Objective. This study, based on recent advances in person-centered analytic methodology, aims to examine psychiatric outpatients’ preferences of participation on decision making.
Methods. Multiple Correspondence Analysis (MCA) was performed with socio-demographic and clinical data for 1,291 consecutive psychiatric outpatients who completed the Control Preferences Scale (CPS).
Results. MCA showed a bidimensional solution. First dimension (eigenvalue 1.723; Cronbach’s alpha .525; inertia value .345) explained 57% of variance while second dimension (eigenvalue 1.282; Cronbach’s alpha .275; inertia value .256) corresponds to 42% of explained variance. The MCA revealed that psychiatric diagnosis was the main factor explaining the variance. Three profiles of patients were evidenced. Passive patients were more often women, over 55 years, with a primary education, and more frequently diagnosed with depressive disorders. Collaborative patients usually were women between 44 and 55 years, with secondary education, and mainly diagnosed with anxiety disorders, but also with bipolar disorder (in one case with a university degree). Active patients were male under 44 years most often diagnosed with schizophrenia.
Conclusion. MCA has proven its usefulness and applicability to study control preferences of psychiatric patients. The differentiation of patient profiles should lead to the development of profile-specific interventions to promote participation in decision making.
Keywords: Psychiatric outpatient, multiple correspondence analysis, control preference scale, preferences of participation, decision making.
Citation: De las Cuevas C, Peñate W, Betancort M. Person-centered approach to control preferences in psychiatric outpatients. J Participat Med. 2016 Oct 12; 8:e13.
Published: October 12, 2016.
Competing Interests: The authors have declared that no competing interests exist.
Funding Source: This work was supported by Instituto de Salud Carlos III, FEDER (Fondo Europeo de Desarrollo Regional) Unión Europea (PI10/00955).


In the last thirty years, patient empowerment has become a strategic issue in health policy of many western developed countries. [1] Shared decision making (SDM) is an evidence-based approach to treatment decision making that also allows for client preferences to be considered. [2] In psychiatric clinical practice, the SDM model supports the values of choice, self-determination, and empowerment for psychiatric patients and puts into use the basic rights of this group of patients who have not sufficiently benefited from consumer empowerment apparent in other medical fields. [3] In addition, given the evidence that active patient participation in decision making and management of chronic conditions such as psychiatric disorders improves health outcomes, [4][5] psychiatrists should routinely take a supportive and encouraging approach to shared decision making and treatment planning. However, the literature shows a wide variation in reporting the extent to which psychiatric patients prefer to be involved. [2][6][7][8] For these reasons, the way treatment decisions are made and the degree of involvement that patients wish to have in this process is thus an important area of research.[9]

Previous research has shown that SDM patterns are different in the primary care and psychiatric outpatient care settings, reflecting quite a different perspective on a complex decision making process that needs to be analyzed according to its different steps. [10] It has also been shown that there is low concordance between psychiatric outpatients’ preferred role and the roles they experience, indicating that patients may desire a more active role in participation on clinical decisions than they actually play. [11][12] Congruence between patients’ preferences and actual experiences for their level of participation in shared decision making is relevant for their adherence to treatment. [13]

However, previous studies on this subject have used variable-centered analyses that treat each variable (or characteristic) studied as related to another variable in contrast to new approaches as person-centered analyses that investigate how these variables group within individuals. Variable-centered approaches presume that variables operate the same way for all individuals in a largely homogenous population. However, person-centered models reject this assumption and seek to identify variations in how variables are associated within groups of individuals, considering individuals as the ultimate predictors of outcomes; variables describe individuals but do not explain development. [14] In summary, person-centered approaches focus on identifying groups of persons with important common characteristics and considering how these groups differ. [15]

The present study based on recent advances in person-centered analytic methodology aims to examine psychiatric outpatients’ preferences of participation on decision making paying special attention to patients’ socio-demographic and clinical variables.



From October 2013 to November 2015, one thousand six hundred consecutive psychiatric outpatients at two Community Mental Health Centers of Canary Islands Health Service on Tenerife Island (Canary Islands, Spain) were invited to participate in a cross-sectional study; a total of 1291 accepted. Patients were eligible for inclusion in the study if they were aged 18 and over and were diagnosed by their psychiatrists using the International Classification of Diseases, Tenth Edition (ICD-10) as F20 (schizophrenia), F31 (bipolar affective disorder), F32-33 (depressive disorder), F40-48 (anxiety disorders), and F60-69 (disorders of adult personality and behavior). Each participant received a full explanation of the study, after which they signed an informed consent document approved by the clinical research ethics committee of Nuestra Señora de Candelaria University Hospital in Santa Cruz de Tenerife. Each participant then filled out a brief socio-demographic and clinical survey and the control preferences scale. Age, sex, educational level (primary studies, secondary studies, and university degree), diagnoses, time under psychiatric treatment, and number of different drugs used were registered.


The Control Preferences Scale (CPS) is the most frequently used measure of patients’ preferred roles in treatment decisions. [16] It was developed by Degner and colleagues to evaluate the amount of control patients want to assume in the process of making decisions about the treatment of their diseases. [17] A card-sorting version of the scale was used in our study. It consists of five “cards” on a board, each illustrating a different role in decision making by means of a cartoon and short descriptive statement. The examiner asks the respondent to choose the preferred card, which is then covered up and cannot be chosen again; the examiner then asks the respondent to choose the preferred card from the remaining four cards. The procedure continues (four choices) until one card is left. If the second preference is incongruous with the first (nonadjacent pairing, such as card A with card C), the test is explained again, and immediately readministered. In the event of a further incongruity, the test is not readministered, and a preference is not assigned. Administration requires about 5 minutes. Six scores are possible based on the subject’s two most preferred roles: active–active, active–collaborative, collaborative–active, collaborative–passive, passive–collaborative, and passive–passive. These scores are grouped as: active (active–active or active–collaborative), collaborative (collaborative–active or collaborative–passive), or passive (passive–collaborative or passive–passive).

Statistical Methods

Chi-square analyses were performed to contrast the differences among categorical variables. In order to analyze the pattern of relationships of the several categorical dependent variables considered, a multiple correspondence analysis (MCA) was carried out. MCA is part of a family of descriptive methods (eg, clustering, factor analysis, and principal component analysis (PCA)) that reveal patterning in complex datasets. However, specifically, MCA is used to represent and model datasets as “clouds” of points in a multidimensional Euclidean space; this means that it is distinctive in describing the patterns geometrically by locating each variable/unit of analysis as a point in a low-dimensional space. The results are interpreted on the basis of the relative positions of the points and their distribution along the dimensions; as categories become more similar in distribution, the closer (distance between points) they are represented in space. [18]


We recorded a high response rate of 80% resulting in a sample of 1291 psychiatric outpatients. Table 1 shows the sample distribution according to socio-demographic and clinical variables included in the study as well as the preferred roles according Control Preferences Scale.

Table 1. Sociodemographic and clinical characteristics of the sample studied.
Abbreviations: ICD, International Classification of Diseases

Most psychiatric outpatients (752 patients, 58.2%) preferred shared decisional control, while 418 (32.4%) preferred a passive approach and only 121 (9.4%) an active decisional control. The most common preferred role was the collaborative-passive (53.1%) where doctor and patient share responsibility for deciding what treatment is best with the doctor making the final decision after considering the patient’s opinion.

The study of the relationship between preference options (the CPS responses were categorized as Active, Collaborative and Passive), and the rest of variables was performed using the chi-square test. Table 2 shows the contrasts of independence between the variables of interest and CPS. Test cases for significant relationships found were checked.

Table 2. Chi-square values (d.f. in parenthesis) for each variable of interest.
Note: CPS Control Preference Scale; ** less than .01; n.s. no significance

In order to establish profiles of response from Control Preference Scale (CPS) and the significant variables (gender, age, educational level and diagnoses), a Multiple Corresponde Analysis (MCA) was conducted. The CPS responses were categorized as Active, Collaborative and Passive. The age variable was categorized into three groups (18-43, 44-55, and 55 years older) according to percentiles in order to balance the sample in those categories. Diagnoses variable included five categories: schizophrenia, bipolar disorder, depressive disorder, anxiety disorder, personality disorder, and other diagnoses. The variable level of education had three levels: primary studies, secondary studies and university studies.

Table 3. Discrimination measures in dimensions of Multiple Correspondence Analysis.

MCA showed a bidimensional solution after 64 interactions from a useful sample that included 1150 cases. This analysis isolated two dimensions. The first dimension with an eigenvalue of 1.723 reached a value of Cronbach’s alpha of 0.525 and inertia value (variability index) of 0.345 which corresponds to 57% of explained variance. The second dimension with an eigenvalue of 1.282 reached a value of Cronbach’s alpha of 0.275 and 0.256 inertia value corresponding to 42% of explained variance. Table 3 shows the contribution of each variable to isolated dimensions. As can been observed the variables that most contribute to dimension 1 are the gender, age and diagnoses while the variables that most contribute to the dimension 2 are the level of education and CPS.

Correspondence analysis of these data yields the graphical display shown in Figure 1 that represent the two-dimensional graph generated from the estimated model showing the weight of the above variables and their contribution to explain the spread of the data. Three profiles of patients were evidenced when considering the relationships of control preference scale with the variables of interest. Passive patients were more often women, older than 55 years, with a level of primary education, and more frequently diagnosed with depressive disorders. Collaborative patients were usually women, between 44 and 55 years old, who had completed secondary studies. Most of these patients had anxiety disorders, but some also had bipolar disorder (in this case the level of education rises to university degree). Finally, we found that the active patients were usually men, with an age below 44 years, a level of secondary education and most often diagnosed with schizophrenia.

Figure 1. Dimensional representation of the multiple correspondence analysis of Control Preferences Scale Data. The graphic represents the frequencies of co-occurrences between categorical variables in a bi-dimensional space. The criterion used to develop profiles was the proximity of such joint frequencies. The closer (distance between points) the variables (their response categories) they are grouped in a co-occurrence profile.


The goal of this study was to determine the applicability and usefulness of multiple correspondence analysis (MCA) in detecting and representing underlying profiles in a large dataset used to investigate sociodemographic and clinical variables that could explain psychiatric outpatients preferences for participation in decision making about their treatment. MCA has proved to be a powerful exploratory technique that provides groupings of categories of variables in the dimensional spaces, offering important insights on relationships between categories or, in other words, multivariate treatment of the data through simultaneous consideration of multiple categorical variables. [18]

Previous studies have shown that patients’ preferences for involvement in treatment decision making change depending on the context, as do factors associated with role preferences. [19] The process of developing them is likely to be highly complex. [20]

The MCA revealed that psychiatric diagnosis was the main factor explaining the variance. The visual representation of this quantitative analysis led to the distinction between the three groups of preference of participation. According to our results, women prefer passive and collaborative roles in medical decision making. The difference is determined by the age and the diagnoses of the patient: depression in older ages than 55 years was associated with a passive profile, and anxiety in ages between 44 and 55 with the collaborative one. Active role preference was associated with men younger than 43 years diagnosed with schizophrenic disorders.

With the exception of age, our results do not agree well with previous literature showing that younger age, female gender and higher education are the most frequently reported predictors for active participation preferences. [21]

Given the range and complexity of evaluating SDM, there are so far no generally applicable primary measurement tools or standard outcome measures, which results in inconsistent measurement, making the comparability of research results difficult. [23]

Our study shows that patients with schizophrenia have a greater desire to participate in decision making about their treatment than patients with other psychiatric conditions. At first sight, this result could be in accordance with the hypothesis that chronic psychiatric conditions might yield higher participation preferences than acute ones. However, our results also indicate that neither time under psychiatric treatment nor number of psychoactive drugs used was associated with preferences of participation. Moreover, other chronic conditions such as bipolar disorder were linked to a more collaborative preference of participation. The preference for an active role by schizophrenic patients is consistent with previous results from Hamann and colleagues. [24] They found that inpatients with schizophrenia expressed a desire for active participation in decisions about their treatment that was partly explained by their negative attitudes toward neuroleptic medication and possibly by the nature of schizophrenia (eg, the fear of loss of control that patients could experience).

This study has certain limitations: First, this is a correlational cross-sectional study devised to investigate psychiatric patients’ preferences for participation in decision making at a single point in time. This design limits the degree to which causal inferences and generalizations can be made from the research findings. Secondly, another limitation could be the use of self-reported questionnaires to assess patients’ control preference. This might be affected by participant motivation, poor recall and social desirability in responding. On the other hand, the strengths of the present study are founded in the large sample size studied, the possibility of adjusting for socio-demographic and clinical variables, and in the existence of well-defined groups of patients with different psychiatric diagnoses.


The person-centered analytic methodology using “the person” as the unit of analysis and focused on the search for holistic patterns among mutually similar person profiles is a promising approach to investigate the association structures and for fast clustering purposes in the field of mental health. The differentiation of patient profiles should lead to the development of profile-specific interventions to promote participation in decision making.


  1. Edgman-Levitan S, Brady C, Howitt P. Partnering with Patients, Families, and Communities for Health: A Global Imperative. Doha: World Innovation Summit for Health (WISH); 2013.
  2. Simmons M, Rice S, Hetrick S, Bailey A, Parker A. Evidence summary: Shared decision making (SDM) for mental health—what is the evidence? Melbourne: Orygen Youth Health Research Centre; 2012.
  3. Hamann J, Leucht S, Kissling W. Shared decision making in psychiatry. Acta Psychiatr Scand 2003;107:403-9.
  4. Joosten EA, DeFuentes-Merillas L, de Weert GH, Sensky T, van der Staak CP, de Jong CA. Systematic review of the effects of shared decision-making on patient satisfaction, treatment adherence, and health status. Psychother Psychosom 2008;77:219–226.
  5. Shay LA, Lafata JE. Where is the evidence? A systematic review of shared decision making and patient outcomes. Med Decis Making 2015;35:114-31.
  6. Flynn KE, Smith MA, Vanness D. A typology of preferences for participation in healthcare decision-making. Soc Sci Med 2006;63:1158-1169.
  7. O’Neal EL, Adams JR, McHugo GJ, Van Citters AD, Drake RE, Bartels SJ. Preferences of older and younger adults with serious mental illness for involvement in decision-making in medical and psychiatric settings. Am J Geriatr Psychiatry 2008;16:826-33.
  8. De las Cuevas C, Rivero A, Perestelo-Perez L, Gonzalez M, Perez J, Peñate W. Psychiatric patients’ attitudes towards concordance and shared decision making. Patient Educ Couns 2011;85:e245-50.
  9. Efficace F, Gaidano G, Sprangers M, Cottone F, Breccia M, Voso MT, Caocci G, Stauder R, Di Tucci AA, Sanpaolo G, Selleslag D, Angelucci E, Platzbecker U, Mandelli F. Preference for involvement in treatment decisions and request for prognostic information in newly diagnosed patients with higher-risk myelodysplastic syndromes. Ann Oncol 2014;25:447-54.
  10. De Las Cuevas C, Peñate W, Perestelo-Pérez L, Serrano-Aguilar P. Shared decision making in psychiatric practice and the primary care setting is unique, as measured using a 9-item Shared Decision Making Questionnaire (SDM-Q-9). Neuropsychiatr Dis Treat 2013;9:1045-52
  11. De las Cuevas C, Peñate W, de Rivera L. Psychiatric patients’ preferences and experiences in clinical decision-making: examining concordance and correlates of patients’ preferences. Patient Educ Couns 2014;96:222-8.
  12. De las Cuevas C, Peñate W. To what extent psychiatric patients feel involved in decision making about their mental health care? Relationships with socio-demographic, clinical, and psychological variables. Acta Neuropsychiatr 2014;26:372-81.
  13. De Las Cuevas C, Peñate W, de Rivera L. To what extent is treatment adherence of psychiatric patients influenced by their participation in shared decision making? Patient Prefer Adherence 2014;8:1547-53.
  14. Laursen B, Hoff E. Person-Centered and Variable-Centered Approaches to Longitudinal Data. Merrill-Palmer Quarterly 2006; 52:377-389.
  15. Magnusson, D. The logic and implications of a person-oriented approach. In R. B. Cairns, L. R. Bergman, & J. Kagan (Eds.), Methods and models for studying the individual (pp. 33–64). Thousand Oaks, CA: Sage, 1998.
  16. Solari A, Giordano A, Kasper J, Drulovic J, van Nunen A, Vahter L, Viala F, Pietrolongo E, Pugliatti M, Antozzi C, Radice D, Köpke S, Heesen C. Role Preferences of People with Multiple Sclerosis: Image-Revised, Computerized Self-Administered Version of the Control Preference Scale. PLoS One 2013;8:e66127.
  17. Degner LF, Sloan JA, Venkatesh P. The Control Preferences Scale. Can J Nurs Res. 1997;29:21–43.
  18. Costa P, Santos N, Cunha P, Cotter J, Sousa N: The Use of Multiple Correspondence Analysis to Explore Associations between Categories of Qualitative Variables in Healthy Ageing. J Aging Res 2013;2013:12.
  19. O’Donnell M, Hunskaar S: Preferences for involvement in treatment decision-making generally and in hormone replacement and urinary incontinence treatment decision-making specifically. Patient Educ Couns 2007;68:243-251.
  20. Say R, Murtagh M, Thomson R: Patients’ preference for involvement in medical decision making: A narrative review. Patient Educ Couns 2006;60:102-114.
  21. Benbassat J, Pilpel D, Tidhar M. Patients’ preferences for participation in clinical decision making: a review of published surveys. Behav Med 1998;24:81–8.
  22. Hamann J, Neuner B, Kasper J, Vodermaier A, Loh A, Deinzer A, Heesen C, Kissling W, Busch R, Schmieder R, Spies C, Caspari C, Härter M. Participation preferences of patients with acute and chronic conditions. Health Expect 2007;10:358-63.
  23. Scholl I, Koelewijn-van Loon M, Sepucha K, Elwyn G, Légaré F, Härter M, Dirmaier J. Measurement of shared decision making – a review of instruments. Z Evid Fortbild Qual Gesundhwes. 2011;105(4):313-24.
  24. Hamann J, Cohen R, Leucht S, Busch R, Kissling W. Do patients with schizophrenia wish to be involved in decisions about their medical treatment? Am J Psychiatry 2005;162:2382-4.

Copyright: © 2016 Carlos De las Cuevas, Wenceslao Peñate, and Moisés Betancort. Published here under license by The Journal of Participatory Medicine. Copyright for this article is retained by the authors, with first publication rights granted to the Journal of Participatory Medicine. All journal content, except where otherwise noted, is licensed under a Creative Commons Attribution 3.0 License. By virtue of their appearance in this open-access journal, articles are free to use, with proper attribution, in educational and other non-commercial settings.