Title | Epi Test 2 Study Guide - Dr. Merrell |
---|---|
Author | Sarah Naeher |
Course | Epidemiology |
Institution | James Madison University |
Pages | 12 |
File Size | 113.3 KB |
File Type | |
Total Downloads | 69 |
Total Views | 158 |
Dr. Merrell...
HTH 450 Exam 2 Study Guide Mortality Rate Adjustment · Purpose ● Can use mortality data to compare two or more populations, or the same population, during different time periods. ○ We know that different groups have different risks for dying. ○ Sometimes overall mortality rate (crude or unadjusted mortality) can mask these differences. · Adjusted Rates ● A rate of mortality or morbidity in a population in which statistical procedures have been applied to permit fair comparisons across by removing the effect of the differences in the composition of various populations. ○ 90% of the time it has to do with age. · Direct Adjustment calculation ● We apply the mortality rates of our populations of interest to a third hypothetical population. ● **Look at practice worksheet** · Indirect Adjustment calculation ● ● ●
We use indirect adjustment when the age (or other characteristics) of specific mortality ranges are not available or when they are really small. **Look at practice worksheet** SMR = (observed total deaths / total expected deaths) x observed deaths.
Cross-sectional Study Design · What can it do and not do ● Helps to investigate causes of disease. ● Study design looks at both the exposure and outcome at the same time. ○ Collect data about both things at the same time. ● Gives a snapshot of a population. ○ We can identify prevalent cases: we know they existed at the time of the study. ○ We cannot identify incidence cases or exposure: we don’t know when the condition started. ● To see an association, there are two approaches: ○ Calculate the prevalence of the disease/outcome ○ Calculate the prevalence of the exposure/risk factor · Set up of 2x2 table · Issues with cross-sectional studies ● ●
Generalizability: ○ They don’t represent all people with that disease/illness Identifying prevalent cases could exclude people who already died from the
●
disease. ○ Because both the exposure and outcome are determined at the same time, it's hard to determine the temporal relationship. Not enough to determine causality. ○ Only mathematical equations.
· Benefits of cross-sectional studies ● ●
Can be very suggestive of a possible exposure or risk factor for a disease. Easy to design, execute, and are cheap. ○ After a cross-sectional study, the next step would be to determine if there is a temporal cause.
Case-Control Studies · What is the purpose of case-control studies? ● Used to examine the possible relationship of an exposure to a certain disease. ○ We identify a group of individuals with that disease or outcome of interest (cases). ○ For comparison, a group of people without that disease or outcome of interest (controls). ● If there is an association between the exposure and disease: ○ The prevalence of exposure should be higher in the cases than the controls. o How are they designed ● Defining cases: ○ How will you define the cases (disease/condition)? ○ Cases should be representative of all those diseased. ○ Correctly defining cases helps us to select cases and limit bias. ● Selecting cases: ○ Selecting cases is important because it can introduce bias. · Who are cases and who are controls? o Issues with selecting cases ● Selection bias of cases: ○ From hospitals: ■ Risk factors unique to the hospital due to referral patterns or other factors. ● Some tertiary facilities selectively admit severely ill patients. ● Risk factors: only individuals with severe forms of the disease. ● Not generalizable to all patients with the disease. ■ If hospital cases are used, it is desirable to select the cases from several hospitals in the community.
○
Using incidence or prevalence cases: ■ Problems with using incidence cases: ● Must often wait for new cases to be diagnosed. ● Prevalent cases have already been diagnosed and represent a larger number of cases available for the study. ■ Problems with using prevalent cases: ● Risk factors that we identified in a study using prevalent cases may be related to survival with the disease rather than to the development of the disease. ● Likely to include long-term survivors. ● Non-representative group of cases. o Issues with selecting controls ● Non-hospitalization population: ○ School rosters, selective service lists, insurance company lists, residents of a defined area, random digit dialing. ● Hospital Selection: ○ Hospital inpatients are a “captive population” and are clearly identified. ○ Selecting from hospitals may mean that controls don’t represent the general population. § Multiple controls vs. controls of the same type ● Same type: ○ All controls come from the same source. ● Different types: ○ “Neighborhood” controls ○ Used when we’re concerned about the controls not being representative of the rate of exposure that is expected in the population of non-diseased persons. o Issues with ascertaining (measuring) the exposure ● Limitations of recall: ○ Data collected from subjects by interviews. ○ Humans are limited in their ability to recall information. ○ People may also not have the information being requested. ○ Causes misclassification of exposure: ■ Cases or controls who were actually exposed will be classified as unexposed. ■ Leading to an underestimate of the true risk of a disease associated with an exposure. ● Recall Bias: ○ More serious than information recall ○ Differential recall of important exposed between cases and controls.
