Chapter 7: Postanalysis Flashcards
Indicate problem with the specimens or an issue with the result
Prevent the release of erroneous laboratory
Alarms and Flags
Additional Steps in Flagging Specimens
- by automated instrument itself
- by specialized middleware
- by laboratory information system
Flags for Problem Specimen
- sample contain amounts inadequate for reliable analysis
- presence of high concentration of interfering substances in the specimen
- instrument that can often analyze more than 100 samples per hour
Automated Cell Counters
What type of samples can be reported immediately?
Normal Sample or Samples that show Quantitative Abnormalities
Flagged for preparation of blood smear and further evaluation
Samples that could contain qualitative abnormalities
What happens to an analyte concentration outside the validated linear range?
- increase in signal linearity related to inc. in conc.
- above the linear range, the machine dilutes and renanalyze sample
- below LR, sample is reported as “less than the limit of detection”
The process of conparing a current laboratory result with results obtained on a previous specimen from the same patient
Delta Checks
Types of Errors Detected by Delta Checks
- preanalytic and analytic issues
May not be readily detectable
Must be reported immediately to a healthcare provider
Critical Values
Required rapid communication of laboratory results
Federal Law, Regulatory Agencies, and Joint Commission
Develops a crticial values policy that meets the needs of the patients and staff served by the laboratory
Medical Director of the Laboratory
Definition of Reference Intervals
Laboratory Results vs Reference or Normal Range
- Range of values into which 95% of non-diseased individuals will fall
- *some of these analytes is defined as “greater than” or “less than” a certain value
- some have been defined by prof. org. w/out adherence to the 95% rule
Reference Ranges
Does not require reference intervals except if the patient population was clearly distinct and exhibited range of values
Standardization
The Joint Committee for Traceability in Laboratory Machine establishes process for standardization by:
- Identifying
- Reviewing against agreed criteria
- Publishing list(s) of Higher Order Certified Reference Materials and Reference Measurement Procedures
Calibrating and using materials traceable to isotope dilution mass spectrometry reference measurement procedure
Reduce Interlaboratory Variability
Factors that Influence Reference Ranges
Age Genetic background Exposure to environmental factors Sample collection container Sample transport Time between specimen collection and analysis Sample storage before analysis
Determination of Reference Ranges
- Testing at least 120 SAMPLES from nondiseased individuals in each “partition”
- Transference
- Verication by another lab’s or the manufacturer’s reference interval if the analyte was not previously tested for the lab
Verification of a reference interval that was previously established for a different method
Transference
Result of assay imprecision
Analytic Variability
Due to biologic changes that cause analyte levels to fluctuate over time
Intraindividuality Variability
Occurs becauuse of factors specific to individual patients
Interindividuality Variation
The sum analytic and intraindividuality variability
Random Variability
Used for disease classification (positive or negative)
Threshold
Based on the results that are seen in 95% of the healthy population
Reference Range / Interval
Determined by comparing test’s ability to discern true disease from nondisease
Diagnostic Accuracy
Used to discern true disease from nondisease
Diagnostic Gold Standard
The non-overlapping areas of the two patient distribution.
True Results
Classified as Abnormal
True Positive (TP)
Classified as Normal
True Negative (TN)
Used to discriminate disease from normal populations
Single Cutoff
The overlapping areas of the two patient distribution
False Results
INCORRECTLY classified as NORMAL
False-Negative (FN)
INCORRECTLY classified as ABNORMAL
False-Positive (FP)
The error of True Positive
False-Negative
The error of True Negative
False-Positive
Measures of diagnostic accuracy
Indicators in distinguishing the presence and absence of disease at a chosen cutoff
Sensitivity and Specificity
● The ability of a test to detect disease.
● The proportion of persons with the disease.
● TRUE POSITIVE (TP)
● Identifies a greater proportion of persons with the disease.
Sensitivity
● The ability of a test to detect the absence of disease
● The proportion of persons without the disease.
● TRUE-NEGATIVE (TN)
● Excludes a greater proportion of persons without the disease.
Specificity
Effect of Altering the Test Cutoff
● Altering the cutoff changes a test’s sensitivity and
specificity.
● An inverse relationship between SENSITIVITY
AND SPECIFICITY is noted.
For tests where high values indicate disease,
lowering cutoff (cutoff line moved to the left)
will lead to more diseased patients being
classified as abnormal.
HIGH SENSITIVITY
If the cutoff is raised (the cutoff line moved to the
right), more non-diseased patients are classified
correctly.
HIGH SPECIFICITY.
The Need for High Sensitivity vs High Specificity
● False results such as FP (false-positive) and FN
(false-negative) can lead to misdiagnosis and
inappropriate clinical management.
● SENSITIVITY should be HIGH to capture the
majority of cases. (Lowering cutoff)
● SPECIFICITY can be INCREASED to exclude all
persons without the disease. (Increasing cutoff).
sometimes referred to as positive predictive
value
may be understood as the probability that a positive test indicates disease
it is the proportion of persons with a positive
test who have truly the disease.
PREDICTIVE VALUE OF A POSITIVE TEST
Referred to as negative predictive value
Is the probability that a negative test indicates
absence of disease
It is the proportion of persons with a negative test who are true without disease
PREDICTIVE VALUE OF A NEGATIVE TEST
Relationship of Prevalence and Posttest Probability
The higher the prevalence, or pretest probability, the higher the posttest probability, or predictive value of a positive test.
- Shows that sensitivity and specificity influence the predictive value.
- Describes the relationship between posttest and pretest probability of disease or no disease based on the sensitivity and specificity of the test
Bayes Theorem
Also known as priori probability, is the prevalence of the disease in the patient’s clinical setting
- used in conjunction with the characteristics of diagnostic accuracy as summarized in the sensitivity and specificity of the test.
Pretest probability
Also known as posteriori probability, is the probability of disease in the posttest situation and is commonly referred to as the predictive value of the test.
Posttest probability
● A convenient measure that combines sensitivity
and specificity into a single number
specificity into a single number.
● An assessment of test performance, and not of
disease status, in the patient being tested.
Likelihood Ratio
The LR+ is the ratio of two probabilities: the probability of a positive test result when the disease is present (TP) divided by the probability of the same test result when the disease is absent (FP).
Likelihood Ratio of a positive test (LR+)
A convenient graphic tool that uses a logarithmic
scale to determine posttest probability, given
the LR at a specified cutoff and the pretest
probability
FAGAN NOMOGRAM
a useful tool for identifying the optimal cutoff
for a diagnostic test by calculating
the sensitivity and specificity combinations
across the entire range of cut off values
RECEIVER OPERATOR CHARACTERISTIC CURVE
a method that allows one to assess the optimal cutoff
with numeric estimates for clinical impact, or
consequences, of test results.
POSITIVITY CRITERION
• Is a process by which medical decision can be
made by using as many objective tools as
possible.
• Can help to reduce the uncertainty of medical
decision making.
EVIDENCE-BASED MEDICINE
5 Steps of EBM
- Ask a clinical question based on a patient
encounter - Acquire information by searching for resources
- Analyze and critically evaluate the information and reach a conclusion
- Apply the information to individual patients
- Audit effectiveness and monitor the literature
The clinical question can be described in four parts with
acronym of PICO
- Problem
- Intervention
- Comparison
- Outcome
