Background
This chapter reports on the methods and findings from the qualitative work package (WP), which explored patient and healthcare professional (HCP) views and experiences of the current monitoring strategy and views on the monitoring recommendations emerging from the prognostic and health economic modelling. This has been published previously.5 We also describe patient adherence to current monitoring recommendations. It is important that the proposed monitoring strategy for the early detection of liver, kidney and blood disorders due to immune-suppressing drugs used to treat inflammatory conditions is acceptable to both patients and HCPs. These stakeholders were therefore included in determining the acceptable monitoring strategy. This ensured that the final recommendations incorporated their views so they may be better accepted into clinical practice.
The specific objectives of the qualitative WP were to:5
- explore patient adherence with current monitoring recommendations
- assess patients’ and HCPs’ views and experiences of current monitoring strategies, and
- explore acceptability of different monitoring strategies.
Methods
Study design
Qualitative interview study
Participant recruitment and eligibility
Patients: Eligible5 patient participants were adults (aged 18 years and above) with the ability to give informed consent, and were:
- diagnosed with one of the following inflammatory diseases: RA, CD, UC, PsO with or without PsA, AS and/or SLE, and
- currently treated with one of the following immune-suppressing treatments for at least 6 months: methotrexate, azathioprine, 5-ASAs, sulfasalazine, mycophenolate, leflunomide or anti-TNF-α agents, that is etanercept, adalimumab, infliximab, golimumab, or certolizumab or their biosimilar.
Patient participants were invited to take part in the study either through patient organisations or their secondary care clinic.5 For recruitment through patient organisations, charities that work on behalf of patients with the relevant inflammatory conditions (National Rheumatoid Arthritis Society, Crohn’s and Colitis UK, National Ankylosing Spondylitis Society, Lupus UK, and Psoriasis Association) were contacted and asked to distribute information about the study through their mailing lists, social media platforms, websites and/or online forums, as appropriate to their established methods of communication. The initial information about the study presented to their members included a short explanation of the study inviting them to participate, a link to the participant information sheet followed by a link to complete the questionnaire online.5
For recruitment through secondary care, dermatology, rheumatology and/or gastroenterology clinics at three hospital sites in the Midlands, North and East of England were asked to identify eligible patients in their care and provide them with a study pack at their usual appointment or through the post.5 The study pack contained an invitation letter explaining the study, a participant information sheet, questionnaire and a stamped, addressed envelope to return the completed questionnaire.5
The questionnaire was used to assess eligibility for interview and measure adherence to current monitoring. Participants were asked to self-report their age, gender, postcode (for calculating Index of Multiple Deprivation), ethnicity, height and weight, inflammatory condition diagnosis and comorbidities, current immune-suppressing treatment(s) and length of treatment (less than or more than 6 months).5 We asked participants to report how often they had been advised to have a monitoring blood test for their immune-suppressing treatment, and how often out of every 10 appointments they had either missed or delayed attending a test by more than a month.5 Questionnaire completers were asked to indicate their willingness to participate in an interview and, if so, to provide their contact details at the end of the questionnaire.5
To select patient participants for interview, a combination of purposive stratified and maximum variation sampling was employed to ensure we talked to people with a mix of the following attributes:5
- Diagnosed with either inflammatory skin, rheumatic or gastrointestinal diseases and treated with different immune-suppressing medicines.
- Good and poor self-reported adherence to current monitoring recommendations. Good adherence was defined as missing or delaying two or fewer monitoring tests out of every 10 appointments (at least 80% attendance). Poor adherence was defined as missing or delaying more than three monitoring tests out of every 10 appointments (below 80% attendance).
- With and without additional risk factors for drug toxicity.
Healthcare professionals
Eligible HCP participants were consultants, GPs, nurse specialists and pharmacists with experience of prescribing and/or monitoring long-term immune-suppressing treatment for the inflammatory conditions listed in the patient eligibility criteria.5
Healthcare professional participants were recruited using a snowballing technique to find participants with the relevant experience and understanding of the monitoring blood tests. In addition to the above, the BAD also sent an e-mail to their professional membership.
The initial questionnaire asked participants for their role, year of qualification, primary specialty and how long they had worked in this specialty, name and location of their hospital or GP practice, and contact details. We then used purposive sampling to select different types of HCPs (three to four consultants, two to three GPs and four to five nurses or pharmacists) working within the different inflammatory disease specialties and with different levels of experience.5
Sample size: We aimed to recruit up to 24 patient participants with an even spread of the attributes, and approximately 12 HCPs.5 However, the exact sample sizes and proportion of participants with different attributes were dependent upon saturation.384 Therefore, recruitment was flexible.
