- Email: [email protected]
Patient information leaflets for antidepressants: Are patients getting the information they need?
Journal of Affective Disorders 128 (2011) 165–170
Contents lists available at ScienceDirect
Journal of Affective Disorders j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j a d
Brief report
Patient information leaflets for antidepressants: Are patients getting the information they need? Camilla Haw a,⁎, Jean Stubbs b a b
Consultant Psychiatrist, St Andrew's Healthcare, UK Research Pharmacist, St Andrew's Healthcare, UK
a r t i c l e
i n f o
Article history: Received 11 May 2010 Revised in revised form 26 June 2010 Accepted 27 June 2010 Available online 24 July 2010
Keywords: Patient information leaflets Antidepressants Patient information
a b s t r a c t Background: According to European law a comprehensive patient information leaflet (PIL) has to accompany all medicines. In this study we examined the uniformity, adequacy and balance of information contained in UK antidepressant PILs. Methods: We studied antidepressant PILs available in the Electronic Medicines Compendium and subjected each to a content analysis. Words were assessed as being positive, negative or neutral. Results: Forty-two PILs concerning 21 different antidepressants and 23 pharmaceutical companies were studied. PILs presented information about side effects in a strikingly heterogeneous way, making it difficult for patients to find the required information. Half the PILs provided no information about how the antidepressant is thought to work. Over 90% stated the antidepressant would take 2–4 weeks to work, although a few PILs indicated earlier onset of improvement. Not all PILs warned about discontinuation syndrome and advice about alcohol was generally that it was prohibited. Almost half of PILs made no mention of St John's wort and its potential for interaction with the antidepressant. Two-thirds of PILs provided no information about the likely duration of treatment. PILs contained far more words judged to be negative rather than positive or neutral. Limitations: Data were extracted by a single researcher, although inter-rater agreement was high. Conclusions: Further guidance and tightening of the approval process for PILs are needed to ensure they are more standardised in content and contain more information that is wanted by and is useful to patients. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Patients need information about proposed medication so that they can give informed consent to drug treatment. They may also need written information to refer to while they are taking the medication (Raynor et al., 2007). Depressed patients require information about antidepressants to dispel commonly held erroneous beliefs, for example that it is only necessary to take the antidepressants on those days when you feel depressed (Jorm et al., 2005). Verbal information is ⁎ Corresponding author. Consultant Psychiatrist, St Andrew's Healthcare, Billing Road, Northampton, NN1 5DG, UK. Tel.: + 44 1604 616 186. E-mail address: [email protected] (C. Haw). 0165-0327/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2010.06.034
often poorly retained by those who are stressed or anxious and in addition may be remembered incorrectly. Written information is better remembered than that given orally and may lead to better treatment adherence (Kessels, 2003). Psychiatric patients have access to written information about their medication from a variety of sources, including their psychiatrist and other healthcare professionals, the internet, media and health charities. One source of information that most patients have ready access to is the Patient Information Leaflet (PIL) that by law has to be issued with each medicine. According to a European Union Directive which came fully into effect in 1999 after a five year phasing in period, all medicines must be accompanied by a PIL setting out comprehensive and accessible information so
166
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
that patients can use their medicines safely and appropriately (Council Directive 92/27/EEC of 31 March 1992). This Directive was subsequently superseded by Directive 2001/83/EC of 6 November 2001 which was then partly amended by Directive 2004/27/EC of 31 March 2004. PILs are required to have a standard content in a specified order and to contain all the information in the Summary of Product Characteristics (SPC). Additional, educational information can be included provided it is non-promotional. PILs are required to be tested on users to ensure they are readable, understandable and comprehensive. The requirement for testing was contained in Directive 2004/ 27/EC of 31 March 2004 and came into force in 2005 for new medicines with a three year window in the UK for existing leaflets to be passed which ended in July 2008. Directive 2004/ 27/EC of 31 March 2004 stipulates [in article 63(2)] that leaflets must be clearly written and designed to be clear and understandable, enabling the users to act appropriately. The Medicines and Healthcare Products Regulatory Authority (MHRA) has written detailed guidance on communicating the risks and benefits in PILs (MHRA, 2005). All PILs have to be approved by the MHRA prior to distribution to the public. Despite efforts to produce a useful source of information, patients can find it difficult to extract essential information from PILs (Dickinson et al., 2001). Patients value information containing a good balance of benefit and harm information (Raynor et al., 2007). They want information about a medicine's indications, benefits, likely duration of treatment as well as side effects (Shrank et al., 2007). The aim of this study was to look at the uniformity, adequacy and balance of information contained in antidepressant Patient Information Leaflets available in the UK and hence how well current PILs inform patients about whether or not they should take their antidepressants.
2. Method We studied PILs for all antidepressant drugs listed in section 4.3 of the British National Formulary (BNF) No 58 (British Medical Association and Royal Pharmaceutical Society, 2009). The PILs available in the Electronic Medicines Compendium (eMC) at www.Medicines.org.uk on 1 January 2010 were investigated. Where an antidepressant was manufactured by more than one pharmaceutical company, we examined the PILs that each of the manufacturers had produced. If a manufacturer had produced different PILs for different formulations of the same drug (for example tablets and liquid), we studied the PIL for the most commonly used product. We investigated the following items which we as clinicians considered were important if patients are to be properly informed about their antidepressant medication: 1. Readability using Flesch reading ease and Flesch-Kincaid Grade Level (Microsoft Office Word, 2003). We also examined passive sentences. 2. Information given about the potential benefits of taking the antidepressant. 3. How the antidepressant works. 4. Speed of onset of action. 5. Likely duration of treatment and time to recover fully from depression.
