PHAR 628 Research Methods & Biostatistics Lecture Notes PDF

PHAR 628 Research Methods & Biostatistics Lecture 6: Randomized Controlled Trial I Lixian Zhong, Ph.D. Feb 5th, 2025 1 Learning Objectives ❖ Define terms related to experimental study d...

PHAR 628 Research Methods & Biostatistics Lecture 6: Randomized Controlled Trial I Lixian Zhong, Ph.D. Feb 5th, 2025 1 Learning Objectives ❖ Define terms related to experimental study design - Variables: Dependent, Independent, Confounding - Random assignment - Control ❖ Identify the three characteristics of a true experiment 2 Research Design Classification Research Design Time Experimental Study Purpose Orientation Setting 1 Descriptive Analytical Prospective Retrospective Experimental Observational 3 Study Design Observational studies -Case report/series -Cross-sectional studies -Case-control studies -Cohort studies Experimental studies -Randomized Controlled Trials (RCTs) sometimes also referred to as Randomized Clinical Trials 4 Important Terminology: “Variables” Variables are characteristics that can vary. The “building blocks” of research questions and protocols. A variable represents a concept, factor, element or feature that can have more than one value. - drug dose - outcomes (e.g., weight, HbA1c level etc.) - … 5 Pose A Research Question and Hypothesis Research Question is a question that states the research problem to be studied. E.g., In patients with diabetes, do clinical services by pharmacists improve short- term clinical outcomes compared with traditional care in an outpatient setting? Hypothesis specifies an expected relationship that is being evaluated between intervention and outcome or two or more variables. E.g., Clinical services by pharmacists will improve short-term clinical outcomes in patients with diabetes in outpatient settings compared to traditional care. 6  Independent Variable A variable whose value is not dependent on that of another. It is a condition, intervention or characteristic that will predict, or cause a given outcome. An independent variable is the variable you have control over, what you can choose and manipulate. It is usually what you think will affect the dependent variable. 7  Dependent Variable The dependent variable responds to the independent variable. It is called dependent because it "depends" on the independent variable. The outcome variable is the dependent variable. It is a response or effect that is presumed to vary depending on the independent variable 8 Independent and Dependent Variables Cause Effect/outcome (independent (dependent variable) variable) 9 Can you identify the DV and IV? 10  Extraneous & Confounding Variables An extraneous variable is any factor that is not directly related to the purpose of the study, but that may affect the dependent variable. A confounding variable (also confounding factor) is an extraneous variable that correlates with both the dependent variable and the independent variable. 11 Confounding Variable Cause Effect/outcome (independent (dependent variable) variable) Other factors Other factors (extraneous (confounding variable) variable) 12 E.g. #1 As ice cream sales increase, the rate of drowning deaths increases sharply. Therefore, ice cream consumption causes drowning. E.g. #2 Since the 1950s, both the atmospheric CO2 level and obesity levels have increased sharply. Hence, atmospheric CO2 causes obesity. 13 Confounding Variable E.g. As ice cream sales increase, the rate of drowning deaths increases sharply. Therefore, ice cream consumption causes drowning. Ice cream sales Drowning deaths (independent (dependent variable) variable) Summer (high temperature) (confounding variable) 14 Confounding Variable E.g. Since the 1950s, both the atmospheric CO2 level and obesity levels have increased sharply. Hence, atmospheric CO2 causes obesity. Atmospheric CO2 obesity (independent variable) (dependent variable) Increased productivity /living standard (confounding variable) 15 A confounding problem in research Observational studies (such as the Nurses Health Study) suggested a cardiovascular protective benefit of hormone replacement therapy (HRT) for postmenopausal women. Randomized, controlled trials (such as the Women’s Health Initiative and the Heart and Estrogen/Progestin Replacement Study) pointed to no benefit, or even harm from HRT. 16 The Nurses Health Study 17 The Women’s Health Initiative 18 19 Which conclusion do you believe more? A. The Nurses Health Study B. The Women’s Health Initiative C. Both D. Neither 20 Confounding Variable Observational studies have demonstrated that women who have used postmenopausal hormone replacement therapy (HRT) are at reduced risk of coronary heart disease. Thus, HRT can reduce risk of coronary heart disease. HRT Reduced risk of (independent CHD variable) (dependent variable) Other factors? (confounding variable) 21 Confounding Variable Observational studies have demonstrated that women who have used postmenopausal hormone replacement therapy (HRT) are at reduced risk of coronary heart disease. Thus, HRT can reduce risk of coronary heart disease. HRT Reduced risk of (independent CHD variable) (dependent variable) Higher socioeconomic status (confounding variable) 22 Confounding Variable When confounding variables are not controlled, they exert a confounding influence on the independent variable – That is, they contaminate the independent variable in such a way that their separate effects are obscured. Experiments are designed to control for this type of confounding. 