■
Can artificially suggest a relationship between exposure and outcome that is not there or is not strong.
· Matching o Group vs. Individual ● Group matching (frequency matching): ○ Selecting controls so they have the proportion of a characteristic identical to that of the case group. ○ Requires all cases are selected first, so that proportions can be calculated before selecting controls. ● Individual matching (matched pairs): ○ For each individual case selected, a control is selected who is similar to the case in terms of the specific variable or variables of concern. ○ Often used for hospital controls. o Problems with matching ● Matching cases on too many characteristics makes it difficult to find controls. ○ Gender, age, race, education, income, geography ● Once we have matched controls to cases to a given characteristic, we can’t study that characteristic. ○ Why? ■ Matching creates an artificially identical proportion of a factor in the control group. We can’t look at a factor whose proportion has been artificially established. · Advantages and Disadvantages of Case-Control Studies ●
●
Advantages: ○ Can be used to study low-prevalence conditions. ○ Relatively quick and easy to complete. ○ Inexpensive ○ Smaller number of subjects ○ Multiple risk factors can be assessed at the same time Disadvantages: ○ Measurement of exposure may be inaccurate ○ Representativeness of cases and controls may be unknown ○ Only provides indirect estimate of risk ○ Data quality may be a factor (information bias) ○ The temporal relationship between exposure factors and outcome cannot always be ascertained.
Cohort Studies · What is the purpose of a cohort study? ● Investigators selects a group of exposed individuals and a group of nonexposed individuals. ● Follows them over time to see who in each of the groups develops the
disease. Compares the two groups based on exposure. o Design of a cohort study (type) ● Prospective study (concurrent cohort): ○ Investigator identifies the original population at the beginning of the study. ○ Exposure and non-exposed are ascertained during the study. ○ Groups are then followed (concurrently) into the future, and incidence is measured. ● Retrospective study (historical cohort): ○ Uses historical data from the past to obtain our results sooner. ■ No longer prospective because we’re beginning with a pre-existing population to reduce the duration of the study. ■ Exposure is ascertained from past records and outcome ascertained at the same time the study began. · Incidence rate and cohort studies ● Boxes A & B represent all those exposed. ○ Incidence of the disease in the exposed group is expressed as A/ (A+B) x 10^n. ● Boxes C & D represent all unexposed. ○ Incidence of the disease in the unexposed group is expressed as C/ (C+D) x 10^n. ● Incidence in entire population: ○ (A/C) / (A+B+C+D) x 10^n · Selection of study participants ● Option 1: ○ Create a study population by selecting groups for inclusion in the study on the basis of whether or not they were exposed. ■ Ex: find equal numbers of people with good and bad flossing and mouth washing habits and follow up with them for a whole year. ● Option 2: ○ Select a defined population before any of its members become exposed or before their exposures are identified. ○ Select a population on the basis of some factor not related to exposure. ■ Ex: Framington Study · Problems (Bias) with cohort studies o Selection Bias ● Nonparticipant bias: ○ Were the people who participated in the cohort study ●
different than those who didn’t? Nonresponse bias: ○ Study participants are unwilling to participate in all of the study procedures. ■ Ex: surveys where questions are skipped. ● Loss to follow-up (attrition): ○ Participants who at one time were actively participating in the study, but have become “lost” at the time of follow-up. ○ TOP Project (Teen Outreach Project): ■ Students move schools or leave the state. o Information/Investigator Bias ● Quality and extent of information obtained is different for exposed person’s than for unexposed person’s. ○ Likely to happen in retrospective cohort studies. ● Quality of information is an issue and is understandable to a degree. ○ There should be no difference in information between exposed and unexposed individuals. · Famous cohort studies ●
●
●
Prospective: Farmington Study (1948) ○ Took a sample of 5,127 men and women, who did not have heart disease at the time the study began. ■ Exposures: smoking, obesity, HBP, elevated cholesterol, low levels of physical activity. ○ Coronary events were identified by: ■ Examining the study population every two years. ■ Daily surveillance of hospitalization at the only hospital in Farmington. ○ Used second option for selecting the population: ■ Population was selected on the basis of location of residence or other factors not related to the exposures in question. ■ Population was observed over time to determine which individuals developed or already had the exposures of interest. ■ Advantage: ● Permitted the investigators to study multiple exposures at one time. ● Complex interactions among those exposures by using advanced statistical procedures. ○ Origin of the word ‘Risk Factor’ Retrospective: Breast Cancer & Progesterone Deficiency (1978) ○ Breast cancer most common in women who are older at the time of their first pregnancy. ○ Cowan identified a population of women who had been patients at Johns Hopkins Infertility Clinic from 1945-1965 ■ All had a late first pregnancy ■ Patients were divided into exposure groups based on their past medical records (hormonal abnormality vs. normal hormones).