Interview procedure: Semistructured interview guides were developed for patients and HCPs separately.5 A semistructured format was chosen to ensure the conversation was focused upon issues related to the current and proposed frequency of blood testing, while allowing participants to reveal any relevant issues not identified a priori and for the interviewer to explore pertinent responses in more depth. All interviews were conducted by Dr Amy Fuller, an experienced qualitative researcher.5 The patient interviews and analysis were conducted prior to the HCP interviews.5 All interviews were conducted remotely by telephone or video call in a private room, and digitally audio-recorded. Before each interview began, it was explained to participants that their responses would remain confidential, they would not be identified in any way and were free to withdraw from the interview at any time without giving a reason.5
Patients
The patient interview guide first considered participants’ perceptions of the risk, benefits and experiences of their current monitoring blood tests, reasons for adherence or non-adherence and perceived importance of continuing monitoring blood tests at their current frequency.5
The findings of the prior risk-stratifying and health economic exercises and proposed monitoring plan were then introduced. AF then entered the participants’ information into the calculator to determine their individual risk score, gathered by the information provided in their questionnaire (medication type, height, weight and sex) or asked during the interview (alcohol consumption, comorbidities, other medication, etc.).5 Participants were presented with their risk score and the three monitoring frequencies (6-monthly, annual or biennial) proposed by the health economic modelling exercise. In the absence of a risk-score calculation for anti-TNF-alpha medications, participants on anti-TNF-alpha monotherapy were not given information about this. The risk score was used with those on combination DMARD/anti-TNF-alpha therapy.5 Participants understanding of the new strategy was first explored, followed by the acceptability of the different monitoring frequencies. Concerns and potential benefits of the proposed plan, as well as how they would prefer changes to the monitoring strategy to be communicated, were also discussed.
Interpreting the risk score
The risk score tells us the predicted risk (as a percentage) of stopping treatment due to an abnormal blood test result in the next 12 months.5 A risk score of one indicates that 1% of people with the same characteristics as those entered into the calculator would have to stop treatment, and so forth. It was explained to patients as ‘the number of people having to stop treatment due to an abnormal test result in the next year, out of 100 people with the same characteristics as them’.
Healthcare professionals
The HCP interview guide first explored perceptions of current blood testing frequencies, including practicalities of implementation, sensitivity and specificity of the tests, perceptions of risk and benefits, and risk management.5
The findings of the prior risk-stratifying and health economic exercises and proposed monitoring plan were then introduced. Participants were presented with four to five anonymised, descriptive scenarios representing real-world patients included in the prognostic modelling.5 Each scenario described a patient’s inflammatory condition, medication type, the presence of any risk factors in the risk calculator, their risk score and the proposed frequencies of monitoring. For anti-TNF-alpha medication, the scenarios described the suggested frequencies and how they were determined. Participants were shown scenarios representing a range of scores, those from secondary care specialties were shown scenarios within their specialism, and participants from primary care were shown scenarios across specialisms. Follow-up questions then explored acceptability of the monitoring frequencies, perceived risks and benefits of the different frequencies, and how HCP participants would prefer changes to the monitoring strategy to be communicated to them.5
Consent: Return of the completed questionnaire was taken as informed consent. With interviews being conducted remotely, participants gave their informed consent online prior to the interview using a Microsoft Office Forms consent form and verbally at the start of each interview, which was digitally audio-recorded.
Data analysis: Qualitative data from the interview recordings were transcribed verbatim by an external transcription company, anonymised and managed using the data management software package, NVivo (version 12). Data were analysed using inductive thematic analysis,385 which took place in parallel with the data-collection process so that initial results could inform subsequent sampling and data collection.
For patient interviews, initial analysis of the first two interviews was performed independently by AF and JH, who familiarised themselves with the data via ‘repeated reading’, noting initials meanings and patterns and generating an initial coding framework of data-driven themes.5 The researchers then discussed their initial coding and identified any further areas for exploration. This process was repeated with the following two interviews. As there was good agreement between the researchers’ initial codes and themes, AF analysed the following interviews and further developed the coding framework using the principle of constant comparison.5
Analysis did not reveal a difference in views between the various sampling attributes (described above); therefore, data collection continued keeping in mind to recruit a mix of participants from these groups.5
The research team came together on two more occasions to discuss analysis and agree upon the final framework.5 Emerging results were also discussed with an experienced senior clinical investigator (AA) for clarification of any clinical concepts which guided coding and theme development. Following analysis of the 18th interview, no further changes were made to the codebook and so it was considered that the sample had reached a sufficient level of data saturation and recruitment ended.5
The same approach was followed for HCP interviews, the only difference being that double coding was performed for the first six interviews to allow the researchers to discuss coding and theme generation for interviews with different HCPs, in case of any key differences by HCP type which may have guided sampling or data collection. As with patient interviews, analysis revealed areas for further follow-up, and informed whether to recruit any further HCPs within each role or specialty.5 Following analysis of the 13th interview, with no further changes to the codebook, recruitment ended.5
Ethical approval: This study received ethical approval by the West Midlands-Black Country Research Ethics Committee (21/WM/0285) on the first of December 2021.
