Achieving a desired outcome while optimizing a resource
Back
What can regression do in terms of data?
Front
Try to predict values of one variable given another variable
Back
What is the formula for logistic regression?
Front
P(y/x) = (e^a+bx)/(1-e^a+bx)
Back
What is CEA?
Front
Cost effectiveness analysis, a way of quantifying a benefit in dollars if CBA not feasible or not needed, assess "bang for buck", find the most efficient therapy
Back
What are the two types of outcomes that are analyzed in a CEA?
Front
Clinical or humanistic
Back
What is sensitivity analysis?
Front
A way to see if a CEA is responsive to certain changes, such as treatment efficacy rates, discount rates, or cost estimates
Back
What is the equation for CUA?
Front
Cost/ (QALY x life years gained)
Back
What is logistic regression most commonly used for?
Front
Predicting the probability of a dichotomous outcome from occurring
Back
When is it appropriate to do a CUA?
Front
When quality of life is important, when both morbidity and mortality is affected, when a wide variety of variables are in play, and when results of other CUAs need to be compared
Back
When should CMA be used?
Front
If outcomes are the same (considering side effects, efficacy), CMA is easily interpretable, is done beyond just the price
Back
How is regression calculated?
Front
By taking the sums least squares and finding the best fit line for a certain set of data
Back
What is beta in multiple regression?
Front
Slope
Back
What are the three types of outcomes?
Front
Economic, clinical, humanistic
Back
What is CMA?
Front
Cost minimization analysis, trying to find the cheapest therapy that will provide the same benefit
Back
What is a residual?
Front
The difference between the point on the regression and the actual point
Back
When is doing a CEA relevant?
Front
When the new therapy is more effective and more expensive or less effective but less expensive
Back
What assumption must be made when doing a CMA?
Front
Outcomes are the same or very similar
Back
What is the formula for CER?
Front
(Cost of new tx - cost of standard tx)/(effectiveness of new tx - effectiveness of standard tx)
Back
What is an intangible cost?
Front
Pain and suffering
Back
What is CBA?
Front
Cost benefit analysis, a straightforward method to evaluate the costs and benefits of interventions
Back
What is the equation for sums of squares?
Front
Sigma(y - y[hat])^2
Back
What is a QALY?
Front
Quality-adjusted life year, a way of analyzing both quantity and quality of life in regards to health outcomes
Back
What is the range of an HRQoL?
Front
0 (death) to 1 (perfect health)
Back
What is the generally accepted annual discount rate?
Front
3%/year (e.g. $1 today = $0.97 next year)
Back
What is logistic regression?
Front
A way of analyzing dichotomous variables to another variable
Back
What type of analysis will analyze "bang for buck" of various interventions?
Front
Cost effectiveness analysis
Back
What is regression?
Front
Drawing the line of best fit through a scatterplot
Back
What is a discount?
Front
Time preference of value, the belief that future dollars are worth less than current dollars
Back
What is the importance of a timeframe in pharmacoecon?
Front
Costs and outcomes may be different in the future, analysis must be done to take this into consideration
Back
Why is multiple regression often a better model than linear regression?
Front
Most dependent variables are affected by more than one independent variables
Back
What is a HRQoL?
Front
Utility of health-related quality of life, allows patients to define their perceptions of health
Back
What is utility?
Front
The value or worth placed on health status
Back
What is measured in a CUA?
Front
Quality-adjusted life years (QALYs)
Back
What is ICER?
Front
Incremental cost-effectiveness ratio, a way to compare costs and benefits of competing therapies in a single summary measure
Back
What is the goal in a CBA?
Front
Cost ratio >1
Back
What is the type of result in a CEA?
Front
Uni-dimensional outcome (e.g. life years saved, cases cured, % reduction of lipids
Back
What is the effect of adding a new X variable to R in multiple regression?
Front
Never decreases - usually fits better
Back
What is CER?
Front
Cost-effectiveness ratio, a way of measuring the incremental benefit of a therapy
Back
What is a dichotomous outcome?
Front
An outcome with two options (yes/no)
Back
What is a decision tree?
Front
A tool that can be used to make a decision in pharmacoeconomics or other fields, graphically displays options
Back
What is measured in a CBA?
Front
Benefits and costs in terms of dollars, even if it's not a direct cost (eg days at work, costs avoided)
Back
What is multiple regression?
Front
The analysis of more than one variable for a relationship
Back
What is CUA?
Front
Cost utilization analysis, a type of CEA that measures the quality of life
Back
Section 3
(50 cards)
What is beta in logistic regression?
Front
An increase in the log of odds for a one unit increase in X
Back
What does PICO stand for?
Front
Population, intervention, comparator, outcome
Back
What is the Cochrane collaboration?
Front
An international nonprofit that lists and prepares systematic reviews
Back
What is the Kaplan Meier Method?
Front
Compares the survival of 2 groups in survival analysis, uses a stair-step curve
Back
What is the rationale for doing a meta-analysis?