○
Results showed that women who had hormonal abnormalities had a 5.4 times greater risk of developing premenopausal breast cancer.
Estimating Risk · Measures of Risk o Odds Ratio ● Compares the odds of exposure in the cases to the odds of exposure in the controls. § Calculation, Interpretation ● (A/C) then (B/D) then (A/C)/(B/D) ● OR = 1: exposure does not affect the odds of the disease occurring. ● OR > 1: exposure is associated with higher odds of the disease occurring (exposure is a high risk). ● OR < 1: exposure is associated with lower odds of the disease occurring (exposure is a protective factor). ● Interpretation: ○ Those with a flu infection have a (1 - 0.6 = 0.4) 0.4 lower odds of having had a flu vaccine than those without the flu infection. o Absolute Risk ● The incidence of a disease in a population. ● Indicates the magnitude of the risk in a group of people with all exposures. ● Since it doesn’t take into consideration the risk of disease in unexposed individuals, we cannot indicate whether the exposure is associated with an increased risk of disease. ● Has important health implications in both clinical and public health policy. ○ Though absolute risk doesn’t stipulate any explicit comparison, an implicit comparison is often made whenever we look at the incidence of a disease. o Relative Risk Ratio ● Probability of an event (developing a disease) occurring in exposed people compared to the probability of the event in unexposed people. § Calculation, Interpretation ● RR = (incidence in exposed) / (incidence in unexposed) ● RR = (A / (A+B)) / (C / (C+D)) ● Interpretation: ○ RR = 1: the risk in exposed persons equals the risk in unexposed. No relationship. ○ RR > 1: the risk in exposed persons is greater than the risk in unexposed persons.
■
○
○
Evidence of a positive association and may be a causal. RR < 1: the risk in exposed persons is less than the risk in unexposed persons. ■ Evidence of a negative association and may be a positive effect. Ex: Those who had a flea bite had 2.7 times the risk of developing the plague than those who did not have a flea bite.
o Attributable Risk ● The amount or proportion of disease incidence (or disease risk) that can be attributed to a specific exposure. ● Addresses how much of a risk (incidence) of disease can we hope to prevent if we are able to eliminate the exposure to the agent in question. · When/which to use for the study design? ● AR for exposed groups: ○ Risk of disease is higher in exposed. ○ Unexposed people have some risk. ■ Risk of a disease is not zero even in the unexposed persons. ■ Risk is called background risk: everyone shares the background risk, regardless of whether or not they have the exposure in question. ○ If we want to know how much of the total risk in exposed persons is due to the exposure, we should subtract the background risk from the total risk. ● Population Attributable Risk (PAR) ○ What proportion of the disease incidence in the total population (both exposed and unexposed) can be attributed to a specific exposure? ○ Incidence in total population - incidence in unexposed group ○ PAR Percent: ■ (ITP - IUG) / (ITP) · What can this tell us? ●
Measure of the strength of association and the possibility of a causal relationship. ○ AR expresses the most we can hope to accomplish in reducing the risk of disease if we completely eliminate the exposure of interest.
Randomized Controlled Trials · Purpose ● Evaluating new drugs and medical technology. ● Assessment of screening programs. ● Assessment of intervention screening programs.
