Results
Themes that emerged from the focus groups were perceived learning effects, the tutor's background and role, the peer CME groups' structure and atmosphere, the experience of receiving and discussing feedback reports, explanations for an unfavourable prescribing profile. We elaborate further on these categories below.
GPs' perceived learning effects
Peer group academic detailing was experienced as a suitable method to learn more about pharmacotherapy, though there were participants who argued that the scheme was time-consuming. The participating GPs experienced the CME group meetings as an important arena for learning. They reported picking up good advice from others and learning practical alternatives to drugs that should not be used. GPs said their prescription data would not mirror all learning effects:
The whole point is to reflect more, that you think twice, and with respect to this it has been a good project. It should have been done within other areas too. (GP, FG 3).
One important outcome for the GPs was an experience of being more reflective in decision-making about prescriptions.
The tutor's background and role
Tutors did not consider themselves as "experts" but as "one of them". Being open about their background as GPs was an agreed upon strategy, and tutors deliberately tried to avoid being perceived as experts:
When I presented the [quality] indicators, I said: "The next one actually surprised me a lot ... I didn't know about this before", and then they understood that I didn't think I was clever than them in the first place (Tutor, FG 2).
The tutors experienced that their own background was important for GPs' trust and acceptance of the project:
It was very important ... if the Directorate for Health had sent the inspectors along with the feedback reports, then I think it would have been much more difficult ... When it comes down to the touchy parts, I think it's crucial that we're on equal terms, that we're in the same boat. (Tutor, FG 2).
GPs confirmed that it was important for them that the tutors were "one of them" and were independent both of the pharmaceutical industry and the health authorities. In their view, the tutors shared an understanding of the complex decision-making involved in prescribing in general practice. Guidelines could give rise to dilemmas when they considered all aspects of the patient's medical condition and situation, and thus it was easier for them to disclose reasons for inappropriate prescribing to a "GP colleague":
We know that as a GP you have experienced these problems, and know exactly how we feel ... to disclose yourself ... and who knows how difficult it can be. (GP, FG 8).
Both GPs and tutors experienced that sharing the experience of being a GP contributed to an open and constructive discussion.
Group structure and group atmosphere
Tutors reported that CME groups varied in terms of structure and atmosphere. Some groups knew little about the project, but tutors reported that there was a constructive atmosphere in most groups. Groups could still have very different "cultures", and a group with a rigid structure could be a challenge for a tutor if formal and informal group leaders insisted on running the group meetings in the usual way. Tutors introduced a new agenda and structure in existing CME groups and experienced that they had to be very explicit about the scheme:
The groups were different ... we thought that a group is a group and all we have to do is to run the scheme ... and then I experienced that groups have their own cultures. These groups have existed for a while, which we probably have to consider in a programme like this. (Tutor, FG 2).
Tutors experienced lack of structure and scepticism as a challenge in some of the groups:
At a private clinic, I had one group which was unstructured, with sceptics. I felt a great deal of "industrial adherence" rather than a critical attitude towards one's own prescription practice (Tutor, FG 2).
Both tutors and GPs emphasised that a "good atmosphere" in the group, and "a sense of security" among group members was essential for an open and constructive discussion.
Receiving and discussing feedback reports
GPs' "hits" for inappropriate prescriptions in the elderly, or an unfavourable antibiotic prescription profile, was the starting point for group discussions at the second meeting. Tutors reported that GPs' tried to justify and explain their practice:
They tried to justify their prescription practices, but I also experienced this as reflections about why they had ended up in a particular situation. (Tutor, FG 1).
GPs said that the feedback on their prescription profile motivated them for reflection, learning, and change. The majority of GPs shared their prescription profile in a straight-forward fashion with the other members of their CME group:
I was surprised to see how willing people were to reflect on their own behaviour and practice ... and constantly comment like: "Well, did I really do that? I surely have to pull myself together". Very strong will, apparently, to make changes. (Tutor, FG 2).
GPs generally experienced the CME group as a safe setting to present and discuss their feedback reports:
It would have been more embarrassing if it had been in a large lecture hall or a large seminar. I am not the kind of person who would present all sorts of data ... but to discuss it with three to four persons ... that's okay. (GP, FG 5).
Some GPs experienced disclosing their prescription profile or "hits" as frightening. Observations both from tutors and GPs indicated that discussing feedback reports had caused distress for some GPs:
The older ones were silent, because they had a prescription profile that was far from what was recommended. The young ones dominated the discussion, and they were much more familiar with the guidelines in the first place. My impression was that the old [GPs] felt distress when disclosing their profiles ... The old ones have repeated their errors for several decades ... old habits are difficult to change. (Tutor, FG 1).
