Reassurance

All 20 participants acknowledged that reassurance was an important factor in the decision to refer for a scan. If they had difficulty alleviating the patient’s worry over a serious cause for their headaches, participants typically referred patients to a neurologist for a scan or requested a scan directly.

Most patients actually take your word for it, as long as you say ‘yes I am pretty sure this is definitely not anything in the brain’ … but for those patients who are sort of intractable it is quite helpful. (P12)

However, a normal scan result was not the sole factor cited as influencing patient reassurance. Participants raised the issue that if patients perceive GPs as having less expertise than a specialist, this limits the reassurance they can provide. Six participants indicated that a key aspect of reassurance may be lost through direct-access scanning, if the patient does not see a specialist. One participant stated direct-access scanning may not save resources as patients may still request to see a specialist postscan:

… everybody wants to see a specialist, about whatever problem they have got… I am still trying to master that art of telling them they don’t need to and reassuring them… But it is hard… with headache, people want to see a specialist (P15)

Participant (P10) stated they would refer for a scan directly if it was primarily for reassurance and to a neurologist if they suspected a sinister underlying cause.

Shared decision making

Although participants had the intention of limiting unnecessary investigation, they also perceived themselves as more facilitators than gatekeepers. All participants found honesty about the limits of their knowledge aided collaboration with patients. One participant stated:

I am quite comfortable articulating… on my part not knowing what the problem is, but don’t have an issue with articulating that to patients, and explaining how I’m then going to manage that… (P7)

Four participants spoke of patient empowerment as important to their management (P1, P11, P12 and P13). This was particularly crucial for patients who were likely to consult their GP on multiple occasions. For example:

I’d hope that (sharing treatment options) would empower the patient… I think that’s a disadvantage… when you refer patients to a specialist, they might be seen once or twice in a specialist clinic, or the headache might be a long term problem, so they need somebody that can be relaxed and help them through it. (P11)

To facilitate patient empowerment, participants made themselves available as much as possible. One participant remarked:

… it’s again leaving the door open, so… there is always a bit of safety netting about… ‘if that happens then… just because I have said today I have not found anything when I have examined you, don’t dismiss it and come back and the door is open’. (P19)

Appointment length

All participants said that patients with severe and/or chronic headaches often required longer appointments or that the short appointment time was typically not sufficient to cover everything. Six participants (P1, P5, P6, P13, P14 and P15) said that multiple consultations were a way to circumvent short appointment times. GPs typically left the door open for patients to book further appointments, with one participant (P15) reporting that they would sometimes book the subsequent appointment during the first consultation. Another explained that multiple consultations can help monitor their patient’s evolving condition and facilitate shared decision making:

So the short answer is that you manage it longitudinally, across several appointments… something like a headache, you might start off by doing the simpler tests and making sure that you are (not) dealing with potentially sinister issues upfront… but on the basis that I will see you… either in a few days’ time or next week and we just continue that conversation. (P5)

Incidental findings

All participants acknowledged that IFs were a source of anxiety for both patients and GPs, particularly for MRI rather than CT. Fourteen participants described warning patients of this possibility when discussing whether to scan, particularly if the scan was for reassurance. Negative consequences were patients subsequently become anxious or fixated on an IF, hindering reassurance.

Radiology reports themselves presented a potential source of difficulty for GPs, due in part to their lack of training in neurology. Although reports frequently came with a consultant’s letter summarising and interpreting the findings, anxious patients sometimes asked for further explanation of reports. Five participants expressed a lack of clinical confidence in being able to explain these reports in detail to patients. Three participants had not had to do this in their careers and said they would not wish to do so. Nine participants would occasionally seek advice from the reporting radiologist if something was unclear in a report.

Four participants (P1, P3, P5 and P11) indicated that they would occasionally telephone a radiologist seeking further explanation for scan results. Two participants (P5 and P11) were aware the hospital has a dedicated room from which radiologists could answer any calls regarding reports.

