, effect) and maximum dose causing adverse effects in specific
which were found to deteriorate heart failure (HF) if used as animal species. The investigator should have data on acute,
antidiabetics in HF-patients. Therefore, they are now sub-acute and chronic toxicity studies. The investigator can
contraindicated in diabetic patients with HF. initiate clinical experiment(s) only after he/ she gets all the
following information and data about a new drug from animal
Clinical trials are conducted with the purpose of
studies:
commercialization while clinical researches/ studies do not
necessarily aim at commercialization. Clinical research can 1) The need of the new molecule over the existing drug
be based on any of the following four concepts: range- the need can be either because of its different
mechanism of action and hence additional benefit or due to
1) treatment of a disease
less adverse effect or due to its greater efficacy and less cost at
2) diagnosis of a disease or disorder or dysfunction least.
3) systematic review of several clinical studies 2) Specific pharmacological actions i.e. those with
4) prognosis of a particular disease. 16 therapeutic potential for humans and general
pharmacological actions i.e. those on other organs and
Clinical research based on therapy/ treatment can be focused systems, especially cardiovascular, respiratory and central
on any of the two areas; 1) pharmacology of a drug and 2) nervous system.
effects of a non-pharmacological intervention such as a
surgery or a device. Clinical pharmacology is a branch of 3) Pharmacokinetic data.17
pharmacology dealing with study of a drug in humans (either 4) Therapeutic index (i.e. ratio of dose lethal in 50% of the
healthy volunteers or patients). A large number of new drugs study population (LD50) to dose effective in 50% of the study
or molecules are synthesized in laboratories or are extracted population (ED50) which is denoted as TI= LD50/ED50) along
from natural sources viz. plants or animal organs. However, with the minimum dose required to produce the desired
only a small number enters pharmaceutical market as a therapeutic effect and the maximum dose producing the
successful drug for treatment of a disease. Many devices, as a unwanted or lethal effect.
disease-intervention or treatment, are also manufactured
5) Toxicity data.
every year but only a few find a place in healthcare system.
This is because of the stringent procedure for a drug or device 6) Nature of the adverse effects of the drug along with
to pass through the pre-clinical and clinical phases to prove its identified signs and symptoms in order to avail the possible
safety and efficacy as well. A drug can not be marketed and treatment if the same occur in humans.
hence not be used for treating a disease until and unless it Clinical research based on clinical pharmacology includes
passes all the phases and proves its effectiveness. Further, exclusively the effect(s) of a drug in human and outweighs the
once a drug is registered and approved for general clinical use other types of clinical studies in number. Clinical studies are
for one indication i.e. one therapeutic use, expansion of its further typified and sub-typified as follows (Fig.1)18,19:
therapeutic-use range requires additional clinical research
and trial. Similarly, for any surgical intervention to be Clinical studies
recommended by the healthcare system, it has to pass through Descriptive studies
various stages of pre-clinical and clinical studies. Descriptive studies report unusual or new events such as the
Even though most of the new drugs are either structurally occurrence of sudden infant death syndrome (SIDS) in
similar to or are the derivatives of already-existing drug several siblings within a single family, prevalence of albinism

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 201
 Purvi Gandhi - Clinical Research Methodology

in a single family etc. The researcher simply records the Explanatory studies
observations and co-relates the events observed with possible Observation studies
reason. These are neither randomized nor pre-designed
researches. They may be presented as case reports whereby In an observational study, the subject to be observed chooses
certain individual patients with distinguished clinical whether or not to take the drug or to have the surgery being
characteristics are included in the study. All the baseline studied.20 Errors that are likely to occur include the differences
characteristics are recorded and the individual patient is in profile of the subjects since variables such as age, family
treated as unique case with control over all the variables. The history of disease, cause and severity of disease etc. may not
be defined. For example, two patients have left ventricular
patient is observed and evaluated for the possible outcome.
(LV) dysfunction, in one it is because of ischemic heart
The results are compared with baseline values or are
disease (IHD) and in another it is because of severe mitral
expressed as success or failure of the treatment given. If the
valve stenosis. Thus, the therapy of both the diseases differs
treatment succeeded, a hypothesis is generated for an
due to different oetioes and hence both the patients can not be
expanded and more rigorous study to find the relationship
compared in one study. Another example is of two patients
between the treatment and the outcome observed. In case-
suffering from headache, one because of migraine and the
series, observations are documented at regular intervals from other because of common cold. These two patients can not be
patients exposed to a particular drug or a group of drugs. They compared for the analgesic activity of one drug since the
may also cover prior histories of patients with the same cause and the severity of headache and hence the analgesic
outcome, to find a possible cause-effect relationship if exists. activity of the drug would vary greatly. Observational studies
These are useful in predicting the incidence of an adverse can never be blinded. Hence, biases from patients, observer
event of newly-marketed drug when reports on such events and experimenter may result into systematic and random
are limited. errors.