Results
Patient participant characteristics and adherence to current monitoring
Six-hundred and forty-nine people with inflammatory conditions completed the questionnaire (see Appendix 4, Table 132).5 Of the 649 participants, 30.4% reported missing or delaying at least 1 out of 10 monitoring tests and 8.1% reported poor adherence to monitoring (see Appendix 4, Table 133).5 The proportion of people under 60 years of age missing or delaying a test was higher than those over 60. The results including quotes have been published previously.5
Patient participant interviews: Eighteen patient interviews were conducted (see Appendix 4, Table 134). Interviews lasted for 50 minutes on average (range 38–57 minutes).5 Five overarching themes, some with accompanying subthemes, were identified and interpreted (see Appendix 4, Table 135).5
Benefits and challenges to current monitoring frequency
Providing reassurance
Patients described how the regularity of monitoring blood tests gave them reassurance that asymptomatic side effects would be detected early.5 Many were nervous to commence immune-suppressing treatment given the seriousness of the potential side effects. The very frequent monitoring in the first few months allayed initial fears.5
It felt like a good thing because it’s almost like a safety net on you. If anything was going wrong, it would be detected because they were so frequent.
Long-term, regular monitoring then provided ongoing peace of mind that treatment was not causing any problems.5 Keeping a check on potential side effects was important to most patients, and a key reason for adhering to their recommended testing frequency.5
It’s more the peace of mind and just knowing that everything is okay… it is really important for me.
Regular blood test monitoring also reassured patients as it kept a check on their inflammatory condition activity.5 For a minority, this was as important as or even more so than checking for side effects.5
Useful obviously for monitoring the side-effects of the medication but also useful to have other markers of disease. And the consultant would always discuss my blood results when I went to the appointment so it felt like at least he was getting useful feedback and I was getting useful feedback about how the disease process was, sort of in a flare or not a flare.
Several commented that regular monitoring feels like someone is keeping an eye on them and their inflammatory condition in the absence of regular consultant appointments.5
It does give me that reassurance that you’re still part of that system, that someone’s still looking out for you… because obviously we can go quite long times without consultations, because they don’t need to, because everything’s fine… But it just gives you a little bit of reassurance from that side of things.
Conversely, some patients saw monitoring tests as a tick-box exercise for their medication to be prescribed and any potential risks of the medication was an issue secondary to treating their inflammatory condition or preventing symptoms returning.5
I saw it as the medication was what was going to help me, and if one of the requirements was to have regular blood tests then so be it.
How much patients felt reassured by monitoring depended somewhat on whether they were informed of the results.5 Those who were told their blood results were stable said it was reassuring to know they were being checked.5 Some were able to access their results, which they liked as it made them feel in control of their disease management,5 although others who received no explanation of their results or feedback that they were stable were left concerned about the possibility they were not being checked, unclear about how the monitoring process worked, and questioned the point and robustness of it.5
At the minute, it does seem, sometimes seems a little bit pointless because I mean they might be looking at it and that might be useful information but I don’t know that.
We do get them online and that’s been very good. But then the difficulty is knowing what something means. There are certain things on there that’s worried me in the past, where it says things are abnormal and it’s a little bit not quite frightening but if everything’s fine, it’s fine, you can see that and that flags that up. But some things they don’t show, you have to do a lot of research yourself to find out whether they’re okay or not. You assume the consultant or somebody at the hospital is doing the same sort of thing, but you don’t always know.
A practical inconvenience
Most participants were monitored in primary care requiring them to organise their own blood tests via their GP surgery.5 The location of blood tests being their GP surgery was considered convenient unless patients had work or educational commitments, which made it difficult to secure appointments at a suitable time.5
It was awkward in my last job, where I had to be in a certain place at a certain time. To book time away from that to get to the doctor’s surgery. The doctor’s surgery usually is closer to home rather than closer to work, and then having often to drive back, get there and go back, yeah, it was definitely more logistically a problem.