6. How the risks of taking the antidepressant were described, including increase in suicidality after starting antidepressants, the dangers of abrupt withdrawal, advice about consuming alcohol, presentation of side effects and interactions with St John's wort. 7. The overall balance of information in PILs. The amount (word count) of information that was judged by us to be (i) positive or neutral or (ii) negative about the medicine in each PIL. Information was judged to be ‘positive’ or ‘neutral’ if it explained what the medicine is for, how it works, what benefits will taking it bring, how quickly it works, how long it is necessary to take it for, how and when to take the medicine, the recommended dosage and how to store the medicine. Information was regarded as ‘negative’ if it stated the risks of taking the antidepressant, including side effects, contra-indications, increased risk of suicide and self harm, not advised for certain patient groups, interactions, contra-indicating consumption of alcohol, possible detrimental effect upon ability to drive, effect of taking an overdose and the dangers of abrupt withdrawal. The number of positive and neutral and also negative words was calculated for each PIL by JS and expressed as a percentage of the total word count (information about ingredients and the manufacturer's address were excluded). CH independently rated the positive, neutral and negative words in five randomly selected PILs and compared her ratings with those of JS. There was agreement between the two raters over the allocation of 98.1% of words. 3. Results Forty-two PILs concerning 21 different antidepressants were listed in eMC. Using the classification given in the BNF No 58, 12 PILS were for tricyclics, 3 for MAOIs, 16 for SSRIs and 11 were for other antidepressants. In total, 23 different pharmaceutical companies were involved in their manufacture. Fluoxetine and venlafaxine were manufactured by 5 different companies; amitriptyline, trazodone, citalopram and paroxetine by 3; imipramine, fluvoxamine, sertraline and mirtazapine by two and the remaining 12 antidepressants by single companies. All but one PIL had undergone MHRA testing. 3.1. Readability statistics Median score on the Flesch–Kincaid Reading Ease Scale was 55.2 (inter-quartile range 52.4–58.5). On the Flesch– Kincaid Grade Level the median score was 7.3 (inter-quartile range 6.8–7.9). Use of passive sentences was uncommon in PILs (median 5%; inter-quartile range 2.3–7.0%). Thus, most PILs scored well on these statistics. However, it must be borne in mind that readability formulae only measure two aspects of readability, namely word length and sentence length and there are many other factors that influence readability. 3.2. Information about how antidepressants work and speed of onset of action Twenty-two (52.4%) PILS gave some sort of an explanation as to how the antidepressant worked, some mentioning a
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
specific mechanism of action, such as inhibition of the reuptake of serotonin or blockade of the action of monoamine oxidase. In some cases the explanation was detailed, for example a PIL for citalopram stated that ‘it increases the effect of the body's naturally occurring hormone, serotonin, by inhibiting its re-uptake into cells’. In other cases the explanation was more rudimentary, for example, a leaflet for lofepramine stated that it ‘alters the levels of chemicals in your brain to relieve symptoms of depression’. Twenty (47.6%) PILs gave no explanation at all (see Table 1). Thirtyeight (90.5%) PILs indicated that the antidepressant would
167
take between 2 and 4 weeks to take effect, with just two leaflets indicating that improvement might occur in a matter of days. Only one leaflet provided advice on what to do if no improvement had taken place after several weeks of treatment. 3.3. Presentation of information about side effects and suicidality Information about side effects was presented in a strikingly heterogeneous way. The majority (31; 73.8%) of
Table 1 Information presented in antidepressant Patient Information Leaflets (N = 42). Information presented in Patient Information Leaflet
N
(%)
How the antidepressant is thought to work No explanation Alters brain chemicals to relieve depression Increases serotonin and/or noradrenaline in the nervous system thus correcting chemical imbalance Blocks the action of enzyme monoamine oxidase in the brain which controls mood
20 7 13 2
(47.6) (16.7) (31.0) (4.8)
Speed of onset of action No information May feel an improvement in 2–3 days/take several days Two to four weeks If you don't feel better after 4 weeks go back to the doctor as you may need a higher dose
1 2 38 1
(2.4) (4.8) (90.5) (2.4)
Presentation of side effects Listed in decreasing order of seriousness Serious side effects first, then by frequency e.g. up to 1 in 10 people Serious side effects first, then by bodily system e.g. skin, cardiovascular Serious side effects first, then a random list Random list of side effects Listed in decreasing order of frequency e.g. up to 1 in 10 people Listed by bodily system e.g. skin, cardiovascular
14 10 5 2 5 4 2
(33.3) (23.8) (11.9) (4.8) (11.9) (9.5) (4.8)
Suicidality No information Suicidal thoughts just listed as a side effect. No information about suicidality in young people May get worse when start the antidepressant If become suicidal tell the doctor — it may be necessary to reduce the dose or stop the drug
1 1 38 2
(2.4) (2.4) (90.5) (4.8)
Discontinuation syndrome No information Warning not to stop antidepressant abruptly but no reason given or possible symptoms described Warning not to stop antidepressant abruptly and possible symptoms described
7 3 32
(16.7) (7.1) (76.2)
Advise about consuming alcohol No information Alcohol is contra-indicated Avoid alcohol as it may increase the effects of the antidepressant Avoid alcohol as the antidepressant may increase the effect of the alcohol Sensible to avoid alcohol but this antidepressant is not expected to interact with alcohol
5 25 4 6 2
(11.9) (59.5) (9.5) (14.3) (4.8)
St John's wort No information Tell your doctor or pharmacist if you are taking St John's wort St John's wort should not be taken at the same time as this antidepressant St John's wort can interact with the antidepressant St John's wort can lead to serotonergic syndrome, potentially life-threatening
20 8 9 2 3
(47.6) (19.0) (21.4) (4.8) (7.1)
Likely duration of treatment No information Don't stop the treatment if you feel better — the illness may come back Less than 3 months Usually at least 4–6 months after you feel better Maintenance treatment may be needed
27 2 1 11 1
(64.3) (4.8) (2.4) (26.2) (2.4)
168
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
