A One-Day Survey of Characteristics of Off-Label Hospital Prescription of Psychotropic Drugs

• Abstract
• Introduction
• Methods
• Statistical analyses
• Results
• Patients characteristics
• Frequency and characteristics of off-label prescriptions
• Off-label prescriptions of 1st and 2nd generation antipsychotics
• Comparison of subjects with and without off-label prescription
• Discussion
• Acknowledgements
• References

Abstract

Introduction: The aim of this study was to explore the characteristics of off-label prescriptions of psychotropic drugs among patients hospitalised in psychiatry.

Methods: A one-day cross-sectional study was used to explore off-label prescribing in patients hospitalised in four wards of a state psychiatric hospital. Information regarding patients and treatment characteristics was collected from medical records. Off-label prescribing was defined according to the Summary of Product Characteristics.

Results: Seventy-five patients were included. Of the 261 prescriptions, more than one out of three (n=104, 39.8%) were off-label. Off-label prescribing was particularly high for anticonvulsants (97%), and to a lesser extent for mood stabilisers (33%) and anxiolytics (31%). The rate of off-label prescription was higher for 2nd compared to 1st generation antipsychotics. Most patients (84%) were prescribed at least one psychotropic drug outside of the licensed indication or recommended dosage. Patients with and without off-label prescription did not differ regarding history of prior admission, type of current admission (voluntary vs. compulsory) and diagnosis.

Conclusion: In spite of its frequency in routine practice, off-label prescribing should not be trivialised. Prescribers should be better informed about the potential legal consequences of off-label prescribing.

Introduction

The target population of a drug is theoretically restricted to the licensed indications specified in the marketing authorisation issued by the regulatory agencies [3]. Several factors unrelated to the pharmacological properties of a drug influence the range of licensed indications. Particularly, companies may be reluctant to apply for marketing authorisation for old products due to the cost of pre-marketing trials. Hence, drugs with comparable pharmacological profiles do not always have the same licensed indications. This may favour discrepancies between marketing authorisation and utilisation in naturalistic conditions [14] [20]. Indeed, prescribers consider mostly the pharmacological effect of a medication, and are not always fully aware of the exact list of its licensed indications.

Although off-label prescribing can be found in all the fields of medicine [1] [6] [13] [16], it may be particularly frequent in psychiatry. The stability and reliability of psychiatric diagnoses may be low even when standardised diagnostic criteria are used, for example, to differentiate between conditions such as mood disorder with psychotic features, schizoaffective disorder or schizophrenia with mood episodes. Furthermore, psychotropic medications have little diagnostic specificity, favouring their use out of the licensed diagnoses. Indeed, prior studies have reported that off-label prescribing was frequent in psychotropic treatments, particularly regarding antipsychotics and anticonvulsants, and mainly among those with recent marketing authorisation [2] [3] [4] [5] [10] [11]. However, little is known about the patient characteristics that may influence off-label prescribing. With respect to the potential legal consequences of off-label prescribing [9] [22], it is of interest to assess if prescribers tend to restrict this kind of practice in first-admitted subjects, for whom they have little or no information on tolerance to psychotropic medication, or in compulsory admitted subjects who are more likely to contest their treatment.

The aims of this study were to explore (i) the characteristics of off-label prescriptions for psychotropic drugs among patients hospitalised in psychiatry; (ii) whether subjects with off-label prescriptions differed from subjects without.

Methods

We conducted a one-day cross-sectional study in March 2005 in four wards of the state psychiatric hospital of Bordeaux. Three were “sectorised,” i.e., recruiting subjects on the basis of their residence in a catchment area irrespective of the diagnosis, and the fourth was a specialised unit recruiting patients presenting with bipolar disorders irrespective of the place of residence; all except one “sector” ward were parts of a university department. All full-time hospitalised patients were included if they were prescribed at least one psychotropic drug. The following information was collected in medical records and completed where necessary by interviewing the ward psychiatrists: (i) demographic characteristics; (ii) number of prior psychiatric hospitalisations, duration and type of current admission (voluntary or compulsory); (iii) ICD-10 diagnosis [23] given by the ward psychiatrists; (iv) psychotropic treatment: type of drug, indication and daily dose.

