Characterizing outcome preferences in patients with psychotic disorders: a discrete choice conjoint experiment☆
Introduction
Optimizing the treatment of individuals experiencing a first episode of schizophrenia has become an important mental health priority internationally (Jackson and McGorry, 2009, Kane et al., 2016, Zipursky and Schulz, 2002). We now appreciate that approximately 80% of individuals who experience a first episode of schizophrenia will achieve a remission of their symptoms in their first year of treatment (Lieberman et al., 1993). However, it is estimated that only one out of seven patients will meet criteria for recovery when it is defined as requiring both sustained improvement in symptoms and in functioning (Jaaskelainen et al., 2013). Greater understanding of the factors that contribute to the gap between rates of remission and recovery is needed (Zipursky et al., 2013). Characterizing the outcomes that are of importance to patients may contribute to our understanding of the outcomes currently observed and provide new insights about how to improve outcomes in the future.
Outcomes found to be of highest priority to patients with schizophrenia have varied greatly across studies and have included better social support and housing (Fischer et al., 2002), improved functioning (Shumway et al., 2003), reduced confusion (Rosenheck et al., 2005), and decreased positive symptoms (Levitan et al., 2015). Previous studies have often elicited priorities using ranking or rating tasks (Fischer et al., 2002). These approaches are vulnerable to demand characteristics that lead respondents to answer in ways that they feel are correct or socially desirable (Streiner and Norman, 2008). Conjoint analysis surveys, on the other hand, are more likely to reflect unconscious values that contribute substantially to real-world decision making (Caruso et al., 2009).
Conjoint analysis or Discrete Choice Experiment (DCE) methodology was developed in the fields of mathematical psychology, marketing, and economics and is increasingly utilized for understanding the healthcare preferences of consumers (Cunningham et al., 2008). Bridges et al. (2011b) demonstrated, in a “proof of principle” study, that when patients with schizophrenia were asked to participate in a choice-based conjoint task, they were able to complete the tasks, to articulate their preferences, and to make meaningful trade-offs between choices. In this study, we developed a DCE to characterize outcome preferences in patients receiving treatment for psychotic disorders. We were interested in investigating whether respondents were distributed into classes or segments with different outcome preferences (Hauber et al., 2016) and whether segment membership was associated with demographic and clinical measures.
Section snippets
Subjects
This study was conducted at St. Joseph’s Healthcare Hamilton, a tertiary care teaching hospital affiliated with the Michael G. DeGroote School of Medicine at McMaster University in Hamilton, Ontario, Canada. Participants were recruited between June and November of 2015 from the waiting rooms of two clinics that specialize in the treatment of psychotic disorders: the Schizophrenia Outpatient Clinic and the Cleghorn Early Intervention Clinic. As the study aimed to sample a representative group of
Patients
A total of 483 participants were approached for study participation; 305 participants provided informed consent and 300 completed the survey (62%). One third of the participants were recruited from the Cleghorn Early Intervention Clinic (n = 100) and two thirds from the Schizophrenia Outpatient Clinic (n = 200).
Latent class analysis
Latent class analysis demonstrated that a three-class model resulted in the lowest BIC and adjusted BIC values (Supplement Table 1). A minus 2 log-likelihood bootstrap difference test showed
Discussion
Advances in the treatment of schizophrenia have yet to translate into dramatic changes in the rates of recovery. This study was designed to investigate whether differences in patient preferences and priorities might contribute to the persisting gap between the high rates of symptomatic remission and low rates of recovery observed in patients with schizophrenia (Revier et al., 2015). In this study, we have found using a DCE that there is substantial variability in the outcome preferences
Conflict of interest
Dr. Zipursky has served as a consultant to Janssen, Roche, Otsuka, and Lundbeck and received research grant support from Janssen and Roche. Dr. Cunningham, Dr. McDermid Vaz, Ms. Stewart, Ms. Rimas, and Ms. Cole have no financial relationships with commercial interests.
Contributors
Dr. Zipursky and Dr. Cunningham were responsible for all aspects of the study that include designing the study, obtaining funding, analyzing the data, and writing the manuscript. Dr. McDermid Vaz, Ms. Stewart, and Ms. Cole were responsible for study implementation and management, and data collection. Ms. Stewart and Ms. Rimas contributed to the survey development and performed the statistical analyses. All authors contributed to the writing of the final version of the manuscript, and read and
Funding body agreements and policies
Support for this research was provided by the CARSTAR Automotive Canada Research Innovation Fund, St. Joseph's Healthcare Hamilton Foundation, and the Jack C. Laidlaw Chair in Patient-Centered Health Care (Dr. Cunningham), McMaster University.
Acknowledgments
We gratefully acknowledge the time and effort devoted to this study by all of the participants in our survey and focus groups. We would also like to thank the staff of the Schizophrenia Outpatient Clinic and Cleghorn Early Intervention Clinic at St. Joseph’s Healthcare Hamilton whose tremendous support and enthusiasm made this study possible.
References (37)
- et al.
Happiness in first-episode schizophrenia
Schizophr. Res.
(2012) - et al.
Conjoint analysis applications in health - a checklist: a report of the ISPOR Good Research Practices for Conjoint Analysis Task Force
Value Health
(2011) - et al.
Statistical methods for the analysis of discrete choice experiments: a report of the ISPOR Conjoint Analysis Good Research Practices Task Force
Value Health
(2016) Measuring happiness with a single-item scale
Soc. Behav. Personal.
(2006)- et al.
An introduction to latent variable mixture modeling (part 1): overview and cross-sectional latent class and latent profile analyses
J. Pediatr. Psychol.
(2014) - et al.
Can patients diagnosed with schizophrenia complete choice-based conjoint analysis tasks?
Patient
(2011) - et al.
Using conjoint analysis to detect discrimination: revealing covert preferences from overt choices
Soc. Cogn.
(2009) - et al.
Measuring socioeconomic status: reliability and preliminary validity for different approaches
Assessment
(2002) - et al.
Using conjoint analysis to model the preferences of different patient segments for attributes of patient-centered care
Patient
(2008) Understanding scores on the Satisfaction with Life Scale
The Satisfaction With Life Scale
J. Pers. Assess.
Sensitivity and specificity of information criteria
PeerJ
The Revised Behavior and Symptom Identification Scale (BASIS-R): reliability and validity
Med. Care
Priorities of consumers, providers, and family members in the treatment of schizophrenia
Psychiatr. Serv.
A systematic review and meta-analysis of recovery in schizophrenia
Schizophr. Bull.
The recognition and management of early psychosis: a preventive approach
Constructing experimental designs for discrete-choice experiments: report of the ISPOR Conjoint Analysis Experimental Design Good Research Practices Task Force
Value Health
Comprehensive versus usual community care for first-episode psychosis: 2-year outcomes from the NIMH RAISE early treatment program
Am. J. Psychiatry
Cited by (0)
- ☆
This paper was presented at the 5th Biennial Schizophrenia International Research Society Conference in Florence, Italy, April 2–6, 2016.