Fig.1: Types of clinical studies

Clinical Studies

Descriptive Explanatory

Case Reports Case Series

Experiment Observation

Randomized Non-Randomized Individual Aggregate

Randomized No- Historical Active

Clinical Trial Treatment Control Control
with Placebo Control
Control
Group Assembled on basis of

Exposure Both Outcome

Cohort Cross-Sectional Case-Control

Groups defined Groups defined

in past in present

Retrospective Cohort Prospective Cohort

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 202
 Purvi Gandhi - Clinical Research Methodology

Aggregate observation studies money consumption. In prospective cohort study, the groups
Pandemic and epidemic studies on communicable diseases are observed for outcomes at particular, pre-decided time
and their treatments are generally carried out as aggregate intervals. Thus, it finds firmly whether a particular exposure
observation studies e.g. occurrence and effective treatment of or sign or symptom is related with the outcomes. If the
malaria and its relapse in particular geographical area. outcome is rare, the study requires inclusion of large number
of patients and longer follow-up. Thus, it is expensive in terms
Individual observation studies of time and money. If the patients are not randomized and
In individual observational study, the patients/ subjects are blinded, the outcomes may be influenced by bias and
individually observed and they are assembled in groups on the confounding.
basis of outcome or exposure or both. Depending upon the 3) Cross-sectional
basis of the grouping, the individual observational study is
sub-classified as 1) Case-control; 2) Cohort and 3) Cross- Cross-sectional study assesses both the exposure and
sectional.19 outcome concurrently. Generally it is survey- or review-
based. Cross-sectional study is, therefore, good for
1) Case-control study prevalence research. However, it is not suitable for causal-
Case-control study involves assembling of subjects in groups outcome assessment.
on the basis of the outcome found in those subjects. It Experimental studies
compares the subjects with outcome in question (the group
behaves as a case group) with the subjects without the Non-randomized studies
outcome (the group acts as a control) e.g. occurrence or non- Patients are selected on the basis of selection criteria. They are
occurrence of myocardial infarction (MI) in patients with not randomized to the particular treatment(s) and are given a
hypertension (HT). It generally follows the retrospective treatment depending upon course of disease. Generally, phase
design and evaluates how the exposure is related to the well- IV of clinical trial follows this way. Further, in many
defined outcome using control group.19 However, grouping on experimental studies in humans, randomization is not
the basis of outcome incorporates subjects with variety of possible. Many of the surgical experiments have evolved with
distinguished characteristics. It is quick and inexpensive. specific indication and application. They have a focused-
Further, patients with rare outcome can be assembled in a patient group and therefore, randomization is not possible or
group to study oetioes, pathophysioes and prognosis of a is unethical. For example, patients with both the kidneys
disease. Results are generally expressed in terms of odds ratio failed require undergoing kidney transplantation. Although,
(OR) and risk ratio/ relative risk (RR). Although multiple dialysis is an available option it is not comparable with renal
exposure variables can be correlated with outcome, it does not transplantation and hence patients can not be randomized to
allow the correlation of temporal sequence of cause and effect such options.
with the final outcome.
Randomized controlled trials
2) Cohort
In the studies which are randomized, controlled clinical trials
It includes groups assembled on the basis of exposure. Here (RCTs), human subjects (either healthy volunteers or
the exposure is well-defined but the outcome is variable. patients) do not choose the therapy being studied or
Thus, it allows study of one exposure with many more compared. Experimental clinical studies are generally RCTs.
outcomes.19 Cohort study can be retrospective wherein the Randomized controlled trials are, as the name indicates, based
groups are defined in past or it can be prospective wherein the on randomization. When a new drug successfully passes the
groups are defined in present. The retrospective cohort pre-clinical studies, it is challenged to clinical experiments
correlates the exposure occurred in past with the outcome that follow random assignment of subjects to two or more
resulted just in recent past. Here the patients have been groups one of which behaves as control group and therefore,
followed forward and hence it associates the exposure with such clinical experiments are called RCTs. The several
some temporal outcomes though not all. If the patients have components to be considered include 1) Study design; 2)
been treated with different treatments to control outcome- Patient population; 3) Control group; 4) Randomization; 5)
related variables, it limits the correlation between exposure Blinding or non-blinding/ open-labeling; 6) Treatment
and one outcome only. Like case-control study, it is also quick considerations and 7) Outcome measures.
and inexpensive. If carried out on the basis of well-defined, -
1) Study design
controlled exposure and followed with control over variables,
retrospective cohort study suffices the requirements of The common study designs employed in RCTs include
prospective study with additional advantage of less time and parallel group design, matched pairs and cross-over designs.21