Patients also reported difficulties getting through to GP receptionists on the telephone, and/or GP receptionists questioning their eligibility or reasoning for needing a blood test, which made it challenging to book tests.5
I have to call my doctor’s surgery… obviously they can be really busy and it can be difficult to get through.
Several reported how other organisational aspects of monitoring blood tests, which would usually be fulfilled by NHS staff, had been left to them.5 This included finding out which type of blood tests was required, gaining access to and sending their results to their consultant, and chasing up on results so their GP could order their prescription on time.5
When it’s time to get my prescription every third month, there’s always a hiccup, it’s always late and I always have to chase it.5 So I’m at the stage of running out or not having any [medication] to take, because the results aren’t filtering through.
While the blood test procedure was a quick and easy process for most, some patients found them painful and stressful.5
Remembering to get a test
Given the frequency of monitoring blood tests, some participants found it difficult to remember to have their test on time and occasionally missed one having lost track of time, being busy and/or prioritising other health issues.5
In my case, you’ve got other things wrong with you as well. And depending on how severe they are at the time I can sort of prioritise those, so you then suddenly forget about the need to do your blood test.
If a lot of time had passed since their last test was due, some waited for the next time point to come around.5
After a couple of weeks you think, well maybe I’ll just wait till the next one.
Among those adhering to the recommended monitoring frequency, many said they remembered to have a blood test on time because they knew their prescription would be withheld and it was important for them to not miss any doses.5
I know that getting my medication is contingent, having my prescription renewed is contingent on the blood test, and I really want to keep my prescription.
For others, receiving prompts from their GP or hospital team reminded them to book their next blood test.5 However, prompts were ineffective if they could not immediately book an appointment, for example where a text prompt was received in the evening when the GP surgery was closed.5 Further, many participants said prompts came only after a missed test, in the form of a consultant letter.5
My current consultant if you don’t have a regular blood test he will send a letter … so that acts as a reminder.5
Some even waited for this letter to arrive before booking a blood test, as it contained details they needed to book a blood test.5
Appended to that letter is the set of blood tests, specific blood tests that need to be undertaken.5 It’s not sent out to me before I get what is effectively a reminder letter. So that then prompts me. As soon as I get that I think, right okay, well I need to get this blood test. So I’ll go and try and get one straightaway, because I can’t get one earlier than that. There’s no point, because I haven’t got the actual test that they need.
Questioning the need for the current frequency of monitoring
Although it was important to patients to continue monitoring, many questioned the need to do so at their current frequency.5 Reasons given were that their inflammatory condition and medication dosage were stable, they had normal test results, or if their blood markers did fluctuate, they were minor and did not result in changes to their treatment.5
Because I’ve been on azathioprine for a few years now, you know, I seem to have settled with it. I think a six monthly blood test would be better for patients like myself who are settled with it.
I think maybe four monthly, three times a year, or like even six monthly would be nicer… They’ve always been, I mean they do tend to float around and there’s some borderline ones, but they’ve never said it needed any action. So, from that perspective I’ve been lucky and therefore haven’t seen the benefit [of ongoing three-monthly monitoring].
Before participants were given information about the risk stratifying and health economic modelling studies and proposed monitoring plan, we asked their opinion on basing the frequency of monitoring on their own clinical need.5 This was considered acceptable by all and even preferred.5
I think that would be more useful. Because I think otherwise it can just turn in to a bit of a tick-box exercise rather than actual, based on clinical need really. I think that’s what it should be based on.
Views on adopting a risk-stratified monitoring plan
Acceptability of the proposed frequencies
Before exploring acceptability of the proposed strategy, participants were asked to recall their understanding of the rationale for the research being conducted that was presented to them at the start of the interview.5 They demonstrated an understanding that once on a stable dosage of medication, the risk of side effects reduces, so it is not necessary to receive monitoring blood tests as frequently as currently recommended.5 And that continuing with 3-monthly monitoring could be an unnecessary use of resources.5
Obviously if people don’t need to be tested every three months, why would you be wasting money on doing it?
