HCM 505 SEU Research Methodology in Health
HCM 505 SEU Research Methodology in Health Discussion
Discussion
Locate a research study that utilized experimental or quasi-experimental methods. Briefly summarize the study. For example, discuss the inclusion of 2-group tests, regression analysis, and time-series analysis in terms of the study design’s strengths, weaknesses, or limitations. What challenges or limitations did the researcher identify they encountered by choosing this method? Embed course material concepts, principles, and theories (which require supporting citations) in your initial response along with at least one scholarly, peer-reviewed journal article. Keep in mind that these scholarly references can be found in the Saudi Digital Library by conducting an advanced search specific to scholarly references. Use Saudi Electronic University academic writing standards and APA style guidelines. CHAPTER 12 Experimental Studies Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com. 12.1 Overview ▪ Experimental studies are the gold standard for assessing causality ▪ In a randomized controlled trial (RCT), some participants are randomly assigned to an active intervention group, the remaining participants are assigned to a control group, and all participants from both groups are followed forward in time to see who has a favorable outcome and who does not Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A controlled trial is an experiment in which some of the participants are assigned to an intervention group and some are assigned to a non-active comparison group Figure 12-1 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com Figure 12-2 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.1 Overview (Cont.) ▪ Experimental study designs require careful definitions of: – The type of control that will be used and why it is appropriate – How participants will be assigned to exposure groups – The end point that will constitute a favorable outcome for the trial ▪ Experimental studies also require careful consideration of the ethical challenges associated with assigning participants to an exposure Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com – The intervention 12.2 Describing the Intervention – Prevention science: the scientific study of which preventive health interventions are effective in various populations, how successful the interventions are, and how well they can be scaled up for widespread implementation Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Intervention: a strategic action intended to improve individual and/or population health status 12.2 Describing the Intervention (Cont.) ▪ The description should state exactly: – What the intervention will be – The eligibility criteria for participants – Where and how participants will receive the intervention – When, how often, and for what duration participants will receive the intervention Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ The first step in an experimental study is to carefully define the intervention that participants assigned to the active intervention group will receive and to decide on the person, place, and time (PPT) criteria for the study 12.3 Defining Outcomes ▪ An equivalence trial aims to demonstrate that a new intervention is as good as some type of comparison ▪ A noninferiority trial aims to demonstrate that a new intervention is no worse than some type of comparison Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A superiority trial aims to demonstrate that a new intervention is better than some type of comparison, not merely as good as the comparison Figure 12-3 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.3 Defining Outcomes (Cont.) ▪ These measures of success must be stipulated prior to the initiation of the study Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Because the term “better” can be defined in so many ways, the researcher must carefully define what constitutes a favorable outcome for an individual participant and for the experimental study as a whole Figure 12-4 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.4 Selecting Controls ▪ One commonly used type of control is a placebo, an inactive comparison that is similar to the therapy being tested ▪ When the goal of the experiment is to test whether a new therapy is better than a current one, it is appropriate to compare the new therapy to some standard of care, an existing therapy that is used as a comparison for a new therapy being experimentally tested Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Experimental studies usually assign some participants to the active intervention and the remainder to a control group Figure 12-5 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.4 Selecting Controls (Cont.) ▪ A factorial design tests several different interventions in various combinations within one trial Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Sometimes the goal is to determine how much of an intervention is required, so varying doses and durations may be tested and compared to one another Figure 12-6 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.4 Selecting Controls (Cont.) ▪ Hawthorne effect: a type of bias that occurs when participants in a study change their behavior for the better because they know they are being observed Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Although experimental studies sometimes include a control group of participants who are randomly assigned to maintain their usual routines, this method is usually not preferred 12.4 Selecting Controls (Cont.) ▪ Before-and-after study: a non-randomized experimental study that measures the same individuals before and after an intervention so that each participant’s “before” status can serve as that individual’s control Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ When there are ethical concerns about withholding a potentially lifesaving intervention from some participants, it may be possible for participants to serve as their own controls 12.4 Selecting Controls (Cont.) ▪ In a crossover study, both groups may take a break between the two arms of the experiment (a washout period when patients receive no treatment) to reduce carryover effects, the residual effects from the first part of an experimental study that may bias the results of the second part of a crossover study if a sufficient washout period between the two arms of the study is not implemented Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A crossover design that randomly assigns some participants to receive the active intervention first and then the control, and assigns the others to receive the control first and then the active intervention 12.5 Blinding – Single-blind: participants do not know whether they are in an active group or a control group – Double-blind: neither the participants nor the people assessing the participants’ health status know which participants are in an active or control group Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Blinding (or masking): an experimental design element that keeps participants (and sometimes some members of the research team) from knowing whether a participant is in the active intervention group or the control group 12.5 Information Bias ▪ Reporting bias: information bias that occurs when members of one study group systematically underreport an exposure or outcome ▪ Detection bias (surveillance bias) information bias that occurs when a population group that is routinely screened for adverse health conditions incorrectly appears to have a higher-than-typical rate of disease because more frequent testing enables a higher case detection rate in that population than in the general population Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Information bias: bias in an epidemiological study that arises due to systematic measurement error 12.5 Blinding ▪ Blinding prevents participants and assessors from being able to evaluate outcomes differently based on the results they expect for an exposure, but blinding is usually possible only when all participants are assigned to similar exposures ▪ To minimize the risk of bias in studies that are not blinded, it is helpful to identify objective outcome measures such as laboratory tests rather than relying on subjective outcome measures such as participants’ self-reported feelings Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Observer bias: information bias that occurs when an observer intentionally or unintentionally evaluates participants differently based on their group membership 12.6 Randomization ▪ Randomization also mitigates the allocation bias that might occur as a result of non-random assignment of participants to experimental study groups, such as when people with different exposure histories are not equally distributed across treatment arms Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Randomization: the assignment of participants to an exposure group in an experimental study using a chance-based method that minimizes several types of possible bias 12.6 Randomization (Cont.) ▪ Simple randomization: the use of a coin toss, a random number generator, or some other simple mechanism to randomly assign each individual in an experimental study to one of the exposure groups ▪ Stratified randomization: the division of a population into subgroups prior to randomly but systematically assigning each individual within each subgroup to one of the exposure groups in an experimental study ▪ Block randomization: an allocation method that randomly assigns groups of people to an intervention group and other groups of people to a control group Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A variety of approaches can be used to randomly allocate participants to an active intervention group or a control group Figure 12-7 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.6 Randomization (Cont.) ▪ A quasi-experimental design is an experimental study that assigns participants to an intervention or control group using a non-random method ▪ A natural experiment is a research study in which the independent variable is not manipulated by the researcher but instead changes due to external forces Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Some experimental studies use non-randomized approaches because randomization is unethical or is not feasible 12.7 Ethical Considerations ▪ Equipoise: experimental research should be conducted only when there is genuine uncertainty about which treatment will work better ▪ Researchers must put in place a system for monitoring adverse reactions and must identify the conditions under which an experiment would be discontinued early either because the exposure proves to be risky or because the new intervention appears to be so beneficial that keeping it from the control group would be unethical Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Experimental studies involve a particularly high level of ethical risk because the researcher assigns participants to exposures that the participants do not choose and may have been unlikely to encounter had they not volunteered to participate in a research project Figure 12-8 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.7 Ethical Considerations (Cont.) ▪ Adverse event: a negative outcome that may be the direct result of a studyrelated exposure or may be a coincidental occurrence that is not directly related to the study but happens after an individual receives a study-related exposure ▪ Adverse events that occur during a research study must be immediately reported to the appropriate institutional review board (IRB) Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Adverse reaction: a negative side effect of a medication, vaccination, or other exposure, or another bad outcome related to a study 12.8 Efficacy – A high efficacy is an indication that an intervention is successful Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Efficacy: a measure of the success of an intervention that is calculated as the proportion of individuals in the control group who experienced an unfavorable outcome but could have expected to have a favorable outcome if they had been assigned to the active group instead of the control group 12.8 Efficacy (Cont.) – A small NNT indicates a more effective intervention ▪ Number needed to harm (NNH): the number of people who would need to receive a particular treatment in order to expect that one of those people would have a particular adverse outcome – A large NNH indicates a safer intervention Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Number needed to treat (NNT): the expected number of people who would have to receive a treatment to prevent an unfavorable outcome in one of those people (or, alternately stated, to achieve a favorable outcome in one person) Figure 12-9 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.8 Efficacy (Cont.) ▪ Efficiency is an evaluation of the cost-effectiveness of an intervention that is based on both its effectiveness and resource considerations Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Effectiveness is calculated with the same equation as efficacy, but refers to results obtained under real-world, less-than-ideal conditions 12.8 Efficacy (Cont.) – Used for calculating efficacy ▪ A treatment-assigned analysis (or intention-to-treat analysis) includes all participants, even if they were not fully compliant with their assigned protocol – Used for calculating effectiveness Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A treatment-received analysis of experimental data includes only the participants who were fully compliant with their assigned intervention or comparison protocol Figure 12-10 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.9 Screening and Diagnostic Tests Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Screening: a type of secondary prevention in which all members of a welldefined group of people are encouraged to be tested for a disease based on evidence that members of the