We interviewed 13 professionals: eight psychiatrists and five nurse practitioners. The average participant age was 49, seven were female and their average experience with working in psychosis care was 17 years (Table 1). Eight clinicians experienced the application as overall benefiting their clinical encounters, whereas five experienced no or even a negative impact on their daily clinical practice. The research group identified five recurrent themes in the interviews: 1) visual structuring, 2) guideline based treatment recommendations, 3) contextual factors, 4) effects on patients and 5) effects on shared decision-making. These themes were appraised differently by the respondents and provided new insights into the way TREAT was used during consultations and contained feedback that can be used to improve TREAT for future use.
Table 1. Clinician demographics
Clinician
|
Occupation
|
Age range
|
Years in psychosis care
|
1
|
Psychiatrist
|
61-65
|
23
|
2
|
Psychiatrist
|
41-45
|
3
|
3
|
Nurse practitioner
|
41-45
|
21
|
4
|
Psychiatrist
|
66-70
|
33
|
5
|
Psychiatrist
|
41-45
|
11
|
6
|
Nurse practitioner
|
41-45
|
20
|
7
|
Psychiatrist
|
56-60
|
23
|
8
|
Psychiatrist
|
36-40
|
5
|
9
|
Nurse practitioner
|
36-40
|
20
|
10
|
Psychiatrist
|
51-55
|
19
|
11
|
Nurse practitioner
|
61-65
|
15
|
12
|
Psychiatrist
|
41-45
|
15
|
13
|
Nurse practitioner
|
46-50
|
12
|
Theme 1: Views on TREAT’s graphic representation
Before the introduction of TREAT, ROM results were summarized in a letter to the clinicians and the general practitioner (i.e. ‘ROM-letter’). It contains a written description of the ROM results. The TREAT application presents ROM results graphically and structures it in three areas (symptoms, physical health and psychosocial wellbeing). This representation was frequently discussed. The majority of the respondents indicated that, compared to the ROM-letter, TREAT reports were an improvement. Data is better structured and more appealing. The graphs made it easier to identify and interpret issues and the visualization improved the discussion with patients. One clinician noted:
“It’s a really good instrument to interpret the ROM-results and to take action if needed. It makes things a lot easier. With the ROM-letter, “You had to figure out what should be discussed with the patient, and which matters were less important. With TREAT it’s much more obvious, so yeah, it’s much easier.” [C1]
Another clinician emphasized the visualization of outliers in the results:
“TREAT is very user-friendly and the graphs also make it very visual. People were able to really see the outliers in their results, which gives me the opportunity to specifically discuss them. It gives people guidance and support during the consult.” [C8]
However, some clinicians indicated that TREAT added little value to their already structured routine:
“Let me start by saying that our ROM-letter, which we have been using for years, has a clear overview of all ROM results. Therefore, I am already used to evaluate these results systematically with my patients. With TREAT this remains the same albeit in a different visual structuring with graphs and treatment recommendations.” [C12]
While some clinicians felt TREAT complicated their routine:
“I always used the ROM-letter myself to check for any particularities, somehow there always seemed to be less than with TREAT. Now the focus is on many more areas, so you almost need to prepare ahead of time.” [C5]
Overall, the graphical representation of the ROM-results was well-appreciated by most clinicians and seen as an improvement compared to the previous ROM-letter. Graphical representation of the ROM-results made pressing issues in treatment more visible and therefore easier to discuss. Based on the UTAUT model, TREAT’s visual representation positively affected the predictive factors of effort and performance expectancy.
-Please insert figure 2 here-
Theme 2: Views on TREAT’s treatment recommendations
TREAT offers several treatment recommendations (Figure 3) for clinicians to consider with their patients. Some clinicians found these recommendations helpful:
“That’s what I like about TREAT; you are not forced to follow for example a recommendation to start an anti-depressant in case of persistent negative symptoms. You just discuss it, like is this something you would prefer or not. Maybe you both decide to try something else. Either way the recommendation is still valid, it’s just not mandatory.” [C6]
However, others found the recommendations bothersome or felt pressured:
“I mean, I know it’s not mandatory to follow the recommendations, but it still feels that way. Sometimes, you’re just happy that somebody is using the medication you prescribe at all, and then you get the recommendation to switch the medication. TREAT seems to always tell you that it’s not good enough. It’s never good enough.” [C13]
Even though the tone of the treatment recommendations was experienced in different ways, all clinicians agreed that the actual content of the recommendations was sound. However, opinions on the applicability varied. Several clinicians experienced the suggestions as generic and comprehensive, sometimes even too comprehensive. Not all recommendations were suited for the clinical complexity of the patient, or had already been tried before:
“The treatment recommendations are sound but you always need to tailor them to a specific patient or circumstance and see if they still apply.... It’s difficult because sometimes certain recommendations from guidelines have already been tried or are not applicable anymore.” [C2]
Although several clinicians raised the issue of utility, most of them had checked the utility and relevancy of the recommendations for each individual patient and found ways to incorporate them into their consultations. The recommendations were used to evaluate previous steps and to discuss and decide on current treatment plans, as this respondent explained:
“TREAT is helpful in aligning treatment with the evidence-based recommendations. It can be used to start a conversation about treatment options and help explain why alternative treatment options might be more preferable.” [C11]
Other clinicians used the recommendations to discuss possible future steps in the treatment process:
“TREAT also provides information about possible future steps in treatment such as for example electroconvulsive therapy. if recommendations are presented on a screen it feels more natural to address it as an option. You can inform patients of different options in case the current treatment doesn’t work.”
[C8]
However, some clinicians did not see the recommendations as beneficial. They argued that they were well aware of the content of existing guidelines and therefore did not need an overview of the different guideline-recommended treatment options, as this respondent stated:
“TREAT was not beneficial in reminding me of new things we could try for a specific problem. It’s not really a lack of knowledge I experience when drafting a treatment plan or when starting a new treatment.” [C12]
Some respondents even experienced the recommendations as irritating:
“I feel guidelines are necessary as a foundation but we can also assume they are well-known. To build a system just to beat people over the head with guidelines defeats its purpose. It irritates.” [C10]
Overall, opinions on the recommendations varied. Some respondents actively used the recommendations during their consultations while others felt no need for guideline implementation. Multiple suggestions were made to shorten the text and to make recommendations more personalized. It is important to look for ways to improve these recommendations, as a perceived lack of utility could potentially prevent clinicians from working with TREAT after this clinical trial.
-Please insert figure 3 here-
Theme 3: Views on TREAT’s contextual factors
All clinicians agreed that TREAT was properly imbedded into the existing technical infrastructure of the electronic patient record. Therefore, the facilitating factor referred by the UTAUT model was perceived as optimal and enhanced TREAT’s use. Clinicians who experienced TREAT as benefiting their practice found it easier to incorporate the application into their routines. Teams that used a strict screening routine, organized processing of the screening data and structured scheduling of treatment plan evaluations, were most successful at implementing TREAT. Some teams used the opportunity of the TREAT study to improve their screening process and feedback procedure, as this clinician highlighted:
“We chose to participate in the TREAT study and to make TREAT the driving force behind our evaluations and yearly screenings.” [C9]
Getting used to a tool such as TREAT, even if ultimately intended as a time-saver, takes time (15). Most teams indicated, however, that they simply did not have that time, as they were understaffed. Some clinicians indicated that the TREAT application made their consultations more time efficient. However, the majority experienced either no difference or reported increased consultation times. Most clinicians had to become familiar with TREAT and find ways to use it effectively during consultations:
“You really need to work with it [TREAT] a few times because you can get questions for which you were not prepared or reminded of things you might have missed.” [C7]
Apart from novelty, TREAT also increased consultations times by bringing up a larger array of topics for discussion:
“I think my consultations became longer, because I noticed some time shortage. Therefore, you probably take or just need some more time to discuss all the results. It depends of course, on what ends up in TREAT. If someone has few problems you are quicker to discuss everything.” [C12]
This is in part because the ROM-PHAMOUS screening is extensive and patients often experience issues in multiple areas. Furthermore, an incomplete screening was mentioned several times as a limiting factor. Clinicians sometimes chose not to use TREAT during a consultation because questionnaires were missing. In addition, more than half of the clinicians indicated that some of the recommended interventions were not part of the available treatment resources within their team. In some teams, nearly all of the recommended interventions were unavailable. Psychomotor therapy (PMT) was mentioned most frequently as a missing resource, followed by cognitive behavioral therapy (CBT), eye movement desensitization and reprocessing (EMDR) and individual placement and support (IPS). Clinicians generally ignored unavailable recommendations during consultations, thereby potentially decreasing TREAT’s efficacy. In some cases, TREAT motivated clinicians to recruit professionals for missing resources elsewhere in their organization:
“I really feel a lot of the added value lies in the fact that we are used to recommending treatments we have available. TREAT reminds you of treatments you do not have available directly, so you can try and find those treatments elsewhere within the organization.” [C8]
Compared to most CDAs in fields such as oncology, cardiology or orthopedics (2), psychiatry differs from settings in which CDAs have found mass adoption because care for patients in FACT teams is mostly integrated in long lasting recovery based processes (19). Treatment decisions fit in an approach in which timing of interventions is important. Interventions should be available at various times throughout the treatment process. Moreover, psychotic illness is periodic and the decision-making process should match this process-based variability. It was mentioned several times that it is not always straightforward to turn treatment recommendations into behavioral changes for this patient group:
“Most people have been in care for a long time and suffer from several disabilities. Sometimes you are able to initiate something new by putting in a lot of effort, but sometimes it just does not work because some patients have been doing things in a certain way for so long it’s difficult to motivate them to try things in a new way.” [C7]
On the other hand, some clinicians actually used TREAT as a driving force to try new steps in treatment without postponing them:
“I think it [TREAT] helps clinicians to stay closer to and be more professional in chronic treatment while remaining evidence-based without postponing the next step in treatments.” [C8]
To summarize, structured and complete ROM screenings facilitate the use of TREAT. Consultations need to be strictly planned after screenings and might take more time. Missing treatment resources within teams can lower the efficacy of TREAT and hamper its implementation.
Theme 4: Views on TREAT’s effects on patients
The effects of TREAT on individual patients were a recurring theme. The extent to which patients were engaged in the use of TREAT varied. In most cases, patients and clinicians sat together in front of a computer screen to review the TREAT report. However, some clinicians preferred to use the printed version. Overall respondents noticed that sharing information with TREAT did not work equally well for all patients. Some had cognitive problems and were easily overwhelmed by the complexity of the data presented in the application:
“I noticed that if patients are not able to process a lot of information at the same time or if they are very much stuck in their own line of thinking, TREAT’s systemic approach doesn’t really work that well.” [C2]
Most clinicians indicated that they did not notice significant changes in the therapeutic relation with their patients when using TREAT. However, some clinicians did:
“We think that the traditional treatment relationship between patient and clinician is fundamentally changing, it is becoming more horizontal, not in every aspect but in many. That is where it is supposed to go. I really think TREAT can facilitate this because it increases commitment and a feeling of ownership.” [C4]
Another clinician noticed a greater sense of ownership for patients while using TREAT:
“It really has to do with ownership of the data. If I have a ROM-letter with a lot of text, it feels like I own the data. With TREAT there is a subtle nuance in how it feels, like you give the patient more ownership and make them the owner of the data.” [C2]
Most clinicians viewed TREAT as an effective tool to engage in conversation with patients about specific areas of interest or suggested treatment recommendations:
“You can show your patients the different treatment options during the consultation and explain the risks and benefits. I see it as a useful tool to engage in a conversation about the available treatment options.” [C3]
It was often mentioned that TREAT prevents you from missing certain issues during consultations. This opens up the opportunity to discuss these issues with patients, as this respondent revealed:
“Of course, that’s the beauty of this system. TREAT suggests things that you otherwise might have forgotten or wouldn’t have thought of. Sometimes it can be used to engage in conversation. For instance saying something like: “According to the guidelines, you would have to start with an antidepressant. What do you think? Oh, you don’t want another pill? Okay.” [C6]
Several clinicians indicated that it became easier to discuss intimate topics because they were explicitly stated in TREAT. One respondent pointed out sexuality as an example:
“For example sexuality. That is not something you would immediately discuss, I mean you should of course, so that is my fault, but with TREAT, it is explicitly stated. Also intimacy. It therefore brings itself up, which makes you talk about it. So that’s an improvement.” [C4]
Some clinicians expressed concerns that TREAT focusses more on problems instead of strengths. Highlighting the positive trends and aspects of treatment was mentioned several times as a potential improvement. Most clinicians have a recovery-oriented view on patient care, which sometimes contradicted the alarming nature of TREAT as this respondent explained:
“Our intention in our patient contact is to try to focus on recovery and strengths. However, TREAT draws the attention mostly to the negative points.” [C13]
In a few cases, patients experienced TREAT as confrontational and it even scared some:
“Sometimes I would notice a negative atmosphere, caused by the results and how they are displayed. That’s because it mostly highlights problems which pop up in red graphs. That scared some patients.” [C12]
In sum, some clinicians noticed some patients did not respond well to TREAT because it confused or scared them. However, clinicians were able to use TREAT effectively during consultations with most of their patients. Important and sensitive issues became apparent and were therefore less likely to be forgotten which strengthened clinicians’ performance expectancy as referred to in the UTAUT model.
Theme 5: Views on TREAT’s effects on shared decision-making
A majority of clinicians indicated that TREAT supported their clinical reasoning. It did not change the outcome of most treatment decisions, but improved the way these decisions were made. Even though clinicians held different opinions regarding the benefits of TREAT, nearly all of them agreed that it contributes to shared decision-making (SDM):
“It [TREAT] did have a positive influence on shared decision-making. You have multiple options to choose from. That was most obvious with things like negative symptoms. You can tell someone music therapy or cognitive behavioral therapy is available, but scrolling through these options together makes it easier for patients to say: ‘that doesn’t suit me, but this is something I’d like to try.” [C8]
Another clinician provided a practical example of TREAT contributing to SDM during a consultation:
“It [TREAT] improves your thinking. For example with a patient suffering from depression and a guilt delusion. For the delusion, it was recommended to start clozapine, but for the depression, the recommendation was to start a lithium addition. You explain and discuss these options. Eventually we both agreed to start with the depression protocol, before starting clozapine. We also agreed it was a mood congruent delusion. TREAT really helps to show things in this way.” [C11]
In conclusion, although respondents have different opinions about the benefits and different aspects of TREAT, they all agree that the application facilitates shared decision-making. In total, five themes explain the use and evaluation of TREAT by clinicians (Table 2). In addition, all four predictors of the UTAUT model were positively evaluated by the majority of the respondents.
Table 2. Summary of results
Views on TREAT’s:
|
TREAT experienced as promoting
|
TREAT experienced as limiting
|
1. Visual structuring
|
Improved representation versus ROM-letter
|
Successful feedback routines not in need of change
|
2. Treatment recommendations
|
Supported clinical reasoning and discussions with patients
|
Too generic, comprehensive or inapplicable
|
3. Contextual factors
|
Structured and complete screening routines
|
Time pressure and unavailable resources
|
4. Effects on patients
|
Sense of ownership and increased commitment
|
Overwhelming and difficult
|
5. Effects on shared decision-making
|
Facilitated shared decision-making
|
Clinical decisions often remained unchanged
|