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 203
 Purvi Gandhi - Clinical Research Methodology

In parallel group design, the patients are enrolled, followed 3) Control group
and observed for outcomes on parallel basis. Parallel group
Randomized controlled trial also includes control group
design requires large number of patients. In matched pairs,
(either placebo control or active control) to show the control
patients are matched for different variables and those
and effect over dependent variables and to obtain clear effects
matching the required variables are then randomized to
of drug under consideration. Control group can be placebo
various treatment groups. This type of study design
control, no-treatment control, historical control or active
overcomes the influence of variables on outcomes, although it control. The placebo means dummy to the drug under
is difficult to follow. Cross-over design is particularly used evaluation with regard to organoleptic properties but lacking
when the effect of a drug is reversible and transient. In cross- any pharmacological actions. Thus, it is to overcome the
over design, the patients are given more than one treatment psychological impact of drug administration manifested by
but in sequence i.e. one after another when the effect of an individual on disease progression. It allows the
previous treatment is washed out. Cross-over design, thus, investigator to determine the true efficacy of the treatment
requires less number of patients. being researched for a particular condition. Some studies also
2) Patient population include no-treatment control or historical control as types of
controls. In no-treatment control group, the patients do not
As a common and required method, the RCTs are carried out
receive the placebo even. Therefore, they know that they do
on specific subject population selected on the basis of
not receive any treatment and hence, individual bias due to
“selection criteria” which are derived in line with various psychic factors affects the study outcomes. In other words, it
fixed, independent and dependent variables. This is to is least preferred type of control. Historical control is the
overcome the misleading by variables. For example, if control group of previous study that was a different with
effects of angiotensin-converting enzyme inhibitor (ACEI) respect to treatment group. Here control group of one study is
on cardiac function are to be studied in patients with LV utilized for another study and both the studies differ with
systolic dysfunction, variables like family history of cardiac regard to treatment only. This is done for studies not allowing
disease, presence of other cardiac diseases such as heart block placebo control or no-treatment control and involving high
or valve failure etc. should be avoided as patients with these mortality disease even after availability of effective treatment
variables are different from those not having the variables. e.g. studies on treatment for cancer and human immune-
Further, they may reveal different outcomes viz. the cardiac deficiency virus (HIV) infection.
function and even survival. Depending upon the defined
Inclusion of placebo in drug research and sham surgery has
criteria, patients or healthy subjects are included in the study
been debated.6 Moreover, when an effective established
to randomize to various treatments for the comparison of
treatment is available, use of such placebo control group is
outcomes and thus, to conclude. The criteria are namely (a)
unethical. For examples, a drug is to be assessed for its effects
inclusion criteria; (b) exclusion criteria and (c) withdrawal
on cardiac function in patients with LV systolic dysfunction,
criteria.
as per American College of Cardiology/ American Heart
(a) Inclusion criteria: Specifications of subjects (patients or Association (ACC/ AHA) guidelines all the patients would be
healthy volunteers) with regard to age, gender, ethnic groups, necessarily receiving ACEI, if not contraindicated.22
body mass index, prognostic factor, diagnostic admission Therefore, in this type of study all the patients receive the
criteria, should be clearly mentioned wherever relevant. recommended drug which has already proved its beneficial
(b) Exclusion criteria: These specify the characteristics of the effect on cardiac function. Thus, one can not have a placebo
control group but will have an active control receiving the
subjects on the basis of which they are excluded from the trial.
best current therapy. It provides information about relative
For example, severity of the disease, concurrent medication
efficacy of the investigational drug over existing one. In the
etc.
present example, the patients would be randomly assigned to
© Withdrawal criteria: These specify the subjects on who a group receiving ACEI or to a group receiving ACEI in
the trial shall be terminated and mention when and how to addition to the drug being evaluated- the former behaving as
withdraw the subjects from the study and to stop further an active control and the later as a treatment group.
follow-up in those.10
4) Randomization
To have comparison possible between or among various Randomization is an optimal method of distributing the
treatment groups, selection of patients must be done on the
variables between the treatment and control groups.9
basis of inclusion and exclusion criteria. This allows Therefore, the bias of selecting specific treatment does not
enrollment of subjects with corresponding clinical occur. Random assignment of subjects to various groups
characteristics.9 provides equal distribution of all variables in all the groups

