Summary

The example of QT prolongation monitoring is critically examined because the proposed value of doing it is being promoted through various channels including via Pharmaceutical company ‘Product Information Sheets (PIs)’. This commentary illustrates that much of the monitoring advice and instruction content which makes up ‘MIMS’ and the ‘Physicians’ Desk Reference’ etc. which are widely relied on is poor and of little clinical value. PIs contain biased, superficial and poor science. Yet they purport to advise and instruct doctors to monitor various clinical and laboratory variables at great cost of time and money and often with minimal evidence of benefit.

Introduction

The ‘Product Information Sheets’ (PIs), also referred to as ‘Summary of Product Characteristics’ (SmPC, or just SPC) are written by drug companies and provided to regulatory agencies (and published in book and now electronic form e.g. ‘MIMS’ and ‘Physicians’ Desk Reference’) in order to satisfy legal requirements for gaining regulatory approval for their drug.

They contain extensive suggestions and requirements for the monitoring of drug side effects and adverse interactions before and during treatment: ranging from clinical measures like weight and blood pressure to specific laboratory tests and electro-cardiograms (ECGs). Further background and details about various aspects of general PI information usefulness and reliability are discussed below. 

First, the topical question of QT prolongation is examined because for many psychotropic drugs (about 25%) monitoring of QT is suggested or ‘mandated’ (wording varies in different countries), but with dubious justification and little consideration of the time, trouble and cost of doing that. For instance, in the PIs reviewed recently Warnier et al. found: ‘in almost half of the drugs [there was] no clear message on QT prolongation’ (1).

There are complex questions to be addressed if QT prolongation is going to be considered a recommended or necessary step with a significant number of newer generation psychotropic drugs. The degree of risk is currently unclear both with therapeutic doses and over-doses. The usefulness of monitoring, how best to go about it, and how to assess its cost are questions that research has barely started to address. If the occurrence of sudden cardiac death (SCD) is deemed sufficiently frequent to require action then one simple approach will be to simply not use drugs which cause QT prolongation. Since, generally speaking, there is little or no evidence that one drug is significantly therapeutically superior to others there seems an argument for simply not using such drugs at all. Considerably more epidemiological data needs to be gathered before such questions can be answered at all reliably and this is likely to take some years, so this problem will be with us for a while.

The QT Interval and its Prolongation

Before looking at a wider selection of examples assessing the quality of the PI content (discussed in more detail below) it is illustrative to consider the QT prolongation issue because it is a good example of the clinical problems that result from a combination of incomplete knowledge and the critical application of science to clinical medicine. Drug companies and the FDA (and other similar agencies) are in a dither because they do not really know what the risks are, nor what best to do. They tend to (indeed may be obliged to) play safe because there is insufficient information. Therefore they issue a ‘Black Box’ warning which may then be misconstrued and may elicit inappropriate and unintended reactions from doctors, other health professionals and auditing authorities etc. This shows why one should not accept uncritically what the PIs recommend, nor how others interpret them. These and similar peripheral but pertinent matters are covered in part 2 of this commentary.

Many research papers in the QT field report measurements of the QT interval that are less useful than they might be because they use the less accurate, but easy, automated measurements from standard electro-cardiograms (2-4). Therefore, from just this one of various other inaccuracies, much of the data being considered are made imprecise from their very source.

Remember, QT Prolongation is only a surrogate marker for serious arrhythmia and sudden cardiac death (SCD) and surrogate markers are notoriously quixotic and misleading. They are nonetheless popular because they are easy, cheap and quick.

There are a series of issues concerning scientific and measurement technique and logical reasoning that are pertinent to this and I will list them briefly, then consider a couple of them in greater detail.

1) Measurement. The QT interval, as measured in common clinical practice and research, is taken from the computerised ECG which gives a significantly inaccurate estimation of the real QT interval (e.g. (4)). Even medical toxicologists use the computerised ECG output half of the time (5), and half also use the unsatisfactory Bazett's QT rate correction. Only a small minority used the benchmark QT normogram (6).

2) Rate correction. The corrected QT interval (QTc) which is intended to correct for its length as the heart rate varies, usually uses the Bazett formula which is inaccurate and not commensurate with current best practice. Bazett’s formula is only useful for a narrow range of heart rates and significantly over-corrects for fast heart rates and under-corrects for slow heart rates (6).

