Research Methods Notes - Karina Wilson.pdf

Research Methods Notes: Chapter 1 - Selecting a Research Topic & Focusing a Research Question: Health = state of complete physical, mental and social well-being, not merely the absence of disease or infirmity Determinants of Health = biolo...

Research Methods Notes: Chapter 1 - Selecting a Research Topic & Focusing a Research Question: Health = state of complete physical, mental and social well-being, not merely the absence of disease or infirmity Determinants of Health = biological, social, environmental, political and other factors that influence the health status of individuals and populations Health Research = investigation of health/disease or any factors that contribute to presence of absence of physical, mental and social health among individuals, families, communities, nations or the world populations as a whole ○ Involves both clinical and population health research ○ If we understand the cause, we can intervene and create treatments Clinical Research = investigates how to prevent, diagnose and treat health issues ○ Ex. examine progression of disease over time; compare efficacy of medication treatments; test safety of new diagnostic tests, medications, or medical devices Longitudinal studies in healthy brain health → following large groups of people from middle age tracking physical activity to see how that could prevent cognitive impairment Alzheimer’s - APOE gene = risk genotype for alzheimer development (genetic component but also pairing environmental factors) ○ Can be physicians, specialists (connected to a research hospital) ○ Generally uses humans as the unit of interest Basic Medical Research = studies molecules, genes and cells and smaller biological components related to human function & health ○ Ex. dopamine transporter and the brain & how that influences human health Translational Research = bridges basic (microscope, lab) and clinical research (cells & molecules as it relates to a disease/disorder) to improve clinical outcomes ○ Bridges the gap between bench (basic) and bedside (clinical) ○ Aims to move research from lab to clinical settings Population Health/Health Research = examines health outcomes at the community, regional, national and worldwide levels (determinants or “causes”; health outcomes) ○ Includes evaluation of: biology, immunology, nutrition, genetics; social & psychological science; economics, politics ○ Often uses tools from demography → study of the size and composition of populations and of population dynamics such as birth and death rates Medicine = process of preventing, diagnosing and treating health problems ○ Clinicians → provide direct care to patients + clients Public Health = actions taken to promote health and prevent illness The Research Process: Research = process of investigating a topic to discover insights about the world Steps involve: ○ Identifying a study question (assignment 1 - individual) ○ Selecting appropriate study approach ○ Designing a study and collecting data ○ Analyzing data ○ Report findings Identifying a Research Question: 1. Your interests (personal life) 2. Your aptitudes (knowledge, skills, etc.) 3. Applications (what gaps in the literature can be filled?) 4. Mentors (practical: what lab can I begin with?...Often depends on supervisor and access to populations and resources and funding) Chapter 2 - Selecting a Research Question: Practical Research: When selecting a research question, consider the who, where, what, when, why and how as those often indicate unmet demands for descriptive studies, needs assessments, program evaluations, clinical effectiveness studies and other types of health-related research Brainstorming & Concept Mapping: Brainstorming: process of generating long lists of spontaneous ideas/topics about possible research questions ○ Not the stage of eliminating ideas ○ Ideas don’t need to be yet well formed ○ Consult colleagues, journals, articles, books Concept mapping: visual method for listing ideas then grouping them to reveal relationships ○ First: brainstorm to generate list of words/phrases that describe topics ○ Second: circle related ideas and group things together ○ Note: similar process can be used as a part of qualitative data analysis Keywords: Keyword: word or short phrase that can be used in a database search MeSH (Medical Subject Headings): vocabulary thesaurus used for searched of MEDLINE and other health science databases ○ Developed by the US National Library of Medicine part of NIH ○ Helps to narrow the scope of the research inquiry Exposure, Diseases and Populations (EDPs):  Exposure: personal characteristic, behaviour, environmental encounter, or intervention that might change the likelihood of developing a health condition Risk factor: exposure that increases an individual’s likelihood of subsequently experiencing a particular disease or outcome Protective factor: exposure that reduces an individual’s likelihood of subsequently experiencing a particular disease or outcome Non-modifiable risk factor: risk factor for a disease that cannot be changed through health interventions Modifiable risk factor: risk factor for a disease that can be avoided or mitigated → enables effective preventative interventions to be developed Three levels of prevention: ○ Primary prevention → health behaviours + protective actions that help keep an adverse health event from occurring in people who do not