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 204
 Purvi Gandhi - Clinical Research Methodology

and does not let them influence the final outcomes. by Neutel and Smith (2003).23 This type of trial is easier to
Randomization techniques mainly used in RCTs are simple carry out than a double-blinded placebo controlled design
randomization, cluster randomization and stratified (DBPC) because it does not require the “matched placebo
randomization. In simple randomization, patients matching group” and the “open-label” allows the enrolled patients to
the selection criteria are randomized to various treatment receive a marketed preparation of the drug. However, the
groups. In cluster randomization various groups of patients PROBE studies have only the end-point blinded i.e. observer
matching the criteria are randomized to treatment under is unaware of the treatment being studied while investigator
investigation. This kind of trial is especially used to find the and patients are aware of it. Therefore, the investigator or the
geographical, genetic variations. In stratified randomization patient bias may be introduced and thus, the results obtained
technique, subjects are classified in groups i.e. strata and then are less reliable than those with double- or triple-blind study.
within a group they are randomized to various treatment
6) Treatment considerations
groups. In RCTs, three main methods of randomization
include 1) Tables of random numbers; 2) Mathematical While conducting RCTs, the treatment (either being studied
algorithms for pseudorandom number generators and 3) or behaving as active-control) must be considered with regard
Physical randomization devices such as coins, cards or to its dosages, dosing frequency and other concurrent
sophisticated devices such as Electronic Random Number medication. A drug is generally available in various dosage
Indicator Equipment (ERNIE). forms viz. tablet, capsule or injectable etc. and it varies in
strength. Moreover, depending upon the dosage form, the
5) Blinding
route of administration differs and hence, the amount of
To avoid bias, trial is carried out in blind fashion. Blinding administration and dosing frequency also. All these factors
means “concealing or masking of the patients-assignment to a together affect the plasma concentration of drug and thereby
study group (control or treatment) from those participating in the effects of the drug and hence the final outcome. Therefore,
the study i.e. patients, observer and experimenter”.9 RCTs can except dose and frequency of drug(s), all above-mentioned
be blinded or non-blinded. The non-blinded experiment is factors are kept unique and constant through out the study.
also called open-label study. In this type of study all three- the Whenever dose and frequency need to be changed, it is done
patient, the physician or the observer and the experimenter or gradually and stepwise. If two drugs are to be administered
the researcher, are aware of the treatment used. In many one of which is likely to interfere with the other either
instances it is unethical to hide the treatment module from the pharmacokinetically or pharmacodynamically, the dosage
patients especially those suffering from life-threatening must be reconsidered to overcome the influence of such
disease such as cancer, AIDS, end-stage HF etc. Additionally, interference on study outcomes. Patient compliance is
open-label study permits the patients to buy brand of the drug another important part of the treatment consideration. A
of his choice independently. However, it has the biggest treatment should not be non-compliant as the patient avoids or
disadvantage of introducing bias from any of the three less prefers to take such medication resulting in erroneously
components of the RCTs. less efficacious outcomes than those obtained with the other
Blinding is carried out at the beginning of study. The blind treatment group.
RCT can be single-, double- or triple- blind. In a single-blind 7) Outcome measures
experiment, the participants either the patient or the healthy
volunteer does not know whether he receives the test The objective of the study determines the outcomes of interest
intervention or placebo. In double-blind trial, neither the to be measured. These measures are nothing but the points of
patient/ subject nor the experimenter knows who belongs to checking and recording to accomplish the comparison. In
the control group and who belongs to test group but the experiments the outcomes are measured in terms of efficacy
observer knows. In triple-blind RCT, none of the three end-points i.e. primary end-points and surrogate end-points
components of study knows name or nature of the treatment which are also called secondary end-points. For examples, in
given. Therefore, the triple-blind RCT is totally devoid of any an experiment evaluating an antihypertensive agent, the
kind of biases and allows the outcomes to be free from any clinical end-point of real interest is whether the treatment
such influence. In double- and triple-blind experiment the under investigation can reduce cardiovascular events; a
keys identifying the patients/ human subjects and the group surrogate is the ability of the treatment to reduce blood
they belonged to are preserved by a separate another party and pressure.24 The primary end-points of the study are the main
given to the researcher only at the end of the study. measures to support or refute the hypothesis of the study.
Randomized controlled trial can also be conducted as They must be defined and specified by the investigator at the
PROBE. PROBE is an acronymus of Prospective, beginning of the study. Secondary measures, although pre-
Randomized, Open-label, Blinded-End point as used earlier specified before the commencement of the study, can be

