Advances in medical laboratory technology have driven major changes in the practice of laboratory medicine over the past two decades by the development of automated, cross-disciplinary single platform analysers. This has led to the blurring of boundaries between traditional disciplines and the emergence of core automated or blood science laboratories. This paper was commissioned by the Union of European Medical Specialists to examine the changing role of laboratory-based physicians in the light of these advances by focusing on the added value of expert interpretation of test results and resultant improvements in clinical outcomes. The paper also considers the broad range of responsibilities of laboratory-based physicians and the difficulties in precisely measuring how this translates into improved clinical outcomes. Given its provenance, the paper concentrates predominantly on the role of laboratory-based physicians while acknowledging the essential and vital role of scientists in running diagnostic laboratory services.
- Quality Assurance
Statistics from Altmetric.com
A predictable and welcome consequence of advances in understanding of the scientific basis of medical practice in the 20th century was the development of physicians specialising in laboratory medicine (LM) or pathology. While the early pioneers were pluripotent in their approach, embracing the composite breadth of LM, the pace of advance in tissue-based diagnosis and its immediately verifiable impact on disease management meant that histopathology (also known as morbid anatomy or cellular pathology) was the first laboratory-based specialty to carve out a distinct identity. Subsequently, advances in diagnostics in other areas of LM and the need for particular expertise in distinct subject areas led to the emergence of laboratory-based physicians in haematology (including transfusion medicine), clinical biochemistry (chemical pathology), microbiology (including virology) and immunology.
Across Europe, the pace of specialisation has varied with the practice of LM in its broadest sense, with the continuation of general pathology or polyvalent LM in some countries alongside the single specialties listed above. The development of a Section of LM (Medical Biopathology) within the European Union of Medical Specialist (UEMS) in 1962 in order to promote and harmonise high standards of training and practice across LM in the constituent countries of the European Union and its subsequent evolution has followed a similar vein. Although at its inception the section of LM acted as an umbrella body for all laboratory-based specialties, the presence of a critical mass of practitioners and the distinctive nature of practice in histopathology and microbiology led to the creation of individual sections for these specialties in 1988 and 2009, respectively.
Over the past few decades advances in medical laboratory technology have significantly influenced and directed changing roles for physicians in many laboratory disciplines with the possible exception of histopathology. Despite advances in image analysis, the role of the histopathologist in the critical analysis of diseased tissue using a variety of conventional and molecular biological techniques, has to date, not been significantly overtaken by technology. By contrast, the automated analysis of blood and other fluids in clinical biochemistry, haematology, microbiology and immunology laboratories using a variety of techniques is not dependent on direct physician involvement. What added value then, does a laboratory-based physician bring to practice in these disciplines?
In the light of the organisational changes driven by these advances and attempts to contain healthcare costs, this paper was written at the invitation of the Section of LM of UEMS to examine the role of laboratory-based physicians in improving clinical outcomes for patients undergoing laboratory tests. By so doing, it does not diminish the vital role of laboratory scientists but highlights the complementary nature of their respective roles.
Influencing clinical outcomes
The integral importance of laboratory tests to the day-to-day practice of medicine is widely accepted. Although the assertion that test results influence approximately 70% of clinical decisions has recently been questioned,1 it is clear that a comprehensive LM service is essential to deliver high quality evidence-based care. The difference that such laboratory tests make to clinical outcomes is beyond dispute but quantifying the effect of testing remains a challenging task. Although the importance of laboratory-related outcomes is now well recognised,2 difficulties in designing rigorous studies of the effects of laboratory tests on clinical outcomes has resulted in a limited evidence-base in contrast to the plethora of randomised controlled trials underpinning therapeutic interventions. Even more challenging is the need to measure the added value of the contributions made by laboratory-based physicians (box 1). As we increasingly move towards patient-centred healthcare systems, it could be argued that the only justification for laboratory-based physicians would be to ensure that they make a clear difference to patient outcomes. In this regard, three levels of laboratory-related patient outcomes have been defined (box 2).3
Clinical quality indicators for LM have recently been defined using focus group consultation and subsequent ranking by an expert panel of primary and secondary care physicians.4 The top 10 indicators defined by this approach included the communication of critical results, education of users and quality assurance but crucially did not include the quality of interpretative comments which is the focus of this paper.