Additionally, radiology reports varied in clarity. Three participants described instances where reports were very thorough but lacking in a clear guideline for interpretation. Some patients requested a copy of their report, seeking further explanation from their GP. One participant suggested that a more standardised format for reports would be helpful:

… it is quite helpful if it says ‘normal intracranial appearances’. If it says:- ‘abnormal appearance suggests discuss with neurologist’; … Or ‘ring for further advice’ or so to highlight something needs further, ‘in the opinion of a radiologist’, or ‘to have further follow up’, I think is good. (P11)

One participant (P12) suggested that even if the radiologist’s report did not recommend a scan, IFs or ‘incidentalomas’ may lead patients to request another in the future. One participant (P18) had used a National Health Service (NHS)-commissioned service provided by a private provider of imaging. The service reduced wait times, but the report was non-committal, as though the radiologists were not wanting to take responsibility for giving advice if an incidental abnormality was found.

Mental health

All but one participant reported that psychological comorbidity was common among patients presenting with headaches. Eighteen participants reported that their formal psychiatry training had been minimal, although they had received substantial on-the-job experience. They acknowledged finding it challenging to address psychological symptoms, when the patient had consulted for a physical concern. For example:

I personally find it quite a difficult therapy to sort of suggest to patients. I think it’s quite difficult to get them on board with the, the fact that, sort of the psychosocial things might be a contributor to their headache or pain problems. (P10)

The two participants who did not experience difficulty in addressing psychological symptoms had received some formal training. Nonetheless, all participants acknowledged the necessity of addressing mental health difficulties. One participant observed:

You actually you can do more harm to somebody, I think by going down a very physical route and not addressing their psychological (issues)… (P8)

Potentially facilitating this process was them having good access to mental health services. All participants were aware of and referred patients to Increasing Access to Psychological Therapies (primary care service delivering evidence-based treatments for anxiety and depression) and/or other mental health services. Three participants reported that their practice had an in-house psychologist. Two participants responded that a normal scan result can help move the management discussion forwards by relieving anxiety (P19 and P6):

… once patients have a normal scan, then they can focus on kind of moving forward with management and are more willing to try tablets and medication regimes, because lots of people don’t like taking tablets. (P19)

However, this benefit did not necessarily extend to more severe cases. One participant observed:

… headache can be part of an overall difficult problem amongst a number of other complex problems the patient has. (P11)

Another participant observed that in complex cases, such as patients with medically unexplained symptoms (MUS), neither a scan nor a specialist referral help resolve problems:

I have got a (patient) with chronic fatigue syndrome, back-ache, that comes backwards and forwards, with headache and (she) can’t see that it is related to her mood. She is definitely depressed. So she has actually been to see a neurologist already, and still comes back with a headache. So it is all related to her mood and she is not willing to accept that. (P15)

For these patients, shifting the discussion towards mental health, while difficult, was more helpful in the long term, with one participant reporting the following:

… this particular person that I mentioned, it took me a year to actually get him to accept that it was a mental health issue rather than a physical problem. But the moment I got him to accept that, his A & E (Accident and Emergency) attendance and GP attendance fell by about 250%… (P3)

Impact of family and media

Seven participants (P11, P12, P13, P15, P18, P19 and P20) acknowledged the impact of web-based information can have on patients’ worry. However, this also provided empowerment when used discerningly, something that participants could facilitate. For example:

… it is part of general practice… there are lots of… verified web sources as well so it is directing patients to things that you know are valid. The internet is full of information and not all of it is appropriate… (P19)

Five participants described the influence of family on patient worry (P13, P15, P18, P19 and P20), for example, a relative having an aneurysm after suffering headaches for a period of time:

… people might know a friend or have a family history… who has had… some sort of brain problem like an aneurism… those are people who might come more directly and say… I’m worried what this headache could be and… they are thinking ‘this is not just a migraine’. (P18)