● ●
Evaluation of new ways of delivering health services. The ultimate objective is to generalize the results beyond the study population itself. ○ Generalizability (aka External Validity) ■ Ability to apply the results obtained in our study population to a broader population. ○ Internal validity: ■ How well an experiment is done, especially if it avoids confounding. ■ The less change of confounding in a study, the more internal validity in a study. o Pros ● Prevent potential biases on the part of the investigators from influencing the assignment of participants to different treatment groups. ○ Investigators make no decisions about which individuals are assigned to which groups. o Cons · History ● Ambrose Pare ○ Unplanned trial ○ Cauterization and tropical treatment ● James Lind ○ Conducted the first ever planned clinical trial ○ Scurvy in sailors (made different juices) ○ Advanced: preventive medicine and nutrition ● Historically used more in psychology, education, and agriculture. ● First RCT was published in 1948 by Austin Bradford Hill. ● Common / golden standard · Design ● Begin with a defined population. ● Randomized to receive new treatment or current treatment. ● Follow subjects in each group to see how many are improved in the current treatment group. ○ If new treatment is associated with a better outcome, we would expect to find better outcome in more of the new treatment group than the current treatment group. · Selection of participants ● Criteria for those who will/will not be included must be specific. ○ Key for replication ● Inclusion criteria: ○ Characteristics that prospective subjects must have if they are to be included in the study. ● Exclusion criteria:
○
Characteristics that disqualify prospective subjects from participating in the study. ● Common inclusion/exclusion criteria ○ Sex, age, race/ethnicity, type and stage of disease, previous treatment history, presence or absence of other medical, psychological, or emotional conditions. · Studies without comparison ● Important to derive causal inferences regarding relationships between treatment and outcome. ● Participants are assigned to the treatment group by a method that is not random. · Randomization ● Process by which each subject has the same chance of being assigned to either intervention or control. ● Critical element of randomization is unpredictability of next assignment. o Why? ● Increases comparability between groups in regard to characteristics that we are concerned about. ○ Randomization is not a guarantee of comparability, since chance plays a role in the process of random treatment assignment. o Stratified Randomization ● Divide the population into groups that differ in important ways. ● Select random sample from within each group. ● Basis for grouping must be known before sampling. · Data collection ● Important that we have quality data collection in both the treatment and control groups. ● Outcome: ○ Measuring the outcome of interest in a study is partially important. ○ Outcome criteria need to be specifically stated from the beginning. o Blinding ● Don’t want participants to know which group they’re assigned to. ● Can mask participants using a placebo. ○ Used to study rates of side effects. ● Double blinding: when we mask both the researcher and the participant. ● Feasibility: ○ Ethics: double-blind procedure should not result in harm. ■ May be unethical to give repeated injections of a placebo to a control group. ○ Practicality: may be impossible to blind some treatment. ■ Ex: how do you create a placebo for radiation treatment?
○
Avoidance of bias: ■ Considers the sources of bias to decide if the reduction in bias is worth the effort.
· Types of RCTs ● Observational: ○ Researchers observe individuals and record information. ● Experimental: ○ Researchers intentionally impose treatments on individuals. ● Historical Control Study: ○ When a disease is uniformly fatal and a new drug becomes available. ● Factorial Design: ○ Use the same study if we want to look at more than one treatment. · Recruitment and Retention of Participants ● Increasing sample size, increases power. ● Participants must we willing to be randomized. ● Must ensure there is no coercion from the investigator. ● Participants must be informed of risks. · Expressing the results of RCTs o Efficacy ● The extent of the reduction in the disease caused by the use of the treatment in a trial. ● Efficacy = ((rate in control group - rate in the treatment group) / (rate in the control group)) x 100 o Effectiveness ● The ability of an intervention to have a meaningful effect on patients/individuals in normal/community conditions. ● Need to consider how well a treatment works under “ideal” conditions may be very different from how well a treatment works in “real life” situations. · Clinical trials in the US ● Phase 1 trials: ○ Clinical pharmacologic studies ■ Toxic and pharmacologic effects are examined, including safety, safe ranges of human dosage, and the side effects observed with the new drug. ● Phase 2 trials (weeks to months): ○ Consists of clinical investigators of 100-300 patients in order to evaluate its relative safety. ● Phase 3 trials (several years): ○ Large-scale RCTs for effectiveness and relative safety. ■ Recruiting large numbers of participants may be difficult and often necessitate recruiting from more than one study center. ■ When recruitment difficulti...