PILs listed the most serious side effects, such as allergic reactions, first but there was then no standardised way of providing information about the remaining less serious side effects (see Table 1). These could be listed by severity, frequency, and bodily system or were in random order. Fourteen (33.3%) PILs made use of frequencies (e.g. less than one in ten people but more than one in a hundred people) to provide patients with information about how common particular side effects were. Forty (95.2%) PILs contained a prominent section warning patients that suicidal thoughts might get worse when starting the antidepressant and that this was particularly the case in young people. Advice was given as to what to do in the event of this occurrence. However, two leaflets contained no warning about antidepressants potentially causing an increase in suicidal thoughts or actions. This is likely to reflect the SPC. 3.4. Information about discontinuation syndrome Although all antidepressants can be associated with a discontinuation syndrome, a minority (7; 16.7%) of PILs contained no warning about discontinuation syndrome and a further three merely warned patients not to stop taking the antidepressant abruptly but did not give a reason or describe the symptoms of discontinuation syndrome (see Table 1). Thirty-two (76.2%) PILs gave adequate information about discontinuation syndrome. 3.5. Advice about consuming alcohol Five (11.9%) PILs did not provide any advice about drinking alcohol while taking antidepressants (see Table 1). Twenty-five (59.5%) contained a statement without any explanation that alcohol was contra-indicated while taking the antidepressant. Ten (23.8%) PILs advised against drinking alcohol stating that either alcohol would increase the effect of the antidepressant or vice-versa. Two PILs expressed caution about drinking alcohol while taking the antidepressant but said that no interaction was expected. 3.6. St John's wort While 20 (47.6%) PILs contained no information about use of St John's wort, the remaining leaflets indicated that concurrent usage with the prescribed antidepressant might be problematic (see Table 1). Warnings ranged from informing your doctor or pharmacist that you are taking St John's wort, to statements that St John's wort may interact with the antidepressant or even that St John's wort when combined with a serotonergic antidepressant can cause a potentially life-threatening serotonergic syndrome. 3.7. Likely duration of treatment No PIL provided information about how long it would take to recover from a depressive episode and only 12 (28.6%) provided information about how long patients were likely to have to take the antidepressant (see Table 1). One PIL stated that the likely duration of treatment was less than 3 months. Eleven (26.2%) leaflets stated the usual duration of treatment was at least 4–6 months after symptom resolution, while one
leaflet mentioned that maintenance treatment might be necessary. 3.8. Balance of information The median positive and neutral word count of the PILs was 785 words (inter-quartile range 652–967) and the median negative word count was 1393 (inter-quartile range 1061–1881). 4. Discussion Although all of the PILs scored well on the readability statistics, as Dickinson et al. (2001) have pointed out, such readability formulae are based on word and sentence length and therefore have their limitations. The same sentence written backwards would have the same readability score as one written forwards. Other factors, including layout, font size and style are also important. A limitation of the study is that it considered only the content of leaflets and not their format. The balance of information provided in the PILs we examined was predominantly negative and this is almost inevitable given the legal requirement to contain all the information in the SPC unless companies are going to include additional positive information in antidepressant PILs. Only half of the PILs studied gave any explanation as to how the antidepressant worked, with very few leaflets giving a detailed explanation of the mechanism of action. This is an important omission given our anecdotal impression of how frequently depressed patients ask about the mechanism of action of antidepressants. Patients need a balance of information about antidepressants so that they can decide whether or not to take (or continue to take) them. The regulator in their guidance to the pharmaceutical industry has stated that PILs need to include more benefit information (MHRA, 2005). Although most PILs stated the antidepressant would take between 2 and 4 weeks to work, it is now generally recognised that some improvement after starting antidepressants can occur within the first week of treatment (although a longer period of administration is needed before pronounced clinical effects occur) and regularly occurs within a few days but only a few leaflets acknowledged this early improvement. There is no evidence that particular antidepressants have a more rapid speed of onset than others, although amongst the SSRIs fluoxetine appears to have a slightly slower speed of onset of action (Edwards and Anderson, 1999). It is surprising that only one leaflet mentioned going back to the doctor to discuss a lack of progress in treatment and possible need for a dosage increment if no improvement had taken place after several weeks and none mentioned the possibility of change to a different antidepressant. The lack of information on the likely duration of treatment by two-thirds of PILs is unfortunate, since this is important information that patients need to know from the outset in order to try and prevent premature discontinuation. Of course, they should be told this by their psychiatrist but it is good for them to have this reiterated by the PIL. It is well established that the minimum duration of treatment after an episode of depression should be 4–6 months after complete recovery (Loonen et al., 1991;
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
