Laboratories in clinical medicine: Ordering tests, interpreting results
Times have changed. There were days when we visited Hakeem-Sahib, who on first look made a diagnosis, prescribed detailed diet advice, and ordered a decoction of an herb for a drink. In later days Hakeem-sahib learned to feel our pulse to tell the whole story.
Doctors developed the protocol of history taking & physical examination to make the most likely diagnosis. With the introduction and expansion of laboratories and imaging tools, Western medicine has become more and more exact and evidence-based and history & physical has gone into the background.
However, it is a dictum that clinical medicine continues to be an art and the best test in making the diagnosis of a patient in modern medicine starts with history taking and physical examination.
So, time spent with the patient and the family during consults is essential in clinical medicine and cannot be replaced by any lab, machine, or other tools.
Of course, laboratory tests and imaging tools are needed and without these, clinical medicine is akin to an army fighting a war without arms. This write-up will address the issue of analytical/biochemical lab tests and will not include the role of pathology and microbiology in clinical medicine wherein the results are documented as reports by the pathologists and microbiologists respectively.
Scope of Laboratory Medicine
Laboratory medicine is expanding and spends around 5% of the health care budget in the West; influences around 50% of patient care decisions and invokes around 25% of malpractice suits as against 15% against prescription errors.
Laboratory tests help clinical medicine in several ways which include: (i) Case definition: In many situations, the abnormal test is the way to name a disease. For example, HBsAg positive and HIV positive defines the disease as Hepatitis B and AIDS respectively. (ii) Diagnosis: Lab data are essential in making a diagnosis and in most situations are a surrogate marker to distinguish between disease free from disease groups. (iii) Disease severity: Here lab results and their level help to differentiate between mild, moderate, and severe disease and influence treatment policies, prognosis, morbidity, and mortality. (iv) Response: Serial lab results on a patient with disease help to evaluate response to treatment, determine the course of the disease, and will affect follow-up policies. (v) Disease prevalence: Prevalence refers to disease occurrence (percentage) in the population. For example, HBV is 2.5% in India. (vi) Screening: Look for disease in a high-risk group. Examples diabetes in obese, and anemia in pregnancy. In addition, lab tests have a major role to play in jurisprudence (for example DNA samples of the culprit can be obtained from the scene of the crime itself.) and as a research tool.
Ordering Lab tests
Ordering a lab test is the sole responsibility of the physician and needs considerable knowledge and experience. It has been proven beyond doubt that correctly ordering lab tests has a significant advantage in clinical practice as it helps to make diagnosis faster and with better precision, shortens the hospital stay, reduces the number of clinical visits, and improves patient satisfaction.
Unfortunately, up to 50% of lab tests in a healthcare setting are ordered inappropriately which has adverse effects on patient care and hospital practice.
Another phenomenon (called Ulysses syndrome) is ordering an unnecessary test which leads to a long sequence of further unnecessary tests as some test results may show mild abnormality of no significance or may represent normal variation.
Because of this many patients and their families get exhausted and lose confidence in the system and have to suffer due to delayed diagnosis.
There are several distinct patterns of ordering a lab test which include: (i) Individual lab tests: Here physician orders an individual test for a specific problem like hemoglobin for anemia, bilirubin for liver disease, and creatinine for kidney disease.
This type of ordering lab tests is inappropriate and rarely helps the physician to make a diagnosis and unless necessary should be discouraged. (ii) Test module namely complete blood count (20 parameters) for evaluating anemia, liver function tests (11 parameters) to evaluate liver disease, and lipid profile (8 parameters) to evaluate dyslipidemia. This type of ordering a test is appropriate if we need a specific answer to a disease state.
(iii) Test panel: namely a set of many tests which help to give insight into the diagnosis and etiology of a disease like HBsAg, anti-HBc IgM, anti-HCV, IgM anti-HAV, and IgM anti-HEV for a patient presenting with acute hepatitis syndrome. In tertiary care setting most of the time, we need to order a test panel as diseases are complex and have varied etiologies. (iv) Algorithm: While investigating complex illnesses which need a battery of tests it is advisable to use lab tests in an algorithm wherein test panels are ordered sequentially (in order of importance and preference) after evaluating the previous panel ordered.
This becomes cost-effective as it reduces the number of tests, though it may take a longer time to reach a diagnosis. (v) Health check-up panel: a set of routine tests like CBC, LFT, KFT, glucose, lipids, HBA1C & TSH at discounted rates to promote business. This of course helps the laboratory as well as the public who get tests done at cheaper rates.
Interpreting test results
Interpreting the test results is the duty and responsibility of the ordering physician. Of course, the laboratory can help the clinician in understanding the basis of the test, the methodology used, normal values and range, and guidelines to interpret the test in the clinical setting.