3) Reference QT values. Population averaged QT values are less appropriate for individuals because an individual’s QT intervals are stable (i.e. there is low intra-individual variation), but inter-individual values in a population vary much more (7). It is therefore the change from an individuals baseline value which is the more reliable indicator, not deviation from a population average. Relatively few studies have assessed changes from baseline values, those that have not are of diminished predictive validity.

4) Cutoff values. There are very few data from which good predictions about torsades des pointes (TdP) or fatal arrhythmias can be made based on particular cut-off values. There is no consensus definition of ‘clinically significant’ prolongation of QTc interval, 450 ms or an increase from baseline of 30 ms are commonly used minimal values: note ICH document (2014) now states ‘Bazett’s corrections is no longer warranted’ (8). Different centers adopt different values, ranging between 440 and 500 ms. An upper 95% confidence limit of the average 24-hour QTc interval was 440 msec in men and 460 msec in women in one study.

5) The relationship between drug induced QT prolongation and TdP and fatal cardiac arrhythmias (FCA) is not reliably established. In case reports at least one other established risk factor for QTc prolongation was present in over 90 % of cases. Hasnain and Vieweg’s conclusion was ‘There is little evidence that drug-associated QTc interval prolongation by itself is sufficient to predict TdP. (9)’. An assessment of drugs that affect QT prolongation is maintained at  www.CredibleMeds.org.

6) The relationship between QT prolongation and the risk of TdP)* is different for different drugs. In other words, some drugs cause QT prolongation but not TdP and other drugs cause TdP with less significant prolongation. E.g. amiodarone and most tricyclic anti-depressants cause QT prolongation but not TdP (10, 11).

*Note: TdP is spelled in various ways in both French and English texts, but it is plural, so torsades des pointes is preferable (12).

7) TdP itself is not necessarily a good predictor of the occurrence of potential or actual fatal cardiac arrhythmias (FCA). Other factors seem to be significant or even dominant determinants of the occurrence of arrhythmias (ironically, ones which are sometimes not recognized or considered). E.g. alpha-1 adrenergic antagonism (13), channelopathies (14, 15), various other drugs (e.g. sotalol), obesity, electrolyte abnormalities (hypokalaemia, hypocalcaemia, hypomagnesaemia), cardiac ischaemia, cardiomyopathies, hypothyroidism and hypoglycaemia (3). Also age, female gender and even time of day influence QT length.

8). The hydrophobic central cavity of the HERG-K+ channels, allows a large number of structurally unrelated drugs to bind and cause direct channel inhibition (16). There are in fact about 10 separate K+ channels expressed in various parts of heart muscle which have slightly different properties, three drive ventricular repolarization, with some redundancy (17). It may be that functional impairment of more than one of them is required before fatal arrhythmias are likely to occur. Genetic abnormalities of several of these channels have been linked to various syndromes, e.g. Brugada and long QT (18).

QT: Predictive Value

Reliable data concerning which drugs really do cause an increase in FCA are sparse and since this is the key question substantial predictive uncertainty remains (11, 19, 20). It will be some time before the complexities of the situation can be untangled because (as a hypothetical scenario) some drugs may cause FCA even at therapeutic doses, but only if a particular genetic abnormality is present (or another drug), but not cause FCA in over-dose in the normal population, whereas others may be the reverse i.e. cause no problems at therapeutic doses but FCA in overdose. A Gordian knot to be untangled: who will rule Asia?

Just as is the case in my area of expertise, serotonin toxicity, one of the crucial elements required to clarify the picture is good systematically collected data on human toxicology, as has been pointed out in the past (21). The tremendous value of such an endeavour has been highlighted in a recent paper summarizing more than 25 years-worth of prospectively gathered data from the Hunter Area Toxicology Service (HATS) in Australia (22). This group have gathered data which have been valuable in assessing the issue of QT prolongation and FCA (many key papers referenced herein emanate from the HATS group). The HATS data have also been the most valuable single source of information for my research on ST. We all owe the HATS group members, past and present, a considerable debt of gratitude for what must have required a great deal of hard work.

So I would like to publically thank all who have been members of the HATS team over the years who surely are justifiably proud of their collective achievements.

Note: key reference, The Hasnain and Vieweg paper has a wealth of information and references (9). Likewise (17).