Reimherr et al., 1998). Stopping antidepressants before this time carries a substantial risk of relapse. The PILs we studied presented antidepressant side effects in diverse ways and it was difficult to locate the same side effect in different leaflets. For this reason it would be better for patients if a standardised format were to be used for all PILs so as to guide readers. Listing by bodily system is helpful when a patient is experiencing a particular side effect since they should be able to easily locate the side effect on the PIL if they are listed in this way. On the other hand, listing the most serious side effects first followed by grouping according to frequency is an easily understood system which may be more helpful to patients when checking the leaflet to see what to expect (Carrigan et al., 2008). This latter method of listing side effects is that recommended by both the UK and EU regulators and this seems appropriate, given that serious side effects require immediate action. All antidepressants when taken for more than a few weeks can be associated with a discontinuation syndrome, although some are more strongly associated with it (for example venlafaxine) than others (for example fluoxetine) (Lejoyeux and Adès, 1997). It was, therefore, surprising that not all PILs contained a warning together with an explanation as to why antidepressants should not be stopped abruptly. Furthermore, not all PILs gave a description of symptoms associated with abrupt withdrawal. Most PILs advised patients to avoid alcohol but unhelpfully did not provide them with a reason for this. A minority did provide a rationale; either that alcohol would enhance the action of the antidepressant or vice-versa. Patients very commonly ask psychiatrists if it is alright to have a drink while taking antidepressants. Anecdotally clinicians are aware that statements that alcohol is contraindicated can have the effect of patients choosing not to take the antidepressant preferring instead to continue to drink alcohol for symptomatic relief. This is a difficult area but a few PILs were helpful in that although they advised against drinking alcohol while taking antidepressants they stated that an interaction was not expected. Of course, clinicians should be aware of patients who are abusing alcohol, comorbidity of alcohol use disorders being very common with depressive disorder (Grant and Harford, 1995). The presence of an alcohol disorder may influence the choice of antidepressant. It would be helpful if PILs could express caution when drinking alcohol while taking antidepressants because of the potential for interaction but also because alcohol itself can have a negative effect upon mood. Many patients with depression try St John's wort (Hypericum) before consulting a medical practitioner about conventional antidepressants. An estimated 1.5 million Britons bought St John's wort in 2009 (Daily Mail, 2010). St. John's wort should not be taken at the same time as other antidepressants since it has the potential to cause a serotonergic syndrome with serotonergic agents and can reduce the efficacy of others by its enzyme-inducing activities (Hammerness et al., 2003). It was, therefore, surprising that only half the PILs studied contained a warning about St John's wort and again, of those that did only a few explained that St John's wort should not be taken with a conventional antidepressant. The antidepressant PILs studied varied in quality. Although PILs are supposed to contain all the information in the
169
SPC in a specific order, in practice their content varied quite widely. The same antidepressant produced by several different companies could have PILs of diverse content, which could be very confusing to patients should they read the different leaflets. One reason for the differing content is, of course, that they may have different SPCs. With generic prescribing this scenario is likely to arise quite frequently. We suggest that further guidance on the content of PILs and also a tightening of the approval process are needed to ensure they are more informative to patients, and as well as containing a summary of the SPC they should contain important items such as those we have highlighted in this study, for example what to do if there is no improvement in symptoms after 4 weeks or so of treatment and the likely duration of treatment once depressive symptoms have resolved. Role of funding source None.
Conflict of interest None.
Acknowledgement Our thanks to Professor PJ Cowen for his helpful comments on the manuscript.
References British Medical Association & Royal Pharmaceutical Society, 2009. British National Formulary No 58. BMJ Publishing group, London. Carrigan, N., Raynor, D.K., Knapp, P., 2008. Adequacy of patient information on adverse effects: an assessment of patient information leaflets in the UK. Drug Safety 31, 305–312. Council Directive 92/27/EEC of 31 March 1992 on the labelling of medicinal products for human use and on package leaflets (OJ No. L 113 of 30.4.1992, p.8). Directive 2001/83/EC of 6 November 2001 on the community code relating to medicinal products for human use. (OJ No. L 311 of 28.11.2004, p67128). Directive 2004/27/EC of 31 March 2004 amending Directive 2001/83/EC on the community code relating to medicinal products for human use (OJ No. L136 of 30.4.2004, p34-57). Daily Mail, 2010. Alternative medicine sales soar as consumers shake off cynicism. April 6th. Dickinson, D., Raynor, D.K., Duman, M., 2001. Patient information leaflets for medicines: using consumer testing to determine the most effective design. Patient Education and Counselling 43, 147–159. Edwards, J.G., Anderson, I., 1999. Systematic review and guide to selective serotonin reuptake inhibitors. Drugs 57, 507–533. Grant, B.F., Harford, T.C., 1995. Comorbidity between DSM-IV alcohol use disorders and major depression: results of a national survey. Drug and Alcohol Dependence 39, 197–206. Hammerness, P., Basch, E., Ulbricht, C., Barrette, E.P., Foppa, I., Basch, S., Bent, S., Boon, H., Ernst, E., 2003. St John's Wort: a systematic review of adverse effects and drug interactions for the consultant psychiatrist. Psychosomatics 44, 271–282. Jorm, A.F., Christensen, H., Griffiths, K.M., 2005. Belief in the harmfulness of antidepressants: results from a national survey of the Australian public. Journal of Affective Disorders 88, 47–53. Kessels, R.P.C., 2003. Patients' memory for medical information. Journal of Royal Society of Medicine 96, 219–222. Lejoyeux, M., Adès, J., 1997. Antidepressant discontinuation. A review of the literature. Journal of Clinical Psychiatry 58 (Suppl 7), 11–16. Loonen, A.J., Peer, P.G., Zwanikken, G.J., 1991. Continuation and maintenance therapy with antidepressant agents. Meta-analysis of research. Pharmacy Weekly Science 23, 167–175. Medicines and Healthcare Products Regulatory Authority, 2005. Always read the leaflet: getting the best information with every medicine. Available at: http://mhra.gov.uk/Howweregulate/Medicines/Medicinesregulatorynews/CON09410. Microsoft Office Word, 2003. Microsoft Corporation, Washington.
170
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
Raynor, D.K., Blenkinsopp, A., Knapp, P., Grime, J., Nicolson, D.J., Pollock, K., Dorer, G., Gilbody, S., Dickinson, D., Maule, A.J., Spoor, P., 2007. A systematic review of quantitative and qualitative research on the role and effectiveness of written information available to patients about individual medicines. Health Technology Assessment, Vol. 11. No. 5. Reimherr, F.W., Amsterdam, J.D., Quitkin, F.M., Rosenbaum, J.F., Fava, M., Zajecka, J., Beasley, C.M., Michelson, D., Roback, P., Sundell, K., 1998.