Front
Increased sample size, power, and precision of treatment effect
Back
What is survival analysis?
Front
The statistical methods for analyzing longitudinal data of survival, such as death over time, onset of illness over time, etc
Back
What are the results of a meta-analysis displayed as?
Calculate measure of treatment effect (odds ratio or risk reduction), calculate treatment effect
Back
How is survival analysis plotted?
Front
S(t), cumulative incidence which ranges from 0 to 100%
Back
What is the relationship between survival and hazard functions?
Front
Mathematically related - survival goes down, hazard goes up (usually)
Back
What are the two common ways of analyzing survival analysis?
Front
Kaplan Meier and Cox proportional hazard
Back
What scale is used to assess systematic review quality?
Front
Jadad scale (0-5)
Back
What is recidivism?
Front
The act of a person repeating an undesirable behavior (common in substance abuse and crime)
Back
What are the eligibility criteria in systematic reviews?
Front
Studies, not patients
Back
What is a meta-analysis?
Front
Data from individual studies that are pooled quantitatively and reanalyzed
Back
What are cox models typically used for?
Front
Multivariable surival analysis
Back
What is the relationship between systematic review and meta-analysis?
Front
Systematic review is the step by step process of data analysis, meta-analysis is the summarized data
Back
What is the most popular method of analyzing Kaplan-Meier data?
Front
Log-rank test
Back
What is the best way of abstracting the systematic review search?
Front
Using 2 or more independent blinded reviewers to assess trials, resolve discrepancies between reviewers
Back
What type of analysis is best to analyze a Kaplan-Meier if survival time in the earlier part of the study is important?
Front
Peto test (best if mortality is high)
Back
How should the validity of a meta analysis be assessed?
Front
Do not assume validity, determine validity of methodology, magnitude of treatment effect, applicability of results to population
Back
What is censoring?
Front
Subjects that are lost to follow up or that drop out of a study, or if the study ends before they die or have an outcome of interest
Back
What should the search strategy be like in systematic review?
Front
Identify key terms for PICO elements, think of misspellings or synonyms, search for everything individually and then link those searches
Back
What is adjusted odds ratio?
Front
An output of a logistic regression that adjusts for other factors
Back
What is a hazard function?
Front
The continuous increase in incidence compared to an independent variable
Back
What is the I^2 stat?
Front
% variation caused by heterogeneity (0-1, >0.5 is very large)
Back
How should the search strategy of a systematic review be analyzed?
Front
Search for published and unpublished trials, avoids biases,
Back
What are the advantages of a Logit?
Front
Makes logistic regression linear, makes the probability scale -infinity to infinity
Back
What are limitations of Kaplan-Meier curves?
Front
Survival analysis becomes unreliable with fewer patients, more than 2 groups becomes difficult, need another approach for >3 variables
Back
What is beta in multiple logistic regression?
Front
Increase in log odds for a one unit increase in xi with all other xis constant
Back
If confounders exist in a logistic regression equation, what can be done to reduce their effect?
Front
Multiple logistic regression
Back
What are the strengths of a logistic model?
Front
Ability to calculate odds ratios that are adjusted for other variables
Back
What is Cochran's Q stat?
Front
Looking at the significance of review heterogeneity
Back
What is PRISMA?
Front
Preferred reporting items for systematic reviews and meta-analyses, a 27-item checklist to assess content of SR and MA
Back
What are the goals of survival analysis?
Front
Estimate time-to-event for a group, compare time-to-event between two groups, assess the relationship of co-variables to time-to-event
Back
What is the equation for calculating odds ratio from beta?
Front
LR = e^Beta
Back
What is time-to-event?
Front
The time from entry into a study until a subject has a particular outcome
Back
What does the slope of a survival curve show?
Front
How fast survival changes and how high mortality will be
Back
How are systematic review search results analyzed?
Front
Organize them by sample size and outcomes per interest group, assess for randomization, bias, blinding, type of analysis, and excluded poor trials according to criteria
Back
What is Cox Proportional Hazards?
Front
Compares the relationship between the hazard rate and an IV
Back
What is the result of a cox analysis?
Front
Hazard ratio
Back
What is a systematic review?
Front
Form of research that provides a summary of trials on a specific clinical question, using an explicit protocol
Back
What is beta related to in logistic regression?
Front
Increase in probability of outcomes associated with a risk factor
Back
What is the kappa stat in systematic review?
Front
The level of agreement between reviewers, ranges from 0 to 1
Back
What is the LOGIT?
Front
Another term for logistic transformation, converts logistic regression to a line
Back
Who decides whether meta-analysis are generalizable?
Front
The practitioner
Back
Can logistic regression be used for ordinal variables?
Front
No, difference between variables is not the same
Back
What is logistic transformation?
Front
Dependent variable in a logistic model is the log of the odds
Back
How should the literature be reviewed in systematic reviews?
Front
Search everywhere, not just PubMed or EMBASE, contact authors and manufacturers to try to find failed studies too