Product Information Sheets: Background and Context

The compendium made up of ‘Product Information Sheets’ provided by drug companies to regulatory agencies (e.g. MIMS and ‘Physicians’ Desk Reference’) is a source of information about drugs that, apparently, is frequently used both by doctors and other healthcare professionals (23). However, anecdotal information suggests there is great regional variation in its utilization and that specialists and more experienced doctors regard it as of little value and rarely use it. It would appear to have more prominence in ‘commercial’ software packages for practice management and clinical records which are used more in private practice. In the UK most doctors (appropriately) prefer the BNF which is independent and is compiled by practicing doctors and pharmacists, not drug company employees.

http://www.bnf.org/bnf/index.htm

PI based compendiums like MIMS generally describe themselves in the following terms … ‘an independent medicines information resource used and trusted by health professionals’. They even claim to be ‘evidence based’. Well, they may be used and trusted by some, but are they worthy of that trust? And ‘independent’? ‘evidence based’? those are an outrageous and affronting uses of those words which strain their definitions to breaking point. Since when, exactly, have pharmaceutical companies been independent sources? There is little or no substantiation of PI monitoring information being evidence based.

The above phrasing betrays my answer to the question ‘are they worthy of that trust?’ It is an unequivocal and emphatic no. No, the PI is not reliable; no, it is not unbiased and no, it is not trustworthy. The comments herein relating to QT prolongation demonstrate this clearly.

The unhelpful recommendations and instructions in PIs are a matter of some consequence since these compendiums may contain extensive suggestions and ‘requirements’ (the word ‘mandatory’ is used) for the monitoring of drug side effects and adverse interactions before and during treatment. These, especially black box warnings, are being written about, taken notice of and acted on, at least by some, e.g. Moore describes how authorities in California have threatened to remove re-imbursements or accreditation from hospitals where continued use of droperidol occurs (24). The unease and even fear that some doctors experience in relation to not adhering to such guidelines and mandates is real and has a negative influence on medical practice.

Note: Further general discussion about various aspects of PIs including the status and impact of black box warnings will be contained in part 2 of this commentary.

Suggestions and requirements for monitoring in the PIs include checking the patient's weight and blood pressure, or doing specific laboratory tests and other investigations, such as electro-cardiograms (ECGs), before and during treatment. Such tests often need to be done by other specialists and repeated on a number of occasions. It is obvious that the costs will be considerable and will also mount rapidly.

It might be generally agreed that advice and guidance concerning the monitoring of drug effects and interactions is a necessary and laudable objective. However, it must be done by independent reviewers and clinicians with the patient's interests and welfare at the forefront. Having it dictated and directed by those who are probably both the most partisan and the least qualified contributors in this arena i.e. the companies who manufacture the drugs, is clearly utterly inappropriate. It is about as sensible as asking a second-hand car dealer about the reliability of the car he proposes to sell you.

These PIs are written by the company seeking approval for their drug and they are called by various names like ‘Product Information Sheets’ (PIs), ‘Summary of Product Characteristics’ (SmPC, or just SPC).

It goes without saying that pharmaceutical companies have a strongly vested interest in the impression created by the contents and the message. These documents do not receive any declared or verified expert review or peer review (as does a typical scientific paper or review). Regulatory agencies do not have either the capability or the remit of performing those quality control and reviewing functions. It is not appropriate to regard PIs as pharmacology or clinical texts; i.e. the are not scientific documents.

Although regulatory agencies provide guidelines on what should be included in the PI the drug companies have a lot of freedom about what to put in, or leave out, and how to phrase it. They sometimes argue robustly about the exact wording if warnings or qualifications are asked for, usually in order to show their product in the most favourable light possible. The may simply refuse to produce data (see part 2 of this commentary for elaboration on this and similar points). For many years drug companies completely omitted or minimised the impact of their drugs on causing drug interactions via cytochrome P-450 inhibition. I wrote an editorial in ‘Expert Opinion on Drug Safety’ years ago (25) pointing this out in relation to fluoxetine and criticising an update about the ‘safety’ of fluoxetine published by a drug company employee, which was something of a whitewash (26).

When PIs are revised these agencies may occasionally mandate certain warnings, like they did with the black box warnings about TdP with droperidol (that was partly triggered by the company declining to produce requested data) and citalopram. History shows that the FDA struggle to make good decisions, sometimes through no fault of their own. Both those above black box decisions have been questioned by those who are regard them as inappropriate advice (9, 24, 27-33).