already have the condition (ie. nutritious diets, exercise, sleep, vaccinations ○ Secondary prevention → detection of health problems in asymptomatic individuals at an early stage when the conditions have not yet caused significant damage to the body and can be treated more easily (ie. cancer screening, blood pressure checks, routine vision and hearing tests) ○ Tertiary prevention → interventions that reduce impairment, minimize pain and suffering and prevent death in people with symptomatic health problems (ie. rehabilitation, palliative care, medications and surgery) Outcome: observed event such as the presence of disease in a participant in observational or experimental study ○ Often the outcome studied is the disease = presence of signs & symptoms of poor health, a pathophysiological condition ○ Disorder = functional impairment that may or may not be characterized by measurable structural or physiological changes Outcome of interest associated with a disease could be mortality (deaths), morbidity (nonfatal illnesses), or comorbidity (two or more adverse health conditions occurring at the same time) Population = group of people, communities or organizations Picot: Evidence-based medicine (EBM): uses results of rigorous research studies to optimize clinical decision making ○ Enables skilled clinicians to integrate the best research into their assessments of the most appropriate ways to care for individual patients ○ Similar practice of evidence-based practice is used in a variety of fields to encourage experienced professionals to integrate research into their decision-making process PICOT is used to optimize a research question: ○ P = patient or population group and problem that will be studied ○ I = intervention that will be tested ○ C = what will the intervention be compared to? ○ O = what is the outcome of interest? ○ T = time-frame for follow-up Research to Inquiry + Testable Questions: Inquiry: process of finding answers to questions that arise from personal experiences ○ Grounded in curiosity about a problem or idea Testability: ability of a research question to be answered using experiments or other types of measurements ○ Should be testable Chapter 3 - Reviewing the Literature: Statistical Reports: Statistics such as prevalence of an exposure in a country, incidence of a disease or demographics of a pop. can help define certain exposures Reports from national governments → provide info on demographics, socioeconomics and environmental data For health-specific data, national health agencies may publish routine epidemiological updates & reports about vital statistics = population-level measurements related to births, deaths and other demographics characteristics Abstract Databases: Abstract = one paragraph summary of an article, chapter or book ○ In health sciences → abstracts provide a brief description of the study methods, study population, key findings and main conclusion or implication of the study Abstract database → collection of abstracts that allow researchers to search for articles using keywords or other terms ○ Most popular public health sciences database = pubmed (30 million abstracts) Critical Reading: Once acquiring a full-text article, a plan of action should be: ○ Reread abstract ○ Look at tables and figures that display most important results ○ Read entire text of article ○ Review the reference list for any additional sources that should be read Critical reading involves: ○ Asking questions about internal validity = evidence that a study measured what it intended to measure → considers how well a study was designed, conducted, interpreted and reported ○ Questioning the external validity = likelihood that the results of a study with internal validity can be generalized to other populations, places and times ○ Generalizability = results of one study are considered to be applicable to a broader target audience Annotated Bibliographies:  List of publications that includes a full reference for each document being reviewed, a brief summary of the article or report and a note about the resource’s potential relevance to the new study ○ Involves summarizing the content in that article most pertinent to the new investigation What makes research original? Originality = aspects of a new research project that are novel and will allow it to make a unique contribution to the health science literature ○ Needs to have 1 main difference from the previous work ○ Does not require that one identify a new exposure or new disease After identifying a novel research question, a replication study can also be considered ○ Replicability = study protocol implemented in a new study population should generate results similar to those of the original study as long as the exact same protocol is used ○ Replication study = repeat a study protocol in a new pop. to confirm original findings were not due to chance They use the same protocol for collecting data about a selected exposure and outcome but implement the study in a new pop. Chapter 4 - Defining Specific Aims: Refining the Study Question + Create a Study Goal: The principle of purposiveness states that research projects should be designed to answer one well-defined research question Study goal: describes the single overarching objective of a research project or the main question that a project seeks to answer Specific Aims for the Study: Specific aim: described action that will help the