TABLE 3.
Doubts about the necessity to modify antibiotic prescribing in accordance with rapid molecular test resultsa
| Sub-theme | Quote number | Supporting quotations |
|---|---|---|
| Treating the patient not the result | 1 | “Um I would always treat the patient and not the result. So regardless of what type of sample analysis has been used, I would treat the patient, so if I felt the patient had clinical features of infection, I would treat them for infection unless I felt it was going to be harmful to do so.” -P8, ICU consultant, Hospital 7 |
| 2 | “If the patient’s super sick, I don’t care what the test says, I’m prescribing antibiotics because, you know, that fits with the clinical picture. If the patient is super well and… and the… and the test result doesn’t corroborate with all the other evidence that I’m triangulating, because no one test is perfect, I’m not going to prescribe antibiotics.” -P20, ICU consultant, Hospital 1 | |
| 3 | “If the organism did not… was not detected, but there was a clinical suspicion for ventilator-associated pneumonia, we would carry on with the antibiotics anyway.” -P10, ICU consultant, Hospital 8 | |
| 4 | “I have to say as a clinician I don’t follow guidelines very well. I tend to go by my gut instinct and by what I see by the patient’s physiology, by the bed space. And frequently, even if the guidelines suggest a different antibiotic, sometimes I change my… my plans. Not on the basis of either the BioFire or… or… It’s all in the whole kind of holistic view about what’s going on.” -P14, ICU Consultant, Hospital 4 | |
| Negative results create dilemmas | 5 | “a negative [molecular diagnostic] test, if it’s well performed, is trying to say to you we cannot identify any bacterial DNA. […] we’ve got no evidence that there is some sort of… one of the common pathogens here, in that [sample]. And so that’s sort of saying to you: Look, you’ve got little evidence to support active infection.” -P4, ICU consultant, Hospital 1 |
| 6 | “[I’ve] chosen to stop them [antibiotics] as a result of the negative BioFire result. So, in a sense, saving two or three days or potentially more of an antibiotic. We do… and I've just seen that as an example in my own mind of good practice, you know, good antibiotic stewardship.” -P3, ICU consultant, Hospital 2 | |
| 7 | “We refocused the antibiotics on sepsis rather than chest sepsis. So, the antibiotics were not stopped, but the BioFire… the results of the BioFire were negative...” -P11, ICU consultant, Hospital 5 | |
| Initial skepticism and unfamiliarity | 8 | “So that even if it’s an antibiotic we’re unfamiliar [with], we don’t routinely use like ceftriaxone for pneumonia, we only tend to use it for things like CNS infection […] 'Cause usually it’s unfamiliarity. It’s the situation where [changes voice] “We don’t normally do this, so I don’t want to do it”, which is how a quite a few of my colleagues still practice. […] I think initially there’s a degree of skepticism because again, the department, well most departments I suspect is slightly split between people who are interested in new things and people [who] are not really that bothered by new things. And I think it was a little bit split.” -P1, ICU consultant, Hospital 3 |
| 9 | “Most intensive care doctors come with a healthy streak of skepticism about a new machine. Is it really going to add something that’s going to change practice?” -P16, ICU consultant, Hospital 10 | |
| 10 | “I haven’t used it [molecular diagnostics] enough […] I really would need more involvement with it.” -P17, ICU consultant, Hospital 1 | |
| 11 | It’s a matter of exposure. So if you have the machine and you use the machine, finally you are used to make decisions with that information. If that machine is available, but it’s not integrated in their routine because you have few patients [that] makes you less confident of using information from the machine. I think it’s a matter of exposure, is… is… is not a… the machine are such that is triggering your decision or your confidence with devices. If something is integrated that it’s part of a pathway and you’ve got enough volume of patients to… to be exposed to… to that pathway decision, definitely you will have an opportunity to be more confident with the machine. So I think it’s not the machine as such. It’s how much this machine is using the context of making decisions.” -P6, ICU consultant, Hospital 9 | |