Responsibilities of a laboratory-based physician
Direction of clinical laboratories
Provision of appropriate test repertoire
Clinical liaison and interpretation of results
Attendance at multi-disciplinary meetings
Assay development and validation
Defining the utility of existing and emerging biomarkers for disease diagnosis/monitoring/prognosis, screening, risk profiling, treatment monitoring and use of targeted therapies
Education and training
Hands on laboratory work (in some disciplines)
Coordinating direct patient care (in some disciplines) with test requesting, interpretation and reporting
Laboratory-related patient outcomes3
Operational performance of a test—sensitivity, specificity
Predictive value of the test using Bayes’ theorem—probability of disease in a patient
Probability of the test result influencing a change in health status resulting from a change in disease management
Given the difficulties in precisely measuring the contribution of laboratory-based physicians to clinical outcomes, this paper highlights the vital contribution made to patient care by expert interpretation of laboratory tests as exemplified in the following case histories.
Integration of laboratory results in the clinical context—qualitative examples of laboratory-related patient outcomes
Case history 1
A peri-menopausal woman with type 2 diabetes was being investigated by her general practitioner (GP) for possible menopausal symptoms. The results of initial biochemical investigations are as follows: plasma luteinising hormone (LH) <0.1 IU/l (menopausal reference range >30), follicle stimulating hormone (FSH) 0.6 IU/l (menopausal ref range >30). The surprisingly low LH and FSH in a peri-menopausal woman prompted laboratory physicians to raise the possibility of pituitary disease. Further investigations instigated by the biochemistry laboratory revealed an elevated prolactin at 6.1 IU/l (reference range 0.09–0.52). The possibility of a macro-prolactin causing artefactual elevation of plasma prolactin was excluded by treatment with polyethylene glycol to precipitate complexed prolactin.
Reassessment of the clinical history following the detection of elevated prolactin revealed chronic galactorrhoea which had previously been thought to be insignificant. A cranial MRI scan revealed the presence of a large pituitary tumour extending into the suprasellar area measuring 23×20 mm. Her vision was fortunately unaffected and imaging revealed no compression of the optic chiasm.
Lessons from this case
Correct interpretation of the significance of an unusually suppressed plasma FSH in a peri-menopausal woman by a laboratory physician and instigation of additional tests revealed the underlying diagnosis of a pituitary tumour.
Case history 2
A previously well adult man was brought to the Emergency Department (ED) by his friends. He had taken part in a martial arts competition on a hot day. He had been using a commercial sugary solution to keep himself hydrated, and had decided not to accompany his friends to a dinner at the end of the competition. They returned to find him semiconscious and wondered whether he might have had a seizure. The tests done in the ED showed a plasma sodium of 116 mmol/l, plasma osmolality of 242 mOsm/kg (reference range 290–300), urine sodium of 50 mmol/l and urine osmolality of 425 mOsm/kg, suggestive of dilutional hyponatraemia associated with the syndrome of inappropriate anti-diuretic hormone secretion (SIADH). The rest of his metabolic profile, including plasma glucose and renal function was normal. During an attempted lumbar puncture, he had a seizure, during which he vomited and inhaled. He was admitted to the intensive care unit, and required antibiotic treatment for aspiration pneumonia.
In this case, hyponatraemia was associated with exercise and ingestion of glucose-containing fluid. It arises in individuals who activate anti-diuretic hormone (ADH) secretion on exercise but then do not switch it off when overhydrated. Hyponatraemia occurs with glucose containing fluids because as cells take up glucose under the influence of insulin and exercise, free water remains and causes dilutional hyponatraemia. ADH prevents the free water being excreted. Investigation includes measurement of plasma and urine sodium and osmolality, and reveals a picture characteristic of SIADH. Conservative management, with airway protection and fluid restriction with addition of an ADH-receptor antagonist if necessary, is usually sufficient.
Lessons from this case
The treatment of acute hyponatraemia is a medical emergency and a physician in the biochemistry laboratory was directly responsible for the diagnosis of exercise and exogenous glucose-driven SIADH, which had initially been overlooked. This diagnosis guided correct management and thus obviated the need for outpatient neurological assessment and further investigation of his seizures.