researcher make progress toward achieving the big-picture goal ○ Most studies have 2-4 specific aims ○ Enumerated items should take the form of a question or a “to” statement ○ For observational studies, the specific aims need to be testable ○ For experimental studies, the specific aims may take the form of a series of hypotheses that will be tested Hypothesis: informed assumption about the likely outcome of a well-designed investigation that can be tested using scientific methods (uses an “if-then-because” format) Specific aims can be written in a sequential (chronological actions list of actions that will achieve the main goal) or independent format (specific aims are related but independent of each other) Characteristics of a specific aim are summarized by this acronym: ○ SMART (S = specific, M = measurable, A = attainable (achievable), R = relevant (realistic), T = timely (time-bound)) Conceptual & Theoretical Frameworks:  Conceptual framework: model that a researcher sketches using boxes and arrows to illustrate the various relationships that will be evaluated during a study ○ Ie. directed acyclic graph uses nodes (vertices) and arrows (edges) to illustrate hypothesized causal pathways from distal exposures (determinants) to proximal exposures (immediate outcomes) to outcomes Theoretical framework: set of models in the published literature that can inform the components and flows of the conceptual framework for a new research study ○ Most common in nursing, social science and educational research literature Health belief model: considers health behaviour change to be a function of perceived susceptibility to an adverse health outcome, perceived severity of the disease, perceived benefits of behaviour change, perceived barriers to change, cues to action, and self-efficacy Social ecological model: considers individual health and health behaviours to be a function of the social environment, which includes individual, interpersonal, institutional, community and public policy dimensions Feasibility: Feasibility study: evaluation of the likelihood that a task can be completed with the time, money, tech, and other resources that are available for the activity ○ Valuable for making decisions about whether to pursue a research idea Characteristics of a viable research project use the acronym FINER: ○ F = feasible, I = interesting, N = novel, E = ethical, R = relevant Chapter 7 - Overview of Study Designs: Primary, Secondary & Tertiary Studies: First step in choosing a research approach is deciding if new data will be gathered Primary study: collects new data from individuals ○ Allows researchers to design studies that will answer their preferred research questions ○ May take months or years of recruitment and data collection ○ Risk of failure if participants cannot be recruited ○ Can become expensive, resource constraints Secondary study: analyzes an existing data set or existing health records ○ Research may be able to move quickly from the definition to the analysis of a study Tertiary study: reviews and synthesizes the existing literature on a topic ○ Allow a researcher to move quickly to the analysis stage Observational & Experimental Studies: To select a appropriate design for a primary study, must determine whether the study can be observational or experimental Experimental study (aka intervention study): assigns participants to a specific exposure ○ Used to determine whether an intervention is effective ○ Used to test the efficacy of new medical vaccines, devices, pharmacological products ○ Used to test effectiveness of preventative interventions, diagnostic methods, treatments ○ Randomly assign some to intervention and others to comparison groups  Observational study: does not intentionally expose participants to an intervention or ask any participants to change their behaviours ○ Participants report perceptions + experiences ○ Used when the goal of the research is to describe the health profile of a population or to examine a certain exposure if associated with an increase risk of a health outcome Epidemiology: study of distribution and determinants of health in human populations (most are observational studies) ○ Descriptive epi: quantify how often health-related exposures and outcomes occur in a population Used to characterize person, place, time associated with a health outcome (what, who, where, when) ○ Analytic epi: seek to identify the risk factors for various health outcomes or to test effectiveness of interventions to improve health status (why, how) Types of Observational Studies: Case-Control Study: Purpose: compares individuals with a particular condition (cases) to those without the condition (controls) to identify potential causes or risk factors Retrospective: looks back in time to assess exposure to risk factors Advantages: efficient for rare diseases or outcomes, relatively quick and inexpensive, can study multiple risk factors for one disease Disadvantages: participants may not accurately remember past exposures, cannot determine incidence or prevalence (only odds). time relationship between exposure and outcome can be unclear Cross-Sectional Study: Purpose: measures both exposure and outcome simultaneously in a specific population at a single point in time Advantages: quick and inexpensive, useful for assessing prevalence and relationships between variables, can generate hypotheses for further studies Disadvantages: cannot determine cause and effect (only associations), does not account for changes over time, susceptible to selection bias and confounding Cohort Study: Purpose: follows a group of people over time to study the development of disease or outcomes in relation to exposures Prospective (forward-looking): follows individuals over time from exposure to outcome Retrospective (backward-looking): looks at historical data to assess past exposures and outcomes. Advantages: can determine incidence, can study multiple outcomes for a single exposure, less recall bias compared to case-control. Disadvantages: time-consuming and expensive, loss to follow-up can bias results, inefficient for studying rare diseases Chapter 9 - Cross-Sectional Studies (snapshot of the health status of a pop at one point in time) Overview: Cross Sectional study (prevalence study) = measures the prevalence of the exposure/disease in proportion of a population at a specific point in time. ○ Uses: describes communities, assess population needs, supports program planning, establishes baseline data for longitudinal studies ○ Objective: describes exposure and/or disease status of a population ○ Population: Study participants must represent the source population. Representative Populations: A large, well-defined pop must be sampled (usually a few hundred people) and completes a questionnaire Participants must reflect the diversity of the source population. Cannot generalize results from a specific group (e.g., friends or attendees of an event) to the entire population KAP Surveys: KAP Survey: A tool used in cross sectional studies that asks about Knowledge, Attitudes (beliefs/perceptions), and Practices (behaviors) → identifies gaps between knowledge and action ○ Adults may know about the benefits of exercise but may not act on this knowledge due to perceived barriers Repeated CrossSectional Surveys: Repeated Cross Sectional Study: resurveys new representatives from the same source population over time (not the same individuals) ○ Shows population level trends but not individual level changes ○ Often used in health surveillance programs (e.g., CDC's BRFSS, NHANES) Other Types of Surveillance: ○ Active Surveillance: Public health officials ask clinicians about disease diagnoses. ○ Passive Surveillance: Medical labs submit reports on notifiable diseases. ○ Syndromic Surveillance: Tracks disease outbreaks based on reported symptoms. ○ Sentinel Surveillance: High Quality data from a few clinics/hospitals to detect larger population trends Prevalence: Prevalence: Measures the proportion of a population with a certain trait at a specific time. Point Prevalence: Snapshot of a population's status at one point in time (e.g., percentage of smokers on July 1, 2020). Period Prevalence: Proportion with a characteristic over a period (e.g., percentage of students with asthma in the past 6 months). Comparative Measures: ○ Prevalence Ratio (PR): Compares prevalence rates between independent populations (e.g., smokers vs. nonsmokers) Chapter 11: Cohort Studies Overview: Cohort: a group of similar people followed through time together Cohort Study: observational study following people forward in time so the rate of incident (new) cases of disease can be measured Key Characteristics of Cohort Studies: Cohort studies have at least two measurements times: ○ An initial survey determines the baseline exposure and disease status of all participants ○ Baseline: an initial measurement used as a benchmark for examining changes over time One or more follow-up assessments can determine how many participants who did not have the disease of interest at baseline have developed incident disease during the study Since data for a cohort study are collected from individuals at multiple points in time, researchers know with certainty which exposures were present before the onset of the new disease ○ This allows for the identification of potentially causal exposures Framework for a Cohort Study: Prospective Cohort Study: recruits participants that have or don’t have the exposure then follows both exposed and unexposed people forward in time to look for incident cases of disease ○ Useful to study uncommon exposures ○ Difference between the two groups = exposure status Retrospective cohort study (historic cohort study): participants are recruited based on data about their exposure status at some point in the past + measures outcomes Times of Baseline and Follow-Up Data Collection for Cohort Studies: Retrospective cohort study the first step in study design is identifying existing records (ex. School, medical files) that can provide baseline data ○ Then, contact the sampled individual (if possible from historic records) Longitudinal cohort study: follows a group of individuals who are representative of a selected population forward in time ○ Does not recruit members based on exposure status but instead based on their membership in a well-defined source population ○ People are assessed at baseline for exposures and diseases and then followed to determine the incident rates for the outcomes of interest Longitudinal Cohort Studies May Use: Fixed population: all participants start the study at the same time and no additional participants are added after the study’s start date Dynamic population (open population): allows new participants to be recruited after the start of data collection → allows for lost participants