Two types of off-label prescriptions were defined according to the Summary of Product Characteristics (SPC): (i) off-label prescription with respect to indication; (ii) off-label prescription with respect to daily dose. The categorisation between label and off-label prescription was done by strictly applying the SPC: for example, divalproate was categorised as off-label if prescribed to a subject with mania with no documented history of “bipolar disorder refractory or with lack of tolerance to lithium”; amisulpride was categorised as off-label in psychotic mood disorder since the indications mentioned in SCP are “psychotic disorders, particularly acute or chronic schizophrenic disorders”.

Statistical analyses

Statistical analyses were performed using the STATA statistical package [17]. The non-parametric Mann-Whitney rank-sum test was used to explore the associations between off-label prescription and continuous variables at a 5% level of significance. Logistic regression giving odds ratio (OR) and 95% confidence intervals (95CI) were used to explore the associations between off-label prescription and categorical variables.

Results

Patients characteristics

Seventy-five patients were included in the study, of whom 41 (54. 7%) were males. The mean age was 38.3 years (SD 13.7; range: 17-82). One patient out of 10 (n=8, 10.7%) was hospitalised for the first time, and two out of three (n=48, 64%) were voluntary admitted. ICD-10 diagnoses were categorised into: (i) mood disorders (n=34, 45.3%): subjects with bipolar disorders (n=32) of whom most (n=27) were presenting with hypomanic, manic or mixed episode, and subjects with depressive episode (n=2); (ii) psychotic disorders (n=34, 45.3%): subjects with schizophrenia (n=20), schizo-affective disorders (n=11) or transient psychotic disorder (n=2, 2.7%); (iii) other diagnoses (n=7, 9.3%): alcohol-related disorders (n=2), personality disorder (n=2), adjustment disorder (n=1), conduct disorder (n=1) and mental retardation (n=1).

Frequency and characteristics of off-label prescriptions

A median number of four psychotropic drugs (interquartile range 3-4, range: 1-6) was prescribed per patient, corresponding to 29 different drugs ([Table 1]). The SPC and proportions of off-label prescription per drug are given in [Table 1]. Of the 261 prescriptions, more than one out of three (n=104, 39.8%) were off-label according to SPC: these percentages were particularly high for anticonvulsants, and to a lesser extent for mood stabilisers and anxiolytics.

Off-label prescriptions were mostly found with respect to indication (n=95, 92.2%) and in a few cases with respect to higher than recommended dosage (n=8, 7.8%). In 62.2% of the cases, drugs with off-label prescription regarding indication had pharmacological properties in accordance with the indication but had no marketing authorisation for it: for example, clonazepam commercialised for epilepsy, was used as an anxiolytic and sedative drug; oxcarbazepine, also commercialised for epilepsy, was used as a mood stabiliser in bipolar disorders. In 16.5% of the cases, off-label prescribing regarding indication was related to prescription for psychiatric diagnoses other than those listed in SPC: for example, amisulpride was prescribed for manic episodes. Lastly, in 12.6% of the cases, it was related to a lack of respect to the recommended strategy: for example, valpromide approved for bipolar disorder refractory to lithium or carbamazepine, was prescribed as a first line treatment. Off-label prescribing regarding dosage was only found for two anxiolytics (oxazepam and hydroxyzine).

Most patients (n=63, 84%) had at least one off-label drug prescribed: one off-label prescription n=32 (42.7%); 2 off-label prescriptions n=22 (29.3%); 3 off-label prescriptions n=9 (12%). The frequency of off-label prescribing was 56.7% after exclusion of the 22 patients with clonazepam as the only off-label drug prescribed.

Off-label prescribing was found in 90% (n=18) of subjects admitted in the non-university ward vs. 81.8% (n=45) subjects admitted to the 3 university wards, and in 81.3% (n=13) of subjects admitted to the specialised unit vs. 84.8% in the 3 “sector” wards.