23 How to Design a Perfect Experiment The most rigorous form of scientific investigation for testing hypotheses. Randomized controlled trials, which follows a true experimental study design, are considered to be the gold standard for clinical research. The purpose of an experiment is to support a cause-and-effect relationship between a particular action or condition (independent variable) and an observed response (dependent variable). 24 The Women’s Health Initiative What does this mean?!? 25 Characteristics of Experiments To be considered a true experiment, a study must have three essential characteristics: 1. The independent variable must be manipulated by the experimenter 2. The subjects must be randomly assigned to groups 3. A control group must be incorporated within the design 26 Characteristics of Experiments To be considered a true experiment, a study must have three essential characteristics: 1. The independent variable must be manipulated by the experimenter 2. The subjects must be randomly assigned to groups 3. A control group must be incorporated within the design 27  Manipulation of Variables Manipulation of variables refers to a deliberate operation performed by the experimenter The experimenter imposes a set of predetermined experimental conditions (the independent variable) on at least one group of subjects - Usually administering the intervention to one group and withholding it from another. The experimenter can also manipulates the levels of the independent variable by assigning subjects to varied conditions (e.g. different drug doses) 28 Characteristics of Experiments To be considered a true experiment, a study must have three essential characteristics: 1. The independent variable must be manipulated by the experimenter 2. The subjects must be randomly assigned to groups 3. A control group must be incorporated within the design 29  Random Assignment Random assignment means that each subject has an equal chance of being assigned to any group Assignments will be independent of personal judgment or bias With randomization, different groups will achieve equivalence - inter-subject differences should balance out If we can assume that groups are equivalent at the start of an experiment, then we can have confidence that differences observed at the end of the study are not due to subject variability that existed before the experiment began 30 32  Characteristics of Experiments To be considered a true experiment, a study must have three essential characteristics: 1. The independent variable must be manipulated by the experimenter 2. The subjects must be randomly assigned to groups 3. A control group must be incorporated within the design 33 Control Groups Random assignment and the use of a control group is an effective way of ruling out extraneous effects Subjects in a control group may receive placebo, no treatment, different dose or standard care and will act as a basis of comparison for a new intervention To draw valid comparisons, we must be able to assume a reasonable degree of equivalence between the control and experimental groups 34 Control group 35  Placebo Placebo is a substance or treatment which has no active therapeutic ingredients E. g.: saline solution, lactose capsules Placebo may have a therapeutic effect based solely on the power of suggestion (placebo effect). Placebo Effect 36 Control Groups The use of a placebo/ no treatment control is sometimes unfeasible in clinical situations, for practical or ethical reasons Therefore, clinical researchers often evaluate a new experimental treatment against conventional methods of care – This does not diminish the validity or usefulness of the study, but it does change the question that can be asked of the data Instead of assessing whether the new treatment works, this approach assesses whether the new treatment is more effective than standard methods 37  Experiments Experiments are based on a logical structure, or design, within which the investigator systematically introduces changes into natural phenomena and then observes the consequences of those changes. The essence of an experiment lies in the researcher's ability to manipulate and control variables, so that rival hypotheses are ruled out as possible explanations for the observed response 38  Reality Check… In reality, of course, clinical experiments seldom have the ability to completely eliminate confounding effects Even though causality can never be demonstrated with complete certainty, the experimental method provides the most convincing evidence of the effect one variable has on another 39