BNUR 2003 Biomedical Chemistry & Lab Diagnostics Student Notes PDF

BNUR 2003 Biomedical Chemistry & Lab Diagnostics Unit 1- Introduction to biomedical chemistry 1-1 Cell structure and function 1-2 Patient care cycle  2 Learning Objectives 1. Understand how clini...

BNUR 2003 Biomedical Chemistry & Lab Diagnostics Unit 1- Introduction to biomedical chemistry 1-1 Cell structure and function 1-2 Patient care cycle  2 Learning Objectives 1. Understand how clinical chemistry relates to the patient care cycle and describe what differential, provisional, and definitive diagnoses are. 2. Describe the three major steps involved in quality assurance in clinical chemistry labs. 3. Describe, in detail, the analytical phase of QA, differentiate between accuracy and precision and explain how laboratories evaluate them during the analytical phase of QA. 4. Describe what a reference interval is and how a reference interval is established. Calculate and interpret various statistics of diagnostic relevance (true Department of Laboratory Medicine 3 The primary goal of the laboratory is to do the right test as accurately as possible on the correct patient in an appropriate time frame in order to affect a positive patient outcome. y Micro g bi o l o gy to lo ma a l He Cl i n i c i s try Histology chem o h ema to l o Cyto lo Immun gy gy  4 Clinical Chemistry Clinical Chemistry is a branch of laboratory medicine that examines chemical or biochemical information about the health status of a patient. We seek to understand the relationships between: Patients’ clinical signs and symptoms Underlying biochemical/physiological phenomena Laboratory results The Patient Care Cycle 5 Step 1 Individual becomes aware of a problem by assessing their state of health (symptoms). Step 2 Individual seeks medical care and patient information is obtained (patient interview). Step 3 Clinician seeks objective data to establish a diagnosis (tests). In 20% of patients, there is an uncertainty in a diagnosis established from signs & symptoms alone The Patient Care Cycle 6 Step 4 Laboratory tests are performed. Step 5 Laboratory tests are interpreted. Step 6 Therapy implemented and evaluated. Types of Diagnosis 7 Differential Diagnosis: list of most probable disease entities that are consistent with observed signs and symptoms. Many conditions could be consistent with the S&S Provisional Diagnosis: The most likely of possible diagnoses. Definitive Diagnosis: Established by confirmatory data such as laboratory tests. Quality Assurance (QA) 8 Q: Are laboratory tests always “correct”? A:  9 Quality Assurance (QA) Labs must carefully monitor the quality of their results. This occurs in three phases: Pre-analytical phase Assess controllable _________ that influence test results e.g., patient preparation, specimen collection, interferences Analytical phase Assess accuracy and precision due to processing of the actual test Post-analytical phase Compare test results to external data after the test is complete.  10 QA: Analytical Phase The laboratory must: use appropriate equipment and methods that can attain goals of accuracy and reproducibility implement an internal ____________ system for each test that monitors the reproducibility and accuracy of test results being reported QA: Analytical Phase 11 Accuracy The closeness of a result to the true value Precision The degree to which repeated measures under the same conditions return the same result Repeated tests on a laboratory standard of known glucose serum Accuracy Test 1: 3.7 mmol/L Test 1: 2.0 mmol/L Test 2: 3.6 mmol/L Test 2: 2.1 mmol/L Test 3: 3.8 mmol/L Test 3: 2.0 mmol/L Test 4: 3.7 mmol/L Test 4: 2.0 mmol/L Precision Test 1: 4.0 mmol/L Test 1: 0.2 mmol/L Test 2: 3.3 mmol/L Test 2: 8.3 mmol/L Test 3: 4.0 mmol/L Test 3: 5.0 mmol/L Test 4: 3.4 mmol/L Test 4: 1.7 mmol/L 12 Laboratory Standard = 3.7  QA: Analytical Phase – Accuracy Accuracy is established when a new method is introduced for the first time Three ways to establish accuracy: 1. Compare a patient’s test result using the new method to a result of same sample using a different, well established method. 2. Perform the new method on a series of ___________________ (e.g., linearity check samples). 