Author's Accepted Manuscript

Baclofen For Alcohol Dependence: Relationships Between Baclofen And Alcohol Dosing And The Occurrence Of Major Sedation Benjamin Rolland, Julien Labreuche, Alain Duhamel, Sylvie Deheul, Sophie Gautier, Marine Auffret, Baptiste Pignon, Thomas Valin, Régis Bordet, Olivier Cottencin www.elsevier.com/locate/euroneuro

PII: DOI: Reference:

S0924-977X(15)00164-9 http://dx.doi.org/10.1016/j.euroneuro.2015.05.008 NEUPSY11040

To appear in:

European Neuropsychopharmacology

Cite this article as: Benjamin Rolland, Julien Labreuche, Alain Duhamel, Sylvie Deheul, Sophie Gautier, Marine Auffret, Baptiste Pignon, Thomas Valin, Régis Bordet, Olivier Cottencin, Baclofen For Alcohol Dependence: Relationships Between Baclofen And Alcohol Dosing And The Occurrence Of Major Sedation, European Neuropsychopharmacology, http://dx.doi.org/10.1016/j.euroneuro.2015.05.008 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

TITLE:

BACLOFEN FOR ALCOHOL DEPENDENCE: RELATIONSHIPS BETWEEN BACLOFEN AND ALCOHOL DOSING AND THE OCCURRENCE OF MAJOR SEDATION

_ AUTHORS : Benjamin ROLLAND1,2,*; Julien LABREUCHE3; Alain DUHAMEL3,4; Sylvie DEHEUL5; Sophie GAUTIER2,6; Marine AUFFRET6; Baptiste PIGNON1; Thomas VALIN7; Régis BORDET2,5,6; Olivier COTTENCIN1,8

AFFILIATIONS : 1. Department of Addiction Medicine, CHU Lille, F-59037 Lille, France 2. Department of Pharmacology, INSERM U 1171, Univ Lille, F-59045 Lille, France 3. Department of Biostatistics, CHU Lille, F-59037 Lille, France 4. CERIM, EA 2694, Univ Lille, F-59045 Lille, France 5. CEIP, CHU Lille, F-59037 Lille, France 6. Department of Pharmacovigilance, F-59037 Lille, France 7. Department of Psychiatry, CH Douai 59507, Douai, France 8. SCALAB UMR CNRS 9193, Univ Lille, F-59045 Lille, France

CORRESPONDING AUTHOR : * Dr Benjamin ROLLAND, Service d’Addictologie, Hôpital Fontan2, CS 70001, CHRU de Lille, 59037 Lille cedex, France ; phone : +33 320 445 838 ; email : [email protected]

Number of Words: 2533 (+ Abstract =250) Number of Tables: 2 Number of Figures: 1 MeSH Keywords: [Baclofen] ; [Alcoholism] ; [Drug-Related Side Effects and Adverse Reactions] ; [GABA Agonists]; [Ethanol]; [Alcohol-related Disorders]

BACLOFEN FOR ALCOHOL DEPENDENCE: RELATIONSHIPS BETWEEN BACLOFEN AND ALCOHOL DOSING AND THE OCCURRENCE OF MAJOR SEDATION

ABSTRACT High-dose baclofen, i.e., 300 mg/d or more, has recently emerged as a strategy for treating alcohol dependence. The impact that the co-exposure of large amounts of alcohol and baclofen has on sedation is unclear. In a prospective cohort of 253 subjects with alcohol dependence, we collected daily alcohol and baclofen doses across the first year of baclofen treatment and the monthly maximum subjective sedation experienced by each patient (0-10 visual analog scale). For each patient-month, we determined the average weekly alcohol consumption (AWAC; standard-drinks/week) and the maximum daily dose of baclofen (DDB; mg/d). The occurrence of an episode of major sedation (EMS) during a patient-month was defined as a sedation score ≥7. The relationship between the EMS occurrence and the concurrent AWAC and DDB was investigated using a generalized estimating equation model. In total, 1528 patient-months were compiled (70 with an EMS). Univariate analyses demonstrated that the rate of patient-month to EMS increased gradually with AWAC (p35. There was also a significant gradual risk for EMS associated with DDB (35 of 1.22 (95%CI, 1.08-1.38) versus 1.11 (95%CI, 0.96-1.29) in AWAC=1-35, and 0.95 (95%CI, 0.76-1.19) in AWAC=0. The level of sedation observed in patients using baclofen for alcohol dependence appears to directly depend on the immediate doses of both the baclofen and the alcohol.