Some recalled that a longer gap between monitoring blood tests may mean side effects are left undetected for a long time.5 While some expressed an understanding that treatment recommendations may change as new evidence is gathered about diseases and their treatments, one participant interpreted the proposals purely as a cost-saving exercise.5
You said cost-effective and so that immediately makes me think this is all down to the cost.5 I’m thinking, ‘Gosh this should be much more in-line with medical research rather than just the cost of the procedures’, so I would not support a change in that direction.5
Nearly all participants who were given a risk score (i.e. those prescribed DMARDs) perceived their risk of stopping treatment to be low, including those with a score of three.5
One out of 100, that doesn’t seem like a large amount. That gives me a bit of reassurance.5
So, I’m quite low risk then really.5
A couple of participants said their risk scores of 1 and 2 seemed quite high, although their views on the proposed frequencies were in line with most other participants, as presented below.5
Considering the information presented, all participants were happy with the proposals to reduce the recommended frequency of monitoring blood tests; however, acceptable frequencies differed between participants. Six-monthly monitoring was acceptable to all and considered a comfortable step down from their current monitoring frequency to still feel like a regular check-up.5
Six months would feel like the next step that you could go to and still feel confident in it.5
I think that’s a fair amount once every six months … I think again its back to things being picked up, it just seems, six months is kind of regular.5
A couple of participants noted they were in active disease with their medication still being adjusted but would be comfortable with 6-monthly monitoring once these were stable.5 Some noted that 6 monthly was in line with how often they currently attended monitoring, given they miss some tests, or it was their currently recommended frequency, and so would have no problem with a change in recommendation to this.5
I wouldn’t see that much different to what I’m doing now. If you count the fact that I’m missing four out of 10, it’s pretty much six monthly.
Views on annual monitoring were mixed.5 Some were comfortable moving from 3-monthly to annual monitoring.5
I think yearly doesn’t feel like that massive a jump when I feel like I’m probably already around six months.
Once a level dose had been acquired and there were no issues with it, I would be quite happy, given the information and whatnot that I’ve said that, would be beneficial to go to yearly then.
Some said this could then ideally be matched up with other annual health checks for their inflammatory condition.5
They do a liver scan every year. So, perhaps if it was one big appointment to do bloods and liver scan which is what happens now, I do it all on one go, that would be fine, yes.
Several were uncomfortable moving straight to annual monitoring but happy to reach it via a graduated reduction, having a 3-monthly gap between monitoring blood tests upon reaching a stable dosage, which was then increased in stages assuming no abnormalities were detected.5
I would say go from three months to maybe, six months for two or three doses, and then increase it up to one year for maybe another one or two goes, and judge by their response rather than a blanket one. So, a graduated reduction in frequency … that way, as time is going on you’re getting a better indication of your own personal situation.
Rather than quite sudden. Like oh, monitored constantly for the first six months and then boom, see you in a year. That would be a bit panicky.
However, a few participants were not comfortable with annual monitoring in any scenario, being perceived as too long an interval for potentially serious side effects to be left undetected and prevented.5
I think it’s too big a jump and I think if there were any issues, one year since the last test, who knows what could have happened. Whereas if you have the test sooner, things are picked up and can be treated.
This was because they had previously experienced side effects or raised biomarkers requiring investigation from medication or had concerns about other health conditions that may not have been factored into the proposals.5
I would want to know and particularly because I’m being monitored for other things, so it’s a holistic view that’s needed isn’t it to say, ‘Well hang on a moment, I’ve got X or Y condition and they could be affected by this, by the adalimumab, so therefore why change it from six months?’
Other concerns included a risk of their results not being checked and therefore having to wait another year for potential side effects to be detected, as well as concerns that abnormal results would not be actioned quickly enough.5
In this area trying to get to see somebody, you know, you’re looking at months. So, if a change was picked up there wouldn’t be an appointment available for you, I wouldn’t be happy. I think a year is too long.
Nearly all participants felt biennial monitoring was too long a gap between tests for serious side effects to be left untreated.5
Two years is a long time with any health condition, you know, and I have multiple. So, I’d be concerned at two years, it’s just a no go.
In addition, it was considered counterintuitive given how often they currently had blood tests and felt it concerning to have such infrequent monitoring during the first few years of treatment.5
Knowing I’ve got the condition and having the medication and getting it under control, that whole experience for me has been four years. So, in that time I’ve had endless blood tests, and various other things that have gone on. So, two years seems like quite a big jump to me. I don’t know if I’m far enough down my journey to know how I would feel about that. It’s quite a long time in between testing.
A couple of participants were surprised by biennial monitoring as an option but said they would consider 2-yearly if sufficient clinically reasoning and assurances could be provided.5
I suppose I would just be a bit like befuddled by it! Or just think, ‘What?’ but yeah, I think as long as somebody explained, ‘Oh, this is the rationale, this is the evidence’ then I would be happy with that.