Optimal length of continuation therapy in depression: a prospective assessment during long-term fluoxetine treatment. American Journal of Psychiatry 155, 1247–1253. Shrank, W., Avorn, J., Rolon, C., Shekelle, P., 2007. Effect of content and format of prescription drug labels on readability, understanding, and medication use: a systematic review. Annals of Pharmacotherapy Available at: http://www.theannals.com/cgi/content/aph.1H582v1
Contents lists available at ScienceDirect
Journal of Affective Disorders j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j a d
Brief report
Patient information leaflets for antidepressants: Are patients getting the information they need? Camilla Haw a,⁎, Jean Stubbs b a b
Consultant Psychiatrist, St Andrew's Healthcare, UK Research Pharmacist, St Andrew's Healthcare, UK
a r t i c l e
i n f o
Article history: Received 11 May 2010 Revised in revised form 26 June 2010 Accepted 27 June 2010 Available online 24 July 2010
Keywords: Patient information leaflets Antidepressants Patient information
a b s t r a c t Background: According to European law a comprehensive patient information leaflet (PIL) has to accompany all medicines. In this study we examined the uniformity, adequacy and balance of information contained in UK antidepressant PILs. Methods: We studied antidepressant PILs available in the Electronic Medicines Compendium and subjected each to a content analysis. Words were assessed as being positive, negative or neutral. Results: Forty-two PILs concerning 21 different antidepressants and 23 pharmaceutical companies were studied. PILs presented information about side effects in a strikingly heterogeneous way, making it difficult for patients to find the required information. Half the PILs provided no information about how the antidepressant is thought to work. Over 90% stated the antidepressant would take 2–4 weeks to work, although a few PILs indicated earlier onset of improvement. Not all PILs warned about discontinuation syndrome and advice about alcohol was generally that it was prohibited. Almost half of PILs made no mention of St John's wort and its potential for interaction with the antidepressant. Two-thirds of PILs provided no information about the likely duration of treatment. PILs contained far more words judged to be negative rather than positive or neutral. Limitations: Data were extracted by a single researcher, although inter-rater agreement was high. Conclusions: Further guidance and tightening of the approval process for PILs are needed to ensure they are more standardised in content and contain more information that is wanted by and is useful to patients. © 2010 Elsevier B.V. All rights reserved.
1. Introduction Patients need information about proposed medication so that they can give informed consent to drug treatment. They may also need written information to refer to while they are taking the medication (Raynor et al., 2007). Depressed patients require information about antidepressants to dispel commonly held erroneous beliefs, for example that it is only necessary to take the antidepressants on those days when you feel depressed (Jorm et al., 2005). Verbal information is ⁎ Corresponding author. Consultant Psychiatrist, St Andrew's Healthcare, Billing Road, Northampton, NN1 5DG, UK. Tel.: + 44 1604 616 186. E-mail address: [email protected] (C. Haw). 0165-0327/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.jad.2010.06.034
often poorly retained by those who are stressed or anxious and in addition may be remembered incorrectly. Written information is better remembered than that given orally and may lead to better treatment adherence (Kessels, 2003). Psychiatric patients have access to written information about their medication from a variety of sources, including their psychiatrist and other healthcare professionals, the internet, media and health charities. One source of information that most patients have ready access to is the Patient Information Leaflet (PIL) that by law has to be issued with each medicine. According to a European Union Directive which came fully into effect in 1999 after a five year phasing in period, all medicines must be accompanied by a PIL setting out comprehensive and accessible information so
166
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
that patients can use their medicines safely and appropriately (Council Directive 92/27/EEC of 31 March 1992). This Directive was subsequently superseded by Directive 2001/83/EC of 6 November 2001 which was then partly amended by Directive 2004/27/EC of 31 March 2004. PILs are required to have a standard content in a specified order and to contain all the information in the Summary of Product Characteristics (SPC). Additional, educational information can be included provided it is non-promotional. PILs are required to be tested on users to ensure they are readable, understandable and comprehensive. The requirement for testing was contained in Directive 2004/ 27/EC of 31 March 2004 and came into force in 2005 for new medicines with a three year window in the UK for existing leaflets to be passed which ended in July 2008. Directive 2004/ 27/EC of 31 March 2004 stipulates [in article 63(2)] that leaflets must be clearly written and designed to be clear and understandable, enabling the users to act appropriately. The Medicines and Healthcare Products Regulatory Authority (MHRA) has written detailed guidance on communicating the risks and benefits in PILs (MHRA, 2005). All PILs have to be approved by the MHRA prior to distribution to the public. Despite efforts to produce a useful source of information, patients can find it difficult to extract essential information from PILs (Dickinson et al., 2001). Patients value information containing a good balance of benefit and harm information (Raynor et al., 2007). They want information about a medicine's indications, benefits, likely duration of treatment as well as side effects (Shrank et al., 2007). The aim of this study was to look at the uniformity, adequacy and balance of information contained in antidepressant Patient Information Leaflets available in the UK and hence how well current PILs inform patients about whether or not they should take their antidepressants.
2. Method We studied PILs for all antidepressant drugs listed in section 4.3 of the British National Formulary (BNF) No 58 (British Medical Association and Royal Pharmaceutical Society, 2009). The PILs available in the Electronic Medicines Compendium (eMC) at www.Medicines.org.uk on 1 January 2010 were investigated. Where an antidepressant was manufactured by more than one pharmaceutical company, we examined the PILs that each of the manufacturers had produced. If a manufacturer had produced different PILs for different formulations of the same drug (for example tablets and liquid), we studied the PIL for the most commonly used product. We investigated the following items which we as clinicians considered were important if patients are to be properly informed about their antidepressant medication: 1. Readability using Flesch reading ease and Flesch-Kincaid Grade Level (Microsoft Office Word, 2003). We also examined passive sentences. 2. Information given about the potential benefits of taking the antidepressant. 3. How the antidepressant works. 4. Speed of onset of action. 5. Likely duration of treatment and time to recover fully from depression.