I.

INTRODUCTION

Baclofen is an agonist of gamma amino-butyric acid type B (GABA-B) receptors and has been used for spasticity since the 1970s (Sachais et al., 1977). Approved oral doses of baclofen never exceed 120 mg/d. In France, the use of high-dose baclofen, i.e., up to 300 mg/d or sometimes more, has recently become a popular choice for treating alcohol dependence (Dupouy et al., 2014). Baclofen is essentially used to control craving for alcohol, thereby reducing consumption (Rolland et al., 2012). The current regulatory situation of this prescribing practice is complex. Although baclofen remains off-label for alcohol dependence, its use was officially framed by the French health authorities in 2014 (ANSM, 2014) in anticipation of the upcoming results of two efficacy clinical trials conducted in France (ClinicalTrials.gov- a,b).

Several safety concerns have been associated with using baclofen to treat alcohol dependence. Notably, the toxicity profile of baclofen is known to include dose-related sedative effects that can induce deep coma and respiratory depression (Kiel et al., 2014). These are also the main toxicological risk factors of alcohol, which raises concerns about prescribing high-dose baclofen to patients who may potentially drink large amounts of alcohol (Rolland et al., 2012). Previous studies have not assessed whether the simultaneous exposure to baclofen and alcohol elicits additional or even potentiating effects on patient sedation levels or the experience of related side effects. Additionally, the forthcoming results of the efficacy trials will only partially resolve this issue, as no systematic assessments of either patient sedation levels or the experience of related side effects were performed in these studies. Here, we conducted a prospective cohort study of subjects with alcohol dependence that began baclofen treatment. From these patients, we collected the daily reported alcohol consumption, dose of baclofen taken, and monthly maximum subjective level of sedation throughout the first year of treatment. Our objective was to determine if episodes of major sedation were related to concurrent use of alcohol and baclofen.

II.

EXPERIMENTAL PROCEDURES

II.1

Subjects

In total, 253 subjects were included between January 2011 and June 2013 among 24 centers. All patients met the DSM-IV-Tr criteria for alcohol dependence, and all were just starting on baclofen treatment with the aim of reducing drinking, ideally to reach protracted controlled drinking or even abstinence, as defined in the French recommendations (Société Française d’Alcoologie, 2015). Exclusionary criteria included patients who depended on drugs other than alcohol (except for tobacco dependence) and patients with current DSM-IV-Tr Axis-I diagnoses.

II.2

Data collection

The data collected pertained to the first year of baclofen treatment. During this period, patients were provided clinical self-report sheets to record their daily consumption of alcohol in standarddrinks (sd, i.e., 10 g of alcohol) and their daily dose of baclofen (in mg) (Rolland et al., 2010).

Patients met with their baclofen prescribers for a monthly medical consultation during which the clinical data were collected. If patients neglected to log their daily consumption of alcohol and dose of baclofen on the self-report sheets, the missing data were completed retrospectively with the prescriber according to the validated model of the alcohol timeline follow-back (Sobell et al., 1988). During the visit, patients also had to report the maximum subjective sedation that they had experienced across the previous month using a visual analog scale (VAS) rated from 0 to 10. For cases in which baclofen treatment was discontinued, the prescriber noted the date of baclofen cessation and the reason: 1) inefficacy, 2) adverse event (and type), or 3: other cause.

Based on this systematic clinical recording, we constituted a database of patient-months, i.e., each month of follow-up for every subject throughout the first year of treatment. For each patient-

month, we determined whether the maximum subjective sedation score reported by the patient across the previous month was rated as 7 or higher. In this case, as has been previously defined elsewhere (Rolland et al., 2015), we concluded that an episode of major sedation (EMS) occurred during that patient-month. If the maximum subjective sedation score was below 7, we concluded that no EMS occurred during the patient-month (noEMS). For each patient-month, we also defined the daily dose of baclofen at the time of the consultation (DDB; mg/d) and the average weekly alcohol consumption across the previous patient-month (AWAC; sd/week). The initial AWAC, i.e., the AWAC across the month preceding the introduction of baclofen, was the only variable not determined by prospective self-reporting but was instead evaluated with the patient during the initial medical consultation.

AWAC and DDB were categorized into three pre-defined clinical subgroups. AWACs were divided into three dose categories: 1) AWAC=0 for patient-months with a reported lack of alcohol use; 2) AWAC=1-35 for patient-months with an AWAC between 1 and 35 sd/week; and 3) AWAC>35 for patients with an AWAC over 35 sd/week. The cut-off of 35 sd/week corresponds to an averaged definition of heavy-drinking according to the NIAAA, i.e., 5 sd or more per drinking day (NIAAA, 2015). Similarly, DDBs of the different patient-months were divided into three dose categories: 1) DDB200 mg/d.

II.3

Statistical Analyses

Data are expressed as median values (interquartile range, IQR) for quantitative variables or as number and percentages for qualitative variables. We performed a visit-level analysis to assess the association between the occurrence of EMS and the concurrent scores of DDB and AWAC by evaluating all available consultations during the 12-month follow-up period. The EMS rate was expressed as the number of episodes per patient-month and, in bivariate analysis, was compared

between the pre-specified DDB and AWAC subgroups using a generalized estimating equation (GEE) approach (Zeger et al., 1988). A GEE model for binary data with a logit link function and a first-order autoregressive correlation structure was used to account for the correlation between repeated measures over time. The choice of the covariance matrix was made using the Akaike Information Criterion (AIC) information. Odds ratios (OR) of EMS were calculated for each DDB and AWAC subgroup using the lowest level as a reference. Because ORs increased gradually with increasing DDB and AWAC levels, we also computed ORs per 20-unit increase in DDC and AWAC levels. We assessed the interaction between DDB and AWAC by including into a multivariable GEE model the corresponding multiplicative term (both DDB and AWAC introduced as continuous variable). Because we found a significant interaction between DDB and AWAC, we calculated the OR of EMS per 20-unit increase in DDC according to AWAC subgroups. Statistical testing was conducted using the two-tailed α level of 0.05. Data were analyzed using the SAS software package, release 9.3 (SAS Institute, Cary, NC).

II.4

Ethical considerations

The procedures for collecting, de-identifying, and analyzing data were previously submitted to and approved by a national ethics committee (Avis CCTIRS #13.290). All of the participants signed a written consent form before participating in the research.

III.

RESULTS

III.1

Descriptive results

Initially, 253 patients received a first prescription of baclofen. Of these, 205 (81%) patients attended their month 1 consultation and were thus included in the further analyses. Subjects included 149 males (73%) with a median age of 47 years (IQR, 41-57) and a median AWAC of 55 (IQR, 11-105) sd/weeks prior to treatment.

Of the 205 included patients, 121 (59.0%) discontinued baclofen before the end of the 12th month of treatment. The reasons, in order of frequency, included the following: loss of follow-up (n=71; 59.7%); experience of adverse event (n=26, 21.5%), including 8 EMSs; inefficacy (n=19, 15.6%); and other causes (n=5, 4.1%). Including all available consultations, a total of 1528 patient-months were compiled, which comprised 1443 noEMS patient-months, 70 EMS patient-months, and 15 undocumented patient-months. The month-by-month reporting of the number of patients in ongoing treatment, the number lost to follow-up, the number of treatment cessations, and the median values for DDB, AWAC, and rate of EMS, can be found in Table 1. According to patient-level analysis, the 70 EMSs were documented in 40 patients (20.2%), of whom 24 reported 1 EMS, 9 reported 2 EMSs, 3 reported 3 EMSs, 3 reported 4 EMSs, and 1 patient reported 7 EMSs.

III.2

Association between the occurrence of EMS and the concurrent DDB and AWAC

As shown in Table 2, the rate of EMS per patient-month increased gradually with increasing amounts of AWAC (p for trend across AWAC categories 35. A 20 sd/week increase in AWAC was found to be associated with an OR of 1.14 (95%CI, 1.04-1.26) for experiencing EMS.

A significant gradually increasing risk for EMS with DDP was also found (Table 2, p for trend across DDB categories 35 sd/week, as compared to 1.11 (95%CI, 0.96-1.29) in patient-months with AWAC=1-35 and 0.95 (95%CI, 0.76-1.19) in patient-months with no alcohol consumption.

IV.

DISCUSSION

The aim of the present study was to determine whether the experience of EMS among patients treated with baclofen for alcohol dependence was related to the quantity of alcohol and baclofen consumed. Overall, we found that both baclofen and alcohol consumption are directly associated with the occurrence of EMS. It appeared that the amount of alcohol consumed was more globally associated with EMS occurrence than the dose of baclofen taken, even if the maximum EMS risk occurred in cases in which subjects concurrently used both high-dose baclofen and large quantities of alcohol. However, taking high doses of baclofen with low or no drinking produced relatively few associations with the occurrence of EMS. The intricacies of these relationships highlights the complexity of assessing the exact level of causality that baclofen has on the frequency of numerous types of adverse neuropsychiatric events that may occur under the context of heavy drinking. To our knowledge, this study is the first to investigate the interaction between alcohol consumption and high-dose baclofen treatment in a clinical cohort of patients. Indeed, only a single previous study has even addressed the combined sedative effects of alcohol and baclofen (Evans and Bisaga, 2009). This study also found that baclofen increased sedation in a dose-dependent manner; however, its design was radically different from the one employed here, as it was a cross-sectional toxicological study on heavy drinkers without a criteria for dependence. Furthermore, only a single dose of baclofen was delivered to these patients, and the maximum dose administered was 80 mg. Although the results of the previously mentioned study are consistent with our own, our study provides additional data that are more relevant to clinical practice by directly addressing the specific issue of prescribing high-dose baclofen for alcohol dependence.

Our findings included several limitations that should be addressed. First, almost all of the data were provided using subjective self-report assessments, which may raise some scientific concerns. However, patient-reported outcome is now considered to be very useful scientific information for health assessment (Deshpande et al., 2011). In the specific field of alcohol treatment, the relevance

of self-report data may specifically be addressed because some patients can minimize their actual alcohol use. However, treatment-seeking individuals with alcohol dependence are usually considered very reliable with regards to their self-reported alcohol use (Bertholet et al., 2014). In our study, all patients were treatment-seeking individuals with alcohol-dependence, which let assume that the soundness of our data is sufficient. Another potential limitation of our study is the high rate of patients lost to follow-up, although it should be noted that follow-up attrition is a long-known issue in studies using patients with substance use disorders (Mackenzie et al., 1987). It is unknown whether some of our included patients chose to stop treatment because of an EMS, which would have implications on the statistical findings provided herein. Moreover, previous research has shown that the patients suffering from alcohol dependence who are lost to follow-up are statistically more likely to exhibit heavier drinking patterns (Hedden et al., 2009). Consequently, it is our opinion that the missing data due to follow-up attrition have primarily reduced the significance of our findings, and could hardly be responsible for generating erroneous statistical associations. Another possible limitation of the study is the methodology used to evaluate sedation. Choosing to use a VAS for these evaluations produces only purely subjective assessments by the patient. Some authors have proposed using objective measures of sedation, e.g., actigraphy (Lafleur 2005). Simple and wearable actigraphic devices could be useful in the future to complete subjective feedbacks, with limited cumbersomeness for patients (Case et al., 2015). Regardless, the occurrence of EMS was associated with 8 out of 26 cases of treatment cessation due to the experience of an adverse event. Even if we did not analyze the type of adverse event in our study, this rate reflects that actual safety issues may directly impact treatment outcomes. Future studies should more precisely evaluate whether serious adverse events induced by sedation, i.e., hospitalization, life-threatening incidents, or even death (FDA, 2015), are related to the concurrent use of alcohol and high-dose baclofen. It should also be noted that we did not include data on whether any additional medications were being used by our cohort of patients (e.g., benzodiazepines). Similarly to what was mentioned above, it is our opinion

that the inclusion of such information would have merely reduced the significance of our findings rather than generated false associations.

The off-label use of high-dose baclofen to treat alcohol dependence is currently framed only in France, where it has become a common treatment for alcohol dependence (Rolland et al., 2014). Furthermore, prescriptions of baclofen have been clearly increasing in several European countries over the past few years (Dupouy et al., 2013). A very recent randomized clinical trial conducted in Germany found a significant effect of high dose baclofen on abstinence maintenance (Müller et al., 2015). If these positive efficacy results are confirmed by the two French trials that have just ended, it is very likely that an official approval field will be filed at the European Medicines Agency. In this case, the use of high-dose baclofen could quickly spread across Europe in the near future, just as it has already spread across France. If this scenario were to occur, the pharmacoepidemiological tolerability data that we have provided herein would be of great importance in delineating the best safety conditions for baclofen use and monitoring. The main message that should be delivered to both prescribers and patients is that baclofen and alcohol dosing potentiate themselves to induce EMSs. However, combined high amounts of alcohol and low doses of baclofen appear to more likely induce EMSs than high doses of baclofen with no or low amounts of alcohol.

AUTHOR DISCLOSURES Benjamin ROLLAND was the main investigator of a study on high dose baclofen funded by Ethypharm, and he received fees for lectures or consultancy from Lundbeck and Ethypharm. Regis BORDET has received fees for lectures from Lundbeck Olivier COTTENCIN was an associate investigator in a clinical trial funded by Ethypharm. He has received sponsorship to attend scientific meetings from Lundbeck, BMS Otsuka, Janssen Cilag, Reckitt Benckiser, Pfizer, and Bouchara Recordati. Other authors declare no conflict of interest related with this article.

CONTRIBUTORS Benjamin ROLLAND, Régis BORDET and Olivier COTTENCIN designed the study. Benjamin ROLLAND, Sylvie DEHEUL, Thomas VALIN, and Baptiste PIGNON recruited the participants and acquired the data. Julien LABREUCHE and Alain DUHAMEL performed the statistical analyses. Benjamin ROLLAND and Julien LABREUCHE wrote the first draft of the manuscript. All authors have read and approved the final manuscript.

REFERENCES - Agence Française du Médicament et des produits de santé (ANSM), 2014. [Recommandation temporaire d’utilisation (RTU) pour le baclofène - Point d’information] Temporary recommendation for use (TRU) for baclofen. Available online at: http://ansm.sante.fr/S-informer/Actualite/Unerecommandation-temporaire-d-utilisation-RTU-est-accordee-pour-le-baclofene-Point-dinformation (accessed March 10, 2015). - Bertholet N, Winter MR, Cheng DM, Samet JH, Saitz R, 2014. How accurate are blood (or breath) tests for identifying self-reported heavy drinking among people with alcohol dependence? Alcohol. Alcohol. 49(4):423-9. - Case MA, Burwick HA, Volpp KG, Patel MS, 2015. Accuracy of smartphone applications and wearable devices for tracking physical activity data. JAMA. 313(6):625-6. - ClinicalTrials.gov-a. Baclofen for the Treatment of Alcohol Drinkers. http://www.clinicaltrials.gov/ct2/show/NCT01604330?term=bacloville&rank=1 (accessed March 10, 2015). - ClinicalTrials.gov-b. Efficacy and Safety of Baclofen for Maintenance of Abstinence in Alcohol Dependent Patients. http://www.clinicaltrials.gov/ct2/show/NCT01738282?term=alpadir&rank=1 (accessed March 10, 2015). - Deshpande PR, Rajan S, Sudeepthi BL, Abdul Nazir CP, 2011. Patient-reported outcomes: A new era in clinical research. Perspect. Clin. Res. 2(4):137-44. - Dupouy J, Fournier J-P, Jouanjus E, Palmaro A, Poutrain J-C, Oustric S, Lapeyre-Mestre M, 2014. Baclofen for alcohol dependence in France: Incidence of treated patients and prescription patternsA cohort study. Eur. Neuropsychopharmacol. 24(2):192-9. - Evans SM, Bisaga A, 2009. Acute interaction of baclofen in combination with alcohol in heavy social drinkers. Alcohol. Clin. Exp. Res. 33(1):19-30. - Food and Drug Administration (FDA), 2015. What is a Serious Adverse Event? Available at: http://www.fda.gov/safety/medwatch/howtoreport/ucm053087.htm (accessed March 10, 2015). - Hedden SL, Woolson RF, Carter RE, Palesch Y, Upadhyaya HP, Malcolm RJ, 2009. The impact of loss to follow-up on hypothesis tests of the treatment effect for several statistical methods in substance abuse clinical trials. J. Subst. Abuse. Treat. 37(1):54-63. - Kiel LB, Hoegberg LCG, Jansen T, Petersen JA, Dalhoff KP, 2014. A nationwide register-based survey of baclofen toxicity. Basic. Clin. Pharmacol. Toxicol., doi: 10.1111/bcpt.12344. - Lafleur KJ, 2005. Will adequate sedation assessment include the use of actigraphy in the future? Am. J. Crit. Care. 14(1):61-3. -

Müller CA, Geisel O, Pelz P, Higl V, Krüger J, Stickel A, Beck A, Wernecke KD, Hellweg R, Heinz A, 2015. Eur. Neuropsychopharmacol. doi:10.1016/j.euroneuro.2015.04.002.

- Mackenzie A, Funderburk FR, Allen RP, Stefan RL, 1987. The characteristics of alcoholics frequently lost to follow-up. J. Stud. Alcohol. 48(2):119-23.

- National Institute on Alcohol Abuse and Alcoholism (NIAAA). What is « at-risk » or « heavy » drinking? Available at: http://rethinkingdrinking.niaaa.nih.gov/IsYourDrinkingPatternRisky/WhatsAtRiskOrHeavyDrinking.a sp (accessed March 10, 2015). - Rolland B, Deheul S, Danel T, Bordet R, Cottencin O, 2010. [A System of Prescriptions without Drug Approval: Example of Baclofen.]. Therapie. 65(6):511-8. - Rolland B, Bordet R, Cottencin O, 2012. Alcohol-dependence: the current French craze for baclofen. Addiction. 107(4):848-9. - Rolland B, Paille F, Fleury B, Cottencin O, Benyamina A, Aubin HJ, 2014. Off-label baclofen prescribing practices among French alcohol specialists: results of a national online survey. PLoS One. 9(6):e98062. - Rolland B, Valin T, Langlois C, Auffret M, Gautier S, Deheul S, Danel T, Bordet R, Cottencin O, 2015. Safety and drinking outcomes among patients with comorbid alcohol dependence and borderline personality disorder treated with high-dose baclofen: a comparative cohort study. Int. Clin. Psychopharmacol. 30(1):49-53. - Sachais BA, Logue JN, Carey MS, 1977. Baclofen, a new antispastic drug. A controlled, multicenter trial in patients with multiple sclerosis. Arch. Neurol. 34(7):422-8. - Sobell LC, Sobell MB, Leo GI, Cancilla A, 1988. Reliability of a timeline method: assessing normal drinkers’ reports of recent drinking and a comparative evaluation across several populations. Br. J. Addict. 83(4):393-402. - Société Française d’Alcoologie, 2015. Good Practice Recommendation. Alcohol misuse: screening, diagnosis, and treatment. Available at: http://www.sfalcoologie.asso.fr/download/SFA-GPRAlcoholMisuse.pdf?PHPSESSID=5b26fb3d20947af67a5f63e5f6689df8 (accessed March 10, 2015). - Zeger SL, Liang KY, Albert PS, 1988. Models for longitudinal data: a generalized estimating equation approach. Biometrics. 44(4):1049-60.

Mont h

Numbe r of patient s

Cumulate d number of loss to follow-up (%)

Cumulative number of medical discontinuatio n (%)

Cumulate d number of patientmonths

DDB (mg/d; media n [IQR])

Previousmonth AWAC (sd/week ; median [IQR])

Previous -month EMS (n; %)

0

253

_

_

_

0

-

_

1

205

48 (19.0)

0 (0.0)

205

38 (3840)

21 (0-60)

9 (4.4%)

2

186

60 (23.7)

7 (2.8)

391

75 (6090)

13 (0-42)

6 (3.2%)

3

168

72 (28.5)

13 (5.1)

559

88 (70120)

8 (0-40)

12 (7.2%)

4

147

83 (32.8)

23 (9.1)

706

100 (70142)

6 (0-30)

8 (5.5%)

5

134

92 (36.4)

27 (10.7)

840

100 (60150)

9 (0-28)

6 (4.5%)

6

120

100 (39.5)

33 (13.0)

960

104 (75150)

7 (0-23)

8 (6.8%)

7

108

106 (41.9)

39 (15.4)

1068

120 (80150)

5 (0-27)

5 (4.7%)

8

105

107 (42.3)

41 (16.2)

1173

120 (75155)

3 (0-23)

4 (3.9%)

9

94

115 (45.5)

44 (17.4)

1267

125 (80160)

8 (0-28)

3 (3.2%)

10

92

116 (45.8)

45 (17.8)

1359

135 (75160)

4 (0-20)

1 (1.1%)

11

85

119 (47.0)

49 (19.4)

1444

135 (80160)

3 (0-20)

4 (4.8%)

12

84

119 (47.0)

50 (19.8)

1528

120 (75165)

2 (0-21)

4 (4.8%)

Table 1. Longitudinal description of data over the 12-month follow-up period For each month of follow-up, the following data are provided: the number of patients undergoing treatment (n); the number of patients lost to follow-up; the number of treatment cessations based on medical decisions over the preceding month (reasons for treatment cessation are described in the text); the cumulated number of patient-months; the daily dose of baclofen (DDB); the average weekly alcohol consumption (AWAC) across the preceding month; the maximum sedation score (0 to 10) during the previous month based on a visual analog scale; and the occurrence of EMS (i.e., average maximum sedation of 7 or more on the visual analog scale) during the previous month.

Table 2. Univariate association between the occurrence of major sedation and the amount of alcohol or the dose of baclofen consumed Number Patientmonths

EMS, n (%)

OR (95%CI)

P

0

556

5 (0.9)

1.00 (reference)

-

1-35

625

34 (5.4)

5.52 (2.22-

35

330

31 (9.4)

9.90 (4.28-