Interpretative Commenting

Summary

• Clinical laboratories should be able to offer interpretation of the results they produce.

• At a minimum, contact details for interpretative advice should be available on laboratory reports.

• Interpretative comments may be verbal or written and printed.

• Printed comments on reports should be offered judiciously, only where they would add value; no comment preferred to inappropriate or dangerous comment.

• Interpretation should be based on locally agreed or nationally recognised clinical guidelines where available.

• Standard tied comments (“canned” comments) can have some limited use.

• Individualised narrative comments may be particularly useful in the case of tests that are new, complex or unfamiliar to the requesting clinicians and where clinical details are available.

• Interpretative commenting should only be provided by appropriately trained and credentialed personnel.

• Audit of comments and continued professional development of personnel providing them are important for quality assurance.

Introduction

Medical laboratories offer diagnostic services that are essential to patient care. In addition to analysis of patient samples, validation and reporting of those results, a diagnostic service may offer advice on appropriate test selection, interpretation as well as advice on further appropriate investigation.^1–3 In fact the provision of interpretative comments on reports is an essential role of the diagnostic laboratory service and should be guided by the needs and requirements of users which should be established in consultation with them.⁴ As a minimum, reports should carry contact details of laboratory personnel who are available to discuss the results and offer interpretative advice.⁴

The typical Clinical Biochemistry laboratory today is highly automated and produces a large amount of numerical data and a much smaller volume of qualitative results. The ability of the laboratory to give meaning to these data by interpreting the results is limited by a number of factors including;

1. Lack of knowledge of the context in which the tests were requested, i.e. the clinical question.

2. Lack of knowledge of patient factors which may influence the results e.g. medication.

3. The sheer volume of tests (that may to some extent be amenable to expert systems which are dealt with elsewhere in this Supplement).

The provision of interpretative comments on reports varies widely between labs. It may be determined by the management philosophy which ranges from the view that a request to perform a laboratory test is an implied request for specialist opinion to the view that clinicians who order tests only want the raw data and know how to interpret the results themselves. In fact they may resent as an intrusion into their territory any interpretative comments by laboratory personnel. In reality most laboratories offer some form of interpretative comments on a fraction of the Biochemistry reports they produce. The type of interpretative comment attached to the reports may vary in depth and degree of individualisation to the particular result depending on the complexity of the test, the requesting doctor, the requester’s perceived ability to interpret the test results, and the expertise available within the laboratory.

Guidelines for commenting

The RCPath (UK) ‘guidelines for the provision of interpretative comments on biochemical reports’⁵ state that whether a comment is required will depend on:

• the clinical details provided

• the clinical implication of the results

• the likely familiarity of the requesting clinician with the tests and their interpretation.

The guidelines suggest that comments might be appropriate when:

• - a decision on management or treatment is indicated by the results in combination with the clinical details provided

• - a result is unexpected

• - a specific question has been posed but it is not obvious whether the results provide the answer

• - a clinician has requested a test with which they are not likely to be familiar.^5,6

Brevity of comments is vital as clinicians are busy, they see large number of reports and will be put off from reading longwinded comments. At the same time the message should be clear and unambiguous.⁴

Given below is a spectrum of comments that may be used by the laboratory.

Analytical quality and pre-analytical interference are dealt with elsewhere in this Supplement. Suffice to mention here that personnel commenting on results should always consider pre-analytical and analytical artefacts which may account for the results; comments specifically related to such artefacts are beyond the scope of this review.

However, analytical limitations or interferences inherent in an assay, which may affect the interpretation of a result need to be highlighted in the report. These can be attached automatically to all results produced using the particular assay (Table 1).

Table 1.

Example 1.

Request for thyroglobulin measurement

Since anti-thyroglobulin antibody can interfere with thyroglobulin measurement, a thyroglobulin request would need to generate an anti-thyroglobulin antibody measurement and, if positive, the following standard comment is attached to the report.

Comment: The immunometric assay used for thyroglobulin may underestimate serum thyroglobulin concentrations in the presence of thyroglobulin antibodies.