6. How the risks of taking the antidepressant were described, including increase in suicidality after starting antidepressants, the dangers of abrupt withdrawal, advice about consuming alcohol, presentation of side effects and interactions with St John's wort. 7. The overall balance of information in PILs. The amount (word count) of information that was judged by us to be (i) positive or neutral or (ii) negative about the medicine in each PIL. Information was judged to be ‘positive’ or ‘neutral’ if it explained what the medicine is for, how it works, what benefits will taking it bring, how quickly it works, how long it is necessary to take it for, how and when to take the medicine, the recommended dosage and how to store the medicine. Information was regarded as ‘negative’ if it stated the risks of taking the antidepressant, including side effects, contra-indications, increased risk of suicide and self harm, not advised for certain patient groups, interactions, contra-indicating consumption of alcohol, possible detrimental effect upon ability to drive, effect of taking an overdose and the dangers of abrupt withdrawal. The number of positive and neutral and also negative words was calculated for each PIL by JS and expressed as a percentage of the total word count (information about ingredients and the manufacturer's address were excluded). CH independently rated the positive, neutral and negative words in five randomly selected PILs and compared her ratings with those of JS. There was agreement between the two raters over the allocation of 98.1% of words. 3. Results Forty-two PILs concerning 21 different antidepressants were listed in eMC. Using the classification given in the BNF No 58, 12 PILS were for tricyclics, 3 for MAOIs, 16 for SSRIs and 11 were for other antidepressants. In total, 23 different pharmaceutical companies were involved in their manufacture. Fluoxetine and venlafaxine were manufactured by 5 different companies; amitriptyline, trazodone, citalopram and paroxetine by 3; imipramine, fluvoxamine, sertraline and mirtazapine by two and the remaining 12 antidepressants by single companies. All but one PIL had undergone MHRA testing. 3.1. Readability statistics Median score on the Flesch–Kincaid Reading Ease Scale was 55.2 (inter-quartile range 52.4–58.5). On the Flesch– Kincaid Grade Level the median score was 7.3 (inter-quartile range 6.8–7.9). Use of passive sentences was uncommon in PILs (median 5%; inter-quartile range 2.3–7.0%). Thus, most PILs scored well on these statistics. However, it must be borne in mind that readability formulae only measure two aspects of readability, namely word length and sentence length and there are many other factors that influence readability. 3.2. Information about how antidepressants work and speed of onset of action Twenty-two (52.4%) PILS gave some sort of an explanation as to how the antidepressant worked, some mentioning a
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
specific mechanism of action, such as inhibition of the reuptake of serotonin or blockade of the action of monoamine oxidase. In some cases the explanation was detailed, for example a PIL for citalopram stated that ‘it increases the effect of the body's naturally occurring hormone, serotonin, by inhibiting its re-uptake into cells’. In other cases the explanation was more rudimentary, for example, a leaflet for lofepramine stated that it ‘alters the levels of chemicals in your brain to relieve symptoms of depression’. Twenty (47.6%) PILs gave no explanation at all (see Table 1). Thirtyeight (90.5%) PILs indicated that the antidepressant would
167
take between 2 and 4 weeks to take effect, with just two leaflets indicating that improvement might occur in a matter of days. Only one leaflet provided advice on what to do if no improvement had taken place after several weeks of treatment. 3.3. Presentation of information about side effects and suicidality Information about side effects was presented in a strikingly heterogeneous way. The majority (31; 73.8%) of
Table 1 Information presented in antidepressant Patient Information Leaflets (N = 42). Information presented in Patient Information Leaflet
N
(%)
How the antidepressant is thought to work No explanation Alters brain chemicals to relieve depression Increases serotonin and/or noradrenaline in the nervous system thus correcting chemical imbalance Blocks the action of enzyme monoamine oxidase in the brain which controls mood
20 7 13 2
(47.6) (16.7) (31.0) (4.8)
Speed of onset of action No information May feel an improvement in 2–3 days/take several days Two to four weeks If you don't feel better after 4 weeks go back to the doctor as you may need a higher dose
1 2 38 1
(2.4) (4.8) (90.5) (2.4)
Presentation of side effects Listed in decreasing order of seriousness Serious side effects first, then by frequency e.g. up to 1 in 10 people Serious side effects first, then by bodily system e.g. skin, cardiovascular Serious side effects first, then a random list Random list of side effects Listed in decreasing order of frequency e.g. up to 1 in 10 people Listed by bodily system e.g. skin, cardiovascular
14 10 5 2 5 4 2
(33.3) (23.8) (11.9) (4.8) (11.9) (9.5) (4.8)
Suicidality No information Suicidal thoughts just listed as a side effect. No information about suicidality in young people May get worse when start the antidepressant If become suicidal tell the doctor — it may be necessary to reduce the dose or stop the drug
1 1 38 2
(2.4) (2.4) (90.5) (4.8)
Discontinuation syndrome No information Warning not to stop antidepressant abruptly but no reason given or possible symptoms described Warning not to stop antidepressant abruptly and possible symptoms described
7 3 32
(16.7) (7.1) (76.2)
Advise about consuming alcohol No information Alcohol is contra-indicated Avoid alcohol as it may increase the effects of the antidepressant Avoid alcohol as the antidepressant may increase the effect of the alcohol Sensible to avoid alcohol but this antidepressant is not expected to interact with alcohol
5 25 4 6 2
(11.9) (59.5) (9.5) (14.3) (4.8)
St John's wort No information Tell your doctor or pharmacist if you are taking St John's wort St John's wort should not be taken at the same time as this antidepressant St John's wort can interact with the antidepressant St John's wort can lead to serotonergic syndrome, potentially life-threatening
20 8 9 2 3
(47.6) (19.0) (21.4) (4.8) (7.1)
Likely duration of treatment No information Don't stop the treatment if you feel better — the illness may come back Less than 3 months Usually at least 4–6 months after you feel better Maintenance treatment may be needed
27 2 1 11 1
(64.3) (4.8) (2.4) (26.2) (2.4)
168
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