3. External quality control programs 13 QA: Analytical Phase – Precision 14 The precision of a method is determined as part of internal quality assurance programs. These programs are critical in determining whether patient results are ________________. Step 1: Establish precision of the new method by Example: performing the new method on two commercial control samples (normal and abnormal) to determine mean values and standard deviation. Step 2: Run a patient’s sample with commercial controls. Step 3: Determine whether the patient’s results are reportable by using both normal and abnormal results  15 QA: Analytical Phase – Precision Step 1: Establish precision Control Values over 1 month Mean Value = 2.63 2.0 3.1 2.3 2.4 3.5 Standard Deviation 2.2 2.2 2.9 2.7 3.3 (S.D.) = 0.37 2.4 2.5 2.5 2.7 3.0 Mean +/- 1 S.D. is the 2.6 2.6 2.8 2.0 3.2 range between 2.26 an 3.0 2.7 2.6 2.6 2.6 3.00 2.9 standard The 2.7 2.3 2.3 is2.4 deviation a useful statistic. Statistically speaking, approximately 66% of points fall within one standard deviation of the mean. E.g. 2.63 - 0.37 = 2.26 2.63 + 0.37 = 3.00  16 QA: Analytical Phase – Precision Step 1: Establish precision Control Values over 1 Mean Value = 2.63 month 2.0 3.1 2.3 2.4 3.5 Standard Deviation 2.2 2.2 2.9 2.7 3.3 (S.D.) = 0.37 2.4 2.5 2.5 2.7 3.0 2.6 2.6 2.8 2.0 3.2Mean +/- 2 S.D. is the 3.0 2.7 2.6 2.6 range between 1.89 and 2.6 2.9 2.7 2.3 2.3 2.4 3.37 Approximately 95% of points fall within two standard deviations of the mean. E.g. 2.63 – 2(0.37) = 1.89 2.63 + 2(0.37) = 3.37 QA: Analytical Phase – Precision 17 Step 2: Run a patient sample Run the quality control samples (normal + abnormal) with patient samples. The frequency that this is done depends on instrument, technologist, and laboratory policy. At the minimum once per 8 hour shift. QC data (normal + abnormal) are plotted on a Levey- Jennings chart*, which is based on the data from Step 1. *The LJ chart does not include patient samples  18 QA: Analytical Phase – Precision Step 2 – Levey-Jennings charts ‘3.37’ "2sd" "2sd" ‘3.00’ "1sd" "1sd" ‘2.63’ mean mean ‘2.26’ -1sd -1sd -2sd -2sd ‘1.89’ Day 1 2 34 5 6 7 Day 1 23 4 5 67 normal abnormal Mean Value = 2.63 (Abnormal would have a different +/- 2 S.D. =1.89 to mean value) 3.37  19 QA: Analytical Phase – Precision Levey-Jennings chart over the course of one month Decreasing trend at end of August  20 QA: Analytical Phase – Precision Step 3: Decide whether to report patient’s result We can use a series of rules known as the Westgard Rules to decide whether or not a patient’s result should be reported (i.e., is reliable).  21 QA: Analytical Phase – Precision Step 3: Westgard Rules Rule 1: If both controls are within 2SD, accept the run and report patient result +3 +3 sd +2 sd +2 sd +1 sd +1 sd Mea sd Mea n -1 n -1 sd -2 sd -2 sd -3 sd -3 sd sd  22 QA: Analytical Phase – Precision Step 3: Westgard Rules Rule 2: If one control is within 2SD and the other is no more than 3 SD from the mean, accept the run and report the patient results. BUT, check next controls carefully. +3 +3 sd +2 sd +2 sd +1 sd +1 sd Mea sd Mea n -1 n -1 sd -2 sd -2 sd -3 sd -3 sd sd QA: Analytical Phase – Precision 23 Step 3: Westgard Rules Rule 3: If either control is outside of 3SD, reject the run and do not report the patient’s results +3 +3 sd +2 sd +2 sd +1 sd +1 sd Mea sd Mea n -1 n -1 sd -2 sd -2 sd -3 sd -3 sd sd QA: Analytical Phase – Precision 24 Step 3: Westgard Rules Rule 4: If both controls are outside 2SD, reject the run and do not report the patient’s result. +3 +3 sd +2 sd +2 sd +1 sd +1 sd Mea sd Mea n -1 n -1 sd -2 sd -2 sd -3 sd -3 sd sd  25 QA: Analytical Phase – Precision Step 3: Westgard Rules Rule 5: If four quality control results in a row, for either the normal or abnormal control, fall on the same side of the mean above or below 1SD, reject the run. +3 +3 sd +2 sd +2 sd +1 sd +1 sd Mea sd Mea n -1 n -1 sd -2 sd -2 sd -3 sd -3 sd sd QA: Analytical Phase – Precision 26 Step 3: Westgard Rules Rule 6: If ten consecutive results for either quality control sample fall on the same side of the mean, reject the run. +3 +3 sd +2 sd +2 sd +1 sd +1 sd Mea sd Mea n -1 n -1 sd -2 sd -2 sd -3 sd -3 sd sd QA: Analytical Phase – Precision 27 Levey-Jennings chart over the course of one month at 2 SD below 2 SD QA: Post-Analytical Phase 28 Two examples of post-analytic QA: 1. Test patterns/grouping: Multiple markers respond to disease at the same time. e.g., renal failure is accompanied by both elevated creatinine and urea nitrogen 2. Delta Values: Compare test result to previous test results for the same patient.  