Managing uncertainty

Seventeen GPs indicated that requesting a scan could decrease their uncertainty. For example:

… when there are some features about the headache… that I think are slightly unusual and don’t fit a particular pattern, and I can’t clinically say ‘oh I am really sure it is that type of headache’, that’s when it would be useful. (P14)

There was variation in their tolerance of uncertainty. Eight experienced GPs valued their sense of toleration for uncertainty. One of these eight participants illustrated a typical scenario where this changed their management:

They walk out of your room no better than when they walked in. They may need a scan, they may be right. And it’s just always keeping an open mind… let’s live with the uncertainty and let’s see you in three weeks’ time, maybe you could keep a diary between now and then, I really want see the pattern of these headaches. And then it is, whoops, it is every morning… and it’s a consistent pattern and it does seem to be relieved by paracetamol which you would expect chronic daily headache to be… Maybe I need to reconsider my own initial working diagnosis. (P1)

One participant who reported lower tolerance for uncertainty also expressed lower clinical confidence regarding headache management:

Headaches I suppose, if a GP has ‘neurophobia’ it is probably because whatever treatment they are giving isn’t working, or because they are worried that every patient that comes through the door with a headache has a brain tumour for whatever reason. (P16)

The relationship between tolerance for uncertainty and a tendency to refer for a scan was complex. Two participants indicated that due to a low tolerance for uncertainty, they primarily stuck to guidelines that discourage scans for reassurance:

I tend to be quite conservative… looking for red flags and, and just sort of examining and managing that way. (P6)

Conversely, four reporting a low tolerance for uncertainty said they referred for a scan for fear of missing something serious. This variation across participants may be partly explained by a negative influence from a prior experience. One participant described the impact of seeing a patient without red flags that ended up having a brain tumour:

You tend to swing to the very, very risk averse side of things when you have those experiences, and then it slowly… you come back to the middle as it were. (P3)

Another participant observed that difficult or ‘heartsink’ patients, who visited frequently with non-specific difficult-to-treat symptoms, might influence them to order a scan in the absence of clinical indicators. The management of uncertainty for participants was also impacted by their patient’s uncertainty. For example:

… there may be patients who are more difficult to manage, patients who have personality disorders… they can’t cope with uncertainty in any sense… if a patient is very demanding in that way, it puts emotional pressure on the doctor which it inevitably will, then you are more likely to organise a scan. (P11)

Change in services

One participant noted that patients were now more likely to expect imaging technology as part of investigation into their chronic pain:

… patients now come along with all sorts of things wanting a scan, be it headache or knee pain or back pain etc. so I think there is an issue around patient expectation. (P10)

Cost

Participants’ perceptions were mixed regarding potential costs. For example:

if you are the sort of GP that… feels the responsibility on the purse strings… then I suppose that adds an extra layer of complexity to… the decision making process. (P16)

…if I put my ‘wider –health-system-cost-hat’ on I would be a little concerned about that… But… the idea is still a good one because I think that the evidence suggests that actually it is cost-effective, if my recollections are correct. (P3)

Legal

Two participants made explicit reference to GPs’ fears regarding legal action as a factor. For example:

I think increasingly the medical/legal landscape is very clear cut… we are getting… more and more at risk of being complained against and sued, it is not a problem of being complained against or sued but the time taken to resolve these things. (P13)

Discerning use of direct access

Two participants (P11 and P19), who had received further training in headaches, reported that direct-access scanning should come packaged with an educational component on headache diagnosis and management and on interpreting scan results:

I think you will need some guidelines to say, ‘these are the real red flag lines’ ‘these are more yellow flags’ so that…, GPs are aware which patients they should be thinking of… (P11)

They believed that tolerance of uncertainty, difficulty managing ‘heartsink’ patients and lack of clinical confidence could all be ameliorated through further training, with an emphasis placed on adopting a holistic approach.

So there probably needs to be… some kind of protocols about, when it is appropriate to refer or things to try first. (P19)