Note: there is discussion about how black box warnings are widely misconstrued by various groups in the health care field in Pt 2.

PIs have little to do with optimal clinical management (they are not written by practicing clinicians) and are influenced by medico-legal considerations, presumably intended to protect the producer from liability. They are peppered with ‘weasel words’. I know of instances from my areas of particular expertise where information and warnings have had nothing to do with pharmacology or established clinical effects and everything to do with avoiding liability.

There is also the important question concerning the rights, expertise, and responsibilities for drug company employees expressing opinions and unilaterally promulgating (via PIs) ‘mandatory’ requirements concerning the actions of physicians. This is especially so since there is little evidential basis or proper scientific rationale provided for the monitoring actions proposed in PIs.

Wide-Ranging Inadequacies of Advice in PIs

At least five major aspects of poor PI content have been adversely commented on recently: 

QT prolongation

Drug-drug interactions

Dose adjustment in renal impairment

Management of over-doses

Pregnancy and breastfeeding.

QT advice 

Warnier et al. (1) found: ‘in almost half of the drugs [there was] no clear message on QT prolongation’.

Drug-drug interactions

Bergk et al. (34) concluded: ‘To meet the SPCs claim of being the basis of information for health professionals on how to use medicinal products safely and effectively, information on drug interactions should be thoroughly up-dated and expanded’. 

Dose adjustment in renal impairment

Salgado et al. (35) found: ‘current versions of SmPCs are characterised by several information deficits and by containing recommendations that are not relevant to clinical practice in terms of dose adjustment in renal impairment.’

Over-doses

In relation to treatment of over-doses Wall et al. (36) found that ‘Gut decontamination advice in SPC documents with respect to CNS drugs was inadequate.

Pregnancy and breast-feeding

Arguello et al. (37) also found: ‘Important information deficits on the use of medicines during pregnancy and breastfeeding … in European SmPCs.’ 

Overview

A recent paper evaluating the general effectiveness of SmPCs (38) concluded ‘Current content and presentation of SmPCs [PIs], while meeting regulatory approval standards, contribute little to the safe and effective use of medication in practice.’

This poor advice in PIs is a frequent, time-consuming and expensive problem. Most psychotropic drugs have at least several instructions for laboratory monitoring most of those have insufficient information to usefully guide doctors. In some country’s PIs an ECG is required for 25% of listed antipsychotics (39, 40).

The only rational comment that can be made about this uncertain situation is that if such burdensome monitoring measures are required then the drug would automatically be disqualified from being usable in general clinical practice and would only be considered in rare special circumstances where there is no alternative.

Conclusion: a Misleading MIMs (PDR etc.)

The ubiquity and persistence (50+ years now) of the MIMs annual compendium and its various siblings and progeny in different countries (like the ‘Physicians’ Desk Reference’) is perhaps what makes them so insidiously misleading. They have been around for so long that there is a tendency to assume they are respected and valuable, their partisan nature is easily overlooked.

One is reminded of the quote from William James, the 19th century American physician/philosopher (from an interesting family of intellectuals which includes the author Henry James): ‘There is nothing so absurd that if you repeat it often enough, people will believe it.’

PIs present an enduring and formidable façade which conceals a paucity of quality evidence-based content. Not very reassuring for doctors (or pharmacists) on matters which account, inter alia, for the majority of hospital admissions for drug-induced problems.

One cannot help observing that an neat feat has been accomplished whereby drug companies have taken these PIs that they are obliged to produce for regulatory agencies and then bundled them into a publication that they then get health professionals to pay substantial amounts of money for every year. Talk about making a virtue out of necessity!* 

* An expression originating, like so many, in Shakespeare (Two Gentlemen of Verona, Act 4, sc.1).

References

1.Warnier, MJ, Holtkamp, FA, Rutten, FH, Hoes, AW, et al., Quality of drug label information on QT interval prolongation. Int J Risk Saf Med, 2014. 26(2): p. 89-98.

2.Chan, A, Isbister, GK, Kirkpatrick, CM, and Dufful, SB, Drug-induced QT prolongation and torsades de pointes: evaluation of a QT nomogram. QJM, 2007. 100(10): p. 609-15.

3.Isbister, GK and Page, CB, Drug induced QT prolongation: the measurement and assessment of the QT interval in clinical practice. Br J Clin Pharmacol, 2013. 76(1): p. 48-57.