By guiding the treating clinicians in the differential diagnosis and prevention of acute hyponatraemia, this case also highlighted the key educational role played by laboratory-based physicians.
Case history 3
An adult woman with chronic renal failure underwent unilateral renal transplantation with her mother as a donor. Peri-operative thrombo-prophylaxis was undertaken with unfractionated heparin. However, following an uneventful operative procedure, multiple thromboses developed accompanied by a marked fall in platelet count from a pre-operative level of 220 000/mm3 to 80 000/mm3.
During a post-operative multi-disciplinary meeting involving different specialties, the laboratory haematologist raised the possibility of the diagnosis of heparin-induced thrombocytopenia (HIT). On his recommendation, immediate substitution with alternative anti-thrombotic treatment (lepirudin) was commenced while laboratory investigations for confirmation of HIT were performed.
The diagnosis of HIT was confirmed by the demonstration of antibodies to heparin-platelet factor 4 complexes by ELISA. The dosage of anticoagulant therapy and further laboratory monitoring was undertaken by medical staff of the haematology laboratory, who continued to supervise the patient's anti-coagulant therapy until the episode of HIT had resolved. Once the platelet count had normalised, the patient was commenced on oral warfarin.
Lessons from this case
The failure to diagnose HIT by the transplant team and consequent delay in stopping heparin are key learning points. Indeed, had the diagnosis of HIT not been made by the laboratory haematologist, it is likely that heparin would have been continued at a higher dose due to the mistaken assumption of inadequate anti-coagulation in the face of multiple thromboses. In addition to laboratory physicians being pivotal to the diagnosis of HIT, selection of alternative anti-coagulants and monitoring of treatment were also directed by the laboratory.
Case history 4
An adult woman from Greece with a long history of lower abdominal discomfort was admitted via the ED with severe abdominal pain. Her routine blood tests were normal apart from a raised C reactive protein and a mean corpuscular volume (MCV) of 69fL. A contrast enhanced CT scan showed a splanchic vein thrombosis.
There was no family history of thrombotic disease and a thrombophilia screen including genetic testing for Factor V Leiden and prothrombin was negative.
The patient was started on warfarin for 6 months and discharged. On review of her laboratory results including a low MCV with a normal haemoglobin, the laboratory haematologist raised the differential diagnosis of α/α thalassaemia or iron deficient polycythaemia rubra vera (PRV). She requested iron studies and a JAK2 V617F mutation analysis which came back as positive confirming the diagnosis of iron deficient PRV. The patient was referred for upper and lower gastrointestinal endoscopy. These showed a bleeding polyp which could be entirely removed. She was started on life-long aspirin and advised never to receive iron replacement.
Lessons from this case
Splanchnic vein thrombosis is a recognised complication of myeloproliferative diseases such as PRV. Iron-deficient PRV can be easily overlooked, as was the case here, as the only hint might be a low MCV in the presence of a normal haemoglobin.
In this case of a woman originating from the Mediterranean, the obvious explanation would have been a mild alpha thalassaemia trait. Without the input from the haematologist, the diagnosis of PRV would have been missed and the patient left untreated after completion of 6 months warfarin therapy.
Importantly, the patient might have been started on iron replacement, which can lead to a rapid increase in haemoglobin levels and precipitation of thrombotic events in PRV patients. The diagnosis of iron deficiency also led to further gastrointestinal investigations and the removal of the underlying cause, in this case a polyp.
Case history 5
A male infant was hospitalised with acute undefined bacterial pneumonia. He was discharged from the hospital 2 weeks later having apparently made a good recovery. His serum immunoglobulin (Ig) levels were reported to be satisfactory:
IgG 3.0 g/l (Ref range 3.0–9.0), IgA 0.1 g/l (0.15–0.7), IgM 0.3 g/l (0.4–1.6), serum electrophoresis—not done, full blood count—Hb 12.8 g/dL, white cell count (per mm3) 9.0, lymphocytes 2.0, neutrophils 5.9, platelets 256/mm3. Six weeks later, he was re-admitted to another hospital with a further episode of pneumonia proven to be due to Pneumocystis jiroveci on this occasion. His repeat Igs were as follows: IgG 2.7 g/l, IgA 0.1 g/l, IgM 0.1g/l, serum electrophoresis—small monoclonal IgG band, lymphocyte surface marker analysis revealed marked T and NK cell lymphopenia with normal numbers of B cells. Final diagnosis—T-B+NK- severe combined immunodeficiency (SCID).