Provisos of acceptability
Reducing frequency should not only be an NHS cost-saving exercise
Most participants welcomed the cost savings that would result from the proposed monitoring strategy, and for resources to not be used if not medically necessary. However, many stressed that they would expect assurances that any cost saving would not come at a cost to patient care and safety.5
It’s obvious the NHS has to cut its costs, but I think it shouldn’t cut its cost at the cost of people’s care.
A reduction in monitoring frequency would not be supported if its sole purpose was to save money for the NHS.5
You said cost-effective and so that immediately makes me think this is all down to the cost… so I would not support a change in that direction.
Building in flexibility to accommodate additional blood tests when required
Several participants said they would feel more confident in the proposed monitoring strategy or would only accept a reduction in the frequency of their monitoring, if they knew it would be flexible to accommodate additional or more frequent blood testing if needed.5
I’d like to know if I was to feel unwell in that time then I’d be able to ring up and come in to have bloods during that time, like I wouldn’t want it to be, ‘Well you’re not going to get it now for another six months’.
The two examples of when they felt this would be needed were: temporarily increasing the frequency of their blood tests if an abnormality was detected, and requesting a blood test if they felt unwell or suspected a flare of their inflammatory condition. Participants also wanted to know what signs they should be aware of that may require them to request a blood test and how to go about doing so.5
Patient involvement in determining a suitable monitoring frequency
Additionally, several participants said they would want to be involved deciding a monitoring frequency appropriate to them,5 for example, to choose 6 months over yearly monitoring if not comfortable with the latter.5
If it’s an opportunity to sit down with your consultant and then negotiate between the two of you on the best medical approach for the individual, then great. If the consultant said to me, ‘Well the recommendation now is that we’re only going to do this every two years, so that’s the way it’s got to be’, then no, that’s not a good approach at all. What would work for me six months probably, and for other people, might be a year, so it needs to be individualised like that.
Ability for patients to view their results
Another stipulation for a few participants was to be able to view their monitoring blood test results or receive confirmation that they were checked and were stable.5
I think six months with good feedback is acceptable absolutely, I would be happy with that if I was – as long as I could get the results and I know what was going on, I’d be very happy with six months.
Benefits to reduced monitoring
Most participants said that reducing the frequency of their monitoring blood tests would reduce the burden of their inflammatory condition on their everyday lives as the practical inconveniences of arranging and attending blood tests would lessen.5
The less that psoriasis can get in the way of my everyday life, the better, and that includes like hospital appointments and blood tests.
Some also said with fewer tests, there would be fewer occasions waiting anxiously for the results. Those who were not adherent to their recommended monitoring frequency said they would attend tests as recommended, as they would not want to go without any monitoring, but did highlight that they would need prompting.5
They would have to tell me, because I would forget and I’d be thinking, oh what days have I got to do that?
Communicating a change in practice
When communicating a new monitoring strategy, participants said it would be important to provide patients with an explanation of why it was safe to reduce their monitoring frequency and the evidence or logic behind it, similar to that provided to them in the interview.5
If I was completely out of the loop of this conversation, I would want to understand the logic behind that having had blood tests regularly for such a long time… I personally would need that little bit of extra reassurance why I can all of a sudden now go to yearly rather than three monthly.
While several said they would be happy to receive this through written correspondence, for others it would be important to be able to discuss and ask questions about the change with a specialist HCP.5 Patient charities were also highlighted as useful source of information about their condition and therefore an ideal place to provide an explanation of any changes.5
Healthcare professional interviews: Thirteen HCP interviews were conducted (see Appendix 4, Table 122). Interviews lasted for 44 minutes on average (range 20–61 minutes).5 Five themes, some with accompanying subthemes (see Appendix 4, Table 136), were identified and interpreted.
Benefits and challenges of the current monitoring strategy
Acknowledging the importance of monitoring patients with inflammatory conditions on immune-suppressing medications to detect early abnormalities, HCPs identified a number of benefits and challenges in providing monitoring blood tests 3-monthly as currently recommended.5
Patient and practitioner reassurance and continuity of care
Being able to see that patients’ blood (test results/markers) were stable was reassuring for some HCPs to know they were prescribing safely. HCPs said they were comforting for patients who were concerned about taking this medication.5 It was also highlighted that the regularity of monitoring provides continuity of care through regular contact with patients.
It gives a kind of regular sort of contact between the prescriber and the patient.
Picking up incidental findings
Having frequent monitoring meant blood tests also picked up incidental findings such as infections, flares or other medical conditions that required treatment; however, HCPs acknowledged this was not common.5
Incidental findings that the physicians are not aware of and then that allows early interventions and further referrals to be made… but they might be few and far between.