4.Berling, I, Isbister, GK, Calver, L, and Clunas, S, Digital Holter measurement of QT prolongation in ziprasidone overdose. Clin Toxicol (Phila), 2011. 49(7): p. 694-6.

5.Othong, R, Devlin, JJ, and Kazzi, ZN, Medical toxicologists’ practice patterns regarding drug-induced QT prolongation in overdose patients: A survey in the United States of America, Europe, and Asia Pacific region. Clin. Toxicol., 2015. 53(4): p. 204-209.

6.Isbister, GK, Risk assessment of drug-induced QT prolongation. Australasian Prescriber, 2015. 38(1): p. 20-24.

7.Newton-Cheh, C, Larson, MG, Corey, DC, Benjamin, EJ, et al., QT interval is a heritable quantitative trait with evidence of linkage to chromosome 3 in a genome-wide linkage analysis: The Framingham Heart Study. Heart Rhythm, 2005. 2(3): p. 277-84.

8.Anon, International conference on harmonisation of technical requirements for registration of pharmaceuticals for human use. The clinical evaluation of Qt/QTc interval prolongation and proarrhythmic potential for non-antiarrhythmic drugs. E14

http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E14/E14_Guideline.pdf

and

http://www.ich.org/fileadmin/Public_Web_Site/ICH_Products/Guidelines/Efficacy/E14/E14_QAs_R2_Step4.pdf

Accessed May 2015. 2005.

9.Hasnain, M and Vieweg, WV, QTc interval prolongation and torsade de pointes associated with second-generation antipsychotics and antidepressants: a comprehensive review. CNS Drugs, 2014. 28(10): p. 887-920.

10.Hondeghem, LM, Thorough QT/QTc not so thorough: removes torsadogenic predictors from the T-wave, incriminates safe drugs, and misses profibrillatory drugs. J. Cardiovasc. Electrophysiol., 2006. 17(3): p. 337-40.

11.van Noord, C, Sturkenboom, MC, Straus, SM, Witteman, JC, et al., Non-cardiovascular drugs that inhibit hERG-encoded potassium channels and risk of sudden cardiac death. Heart, 2011. 97(3): p. 215-20.

12.Mullins, ME, Mon bete noir (my pet peeve). J Med Toxicol, 2011. 7(2): p. 181.

13.Mow, T, Frederiksen, K, and Thomsen, MB, Assessment of anti-arrhythmic activity of antipsychotic drugs in an animal model: influence of non-cardiac alpha(1)-adrenergic receptors. Eur J Pharmacol, 2015. 748: p. 10-7.

14.Petropoulou, E, Jamshidi, Y, and Behr, ER, The genetics of pro-arrhythmic adverse drug reactions. Br J Clin Pharmacol, 2014. 77(4): p. 618-25.

15.Campuzano, O, Allegue, C, Fernandez, A, Iglesias, A, et al., Determining the pathogenicity of genetic variants associated with cardiac channelopathies. Sci Rep, 2015. 5: p. 7953.

16.Cubeddu, LX, Iatrogenic QT Abnormalities and Fatal Arrhythmias: Mechanisms and Clinical Significance. Curr Cardiol Rev, 2009. 5(3): p. 166-76.

17.Schmitt, N, Grunnet, M, and Olesen, SP, Cardiac potassium channel subtypes: new roles in repolarization and arrhythmia. Physiol. Rev., 2014. 94(2): p. 609-53.

18.Merghani, A and Sharma, S, Identifying patients at risk of sudden arrhythmic death. Practitioner, 2012. 256(1755): p. 15-8, 2.

19.Teodorescu, C, Reinier, K, Uy-Evanado, A, Chugh, H, et al., Antipsychotic drugs are associated with pulseless electrical activity: the Oregon Sudden Unexpected Death Study. Heart Rhythm, 2013. 10(4): p. 526-30.

20.Peacock, J and Whang, W, Antipsychotic medications and sudden cardiac arrest: more than meets the eye? Heart Rhythm, 2013. 10(4): p. 531-2.

21.Gillman, PK, Serotonin Syndrome: History and Risk. Fundam. Clin. Pharmacol., 1998. 12(5): p. 482-491.

