Table 6 End of study interviews with surgeons (n = 5) showing the four emerging themes and subthemes and include corresponding example quotes

Theme: Information and communication

Patients identifying needs and preferences

He knew he wanted to have it one go … so he actually, I think he pretty much told me “I know what I want, I am prepared to go through this process but I know what I want” (Surgeon 2)

Share decision-making, shared decision-making based on information from a variety of sources

Having an infection's a terrible thing to have and is very, very unpleasant and there's a pretty high failure rate. So personally I think it's really essential that the patient is involved in the decision, …It's their body, they can decide whatever they want … because when it fails, I think it's really important they felt, well, we decided together … (Surgeon 5)

Patient equipoise

… Within this study we’ve had experience of patients who are much more sort of involved in the decision making process and ultimately, after they’ve been given all the information, they’ve not found themselves in a position of equipoise ….The question of whether or not the patient has equipoise is very challenging and I think it is dependent on the patient, how much they want to know, you know (Surgeon 3)

Surgeon credibility and expertise, patient scepticism and gaining trust

And some people say, “oh well I don’t want to be a guinea pig”, and those are the ones that—that’s a nonstarter…. but by the time that that sort of conversation comes up and you’ve met them at least once, if not more, they kind of do trust you and so the credibility thing, and because they’re coming to a, you know, I suppose, a more reputable centre, and everyone’s said “we’re going to send you to this place and they’ll cure your infection” (Surgeon 5)

Theme: Recruitment and randomisation

Covid

It was just everything seemed to change after Covid, …with number of previous surgeries, pre and post pandemic… (Surgeon 5)

Wrong sites

…I think looking through our data, I think the cases that came to us were far too complex to basically, I think in hindsight we were the wrong site for this sort of trial. (Surgeon 5)

Increasingly complex (or acutely unwell) cases

I think looking through our data, I think the cases that came to us were far too complex to basically (Surgeon 5)

Changes or challenges with care pathways

It’s always going to be a mixture of planned and ad-hoc work, … people are coming in in emergency settings are definitely eligible, the trick is trying to get to them before someone operates on them (Surgeon 9)

Theme: Equipoise, dissonance and challenging dogma

Equipoise —

genuine uncertainty and living with uncertainty

I suppose that's life as a clinician, isn't it, that you, at the end of the day, if you, [find] your equipoise disabling, you wouldn't be able to treat anybody, so, you know, at the end of the day, when you're treating someone, even if you're not sure what the best option is, you've got to go with what you and the patient think, with what you know … (Surgeon 9)

I guess some of us have quite wide area, a wide area of patients where we have equipoise and some it's much more narrow and I suppose where that point is where, whether you can have an area where virtually every surgeon is uncertain about that question … (Surgeon 6)

… Because I don’t know what the infection eradication rate is, I am, I mean in equipoise (Surgeon 5)

Changing paradigms, challenging dogma

Entirely incorrect the believe to think that two stage should result in lower chances of re-infection but difficult mindset to change (Surgeon 3)

Even though, you know, there’s no science to back that up, I suppose that’s the, you know, that would be the conventional teaching and that’s the ingrained bias that you get (Surgeon 3)

Infection eradication is key concern but many other factors to consider

Better means I suppose getting a good infection eradication rate, so the primary aim of this operation is to get rid of the infection, the second aims are to have better function, … and minimise complications (Surgeon 5)

Dissonance gut feeling and evidence required to change practice

But you can't get away from that fact you're kind of, your gut wants to perform, do one or the other … at the time of randomisation, willing them to be one or the other, you know, well, I know, based on literature, etc., I have equipoise, but you can't help, get away from that sort of experiential feeling about how you treat these cases because they're so complex you have to go with what you think's best, you can't sort of put them into an algorithm and say, this is best (Surgeon 6)

Surgeons sometimes have preferences, which may not be based on evidence but on other things

You know, sometimes it's probably subconscious things that you don't even appreciate you're thinking about, for example, you probably have biases that you didn't realise you had (…) as to whether they're a good host or a bad host, things in MDT that people say that sway you one way or the other (Surgeon 9)

Study design

Attitude to study, study design and study involvement

Super keen to try and see if [it was possible] to get [a] randomised trial actually going, because it’s the ever present question amongst our community (Surgeon 5)

Barriers and examples of recommendations

If we maybe cast our web out wider we could each, but it would take a lot more hospitals and each hospital would contribute a small number I think. And then that’s, well that becomes a national trial which is really expensive (Surgeon 5)

Considering other research designs such as cluster randomisation

So certainly, I mean that would be a possibility, you know, one of the most experienced revision knee surgeons in the country does almost exclusively single stage, so you could recruit that unit to do single stage, […] I guess, very subjectively my view would be that surgeons would say they’ll be very happy to do either but actually you’d find it more straightforward to then find another site of a surgeon who was prepared to do a two stage one, so actually a cluster approach would potentially be… (Surgeon 3)