Lessons from this case
The delayed diagnosis in this case was a consequence of the failure to synthesise and correctly interpret data at first presentation. The significance of lymphopenia as a flag for SCID was overlooked and the lack of serum electrophoresis led to the report of an apparently ‘normal’ Ig profile. Monoclonal bands are extremely rare in infancy, which frequently signify underlying immunodeficiency.5 This finding prompted the laboratory physician to raise the strong possibility of SCID. It is arguable that the significance of lymphopenia in a 4 month old baby with pneumonia and a borderline serum IgG that was likely to be maternal in origin was ignored because of the absence of clinical immunology expertise during his first presentation.
Case history 6
A previously well adult woman presented to her GP with a 6 month history of generalised arthralgia. The results of initial investigations were as follows: rheumatoid factor (RF) 160 IU/l)(reference range <40), anti-nuclear antibody 1/160, C reactive protein <6 mg/l, full blood count normal, Na 142 mmol/l, K 3.8 mmol/l, creatinine 120 µmol/l (reference range 50–145). On the strength of these results, a tentative diagnosis of rheumatoid arthritis (RA) was made by the GP. On rheumatological assessment, she was noted to have a purpuric rash on her legs but no clinical features to suggest RA.
The results of further investigations were as follows: antibodies to double-stranded DNA and extractable nuclear antigens—negative, RF 225 IU/l (reference range <40 units), serum complement C3 1.2 g/l (reference range 0.6–1.8), C4 0.02 g/l (reference range 0.15–0.4), serum IgG 8.4 g/l (reference range 6–13), IgA 1.0 g/l (reference range 0.8–2.5), IgM 4.5 g/l (reference range 0.4–2.0). Electrophoresis—Normal. The combination of a strongly positive RF and a low C4 led to the possibility of hepatitis C—associated mixed cryoglobulinaemia being raised in the laboratory report. Further investigations confirmed the presence of a mixed cryoglobulin in serum accompanied by evidence of glomerular involvement (positive red cell casts in urine), leading to the unifying diagnosis of hepatitis C- associated mixed cryoglobulinaemic vasculitis.
Lessons from this case
The correct diagnosis of mixed cryoglobulinaemic vasculitis associated with hepatitis C infection was a direct result of proactive clinical interpretative comments by the laboratory-based clinical immunologist.
Case history 7
Matthews et al6 reported the case of a 73-year-old man with a history of progressive right-sided facial and periorbital swelling, right-sided nasal blockage, serous nasal discharge and visual blurring. The patient was immunocompromised following treatment for prostatic carcinoma, and at the time of presentation had a disseminated vesicular rash. Based on clinical and radiological findings, a diagnosis of necrotising maxillary and ethmoid sinusitis with periorbital cellulitis and conjunctivitis was made, together with disseminated varicella-zoster virus infection. After surgical debridement, a microbiological diagnosis of acute necrotising sinusitis caused by Staphylococcus lugdunensis was made. No other organisms, including fungi, were grown from the operative samples, and histology was negative for fungal elements. The patient made a good recovery after 6 weeks of appropriate therapy.
S lugdunensis is a coagulase-negative Staphylococcus that is a normal commensal of human skin. Coagulase negative staphylococci are commonly isolated from clinical samples, and are usually disregarded as being non-pathogenic or contaminants unless found in association with a prosthetic device or endocarditis, and repeatedly isolated. S lugdunensis is now well recognised7 as a significant pathogen that can cause invasive disease similar to S aureus, and can also (as in this case) initially be mis-identified as a S aureus. Clinically it is usually associated with infections such as endocarditis following interventions in the groin for example, vasectomy or cardiac catheterisation via the femoral artery.
Lessons from this case
In the laboratory it is not cost-effective to identify all coagulase-negative staphylococci to species level, especially from non-blood culture isolates. The role of the laboratory-based physician in this case was to instigate appropriate further investigations to correctly identify S lugdunensis, liaise with the clinical team to clarify its role as a significant pathogen in the clinical context described and ensure that appropriate treatment was instituted.