population are at risk for the disease and early intervention improves health outcomes 12.9 Screening and Diagnostic Tests (Cont.) – In most situations, this goal involves comparing a new test to an existing one – A reference standard is the test used for comparison when examining the validity of a new diagnostic test – A “gold standard” reference test shows the actual presence of disease in affected people Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ The goal of most studies of new or improved screening or diagnostic tests is to compare two assessments that are supposed to measure the same thing 12.9 Screening and Diagnostic Tests (Cont.) ▪ The diagnostic accuracy is the percentage of individuals who are correctly classified by a test as true positives or true negatives (that is, the percentage for which both the reference test and the new test yield the same result) – An ideal test will have 100% diagnostic accuracy Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A cutpoint or threshold is a value that divides a numeric variable into separate categories 12.9 Screening and Diagnostic Tests (Cont.) ▪ The specificity, or true negative rate, is the proportion of people who do not have the disease who test negative with the new test ▪ An ideal screening or diagnostic test would have 100% values for sensitivity and specificity Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ The sensitivity, or true positive rate, is the proportion of people who actually have a disease (according to the reference standard) who test positive using the new test 12.9 Screening and Diagnostic Tests (Cont.) ▪ The negative predictive value (NPV) is the proportion of people who test negative who actually do not have the disease ▪ An ideal screening or diagnostic test would have 100% values for PPV and NPV Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ The positive predictive value (PPV) is the proportion of people who test positive with the new test who actually have the disease (according to the reference standard) Figure 12-11 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.9 Screening and Diagnostic Tests (Cont.) – The false positive rate can also be calculated as 1 – specificity (that is, the number one minus the specificity, which must fall between 0 and 1) ▪ The false negative rate is the proportion of people who actually have a disease who incorrectly test negative using the new test ▪ An ideal screening or diagnostic test would have a 0% rate of false positive and false negative outcomes Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ The false positive rate is the proportion of people who actually do not have a disease (according to the reference standard) who incorrectly test positive using the new test 12.9 Screening and Diagnostic Tests (Cont.) – Increasing the sensitivity decreases the specificity – Increasing the specificity decreases the sensitivity Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ For tests with a flexible cutoff point for defining positive and negative test results, there is always a trade-off between sensitivity and specificity Figure 12-12 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 12.9 Screening and Diagnostic Tests (Cont.) – This is the same as saying that the ROC curve plots 1 – specificity on the x-axis and sensitivity on the y-axis for a variety of cutoff points Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A receiver operating characteristics (ROC) curve can be used to graphically examine the accuracy of a screening or diagnostic test by plotting the false positive rate (on the x-axis) versus the true positive rate (on the y-axis) for the different possible cutoff points of the test 12.9 Screening and Diagnostic Tests (Cont.) – AUC values can range from 0 to 1, with 0 indicating a test that is incorrect 100% of the time and 1 indicating a test that is correct 100% of the time Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ The area under the curve (AUC) is an aggregate measure of how well a screening or diagnostic test performs across various cutoff points 12.9 Screening and Diagnostic Tests (Cont.) Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Likelihood ratio tests are probability ratios used to evaluate the accuracy of diagnostic tests 12.9 Screening and Diagnostic Tests (Cont.) Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com CHAPTER 14 Correlational Studies Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com. 14.1 Overview ▪ The variables included in correlational analyses are usually aggregate (grouped) statistics such as the proportion of a population with a particular characteristic or the average value of the variable in a population ▪ A correlational study is sometimes called an aggregate study because correlational studies look only at grouped population-level data and they do not include any individual-level data Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A correlational study uses population-level data to look for associations between two or more characteristics that have been measured in several groups Figure 14-1 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 14.1 Overview (Cont.) ▪ Correlational studies then examine exposure–outcome pairs – Statistical methods can be used to control for interactions among related variables ▪ A correlational study that explores an environmental exposure may be called an ecological study Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ For most correlational studies, at least one characteristic of the populations being examined is designated as an exposure and at least one is designated as an outcome or disease 14.2 Aggregate Data – For any one variable in a correlational study, the best option is for all data to come from the same source – When using multiple sources of data for one variable, the researcher should establish a scientifically justifiable set of inclusion criteria for the study and then exclude any data sources that do not meet all of those eligibility requirements Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Information about all the variables of interest must be available for a suitable number of populations, which can be grouped by place or time 14.2 Aggregate Data (Cont.) ▪ Each population should be assigned to its own row in the spreadsheet ▪ Each exposure and each outcome should be assigned to its own column ▪ The data should be filled into the cells in each column so that they line up with the correct population Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com Before conducting a statistical analysis of aggregate data, the data from each population must be entered into a spreadsheet Figure 14-2 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 14.3 Avoiding the Ecological Fallacy ▪ Correlational studies compare groups rather than individuals ▪ Ecological fallacy: the incorrect assumption that individuals follow the trends observed in population-level data – The experience of individuals in a population may vary significantly