29 Interpreting a patient’s result We have one large question remaining “Does the patient’s test result indicate whether the patient can be diagnosed as likely having the disease or not?”  30 Let’s try… Q: Bob has a glucose level of 4.0 mmol/L. Is Bob’s glucose level normal? Q: Chandra has a glucose level of 3.7 mmol/L and the population average is 4.9 mmol/L. Is Chandra’s glucose level normal? Q: Gideon has a glucose level of 7.0 mmol/L and 95% of the population has glucose levels between 3.6 and 6.1 mmol/L. Is Gideon’s blood glucose level normal? Interpreting a patient’s result 31 We need to understand sources of variation: We are usually interested in the effect of disease state on test results, but variation could also be due to: Analytical variation: Variation that is inherent in the test method. Biological variation: Inter-individual variation can include age, sex, race, diet. Within individual variation can include time of day, social environment, diet, etc…  32 Interpreting a patient’s result The most common method of establishing whether a patient’s result should be considered “normal” is the reference interval. Reference intervals are constructed from test populations that must consist of more than ____ individuals, be well described (age, sex, etc.) and be measured with clearly stated techniques. Interpreting a patient’s result 33 The reality is that the diseased and non- diseased populations usually overlap F Non-Diseased R E Q U E N Diseased C Y 10 100 140 140 Test Values  34 Interpreting a patient’s result Overlapping populations lead to the possibility of our test having False positives and False negatives. Ov e rl ap is w he re the false resu lt happ s en ↓ overlap = better test  35 Interpreting a patient’s result Some tests are better than others due to varying degrees of overlap between the reference intervals for the healthy and diseased population. We can use test parameters of diagnostic relevance to quantify how good a test is. Interpreting a patient’s result 36 Test Parameters Sensitivity = Proportion of people with disease that test positive for the disease How good the test is at correctly identifying people with the disease Calculations only involve people who have the disease (doesn’t tell us if people without disease test positive) aka – “True Positive Rate” or “TPR”  37 Interpreting a patient’s result Test Parameters Specificity = Proportion of people without disease that test negative for the disease How good the test is at correctly identifying people without the disease Calculations only involve people who do not have the disease (doesn’t tell us if people with disease test negative) aka – “True Negative Rate” or “TNR”  38 Interpreting a patient’s result Test Parameters Predictive Values are the probability of having disease after the tests results are known. Positive Predictive Value (PPV) = aka – “Predictive Value Positive Result” or “PVPR” (textbook) Proportion of people with a positive test result who have the disease Sometimes referred to as the ‘post-test probability of disease given a positive test’. Interpreting a patient’s result 39 Test Parameters Predictive Values are the probability of having disease after the tests results are known. Negative Predictive Value (NPV) = aka – “Predictive Value Negative Result” or “PVNR” (textbook) Proportion of people with a negative test result who do not have disease Sometimes referred to as the ‘post-test probability of not having disease given a negative test’. Interpreting a patient’s result 40 Test Parameters False Positive Rate (FPR) = Number of positive test results obtained as a percentage of all people without disease. Proportion of people without disease that test positive for disease. Interpreting a patient’s result 41 Test Parameters Efficiency = Number of accurate test results obtained as a percentage of all people tested. Proportion of patients that receive accurate test results. Interpreting a patient’s result 42 Test Parameters Prevalence = Proportion of population that has disease.