Case history 8
A young man presented to his GP with a 3 month history of unilateral cervical lymphadenopathy, general malaise, night sweats and weight loss. On examination, the GP was concerned to find mild hepatosplenomegaly, and given the duration of the history performed some routine investigations, including Epstein-Barr virus (EBV) and cytomegalovirus (CMV) serology. He planned to refer him for urgent haematological assessment, concerned about possible lymphoma. The EBV and CMV serology results were not consistent with recent infection. Given the clinical history, the laboratory physician arranged additional tests for Toxoplasma; the results were consistent with recent Toxoplasma infection (Toxoplasma IgM and IgG strongly reactive). The patient went on to make a full recovery over the next 3 months with no further intervention.
Lessons from this case
In this case recognition of the possibility by a laboratory-based physician and subsequent confirmation of Toxoplasma infection led to the avoidance of outpatient assessment and lymph node biopsy.
Toxoplasma infection is estimated to cause 3–7% of clinically significant lymphadenopathy8 and is commonly confirmed by serology following lymph node biopsy with suggestive histology in such cases.
As the above case histories illustrate, the interpretation of laboratory results in the clinical context by an expert laboratory-based physician makes a significant difference to clinical outcomes for individual patients.9 Laboratory-based physicians with discipline-specific specialist expertise are well placed to provide expert clinical interpretation by virtue of comprehensive training in both clinical and laboratory aspects of diseases in their respective specialties. In some countries of the EU such as the UK and Germany this extends to laboratory-based physicians undertaking training in general internal medicine followed by specialty training in the relevant LM specialty (Biochemistry, Haematology, Immunology, Microbiology) thus providing the requisite competencies for combined laboratory and clinical practice at consultant or senior practitioner level.
The pressure for expert advice on test selection and interpretation of results is likely to grow as the complexity of clinical practice grows and as primary care physicians are called upon to provide care to patients with increasingly complex problems.10 Additionally, the drive to reduce medical errors (from misinterpretation of test results) and contain costs are powerful reminders of the need for informed clinical interpretation of test results leading to effective decision-making. Inappropriate test selection leading to further testing can not only exact a severe financial cost but can also have unintended adverse consequences in other areas for example, exposure to unnecessary irradiation due to inappropriate imaging, as exemplified by the uncritical use of tumour markers.11 Equally, inappropriate clinical decisions may be made in patients as a result of erroneous test results generated through interference by heterophilic antibodies in many routine immunoassays.12–14 An example of the serious adverse consequences of misinterpretation of spuriously elevated serum human chorionic gonadotrophin levels is the needless treatment of young women for ‘occult’ trophoblastic disease with chemotherapy and surgery.15 Conversely, false-negative results may also have adverse clinical consequences due to inappropriate or delayed treatment as exemplified by erroneous thyroglobulin measurements due to interference by anti-thyroglobulin antibodies in patients with thyroid carcinoma.16 In the current climate of economic austerity, it is essential that hospitals and commissioners of LM services guard against the temptation that a results-only laboratory service would be adequate, based on the assumption that the requesting clinician is fully capable of interpreting any laboratory test. Although routine blood results such as blood counts or renal and hepatic function which fall into the reference range may require little or no interpretation, the fallacy that this also applies in more complex situations and the clinical risks surrounding such a blanket proposition are amply illustrated by the afore-mentioned case histories.
Audit designed to improve clinical outcomes is an essential function of a diagnostic laboratory. Physicians play a key role in designing and leading clinical audit projects to ensure appropriate test selection, minimise unnecessary testing17 and validate gating policies for test utilisation. These initiatives form the basis for evidence-based demand management and enable clinical audit to be a powerful tool in improving patient management. For example, the use of a gating policy to ensure testing for anti-neutrophil cytoplasmic antibodies (ANCA) was restricted to patients with a high pre-test probability of ANCA-associated vasculitis18 (AAV) minimises the occurrence of false-positive ANCA and consequently, reduces the risk of misdiagnosis of AAV and instigation of inappropriate immunosuppressive treatment.