PILs listed the most serious side effects, such as allergic reactions, first but there was then no standardised way of providing information about the remaining less serious side effects (see Table 1). These could be listed by severity, frequency, and bodily system or were in random order. Fourteen (33.3%) PILs made use of frequencies (e.g. less than one in ten people but more than one in a hundred people) to provide patients with information about how common particular side effects were. Forty (95.2%) PILs contained a prominent section warning patients that suicidal thoughts might get worse when starting the antidepressant and that this was particularly the case in young people. Advice was given as to what to do in the event of this occurrence. However, two leaflets contained no warning about antidepressants potentially causing an increase in suicidal thoughts or actions. This is likely to reflect the SPC. 3.4. Information about discontinuation syndrome Although all antidepressants can be associated with a discontinuation syndrome, a minority (7; 16.7%) of PILs contained no warning about discontinuation syndrome and a further three merely warned patients not to stop taking the antidepressant abruptly but did not give a reason or describe the symptoms of discontinuation syndrome (see Table 1). Thirty-two (76.2%) PILs gave adequate information about discontinuation syndrome. 3.5. Advice about consuming alcohol Five (11.9%) PILs did not provide any advice about drinking alcohol while taking antidepressants (see Table 1). Twenty-five (59.5%) contained a statement without any explanation that alcohol was contra-indicated while taking the antidepressant. Ten (23.8%) PILs advised against drinking alcohol stating that either alcohol would increase the effect of the antidepressant or vice-versa. Two PILs expressed caution about drinking alcohol while taking the antidepressant but said that no interaction was expected. 3.6. St John's wort While 20 (47.6%) PILs contained no information about use of St John's wort, the remaining leaflets indicated that concurrent usage with the prescribed antidepressant might be problematic (see Table 1). Warnings ranged from informing your doctor or pharmacist that you are taking St John's wort, to statements that St John's wort may interact with the antidepressant or even that St John's wort when combined with a serotonergic antidepressant can cause a potentially life-threatening serotonergic syndrome. 3.7. Likely duration of treatment No PIL provided information about how long it would take to recover from a depressive episode and only 12 (28.6%) provided information about how long patients were likely to have to take the antidepressant (see Table 1). One PIL stated that the likely duration of treatment was less than 3 months. Eleven (26.2%) leaflets stated the usual duration of treatment was at least 4–6 months after symptom resolution, while one
leaflet mentioned that maintenance treatment might be necessary. 3.8. Balance of information The median positive and neutral word count of the PILs was 785 words (inter-quartile range 652–967) and the median negative word count was 1393 (inter-quartile range 1061–1881). 4. Discussion Although all of the PILs scored well on the readability statistics, as Dickinson et al. (2001) have pointed out, such readability formulae are based on word and sentence length and therefore have their limitations. The same sentence written backwards would have the same readability score as one written forwards. Other factors, including layout, font size and style are also important. A limitation of the study is that it considered only the content of leaflets and not their format. The balance of information provided in the PILs we examined was predominantly negative and this is almost inevitable given the legal requirement to contain all the information in the SPC unless companies are going to include additional positive information in antidepressant PILs. Only half of the PILs studied gave any explanation as to how the antidepressant worked, with very few leaflets giving a detailed explanation of the mechanism of action. This is an important omission given our anecdotal impression of how frequently depressed patients ask about the mechanism of action of antidepressants. Patients need a balance of information about antidepressants so that they can decide whether or not to take (or continue to take) them. The regulator in their guidance to the pharmaceutical industry has stated that PILs need to include more benefit information (MHRA, 2005). Although most PILs stated the antidepressant would take between 2 and 4 weeks to work, it is now generally recognised that some improvement after starting antidepressants can occur within the first week of treatment (although a longer period of administration is needed before pronounced clinical effects occur) and regularly occurs within a few days but only a few leaflets acknowledged this early improvement. There is no evidence that particular antidepressants have a more rapid speed of onset than others, although amongst the SSRIs fluoxetine appears to have a slightly slower speed of onset of action (Edwards and Anderson, 1999). It is surprising that only one leaflet mentioned going back to the doctor to discuss a lack of progress in treatment and possible need for a dosage increment if no improvement had taken place after several weeks and none mentioned the possibility of change to a different antidepressant. The lack of information on the likely duration of treatment by two-thirds of PILs is unfortunate, since this is important information that patients need to know from the outset in order to try and prevent premature discontinuation. Of course, they should be told this by their psychiatrist but it is good for them to have this reiterated by the PIL. It is well established that the minimum duration of treatment after an episode of depression should be 4–6 months after complete recovery (Loonen et al., 1991;
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
Reimherr et al., 1998). Stopping antidepressants before this time carries a substantial risk of relapse. The PILs we studied presented antidepressant side effects in diverse ways and it was difficult to locate the same side effect in different leaflets. For this reason it would be better for patients if a standardised format were to be used for all PILs so as to guide readers. Listing by bodily system is helpful when a patient is experiencing a particular side effect since they should be able to easily locate the side effect on the PIL if they are listed in this way. On the other hand, listing the most serious side effects first followed by grouping according to frequency is an easily understood system which may be more helpful to patients when checking the leaflet to see what to expect (Carrigan et al., 2008). This latter method of listing side effects is that recommended by both the UK and EU regulators and this seems appropriate, given that serious side effects require immediate action. All antidepressants when taken for more than a few weeks can be associated with a discontinuation syndrome, although some are more strongly associated with it (for example venlafaxine) than others (for example fluoxetine) (Lejoyeux and Adès, 1997). It was, therefore, surprising that not all PILs contained a warning together with an explanation as to why antidepressants should not be stopped abruptly. Furthermore, not all PILs gave a description of symptoms associated with abrupt withdrawal. Most PILs advised patients to avoid alcohol but unhelpfully did not provide them with a reason for this. A minority did provide a rationale; either that alcohol would enhance the action of the antidepressant or vice-versa. Patients very commonly ask psychiatrists if it is alright to have a drink while taking antidepressants. Anecdotally clinicians are aware that statements that alcohol is contraindicated can have the effect of patients choosing not to take the antidepressant preferring instead to continue to drink alcohol for symptomatic relief. This is a difficult area but a few PILs were helpful in that although they advised against drinking alcohol while taking antidepressants they stated that an interaction was not expected. Of course, clinicians should be aware of patients who are abusing alcohol, comorbidity of alcohol use disorders being very common with depressive disorder (Grant and Harford, 1995). The presence of an alcohol disorder may influence the choice of antidepressant. It would be helpful if PILs could express caution when drinking alcohol while taking antidepressants because of the potential for interaction but also because alcohol itself can have a negative effect upon mood. Many patients with depression try St John's wort (Hypericum) before consulting a medical practitioner about conventional antidepressants. An estimated 1.5 million Britons bought St John's wort in 2009 (Daily Mail, 2010). St. John's wort should not be taken at the same time as other antidepressants since it has the potential to cause a serotonergic syndrome with serotonergic agents and can reduce the efficacy of others by its enzyme-inducing activities (Hammerness et al., 2003). It was, therefore, surprising that only half the PILs studied contained a warning about St John's wort and again, of those that did only a few explained that St John's wort should not be taken with a conventional antidepressant. The antidepressant PILs studied varied in quality. Although PILs are supposed to contain all the information in the
169
SPC in a specific order, in practice their content varied quite widely. The same antidepressant produced by several different companies could have PILs of diverse content, which could be very confusing to patients should they read the different leaflets. One reason for the differing content is, of course, that they may have different SPCs. With generic prescribing this scenario is likely to arise quite frequently. We suggest that further guidance on the content of PILs and also a tightening of the approval process are needed to ensure they are more informative to patients, and as well as containing a summary of the SPC they should contain important items such as those we have highlighted in this study, for example what to do if there is no improvement in symptoms after 4 weeks or so of treatment and the likely duration of treatment once depressive symptoms have resolved. Role of funding source None.
Conflict of interest None.
Acknowledgement Our thanks to Professor PJ Cowen for his helpful comments on the manuscript.
References British Medical Association & Royal Pharmaceutical Society, 2009. British National Formulary No 58. BMJ Publishing group, London. Carrigan, N., Raynor, D.K., Knapp, P., 2008. Adequacy of patient information on adverse effects: an assessment of patient information leaflets in the UK. Drug Safety 31, 305–312. Council Directive 92/27/EEC of 31 March 1992 on the labelling of medicinal products for human use and on package leaflets (OJ No. L 113 of 30.4.1992, p.8). Directive 2001/83/EC of 6 November 2001 on the community code relating to medicinal products for human use. (OJ No. L 311 of 28.11.2004, p67128). Directive 2004/27/EC of 31 March 2004 amending Directive 2001/83/EC on the community code relating to medicinal products for human use (OJ No. L136 of 30.4.2004, p34-57). Daily Mail, 2010. Alternative medicine sales soar as consumers shake off cynicism. April 6th. Dickinson, D., Raynor, D.K., Duman, M., 2001. Patient information leaflets for medicines: using consumer testing to determine the most effective design. Patient Education and Counselling 43, 147–159. Edwards, J.G., Anderson, I., 1999. Systematic review and guide to selective serotonin reuptake inhibitors. Drugs 57, 507–533. Grant, B.F., Harford, T.C., 1995. Comorbidity between DSM-IV alcohol use disorders and major depression: results of a national survey. Drug and Alcohol Dependence 39, 197–206. Hammerness, P., Basch, E., Ulbricht, C., Barrette, E.P., Foppa, I., Basch, S., Bent, S., Boon, H., Ernst, E., 2003. St John's Wort: a systematic review of adverse effects and drug interactions for the consultant psychiatrist. Psychosomatics 44, 271–282. Jorm, A.F., Christensen, H., Griffiths, K.M., 2005. Belief in the harmfulness of antidepressants: results from a national survey of the Australian public. Journal of Affective Disorders 88, 47–53. Kessels, R.P.C., 2003. Patients' memory for medical information. Journal of Royal Society of Medicine 96, 219–222. Lejoyeux, M., Adès, J., 1997. Antidepressant discontinuation. A review of the literature. Journal of Clinical Psychiatry 58 (Suppl 7), 11–16. Loonen, A.J., Peer, P.G., Zwanikken, G.J., 1991. Continuation and maintenance therapy with antidepressant agents. Meta-analysis of research. Pharmacy Weekly Science 23, 167–175. Medicines and Healthcare Products Regulatory Authority, 2005. Always read the leaflet: getting the best information with every medicine. Available at: http://mhra.gov.uk/Howweregulate/Medicines/Medicinesregulatorynews/CON09410. Microsoft Office Word, 2003. Microsoft Corporation, Washington.
170
C. Haw, J. Stubbs / Journal of Affective Disorders 128 (2011) 165–170
Raynor, D.K., Blenkinsopp, A., Knapp, P., Grime, J., Nicolson, D.J., Pollock, K., Dorer, G., Gilbody, S., Dickinson, D., Maule, A.J., Spoor, P., 2007. A systematic review of quantitative and qualitative research on the role and effectiveness of written information available to patients about individual medicines. Health Technology Assessment, Vol. 11. No. 5. Reimherr, F.W., Amsterdam, J.D., Quitkin, F.M., Rosenbaum, J.F., Fava, M., Zajecka, J., Beasley, C.M., Michelson, D., Roback, P., Sundell, K., 1998.
Optimal length of continuation therapy in depression: a prospective assessment during long-term fluoxetine treatment. American Journal of Psychiatry 155, 1247–1253. Shrank, W., Avorn, J., Rolon, C., Shekelle, P., 2007. Effect of content and format of prescription drug labels on readability, understanding, and medication use: a systematic review. Annals of Pharmacotherapy Available at: http://www.theannals.com/cgi/content/aph.1H582v1