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 205
 Purvi Gandhi - Clinical Research Methodology

further elaborated during the study. For example, when Power is the ability of a statistical test to show significance if a
diuretics are used for treating hypertension, serum glucose specified difference truly exists.27 It is essential to minimize
level measurement can also be added though serum these errors at pre-decided levels or below to draw a
electrolytes are usually measured as main secondary end- conclusion in a clinical study. Furthermore, results of a
point. Although various measures are determined as primary statistical analysis are found conclusive only when the sample
and secondary end-points, quality of life is now-a-days size is sufficiently large. However, because of time and cost
becoming main primary end-point.25 factors, it may not be possible to enroll large sample size in a
STATISTICS IN CLINICAL RESEARCH study. In that case, finding power of the study may disclose
and support the inability of a test not to reach a statistically
Statistics play a crucial role in concluding a clinical research. significant difference between the groups, even when the
It is applied in clinical research to analyze data and to infer the clinical difference is significant.
results obtained. It is important to obtain a statistically
significant difference between two or more groups being Sample size
compared in a clinical research, in order to make the outcomes Sample size depends upon the design of study, nature of
acceptable. Statistics is also required at the beginning of the variable and measurement scale.21
trial to calculate the sample size required to reach a statistical
Sample size formula for quantitative response
significance in the findings.
The number of patients required per group can be estimated
'P' value and level of significance
using following formula:
In a clinical research, a null hypothesis is stated and tested by
2 (SD)2
finding the difference between/ among the results of groups
involved in the research. The difference in the results obtained N per Group = ----------------x (Zα + Zβ)2
between/ among various groups should be of statistical (Diff. or ? )2
significance in order to reject the null hypothesis and thus, to Where,
accept the alternative hypothesis i.e. “treatment being study
SD = estimated standard deviation – known or based on some
as effective one”. In many instances, clinical research finds
other study
the difference in the results of clinical significance but fails to
attain a statistical significance and therefore, the null Diff= ? = expected difference between two treatment
hypothesis is accepted. Rejection or acceptance of a null considered as clinically important
hypothesis is based on 'P' value. 'P' value is defined as “the Zα = Z value from statistical tables corresponding to level of
smallest level of significance of the difference in the results
significance (α)
that would reject the null hypothesis”.26 It tells how likely it is
that the difference between/ among groups occurred by Zβ = Z value from statistical tables corresponding to (β)
chance rather than because of an effect of treatment.16 For 2 sided tests with α= 0.05 and β= 0.20
Types of errors and power of study Zα = Corresponding to level of significance (0.05)= 1.96
The 'P' value is based on two types of errors that one may Zβ = Corresponding to (β= 0.20) = 0.842
encounter during experiment. These two errors are designated
as type I error and type II error. The former is also called alpha Sample size formula for qualitative data
(α) error and the later beta (β) error. A type I error occurs if a The number of patients required per group can be calculated
difference is found between A and B when none actually using following formula:
exists.11 Thus, α error indicates the chances of detecting a
difference which does not actually exist i.e. the chances of N = (2 x Joint Success rate x Joint Failure rate) x (Zα + Zβ)2
having False Positive Difference. A type II error occurs if no (Diff)2
difference is found though A and B do actually differ.11 Thus, β
Confidence interval
error indicates the chances of not detecting a difference which
actually exists i.e. the chances of having False Negative The confidence interval (CI) gives a range. It gives a measure
Difference. Alpha error indicates the level of significance of of reproducibility of the results within the obtained range.
the result difference among various treatment groups. The Expression of 'P' value along with CI is clinically more useful
level of significance is usually set at the traditional value of and acceptable by many researchers. Generally, it is kept at
5%. Beta error gives an idea about power (1-β) of a clinical level of 95%. A 95% CI indicates that if the study is repeated
study. Beta error is often chosen to be between 5 and 20%. 100 times, the study results would fall within this interval 95