to be replaced by new recruits  Time series study (panel study): measures participants at multiple points in time Ratios, Rates & Risks: Ratio: comparison of two numbers (displayed as A:B or calculated as A/B) Reference population: group that is used as a comparison for another population Proportion: ratio in which the numerator is a subset of the denominator Percentage: proportion presented in units of “per 100” Rate: ratio in which the numerator and denominator have different units Count: number that enumerates the quantity of similar items Risk: probability of an individual in a population becoming a case during a defined period of time Incidence Rates: Incidence: number of new cases of disease in a population during a specified period of time Incidence rate (absolute risk): number of new cases of disease in a pop during a specific period of time divided by the total number of people within the pop who were at risk during that period of time Population at risk: those who do not already have the disease being tracked in a cohort study Incidence Rate Ratios: Cumulative incidence (incidence proportion/risk): percentage of people at risk in a population who develop new disease during a specified period of time ○ Cumulative incidence is a measurement of risk during a stated time frame Attack rate: cumulative incidence of infection during the course of an epidemic Rate ratio (RR), also known as relative rate: describes any ratio of two rates in which the reference (comparison) group is placed in the denominator Incidence rate ratio: compares the incidence rate among the exposed to the incidence rate in the unexposed 𝑎/(𝑎+𝑏) RR= 𝑐/(𝑐+𝑑) RR = 1 (or close to 1): The incidence rate was the same (or about the same) in the exposed and in the unexposed. RR > 1: The incidence rate was greater in the exposed than in the unexposed, suggesting that the exposure was risky. RR < 1: The incidence rate was lesser in the exposed than in the unexposed, suggesting that the exposure was protective Attributable Risk: Rate difference (excess risk, attributable risk or risk difference): absolute difference in the incidence rate between the exposed and unexposed group  Attributable risk percentage (AR%): is the proportion of incident cases among the exposed population in a cohort study that is due to the exposure (AR% is also called the etiologic fraction in the exposed) Population attributable risk (PAR): rate of new disease in a population that can be attributed to some people in the population having an exposure. Population attributable risk percentage (PAR%): is the proportion of incident cases in the total population that can be attributed to some people having the exposure Person-time: uses units like person-years, person-months, or even person-days to qualify how long participants in a study were observed Censoring: occurs when participants in a prospective or longitudinal study die, drop out, are lost to follow up, or for another reason are removed from further analysis Chapter 12 - Experimental Studies: Overview: Experimental studies → gold-standard for assessing causality ○ Researcher knows when the exposure occurred, allowing health status of each participant to be compared before and after to see whether that exposure caused the outcome Controlled trial: experiment where some participants are assigned to an intervention group and some to a comparison (control) group Randomized controlled trial (RCT): common experimental study design used in health sciences → some participants are randomly assigned to an intervention group while others are assigned to a control group ○ Participants are followed prospectively to see who has a favourable outcome and who does not ○ Some RCTs use a factorial design that tests multiple interventions in diff combinations within one trial RCTs and all experimental study designs require descriptions of: ○ Intervention, type of control, how participants will be assigned to exposure groups, the end point that will constitute a favourable outcome for the trail Step 1: Describing the Intervention: Intervention: strategic action intended to improve individual and/or population health status Prevention science: scientific study of which preventative health interventions are effective in various populations, how successful the interventions are, and how well they can be scaled up for the widespread implementation Description of the intervention should include: ○ What the intervention will be, eligibility criteria for participants, where and how participants will receive the intervention, when, how often and for what duration participants will receive the intervention Defining Outcomes:  Most experimental studies are superiority trials which aim to show that a new intervention is better than some type of comparison and not merely as good ○ Ex. an experimental study can examine whether a cheaper intervention is as good as, or no worse than, a more expensive intervention Equivalence trial aims to show that a new intervention is as good as some type of comparison Non inferiority trial: aims to show that a new intervention is no worse than some type of comparison Selecting Controls: Experimental studies assign a control and experimental group Common control = placebo → inactive comparison that is similar to the therapy being tested ○ Often enough to make recipients feel better There could also be an experimental study where a control