22.Buckley, NA, Whyte, IM, Dawson, AH, and Isbister, GK, A prospective cohort study of trends in selfpoisoning, Newcastle, Australia, 1987–2012: plus ça change, plus c’est la même chose. Med. J. Aust., 2015: p. https://www.mja.com.au/journal/2015/202/8/prospective-cohort-study-trends-self-poisoning-newcastle-australia-1987-2012-plus.

23.Vromans, L, Doyle, G, Petak-Opel, S, Rodiger, A, et al., Shaping medicinal product information: a before and after study exploring physicians' perspectives on the summary of product characteristics. BMJ Open, 2013. 3(8): p. e003033.

24.Moore, J, The Use of Droperidol in Anesthesia Care. California Society of Anesthesiologists Bulletin, 2009: p. 28-34.

25.Gillman, PK, Drug interactions and fluoxetine: a commentary from a clinician’s perspective. Ex Op Drug Saf, 2005. 4: p. 965-969.

26.Wernicke, JF, Safety and side effect profile of fluoxetine. Expert Opin Drug Saf, 2004. 3(5): p. 495-504.

27.Calver, L and Isbister, GK, High dose droperidol and QT prolongation: analysis of continuous 12-lead recordings. Br J Clin Pharmacol, 2014. 77(5): p. 880-6.

28.Calver, L, Page, CB, Downes, MA, Chan, B, et al., The Safety and Effectiveness of Droperidol for Sedation of Acute Behavioral Disturbance in the Emergency Department. Ann. Emerg. Med., 2015.

29.Tracz, K and Owczuk, R, Small doses of droperidol do not present relevant torsadogenic actions: a double-blind, ondansetron-controlled study. Br J Clin Pharmacol, 2015. 79(4): p. 669-76.

30.Faine, B and Hogrefe, C, News flash: Old Mother Hubbard reports the cupboard is bare...time for the FDA to let droperidol out of the (black) box. Ann Pharmacother, 2012. 46(9): p. 1259-61.

31.Spevak, C, Hamsher, C, Brown, CQ, Wedam, EF, et al., The clinical significance of QT interval prolongation in anesthesia and pain management: what you should and should not worry about. Pain Med, 2012. 13(8): p. 1072-80.

32.Vieweg, WV, Hasnain, M, Howland, RH, Hettema, JM, et al., Citalopram, QTc interval prolongation, and torsade de pointes. How should we apply the recent FDA ruling? Am J Med, 2012. 125(9): p. 859-68.

33.Habib, AS and Gan, TJ, Food and drug administration black box warning on the perioperative use of droperidol: a review of the cases. Anesth. Analg., 2003. 96(5): p. 1377-9.

34.Bergk, V, Haefeli, WE, Gasse, C, Brenner, H, et al., Information deficits in the summary of product characteristics preclude an optimal management of drug interactions: a comparison with evidence from the literature. Eur. J. Clin. Pharmacol., 2005. 61(5-6): p. 327-35.

35.Salgado, TM, Arguello, B, Martinez-Martinez, F, Benrimoj, SI, et al., Clinical relevance of information in the Summaries of Product Characteristics for dose adjustment in renal impairment. Eur. J. Clin. Pharmacol., 2013. 69(11): p. 1973-9.

36.Wall, AJ, Bateman, DN, and Waring, WS, Variability in the quality of overdose advice in Summary of Product Characteristics (SPC) documents: gut decontamination recommendations for CNS drugs. Br J Clin Pharmacol, 2009. 67(1): p. 83-7.

37.Arguello, B, Salgado, TM, and Fernandez-Llimos, F, Assessing the information in the Summaries of Product Characteristics for the use of medicines in pregnancy and lactation. Br J Clin Pharmacol, 2015. 79(3): p. 537-44.

38.Raynor, DK, De Veene, P, and Bryant, D, The effectiveness of the summary of product characteristics (SmPC) and recommendations for improvement. Therapeutic Innovation & Regulatory Science, 2014. 48(2): p. 255-265.

39.Geerts, AF, De Koning, FH, Van Solinge, WW, De Smet, PA, et al., Instructions on laboratory monitoring in 200 drug labels. Clin. Chem. Lab. Med., 2012. 50(8): p. 1351-8.

40.Rougemont, M, Ulrich, S, Hiemke, C, Corruble, E, et al., French summaries of product characteristics: content in relation to therapeutic monitoring of psychotropic drugs. Fundam. Clin. Pharmacol., 2010. 24(3): p. 377-84.