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 206
 Purvi Gandhi - Clinical Research Methodology

times.16 For example, if improvement in LV ejection fraction male or female proportion in occurrence of a disease. Ordinal
(LVEF) after revascularization in 95 patients is 6% on an data are expressed as scores and ranks e.g. pain, categorized
average when compared with baseline with a 95% CI of 4.5 to as mild, moderate and severe and can be scored as 1, 2 and 3
9% for the difference, it is concluded that the respectively. The numerical data are observed in form of
revascularization has the specificity of producing interval measurements either continuous (e.g. blood sugar
improvement in LVEF by 4.5 to 9% if this revascularization is level, blood urea level) or discrete (e.g. number of patients
performed in such 100 patient populations i.e. if it is repeated admitted to a hospital, heart rate etc.).
100 times. Statistical tests
Odds ratio and relative risk Application of most suitable statistical test allows analyzing
Odds ratio (OR) and relative risk (RR) both are measures of the data and interpreting the results of the study. Different
the size of an association between an exposure and a disease statistical tests are applied in a clinical research to analyze the
or death.16 For example, association between smoking or HT data and to infer the results obtained depending upon the type
and development of IHD; use of a medication and occurrence of data and their distribution (Table No.1).29
of a side effect; exposure to MI over global LV ischemia and
Status of clinical research in india
mortality etc. are expressed in terms of OR or RR.
Observational studies usually report their results as OR or Clinical research in form of trial is conducted not only as
RR, although experiments also include these types of unicenter but also as muticenter at various clinical research
measurements as safety and efficacy end-point. A RR of 1.0 centers spread over various countries including India. In
indicates that the exposure does not change the risk of disease. India, for international collaborative study, details about
A RR of 1.9 indicates that patients with the exposure are 1.9 foreign collaborators and documents for review of Health
times more likely to develop the disease or have a 90 percent Ministry's Screening Committee (HMSC) or appropriate
higher risk of disease. For example, if the RR of Committees under other agencies/authority like Drug
hyperlipidemia is 1.4 for development of IHD indicates that Controller General of India (DCGI) are implemented and
patients with hyperlipidemia are 1.4 times more likely to followed in line with the guidelines by Indian Council of
develop IHD than those without hyperlipidemia or they have Medical Research. The centers participating in the trial are
a 40% higher risk of developing IHD. taken care by clinical research organizations (CROs), which
play a distinguished role of central facilities. With
Odds ratio is a way to estimate relative risks in case-control
advancement and development of various guidelines to
studies, when the RR can not be calculated specifically.
implement in such trials, more than 20 CROs have come up
Although it is accurate when the disease is rare, the
with many known to conduct trials at international level.30
approximation is not as good when the disease is common.16
Though developing at full swing, further expansion of field to
Data analysis include research on biologics and devices is needed.
Data analysis is carried out after compiling the observations ACKNOWLEDGEMENT
and applying appropriate statistical test. The test to be applied
The author is thankful to Dr. R. K. Goyal, The Vice-
depends on the type of data and their distribution in the study.
Chancellor, The Maharaja Sayajirao University of Baroda,
At large, the data are categorized as parametric or non-
Vadodara for his valuable suggestions.
parametric. However, in clinical study, the data collected for
analysis can alternatively be classified in four classes28: REFERENCES
1) Continuous e.g. blood pressure, blood sugar 1. Part X-A Import or Manufacture of New Drug for Clinical
2) Discrete, associated with numbers and ordered e.g. Trials or Marketing. In: Malik V, editor. The Drugs and
number of anginal episodes per week, number of MI attack in Cosmetics Act, 1940 and Drugs and Cosmetics Act, 1945
past etc. and their Allied Acts and Rules. Lucknow: Eastern Book
Company; 2006. p.154-58.
3) Attributes: categorical, ordered e.g. degree of
overweight, intensity of pain 2. Tripathi KD. Essentials of Medical Pharmacology. 6th ed.
New Delhi: Jaypee Brothers Medical Publishers (P) Ltd.;
4) Attributes: categorical, not ordered e.g. male or female, 2009. p. 59-77.
patients with diabetes mellitus or not
3. International Conference on Harmonisation of Technical
Data can also be typified alternatively as categorical or Requirements for Registration of Pharmaceuticals for
numerical. The categorical data can be nominal or ordinal in Human Use. ICH Harmonised Tripartite Guideline:
nature. Nominal data are expressed as proportion e.g. sex- Guideline for Good Clinical Practice E6(R1) 10 June

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 207
 Purvi Gandhi - Clinical Research Methodology

Table No.1- Type of data and applicable statistical test.