group of participants are assigned to maintain their usual routines but this raises concerns of the Hawthorne effect wherein participants may act in a certain way because they know they are being observed Before and after study: nonrandomized experimental study that measures the same individuals before and after an intervention so that each participant’s “before” status can serve as the individual’s control Cross over design: randomly assigning some participants to receive the active intervention first and then the control while others receive the control and then the intervention Arms: describes treatment and nontreatment groups in an experiment Washout period: time between arms of the study to reduce carryover effect where effects from the first part of the study may bias the results of the second part of a crossover study Blinding: Also called masking is an experimental design element that keeps participants (sometimes even members of the research team) from knowing whether a participant is in the active intervention group or a control group. Single-blind: participants do not know whether they are in an active group or in a control group Double-blind: participants nor the researchers know which participant is in an active or control group Information bias: bias in epidemiological study that arises due to systematic measurement error Reporting bias: members of one study group systematically underreport or overreport an exposure or outcome Detection bias (surveillance bias): when a population group is routinely screened for adverse health conditions incorrectly appears to have a higher rate of disease because more frequent testing enables a higher case detection rate in that population rather than the general population Observer bias: when a researcher intentionally or unintentionally evaluates participants differently based on their group membership Randomization: randomized assignment of participants to an exposure group in a study Allocation bias: result of nonrandom assignment of participants to experimental study groups, such as when people with different exposure histories are not equally distributed across treatment arms Simple randomization: assigning individuals with some kind of random generator (coin toss)  Stratified randomization: dividing pop into subgroups prior to randomly but systematically assigning each individual Blocked randomization: randomly assigning groups of people to an intervention or control group Quasi-experimental design: assigning participants to an intervention or control group using a non-random method Natural experiment: research study in which the independent variable is not manipulated by the researcher but changes due to external forces Policy: set of principles and procedures defined by governments to help in decision making and resource allocation Principle of equipoise: experimental research should be conducted only when there is genuine uncertainty about which treatment will work better Adverse reaction: negative side effect of an outcome (medication) related to a study Adverse event: negative outcome that may be the direct result of a study-related exposure or due to a coincidental occurrence that is not directly related to the study Efficacy: measure of the success of an intervention (as per the proportion of individuals in the control group) Number needed to treat (NNT): expected number of people who would have to receive a treatment to prevent an unfavorable outcome Number needed to harm (NNH): number of people who would need to receive a particular treatment in order expect one of those people would have an adverse outcome Effectiveness: refers to results obtained under real-world, less-than-ideal conditions Efficiency: cost-effectiveness of an intervention based on effectiveness/resource considerations Treatment received analysis: experimental data that only includes participants who are compliant with their assigned intervention Treatment-assigned analysis (or intention-to-treat analysis): includes all participants even those who were not compliant Screening: type of secondary prevention where all members are encouraged to be tested for a disease based on risk factors Reference standard: a test used for comparison when examining the validity of a new screening/diagnostic test ○ A biopsy is considered to be a “gold-standard” as the test shows the actual presence of disease Cutpoint or threshold: a value that divides a numeric variable into separate categories Diagnostic accuracy: percentage of individuals who are correctly classified by the testas true positives/negatives Sensitivity (or true positive rate): proportion of people who actually have a disease and who test positive for the new test Specificity or (true negative rate): proportion of people who do not have the disease and test negative for the new test Positive predictive value (PPV): proportion of people who test positive with the new test and actually have the disease Negative predictive value (NPV): proportion of people who test negative and do not have the disease  False positive rate: proportion of people who do not have a disease and who were incorrectly tested positive using the new test False negative rate: proportion of people who have a disease and who were incorrectly tested negative using the new test Receiver operating characteristics (ROC) curve, or ROC curve: used to graphically examine the accuracy of a screening