Chapter 4 Evidence-Based Patient Care Flashcards
Mean difference or weighted mean difference
Used to compare cases and controls when units of measurements are identical between studies
Standard mean difference
Used to compare cases and controls where differences are normalized with respect to pooled standard deviation of the two groups
Can be used to compare different measurements
Selection bias
-Type of bias in study design
-certain patients are preferentially enrolled into the study and may not represent population at large
Channeling bias
-Type of bias in study design
-patients are steered into study arms by organizers based on preconceived notions
-mitigated by blinded randomization
Interviewer bias
-Type of bias in study design
-investigators asks different questions of the patients, knowing which study arm they are in
-mitigated by double-blind design
Hawthorne effect
-Type of bias in study design
-Subject may behave differently because he knows he is being observed
-mitigated by double-blind design
Chronology bias
-Type of bias in study design
-study patients are compared to historical controls at a time when standard of care may have been different
-mitigated by prospective design
Recall bias
-Type of bias in study design
-subjects may not remember their symptoms or exposures correctly
-mitigated by using objective data sources
Transfer bias
-Type of bias in study design
-certain types of patients are more likely to be lost to follow-up
-also known as attrition bias
-mitigated by careful plan to follow-up with patients
Performance bias
in cases where the intervention involves a procedure, possible for different operators to perform the procedure differently
Publication bias
-Type of bias in study design
-journals tend to report positive findings more readily than negative findings
p-value
Probability that the results found in the study could be a result of pure chance
Odds ratio
used in case-control studies when prevalence is not known
Grading of Recommendations Assessment, Development and Evaluation (GRADE)
-Code A (high quality, further research is unlikely to change our confidence in the estimate of effect), Code B (moderate quality, further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate), Code C (low, further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate), Code D (very low (any estimate of effect is very uncertain
-Assess studies by: study design, study quality (bias, flaws), consistency (subgroup analysis), directness (population differences, use of surrogate markers, no direct comparison), strength of association, evidence of dose response gradient, when reasonable confounding would have pushed effect in one direction
Evidence-based Practice Center (EPC) approach
-Supported by Agency for Healthcare Quality and Research (AHRQ)
-Similar to GRADE
-Assesses domains of: risk of bias, consistency, directness, precision
Precision
Refers to lack of random errors in measurement
Validity
Refers to lack of systematic error
Internal validity
Relates to ability of study’s interventions to explain the findings
External validity
Refers to the ability of the study to generalize to other populations
Strength-of-Recommendation Taxonomy (SORT)
Code A: consistent, good-quality patient-oriented evidence
Code B: Inconsistent or limited-quality patient-oriented evidence
Code C: consensus, disease-oriented evidence, usual practice, expert opinion, or case series for studies of diagnosis, treatment, prevention, or screening
Clinical practice guidelines (CPG)
-Statements that include recommendations intended to optimize patient care that are informed by a systematic review of evidence and an assessment of the benefits and harms of alternative care options
-Typically includes a systematic review of the research evidence, and set of recommendations
National Guideline Clearinghouse (NGC)
Public resource maintained by the AHRQ, to be included guidelines must follow Institute of Medicine (IOM) definition of CPG and include systematic review and assessment of benefits/harms
CPG algorithm
Action (perform specific action)
Conditional (carry out action based on defined criteria)
Branch (direct flow to one or more additional steps)
Synchronization (converge paths back from branches to a common outcome/end point)
National Guideline Clearinghouse practice guidelines levels of evidence and grades of recommendation
-Page 82
-Grades of Recommendations are based on the Levels of Evidence
Comparative Effectiveness Research (CER)
Key elements are:
-Direct comparisons of active treatments
-Study patients, clinicians, and interventions that are representative of usual practice
-Focus on helping patients, clinicians, and policy makers to make informed choices
Appropriate Use Criteria
-Variation on clinical practice guidelines, but differ in various ways (including rely more on expert opinion)
Moulding multistep process for adoption of clinical changes
-Described in 1999
-Knowledge: individuals exposed to innovation
-Persuasion: individual seeks more information about innovation
-Decision: decide whether or not to adopt or reject innovation
-Implementation: tries out innovation
-Confirmation: finalizes decision to continue using innovation
Macdonald 5 Stages of Assessment
-Step 1: Assessment of Practitioners’ Stage of Readiness to Change
-Step 2: Assessment of Specific Barriers to Guideline Use
-Step 3: Determination of Appropriate Level of Intervention
-Step 4: Design of Dissemination and Implementation Strategies
-Step 5: Evaluation
Knowledge-to-Action Framework
-As adapted for implementation of Best Practice Guidelines
1) Identify the problem
2) Adapt knowledge, tools, and resources to local context
3) Assess barriers and facilitators to knowledge use
4) Select, tailor, and implement interventions
5) Monitor knowledge use
6) Sustain knowledge use
Information retrieval
-process of dinging material of an unstructured nature that satisfies an information need from within large collections
-users start with information needs form which they compose queries, search a body of knowledge known as a corpus for individual documents, the individual documents are tokenized and stored into an index for rapid retrieval
Medical Subject Heading (MeSH) dictionary
Assigned by independent indexers from the National Library of Medicine and are hierarchical
Stopwords
words that are deliberated excluded from an index
Tokenization
Process of breaking up document into individually searchable items