Statistics Flashcards
What is sensitivity in diagnostic testing?
TP/(TP+FN)
-Sensitivity (true positive rate): chance that a diseased person will test positive
-Value approaching 100% is desirable for ruling out disease; indicates low false-negative rate
-SN-N-OUT: highly SeNsitive test, when Negative, rules OUT disease
-High sensitivity test used for screening
What is specificity in diagnostic testing?
TN/(TN+FP)
-Specificity (true negative rate): chance that a non-diseased person will test negative
-Value approaching 100% desirable for ruling in disease & indicates low false-positive rate
-SP-P-IN= highly SPecific test, when Positive, rules IN disease
-High specificity test use for confirmation after positive screening test
PPV
Positive Predictive Value, calculated as TP/(TP+FP). It indicates the likelihood that a positive result means the patient actually has the disease.
Predictive value – Depends on disease prevalence while Sensitivity and specificity does not
NPV
Negative Predictive Value, calculated as TN/(TN+FN). It indicates the likelihood that a negative result means the patient does not have the disease.
What is relative risk?
Relative risk is calculated as (Incidence of exposed/Incidence of unexposed) = (a/a+b)/(c/c+d). A relative risk < 1 indicates a negative association, > 1 indicates a positive association, and = 1 indicates no association.
What is the odds ratio?
The odds ratio is calculated as (a/c)/(b/d) = (AxD)/(BXC).
What is a Type I error?
A Type I error is a false positive that rejects the null hypothesis incorrectly, falsely assuming there is a difference when none exists.
What is a Type II error?
A Type II error is a false negative that accepts the null hypothesis incorrectly, often caused by a small sample size.
What is the power of the test?
The power of the test is the probability of making a correct conclusion, calculated as 1 - probability of Type II error (Beta). Larger sample size increases power
What is the null hypothesis?
The null hypothesis states that no difference exists between two groups.
What does a p-value < 0.05 indicate?
A p-value < 0.05 indicates rejection of the null hypothesis, suggesting a > 95% likelihood that the difference between populations is true and did not occur by chance.
What is variance?
Variance is the spread of data around a mean.
What is a cohort study?
A cohort study is a prospective study that compares patients with exposure risk factors to those without, assessing disease occurrence.
What is a case-control study?
A case-control study is a retrospective study comparing patients with and without a disease, looking for exposure risk factors.
What is Kaplan-Meier analysis?
Kaplan-Meier analysis estimates survival in small groups.
What is primary prevention?
Primary prevention aims to avoid disease altogether, such as through vaccination.
What is secondary prevention?
Secondary prevention focuses on early detection, such as colonoscopy and mammograms.
What is tertiary prevention?
Tertiary prevention aims to decrease morbidity related to established disease, such as controlling blood pressure in hypertensive patients.
What are the phases of clinical drug trials?
Phase I studies demonstrate safety on a small group, Phase II assesses effectiveness on a larger group, Phase III compares the new drug to standard care, and Phase IV investigates long-term effects.
What is a normal distribution?
A normal distribution (Gaussian) is a symmetric bell-shaped curve. 1 SD includes 68% of samples, 2 SD includes 95%, and 3 SD includes 99.7% around the mean.
What is a propensity-score-matched study?
-Propensity-score-matched study: observational study that attempts to reduce confounding and be closer to a randomized controlled trial; cases selected so subjects in both groups similar across multiple factors
What is a meta-analysis?
A meta-analysis pools data from multiple studies using quantitative statistical methods.
What is the receiver-operating curve (ROC)?
The ROC analyzes the effect on sensitivity and specificity when adjusting the cut-off value of a test.
Comparing variables
Selection bias
Nonrandom sampling or treatment allocation of subjects such that study population is not representative of target population
Berkson bias—cases and/
or controls selected from hospitals (bedside bias) are less healthy and have different exposures
Attrition bias—participants lost to follow up have a different prognosis than those who complete the study
Reduce bias: Randomization
Recall bias
Awareness of disorder alters recall by subjects; common in retrospective studies
Patients with disease recall exposure after learning of similar cases
Reduce: Decrease time from exposure to follow-up
Hawthorne effect
Participants change behavior upon awareness of being observed
Observer-expectancy bias
Researcher’s belief in the efficacy of a treatment changes the outcome of that treatment (aka, Pygmalion effect)
An observer expecting treatment group to show signs of recovery is more likely to document positive outcomes
Reduce: blinding (masking) and
use of placebo
Confounding bias
Factor related to both exposure and outcome (but not on causal path) distorts effect of exposure on outcome (vs effect modification, in which the exposure leads to different outcomes in subgroups stratified by the factor)
An uncontrolled study shows an association between drinking coffee and lung cancer; however, people who drink coffee may smoke more, which could account for the association
Reduce: matching, crossover study
Lead-time bias
Early detection interpreted as survival, but the disease course has not changed
Breast cancer diagnosed early by mammography may appear to exaggerate survival time because patients are known to have the cancer for longer
Reduce: adjust survival according to the severity of disease at the time of diagnosis
Length-time bias
Screening test detects diseases with long latency period, while those with shorter latency period become symptomatic earlier
Quaternary disease prevention
Quit (avoid) unnecessary medical interventions to minimize incidental harm (eg, imaging studies, optimizing medications to reduce polypharmacy)
NSQIP
Collects outcome data to measure and improve surgical quality; outcomes are reported as observed/ expected ratios
JCAHO
Prevention of wrong site/ procedure/ patient protocol
Preop verification of patient and procedure
Operative site and side marking
Time out: before incision is made: verifying patient, procedure, position site + side, availability of implants or special requirements
RFs for retained object after surgery
MC= sponge
emergency procedure, unplanned change in procedure, obesity, towel used for closure
Sentinel event
Unexpected occurrence involving death or serious injury or the risk thereof; hospital undergoes root cause analysis to prevent and minimize future occurrences (e.g. wrong site surgery)
Gap protection technique
gaps in care (change in caregiver, divisions of labor, shift changes, transfers)
can lead to loss of information and error
Prevention: structured handoffs, checklists (face to face best), standardized orders; reading back orders if verbal
Patient safety errors
Plan-Do-Study-Act
-Plan-Do-Study-Act:
-Develops and implents a quality initiative, and analyzes the results, and adjusts the iniative based on the results
-Plan: develop the initiative
-Do: implement the plan
-Study: analyze the results
-Act: adjust the process based on the results
-Six Sigma:
DMAIC, Define, Measure, Analyze, Improve, Control
-Focus on improving existing processes
-Situation-Background-Assessment-Recommendation:
Provides framework for communication between members of the health care team about a patient’s condition
-Root Cause Analysis:
-Identifies the factors that caused a failure and should be eliminated through the process
-Failure Mode and Effects Analysis:
-Prospectively uses a step-by-step approach for identifying all possible failures in a design or process