Off-label prescriptions of 1st and 2nd generation antipsychotics

Frequency of off-label prescriptions was higher for 2nd compared to 1st generation antipsychotics (22% vs. 3.6%, OR=12.8, 95%CI 2.6-120.1, p=0.001).

Comparison of subjects with and without off-label prescription

A comparison of subjects with at least one off-label prescription to subjects without is given in [Table 2]. The two groups did not statistically differ regarding age and gender, history of prior psychiatric hospitalisation, type of current admission and diagnosis, and duration of current admission.

Owing to the fact that the high frequency of prescription of clonazepam might influence the findings, the analyses were repeated after categorising subjects with clonazepam as the only off-label drug prescribed in the group without off-label prescription. The two groups did not differ regarding age [with off-label median 38 years (interquartile range (IQR) 29-46] vs. without 34 years (IQR 26-47, z=-0.15, p=0.88), gender (off-label: 58.8% male vs. 51.2% female, OR=1.4, 95% CI 0.5-3.4, p=0.66), history of prior psychiatric hospitalisation (off-label: 56.7% prior admission vs. 37.5% 1st admission, OR=2.1 95% CI 0.5-9.9, p=0.31), type of current admission (off-label: 52.1% voluntary admission vs. 59.3% compulsory admission, OR=1.3, 95 %CI 0.5-3.5, p=0.60). Regarding diagnosis, off-label prescribing was more frequent in subjects with diagnoses other than mood or psychotic disorders, but the difference was not significant [off-label: 44.1% psychotic disorders (baseline) 58.8% mood disorder OR=1.8, 95% CI 0.7-4.7, p=0.23, 85.7% other diagnoses OR=7.6, 95% CI 0.8-70.2, p=0.07].

Off-label prescribing was found in 55.0% (n=11) of subjects admitted in the non-university ward vs. 54.5% (n=30) subjects admitted to the 3 university wards, and in 50.0% (n=8) of subjects admitted to the specialised unit vs. 55.9% (n=33) in the 3 “sector” wards.

Discussion

This study conducted in subjects hospitalised in psychiatry showed that more than one out of three prescriptions of psychotropic drugs was off-label with respect to SPC. The percentage of off-label prescriptions was particularly high for anticonvulsants. The rate of off-label prescription was higher for 2nd compared to 1st generation antipsychotics. Most patients were prescribed at least one psychotropic drug outside of the licensed indication or recommended dosage. Patients with and without off-label prescription did not differ regarding history of prior admission, type of current admission (voluntary vs. compulsory) and diagnosis.

These findings have to be interpreted in the light of methodological limitations. First, since patients were mainly recruited in the wards of a university department, generalisation of prescribing practice should be cautiously made [7], particularly due to an over-representation of refractory patients. However, two of the three university units were “sector” wards with a non-specialised recruitment and the frequency of off-label prescriptions was comparable between university and non-university wards, as well as between the “sector” wards and the specialised unit. Secondly, local prescribing practices have to be taken into account, explaining, for example, the large use of clonazepam in the present study. For obscure motives, this product is used as the first-rank anxiolytic/sedative/hypnotic drug at the state psychiatric hospital of Bordeaux, irrespective of diagnosis. Hence, the percentage of patients with off-label prescription was reduced from 84%-57% when clonazepam was not included in the off-label drugs. However, excluding this drug did not change the findings regarding the lack of association between off-label prescription and patient characteristics. Thirdly, we used a stringent definition of off-label prescription by strictly applying SCP while in real life practice the distinction between label and off-label prescription is complex due to possible interpretation of SCP [2]; for example “lack of tolerance to lithium” in the SCP of valproate may be interpreted more or less strictly according to the prescriber. Fourthly, the large proportion of subjects with bipolar disorders is partly explained by the specialised recruitment of the university department, and may have favoured an overestimation of off-label prescriptions for anticonvulsants and mood stabilisers. The low proportion of subjects with diagnoses other than bipolar disorder or non-affective psychotic disorder is also the direct consequence of a reduction in the number of beds in French state hospitals [19], restricting admission to subjects presenting with severe mental illness. However, the rate of off-label prescribing did not differ according to diagnosis, particularly between subjects with mood and psychotic disorders. Lastly, the low number of subjects may have induced a type II error explaining the lack of statistical differences between subjects with and without off-label prescribing. However, we have little arguments to suspect a selection bias of the patients included in the present survey, and that the prescribing practices would be dramatically different if a large sample size recruited over a longer period or in a higher number of wards had been examined. Furthermore, the small effect sizes of the associations suggest that increasing the sample size would not dramatically change the findings. The only exceptions were the higher frequencies of off-label prescriptions excluding clonazepam in subjects with previous admissions, or with diagnoses other than psychotic or mood disorder, which might be significant if a large sample size was examined.

