Bcacp- Trial Design & Biostatistics Exam Questions And Answers 2024.

Bias in study design Systematic , nonrandom variation in study methodology and/or data collection process that results in an incorrect estimate of association between exposure and outcome affecting the true value of the result Ex of bias: - Selection bias: systematic selection of subjects that leads to an imbalance or an advantage in favor of one cohort over the other, leading to an error in the selection, or sampling of individuals for a clinical study -observational or information bias: incorrect determination of exposure and/or outcomes; can be random or systematic -recall bias: cases are more likely to remember exposures than control -publication bias: studies with positive results have a greater likelihood of publication than those with negative results -confounding variable: extraneous variable that influence both the dependent and the independent variable, affecting how the overall result can be interpreted Confounding in Study design A variable that affects the independent variable has not been separated from another and confounds it, altering the ability to determine the true effect on the measure outcome To miminimize the potential for missing a confounding variable, all relevant information should be collected and evaluated and subjects should be randomized, matched (when possible) and/or restricted to specific population. In the analysis perform stratified analysis, and propensity score matching - Multivariate analysis - Propensity score matching: variable that predict treatment assignment (not outcome assignment) create a propensity score. A propensity score is the probability that a subject with certain characteristics will be assigned to the tx group as opposed to the control group. Pairs of treated and untreated subjects are matched on the basis of similar values of the propensity score. Propensity score matching can be used to reduce selection bias in observational studies and compare outcomes across groups that have a similar probability of receiving the treatment of interest Causation (or a casual relationship) The following must be met in order to determine a causal relationship: 1. there is temporarily; the cause occurs before the effect relationship is biologically, and the likelihood of the event increase with the strength of the relationship Case Report/Case Series Describe the experience in the tx of a single patient or a small group of pts with an unusual disease, sx or a member of a population not usually affected -not a study technically Advantages & Disadvantages: - Hypothesis are formed: first step in describing an important clinical problem; easy to perform (inexpensive); may be the first indication of new risks, disease, trends - Difficult to generalize results: may not be truly characteristic of the disease; does not determine causality of association between exposure and outcome; needs to be further investigated and reproduced Cross-Sectional Study Design (or cross-sectional survey) Observational (non experimental), hypothesis-generating Identify the prevalence or characteristics of a condition in a group of individuals and evaluate associations Most surgery and mass screening programs are cross-sectional surveys Useful for measuring current health status and planning for some health services, including setting priorities for disease control Advantages/Disadvantages -Hypothesis are formed: relatively easy and quick; can be inexpensive, assist with health policy - limited information available: low response rate is common, can not establish temporal or causal relationships; reverse causality (can't whether the exposure preceded the outcome) Hierarchy of trial designs Systematic Review & Meta-analysis RCT Cohort Quasi-experimental Case control Cross sectional Case report, case series, qualitative studies Case-control study Observational, hypothesis-generating study design: in specific circumstances; case control studies can be hypothesis generating - start with the outcome of interest Typically retrospective Case control study provides an efficient means to determine the association (or exposure) between the risk factor and the outcome of interest. Study groups are selected on the basis of outcome and evaluate the past exposure to possible RF Data on potential exposure to RF are collected by review of the health record or by interviewing subjects A case-control study cannot determine the actual risk of the outcome (actual risk is the relative risk [RR]). Can only determine an estimate of the risk (OR is an estimate of the risk [the odds ratio] - in some cases the OR can approximate the RR - the OR is interpreted as the odds of exposure to a factor intros with a condition of disease compared with those who do not have the condition of disease. This will be discussed more in the text that follows: Cases with the disease (or outcome of interest) should be representative of all cases in the population, and study inclusion should be independent or exposure status (exposure to past RF) The control group should be selected from the same source population without the disease, the source population is the underlying cohort before acquisition of the outcome of interest Results are analyzed on the basis of the difference in exposure between cases and controls (cohort study evaluates differences in outcomes between case and controls) Advantages/Disadvantages - quick/inexpensive can evaluate many hypotheses (or exposure) for a cause for a disease; rare diseases and diseases with long latencies can be studied; when posing a group to an intervention may be an ethical issues - data quality; matching is often flawed; indirect measure of risk (OR Cohort Study Observational, hypothesis-generating or hypothesis testing A cohort study identifies the relationship between exposure and outcome Uses a defined study population without the outcome of interest at study initiation. This is called a "healthy cohort". Cohort may have other disease but not the disease of interest. The cohort should be eligible to develop the outcome Subjects are followed over time to try to correlate exposure (or RF) and disease occurrences. Study groups are usually defined by their exposure and the outcome is compared Allows for an estimate of ten risk of the outcome (RR between the exposure groups) in those with and without the exposure of interest. RR is the risk of someone developing a condition when exposed compared with someone who has not been exposed Most high-impact cohort studies are prospective. Retrospective cohort studies assemble the cohort by collecting historical data on RD and then evaluating the data for outcomes. Retrospective cohort studies are not preferred, but they can be used if good, comprehensive records are available Advantages/Disadvantages - data collection is typically standardized, and outcomes can be verified (minimal recall bias); risk estimates are a true value for the group studied; many different outcomes can be studied; can investigate a rare exposure; exposure comes before disease (can establish temporarily); strongest design to establish cause-and-effect relationship for risk - Only risks evaluated at baseline can be used; can't investigate a rare disease (can establish temporality) strongest design to establish cause-and-effect relationship for risk - ex:framingham study- the original question was to identify the RF for and characteristics of heart disease and related outcomes. The original cohort was the population of adult male subjects residing in framing Incidence Measure of the probability of developing a disease or a condition in a population during a specified time -incidence time: important in monitoring the course of an epidemic and the success of prevention and intervention measures -incidence proportion: measure proportion of an initially disease-free population that develops disease (or becomes a "case") during a specified period (attack rate); the number of new cases during specified period/size of population at start of period -incidence rate: measure of incidence that incorporates them directly into the denominator; number of new cases during specified period/totaled time each person was observed. An incidence rate describes the rapidity of disease occurrence for newly dx case; typically reported as cases/person-time Prevalence Proportion of the population with the condition/disease at any given time -# of individuals with dx/ # in the population x 100% Incidence vs. Prevalence Incidence vs prevalence: Prevalence includes all cases in which incidence is limited to new cases Incidence: measure the occurrence of new cases of a disease (or event) during a time interval Prevalence: measure the occurrence of new and existing of disease (or event) at a specified point or period time Relative Risk Compares the incidence of one group with the incidence of another group - best estimate of the strength or magnitude of the association between exposure and outcome •The risk in the treatment group divided by the risk in the control group (i.e., the risk of the treatment group relative to the control group) Risk among treated = a/(a+b) Risk among control = c/(c+d) RR=ratio of risk among treated to risk among control = [a/(a+b)]/[c/(c+d)] Interpretation of RR: •RR 1 indicates "risk" is lower among "treated" than "control" •RR 1 indicates "risk" is higher among "treated" than "control" •RR = 1.28 ØPatients receiving 300 mg bupropion were 1.28 times more likely to quit smoking by month 3 compared to patients receiving 150 mg bupropion ØPatients receiving 300 mg bupropion were 28% more likely to quit smoking by month 3 compared to patients receiving 150 mg bupropion Relative risk reduction: another way to report RR; RRR= 1- RR x100%