Fiser Chapter 44 STATISTICS Flashcards
Meta-analysis
combines data from different studies
RCTs avoid ____
avoids treatment biases
Promoting culture of safety
Confidential system of reporting errors
Emphasis on learning over accountability
Flexibility in adapting to new situations or problems
Case-control study
Those who have disease are compared with those who do not
Compare frequency of suspected risk factor
Retrospective
Paired t tests
Before and after studies (weight before and after, drug versus placebo)
Quantitative
p
null hypothesis rejected
> 95% likelihood that the difference between the populations is true
AKA
Parameter
population
Kaplan-Meyer
estimates survival in small groups
NPV
TN / (TN + FN)
Relative risk
incidence in exposed / incidence in unexposed
Variance
spread of data around a mean
Sensitivity
TP / (TP + FN)
Student’s t test
Compares means between 2 independent groups
Quantitative
Accuracy
???
Sentinel event
Unexpected occurrence involving death or serious injury, or risk thereof
Hospital undergoes root-cause analysis to prevent and minimize future occurrences
eg wrong site surgery
Risk factors for retain object
Emergency procedure
Unplanned change in procedure
Obesity
Towel used for closure
Chi-squared test
Compares 2 groups with categorical variables
e.g. number of obese patients with and without diabetes versus number of nonobese patients with and without diabetes
GAP protection technique
gaps in care (eg change in caregiver, divisions of labor, shift changes, transfers) can lead to loss of information and error
Prevention: structured handoffs and checklists (face to face if possible); standardizing orders; reading back orders if verbal
Type 1 error
no difference (but thinks there is)
Mean
average
Incidence
Number of new cases diagnosed over a certain time frame in a population
Specificity
TN / (TN + FP)
Mode
most frequently occurring value
Median
middle value (50th %ile)
Cohort studies
compare disease rate between exposed and unexposed groups
Nonrandom
Prospective
ANOVA
Compares means between more than 2 groups
Quantitative
Power of test
Probability of making correct conclusion
1 - probability of type 2 error
1 - probability of not seeing difference when there is one
Likelihood that conclusion is true
Larger sample size increases power
Nonparametric statistics
compares categorical (qualitative variables)
E.g. race, sex, medical problems, medications
Predictive values depend on ___
disease prevalence
Type 2 error
there’s a difference (but don’t see it)
due to small sample size
Prevalence
number of people with disease in a population
higher in chronic diseases
Double-blind controlled trial avoids ____
avoids observational biases
PPV
TP / (TP + FP)
JCAHO prevention of wrong site/procedure/patient protocol
Time out before incision
Pre-op verification of patient and procedure
Operative site and side
Availability of implants or special requirements