Standard tied comments regarding diagnostic cut-offs and decision levels based on recognised guidelines, while not considered interpretative comments, are used by most laboratories, and may help recipients to be reminded or informed of such agreed cut-offs. These can be automatically generated every time such a result is reported (Table 2).

Table 2.

Examples 2 and 3.

Pregnancy test (hCG) in serum

Comment: A pregnancy test is considered positive if hCG >25 U/L. HCG values between 5 and 25 U/L are equivocal and should be repeated. In the first 8–10 weeks of a normal pregnancy hCG doubles every 3 days.

Troponin request for acute chest pain

Comment: Troponin I ≥0.40 μg/L in association with ischaemic symptoms or ECG changes is compatible with MI. Troponin between 0.10 and 0.40 may suggest increased risk in the setting of Acute Coronary Syndrome.

Taken one step further, the comment can relate the result to the cut-off, and the personnel commenting can choose the appropriate comment from a list of options (as has been done in Table 3). These sorts of comment can also be automatically generated by computer.

Table 3.

Examples 4 and 5.

Patient: 56 year old man
75 gram glucose tolerance test.
0 minutes glucose	5.7 mmol/L
60 minute glucose	12.2 mmol/L
120 minute glucose	6.6 mmol/L
Comment: Normal glucose tolerance test. Suggest repeat fasting glucose in 3 years time.
or
0 minutes glucose	6.7 mmol/L
60 minute glucose	12.2 mmol/L
120 minute glucose	6.6 mmol/L
Comment: Impaired fasting glycaemia. Suggest re-test in a years time. This patient is at increased risk of developing diabetes and cardiovascular disease. Lipid testing and lifestyle advice may be appropriate.

The examples in Table 3 also highlight two issues.

1. Interpreting dynamic test results such as the oral glucose tolerance test (GTT), the Synacthen stimulation test and especially more complex tests such as the water deprivation test which have a set of results rather than a single result may not be straightforward. The laboratory should consider commenting on these tests routinely.

2. The American Diabetes Association guidelines use a cut-off of 5.5 mmol/L for impaired fasting glucose whereas the Australian Diabetes Society guidelines use 6.0 mmol/L; when there is a discrepancy between guidelines national guidelines generally take precedence over overseas ones.

Clinical context is important in commenting and patient details need to be available and, if not, sought before commenting. Thus, in the above case, advice on additional testing and management would be useful to add to the basic interpretation of the GTT result.

The individualised narrative interpretative comment epitomises interpretative commenting. Components of a good comment may include the following:

1. The absence or presence of an abnormality and its degree or severity.

2. Possible clinical implications of abnormality and/or a diagnosis.

3. Suggested follow-up including further testing and specialist referral.⁷

Individualising comments is often difficult with little or no clinical details typically available on the request form. Commenting in such instances should be restricted to what can be offered confidently based on the available information, or such information should be sought before the comment is generated. The Electronic Patient Record with data linkage will help in accessing patient information. Table 4 shows an example of how the same result would attract different comments depending on the clinical context. Inappropriate comments can be dangerous if acted on although, anecdotally, they would often be just dismissed as such and lead to loss of the laboratory’s credibility. Release of results (if necessary as an interim report) should not be delayed while they await comments to be added except in the case of very complex Biochemistry reports which cannot be interpreted without input from the laboratory.

Table 4.

Example 6.

Thyroid Function Tests
Free T4	14 pmol/L	(10–23)
TSH	5.90 mU/L	(0.40–4.00)
The comment, depending on the context, could be:
Mild elevation of TSH may be found in patients with subclinical hypothyroidism or non-thyroidal illness. Thyroid antibody testing may be useful in subclinical hypothyroidism.
or
Elevated TSH suggests inadequate thyroid hormone replacement if the dose has not been changed for at least 6 weeks and the patient has been taking the medication regularly. Suggest review of thyroid hormone treatment and repeat TFTs in 2–3 months.

A typical example from the Royal College of Pathologists of Australasia (RCPA) Quality Assurance Programs (QAP) Case Report Comments Program, an educational self assessment program for interpretative commenting, is given in example 7 shown in Table 5.⁸

Table 5.

Example 7.