from the population average Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com – No individual-level data are included in the analysis, only population-level data 14.4 Correlation ▪ On a scatterplot used to illustrate correlation, each point represents one population in the study – The exposure is plotted on the x-axis – The outcome or disease is plotted on the y-axis – A trend line is fit to the data points, usually using a software program that calculates the line with the best fit Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Correlation: a statistical measure of the degree to which changes in the value of one variable predict changes in the value of another 14.4 Correlation (Cont.) ▪ When the points are not exactly linear but a line for trend can be drawn through them, the correlation is mild or moderate ▪ When the points appear to be randomly placed and no obvious line can be drawn through them, or when the best-fit line is horizontal then the correlation is weak or nonexistent Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ When all the points fall neatly along or very near to a sloped line (with a positive upward or negative downward slope), the correlation is strong Figure 14-3 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 14.4 Correlation (Cont.) – The Pearson correlation coefficient (r) is used for continuous variables and other variables with responses that can be plotted on a number line – The Spearman rank-order correlation (ρ = rho) is used with variables that assign a rank to responses (like 1st place, 2nd place, 3rd place, and so on) or that have ordered categories (such as scales that range from 1 = strongly disagree to 5 = strongly agree) Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Different equations are used to calculate correlations between different types of variables 14.4 Correlation (Cont.) – When r = 0, there is no association between the exposure and outcome ▪ The coefficient of determination, r2, shows how strong a correlation is without indicating the direction of the association – The value of r2 ranges from 0 for no correlation to 1 for perfect correlation Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ The value of r (or ρ) ranges from –1, when all points lie perfectly on a line with a negative slope, to 1, when all points lie perfectly on a line with a positive slope 14.4 Correlation (Cont.) ▪ A correlation can be present even when the relationship is not causal – The term correlation that should never be used as a synonym for causation Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ When more than two variables are being compared or the goal is to understand the relationship between two variables while controlling or adjusting for the effects of other variables, linear regression models can be used to assess the associations 14.5 Age Standardization – A younger population may have more favorable health statistics because fewer population members have developed the chronic diseases associated with aging – An older population may have less favorable health statistics simply because its members are more advanced in years ▪ Age adjustment methods can improve the validity of comparisons of two or more populations with different age distributions Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Sometimes populations that are being compared have very different age structures, with one or more populations considerably younger or older than the others 14.5 Age Standardization (Cont.) ▪ An age-specific rate is a rate for a particular age group – The age standardization method used depends on how much is known about the age distributions in the study populations Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Age standardization applies age-specific rates from one or more study populations to a “standard population,” or vice versa, to generate comparable rates for populations with different age structures 14.5 Direct Age Standardization – For example, the rates from several cities can be standardized to the national population or the rates in several countries can be standardized to the global population ▪ Direct age standardization can only be conducted when age-specific data are available for all of the study populations being compared Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Direct age adjustment (also called direct age standardization) applies agespecific 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 14.5 Direct Age Standardization (Cont.) ▪ The standardized rates are directly comparable because they are based on the age distribution of the same standard population, and they allow for a fairer comparison of disease statue in the two cities Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ To calculate an age-adjusted overall rate for each study population, each population’s age-specific rates are applied, one at a time, to a standard population and then the expected numbers of outcomes for each age group are added together and the sum is used to calculate an all-ages rate in the standard population Figure 14-4 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 14.5 Indirect Age Standardization ▪ Indirect age standardization methods can be used to compare study populations for which the age distributions are known but age-specific rates of exposure and/or disease are not known Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ Indirect age adjustment (also called indirect age standardization) applies agespecific 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 given the population’s age distribution 14.5 Indirect Age Standardization (Cont.) Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ A standardized mortality ratio (SMR) compares the number of deaths observed in the study population to the number of deaths expected in the study population based on the age-specific mortality rates in the standard population Figure 14-5 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com 14.5 Age Standardization ▪ A crude statistic is a raw or unadjusted statistic ▪ An adjusted statistic is a statistic that has been corrected to account for the effects of one or more other variables – Age standardization improves the validity of comparisons of populations with different age structures, but it does not improve the accuracy of the statistics themselves Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com – When only one population is being described, the crude statistic is usually the correct measure to report, because it accurately describes the true experience in the population 14.5 Age Standardization (Cont.) ▪ When two populations are being compared, the standardized rates are the best numbers to use for the comparison Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com ▪ When the value of a statistic for one population is being reported, the crude rate is usually the most appropriate number to report Figure 14-6 Copyright © 2021 by Jones & Bartlett Learning, LLC an Ascend Learning Company. www.jblearning.com