Test results generally are reported in three patterns: (i) Dichotomous test results. Here the test is reported as either positive or negative. (ii) Categorical test results: here test results can fit into categories like mild, moderate, or severe. (iii) Continuous test results: here data are given in a continuous scale and normal values are given as mean/median/mode with a range. Interpreting test results needs a knowledge of data expression and analysis (sensitivity & specificity of test results) and the measure of diagnostic test accuracy (true positive rate, true negative rate, false positive rate & false negative rate).
Finally, it is the task of the physician to determine the presence or absence of disease and the question he has to answer: (i) what is the probability of disease when the test is positive, and (ii) what is the probability of no disease when the test is negative? The measure of disease probability as assessed by test results is based on the prevalence of disease in the population and interpreted on that basis.
Lab test errors
Many of us do not believe our lab results especially when they are abnormal and rush to the next lab to check on the first results. Any variation of results (usually better) infuriates the patient and his family and becomes a source of big difficulty for the original lab. What is the truth about lab test errors and the reproducibility of lab results between laboratories?
There are three phases in which errors can occur in lab testing which includes the pre-analytical phase (physician’s inappropriate ordering of test, poor patient preparation, faulty specimen collection and sample handling by phlebotomist, wrong method of storage and transport of specimens, and faulty processing before analysis), analytical phase (inaccurate test results from the analytical machine due to equipment malfunction or sample mixing in the test plates), and, post-analytical phase (faulty transcription of results, poor method of transmission of the report to client/physician, and physicians inability to interpret test results correctly).
Around 70% of lab test errors occur in the pre-analytical phase, 20% in the post-analytical phase, and only around 10% in the analytic phase (Fig). Thus, for a lab test error, any of the several steps in lab testing can go wrong and the fault may lie at the level of the patient preparation, ordering physician, phlebotomist, supportive staff, and less often with lab technicians, and the machine!
How to suspect a lab test error?
How can we suspect a laboratory test error? These include the following conditions (i) lab values that are very high or very low; (ii) results that vary significantly from previous values; (iii) values that do not correlate with the patient’s clinical condition; (iv) values on samples which are not well kept especially hemolyzed samples or those done after a prolonged period or transport or storage. It is the duty of the clinician (primary physician) to be on guard about the possibility of laboratory test error and take necessary action.
How to report a lab test error?
What action should be taken if a laboratory test error is suspected? A formal request named “confirmation Request for lab test” should be forwarded from the primary physician identifying the patient’s name, unique identification number, and test result.
The laboratory must take the following action: (i) Check original data to detect transcription errors; (ii) Repeat study on the original sample (if available) to detect errors in original determination; (iii) Repeat determination on a new sample on the patient to detect a possible error in patient or specimen identification.
If the test results are found correct, the results of the confirmation are reported to the clinician by the lab. The clinician must review the clinical findings, and drug history is reviewed for possible unsuspected disease or complication and possible drug interferences.
However, if a lab error is detected, erroneous lab test results (whether clerical or analytical errors) which have been entered into the patient’s records need to be replaced with the corrected results.
Reporting alarm test results
All laboratories must report very high or very low laboratory results to the primary physician immediately on phone. The cut-off values of the “Alarm Results” varies from one laboratory to another and from one test to another. Such results are immediately rechecked and if reproduced patient management started accordingly.
This practice saves many lives. One common alarm test result is serum potassium level (normal 3.6 to 5.2 mmol/L). If the lab reports serum potassium above 6.5 mmol/L (alarm test result), it can lead to cardiac arrhythmia and death. So, the test result must be rechecked and the physician informed immediately to start treatment against hyperkalemia (high potassium).
This is a method by which laboratories can establish that the results given are accurate. This is of 2 types. Internal quality control defines that standard procedure is followed as per manufacturer recommendations and needs to be done each morning before starting tests.
External quality control is a process in which a sample is blindly provided to the laboratory for estimation and the validity of results analyzed by an external source. All laboratories must follow both internal and external quality control procedures. One of the major sources of laboratory errors is due to inability to strictly follow quality control procedures.
(The author is MD, DM, FRCP (Edin), FACP, Master American College of Physicians (MACP, Emeritus), Director, Digestive Diseases Centre, Dr.Khuroo’s Medical Clinic, Srinagar, Kashmir, India, Former Director, Professor and Head Gastroenterology, Chairman Dept. Medicine, Sher-I-Kashmir Institute of Medical Sciences, Soura, Srinagar, Kashmir, India. E-mail: email@example.com, firstname.lastname@example.org, Website: www.drkhuroo.com.)
DISCLAIMER: The views and opinions expressed in this article are the personal opinions of the author.
The facts, analysis, assumptions and perspective appearing in the article do not reflect the views of GK.