Patient: 51 year old male.
Patient Location: Presenting to Emergency Department.
Clinical Notes on Request Form: Acute confusional state.
On risperidone.
Case Details
Plasma
Sodium	108 mmol/L	(136–146)
Potassium	4.1 mmol/L	(3.5–5.5)
Creatinine	80 umol/L	(50–110)
Urea	2.0 mmol/L	(2.7–7.8)
Osmolality	227 mmol/kg	(280–300)
Urine
Sodium	12 mmol/L
Potassium	8 mmol/L
Osmolality	0 mmol/kg
A Suggested Comment: Severe hyponatraemia and a hypo-osmotic plasma with appropriately dilute urine and no evidence of salt wasting. This pattern is in keeping with primary polydipsia with water intoxication. While risperidone is not necessarily the cause of the polydipsia, an alternative atypical antipsychotic may be more effective in improving polydipsic behaviour.

There is a school of thought that with electronic reporting, help with interpretation of results should be offered in the form of hyper-linking to knowledge resources (e.g. FAQs, guidelines, laboratory handbooks available on the intra- or internet) rather than as added comments.⁹ Printed reports can also have details of such sources of information. However, electronic hyperlinks allow clinicians to navigate to the information they want in real time, and also allow other communications, including e-mail to the laboratory.⁹ Of course the two approaches are not mutually exclusive. The “hyper-linking to knowledge” approach is beyond the scope of this review, and has been reviewed elsewhere. ⁹

Some pitfalls to avoid with interpretative commenting

1. Restating the obvious; e.g. “normal sodium” “raised potassium”. However, qualifying the degree of abnormality may be useful e.g. “severe’’ or “life- threatening hyperkalaemia”.

2. Restating the clinical question; e.g. Where the clinical notes on the request form state “?hypothyroid” a report comment “consider hypothyroidism” does not add value. A more useful comment in this situation (depending on the results) would be “results confirm hypothyroidism”.

3. Commenting on reports to a doctor who has indicated (s)he does not wish to receive them.

4. Commenting on speciality reports to a specialist in the field (e.g. thyroid function tests sent to an endocrinologist) unless specifically sought, except in the case of complex dynamic function tests (e.g. adrenal vein sampling, water deprivation test).

5. Telling the clinician how to do his job: “Suggest clinical examination”; “Suggest check BP”.

6. Suggestion for invasive investigations should not be given lightly, e.g. “Suggest liver biopsy”.

Proactive communications and clinical meetings

Commenting in a vacuum by a faceless person in an ivory tower will be resented by clinicians. Interpretative commenting should go hand in hand with regular contact with clinicians to develop a dialogue about appropriate testing, developing testing algorithms, agreeing on test protocols as well as diagnostic criteria (which can take the form of formal guidelines in some instances), and to get feedback about the comments that go out on the reports in order to improve and fine-tune them. At the same time, unusual and interesting results could be triggers for direct communication and discussion.³

In fact, regular and systematic communication between laboratory staff and clinical staff is important for the effective utilisation of laboratory services by clinicians as well as for improvement of the laboratory service to meet the needs of the end user.^1,4

Personnel qualified to comment

Pathologists and Scientists with appropriate professional qualification and credentialed as specifically trained in commenting can perform this function adequately in a Clinical Biochemistry laboratory. In addition, laboratory personnel with extensive expertise and experience in a highly specialised area often perform this function in their limited area of expertise. It is for the management of individual laboratories to define the standards for such personnel, ensure they have the appropriate qualification and training to be able to perform these tasks and to define the responsibilities of such services.^1,3,4 Interpretation provided by laboratory professionals with inadequate expertise can be clinically dangerous.^7,9 In addition to formal training, personnel performing interpretative commenting need to regularly audit their activity and be involved in continuing professional development which addresses such functions.⁶ Participation in relevant interpretative external quality assessment schemes is part of continuing professional development.^3,4 ISO 15189 states “External quality assessment programmes should, as far as possible, provide clinically relevant challenges ….. that check the entire examination process including pre- and postexamination procedures”.¹ Programs addressing interpretative commenting in Clinical Biochemistry are offered by RCPA QAP and UK NEQAS.¹¹