Research priorities set by people with OCD and OCD researchers: Do the commonalities outweigh the differences?

Abstract Objective In contrast to research agendas being predominantly set by scientists or funders, a collaborative approach was used to spot future goals for research on obsessive‐compulsive disorder. Methods First, we conducted a meta‐review and then compared the results of two online surveys with OCD professionals and patients on research priorities. The literature search was performed in three comprehensive databases, and ten research goals were extracted. Sixty‐four patients and eight professionals responded to open questions on their five most important goals. Then, they ranked the ten aims extracted from the literature on a 6‐point Likert scale. Results For patients and professionals, research on treatment gains that persist long‐term was most important. Concerning the top five goals listed in an open format, for patients, development and maintenance of the disease was as important as psychotherapy and its efficacy. In contrast, for professionals, the efficacy and the optimization of psychotherapy were the far most important research goals. Conclusions We proposed one possibility to involve patients in OCD research, and the multitude of answers presents a wealth of research ideas. Practice Implications Since consistent research involvement may contribute to its clinical impact, researchers are now invited to translate our findings into empirical studies.

collaborative process, 3 and involvement into research gives credit to patients' perspectives, interests and values. 4 For example, it aims at asking meaningful research questions, improving study feasibility, supporting the dissemination of relevant findings into clinical practice or improving health outcomes. 4,5 Against this backdrop, PI was acknowledged by the Lancet Psychiatry Commission on Psychological Treatments Research as one means for advancing psychotherapy research. 6 The scope of PI varies considerably, from organizational involvement to writing lay summaries, from single to continuous participation. 4 Considering Farin-Glattacker and colleagues' 7 matrix, we aimed at early involvement and at equal collaboration; that is, in our study, research priorities will be identified equally by researchers and patients, and patients will have the final say. This is also in line with results of a former own study, in which some patients encouraged earlier involvement, especially as to defining the research agenda. 4 One of the most prominent associations asking patients to prioritize research goals is the James Lind Alliance, which also published a guidebook to enhance identifying the ten most important research aims from patients, carers and professionals, also to inform funding agencies. 8 Although we had to align to the resources available within our institutions, we still got inspired by their inclusive, transparent and evidence-based approach.
Former PI studies in our field referred to research priorities in mental health in general 9 or to uncertainties in the treatment of specific subgroups, such as patients diagnosed with Parkinson's disease. 10 Regarding mental health, during the online survey conducted by Haarig and colleagues, 11 patients diagnosed with bipolar disorder were asked for the individually most important therapeutic goals. Coping with the symptoms of the disease was most important to the participants, and clearly inferior were improvement of quality of life, participation in psychotherapy itself and the management of adverse effects of the medication. 11 Another study focused on research questions of patients with depression. 12 To our knowledge, there is no PI study specifically on the psychotherapy for obsessive-compulsive disorder (OCD) and respective research priorities to date. Analysing former studies, Ennis & Wykes 5 pointed out that studies on difficult-to-treat populations such as patients diagnosed with developmental and personality disorders were associated with lower PI. Similar prejudices seem to persist on OCD, for example, that patients were rather difficult to treat, 13 which could also explain the scarcity of PI in this field.
OCD is characterized by recurrent and persistent thoughts, urges or images experienced as intrusive and unwanted causing anxiety or distress (obsessions), and consequently, repetitive behaviours or mental acts to reduce discomfort (compulsions). 14 Whereas untreated OCD tends to be a rather chronic disease, 15,16 a number of patients are not reached by existing evidence-based approaches, namely cognitive-behavioural therapy (CBT) or selective serotonin reuptake inhibitors (SSRI), and therapists do not always implement CBT as intended. [16][17][18] Although therapy improved during the last decades, relapse rates are still unsatisfactory. 16 Since there is still room for improvement in the psychotherapy for OCD, 19 PI has the potential to contribute to a meaningful research agenda. Further, at least in Germany, PI into psychotherapy research is still in its infancy. Furthermore, current therapy is stretched to its limits as prominent emotions in OCD, such as disgust, tend to habituate more slowly than anxiety, which may impair the efficacy of therapy if not considered sufficiently. 20 As outlined above, OCD differs from other anxiety disorders as it is still considered difficult to treat, due to other susceptibilities to change, through less appropriate care, and thus higher relapse rates and chronic trajectories. With that said, the aim was to spot future goals for research specifically on obsessive-compulsive disorder (OCD).
Therefore, we examined the current literature through a meta-review and surveyed professionals (Study 1) and patients (Study 2) with regard to psychotherapy and OCD-related research priorities.
Subsequently, we compared patients' and professionals' views in order to examine commonalities and differences and, thus, to support patient-focused research agendas.

| ME THODS
To derive research aims from the current literature, three databases (the International Prospective Register of Systematic Reviews no OCD (n = 7)); thus, n = 13 reviews were included from that search.
Searching the references of the German S3-Guideline on Obsessive- Compulsive Disorders 17 resulted in n = 6 additional reviews, so that N = 19 reviews were included in the meta-study. From these reviews (see Appendix S1), the top ten aims for future research on the psychotherapy of OCD were extracted from the discussion sections by one researcher (xx), discussed with another (xx), and then phrased as concise items for the subsequent surveys (Studies 2 and 3).

| Study 2: patient survey
To anonymously gather data on OCD patients' perspectives on psychotherapy research, we again conducted an independent Internetbased survey. Participants' health expertise was addressed, and

| Data analysis
Due to the different sample sizes between professionals and patients, we mainly used descriptive statistics. According to the means, standard deviations and ranges, the items were ranked in descending order and independently for patients and professionals. In line with Banfield et al 9 and due to skewness, we also dichotomized the scale (ie combined important and very important ratings versus all others) and ranked the items again. Due to the data structure and unequal sample sizes, we then examined differences between the central tendencies of patients and professionals via a non-parametric Mann-Whitney U-test, that is, one analysis over the 10 goals. To determine agreement among participants regarding the 10 goals, we used Krippendorff's α, a flexible reliability measure considering any number of categories, any number of judges and missing data, 21,22 that is implemented within an SPSS macro. 23 Krippendorff's α = 1 presents perfect agreement, whereas α = 0 defines its absence. 22 All analyses were performed using IBM SPSS Statistics 25 and Microsoft Excel at a 0.05 level of significance.
Open questions regarding the top 5 research priorities were analysed qualitatively using inductive content analysis techniques. 24 First, three researchers (xx, xx, xx) read all answers to familiarize with the data. Preliminary categories were developed separately for patients and professionals and then discussed to foster a common understanding. Following this, the final categories were derived by two independent researchers (xx, xx/xx). Inter-rater agreement reached κ = 0.81 (patients' category system) resp. κ = 0.92 (professionals' category system). To generate a hierarchy, the priorities given by the participants in the quantitative part were inverted (ie the individual priority 1 was inverted to 5 to give it the most weight) and summed up within each category. Then, for indications and priorities, their number per category was divided by their total number in order to obtain scores comparable between the two groups. Since for professionals, very similar qualitative answers were given with respect to both research questions (top 5 aims for OCD research in general and for research on psychotherapy with OCD patients in particular), we summarized the two into a joint category system. In the subsequent patient survey, we therefore focused on the psychotherapy question only.

| Goals from the meta-review
The research goals extracted from the current literature (Table 2, Appendix S1) concerned the following overarching topics: comparative effectiveness, mechanisms, moderators, administration, target groups, motivation, dissemination, quality of life and treatment gains that persist long-term.

| Quantitative ranking results
A total of 150 persons followed the link, and of these, N = 63 were included in the final analyses ( Table 1). The others were excluded because they did not give electronic informed consent (n = 22), were younger than 18 years (n = 5), did not proceed until the ranking questions (n = 59), or filled in the questionnaire twice (n = 1).
The professionals (N = 8) were mainly experienced psychologists and licensed psychotherapists practicing research, teaching and patient care.
According to the quantitative rankings of the 10 predefined aims by patients and professionals, the most prominent goal was doing research ensuring treatment gains that persist long-term (goal 7, Table 2). Both groups also agreed on the least important aim for future research, that is, the effectiveness of group versus individual psychotherapy (goal 5). For OCD patients, the comparative effec-  more consistently than the patients did, but this may also be attributable to small sample size. For the ten research goals, differences in the central tendencies between the two groups were non-significant (P>.05); that is, patients and professionals evaluated every goal as comparably important. Neither was there agreement among the patients (Krippendorff's α = .097) nor among the professionals (Krippendorff's α = .124) regarding prioritization of the 10 predefined goals.

| Qualitatively found priorities
Considering the open answers, that is, to name the five individually most important aims for psychotherapy in OCD, six (patients) and five (professionals) relatively comparable categories emerged (Table 3). Most often, patients indicated aspects that fell within the category 'Disease development and maintenance'.

| Discussion
In order to examine priorities for future OCD research, we conducted a meta-review and two subsequent online surveys with OCD patients and professionals. As one of the first OCD studies, we involved patients and combined quantitative and qualitative data. Whereas the development and maintenance of OCD and more therapy-related questions were central to patients, professionals prioritized future research on the efficacy and optimization of psychotherapy. Both patients and professionals indicated treatment gains that persist long-term as the most important research goal. In contrast, most current studies on psychotherapy and medication in OCD make use of short-term durations of approximately 12 weeks 25 or of on average 15-month follow-ups. 15 Since OCD is often, at least without treatment, proceeding chronically, 18   applicability of an intervention (eg in everyday life or in crises), better self-perception (eg coping with oneself) and empowerment (ie knowing what helps oneself) as important treatment outcomes. Interview studies or focus groups would be appropriate to follow these issues.
Regarding the other research goals, there was low agreement, both within and between patients and professionals. Still, most goals were assessed as highly relevant by both groups underscoring the findings from the meta-review. There were no significant differences between the means and ranks of the two groups which again indicate the comparable importance to patients and experts.
Nevertheless, the patients' rankings were distributed a bit more which may be attributable to more scepticism or to more problems with evaluating the relevance of every aspect. Subsequent studies using a forced choice approach may yield clearer results. 9 Associated with their roles, in the open format, patients listed goals that were more therapy-related, and professionals placed more emphasis on advancements in psychotherapy research.
Research trainings adapted to patient participants are a feasible method to enable them to overlook the scope of research. 4  Research on specific symptoms of the disease was more relevant to patients diagnosed with Parkinson's (eg balance problems, dyskinesia or cognitive problems) 10 than for the OCD patients involved in our survey. In the future, studies could target the heterogeneous subgroups of OCD (such as contamination, harming, symmetry/ordering, pure obsessions) 20 to clarify differential needs. Furthermore, patients diagnosed with depressive disorders, their relatives and care providers focused more on self-help issues and the access to adequate care 13 than our participants did. Nevertheless, the efficacy of treatment and its long-term success were essential to both samples.
In order to aim for these patient-derived research topics, funding requirements are one means. 5,27 When research agendas are set up together with patients, when study objectives are discussed with patients, and when grant applications require a statement on how patients have been and will be involved, researchers are guided to more PI in their studies. During the survey, one patient asked 'Who is this "research" and when will "it" answer me?', and called for more and understandable feedback of research results to patients. Thus, we prepared a summary of our results using plain language for publication in the DGZ magazine. Although the current study used indirect involvement, 2 it is targeted at direct participation of patients into the whole cycles of future research projects. 28 According to experiences from other countries with a longer tradition in PI, coordination and collaboration across institutions, identification and evaluation of effective PI strategies for divergent contexts, and eventually, a change in the research culture is strongly recommended. 29 Limitations of the study are firstly attributable to the resources available. The meta-review was conducted only by one researcher,