How Reliable is Laboratory Testing?

What are the indicators of test reliability?

Four indicators are most commonly used to determine the reliability of a clinical laboratory test. Two of these, accuracy and precision, reflect how well the test method performs day to day in a laboratory. The other two, sensitivity and specificity, deal with how well the test is able to distinguish disease from absence of disease.

The accuracy and precision of each test method are established and are frequently monitored by the professional laboratory personnel. Sensitivity and specificity data are determined by research studies and are generally found in medical literature. Although each test has its own performance measures and appropriate uses, laboratory tests are designed to be as precise, accurate, specific, and sensitive as possible. These basic concepts are the cornerstones of reliability of your test results and provide the confidence your health care provider has in using the clinical laboratory.

Accuracy and Precision

Statistical measurements of accuracy and precision reveal a lab test’s basic reliability. These terms are not interchangeable. A test method can be precise (reliably reproducible) without being accurate (measuring what it is supposed to measure), and vice versa.

Precision (Repeatability) A test method is said to be precise when repeated analyses on the same sample give similar results. When a test method is precise, the amount of random variation is small. The test method can be trusted because results are reliably reproduced time after time. Picture a dartboard with darts all clustered together – but not at the bull’s eye – and you see what a precise but inaccurate method produces: the method can be counted on to reach the same target over and over again, but the target may not be the right one!

Accuracy (Trueness) A test method is said to be accurate when it measures what it was supposed to measure – in technical terms, the test value approaches the absolute “true” value of the substance (analyte) being measured. Results from every test performed are compared to known “control specimens” that have undergone multiple evaluations and compared to the “gold” standard for that assay, thus analysed to the best testing standards available. Picture a dartboard with a dart right in the centre of the bull’s eye and you see what an accurate method produces: the method is capable of hitting the intended target.

When the method is both precise and accurate – bull’s eye every time!

Although a test that is 100% accurate and 100% precise is the ideal, in reality, tests,, instruments, and laboratory operations all contribute to small but measurable variations in results. The small amount of variability that typically occurs does not usually detract from the test’s value and statistically is insignificant. The level of precision and accuracy that can be obtained is specific to each test method but is constantly monitored for reliability through quality control and assessment procedures. When your blood is tested repeatedly your test results should change little unless your state of health has improved or deteriorated. A slightly larger difference in precision and accuracy can often be seen between two laboratories and therefore your results may vary somewhat more when the repeat test is performed by a different laboratory.

Sensitivity and Specificity

To be effective, a medical test is expected to detect abnormalities with a very high degree of confidence. How likely is it that an individual who has a positive test has the disease? What are the chances that an individual has a certain disorder even though the test for it was negative?

Sensitivity
Sensitivity is the ability of a test to correctly identify individuals who have a given disease or condition. For example, a certain test may have been shown to be 90% sensitive. If 100 people are known to have a certain disease, the test that identifies that disease will correctly do so for 90 of those 100 cases. The other 10 people tested will not show the expected result for this test. For that 10%, the finding of a “normal” result is a misleading false-negative result.

A test’s sensitivity becomes particularly important when you are seeking to exclude a dangerous disease, such as testing for the presence of the HIV antibody. Screening for HIV antibody often utilizes an ELISA (enzyme-linked immunoassay) test method, which has greater than 99% sensitivity. However, a person may get a false-negative result if they are tested too soon after the initial infection (less than 6 weeks). The result of a false-negative gives a person the sense of being disease-free when in fact they are not. The more sensitive a test, the fewer “false-negative” results will be produced.

Specificity
Specificity is the ability of a test to correctly exclude individuals who do not have a given disease or condition. For example, a certain test may have proven to be 90% specific. If 100 healthy individuals are tested with that method, only 90 of those 100 healthy people will be found “normal” (disease-free) by the test. The other 10 people (who do not have the disease) will appear to be positive for that test. For that 10%, their “abnormal” findings are a misleading false-positive result. When it is necessary to confirm a diagnosis that requires dangerous therapy, a test’s specificity is one of the crucial indicators. A patient who has been told that he’s positive for a specific test when he’s not may be subjected to potentially painful or dangerous treatment, additional expense, and unwarranted anxiety. The more specific a test, the fewer “false-positive” results it produces.

Developers and manufacturers of a new test must provide target values for test results and must provide evidence for the expected ranges as well as information on test limitations and other factors that could generate false results.

What types of things could compromise the quality of a test and what is done to minimise them?

Many things can change laboratory tests which could adversely affect the reliability of the result and prevent the timely reporting of an accurate test result to your doctor. These influences can be placed into three groups which relate to a process from sample request to it being reported.

• Pre-Analytical is the term used to describe things that happen from the time the test is requested to the time the sample arrives in the lab.
• Analytical is the term used to describe the things that happen during the handling and analysis of the sample in the laboratory.
• Post-Analytical is the term used to describe what happens after a result is obtained and includes how and when it is reported to your doctor.

Pre-Analytical Activities
When a test is requested, a chain of events is set into motion. All these steps must be performed correctly to ensure that an acceptable sample arrives at the lab. The following are some of the areas where errors can occur, and standard procedures are in place to prevent such errors.

• Test requesting process – the doctor or nurse must request the correct test using the correct name or code.
• Patient preparation for the test – the patient should have received appropriate instructions about diet, fasting, medications, etc. to assure that the sample will not contain substances that interfere with the test.
• Patient identification – the person collecting the sample needs to confirm that the person is indeed the patient on whom the test was requested. In hospitals, patient identification is usually checked on a wrist band and conversation with the patient.
• Completeness of patient information - the sample must be carefully labelled with three identifiers (usually patient name and date of birth and medical record number) to ensure that the sample is from the correct patient and not one with the same or similar name.
• Specimen collection procedures - the sample must be collected in the correct type of container and mixed with the right preservative. The laboratory staff who collect samples receive special training on how to collect samples for each kind of test.
• Transport to the lab - some samples need to be kept cold and some need to be transported quickly in the laboratory, so careful handling and prompt transport are important for sample integrity.

Automated test requesting systems (“order communications”) are increasingly used to minimize errors in sample collection and test requests. Computer technology allows health care providers to quickly request tests themselves and minimise the chance of misunderstanding and error. These systems can generate labels that include a bar code for tracking and identification and can provide supplementary information on the size of sample needed, the types of containers or blood tubes to be used to collect the sample, and procedures for handling and transport.

One of the key steps to getting correct results is proper identification of the person from whom the sample is collected and the sample itself. Safety guidelines require the use of at least three different identifying methods (such as name, date of birth and patient ID number) for identifying both the person and their sample. When you are asked to give your name and some other identifying information, it is to check that you and your sample are correctly matched. If this isn’t done when a sample is taken, make sure you check to see that your sample is identified correctly. If you collect a sample (for example, an early morning urine collection) at home, make sure you use an appropriate clean container, and that the sample is labelled with your full name, date of birth and the time and date of collection.

For some tests, your diet, medications, and exercise history are important for proper interpretation of results. In some cases, special preparation must be taken to get reliable results for a test (such as fasting overnight before blood glucose and lipids are checked). A patient who does not follow preparation instructions or provides insufficient information to the doctor or laboratory undermines the entire quality assurance effort for a specific test.

Analytical Activities
Once the specimen is received in the laboratory, quality assurance procedures guide and monitor all processes, including the following:

• Instrument operations – all instruments are calibrated and tested regularly according to a standard protocol that covers quality control, maintenance, and operation.
• Valid test reagents – all test materials have an expiration date. The lab has routine procedures to make sure that test materials are in good condition and have not passed their expiration date.

Many test methods use automated analysers which minimise the opportunity for human error. Bar code readers read the sample tube label and associate the result with the correct patient. A robotic system carefully removes the exact amount of blood or urine needed for the test. The onboard computer monitors the test as it is in progress to assure that there are no unexpected occurrences during the analysis. Most instruments have internal computer surveillance systems to detect malfunctions or other discrepancies and bring them to the attention of the laboratory staff.

Post-Analytical Activities
After the test is completed and the result generated in the lab, there are systems to deliver it in a timely fashion to the doctor or other health care provider.

• Report sent to correct place – the report needs to be sent to the provider who requested the test and any other health care professional who needs to receive a copy of the report.
• Timely reporting of data – there is an expected turn-around-time (TAT) for test results, with more critical results and results for monitoring of acute events needed more promptly than routine tests. Labs have a hierarchy of tests that are critical and special codes that can be used to assist in meeting medically necessary TAT.
• Reference intervals included - lab reports do not simply report a test result without providing a framework for interpreting the result – the “reference range.” Some lab computer systems can also suggest diagnoses, follow-up testing, and other insights that may help the provider interpret the result in the context of a particular patient.
• Immediate notification of results exceeding “critical limits” - each laboratory has in place a system for immediate reporting to the requesting doctor any finding that reflects a critical or life-threatening condition.

Most laboratory results are collated and managed by a sophisticated computer system capable of sending electronic reports to the health care provider by direct transfer to the GP’s computer or hospital information system. These computers can alert laboratory staff of an unusual finding, such as a critical value, and can alert the staff to call the physician. Laboratory reports generated by information systems can also highlight values outside the expected or reference range to help the physician focus on the tests that are of most concern.

Evaluating Test Results in a Clinical Context – why do tests sometimes need to be repeated?

The doctor is expected to evaluate all the relevant findings – laboratory test data plus information from other sources, such as physical exam, personal and family histories, signs and symptoms, other diagnostic examinations, i.e., x-rays, ECG, etc. – before settling on a diagnosis and developing a treatment plan. Given the complexities in human physiology and disease response, no diagnosis should ever be made solely on the basis of a single lab test. The clinician must always ask, “Do the test data fit with the other pieces of the puzzle?” Careful evaluation and consideration of test findings increases the reliability of a diagnosis and reduces the chance of medical errors.

As this diagram shows, data from medical tests are part of the information set that needs to be considered when a doctor makes a diagnosis. When a laboratory report indicates abnormal or unexpected results, it is necessary for your health care provider to further evaluate and corroborate the information at hand to ensure an accurate diagnosis. If the data do not correspond with the clinical picture, additional information may be needed and retesting may be appropriate to eliminate transient or spurious results.

Is there anything I can do?

Yes. Sometimes there are things you do or don’t do that can affect your tests results. Following a couple of guidelines can help to ensure that your results are interpreted correctly by your doctor:

• Your doctor or nurse should discuss with you how to prepare for a test to avoid known interferences. You may be instructed to fast or avoid certain foods or activities. Carefully follow these instructions to prepare for the test.

• Since some test results can be affected by medications, vitamins and other over-the-counter health supplements it is important that you provide a complete and honest medical history so that your doctor can correctly interpret the results from the lab.

Conclusion

Laboratory testing is an exact science conducted by trained professionals, with rigorous statistical analysis, quality controls, and extensive oversight. Medical testing is an important component in the diagnostic tool kit. However, laboratory testing is most reliable when used in conjunction with other meaningful data collected in the diagnostic process and when appropriate questions are asked and answered.

You can put great trust in the data generated by today’s exacting medical tests. A caveat to remember is that a diagnosis and treatment plan for a serious disease should never be based on a single medical test. Test findings are just one of the pieces of the diagnostic puzzle. Help your health care provider with the interpretation of the data by providing thorough information on your medical history and keeping track of symptoms you wish to report. To reduce the chance that you are improperly diagnosed and given potentially harmful and unnecessary medical procedures, first follow test preparation instructions carefully and tell your doctor and the person collecting your sample about any variation to these instructions.

If you and your doctor are surprised by a test result, here are some questions you could discuss together.

1. Does the result fit with my symptoms and current state of health?

2. How often do false negative or false positive occur with this test?

3. Is the course of action being considered serious enough that a test should be repeated?

4. Is their any reason to doubt the quality of the test result?

FURTHER READING

Making Sense of Testing: A guide to why scans and health tests for well people aren’t always a good idea. London: Sense About Science, 2008.

Fraser CG (1986) ‘Interpretation of Clinical Chemistry Laboratory Data’ Oxford: Blackwell Scientific Publications.