| Variable knowledge of the tests’ inherent limitations | 12 | “[Molecular diagnostic tests] can’t distinguish between live or dead bacteria, but well, that’s not a concern. That’s like a feature of understanding how the tools you have available to help you, work. And like there isn’t a perfect tool. So, it’s just a piece of knowledge that yeah, you need to have while you’re doing these things.” -P4, ICU consultant, Hospital 1 |
| 13 | “I can’t remember if Stenotrophomonas was on there [panel]. I think it… maybe it wasn’t, I don’t know. And we’ve had Elizabethkingia bacteria.” -P12, ICU consultant, Hospital 5 | |
| 14 | “[Patients] can’t really be on antibiotics, you wouldn’t use it [molecular diagnostic test] then.” -P17, ICU consultant, Hospital 1 |
|
| Respiratory sample unavailability and of uncertain quality | 15 | “the COVID patients, they were just dry [as] a bone. You could never get specimen once they’d been there [ICU] for 3–4 days. […] You can’t do the test if you haven’t got sputum. So. And these patients are on a lot of oxygen, so you’re not inclined to do bronchoalveolar lavages on them either.” -P13, ICU consultant, Hospital 4 |
| 16 | “[In the COVID-19 surges] the ICU staff may not have been familiar with procedures in an ICU in general, let alone what a BioFire is. Particular locations meant that it was more difficult for the research nurses to have time to go and consent a patient and also pick up a sample in a, sort of, in a timely fashion […] there was competing workload from other trials that were running, on the research nurses. So, when you combine all of those three, a quite common event would be that we would identify somebody on the ward round who would meet the criteria to be recruited, but they weren’t. So, they would have a sample sent for MC&S [microbiology culture & sensitivity], but they didn’t get a sample taken for BioFire.” -P19, consultant microbiologist, Hospital 6 | |
| 17 | “I don‘t know how good our quality control was for sampling. […] nurse or research nurse or physio, whoever is collecting samples, [do] they apply anything like the same kind of quality control to ‘That’s proper sample, and that isn’t’?” -P16, ICU consultant, Hospital 10 | |
| 18 | “[The molecular diagnostic test is] on a desktop in intensive care or something like that. Where they’re less used to handling sensitive PCR machines. And has the potential to be contaminated by bugs and flora.” -P18, consultant microbiologist, Hospital 10 | |
| 19 | “deep [BAL-like] sampling is a better… in a sense is closer to the ‘truth’, if you like, in inverted commas, about pneumonia. As opposed to proximal [sputum-like] sampling. And so of course, a lot of these patients had proximal sampling. And so, were we actually just dealing with colonization?” -P3, ICU consultant, Hospital 2 | |
| 20 | “it’s easy to mess up a BAL so the test comes back negative.” -P20, ICU consultant, Hospital 1 | |
| False-positive results encouraging antibiotic overtreatment | 21 | “the temptation for the [ICU] clinician is to try and treat all of those organisms [detected by molecular diagnostics]. Which often mean[s] meropenem […] [intensivists] will be less critical than I am of the results, or if they see a result they will say: ‘Right, what do we give to treat it?’ They won‘t think: ‘Do we need to treat it?’” -P5, consultant microbiologist, Hospital 1 |
| 22 | “I need a quantitative assay as opposed to [a] qualitative assay. So, I'm happy to say that well that’s Klebsiella in sputum. Fine. But is that Klebsiella, is it significant? And that level of significance is what I need.” -P14, ICU consultant, Hospital 4 | |
| 23 | “where I struggle a bit is to understand what the quantitative piece [of molecular diagnostic results] means.” -P4, ICU consultant, Hospital 1 | |
