![]()
1
Requesting and interpreting tests
Learning objectives
To understand:
how sample handling, analytical and biological factors can affect test results in health and disease and how these relate to the concept of a test reference range;
the concepts of accuracy, precision, test sensitivity, test specificity in the quantitative assessment of test performance.
Introduction
Biochemical tests are crucial to modern medicine. Most biochemical tests are carried out on blood using plasma or serum, but urine, cerebrospinal fluid (CSF), faeces, kidney stones, pleural fluid, etc. are sometimes required. Plasma is obtained by collecting blood into an anticoagulant and separating the fluid, plasma phase from the blood cells by centrifugation. Serum is the corresponding fluid phase when blood is allowed to clot. For many (but not all) biochemical tests on blood, it makes little difference whether plasma or serum is used.
There are many hundreds of tests available in clinical biochemistry but a core of common tests makes up the majority of tests requested in clinical biochemistry. These core tests are typically available over a 24 h period. Tests are sometimes brought together in profiles, especially when a group of tests provides better understanding of a problem than a single test (e.g. the liver function test profile). Many of the other more specialist tests are restricted to larger laboratories or specialist centres offering a national or regional service.
In dealing with the large number of routine test requests, the modern clinical biochemistry laboratory depends heavily on automated instrumentation linked to a laboratory computing system. Test results are assigned to electronic patient files that allow maintenance of a cumulative patient record. Increasingly, test requests can be electronically booked at the ward, clinic or in General Practice via a terminal linked to the main laboratory computer. Equally, the test results can be displayed on computer screens at distant locations, even negating the need for issuing printed reports.
In this first chapter, we set out some of the principles of requesting tests and of the interpretation of results. The effects of analytical errors and of physiological factors, as well as of disease, on test results are stressed. Biochemical testing in differential diagnosis and in screening is discussed.
Collection of specimens
Test requests require unambiguous identification of the patient (patient's name, sex, date of birth and, increasingly, a unique patient identification number), together with the location, the name of the requesting doctor and the date and time of sampling. Each test request must specify the analyses requested and provide details of the nature of the specimen itself and relevant clinical diagnostic information. This may be through a traditional request form and labelled specimen or be provided electronically in which case only the sample itself need be sent to the laboratory with its own unique identifier (typically a bar code which links it to the electronic request).
Because of the large number of samples that are processed by most clinical biochemistry laboratories, every step needs to be taken to avoid errors. Regrettably, errors do rarely occur and these can be divided according to the error source:
- Pre-analytical. These arise prior to the actual test measurement and can happen at the clinical or laboratory end. Most errors fall into this category (see Table 1.1).
- Analytical. Laboratory based analytical errors are rare but may occur e.g. reagent contamination, pipetting errors related to small sample volumes, computing errors.
- Post-analytical. These are increasingly rare because of electronic download of results from the analyser but include, for example, transcription errors when entering results from another laboratory into the computer manually; results misheard when these are telephoned to the clinician.
Table 1.1 Some more common causes of pre-analytical errors arising from use of the laboratory.
| Crossover of addressograph labels between patients | This can lead to two patients each with the other's set of results. |
| Where the patient is assigned a completely wrong set of results, it is important to investigate the problem in case there is a second patient with a corresponding wrong set of results. |
| Timing error | There are many examples where timing is important but not considered. Sending in a blood sample too early after the administration of a drug can lead to misleadingly high values in therapeutic monitoring. Interpretation of some tests (e.g. cortisol) is critically dependent on the time of day when the blood was sampled. |
| Sample collection tube error | For some tests the nature of the collection tube is critical, which is why the Biochemistry Laboratory specifies this detail. For example, using a plasma tube with lithiumāheparin as the anti-coagulant invalidates this sample tube for measurement of a therapeutic lithium level! Electrophoresis requires a serum sample; otherwise, the fibrinogen interferes with the detection of any monoclonal bands. Topping up a biochemistry tube with a haematology (potassium ethylenediamine tetraacetic acid (EDTA) sample) will lead to high potassium and low calcium values in the biochemistry sample. |
| Sample taken from close to the site of an intravenous (IV) infusion | The blood sample will be diluted so that all the tests will be correspondingly low with the exception of those tests that might reflect the composition of the infusion fluid itself. For example, using normal saline as the infusing fluid would lead to a lowering of all test results, but with sodium and chloride results that are likely to be raised. |
On the scale of the requesting of biochemical tests, errors are fortunately rare. However, occasional blunders do arise and, if very unexpected results are obtained, it is incumbent on the requesting doctor to contact the laboratory immediately to look into the possibilit...
