BC Jung A Brief Introduction to Epidemiology - XIII (Critiquing the Research: Statistical Considerations) Betty C. Jung, RN, MPH, CHES
BC Jung Learning/Performance Objectives u Quick review –Basics of inferential statistics –Common measures of association u To be able to statistically critique studies –Statistical Caveats –Statistical Issues –Statistical Rules of Thumb
BC Jung Introduction u Refresh your memory –Basics of inferential statistics –Common measures of associations used in epidemiologic studies
BC Jung Measures of Association & Hypothesis Testing Test Statistic = Observed Association - Expected Association Standard Error of the Association u Type I Error: Concluding there is an association when one does not exist u Type II Error: Concluding there is no association when one does exist
BC Jung Measures of Association u Two Main Types of Measures –Difference Measures (Two Independent Means, Two Independent Proportions, The Attributable Risk) –Ratio Measures (Relative Risk, Relative Prevalence, Odds Ratio)
BC Jung Measures of Association: Difference Measures u Two Independent Means u Two Independent Proportions u The Attributable Risk
BC Jung Attributable Risk (AR) u The difference between 2 proportions u Quantifies the number of occurrences of a health outcome that is due to, or can be attributed to, the exposure or risk factor u Used to assess the impact of eliminating a risk factor
BC Jung Measures of Association: Ratio Measures u Relative Risk (RR) u Relative Prevalence (RP) u Odds Ratio (OR)
BC Jung Strength of Association Relative Risk;(Prevalence); Odds Ratio Strength of Association None Weak Moderate Strong 10.0Approaching Infinity Source: Handler,A, Rosenberg,D., Monahan, C., Kennelly, J. (1998) Analytic Methods in Maternal and Child Health. p. 69.
BC Jung Caveats about Classifying Data u All persons in an epidemiologic study must be classifiable u All study reports should clearly state criteria used for classifying variables u Studies that use different criteria for defining the presence of any health state are not comparable with respect to reported rates of that health state
BC Jung Caveats about Quantitative & Categorical Variables u Information on variability between persons is lost when quantitative data are categorized u Collapsing a quantitative variable into a categorical variable with two or more categories may obscure the fact that the underlying variable has a much larger range in one category than in another category
BC Jung Caveats about Quantitative & Categorical Variables (Continued) u Be careful about comparing ranges because a larger sample will generally have a larger range u Collapsing quantitative variables into categories limits the choices of appropriate statistical tests of significance u Try using commonly used categories (as five- or ten-year age bands) to facilitate comparisons across studies
BC Jung Berkson’s Fallacy u Associations based on hospital or clinic data are influenced by differential admission rates among groups of people u Similar source of selection bias occur when associations are based on autopsy data
BC Jung Caveats about P-Values u The size of the p-value has no relationship to the potential practical significance of the findings u The P-value reveals nothing about the magnitude of effect (i.e., how much one group differs from another), or the precision of measurement (i.e., the amount of random error) u The nature of the sample, not the p-value, will determine whether inferences to the population of interest can be made (and the sample must be representative of the population)
BC Jung Confidence Interval Estimation u Uses the sample mean to construct an interval (range) of numbers to estimate the effect u Provides some indication of how probable it is (e.g., 68%, 90%, 95%), or how “confident” one can be, that the true mean lies within the range of numbers in the interval estimate
BC Jung Greenhalgh’s Questions to Ask About the Analysis (A) u Have the authors set the scene correctly? u Have they determined whether their groups are comparable, and, if necessary, adjusted for baseline differences? u What sort of data have they got, and have they used appropriate statistical tests?
BC Jung Greenhalgh’s Questions to Ask About the Analysis (B) u If the authors have used obscure statistical tests, why have they done so and have they referenced them? u Are the data analyzed according to the original protocol? u Were paired tests performed on paired data?
BC Jung Greenhalgh’s Questions to Ask About the Analysis (C) u Was a two-tailed test performed whenever the effect of an intervention could conceivably be a negative one? u Were “outliers” analyzed with both common sense and appropriate statistical adjustments? u Have assumptions been made about the nature and direction of causality?
BC Jung Greenhalgh’s Questions to Ask About the Analysis (D) u Have “P values” been calculated and interpreted appropriately? u Have confidence intervals been calculated, and do the authors’ conclusions reflect them? u Have the authors expressed the effects of an intervention in terms of the likely benefit or harm which an individual patient can expect?
BC Jung Statistical Issues: Epidemiological Studies u Logistic regression for binary outcomes u Cox regression for survival analysis u Poisson distribution for disease incidence or prevalence u Odds ratio approximates relative risk when disease is rare
BC Jung Statistical Issues: Environmental Studies u Good statistical models are hard to come by u Publication bias can exaggerate excess risk u Odds ratios less than two (or greater than 0.5) can be interesting
BC Jung Statistical Issues: Environmental Studies u What is the statistical basis for the environmental standard? u Variability vs. uncertainty u What’s the quality of the metadata u Biomarkers as surrogates for clinical outcomes
BC Jung Statistical Issues: Risk Assessment u Hazard identification u Dose-response evaluation u Exposure assessment u Risk characterization u Risk management
BC Jung Statistical Rules of Thumb u Use a logarithmic formulation to calculate sample size for cohort studies u Use no more than 4 or 5 controls per case for case-control studies u Obtain at least 10 subjects for every variable investigated for logistic regression
BC Jung Statistical Rules of Thumb u Increase sample size in proportion to dropout rate. If dropout rate is expected to be 20%, then increase n/0.80 u If dropout is greater than 20%, review reasons for dropouts u Accept substitutes with caution
BC Jung Statistical Rules of Thumb u Choosing cutoff points u Do not dichotomize unless absolutely necessary u Select an additive or multiplicative model according to: theoretical justification, practical application, and computer implication
BC Jung References u For Internet Resources on the topics covered in this lecture, check out my Web site: u Other lectures in this series: