Psychiatry Research 329 (2023) 115540 Contents lists available at ScienceDirect Psychiatry Research journal homepage: www.elsevier.com/locate/psychres Shared decision making in patients with substance use disorders: A one-year follow-up study Pedro Serrano-Pérez a, b, c, *, Amado Rivero-Santana d, e, Constanza Daigre-Blanco a, f, g, h, Raúl Felipe Palma-Álvarez a, f, g, h, Icía Nistal-Franco b, Josep Antoni Ramos-Quiroga a, f, g, h, Lara Grau-López a, f, g, h a Department of Psychiatry and Forensic Medicine, Autonomous University of Barcelona, Barcelona, Spain b Department of Psychiatry, Hospital Álvaro Cunqueiro, SERGAS, Vigo, Spain c Translational Neuroscience Research Group, Galicia Sur Health Research Institute (IIS-Galicia Sur), SERGAS-UVIGO, CIBERSAM, Vigo, Spain d Canary Islands Health Research Institute Foundation (FIISC) e Network for Research on Chronicity, Primary Care, and Health Promotion (RICAPPS) Spain f Department of Psychiatry, Addiction and Dual Diagnosis Section, Hospital Universitari Vall d’Hebron, Barcelona, Spain g Psychiatry Group, Mental Health and Addiction, Vall d’Hebron Research Institute (VHIR), Barcelona, Spain h Biomedical Network Research Centre on Mental Health (CIBERSAM), Madrid, Spain A R T I C L E I N F O A B S T R A C T Keywords: Patient-centered care in therapeutic processes has been associated with better clinical outcomes, however, it Patient-centered care remains a poorly studied aspect in Substance Use Disorder (SUD). The study aimed to evaluate patient’s pref­ Patient preferences erences, perceived participation in treatment decisions and activation level; and how they predict retention, Addiction pharmacological adherence and substance use during one-year follow-up. Logistic regression models were used Patient activation Patient engagement to analyze the association between independent variables, along with a wide number of sociodemographic and Chemical dependency clinical covariates, and outcomes. Most patients prefer a shared or passive role when making decisions about their treatment, and showed concordance between their preferred and perceived roles. In the univariate models, perceiving more involvement than desired showed a higher likelihood of treatment discontinuation at 12 months, and substance use at 6 and 12 months. No significant associations were found between the remaining decisional variables or the degree of activation with the assessed outcomes. A majority of SUD patients prefer and perceive to be involved in the decision-making process about their treatment. Patients perceiving more involvement than desired might experience an excess of responsibility that could negatively influence treatment continuation and substance use. Limitations of the study preclude any definitive conclusion, and more research is needed to confirm these results. 1. Introduction These figures, together with their economic consequences, have forced public health systems to implement preventive and therapeutic in­ Substance Use Disorders (SUD) are complex, multifactorial, and terventions of proven effectiveness and cost-effectiveness, such as the chronic health conditions in which the patient’s awareness of their International Standards for the Treatment of Drug Use Disorders Draft illness, motivation and the degree to which barriers to treatment are for Field Testing (2020). In fact, only 16 % of the people affected by SUD eliminated, play a relevant role. Substance use is the source of serious receive treatment (Pirie, T., and National Treatment Indicators Working health problems, of increasing frequency, affecting 11.2 % of the world Group, 2015; World Drug Report, 2020) and therefore one of the main population; contributing to 21.1 % of all deaths. The 2020 United Na­ objectives continues to be linking and retaining of the patient to the tions World Drug Report noted that in 2017 there were 167,000 deaths treatment (Marchand et al., 2019). Several studies have shown that associated with SUD; meaning the loss of 21 million years due to greater retention is related to better health outcomes, as well as allowing disability or death (Messas et al., 2019; World Drug Report, 2020, s. f.). therapeutic goals to be achieved. The high rate of dropouts in SUD * Corresponding author at: Department of Psychiatry and Forensic Medicine, Autonomous University of Barcelona, Barcelona, Spain. E-mail address: pedro.guillermo.serrano.perez@sergas.es (P. Serrano-Pérez). https://doi.org/10.1016/j.psychres.2023.115540 Received 13 May 2023; Received in revised form 7 October 2023; Accepted 9 October 2023 Available online 11 October 2023 0165-1781/© 2023 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC license (http://creativecommons.org/licenses/by- nc/4.0/). P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 patients makes studying and improving this parameter crucial (Daigre 2.1. Setting et al., 2021; Friedrichs et al., 2016). To this end, efforts have been made to identify barriers and facilitators to improve access by offering The Vall d’Hebron outpatient center for addiction treatment serves healthcare resources and treatment options of different organizational individuals with SUD who live in the northern neighborhoods of Bar­ complexity (hospitalizations, ambulatory care centers, day hospitals and celona, an area with a lower socioeconomic level than the average of the therapeutic communities) (Muthulingam et al., 2019). Some authors city. The center offers an integrative treatment approach for behavioral have argued that a more focused or Patient-Centered Care (PCC), and all substance addictions based on a combination of individual psy­ defined as a more consistent with the values, needs and wishes of the chotherapy based on motivational interviewing and psychopharmaco­ patient care and carried out involving patients in discussions and de­ logical treatment based on international guidelines. cisions about their health, could help to achieve this goal (Mead and Bower, 2000). One of the main tools to carry out PCC is shared decision 2.2. Participants making (SDM); a bilateral communicative process in which the health care provider and the patient try to reach an agreement on the treat­ We included consecutive patients who met the following criteria: 1) ment, based on the available scientific evidence about the potential age between 18 and 65 years, 2) meeting SUD criteria according to DSM- benefits and risks of the different treatment options, considering also 5; 3) ability to understand and fill out the research questionnaire, in the how these potential consequences are valued by the patient (Charles opinion of the physician; and 4) signature of the informed consent. Both et al., 1999). SDM, therefore, aims to include the patient’s preferences in new patients and patients who already were under treatment at the the decision-making process, and it is especially relevant in the man­ center were included. Exclusion criteria were: 1) present a state of agement of chronic diseases and "preference-sensitive" medical de­ intoxication at the time of the interview, 2) decompensated psychiatric cisions when different treatment options offer a similar balance between disorder, and 3) not fluent in Spanish. risks and benefits. Previous studies indicate that a greater degree of involvement in 2.3. Recruitment and assessment process dealing with addictions is associated with greater treatment satisfaction (Brener et al., 2009), and a reduction in the severity of the SUD and the The recruitment and assessment process was performed by trained comorbid psychiatric pathologies (Joosten et al., 2009). SDM has been psychiatrists and psychologists. A convenience sample method was also linked to better outcomes in the treatment of SUD patients, used. All consecutive new patients and patients who already were under improving self-control and reducing substance use (Joosten et al., 2009, treatment at the center were invited to participate whether they met the 2011). It is not clear, however, how and to what extent patients want to inclusion criteria. Those who accepted, signed informed consent. There be involved in this therapeutic process. Some studies indicate that pa­ was no financial compensation for participation. tients prefer more information than responsibility when deciding Baseline data was collected in at least two visits for patients’ usual (Joosten et al., 2011). Friedrichs et al. (2016), in a systematic review on outpatient treatment, carried out by the psychiatrist and/or the psy­ SDM in patients with SUD, found only two studies reporting on patients chologist in charge of the patient. At 6 and 12 months, computerized desire to play an active role in the choice of treatment, either by clinical records were reviewed regarding attendance at scheduled visits, choosing the treatment option themselves or sharing the decision with discharge or disassociation from the center, and substance use. Patients the clinician (Neuner et al., 2007; Sobell et al., 1992). Besides, some­ who had not discontinued treatment at both time points filled out the times, the patient’s therapeutic objectives could not coincide with those questionnaire on medication adherence. offered by the professional (Alves et al., 2017; Hodgins et al., 1997). In the PCC context, a distinction is made between the patient’s 2.4. Measures preferences regarding their treatment and their participation (Montori et al., 2013). SDM focuses on the process itself and tries to include these All psychometric and clinical scales used have been previously preferences within the clinical encounter. As stated, it is not known to validated in Spain. what extent the preferences of the patients and the SDM are sufficiently Independent variables: implemented during the clinical encounter, and the degree of partici­ pation that the patients with SUD want to have (Marchand et al., 2019). - The Control Preference Scale (CPS): a self-reported scale with two The objective of the study is to evaluate the preferences about items and five response options, that evaluate the patient’s prefer­ participation in decision making, the degree of patients’ perceived SDM ence and perception, respectively, about their involvement in the and activation; and how these factors influence treatment retention, decision-making process (De Las Cuevas and Peñate, 2016; Degner pharmacological adherence and abstinence from the substance during a et al., 1997). Options range from a completely active role (the patient year follow-up in an outpatient center. The hypothesis was that a lower makes the decision) to a completely passive one (the doctor decides perceived participation in the decision-making process, a lower without the participation of the patient), with a shared decision in concordance between perceived and preferred role in participation, and the midpoint. For analysis purposes, scores were collapsed into three a lower degree of activation are associated with worse results. categories: active (considering or not doctor’s opinion), shared and passive (considering or not patient’s opinion). 2. Method - Shared Decision-Making Questionnaire (SDM-Q-9): a self-reported single-factor 9-item scale, which assessed the doctor’s promotion This is a cohort prospective observational study, carried out in the of SDM in consultation as perceived by the patient (De las Cuevas outpatient service for addictions of the Hospital Vall d’Hebron (Barce­ et al., 2015; Kriston et al., 2010). Items are answered on a Likert scale lona, Spain). The study was approved by the hospital’s Ethical Com­ ranging from 0 (completely disagree) to 5 (completely agree). The mittee, and it was carried out following the principles of the Helsinki total score is transformed to a 0–100 scale, with higher scores indi­ Declaration (World Medical Association, 2013). The article has been cating higher levels of SDM. made following the recommendations of the Strengthening the Report­ - Patient Activation Measure (mental health version, PAM-MH): a self- ing of Observational studies in Epidemiology (STROBE) guideline for report questionnaire that assesses patients’ self-perceived knowl­ cohort studies (Cuschieri, 2019). edge, skills, confidence, and involvement in the management of chronic diseases (Green et al., 2010; Moreno-Chico et al., 2017). It consists of 13 items to be answered on a Likert scale ranging from (1) totally disagree to (4) totally agree. The total score is transformed to 2 P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 a 0–100 scale using calibration tables, provided under license. frequent outcome value (Austin and Steyerberg, 2017), this size would Higher scores indicate a higher degree of activation. offer adequate power for univariate models with an outcome probability of 0.10, and for a model with three independent variables (SDM-9-Q, Dependent variables: PAM and one scale from the CPS) and an outcome probability of 0.30. - Retention: clinical records and professionals’ reports were used to 2.6. Statistical analysis assess treatment continuation (yes/no). Discontinuation was defined as not attending a scheduled appointment without prior justification, Descriptive statistics (means, standard deviations, percentages) were or during the subsequent 30 days. This definition allows an objective calculated for the study variables. The association between the three measure of retention according to a specific time frame period, and it independent variables at baseline was analyzed by means of Pearsons’ has been previously used in other studies evaluating retention in SUD correlation and linear regression. patients (Daigre et al., 2021; Palma-Álvarez et al., 2021; Ros-Cu­ The associations of the independent variables, and the included curull et al., 2018). covariates, with the outcomes were analyzed separately for the periods - Medication adherence was assessed with the Spanish version of the 0–6 and 6–12 months. First, univariate analyses were carried out for Simplified Medication Adherence Questionnaire (SMAQ) (Knobel each predictor; we used multilevel logistic regression models, including et al., 2002). It is made up of six questions that evaluate different the health provider as a random effect in order to adjust for patients’ aspects related to pharmacological compliance: forgetfulness, clustering. This model was then compared with the logistic model by routine, adverse effects and number of omissions. A patient is clas­ means of the likelihood ratio test, and the latter was used if the test’s sified as non-compliant if he/she answers any response in the sense result was not significant. Those variables with p-values lower than 0.10 of non-adherence, or reports having missed more than two doses in in the univariate analyses were introduced together in a multivariate the last week, or more than two full days not taking the medication in model (model 1). Multicollinearity was assessed by means of the vari­ the last three months. ance inflation factor (VIF) and those variables with VIF > 4 were - Substance use: it was measured by a multi-panel urine drug test for excluded (model 2). For the multivariate models, a p-value < 0.05 was the most common drugs (opioids, cannabis, cocaine, benzodiaze­ considered significant. pines, and amphetamines) and by a breath alcohol test. Substance For each model, potential misspecification was explored with the use urinalysis was conducted twice weekly during the follow-up. linktest Stata command. This test rebuilds the model using the predicted Active use or relapse was defined as three consecutive positive value and predicted value squared as predictors. A non-significant result urine tests for the main substance or the substance that caused the of the former term or a significant value of the latter indicates potential most problems for the patient (according to clinical judgment and specification errors. supported by the information provided by the ad-hoc questionnaire We report four goodness-of-fit indexes: the pseudo-R2, the Hos­ and the EuropASI on social, medical, psychological and psychiatric mer–Lemeshow test, the Akaike’s Information Criterium (AIC) and the issues). This definition has been used in previous studies Bayesian Information Criterium (BIC). The models’ predictive perfor­ (Grau-López et al., 2012; Roncero et al., 2019). mance was assessed by calculating sensitivity, specificity, correctly classified cases and the area under the receiver operating characteristic Covariables (ROC) curve (AUC). Finally, we further refined the models excluding Sociodemographic variables included were: age, gender, education, non-significant predictors until the AIC and BIC were minimized (model marital status and job status. For the regression analyses, education was 3). dichotomized into no studies/primary vs. secondary/university/other, Missed scores were not imputed. Only in the case of substance use, marital status into married/coupled vs. single/separated/widow, and we carried out an additional analyses assuming that none of the lost job status into active vs. non-active. We also assessed substance use at participants were abstinent at each time point (worst-case scenario). baseline (yes/no), previously under follow-up/new patient and other substance consumption variables. 3. Results The Zuckerman–Kuhlman Personality Questionnaire (ZKPQ) was used to determine the basic personality dimensions. It consists of 99 From March 2019 to June 2021, 214 patients were recruited by 10 items on 5 scales. Neuroticism–Anxiety, Activity, Sociability, Impul­ mental health providers (median = 16, range 10–62). The sociodemo­ siveness/Sensation seeking, and Aggressiveness/hostility graphic characteristics of the recruited sample are detailed in Table 1. (Gomà-i-Freixanet et al., 2008). The mean age was 44 years old, two-thirds were men, and 56.8 % had The European Addiction Severity Index (EuropASI) was used to only primary or no formal studies. Seventy-one percent were active measure addiction severity, main substance and general consumption of substance users at the start of the study, and 37.4 % were new patients. the substance. It is a semi-structured interview that collects medical, Forty-nine percent of the participants were being treated for alcohol occupational, legal, family, social, psychological problems, and sub­ dependence, 28.5 % for cocaine and 7.9 % for opioids, whereas other stance use history (Bobes et al., 2017; Kokkevi and Hartgers, 1995). drugs showed values under 5 %. Half of the sample had been diagnosed The 36-Item Short-Form Health Survey (SF-36) was used to measure with an Axis-I mental disorder, and almost one-third had a personality the physical and psychological quality of life from the patient’s disorder diagnosis. perspective. The questionnaire provides scores in 8 spheres, and these At 6 months, one patient had died and another one was discharged. are aggregated into the physical dimension and the mental dimension Between 6–12 months, five patients died and one was discharged. These (Vilagut et al., 2005; Ware and Sherbourne, 1992). eight patients were excluded from the corresponding analyses. The Spanish versions of the Semi-Structured Clinical Interview for At 6 months, the rate of treatment retention was 81.6 % (173/212). DSM-IV Axis-I Disorders (SCID-I) and Axis-II Disorders (SCID-II), were Among these, 126 had been prescribed medication and there was only used to assess patients’ psychiatric comorbidities. Both have shown high one missed value on this variable. Between 6–12 months, 69.4 % of inter-rater reliability for categorical diagnosis (First and Gibbon, 2004). patients (143/206) were compliant with appointments (with only two missed values on medication adherence for those prescribed). In addi­ 2.5. Sample size tion to these retained patients, data on medication adherence and sub­ stance use were available, respectively, for 20 and 9 out of the 63 We aimed to recruit at least 200 participants. Based on the recom­ patients who discontinued (Fig. 1). mended formulae (20 * Number of predictors) / Probability of the less At baseline, missed values ranged from 0 to 4.7 % across all 3 P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 Table 1 associations were observed between the CPS (perceived role and role Sample Characteristics (n = 214). matching) and the other two scales. Sociodemographic N (%) 3.2. Predictors of the clinical outcomes Age, mean (sd) 43.98 (11.6) Women 72 (33.6 %) Education (n = 213) In all univariate analyses, the multilevel model with the psychiatrist No studies 25 (11.7 %) as a random effect did not significantly differ from the simple logistic Primary 96 (45.1 %) model, with the likelihood ratio test yielding most χ2 values near to zero. Secondary 65 (30.5 %) University 27 (12.7 %) Therefore, we used the more parsimonious logistic model. In the ap­ Marital status (n = 210) pendix, Tables A3-A6 show the univariate models and the initial results Single 84 (40.0 %) of the multivariate models (model 1) (Table 2). Married/coupled 86 (40.6 %) Separated/divorced 39 (18.6 %) 3.2.1. Treatment retention Widow 1 (0.5 %) Job status (n = 209) Decision-related variables (CPS, SDM-9-Q) and patient activation Employed 64 (30.6 %) (PAM) did not significantly predict retention at 6 months (Table 3). In Unemployed 74 (35.4 %) the final model, being abstinent at baseline (OR = 0.13, 95 % CI: 0.04, Sick leave 23 (11.0 %) 0.46), a higher score on the ASI-Drug use subscale (OR = 0.86, 95 % CI: Pensioner 40 (19.1 %) Other 8 (3.8 %) 0.75, 0.98) and lower sociability scores (OR = 0.89, 95 % CI: 0.81, 0.99), Clinical significantly related to continuation (Table 4). The pseudo-R2 was 0.18, New patient 80 (37.4 %) the model correctly classified 84.16 % of the patients and the AUC was Substance use at baseline 153 (71.5 %) 0.78. Under treatment for Retention during 6–12 months was significantly predicted in the Alcohol 105 (49.1 %) Cocaine 61 (28.5 %) final multivariate model by not being a new patient (OR = 0.49, 95 % CI: Opioid 17 (7.9 %) 0.25, 0.95), and having a mood disorder (OR = 2.12, 95 % CI: 1.02, Cannabis 9 (4.2 %) 4.41) (Table 4). Participants who perceived more involvement than Benzodiazepine 6 (2.8 %) desired were less likely to be retained compared to those with matched Amphetamine 4 (1.9 %) Analgesic 1 (0.5 %) preferences (OR = 0.39, 95 %CI: 0.15, 1.98; p = 0.046), but the differ­ Polydrug 11 (5.1 %) ence did not reach significance in the multivariate model (OR = 0.38, 95 Current or past dependence %CI: 0.14, 1.06; p = 0.061). Excluding any of the non-significant pre­ Alcohol 124 (57.9 %) dictors did not improve model fit, assessed by the AIC. The pseudo-R2 Cocaine 98 (45.8 %) was 0.11, the model correctly classified 74.62 % of the patients and the Opioid 28 (13.1 %) Cannabis (n = 212) 54 (25.5 %) AUC was 0.72. Benzodiazepine 18 (8.4 %) Polydrug (n = 212) 44 (20.1 %) 3.2.2. Medication adherence Psychiatric comorbidities The rate of adherent patients at 6 months was 84 % (105/125). The Depressive disorder (n = 211) 68 (32.2 %) Bipolar disorder (n = 210) 8 (3.8 %) CPS, SDM-9-Q and PAM did not significantly predict medication Anxiety (n = 210) 33 (15.7 %) adherence (Table 5). In the multivariate model, significant predictors Adaptive disorder (n = 210) 13 (6.2 %) were: not being a new patient (OR = 0.26, 95 % CI: 0.08, 0.92), having Psychotic disorder (n = 213) 9 (4.2 %) an Axis-I mental disorder (OR = 5.03, 95 % CI: 1.27, 19.87), and not Eating disorder (n = 212) 4 (1.9 %) having a personality disorder (OR = 0.22, 95 % CI: 0.06, 0.79) (Table 6). Personality disorder (n = 205) 63 (30.7 %) After excluding non-significant predictors (model 3), the model’s fit improved and substance use at baseline (OR = 0.19; 95 %CI: 0.04, 0.98) variables, except for the PAM (11.2 %) (Appendix, Table A1). None of and the ASI-Employment and support subscale (OR = 0.78; 95 %CI: the included variables were significantly associated with having a 0.62, 0.98) became significant. The pseudo-R2 was 0.26, the model missed score in the PAM (Appendix, Table A2). correctly classified 88.79 % of the patients and the AUC was 0.82. Internal consistency (Cronbach’s alpha) of the psychometric scales Between 6–12 months, 73.2 % of the patients (93/127) were used was 0.90 (SDM-9-Q), 0.84 (PAM), 0.96 (SMAQ), 0.87 (ZKPQ- adherent to medication. The significant variables in model 2 were not Neuroticism), 0.61 (ZKPQ-Aggression/hostility), 0.72 (ZKPQ-Activity), being a new patient (OR = 0.29, 95 % CI: 0.10, 0.82), having a 0.79 (ZKPQ-Impulsiveness), 0.79 (ZKPQ-Sociability), 0.79 (SF-36 depressive disorder (OR = 3.58, 95 %CI: 1.13, 11.29) and not having a physical) and 0.82 (SF-36 mental). personality disorder (OR = 0.26, 95 % CI: 0.09, 0.75) (Table 6). In model 3, the ASI-Medical status subscale also obtained a significant result (OR = 0.81, 95 %CI: 0.70, 0.95). The pseudo-R2 was 0.19, the model 3.1. SDM-related variables and patient activation correctly classified 75.00 % of the patients and the AUC was 0.79. Most patients preferred a shared (46.9 %) or passive (38.7 %) role in 3.2.3. Substance use the decision-making process, and 38.1 % and 50.1 %, respectively, Between baseline and 6 months, 40 % of participants (70/175) had perceived that they played those roles. Two-thirds (66.5 %) of the pa­ presented substance use. In the univariate models, participants who tients showed concordance between their preferred and perceived roles, perceived more involvement than desired, compared to those with whereas 23 % perceived less control than they wanted, and 10.5 % were matched preferences, were significantly more likely to present substance given more participation than desired. use (OR = 2.63, 95 % CI: 1.04, 6.63) (Table 7). However, in the multi­ Mean scores on the SDM-9-Q and the PAM were respectively 69.0 (sd variate model, the result became non-significant, although its exclusion = 21.6) and 56.8 (sd = 14.4), indicating moderate levels of perceived clearly worsened the model’s fit, assessed by the AIC and BIC (Table 8). SDM and activation. These two variables showed a significant correla­ Significant predictors were substance use at baseline (OR = 7.08, 95 % tion (Pearson r = 0.22, p = 0.003). Patients who preferred an active role, CI: 2.71, 18.51) and a higher score on the ASI employment/support based on the CPS, scored significantly higher on the PAM than those who subscale (OR = 1.19, 95 % CI: 1.03, 1.37). In model 3, being a new preferred a shared role. (B = 5.22, p = 0.003). No other significant patient entered in the model significantly associated to a higher 4 P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 Fig. 1. Flow chart of the study. significantly more likely to present substance use, but the p-value felt Table 2 slightly above the significance level (p = 0.051). In the multivariate Decisional variables and patient activation. models, the result was not significant, but again its exclusion worsened CPS the model’s fit. Significant predictors were substance use at baseline Preferred role (n = 209) (OR = 4.23, 95 % CI: 1.41, 12.69) and the ASI-Family status subscale Active 30 (14.4 %) (OR = 1.18, 95 % CI: 1.00, 1.36). In model 3, the ASI-Legal status Shared 98 (46.9 %) subscale became significant (OR = 1.24, 95 %CI: 1.03, 1.48). The Passive 81 (38.7 %) Perceived role (n = 210) pseudo-R2 was 0.13, the model correctly classified 73.79 % of the pa­ Active 24 (11.4 %) tients and the AUC was 0.76. Shared 80 (38.1 %) In the worst-case scenario, only substance use at baseline was sig­ Passive 106 (50.1 %) nificant (appendix, Table A6). Role matching (n = 209) Less than desired 48 (23.0 %) Matched 139 (66.5 %) 4. Discussion More than desired 22 (10.5 %) SDM-9-Q (0–100) (n = 211), mean (sd) 69.0 (21.6) This study aimed to explore the control preferences, SDM perception PAM (0–100) (n = 190), mean (sd) 56.8 (14.4) and activation of people treated by SUD, as well as the longitudinal CPS: Control Preferences Scale; PAM: Patient Activation Measure; SDM-9-Q: association of these variables with the evaluated behavioral outcomes. Shared decision making questionnaire. The results of the CPS show that most patients prefer a shared (47 %) or passive (39 %) role when making decisions about their treatment. Pre­ likelihood of substance use (OR = 2.18, 95 %CI: 1.03, 4.61). The pseudo- vious research in patients with SUD has focused mainly on treatment R2 was 0.21, the model correctly classified 71.10 % of the patients and preferences (Friedrichs et al., 2016), and studies on participation pref­ the AUC was 0.77. In the worst-case scenario, the significant predictors erences are very scarce and limited in their scope (Neuner et al., 2007; were the same (appendix, Table A6). Sobell et al., 1992). A recent study in Germany with patients with Between 6–12 months, the rate of substance use was 28.9 % (44/ alcohol use disorders showed a quite lower rate of passive role prefer­ 152). Patients who perceived more involvement than desired were also ence (10 %) (Friedrichs et al., 2018), whereas in our study patients with 5 P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 Table 3 diagnosed with SUD. In mixed psychiatric samples, Mahone et al. (2008) Univariate logistic regression models for treatment retention with decisional did not observe a significant association between a preferred role in SDM variables and patient activation as independent variables. and self-reported medication adherence; whereas De las Cuevas et al. 0–6 months 6–12 months (2014) found that patients who preferred a passive role showed better CPS N OR (95 % P N OR (95 % P adherence than those who preferred an active role. Regarding the CI) CI) perception of involvement, these two studies did not obtain significant associations with adherence, as also happened in our study. Preferred role (ref: 207 202 shared) Approximately one-third of the participants perceived more or less Active 0.99 (0.33, 0.982 0.60 (0.25, 0.262 involvement than they preferred, indicating that there is still room for 2.97) 1.46) improvement in the matching of patients’ preferences and professionals’ Passive 0.73 (0.34, 0.450 0.94 (0.48, 0.843 promotion of involvement in this sample. Nonetheless, the observed 1.56) 1.81) Perceived role (ref: 208 203 shared) Active 1.06 (0.32, 0.925 0.65 (0.25, 0.384 Table 5 3.59) 1.71) Univariate logistic regression models for medication adherence with decisional Passive 0.89 (0.42, 0.764 1.05 (0.55, 0.877 variables and patient activation as independent variables. 1.90) 2.02) Role matching 207 202 0–6 months 6–12 months (ref: matched) CPS N OR (95 % P N OR (95 % P Less than 0.84 (0.36, 0.611 0.66 (0.32, 0.251 CI) CI) desired 1.98) 1.35) More than 0.53 (0.19, 0.236 0.39 (0.15, 0.046 Preferred role (ref: 124 127 desired 1.51) 0.98) shared) SDM-9-Q 209 1.00 (0.99, 0.717 203 1.01 (0.99, 0.286 Active 0.60 (0.17, 0.419 0.58 (0.10, 0.539 1.02) 1.02) 2.07) 3.25) PAM 189 1.01 (0.98, 0.499 183 1.00 (0.98, 0.926 Passive 1.80 (0.56, 0.324 0.86 (0.33, 0.769 1.04) 1.02) 5.79) 2.29) Perceived role 124 127 Significant p-values are shown in bold. (ref: shared) CI: Confidence Interval; CPS: Control Preferences Scale; OR: Odds ratio; PAM: Active 0.77 (0.17, 0.734 0.57 (0.10, 0.535 Patient Activation Measure; SDM-9-Q: Shared decision making questionnaire. 3.45) 3.32) Passive 2.24 (0.77, 0.140 1.19 (0.43, 0.736 6.54) 3.31) current or past alcohol dependence showed the same rate that the Rol matching (ref: 124 127 overall sample. The preferences’ distribution reported in our study is matched) similar to that observed in Spanish mixed psychiatric samples (De Las Less than 1.28 (0.39, 0.685 0.94 (0.37, 0.904 Cuevas et al., 2020; De Las Cuevas and Peñate, 2016; Mundal et al., desired 4.25) 2.42) More than 1.58 (0.18, 0.679 0.39 (0.11, 0.155 2021), and therefore this difference with the German study could be desired 13.63) 1.42) explained by cultural differences between countries, as seen in other SDM-9-Q 124 0.98 (0.96, 0.242 126 1.00 (0.99, 0.738 medical conditions (Yennurajalingam et al., 2018). Regarding the as­ 1.01) 1.01) sociation between participation preferences and outcomes, we have not PAM 115 0.98 (0.95, 0.395 117 0.98 (0.95, 0.414 found significant results. Neuner et al. (2007) found a significant asso­ 1.02) 1.02) ciation between tobacco and alcohol self-reported use and a lower Significant p-values are shown in bold. preference for involvement in decision-making, but this was a CI: Confidence Interval; CPS: Control Preferences Scale; OR: Odds ratio; PAM: cross-sectional analysis in trauma patients treated for acute injury, not Patient Activation Measure; SDM-9-Q: Shared decision making questionnaire. Table 4 Final multivariate logistic models on patient retention. 0–6 months (n = 202) 6–12 months (n = 197) Model 2a Model 3b Model 2c Baseline predictors OR (95 % CI) p OR (95 % CI) p Baseline predictors OR (95 % CI) p Married/couple 2.32 (0.96, 5.57) 0.061 2.31 (0.96, 5.56) 0.061 Substance use 0.53 (0.23, 1.25) 0.148 Substance use 0.13 (0.04, 0.50) 0.003 0.13 (0.04, 0.46) 0.002 New patient 0.49 (0.25, 0.95) 0.035 New patient 0.88 (0.38, 2.02) 0.764 – – Opioids 2.73 (0.72, 10.31) 0.139 ASI-DU 0.86 (0.75, 0.98) 0.026 0.86 (0.75, 0.98) 0.023 Mood disorder 2.12 (1.02, 4.41) 0.043 Alcohol 1.95 (0.82, 4.63) 0.130 1.99 (0.84, 4.68) 0.116 Role matching (ref: matched) Mood disorder 1.86 (0.74, 4.66) 0.184 1.87 (0.75, 4.68) 0.181 Less than desired 0.56 (0.25, 1.22) 0.143 Sociability 0.89 (0.80, 0.99) 0.034 0.89 (0.81, 0.99) 0.035 More than desired 0.38 (0.14, 1.06) 0.063 Linktest: p-values of PS and PS2 0.001; 0.249 0.001; 0.249 Linktest: p-values of PS and PS2 <0.001; 0.170 Hosmer–Lemeshow χ2 (p) 7.89 (0.444) 9.74 (0.284) Hosmer–Lemeshow χ2 (p) 7.27 (0.401) AIC / BIC 175.76 / 202.23 173.85 / 197.01 AIC / BIC 230.34 / 253.33 Pseudo-R2 0.18 0.18 Pseudo-R2 0.11 Sensitivity 98.17 % 98.17 % Sensitivity 94.89 % Specificity 26.32 % 23.68 % Specificity 28.33 % Correctly classified 84.65 % 84.16 % Correctly classified 74.62 % AUC 0.78 0.78 AUC 0.72 a Age excluded due to collinearity (VIF > 4). b Non-significant predictors excluded until both AIC and BIC values were minimized. c Age and SF-36 physical component excluded due to collinearity (VIF > 4). The exclusion of any predictor did not improve AIC. ASI: Addiction Severity Index; ASI- DU: Drug use; AIC: Akaike’s information criterion; AUC: Area Under the Curve; BIC: Bayesian information criterion; CI: Confidence Interval; OR: Odds ratio; PS: Predicted score; VIF: Variance inflation factor. Significant p-values are shown in bold. 6 P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 Table 6 Final logistic multivariate models on medication adherence. 0–6 monthsa (n = 116) 6–12 months (n = 119) a b Model 2 Model 3 Model 2 Model 3b Baseline predictors OR (95 % CI) p OR (95 % CI) p Baseline predictors OR (95 % CI) p OR (95 % CI) p Substance use 0.19 (0.04, 0.056 0.19 (0.04, 0.048 Substance use 0.86 (0.25, 0.803 – – 1.04) 0.98) 2.91) New patient 0.26 (0.08, 0.037 0.25 (0.07, 0.029 New patient 0.29 (0.10, 0.020 0.25 (0.10, 0.005 0.92) 0.87) 0.82) 0.66) ASI-ES 0.82 (0.64, 0.109 0.78 (0.62, 0.031 ASI-MS 0.86 (0.72, 0.072 0.81 (0.70, 0.010 1.05) 0.98) 1.01) 0.95) ASI-AU 0.88 (0.70, 0.263 – – ASI-ES 0.89 (0.75, 0.200 – – 1.10) 1.06) ASI-FS 0.93 (0.74, 0.548 – – ASI-AU 0.95 (0.79, 0.547 – – 1.17) 1.13) Axis-I disorder 5.03 (1.27, 0.021 4.61 (1.19, 0.027 Depressive disorder 3.58 (1.13, 0.030 3.43 (1.13, 0.030 19.87) 17.79) 11.29) 10.43) Personality disorder 0.22 (0.06, 0.020 0.24 (0.07, 0.025 Personality disorder 0.26 (0.09, 0.013 0.21 (0.08, 0.002 0.79) 0.83) 0.75) 0.57) SF-36 Mental 1.02 (0.98, 0.350 – – 1.06) Linktest: p-values of PS and 0.003; 0.285 0.004; 0.246 Linktest: p-values of PS and 0.002; 0.799 0.019; 0.278 PS2 PS2 Hosmer–Lemeshow 5.56 (0.696) 12.83 (0.118) Hosmer–Lemeshow χ2 (p) 7.53 (0.480) 5.97 (0.543) χ2 (p) AIC / BIC 91.16 / 113.19 88.73 / 105.25 AIC / BIC 126.22 / 151.23 123.25 / 137.19 Pseudo-R2 0.27 0.26 Pseudo-R2 0.22 0.19 Sensitivity 96.91 % 98.97 % Sensitivity 91.95 % 90.91 % Specificity 31.58 % 36.84 % Specificity 40.63 % 31.25 % Correctly classified 86.21 % 88.79 % Correctly classified 78.15 % 75.00 % AUC 0.84 0.82 AUC 0.81 0.79 a Age excluded due to collinearity (VIF > 4). b Non-significant predictors excluded until both AIC and BIC values were minimized. AIC: Akaike’s information criterion; ASI: Addiction Severity Index; ASI-ME: Medical status; ASI-ES: Employment and support; ASI-AU: Alcohol use; ASI-FM: Family status; AUC: Area Under the Curve; BIC: Bayesian information criterion; CI: Confidence Interval; OR: Odds ratio; PS: Predicted score; VIF: Variance inflation factor. Significant p-values are shown in bold. near-significant associations with the outcomes. Concretely, patients Table 7 who perceived more involvement than desired showed a subsequent Univariate logistic regression models for substance use with decisional variables higher likelihood of treatment discontinuation and substance use. and patient activation as independent variables. However, this result is based only on 22 patients in the mentioned 0–6 months 6–12 months subgroup (16 at 6 months and 14 at 12 months) and the p-values in the CPS N OR (95 % P N OR (95 % P multivariate models were non-significant, and therefore it must be CI) CI) interpreted as a mere hypothesis to be confirmed in future studies. De las Preferred role (ref: 171 150 Cuevas et al. (2014) found better self-reported medication adherence in shared) psychiatric patients with matched preferences compared to those un­ Active 1.34 (0.63, 0.443 1.74 (0.47, 0.407 matched, but they did not compare separately patients with more and 2.85) 6.40) Passive 0.54 (0.14, 0.443 1.16 (0.48, 0.741 less involvement than desired. In certain pathologies, self-esteem or the 2.02) 2.82) patients’ perception of his own abilities to make decisions is reduced. Perceived role (ref: 172 151 This circumstance has been associated with a lower desire to assume shared) responsibilities and a subsequent preference for a more passive role in Active 2.22 (0.56, 0.259 1.80 (0.39, 0.454 8.88) 8.38) the consultation (Hamann et al., 2007). It has been pointed out how Passive 0.73 (0.43, 0.259 0.90 (0.51, 0.717 patients with SUD tend to have lower self-esteem and see themselves as 1.23) 1.60) more submissive subjects compared to other samples of patients without Role matching 171 150 SUD, which could explain the observed relationship between perceiving (ref: matched) more responsibility than desired and subsequent worsening in clinical Less than 1.76 (0.85, 0.129 1.42 (0.79, 0.244 desired 3.64) 2.58) results (Alavi, 2011; Joosten et al., 2011). More than 2.63 (1.04, 0.041 2.35 (1.00, 0.051 A recent systematic review evaluated the association between in­ desired 6.63) 5.53) dicators of PCC and outcomes of substance treatment (Davis et al., SDM-9-Q 174 0.99 (0.98, 0.267 151 0.99 (0.98, 0.071 2020). Results showed positive associations with improved outcomes 1.01) 1.00) (substance and services use, psychological well-being), but only five PAM 161 1.00 (0.99, 0.937 140 1.01 (1.00, 0.161 1.01) 1.02) studies included a patient-centered indicator other than satisfaction. Intervention studies have been very scarce, with mixed results (Davis Significant p-values are shown in bold. et al., 2020). A randomized trial showed that a 5-session SDM inter­ CI: Confidence Interval; CPS: Control Preferences Scale; OR: Odds ratio; PAM: vention with SUD patients significantly improved primary substance Patient Activation Measure; SDM-9-Q: Shared decision making questionnaire. use, addiction severity, quality of life and psychiatric problems at 3-month follow-up (E. a. G. Joosten et al., 2009). More recently, an concordance is greater than the reported in Spanish psychiatric samples, uncontrolled pilot study also showed a reduction of drug use and craving which has been around 50 % (De las Cuevas et al., 2014; De las Cuevas in opioid users under a PCC program, with high retention at 3 months et al.; 2020). This variable is the only one that has shown significant or (Lynch et al., 2021), but a previous randomized trial did not show any 7 P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 Table 8 Final logistic multivariate models on substance use. 0–6 months (n = 168) 6–12 months (n = 143) a b Model 2 Model 3 Model 2c Model 3b Baseline predictors OR (95 % CI) p OR (95 % CI) p Baseline predictors OR (95 % CI) p OR (95 % CI) p Substance use 7.08 (2.71, <0.001 7.18 (2.81, <0.001 Job status 0.65 (0.28, 0.323 – – 18.51) 18.32) 1.53) New patient 2.03 (0.95, 0.068 2.18 (1.03, 0.042 Substance use 4.23 (1.41, 0.010 4.14 (1.52, 0.005 4.33) 4.61) 12.69) 11.28) ASI-MS 1.08 (0.94, 0.294 – – ASI-LS 1.19 (0.97, 0.088 1.24 (1.03, 0.022 1.25) 1.45) 1.48) ASI-ES 1.19 (1.03, 0.021 1.23 (1.09, 0.001 ASI-FS 1.18 (1.00, 0.042 1.17 (1.01 0.037 1.37) 1.39) 1.36) 1.35) ASI-AU 1.06 (0.93, 0.367 – – Opioids 1.50 (0.42, 0.530 – – 1.21) 1.33) ASI-FS 1.04 (0.88, 0.669 – – Polydrug 0.96 (0.34, 0.942 – – 1.21) 2.75) ASI-PS 0.96 (0.81, 0.662 – – Personality disorder 1.30 (0.52, 0.572 1.25 (0.53, 0.613 1.14) 3.22) 2.97) Role matching (ref: matched) Role matching (ref: matched) Less than desired 1.51 (0.61, 0.372 1.58 (0.65, 0.314 Less than desired 1.65 (0.62, 0.319 – – 3.71) 3.85) 4.39) More than desired 1.87 (0.55, 0.319 1.90 (0.56, 0.301 More than desired 1.86 (0.47, 0.378 – – 3.67) 6.42) 7.46) Linktest: p-values of PS and <0.001; 0.802 <0.001; 0.412 Linktest: p-values of PS and 0.011; 0.258 0.142; 0.056 PS2 PS2 Hosmer–Lemeshow, χ2 (p) 8.58 (0.379) 14.38 (0.072) Hosmer–Lemeshow χ2 (p) 9.18 (0.327) 6.11 (0.634) AIC/BIC 196.84 / 228.08 191.66 / 210.44 AIC/BIC 165.60 / 195.23 161.51 / 176.40 Pseudo-R2 0.21 0.21 Pseudo-R2 0.15 0.13 Sensitivity 60.61 % 60.00 % Sensitivity 31.71 % 26.19 % Specificity 80.39 % 78.64 % Specificity 94.12 % 93.2 % Correctly classified 72.62 % 71.10 % Correctly classified 76.22 % 73.79 % AUC 0.80 0.77 AUC 0.77 0.76 a Acitivity trait and SF-36 mental component excluded due to collinearity (VIF > 4). b Non-significant predictors excluded until both AIC and BIC values were minimized. c Age, ASI-Employment and support, ASI-Psychiatric status and SDM-9-Q, excluded due to collinearity (VIF > 4). AIC: Akaike’s Information Criterion; ASI: Addiction Severity Index; ASI-ME: Medical status; ASI-ES: Employment and support; ASI-AU: Alcohol use; ASI-LS: Legal status; ASI-FM: Family status; ASI-PS: Psychiatric status; AUC: Area Under the Curve; BIC: Bayesian information criterion; CI: Confidence Interval; OR: Odds ratio; PS: Predicted score; SDM-9-Q: Shared decision making questionnaire; VIF: Variance inflation factor. Significant p-values are shown in bold. benefit for methadone patients (Schwartz et al., 2017). Neumann et al. No significant differences were obtained regarding the degree of developed an intervention for patients with trauma and risk of alcohol activation. Activation is an important aspect for self-care and manage­ consumption. They applied a computerized tool that provided support in ment of chronic diseases (Rodriguez et al., 2019), however, it requires decision-making. The intervention was associated with a significant prior education and guidance. This could justify the findings of our study reduction in alcohol consumption at 6 months but not at 12 months and would point out the importance of carrying out interventions to follow-up (Neumann et al., 2006). Barrio et al. created an app for pa­ enhance the activation of patients (Newland et al., 2021). It is important tients’ phones that was developed based on the principles of SDM and to point out that there are a multitude of factors that can influence obtained significant reductions in alcohol consumption and greater abstinence, retention and pharmacological adherence and therefore ability to achieve patients’ own therapeutic goals (Barrio et al., 2017). PCC/SDM is just one more element. In somatic health conditions, SDM has been more frequently asso­ Addictions are disorders in which initially there may be little ciated with improvement in affective aspects and knowledge than with awareness of the disease or a tendency to self-deception in order not to outcomes such as adherence, acquisition of healthy habits, quality of completely abandon consumption. Given this particularity, an excess of life, or improvement in biological measures of health such as blood responsibility on the part of the patient in the initial phases could be pressure (Aubree Shay and Lafata, 2015; Saheb Kashaf et al., 2017; Yun counterproductive and be associated with worse results. It is not un­ and Choi, 2019). common for the professional to have to be firm in some of the initial As commented above, research on the preferences of patients with recommendations to counteract the lack of initial motivation. It is SUD has focused on treatment or setting preferences. In some studies, important to note that both SDM and patient preferences can be applied when these preferences were satisfied, there was a tendency to observe, in a multitude of treatment areas and it would be important to clarify at least, less intense consumption tendencies or less substance use in which ones really provide the expected benefits. previous days (Brown et al., 2002; Friedmann et al., 2004; Luty, 2004). This study has several limitations. First, we established the sample In another study, improvements in ASI scores were obtained when the size in order to get adequate statistical power for univariate analyses and type of service offered by the provider matched what the patients for multivariate models with the three independent variables. Therefore, wanted; with greater retention when the offered medical services coin­ the final multivariate models for retention at 6 months and mediation cided with what was desired by patients with cocaine SUD (Hser et al., adherence could be underpowered. This also precludes the analysis of 1999). Nonetheless, these results could not be confirmed by other interactions between the study variables and potential moderators like studies characterized by their heterogeneous methodology and diver­ type of addiction or presence of a mental disorder, since these analyses gent results (Adamson et al., 2005; Marlowe et al., 2003; Sterling et al., of interactions require very large sample sizes. Further, a larger sample 1997); indicating that preference matching is a suitable approach that size would have allowed it to be randomly divided into two subsamples, should be further evaluated. using the second as a validation sample. On the other side, a relevant 8 P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 percentage of eligible patients did not participate due to their psycho­ the online version, at doi:10.1016/j.psychres.2023.115540. pathological state, the language barrier, or the lack of motivation to be included, and therefore the sample is not completely representative of References the target population. Patients were evaluated in the same center where they receive treatment and not by an external assessor, and this could Adamson, S.J., Sellman, D.J., Dore, G.M., 2005. Therapy preference and treatment outcome in clients with mild to moderate alcohol dependence. Drug Alcohol Rev. 24 introduce to some extent a desirability bias. Another limitation is that (3), 209–216. https://doi.org/10.1080/09595230500167502. only pharmacological adherence was evaluated and not the compliance Alavi, H.R., 2011. The Role of Self-esteem in Tendency towards Drugs, Theft and and effect of the psychological interventions carried out in the center. Prostitution. Addict. Health 3 (3-4), 119–124. Alves, P., Sales, C., Ashworth, M., 2017. Does outcome measurement of treatment for Related to this, although we have included numerous covariates, specific substance use disorder reflect the personal concerns of patients? A scoping review of factors such as the medication type, treatment modality or other aspects measures recommended in Europe. Drug Alcohol Depend. 179, 299–308. https:// related to the clinical encounter were not assessed. The process of SDM doi.org/10.1016/j.drugalcdep.2017.05.049. Aubree Shay, L., Lafata, J.E, 2015. Where is the evidence? A systematic review of shared was assessed only from the patients’ perspective, and not by objective decision making and patient outcomes. Med. Decis. Making 35 (1), 114–131. external observers (e.g., by audio-recording of consultations). Finally, https://doi.org/10.1177/0272989X14551638. results on medication adherence and substance use are restricted to Austin, P.C., Steyerberg, E.W., 2017. Events per variable (EPV) and the relative performance of different strategies for estimating the out-of-sample validity of completers, since most patients who did not comply with the appoint­ logistic regression models. Stat. Methods Med. Res. 26 (2), 796–808. https://doi. ments were missed for the analyses. Although lack of retention is a real org/10.1177/0962280214558972. aspect of clinical practice, the results obtained must be interpreted with Barrio, P., Ortega, L., López, H., Gual, A., 2017. Self-management and Shared Decision- caution. Making in Alcohol Dependence via a Mobile App: A Pilot Study. Int. J. Behav. Med. 24 (5), 722–727. https://doi.org/10.1007/s12529-017-9643-6. Despite these limitations, the study adds valuable knowledge to a Bobes, J., Bascarán, T., Bobes-Bascarán, T., Carballo, J.L., Ma, E., Mesa, D., Flórez, G., poorly researched issue, with a prospective design of a one-year follow- García-Portilla, P., & Sáiz, P.A. (2017).Valoración de la gravedad de la adicción: up. There are very few studies that evaluate SDM and activation in a Aplicación a la gestión clínica y monitorización de los tratamientos, ISBN: 978-84- 611-7641-0. clinical population with SUD, and although the results are only Brener, L., Resnick, I., Ellard, J., Treloar, C., Bryant, J., 2009. Exploring the role of exploratory and subjected to the above-mentioned limitations, they consumer participation in drug treatment. Drug Alcohol Depend. 105 (1-2), raise some interesting insights that should be confirmed in future 172–175. https://doi.org/10.1016/j.drugalcdep.2009.06.016. Brown, T.G., Seraganian, P., Tremblay, J., Annis, H., 2002. Matching substance abuse studies. A majority of SUD patients want to be involved to some extent in aftercare treatments to client characteristics. Addict. Behav. 27 (4), 585–604. the decision-making process about their treatment, and perceive that https://doi.org/10.1016/s0306-4603(01)00195-2. their desire is respected by psychiatrists. Results suggest that it is not the Charles, C., Gafni, A., Whelan, T., 1999. Decision-making in the physician-patient encounter: Revisiting the shared treatment decision-making model. Soc. Sci. Med. 49 preference nor the perception of SDM, but their matching, what might (5), 651–661. https://doi.org/10.1016/s0277-9536(99)00145-8 (1982). influence treatment outcomes. Contrary to other health conditions, Cuschieri, S., 2019. The STROBE guidelines. Saudi J. Anaesth. 13 (Suppl 1), S31–S34. where perceiving less involvement than desired may be more detri­ https://doi.org/10.4103/sja.SJA_543_18. Daigre, C., Rodríguez, L., Roncero, C., Palma-Álvarez, R.F., Perea-Ortueta, M., Sorribes- mental than the opposite, SUD patients who perceive more involvement Puertas, M., Martínez-Luna, N., Ros-Cucurull, E., Ramos-Quiroga, J.A., Grau- than desired might experience an excess of responsibility and control López, L., 2021. Treatment retention and abstinence of patients with substance use that could negatively influence their compliance with appointments and disorders according to addiction severity and psychiatry comorbidity: A six-month substance use. More research is needed to confirm these results. In any follow-up study in an outpatient unit. Addict. Behav. 117, 106832 https://doi.org/ 10.1016/j.addbeh.2021.106832. case, exploring patient preferences for participating in treatment de­ Davis, E.L., Kelly, P.J., Deane, F.P., Baker, A.L., Buckingham, M., Degan, T., Adams, S., cisions and the degree of responsibility they wish to assume throughout 2020. The relationship between patient-centered care and outcomes in specialist the treatment,.will help professionals to adapt their promotion of SDM drug and alcohol treatment: A systematic literature review. Substance Abuse 41 (2), 216–231. https://doi.org/10.1080/08897077.2019.1671940. to the patients’ desires and characteristics, and to establish accepted De Las Cuevas, C., Mundal, I., Betancort, M., Lara-Cabrera, M.L., 2020. Assessment of therapeutic objectives. shared decision-making in community mental health care: Validation of the CollaboRATE. Int. J. Clin. Health Psychol. 20 (3), 262–270. https://doi.org/ 10.1016/j.ijchp.2020.06.004. Funding De Las Cuevas, C., Peñate, W., 2016. Validity of the Control Preferences Scale in patients with emotional disorders. Patient Prefer Adherence 10, 2351–2356. https://doi.org/ This work did not receive any specific grant from funding agencies in 10.2147/PPA.S122377. De las Cuevas, C., Peñate, W., de Rivera, L., 2014. Psychiatric patients’ preferences and the public, commercial, or not-for-profit sectors. experiences in clinical decision-making: Examining concordance and correlates of patients’ preferences. Patient Educ. Couns. 96 (2), 222–228. https://doi.org/ Declaration of Competing Interest 10.1016/j.pec.2014.05.009. De las Cuevas, C., Perestelo-Perez, L., Rivero-Santana, A., Cebolla-Martí, A., Scholl, I., Härter, M., 2015. Validation of the Spanish version of the 9-item Shared Decision- Dr. Serrano- Pérez has received fees as spekaer for Angelini, Casen Making Questionnaire. Health Expect 18 (6), 2143–2153. https://doi.org/10.1111/ Recordati and Lundbeck. Dr. Palma-Álvarez has received fees as speaker hex.12183. Degner, L.F., Sloan, J.A., Venkatesh, P., 1997. The Control Preferences Scale. Can. J. for Angelini, Casen Recordati, Exeltis, Lundbeck, MSD, Mundipharma, Nurs. Res. 29 (3), 21–43. and Takeda. Dr. Ramos-Quiroga has received fees as speaker from First, M.B., & Gibbon, M. (2004). The Structured Clinical Interview for DSM-IV Axis I Janssen-Cilag, Shire, Lilly, Ferrer, Medice, and Rubió. He has received Disorders (SCID-I) and the Structured Clinical Interview for DSM-IV Axis II Disorders research funding from JanssenCilag, Lilly, Ferrer, Lundbeck, and Rubió. (SCID-II). En Comprehensive handbook of Psychological assessment, Vol. 2: Personality assessment (pp. 134–143). John Wiley & Sons, Inc. Dr. Grau-López has received fees to give talks for Janssen-Cilag, Lund­ Friedmann, P.D., Hendrickson, J.C., Gerstein, D.R., Zhang, Z., 2004. The effect of beck, Servier, Otsuka, and Pfizer. The rest of the authors have no in­ matching comprehensive services to patients’ needs on drug use improvement in terests to declare. addiction treatment. Addiction 99 (8), 962–972. https://doi.org/10.1111/j.1360- 0443.2004.00772.x. Friedrichs, A., Silkens, A., Reimer, J., Kraus, L., Scherbaum, N., Piontek, D., Röhrig, J., Acknowledgements Hempleman, J., Härter, M., Buchholz, A., 2018. Role preferences of patients with alcohol use disorders. Addict. Behav. 84, 248–254. https://doi.org/10.1016/j. addbeh.2018.05.002. The authors wish to acknowledge the work of the whole team of the Friedrichs, A., Spies, M., Härter, M., Buchholz, A., 2016. Patient Preferences and Shared outpatient center for SUD treatment of Hospital Universitari Vall Decision Making in the Treatment of Substance Use Disorders: A Systematic Review d’Hebron for their support in the recruitment. of the Literature. PLoS One 11 (1), e0145817. https://doi.org/10.1371/journal. pone.0145817. Gomà-i-Freixanet, M., Valero, S., Muro, A., Albiol, S., 2008. Zuckerman-Kuhlman Supplementary materials Personality Questionnaire: Psychometric properties in a sample of the general population. Psychol. Rep. 103 (3), 845–856. https://doi.org/10.2466/ pr0.103.3.845-856. Supplementary material associated with this article can be found, in 9 P. Serrano-Pérez et al. Psychiatry Research 329 (2023) 115540 Grau-López, L., Roncero, C., Daigre, C., Gonzalvo, B., Bachiller, D., Rodríguez-Cintas, L., among patients seeking treatment for opioid use disorder: A needs assessment for Egido, Á., Casas, M., 2012. Risk factors for relapse in drug-dependent patients after decision making support. J. Subst. Abuse Treat. 100, 18–28. https://doi.org/ hospital detoxification. Adicciones 24, 115–122. 10.1016/j.jsat.2019.01.019. Green, C.A., Perrin, N.A., Polen, M.R., Leo, M.C., Hibbard, J.H., Tusler, M., 2010. Neumann, T., Neuner, B., Weiss-Gerlach, E., Tønnesen, H., Gentilello, L.M., Wernecke, K. Development of the Patient Activation Measure for mental health. Adm. Policy Ment. D., Schmidt, K., Schröder, T., Wauer, H., Heinz, A., Mann, K., Müller, J.M., Haas, N., Health 37 (4), 327–333. https://doi.org/10.1007/s10488-009-0239-6. Kox, W.J., Spies, C.D., 2006. The effect of computerized tailored brief advice on at- Hamann, J., Neuner, B., Kasper, J., Vodermaier, A., Loh, A., Deinzer, A., Heesen, C., risk drinking in subcritically injured trauma patients. J. Trauma 61 (4), 805–814. Kissling, W., Busch, R., Schmieder, R., Spies, C., Caspari, C., Härter, M., 2007. https://doi.org/10.1097/01.TA.0000196399.29893.52. Participation preferences of patients with acute and chronic conditions. Health Neuner, B., Dizner-Golab, A., Gentilello, L.M., Habrat, B., Mayzner-Zawadzka, E., Expect. 10 (4), 358–363. https://doi.org/10.1111/j.1369-7625.2007.00458.x. Górecki, A., Weiss-Gerlach, E., Neumann, T., Schlattmann, P., Perka, C., Spies, C.D., Hodgins, D.C., Leigh, G., Milne, R., Gerrish, R., 1997. Drinking goal selection in 2007. Trauma patients’ desire for autonomy in medical decision making is impaired behavioral self-management treatment of chronic alcoholics. Addict. Behav. 22 (2), by smoking and hazardous alcohol consumption—A bi-national study. J. Int. Med. 247–255. https://doi.org/10.1016/S0306-4603(96)00013-5. Res. 35 (5), 609–614. https://doi.org/10.1177/147323000703500505. Hser, Y.I., Polinsky, M.L., Maglione, M., Anglin, M.D., 1999. Matching clients’ needs with Newland, P., Lorenz, R., Oliver, B.J., 2021. Patient activation in adults with chronic drug treatment services. J. Subst. Abuse Treat. 16 (4), 299–305. https://doi.org/ conditions: A systematic review. J. Health Psychol. 26 (1), 103–114. https://doi.org/ 10.1016/s0740-5472(98)00037-3. 10.1177/1359105320947790. Joosten, E.A.G., de Jong, C.A.J., de Weert-van Oene, G.H., Sensky, T., van der Staak, C.P. Palma-Álvarez, R.F., Ros-Cucurull, E., Daigre, C., Perea-Ortueta, M., Martínez-Luna, N., F, 2009. Shared decision-making reduces drug use and psychiatric severity in Serrano-Pérez, P., Sorribes-Puertas, M., Ramos-Quiroga, J.A., Grau-López, L., substance-dependent patients. Psychother. Psychosom. 78 (4), 245–253. https://doi. Roncero, C., 2021. Is alexithymia related to retention and relapses in patients with org/10.1159/000219524. substance use disorders?: A one year follow-up study. Addict. Behav. 113, 106681 Joosten, E.A.G., De Jong, C.A.J., de Weert-van Oene, G.H., Sensky, T., van der Staak, C.P. https://doi.org/10.1016/j.addbeh.2020.106681. F, 2011. Shared decision-making: Increases autonomy in substance-dependent Pirie, T., National Treatment Indicators Working Group, 2015. National Treatment patients. Subst. Use Misuse 46 (8), 1037–1038. https://doi.org/10.3109/ Indicators Report: 2012–2013 Data. Canadian Centre on Substance Abuse, Ottawa, 10826084.2011.552931. Ontario. https://www.researchgate.net/publication/279957627_National_Treatme Knobel, H., Alonso, J., Casado, J.L., Collazos, J., González, J., Ruiz, I., Kindelan, J.M., nt_Indicators_Report_2012-2013_Data. Carmona, A., Juega, J., Ocampo, A., GEEMA Study Group, 2002. Validation of a Rodriguez, H.P., Poon, B.Y., Wang, E., Shortell, S.M., 2019. Linking Practice Adoption of simplified medication adherence questionnaire in a large cohort of HIV-infected Patient Engagement Strategies and Relational Coordination to Patient-Reported patients: The GEEMA Study. AIDS 16 (4), 605–613. https://doi.org/10.1097/ Outcomes in Accountable Care Organizations. Milbank Q. 97 (3), 692–735. https:// 00002030-200203080-00012. doi.org/10.1111/1468-0009.12400. Kokkevi, A., Hartgers, C., 1995. EuropASI: European Adaptation of a Multidimensional Roncero, C., Palma-Álvarez, R.F., Díaz-Morán, S., Grau-López, L., Rodríguez-Cintas, L., Assessment Instrument for Drug and Alcohol Dependence. Eur. Addict. Res. 1 (4), Ros-Cucurull, E., Álvarez, A.I., Casas, M., Daigre, C., 2019. Cocaine relapse and 208–210. https://doi.org/10.1159/000259089. health-related quality of life: A 23 weeks study. Actas Esp. Psiquiatr. 47 (2), 37–44. Kriston, L., Scholl, I., Hölzel, L., Simon, D., Loh, A., Härter, M., 2010. The 9-item Shared Ros-Cucurull, E., Palma-Álvarez, R.F., Cardona-Rubira, C., García-Raboso, E., Jacas, C., Decision Making Questionnaire (SDM-Q-9). Development and psychometric Grau-López, L., Abad, A.C., Rodríguez-Cintas, L., Ros-Montalbán, S., Casas, M., properties in a primary care sample. Patient Educ. Couns. 80 (1), 94–99. https://doi. Ramos-Quiroga, J.A., Roncero, C., 2018. Alcohol use disorder and cognitive org/10.1016/j.pec.2009.09.034. impairment in old age patients: A 6 months follow-up study in an outpatient unit in Luty, J., 2004. Treatment preferences of opiate-dependent patients. Ment. Hosp. 28 (2), Barcelona. Psychiatry Res. 261, 361–366. https://doi.org/10.1016/j. 47–50. https://doi.org/10.1192/PB.28.2.47. psychres.2017.12.069. Lynch, A.C., Weber, A.N., Hedden, S., Sabbagh, S., Arndt, S., Acion, L., 2021. Three- Saheb Kashaf, M., McGill, E.T., Berger, Z.D., 2017. Shared decision-making and month outcomes from a patient-centered program to treat opioid use disorder in outcomes in type 2 diabetes: A systematic review and meta-analysis. Patient Educ. Iowa, USA. Subst. Abuse Treat. Prev. Policy 16 (1). https://doi.org/10.1186/ Couns. 100 (12), 2159–2171. https://doi.org/10.1016/J.PEC.2017.06.030. S13011-021-00342-5. Schwartz, R.P., Kelly, S.M., Mitchell, S.G., Gryczynski, J., O’Grady, K.E., Gandhi, D., Marchand, K., Beaumont, S., Westfall, J., MacDonald, S., Harrison, S., Marsh, D.C., Olsen, Y., Jaffe, J.H., 2017. Patient-centered methadone treatment: A randomized Schechter, M.T., Oviedo-Joekes, E., 2019. Conceptualizing patient-centered care for clinical trial. Addiction 112 (3), 454–464. https://doi.org/10.1111/add.13622. substance use disorder treatment: Findings from a systematic scoping review. Subst. Sobell, M.B., Sobell, L.C., Bogardis, J., Leo, G.I., Skinner, W., 1992. Problem drinkers’ Abuse Treat. Prev. Policy 14 (1), 37. https://doi.org/10.1186/s13011-019-0227-0. perceptions of whether treatment goals should be self-selected or therapist-selected. Marlowe, D.B., Kirby, K.C., Festinger, D.S., Merikle, E.P., Tran, G.Q., Platt, J.J., 2003. Behav. Thera. 23 (1), 43–52. https://doi.org/10.1016/S0005-7894(05)80307-7. Day treatment for cocaine dependence: Incremental utility over outpatient Sterling, R.C., Gottheil, E., Glassman, S.D., Weinstein, S.P., Serota, R.D., 1997. Patient counseling and voucher incentives. Addict. Behav. 28 (2), 387–398. https://doi.org/ treatment choice and compliance. Data from a substance abuse treatment program. 10.1016/s0306-4603(01)00248-9. Am. J. Addict. 6 (2), 168–176. Mead, N., Bower, P., 2000. Patient-centredness: A conceptual framework and review of Vilagut, G., Ferrer, M., Rajmil, L., Rebollo, P., Permanyer-Miralda, G., Quintana, J.M., the empirical literature. Soc. Sci. Med. 51 (7), 1087–1110. https://doi.org/10.1016/ Santed, R., Valderas, J.M., Ribera, A., Domingo-Salvany, A., Alonso, J., 2005. The S0277-9536(00)00098-8. Spanish version of the Short Form 36 Health Survey: A decade of experience and new Messas, G.P., Fukuda, L., Pienkos, E., 2019. A Phenomenological Contribution to developments. Gac. Sanit. 19 (2), 135–150. https://doi.org/10.1157/13074369. Substance Misuse Treatment: Principles for Person-Centered Care. Psychopathology Ware, J.E., Sherbourne, C.D., 1992. The MOS 36-item short-form health survey (Sf-36): I. 1–9. https://doi.org/10.1159/000501509. conceptual framework and item selection. Med. Care 30 (6), 473–483. https://doi. Montori, V.M., Brito, J.P., Murad, M.H., 2013. The optimal practice of evidence-based org/10.1097/00005650-199206000-00002. medicine: Incorporating patient preferences in practice guidelines. JAMA 310 (23), World Drug Report, 2020. https://www.wdr.unodc.org/wdr2020/en/index2020.htmll. 2503–2504. https://doi.org/10.1001/jama.2013.281422. World Medical Association, 2013. World Medical Association Declaration of Helsinki: Moreno-Chico, C., González-de Paz, L., Monforte-Royo, C., Arrighi, E., Navarro-Rubio, M. Ethical principles for medical research involving human subjects. JAMA 310 (20), D., Gallart Fernández-Puebla, A., 2017. Adaptation to European Spanish and 2191–2194. https://doi.org/10.1001/jama.2013.281053. psychometric properties of the Patient Activation Measure 13 in patients with Yennurajalingam, S., Rodrigues, L.F., Shamieh, O.M., Tricou, C., Filbet, M., Naing, K., chronic diseases. Fam. Pract. 34 (5), 627–634. https://doi.org/10.1093/fampra/ Ramaswamy, A., Perez-Cruz, P.E., Bautista, M.J.S., Bunge, S., Muckaden, M.A., cmx022. Fakrooden, S., Sewram, V., Tejedor, A.N., Rao, S.S., Williams, J.L., Liu, D.D., Mundal, I., Lara-Cabrera, M.L., Betancort, M., De Las Cuevas, C, 2021. Exploring patterns Park, M., Lu, Z., Bruera, E., 2018. Decisional control preferences among patients in psychiatric outpatients’ preferences for involvement in decision-making: A latent with advanced cancer: An international multicenter cross-sectional survey. Palliat. class analysis approach. BMC Psychiatry 21 (1), 133. https://doi.org/10.1186/ Med. 32 (4), 870–880. https://doi.org/10.1177/0269216317747442. s12888-021-03137-x. Yun, D.W., Choi, J.S., 2019. Person-centered rehabilitation care and outcomes: A Muthulingam, D., Bia, J., Madden, L.M., Farnum, S.O., Barry, D.T., Altice, F.L., 2019. systematic literature review. Int. J. Nurs. Stud. 93, 74–83. https://doi.org/10.1016/ Using nominal group technique to identify barriers, facilitators, and preferences J.IJNURSTU.2019.02.012. 10