Tutors experienced that some GPs told that they had forgotten to bring their reports, and the tutors suspected that this was an intentional to avoid disclosing their profile or "hits":
The ones that were afraid, didn't bring [the feedback reports] ... there was one who had failed in making any changes, and I think that was why she didn't want to present her profile. (Tutor, FG 1).
GPs were generally more embarrassed if they had hits they knew they should have avoided, such as prescribing flunitrazepam to elderly patients, compared to potentially harmful drug combinations that had not been highlighted in the recommendations.
Patients' needs and demands
Patients' demands for certain drugs were commonly used as explanation for an unfavourable prescription profile. GPs explained that they had tried to stop prescribing certain drugs after they joined the project, but had failed to convince their patients that this was a good solution for them. A typical example was a patient who had been using a benzodiazepine hypnotic for several years:
The patient had used nitrazepam 5 milligrams for five, six, seven years, and then I learned that we should use zopiclone, which does not work, ok, then we try the other, zolpidem or something like that, and he is not happy and demands nitrazepam. And then he gets nitrazepam and he sleeps well. If you are born in 1919 and are 87 years old, then I don't see the reason why we should change this. (GP, FG 6).
GPs feared that changing to another drug in elderly patients could cause distress to certain patients:
It is important that you know your patients well. You may make it worse by withdrawing a drug if they are nervous. (GP, FG 8).
The GPs generally said that it was easier not to prescribe an inappropriate drug in the first place, than to withdraw a drug from a patient. The taste of the drug was a common explanation for not prescribing the recommended first-line drug for respiratory tract infections in children:
One issue when it comes to children is taste [of the antibiotic] ... we have to give them something they [will] take, and then you find a favourite that works well. (GP, FG 3).
Generally, GPs found it easier not to prescribe an inappropriate drug in the first place, than to withdraw a drug a patient had used for some time. GPs in the antibiotic group said there had been an increase in atypical pneumonias the winter the project started, and that this could have influenced their antibiotic prescriptions.
The GP's background and the practice
A common explanation for "hits" was the patient was not really their patient, because one had just issued the prescription on behalf of a colleague who was absent. The notion of "summer holidays prescriptions" denoted this phenomena. If the patient's regular GP was absent, the receptionist would ask one of the other GPs to issue the prescription that the patient needed. As long as the drug was one the patient's list of drugs in the electronic patient record, one rarely questioned the indication for the drug:
When we work in a large health centre, then we sign prescriptions for each other ... when a colleague is absent, we issue prescriptions for him that day. Any prescription I issue is my responsibility, but if you are asked to prescribe a particular drug [for a colleague] then you sign it in the reception. I don't check which other drugs that person uses. (GP, FG 6).
A usual explanation for issuing inappropriate combination of drugs was that the combination had been "imported", for example that they had been prescribed by the patient's previous regular GP. In this situation, one was reluctant to question problematic drugs or combination of drugs before trust had been established in the doctor-patient-relationship. Other explanations were that it could be difficult to know the complete list of all drugs a patient used. GPs explained that also a local "prescription culture" could influence patients' expectations, which in turn could influence GPs' prescription habits.
Other health professionals or health care institutions
Other health care institutions, such as nursing homes, community nursing services, or hospital specialists could have a role in inappropriate prescribing:
How detailed should we check what the community nurses order? My secretary print it and I sign it, and three drugs are listed. It could be a fourth drug the patient uses that I don't know about, right. That was quite unpleasant. (GP, FG 6).
GPs said that hospital specialists influenced their prescribing practice, because the specialists prescribed new treatments without considering whether the new drug was appropriate in relation to the patient's other medication:
Very much is about tradition, and we are influenced by the feedback from other colleagues, especially from hospital. Some of my tetracycline [prescriptions] are prescribed for elderly patients and patients with chronic obstructive lung disease who have exacerbations that the hospital treat with tetracycline ... Obviously this influences us. (GP, FG 4).
A typical example was patients with chronic obstructive pulmonary disease where a hospital specialist both had prescribed the treatment and provided written instructions about this to the GP in the discharge report.
Competing interests
The study was funded by The National Board of Health and The Norwegian Medical Association. JS and ML are principle investigators for The Prescription Peer Academic Detailing (Rx-PAD) study.
Authors' contributions
JCF was principal investigator for this study. JCF designed the study in collaboration with JS and ML. JCF and SH collected the data. JCF and SH coded and analysed the data, and JCF wrote the first draft of the manuscript. All authors have critically revised of the manuscript and participated in the interpretation of findings. All authors have read and approved the final manuscript.