Our findings are in accordance with those of prior studies showing that off-label prescribing is more the rule than the exception in psychiatric settings, particularly for anticonvulsants and antipsychotics [2] [3] [4] [5] [10] [11]. We also replicate the finding reported by Barbui and colleagues [3] that off-label prescription was more frequent for 2nd compared to 1st generation antipsychotics. Several factors may explain the fact that the SCP issued from randomised clinical trials are seldom respected in clinical practice [3]. The most obvious is that the range of legal indications varies widely from one drug to another of the same pharmacological class with similar clinical effects [3] [4] [15], depending or whether or not the drug company marketing the product has applied/obtained a marketing authorisation in the country. For example, the range of indications for 2nd generation antipsychotics is smaller than that of 1st generation, as illustrated in [Table 1]. The lag between marketing authorisation and guidelines issued from psychiatric associations also plays a role, since use of drugs without marketing authorisation may be recommended in treatment algorithms. For example, divalproate may be prescribed only as a second-line mood stabiliser in France, while it may be used as a first-line product according to international guidelines; lamotrigine or oxcarbazepine have no legal indications in the treatment of bipolar disorder in France while they are mentioned in international guidelines treatment algorithms [8] [18].

Few studies have explored whether off-label prescribing was influenced by patient characteristics. A study carried out on patients with a prescription of antipsychotic filed at local pharmacies showed that off-label indications were more frequent in unmarried or divorced compared to married or widowed patients, and in older subjects [21]. To our knowledge, no prior study has investigated the influence of patient characteristics with a potential legal impact in cases of adverse effects induced by an off-label prescribed drug. This point is of importance considering that off-label prescription may be associated with a higher frequency of adverse drug reactions [12], and that the prescriber may be considered as fully responsible for the legal consequences of side-effects for off-label prescriptions [9] [22]. In the present study, a striking finding was that compulsorily admitted patients were as likely as voluntary admitted patients to be prescribed off-label psychotropic medication. Considering that involuntary admitted patients are more prone to contest their treatment and to sue, this finding highlights the fact that prescribers pay little attention to the potential legal consequences of their prescription. To a lesser extent, the same comment applies to off-label prescribing in 1st admitted subjects, for whom prescribers have little information regarding tolerance to psychotropic drugs, implicating that a cautious attitude would be to first test this tolerance using drugs with marketing authorisation.

From a legal point of view, patients have to be informed that they are prescribed off-label medication. A study carried out in subjects with mild intellectual disability showed that patients are not informed of the off-label usage [10]. We did not explore this point in the present study, but we can hypothesise based on our knowledge of local practice that few if no patients with an off-label prescription included in the present study were informed of this.

Off-label prescribing is not necessarily harmful and may be of benefit to the patient. However, it should not be trivialised in spite of its frequency in routine practice, owing to its potential legal consequences, that it is beyond the scope of this paper to discuss. Prescribers should be better informed about the potential legal consequences of off-label prescribing, particularly the fact that they may be considered as responsible in cases of occurrence of an adverse event, and should be advised to restrict this practice to severe conditions for which no other labelled medication is available [9] [22].

Acknowledgements

We acknowledge the ward psychiatrists of Hôpital Charles Perrens for their collaboration. We thank Philip Robinson who kindly supervised the English of this paper.

References

• 1 ASCO . Reimbursement for cancer treatment: coverage of off-label drug indications.  J Clin Oncol. 2006;  24 3206-3208
  CrossrefPubMedSearch in Google Scholar
• 2 Assion HJ, Jungck C. Off-label prescribing in a German psychiatric hospital.  Pharmacopsychiatry. 2007;  40 30-36
  Thieme ConnectPubMedSearch in Google Scholar
• 3 Barbui C, Ciuna A, Nose M, Patten SB, Stegagno M, Burti L. et al . Off-label and non-classical prescriptions of antipsychotic agents in ordinary in-patient practice.  Acta Psychiatr Scand. 2004;  109 275-278
  CrossrefPubMedSearch in Google Scholar
• 4 Barbui C, Danese A, Guaiana G, Mapelli L, Miele L, Monzani E. et al . Prescribing second-generation antipsychotics and the evolving standard of care in Italy.  Pharmacopsychiatry. 2002;  35 239-243
  Thieme ConnectPubMedSearch in Google Scholar
• 5 Chen H, Deshpande AD, Jiang R, Martin BC. An epidemiological investigation of off-label anticonvulsant drug use in the Georgia Medicaid population.  Pharmacoepidemiol Drug Saf. 2005;  14 629-638
  CrossrefPubMedSearch in Google Scholar
• 6 Conroy S, Choonara I, Impicciatore P, Mohn A, Arnell H, Rane A. et al . Survey of unlicensed and off label drug use in paediatric wards in European countries. European network for drug investigation in children.  BMJ. 2000;  320 79-82
  CrossrefPubMedSearch in Google Scholar
• 7 Davids E, Bunk C, Specka M, Gastpar M. Psychotropic drug prescription in a psychiatric university hospital in Germany.  Prog Neuropsychopharmacol Biol Psychiatry. 2006;  30 1109-1116
  CrossrefPubMedSearch in Google Scholar
• 8 Goodwin GM, Young AH. The British Association for Psychopharmacology guidelines for treatment of bipolar disorder: a summary.  J Psychopharmacol. 2003;  17 3-6
  CrossrefPubMedSearch in Google Scholar
• 9 Gromb S, Maurain C, Carbonnel S. Drug prescriptions outside the marketing product license and its responsibilities.  Presse Med. 2000;  29 1053-1057
  PubMedSearch in Google Scholar
• 10 Haw C, Stubbs J. A survey of off-label prescribing for inpatients with mild intellectual disability and mental illness.  J Intellect Disabil Res. 2005;  49 858-864
  CrossrefPubMedSearch in Google Scholar
• 11 Haw C, Stubbs J. A survey of the off-label use of mood stabilizers in a large psychiatric hospital.  J Psychopharmacol. 2005;  19 402-407
  CrossrefPubMedSearch in Google Scholar
• 12 Horen B, Montastruc JL, Lapeyre-Mestre M. Adverse drug reactions and off-label drug use in paediatric outpatients.  Br J Clin Pharmacol. 2002;  54 665-670
  CrossrefPubMedSearch in Google Scholar
• 13 Li VW, Jaffe MP, Li WW, Haynes HA. Off-label dermatologic therapies. Usage, risks, and mechanisms.  Arch Dermatol. 1998;  134 1449-1454
  CrossrefPubMedSearch in Google Scholar
• 14 Martin K, Begaud B, Latry P, Miremont-Salame G, Fourrier A, Moore N. Differences between clinical trials and postmarketing use.  Br J Clin Pharmacol. 2004;  57 86-92
  CrossrefPubMedSearch in Google Scholar
• 15 Percudani M, Barbui C, Fortino I, Petrovich L. Epidemiology of first- and second-generation antipsychotic agents in Lombardy, Italy.  Pharmacopsychiatry. 2005;  38 128-131
  Thieme ConnectPubMedSearch in Google Scholar
• 16 Poole SG, Dooley MJ. Off-label prescribing in oncology.  Support Care Cancer. 2004;  12 302-305
  PubMedSearch in Google Scholar
• 17 StataCorp . Stata Statistical Software: Release 7.0., in College Station. TX: STATA corporation 2001
• 18 Suppes T, Dennehy EB, Hirschfeld RM, Altshuler LL, Bowden CL, Calabrese JR. et al . The Texas implementation of medication algorithms: update to the algorithms for treatment of bipolar I disorder.  J Clin Psychiatry. 2005;  66 870-886
  CrossrefPubMedSearch in Google Scholar
• 19 Verdoux H. The current state of adult mental health care in France.  Eur Arch Psychiatry Clin Neurosci. 2007;  257 64-70
  CrossrefPubMedSearch in Google Scholar
• 20 Verdoux H, Begaud B. Pharmaco-epidemiology: what do (and don’t) we know about utilisation and impact of psychotropic medications in real-life conditions?.  Br J Psychiatry. 2004;  185 93-94
  PubMedSearch in Google Scholar
• 21 Weiss E, Hummer M, Koller D, Pharmd, Ulmer H, Fleischhacker WW. Off-label use of antipsychotic drugs.  J Clin Psychopharmacol. 2000;  20 695-698
  CrossrefPubMedSearch in Google Scholar
• 22 Wetterling T. Legal decisions of german supreme courts about prescription of drugs for not registered indications (“off-label-use”).  Fortschr Neurol Psychiatr. 2004;  72 255-259
  PubMedSearch in Google Scholar
• 23 WHO . WHO coordinated multi-center study on the course and outcome of schizophrenia. Geneva: WHO 1992

Correspondence

Prof. H. Verdoux

Hôpital Charles Perrens

121 rue de la Béchade

33076 Bordeaux cedex

France

Phone: +33/556/561 73 2

Fax: +33/556/563 54 6

Email: helene.verdoux@u-bordeaux2.fr

References

• 1 ASCO . Reimbursement for cancer treatment: coverage of off-label drug indications.  J Clin Oncol. 2006;  24 3206-3208
  CrossrefPubMedSearch in Google Scholar
• 2 Assion HJ, Jungck C. Off-label prescribing in a German psychiatric hospital.  Pharmacopsychiatry. 2007;  40 30-36
  Thieme ConnectPubMedSearch in Google Scholar
• 3 Barbui C, Ciuna A, Nose M, Patten SB, Stegagno M, Burti L. et al . Off-label and non-classical prescriptions of antipsychotic agents in ordinary in-patient practice.  Acta Psychiatr Scand. 2004;  109 275-278
  CrossrefPubMedSearch in Google Scholar
• 4 Barbui C, Danese A, Guaiana G, Mapelli L, Miele L, Monzani E. et al . Prescribing second-generation antipsychotics and the evolving standard of care in Italy.  Pharmacopsychiatry. 2002;  35 239-243
  Thieme ConnectPubMedSearch in Google Scholar
• 5 Chen H, Deshpande AD, Jiang R, Martin BC. An epidemiological investigation of off-label anticonvulsant drug use in the Georgia Medicaid population.  Pharmacoepidemiol Drug Saf. 2005;  14 629-638
  CrossrefPubMedSearch in Google Scholar
• 6 Conroy S, Choonara I, Impicciatore P, Mohn A, Arnell H, Rane A. et al . Survey of unlicensed and off label drug use in paediatric wards in European countries. European network for drug investigation in children.  BMJ. 2000;  320 79-82
  CrossrefPubMedSearch in Google Scholar
• 7 Davids E, Bunk C, Specka M, Gastpar M. Psychotropic drug prescription in a psychiatric university hospital in Germany.  Prog Neuropsychopharmacol Biol Psychiatry. 2006;  30 1109-1116
  CrossrefPubMedSearch in Google Scholar
• 8 Goodwin GM, Young AH. The British Association for Psychopharmacology guidelines for treatment of bipolar disorder: a summary.  J Psychopharmacol. 2003;  17 3-6
  CrossrefPubMedSearch in Google Scholar
• 9 Gromb S, Maurain C, Carbonnel S. Drug prescriptions outside the marketing product license and its responsibilities.  Presse Med. 2000;  29 1053-1057
  PubMedSearch in Google Scholar
• 10 Haw C, Stubbs J. A survey of off-label prescribing for inpatients with mild intellectual disability and mental illness.  J Intellect Disabil Res. 2005;  49 858-864
  CrossrefPubMedSearch in Google Scholar
• 11 Haw C, Stubbs J. A survey of the off-label use of mood stabilizers in a large psychiatric hospital.  J Psychopharmacol. 2005;  19 402-407
