Critical Inquiry - Exam 1

class EBP triangle

current best evidence, clinical expertise, and client/patient values

empiricism

empirical knowlegde derived from direct experience, observation, and experimentation

independent variable

possible causative factor, variable that is varied or manipulated by researcher

dependent variable

quantifiable characteristic, response that is measured

wide range, discrete number of categories, dichotomous classification

internal validity

extent to which change in DV can be attributed to IV , control of extraneous/confounding variable

covariates

possible predictive variable, can be direct interest (IV) or confounding variable

external validity

extent to which results can be generalized

internal and external validity note

lacking internal validity is hard to get published

without external validity lack clinical significance

mean

average

median

middle value

mode

value that occurs the most often

which is more appropriate if outliers

median

qaulitative research

often applied to any research that uses methods for study of human behavior

type 1 error

false positive

type 2 error

false negative

power

probability that test will show a true effect when one exists
inversely related to type 2 error

the higher the power the lower the probability of making type 2 error

can be increased by increased sample size

clinical trial

research project that prospectively assigns human subjects to interventions or comparison groups to study cause effect relationship

RCT

control of extragenous variables (strong internal validity)

random assignment of subjects experimental vs control groups

equipoise

major limiting factor, alternative treatments must be equally acceptable on basis of current knowledge

standard deviation

how much variation from the mean small standard deviation means that values in stats are close to mean

within 1 SD , 2 SD ,3 SD

68%, 95%, 99%

biomedical model

focus on etiologic agents, pathological processes, physiological measurements, understand causation for diagnosis and treatment

social science and QOL model

focus on dimensions of functioning and overall well being

outcome

change in patient current or future health that can be attributed to antecedent health care

• poor internal validity

FOTO

database contains clinical data from clinics to do outcomes

clinical epidemiology

applying principles of population based evidence to the management of individual patients

science of making predictions about individual patients by counting clinical events in groups of similar patients

clinical prediction rules

quantification of relative importance of factors that may effect patients outcome

prevalence

proportion of population that exhibits a specified condition at any given time

incidence

proportion of population that sustains a specified injury within a specific period of time

probability

based on prevalence in particular population, clinical presentation

based on incidence, injury susceptibility

criterion positive case

true positive, false negative

crtierion negative

true negative, false positive

sensitivity

identification of condition that exists

specific

exclusion of condition that does not exist

high specificity and postivie result

rule in

high sensitivity and negative results

rule out

likelihood ratio

reflect magnitude of difference in proportions of true vs false results

likelihood that a given test result would be expected in a patient with the disorder compared to the same result expected without disorder

LR of 1 lacks diagnostic value , over 1 increase odds for criterion positive, less than 1 is decrease odd for criterion positive

relative risk

ratio of probability for specified outcome for one group in relation to that of another group

same= 1

over 1 RR increase risk

less than 1 decreased risk of outcome

odds ratio

ratio of odds for specified outcome for one group in relation to the odds for another group

case control studies

MCID

smallest difference in score in domain of interest which patient perceive as beneficial

efficacy

level of benefit observed, under ideal conditions, defined population of subjects

effectiveness

level of benefit expected when in normal practice setting and typical patients

strongest evidence vs weakest evidence

systematic review and double blind RCT, expert opinion

what is the pyramid of research types starting with highest

systematic review/meta analyses, RCT, cohort studies, case control study, case series/case reports, editorials/expert opinion

meta analysis

quanitative analysiss that is focused on contrasting and combining results from different studies in the hope of identifiy patterns amoung results, sources of disagreement amoung those results, or other interesting relationships that may come to light in the context of multiple studies

forest plot

graphical representation of meta analysis

characteristics of RCT

random assignment of patient, prospective manipulation of variables, precise measurement of variables, rigorous control of extraneous variables, strong internal validity, provides evidence of treatment efficacy

quasi study

Intervention + NO random assignment → compare groups

used when randomization is not ethical

delphi technique

method of group decision making and forecasting that involves successively collating the judgements of experts

practice reccommendation, search categories, PT education

epidemiology

study society’s collective health

study of pattern, causes, effects of health and disease in defined populations

cohort (common in medical research) study

group of defined people with common experience over a defined period of time

-usually prospective (cause to effect hypothesis)

varaibility in subject exposure to risk factors, not exposed then control, large number of subjects and ling duration required

observational (quasi experimental) cohort study

analytic epidemiology, observation to determine differences in incidence of injury in relation to risk factors

observational (quasi experimnental) case control study

usually retrospective (effect to cause hypothesis)

closely match as many factors as possible

odds ratio

measures the association between exposure and a disease, greater value the greater the association

case series

information on series of patient having same injury/illness

no control or comparison group

case report

information about a single patient

good for unusual events or rare cases

what is the gold standard of systematic reviews

cochrane

hierarchy of evidence from originial journal articles

original journal articles, evidence based journal abstracts, systematic reviews, evidence based journal abstract, evidence based clinical practice guidelines/evidence based textbooks, computerized decision support system

what helps external validity

sampling, randomization, large sample size, long term outcomes, methods that are user friendly

nominal data

categories with no order (female and male)

ordinal data

categories with an order but the distance between ranks is not uniform (likert scale)

interval data

ordered precisely and continously, equal intervals between data but no true zero (temperature in centigrade)

ratio data

ordered precisely and continously and has natural zero point

like goniometry

nominal and ordinal vs interval and ratio

nominal and ordinal- quan/qual, nonparametric stats used to analyze

interval and ratio - quan, use parametric stats

what is PICO

patient population or problem

intervention

comparison

outcomes