| False-negative results leading to antibiotic undertreatment | 24 | “if the BioFire is negative and you are still having [a] small possibility that the patient is having [an] infection. Very small possibility, but you might start treatment with antibiotics while you do other things that might not be related with the sepsis. Uhm, you might see that response over the next 24 hours, 48 hours […] [if] the patient dies or have [sic] any complication related with an infection and you did not cover that because you restrained yourself, rightly or wrongly, at that particular time. You might see the situation as a potential litigation problem.” -P6, ICU consultant, Hospital 9 |
| 25 | “a false negative may give you the confidence to stop therapy when actually they’re [patient] still unwell.” -P18, consultant microbiologist, Hospital 10 | |
| 26 | “I don't mind false positives 'cause I'll just treat for a while. Um, that’s not… not such a negative, but the false negative would be the thing I don't want to miss.” -P12, ICU consultant, Hospital 5 | |
| 27 | “there have been a few situations where we’ve not believed a negative result. […] And we've repeated it [molecular test]. And done it with deep [BAL-like] samples and there’s been a… just because of the clinical situation. And we've re-calibrated the machine.” -P3, ICU consultant, Hospital 2 | |
| Concerns about how results influence existing AMS structures and communications | 28 | “if you just use it [molecular diagnostics] on everybody without making a decision beforehand of ‘Do I think they have an infection or not’, you're probably going to end up with a lot of people [getting antibiotics].” -P1, ICU consultant, Hospital 3 |
| 29 | “there’s a limit on the number of the test you can run concurrently. I think that that limits… you know, it’s not for all comers [into ICU] as it were. And I think if people start abusing it then you're gonna have patients that you need the results [for] and you’re not gonna get [them].” -P13, ICU consultant, Hospital 4 | |
| 30 | “[Molecular diagnostic] results come out at 8:00 o’clock [at] night. I don't know why that is particularly, but that’s quite common. […] Sometimes I'm notified and don’t see it till the following morning. So, in the interim, you’ll tend to find they [patients] get put on whatever they [intensivists] think is going to cover it [detected organism].” -P5, consultant microbiologist, Hospital 1 | |
| 31 | “I wrote on the drug chart the result of the BioFire. So, right next to where the antibiotics are with the box on the day after to say ‘Let’s review this’. So, I was giving a plan and clearly labeling it, but that doesn't mean that it got through to the microbiologists […] if you could find a way to get that result onto our in-house system and flagged to the microbiologist paired up with the BAL sample, then I think that would be really useful.” -P12, ICU consultant, Hospital 5 | |
| 32 | “one thing that we could have done would have been a way to, you know, scan or image the result and incorporate it into our clinical notes so that it would be apparent to other colleagues why a patient was de-escalated from a carbapenem to temocillin a week ago.” -P19, consultant microbiologist, Hospital 6 | |
| Uncertainty about the evidence base for molecular diagnostic test results’ clinical usage | 33 | “[Molecular diagnostic tests] gotta really show an impact for it to be worth the hassle and the maintenance and the cost and the variability […] It’s gotta be clearly better for it to be adopted.” -P16, ICU consultant, Hospital 10 |
| 34 | “[Recruiting in COVID has] been good in that many of my colleagues because we were using off… just using it [molecular diagnostics] routinely, liked it, and gained confidence in it.” -P1, ICU consultant, Hospital 3 | |
| 35 | “So the BioFire was negative for any of the common organisms. I guess the thing that influenced me is that I didn‘t stop the antibiotic at that time. I decided to continue them over the first 24 hours. For the reasons I’ve already sort of talked about that I haven't built the confidence in the test yet and I haven’t seen the sort of large validated study yet.”-P12, ICU consultant, Hospital 5 | |
| 36 | “Most clinicians would want to know how accurate is that [molecular diagnostic test] and is it inferior or non-inferior? And we would pour over the evidence for that in some detail.” -P7, ICU consultant, Hospital 7 |
a
BAL, bronchoalveolar lavage; BioFire, BioFire FilmArray pneumonia panel; ICU, intensive care unit; and PCR, polymerase chain reaction.