  CrossrefPubMedSearch in Google Scholar
• 12 Horen B, Montastruc JL, Lapeyre-Mestre M. Adverse drug reactions and off-label drug use in paediatric outpatients.  Br J Clin Pharmacol. 2002;  54 665-670
  CrossrefPubMedSearch in Google Scholar
• 13 Li VW, Jaffe MP, Li WW, Haynes HA. Off-label dermatologic therapies. Usage, risks, and mechanisms.  Arch Dermatol. 1998;  134 1449-1454
  CrossrefPubMedSearch in Google Scholar
• 14 Martin K, Begaud B, Latry P, Miremont-Salame G, Fourrier A, Moore N. Differences between clinical trials and postmarketing use.  Br J Clin Pharmacol. 2004;  57 86-92
  CrossrefPubMedSearch in Google Scholar
• 15 Percudani M, Barbui C, Fortino I, Petrovich L. Epidemiology of first- and second-generation antipsychotic agents in Lombardy, Italy.  Pharmacopsychiatry. 2005;  38 128-131
  Thieme ConnectPubMedSearch in Google Scholar
• 16 Poole SG, Dooley MJ. Off-label prescribing in oncology.  Support Care Cancer. 2004;  12 302-305
  PubMedSearch in Google Scholar
• 17 StataCorp . Stata Statistical Software: Release 7.0., in College Station. TX: STATA corporation 2001
• 18 Suppes T, Dennehy EB, Hirschfeld RM, Altshuler LL, Bowden CL, Calabrese JR. et al . The Texas implementation of medication algorithms: update to the algorithms for treatment of bipolar I disorder.  J Clin Psychiatry. 2005;  66 870-886
  CrossrefPubMedSearch in Google Scholar
• 19 Verdoux H. The current state of adult mental health care in France.  Eur Arch Psychiatry Clin Neurosci. 2007;  257 64-70
  CrossrefPubMedSearch in Google Scholar
• 20 Verdoux H, Begaud B. Pharmaco-epidemiology: what do (and don’t) we know about utilisation and impact of psychotropic medications in real-life conditions?.  Br J Psychiatry. 2004;  185 93-94
  PubMedSearch in Google Scholar
• 21 Weiss E, Hummer M, Koller D, Pharmd, Ulmer H, Fleischhacker WW. Off-label use of antipsychotic drugs.  J Clin Psychopharmacol. 2000;  20 695-698
  CrossrefPubMedSearch in Google Scholar
• 22 Wetterling T. Legal decisions of german supreme courts about prescription of drugs for not registered indications (“off-label-use”).  Fortschr Neurol Psychiatr. 2004;  72 255-259
  PubMedSearch in Google Scholar
• 23 WHO . WHO coordinated multi-center study on the course and outcome of schizophrenia. Geneva: WHO 1992

Correspondence

Prof. H. Verdoux

Hôpital Charles Perrens

121 rue de la Béchade

33076 Bordeaux cedex

France

Phone: +33/556/561 73 2

Fax: +33/556/563 54 6

Email: helene.verdoux@u-bordeaux2.fr