Challenges of providing regular monitoring in practice
The most reported challenge, for both primary and secondary care HCPs, was that monitoring took up a large proportion of their workload involving multiple staff to arrange appointments and blood forms, perform the blood test, check results, chase patients who are not adherent to their testing schedule, and investigate abnormal results.5
A huge amount of our time is spent looking at bloods.
If you get a slight abnormality, you maybe can’t remember exactly what actually you’re meant to take, so you have to open the guidelines and read it, and then you’re like, ‘Oh yeah, of course it’s fine’. It’s just time. It’s not difficult but time consuming.
Some noted the large costs associated with frequent testing. Others highlighted capacity issues within busy NHS settings to provide and review blood tests at the recommended frequency.5
Often that’s a capacity issue. For example, if you’ve got staff sickness or shortages, blood test results might not always be reviewed on time, which is a risk.
Healthcare professionals from secondary care settings reported occasional difficulties with their Shared-Care Agreements, with some primary care centres not providing blood tests or prescribing without an up-to-date test. GPs felt it was not appreciated how much of their time is spent organising and reviewing blood results.5
A couple of years ago we pulled out of it entirely and said it was too much work… The powers that be just don’t consider how many people are involved in the stage of getting the drug to the person and how much time it takes for each person to do that.
On top of the time chasing patients not adherent to their monitoring schedule, HCPs also reported that when patients missed or delayed a blood test, they were unable to issue a prescription.5 This risked the patient flaring or having a poorly controlled disease.5
Patients will often flare in terms of their disease, so that causes issues that could have been prevented if they continued on their regimen.
Additionally, GPs noted that having to convince non-compliant patients to attend their monitoring tests can cause friction in the clinician–patient relationship.5
If patients don’t want to have their blood test in order to get the drug but want to continue the drug, we can’t actually do that and that causes conflict.
Challenges for patients having regular monitoring
Healthcare professionals also noted the inconvenience of regular monitoring on patients. Notably, in having to attend regular blood tests, which was particularly inconvenient for those of working age, and the difficulty of securing a GP appointment.5
We have no evening or weekend phlebotomy in the area, so patients have to take time off work to get it done.
Other challenges for patients that HCPs identified were forgetting to get their blood test causing non-compliance issues as described above, the stress associated with having a blood test and waiting for test results and bruising or weakening of veins from regular testing.
It’s inconvenient for patients, they have to – and who would like to be pricked this many times? After a certain amount of time the veins are bad and there’s a couple of goes, they have to have to have the blood test.
But they do create anxiety for patients though, some patients panic when they know their test is due and they panic if they haven’t heard about the results and they’re phoning in and asking for their results, ‘Why haven’t I heard?’
Interpreting and actioning results
All HCPs discussed how abnormal test results were often false positives or transient abnormalities not caused by immune-suppressing medications, but which require investigation or repeat testing to confirm. Common abnormalities were raised liver markers due to heavy alcohol consumption before a blood test, high potassium levels due to a haemolysed sample or raised WBC count due to an infection. HCPs said this results in unnecessary concern for the patient while being investigated and, if medication is paused during this time, a risk of their inflammatory condition worsening.5
Usually, we catch it early enough and usually it doesn’t cause problems, but sometimes we are late picking it up, a patient turns up to the clinic and their GP told them to stop their azathioprine five months before and they’re now flaring and that’s a problem.
Some also highlighted that threshold for intervening and stopping medications differed among clinicians, depending on how risk averse they were, which resulted in different actions for patients and the perception among some that some bloods are repeated unnecessarily.5
We don’t have any consistent guidelines on what to action and what not to action, so then I feel like a lot of the time we’re probably doing unnecessary bloods… every consultant will do things differently here.
Questioning the need for the current monitoring frequency
Bound by current guidance
Healthcare professionals stated how they were required to follow blood monitoring guidelines as set out by various governing bodies, including the Summary of Product Characteristics (SmPC), specialty bodies such as British Society of Rheumatologists, and local hospital trust policies, which all recommended 3-monthly monitoring at the very least.5
As per BSR guidelines it has to be every three months as the maximum timeframe for doing a blood test.
Some clinicians prescribed monitoring tests more frequently than 3-monthly for patients with comorbidities considered to increase the likelihood of side effects.
A cautious approach which does not tally with clinical experience
However, all HCPs questioned the need to continue monitoring all patients at a 3-monthly frequency. Some viewed the current approach as outdated and overly cautious.