Type of Data
Goal Numerical Binomial (Two
Rank, Score Survival Time
Measurement Possible Outcomes)
Describe one group Mean, SD Median, Proportion Kaplan Meier
interquartile range Survival Curve
Compare one group to One-sample t test Wilcoxon test Chi-square or
a hypothetical value Binomial test
Compare Two unpaired unpaired t test Mann-Whitney Fisher's test Log-rank test or
group test (chi-square for Mantel-Haenszel
large samples)
Compare Two paired paired t test Wilcoxon test McNemar’s test Conditional
group proportional
hazards regression
Compare three or more One-way ANOVA Kruskal-Wallis Chi-square Cox
unmatched group test test proportional
hazards regression
Compare three or more Repeated-measures Friedman Cochrane Q Conditional
matched group ANOVA test proportional
hazards regression
Quantify association Pearson Spearman Contingency
between two correlation correlation coefficients
variables
Predict value Simple linear Nonparametric Simple logistic Cox
from another regression or regression regression proportional
measured variable Nonlinear regression hazards regression
Predict value from Multiple linear Multiple logistic Cox
several measured or regression or regression proportional
binomial variables Multiple nonlinear hazards regression
regression
SD, standard deviation; ANOVA, analysis of variance.

1 9 9 6 [ o n l i n e ] . Av a i l a b l e f o r m : U R L : Research and Methodology. New Delhi: Jaypee Brothers;

http://www.ich.org/LOB/media/MEDIA482.pdf 2007. p. 233-52.
4. World Health Organization Guidelines for good clinical 8. Thatte U. Ethical Issues in Clinical Research. In: Gupta
practice (GCP) for trials on pharmaceutical products SK, editor. Basic Principles of Clinical Research and
[online]. [cited 2010 Jul 29]. Available form: URL: Methodology. New Delhi: Jaypee Brothers; 2007. p. 58-
http://apps.who.int/medicinedocs/pdf/whozip13e/whozi 72.
p13e.pdf 9. Oates JA. The Science of Drug Therapy. In: Brunton LL,
5. U. S. Food and Drug Administration (FDA) Guidelines Lazo JS and Parker KL, editors. Goodman & Gilman's
for Conduct of Clinical trials (July 27, 2010) [online]. The Pharmacological Basis of Therapeutics 11th ed. New
[cited 2010 Jul 29] Available form: URL: York: McGraw-Hill Companies Inc.; 2006. p. 117-36.
http://www.ninds.nih.gov/research/clinical_research/pol 10. Central Drugs Standard Control Organization,
icies/fda.htm Directorate General of Health Services, Government of
6. Indian Council of Medical Research: Ethical Guidelines India: Good Clinical Practices for Clinical Research in
for Biomedical Research on Human Participants [online]. India. Schedule Y (Amended Version -2005) [online].
2006 [cited 2007 Jul 21]. Available from: URL: [cited 2007 Jul 21]. Available from: URL:
http://www.icmr.nic.in/ethical.pdf http://www.cdsco.nic.in/html/GCP1.html
7. Gupta SK. Requirements and Guidelines to Undertake 11. Rang HP, Dale MM, Ritter JM and Moore PK.
Clinical Trials in India- Schedule Y (Amended Version). Pharmacology 5th ed. New Delhi: Elsevier; 2003. p. 80-
In: Gupta SK, editor. Basic Principles of Clinical 90.