by plotting false positive rate (x-axis) versus the true positive rate (y-axis) Area under the curve (AUC): an aggregate measure of how well a screening performs across various cutoff points Likelihood ratio tests: probability ratios used to evaluate the accuracy of screening/diagnostic tests Positive likelihood ratio (LR+) test: examines whether a new test is good at predicting the presence of disease Negative likelihood ratio (LP-) test: examines whether a new test is good at predicting the absence of disease Chapter 14 - Correlational Studies: Overview: Correlational study: uses population-level data to look for associated between 2 or more characteristics that have been measured in several groups ○ Sometimes referred to as an aggregate study since correlational studies look only at grouped level population data and not individual level data ○ Correlational study exploring environmental exposure = ecological study ○ These studies compare groups rather than individuals Ecological fallacy: incorrect assumption that individuals follow the trends observed in population-level data Correlation: statistical measure of the degree to which changes in the value of one variable predict changes in the value of another Pearson correlation coefficient: statistical measure of the degree to which changes in the value of one numeric variable predict changes in the value of another numeric variable Age Standardization: Age adjustment: improves validity of comparisons of two or more populations with different age distributions Age-specific rate: rate for a certain age group Age standardization: applies age-specific rates from one or more study populations to a standard population or vice versa (depends on how much is known about age distribution) Direct age adjustment: applies age specific rates in two or more study populations with different age structures to one standard population so that the rates in the study populations can be more fairly compared Indirect age adjustment: apples age specific rates in a standard population to a study population so that a determination can be made about whether the overall rate in the study population is greater or lesser than expected  ○ Used to compare study populations for which the age distributions are known but age specific rates of exposure and/or disease are not known Standardized mortality ratio: compares the # of deaths in the study pop to the # of deaths expected in the study pop based on the age specific mortality rates in the standard pop Crude statistic: raw or unadjusted statistic → accurately describes the true experience in the pop Adjusted statistic: corrected to account for the effects of one or more other variables Age-standard statistic: fictitious statistic for a study pop that is created by applying age-specific rates to or from a standard population Chapter 15 - Synthesis Research: Synthesis Research: Integrates existing knowledge from previous research projects ○ In health sciences → narrative reviews, systematic reviews and meta-analyses are common ○ These types of research summarize what is already known about a topic, show connections among previous studies and offer new interpretations of previous studies’ contributions to scientific knowledge Review article in health sciences requires: ○ Extensive search of the literature ○ Extraction of key info from relevant articles ○ Clear and concise presentation of this information Limitations: not all journals publish review articles, exhibit less originality than other types of research Selecting a Topic: Select a topic that is narrow enough that all the relevant publications can be acquired ○ Should be greater than 8 articles but less than 352 articles for instance Narrative review = provides unique perspective about a topic by using evidence from the literature to support the author’s commentary ○ Intended to convey a perspective and not merely compile facts ○ Works best when researcher has a unique perspective on a topic and particular expertise in the field Systematic Reviews: Uses a predetermined and comprehensive searching and screening method to identify relevant articles ○ Designed to minimize the bias that might occur when researchers handpick the articles they want to highlight ○ Most important aspects in a systematic review are selecting appropriate keywords and inclusion criteria Important steps: ○ Identifying a narrow study question ○ Selecting a well-defined and valid study strategy  ○ Screening of all potential relevant articles to determine whether they meet the predefined eligibility criteria ○ Extraction of relevant info from all eligible articles ○ Summarizing findings of these articles Meta-Analysis: Involves the calculation of a pooled statistic that combines the results of similar studies identified during a systematic review Steps involve: ○ Defining study question and developing a study protocol ○ Using a comprehensive systematic search strategy to identify eligible articles ○ Using predefined inclusion and exclusion criteria to identify candidates for inclusion in meta-analysis ○ Extracting statistical results from each of the candidate studies ○ Combining comparable statistical results Note: inclusion criteria for meta-analysis is more strict than general systematic reviews Meta-Synthesis: Tertiary analysis that integrates results from different qualitative studies Goal: enhance understanding of a particular phenomenon