EBM Intro/Stats Flashcards
Systematic vs Random error
Bias/Bounding, errors in study design, unavoidable vs
Errors from measurements, biostat provides tools do deal with this
Epidemiology
Study of distribution and determinants of health-related states or events and application of this study to control health related problems
Clinical epidemiolgoy
Making predictions about individual patients by counting clinical events of groups of similar patients
EBM
Using evidence from clinical epidemiology to treat patients
Mantra
Chance, bias, confounding, if rule those out-truth
why not look at individuals and look at what
They vary too much, instead look at probabilites of outcomes in similar patient groups
excess mortality
people who died due to a procedure vs the control group
System I vs System II
Jump to conclusions based on limited evidence, effortless
Takes effort, quantitiative analysis, when we have time to think
Health
State of complete mental, physical, and social well being
Health Outcomes
death, disease, discomfort, disability, dissatisfaction
Epidemiological Transition
Acute, contagious diseases to chronic and noncontagious lifestyle diseases
EBM decision
Current best evidence supports this way and is okay with patient and is congruent with clinical experience
dyspnea
shortness of breath
Prevalence vs Incidence vs incidence rate
Current cases expressed as proportion of total cases to total population-tells about past
New cases of disease over a finite time-risk of getting the disease
Risk of getting disease/time
Rate vs Proportion vs Ratio
Ratio-numerator no included in denominator
Rate-Numerator and time included in denominator-incidence rate
Proportion-numerator included in denominator-prevelance rate
Case definitions
diagnostic criteria for defining cases
Disease prevalence (point/period prevalence)
What proportion of population has disease at specific point in time
Number of people with disease at given point in time/
Total number of people in population
Point is instantaneous and prevalance is over time
Disease Incidence (cumulative incidence)
How quickly people are being diagnosed with disease *new cases
Cum incidence-proportion of a group that develops disease over a certain time=
number of new cases/
Number of people at risk of developing the disease over a defined time
Incidence rate
Rate at which new disease has occurred in population at risk per some unit time
Number of new cases/
Total time experienced by the at risk population
Incidence density
new cases/
incidence rate in dynamic changing population which people are under study and at risk for varying periods of time
Total time experienced b the at risk population
(same as incidence rate?-just population is dynamic)
Add up periods of observation
Number of cases/person years
Point prevalence
start of study-proportion who had sickness
Rate
TIME IN THE FUCKING DENOMINATOR
Steady state
no variables are changing over time
Prevalence
Incidence x Avg Duration
Incidence
Prevalence/Duration
Duration
Prevalence/Incidence
Apparent incidence
Completely dependent on intensity of effort to identify cases
Epidemic
increase in incidence in community or region
Outbreak
small epidemic in limited locale
Pandemic
Epidemic that crosses many international boundaries
Endemic disease
constant presence of disease in particular area
Case fatality rate
proportion of people who suffer from disease
complication rate
proportion of people who suffer complication after trx
infant mortality rate
number of death
perinatal mortality rate
Number stillbirths and deaths in first week of life per 1000 births
maternal mortality rate
Number maternal deaths related to childbirth/number of live births
physician use of prevalence and incidence
Use prevalence to estmiate likelihood of diagnoses before test done
-incidence is not better because it is comparison of groups vs 1 patient
ways to express and compare risk
express- cumulive incidence and incidence rate
Compare-Measures of effect (abs risk difference and relative risk)
Absolute risk/absolute risk difference
Abs risk=Incidence
Absolute risk difference=Incidence of exposed- Incidence of unexposed
ex. some has risk factor has increased risk of 3/1000
- easiest for patients to understand
Relative Risk
I exposed/I unexposed
AKA risk ratio aka crude mortality ratio
How many times more likely are patients to get disease when exposed vs unexposed
Random/Probability sample
Random-each person has equal probabilty of being sampled
Probability-Each person has a known probability of being selected (not always equal)
-allows weighing towards groups of interest/low freq
Probabilty vs Odds
Probabilty:A/A+B
Odds: A:B
Addition/Mult rule
Addition=two independent events occurs (and)
Mult=Probabilty of two mutually exclusive events occurs (or)
Point Prevalence
Proportion of people in population who have disease at particular point in time
%
Period prevalane
Proportion of people who have diease over a specifd period of time
%
Cumulative incidence
proportion of population at risk that develops disease over a period of time
%
Incidence density
Rate of newcases per total amount of time people in a population where at risk
- dynamic population (workers)
- person time
Incidence rate
rate at which new disease has occurred in the population at risk per some unit time