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 208
 Purvi Gandhi - Clinical Research Methodology

12. CenterWatch: Clinical Trials- Overview of Clinical 21. Bowalekar S. Biostatistics in Clinical Trials. In: Gupta
Trials. [online] [cited 2010 Apr 28]. Available from: SK, editor. Basic Principles of Clinical Research and
U R L : h t t p : / / w w w. c e n t e r w a t c h . c o m / c l i n i c a l - Methodology. New Delhi: Jaypee Brothers; 2007. p. 233-
trials/overview.aspx 52.
13. Reagan-Shaw S, Nihal M and Ahmad N. Dose translation 22. Hunt SA, Abraham WT, Mancini DM, Chin MH, Michl
from animal to human studies revisited. FASEB J. 2007; K, Feldman AM, Oates JA, Francis GS, Rahko PS,
22: 659-61 [online]. [cited 2010 April 22]. Available Ganiats TG, Silver MA, Jessup M, Stevenson LW,
from: URL: http://fasebj.org/cgi/reprint/22/3/659 Konstam MA and Yancy CW. ACC/AHA 2005 Guideline
Update for the Diagnosis and Management of Chronic
14. Understanding Clinical Trials [online]. updated 2007 Sep
Heart Failure in the Adult: a report of the American
20 [cited 2010 Apr 27]. Available from: URL:
College of Cardiology/American Heart Association Task
http://clinicaltrials.gov/ct2/info/understand
Force on Practice Guidelines (Writing Committee to
15. Vahle JL and Tashjian Jr AH. Clinical Drug Evaluation Update the 2001 Guidelines for the Evaluation and
and Regulatory Approval. In: Golan DE, Tashjian Jr AH, Management of Heart Failure) [online]. [cited 2007 Jul
Armstrong EJ and Armstrong AW, editors. Principles of 23]. Available from: URL: http://circ.ahajournals.org/
Pharmacology- The Pathologic Basis of Drug Therapy cgi/content/full/112/12/e154
2nd ed. New Delhi: Wolter Kluwer (India) Pvt. Ltd.;
23. Neutel JM and Smith DHG. Evaluation of angiotensin II
2008. p. 863-73.
receptor blockers for 24-hour blood pressure control:
16. POEMs and EBM Glossary: American Academy of meta-analysis of a clinical database. J Clin Hypertens.
Family Physicians- News and Publications [online]. 2003; 5(1): 58-63.
[cited 2007 Jul 21]. Available from: URL:
24. D'Agostino RB Sr and Massaro JM. New developments
http://www.aafp.org/online/en/home/publications/journ
in medical clinical trials. J Dent Res. 2004; 83(Spec Iss
als/afp/afppoems.html
C): C18-C24.
17. The Drugs and Cosmetics Act, 1940 [online]. [cited 2010
25. Pais-Ribeiro JL. Quality of life is a primary end-point in
Apr 23]. Available from: URL:
clinical settings. Clinic Nutrit. 2004; 23(1): 121-30.
http://cdsco.nic.in/html/Copy%20of%201.%20D&CAct
121.pdf 26. Montgomery DC. Introduction to statistical quality
control 4th ed. John Wiley and Sons Inc.: 2003. p. 98.
18. Abate MA and Hildebrand III JR. Clinical Drug
Literature. In: Beringer P, Der Marderosian A, Felton L, 27. Bolton S and Bon C. Pharmaceutical statistics: Practical
Gelone S, Gennaro AR, Gupta PK, Hoover JE, Popovick and clinical applications 4th ed. New York: Marcel
NG, Reilly WJ and Hendrickson R, editors. Remington- Dekker Inc.; 2004. p. 151-72.
The Science and Practice of Pharmacy 21st ed. New 28. Bolton S and Bon C. Pharmaceutical statistics: Practical
Delhi: Wolters Kluwer (India) Pvt. Ltd.; 2009. p. 74-86. and clinical applications 4th ed. New York: Marcel
19. Grossi PM. Study Design and Validity. [online]. 2004 Dekker Inc.: 2004. p. 1-30.
[cited 2007 Jul 21]. Available from: URL: 29. Motulsky H. Choosing a test. In: Intuitive Biostatistics.
http://dukehealth1.org/ surgery/documents /Clinical New York: GraphPad Software, Oxford University Press:
Studies. pdf [online]. 1995 [cited 2007 Jul 21]. Available from: URL:
20. Rosenbaum PR. Observational Study. In: Everitt BS and http://www.graphpad.com/www/book/ Choose.htm
Howell DC, editors. Encyclopedia of Statistics in 30. Patel BB, Tiwari S and Mishra B. Clinical research in
Behavioral Science. Chichester: John Wiley & Sons Ltd; India: An update. Ind J Pharm Pract. 2009; 2(2): 14-24
2005(Vol. 3). p. 1451–62. [online]. [cited 2010 Jul 29]. Available from: URL:
http://ijopp.org/ijpp_3rd_issue/review_articles2.pdf

Ind J Pharm Edu Res, Apr-Jun, 2011/ Vol 45/ Issue 2 209