A total of 5 FG meetings and 2 semi-structured individual telephone interviews took place, and these included 40 GPs (22 HIGPs and 18 LIGPs). The Table 1. shows the characteristics of participating GPs in the two FG categories.
Three of the 41 constructs strongly discriminated between the HI and LIFG (Intervention Complexity, Individual Stage of Change, and Engagement Key Stakeholders) and 7 constructs weakly discriminated these two groups (Intervention Adaptability, External Policy and Incentives, Implementation Climate, Compatibility, Relative Priority, Self-Efficacy and Formally Appointed Implementation Leader Engaging). Fourteen constructs were not discriminatory. Six constructs had insufficient data for evaluation and eleven constructs had no data for evaluation. (Table 2)
We further describe the discriminatory constructs in each CFIR domain and provide brief descriptions of some constructs that were not discriminatory but important to take in account for a future implementation strategy.
Domain I: Intervention Characteristics
Complexity was a strong discriminatory construct. The HIGPs perceived the intervention as having low complexity, wellbeing-conceived, having steps that were clear and concise, and not requiring great effort.
"... I think that the intervention was well thought out, I haven’t found it difficult at all, I don’t see it as a chore, as a specific task that you have to do, the truth is it hasn’t been difficult for me.”(6)HIFG-3
One LIFG perceived the intervention as having high complexity because asking a patient about BZD consumption meant "opening a Pandora's box” that could lead to emotional stress and necessitate recommendations for psychotherapy sessions, delays in the consultation, and difficulties in involving the patient. The other two LIFG stated that they did not find the intervention very complex but LIGPs reported that they were unclear about BZD's discontinuation plan and that they often had to review the instructions.
"... you get to a point when you can’t, you don’t get out the paperwork (intervention instructions) every day, because our day to day goes so fast that you have to plan it, you’re not going to open the drawer every day to see what the intervention was about.”(77)LIFG-1
Adaptability was a weakly discriminatory construct. The HI and LIFGs supported the adaptation of BZD withdrawal to patients by making it slower or faster, but the HIFGs provided more supportive, stronger, and creative comments regarding the tailoring of BZD withdrawal individually for each patient.
“…… .tell them that they had to reduce their dose, we’ll reduce it by a quarter or you would tell them to file down the tablet every week with a nail file, the first week twice, bam bam, the next week, 3 or 4 times, because that will help us to stop it sooner, filing it down with a nail file because it’s very difficult to remove a quarter. They are very old and their eyesight isn’t good, the nail file works well for me…”(19)HIFG-2
" I’ll halve the dose and you come back to see me in a month, I’ll halve the morning dose and you come back in ... or in two months, … "(73)LIFG-1
Analysis of other constructs in this domain indicated that GPs perceived the intervention as an internal intervention (Intervention Source) and they felt valued as enablers because those who designed the intervention were also GPs and understood primary care settings. Most clinicians placed greater value on the benefit of the intervention to the patient than scientific evidence regarding its implementation (Evidence Strength & Quality).
In general, GPs stated that the intervention required more time than their usual in clinical practice. Some comments indicated they believed the intervention offered them advantages over their usual practice (Relative Advantage), but others believed there was no advantage because application of the intervention greatly prolonged the consultation. The training workshop and the quality of the support materials (Design Quality & Packaging) were valued as enablers of implementation, but some LIGPs found the workshop too theoretical and requested a workshop that provided more practical advice.
Domain II: Outer setting (Health District)
External Policy & Incentives was a weakly discriminatory construct.
We distinguished health districts that did and did not have local policies regarding BZD prescribing including indicator and incentives.
One LIFG without district health local policies made reference to the lack of strategies of the health district regarding safe prescription of BZDs (personalized audits of GPs to improve their BZD prescriptions, global strategies aimed at professionals in all levels of care including the hospital, primary care, nursing homes……). HIFGs did not comment on this topic.
GPs in the health district with local policies regarding BZD prescribing stated that knowing that the goals of the intervention were aligned with the goals of their health district encouraged them to implement the intervention. They perceived a greater willingness of patients to participate in the intervention because patients received the same information from multiple sources (brochures in the waiting rooms, comments between patients, dialogue with GPs). HIGPs made stronger supportive comments about the intervention and local policies on BZD and were more willing to adopt it.
"I think that the indicator is good because it’s also an indicator of poor practice, that’s why it’s there, to help you get information about how you’re doing, an indicator can also help you meet the indicator at a particular time.” (45)HIFG-2
"The pharmacy also gave us some leaflets to hand out about benzodiazepines……, and that also helps, you give it to the patient........and then they read it …of course the leaflets I’ve mentioned have nothing to do with this, but I’ve used them, …… and they’ve worked for me…” (12)HIFG-5
“… .that not all the health system has this culture of evaluation, in specialist care, there is no control over prescriptions, unlike us, it doesn’t matter to them if they prescribe one thing or another because nobody is going to check.” (42)LIFG-1
Domain III: Inner setting (Healthcare Center)
Implementation Climate was a weakly discriminatory construct. The HIFGs referred to the high degree of participation of their team in the project, to their high capacity to accept new challenges, and the role of the medical director as promoter. However, comments by the LIFGs were completely opposite.
"We take everything on board, we’re pioneers, everything, we sign up for everything at our center…….” (50)HIFG-2
“In our center the same, it has been very well received, in this respect, there are no problems with implementing it,.....what’s more, our coordinator is very interested in us doing new things, in participating in things like this,........ (51)HIFG-2
“Bad in my center, anything new is a struggle”(59)LIFG-1
Compatibility was a weakly discriminatory construct. Although all GPs believed that implementing the intervention required more time than is usually available in primary care consultations, the HIGPs expressed more interest in the intervention and said they were more likely implement it when their workload was light and vice versa. However, the LIGPs reported that implementing the intervention was not compatible with workflows, because if they did, they caused significant delays.
“First I thought it was going to be a little difficult because of the stress of everyday work, this requires more time than usual, to reduce benzodiazepines you sometimes need to have good conversations......, so, at least in my case, I did see that at the beginning…I was gradually reducing doses, then we had a period with more winter illnesses and the practice was much busier than usual etc., and this was when I stopped doing as much but then,....., sometimes you try, but depending on the time of year, you succeed or you don’t........”(1)HIFG-3
” ... but when it comes to putting it into practice, in doing it in everyday practice when you’re under pressure, when you’re up to your neck like always, it’s very difficult…”(10)LIFG-1
Relative priority was a weakly discriminatory construct. All physicians stated that it was necessary to address the high rate of BZD prescriptions in primary care settings (Tension for Change), but only the HIGPs prioritized implementation of the intervention. Some HIGPs attached great importance to the project because it made them aware of the high consumption of BZDs and the need to address this problem.
"Rather than priority, what I liked most about it was that it raised awareness but I would put it at the same level as so many of the other interventions that we implement,..... but it is true that it has been a wake-up call, it's raised awareness about an important issue. " (59)HIFG-2
"It depends on the case, we’ve tried to prioritize, you have to realize that all the doctors are more motivated, but I don’t think they have prioritized this intervention over other things." (24)LIFG-4
In the Access to Knowledge & Information construct we assessed access to the support web-page, which has a video describing intervention instructions and supporting materials. GPs reported technical problems accessing the website. The GPs who were able to access and use the support materials rated them as high quality.
Domain IV: Individual Characteristics (GPs)
In the Self-efficacy construct we were interested in determining the self-efficacy of the GPs for each component of the intervention, and therefore divided this construct into two sub-constructs:
1. First prescription: This was a non-discriminatory sub-construct. All GPs stated that correctly making the first BZD prescription by limiting the duration of treatment was the easiest part of the intervention and the most effective in preventing chronic BZD consumption.
2.Benzodiazepine withdrawal: This was a strongly discriminatory sub-construct. Despite the difficulty of withdrawing from BZD treatment, HIGPs reported that the intervention offered them a useful tool to assist chronic users to withdraw from BZDs and they saw clear benefits for patients. The LIGPs stated that BZD withdrawal seemed too difficult and in most cases was not even worth trying.
“ This is what it has been useful for me personally, for new treatments, taking things more slowly, trying to prevent patients from becoming chronic users, and yes it’s given me a tool that I can use to help chronic users come off the drug, a difficult task, but it’s given me a system for doing this, little by Little, let’s see if we manage to help them, I’ve definitely changed my attitude towards new patients, it’s really helped me with this” (66)HIFG-2
“Now when you prescribe you explain that this is a medication, for a short time, for problems…. For chronic users, I often don’t even think about it, I leave things as they are.” (67)LIFG-4
Individual Stage of Change was a strongly discriminatory construct. The HIGPs stated that they adopted the intervention and want to keep it in their practices. The LIGPs were enthusiastic about implementing the innovation soon after receiving the training, but they lost enthusiasm over time and began prioritize other activities. None of the LIGPs reported they integrated the intervention into their usual practices.
“It sometimes depends on your caseload, and as it’s normally high, but I think that.....,like I've mentioned before, in the height of winter with so many people who are sick there are times… (you stop implementing) and this is reflected in the numbers of course, but, it's something to keep in mind and regardless or not of whether the study finishes, it’s something that has sunk in and it’s something you get used to doing…”(36)HIFG-3
”…when you do the training at the beginning, you tell everyone, then you go on holidays for a month, the summer comes, and then you forget."(4)LIFG-1
All GPs stated that the doctor-patient relationship and the patient's trust in the GP were key points in being able to initiate BZD withdrawal (Knowledge & Beliefs about the Intervention).
Domain V: Process
Formally Appointed Implementation Leader Engagement was a weakly discriminatory construct. Leaders were valued by HIFGs but by only one LIFG as key elements to drive implementation. The other two LIFGs made positive assessments of the leaders for their personal characteristics, but not as drivers of implementation.
“The person who came to sell us the project has been very important in our center,....he makes everything easier...and then it makes you become more involved, you’re going to try this, you’re going to get better results…..” (67)HIFG-5
“Yes, I think they are very capable people (the implementation leaders), despite that,...I used it (implemented it) but I would like to have used it more…" (6)LI-INTERVIEW
Stakeholders Engagement was a strongly discriminatory construct. The HIGPs were enthusiastic, involved in applying an innovation that offered a clear benefit to patients, incorporated it into their clinical practices, and strove to make it compatible with workflows. In contrast, the LIGPs were not so involved or enthusiastic about the intervention.
"I’ve realized that it doesn’t take a lot of effort,....,just remember, make a little effort, here I’m going to rank number one in terms of users and I’ve realized that I've brought my numbers down simply by making a little effort,....,so yes, you try to prescribe less, ask why they are taking it, try to negotiate with your patient, reduce it a little, it’s not that you have to put up a big fight…” (41)HIFG-3
"I’ve tried to follow it but sometimes you don’t do everything, just a part...., you give them information and aim to continue on other days, and that day then...it often stops there...ask them whatever at the next visit, and then at the next visit if you don’t look at what you wrote down, if they don’t tell you… and that's it, they don't tell you, patients don't see it as a problem, it's really hard for them to see it as a problem, because taking it makes them feel better..”(24)LI- INTERVIEW
All GPs mentioned the difficulty of getting patients to commit to BZD withdrawal (Engaging Innovation Participants) because they had been taking it for many years without noticing adverse effects and expressed fear of not being able to sleep if BZD was withdrawn. One HIFG made strong comments that patients were more aware of the need for BZD withdrawal because of the local policies regarding BZD prescription. These HIGPs were sometimes surprised by their success in engaging patients and achieving withdrawal or dose reductions.
In the Reflecting & Evaluating construct, all GPs agreed on the role of receiving individual feedback to promote implementation because they appreciated the result of their effort. In fact, in one health district, feedback was not sent to each GP's email, but was available on a list of prescription indicators from the primary care organization to which the GP had access. Many GPs did not discover that the feedback was provided in this way, and they criticized this procedure, and even got angry, because they did not have data on their achievements.
Reminder sessions of the intervention were requested and the FG meetings were valued as reinforcers of the implementation.
At the end of the focus group meeting, the GPs were asked to make proposals for improving the implementation (Appendix 1)