The concept of multi-disciplinary team meetings to devise optimal treatment and management plans for individual patients was originally devised for patients with solid organ cancer. Laboratory physicians (in this instance, histopathologists) play a vital role not only in making an accurate diagnosis, but also in molecular typing to enable selection of targeted treatment, as exemplified by the use of Trastuzumab to treat breast cancers over-expressing the epidermal growth factor receptor. Similarly, in haematological malignancies, haematologists have integrated specialist knowledge of cancer with laboratory expertise in developing molecular diagnosis to enable targeted treatment with other therapeutic monoclonal antibodies and small molecule tyrosine kinase inhibitors.
The success of current initiatives on stratified medicine19 to tailor treatment of cancers according to the molecular genetic signature of individual tumours will also be crucially dependent on close physician engagement from both laboratory and clinical ends. In the future, it is possible that conventional histopathological analysis may be replaced by gene expression profiling for some specimens. However, for the foreseeable future molecular analyses are likely to complement rather than replace histopathological interpretation of tissue biopsies. In either scenario, medical involvement will be critical to the interpretation and integration of results into clinical management plans.
In focusing on the role of physicians in the laboratory it is important not to overlook the vital role played by non-medical scientists in LM. Although traditionally many scientists have concentrated on the detailed technical and operational aspects of LM, many have also have successfully undertaken leadership roles within LM, including the directorship of diagnostic laboratories, thus precluding the absolute need for a medical degree in fulfilling most of the responsibilities outlined in box 1. However, it could be argued that a physician with a medical degree complemented by a solid grounding in general internal medicine and further sub-specialty training is better equipped to integrate and interpret laboratory results in the context of complex or unusual clinical case presentations and play an active role in patient management. Thus, while acknowledging the particular skills of scientists and physicians, it is important to emphasise the complementary roles fulfilled by these individuals in running clinically responsive diagnostic laboratory services.
Given the difficulties in precisely measuring the added value of expert interpretation of test results by laboratory physicians, it needs to be acknowledged that the clinical value of such contributions remains persuasive but anecdotal. It may well be that a specialist physician will produce the same level of nuanced interpretation in a particular specialty as a laboratory-based physician. However, the reduction in exposure to LM during medical training makes it unlikely that an individual physician will be fully conversant with the breadth of expertise provided by a laboratory-based physician as detailed in box 1. Equally, speedy pro-active interpretation of test results in the laboratory is likely to minimise diagnostic delay.
The past two decades have seen traditional boundaries between individual LM disciplines being blurred by the ability to assay on a single platform an ever-increasing range of analytes, previously considered to be discipline-specific in biochemistry, haematology, immunology and microbiology. This has led to the emergence of blood science laboratories (core automated laboratories) enabling rapid-throughput of large numbers of samples. While welcoming such technological advances it is essential not to overlook the importance of maintaining systems of expert interpretation of test results emanating from these centralised facilities. Irrespective of the wide variations in the practice of LM across Europe, it is our hope that this paper has highlighted the crucial importance of continuing active physician involvement in clinical diagnostic laboratories, particularly given the diverse and competing demands which are increasingly being placed on this group of professionals.
This paper was written at the invitation of and ratified by the following members of the Board of Medical Biopathology-Laboratory Medicine of the Union of European Medical Specialists (UEMS): Altankova I (Bulgaria), Faure G (France), Frayling I (UK) Kokkinou V (Greece), YV Leftherioti (Greece), Machado A (Portugal), Malenica B (Croatia), Marcelis L (Belgium), Misbah SA (UK), Montserrat G (Spain), Norlund L (Sweden),Nunes AR (Portugal) Oosterhuis W (Netherlands), Twomey P (UK), Zlabinger G (Austria). We thank Dr Richard Herriot, Dr Jonathan Kay and Dr Catherine Swales for their critical comments.
Contributors The idea for the paper was conceived by SAM. SAM wrote the paper, including the Immunology case histories. VK, TT and AS contributed haematology case histories, KJ contributed microbiology case histories, BS and WO contributed biochemistry case histories.
Competing interests All authors have a keen interest in the role of physicians in laboratory medicine.
Patient consent Details have been removed from these case descriptions to ensure anonymity. The editors and reviewers have seen the detailed information available and are satisfied that the information backs up the case the authors are making.
Provenance and peer review Not commissioned; externally peer reviewed.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.