I would say that it is probably, errs too much towards safety belts and braces and that we probably could get away with less frequent monitoring.
Many felt that since the guidelines were created, there was more confidence in the safety of immune-suppressing medications. Further, based on their clinical experience, abnormalities arising from such medications were uncommon in practice once a patient was established on a stable dosage. Thus, they felt a reduction in frequency would be possible.
I’m not sure what data is out there but we’re very happy to reduce the frequency because it seems to be so well tolerated.
I can’t remember the last time I saw an abnormal [result]. So, I’d be quite happy for it to be reduced, because I don’t feel like it would be risky if there’s research supporting that that’s safe to do. Because that would be my experience as well, that we’re just taking tonnes of bloods and nothing happens.
Experience of reducing frequency in current clinical practice
Further, some reported how they provided a reduced monitoring schedule during the COVID-19 pandemic lockdowns, where there were no observable differences in the rate of abnormal results. This prompted them to consider the necessity of 3-monthly monitoring.
During COVID time, we did monitor differently. Our patients could not go in to the blood monitor so that prompted you to say, ‘Do we need to have blood monitoring that frequently?’ … But now we are asked to go back into the pre-COVID situations. So, you need to say for the last two years, how many bad side effects or problems you were faced with? That is negligible.
Others also said how they were comfortable to prescribe medications with a blood test result within a 6-month period, or already less-frequently monitored medications they considered to be less-likely to cause side effects, such as sulfasalazine and biologic medications.
Drugs like sulfasalazine, we use 3 months, 6 months and 12 months because they don’t have much side effects.
We decided with, obviously, discussions with the consultant to reduce the frequency to every six months.
Views on adopting a risk-stratified monitoring strategy
Acceptability of proposed frequencies
In all, HCPs were in favour of adopting a risk-stratified approach to monitoring blood tests.
I absolutely agree that we should take a more nuanced and precise way of, and risk stratifying.
It makes good sense; we shouldn’t be doing tests that are not needed just because that’s the way it was set up 20 years ago … we should be able to get a proper rational approach to what the frequency actually is required.
All were happy for anti-TNFs to be monitored annually. Anti-TNFs were viewed as rarely or never causing the blood test abnormalities that can arise from DMARDs, and most highlighted that monitoring to ensure the medication was still effective was more important.
We treat lots of patients with anti-TNF drugs. Although they have lots of other issues, we seldom stop it because of repeat blood test abnormalities. I would be happy with once a year on that basis.
With the DMARD scenarios presented, the acceptable monitoring frequencies varied between HCPs. A risk score of 1 or 2 was viewed as low risk, with acceptable frequencies being 6-monthly or annually. A risk score of 3 or 4 was viewed as higher risk, and acceptable frequencies were 3-monthly (i.e. remain on currently recommended frequency), 6-monthly and annually. There was one exception who felt 2-yearly was appropriate for all scenarios presented.
Given that 3-monthly monitoring has been used since these medications were introduced, many felt a stepwise reduction to reducing monitoring frequency would be more appropriate and reassuring for HCPs and patients, enabling them to see results are stable at a higher frequency before increasing the gap between blood tests.
You could do it to six months and do that for three years or two years and then move to a year if it seems safe and appropriate.
Bar the exception described above, none of the HCPs were comfortable to leave patients without monitoring for 2 years. Given the potential toxicity of the medications, this was considered too long a period for side effects to build and worsen. Some also pointed out that these patients require check-ups or blood tests annually for their inflammatory condition, comorbidities or other medications that are not as toxic.
For even just reviewing purposes or even patients who have other comorbidities, the expectation is they should have a yearly blood test to make sure that everything’s okay. These are medications with no risk factors and we’re still saying from a guidelines perspective that they should have a yearly check.
Clinical opinion of risk factors in determining acceptable frequencies
When determining which frequency was acceptable, some HCPs used the risk score presented, with more frequent monitoring generally preferred as the score increased because the chance of abnormalities occurring had increased.
I think given that increase from 1 in a 100 to 1 in 50 to 1 in 25, it’s probably now at that sort of rate that actually, probably six monthly would be reasonable.
However, most also considered the medication-type and/or risk factors presented in the scenarios, using their own clinical experience and opinion of these risk factors to determine a monitoring frequency acceptable to them.
If you just gave me the fact that he was a 59-year-old gentleman on leflunomide and his risk score was 3 out of 100, I’d be happy with six monthly. But when I start looking at the individual factors it does make me more reluctant to reduce the frequency.
I suppose if you just questioned the 1 out of 100, then I’d be more inclined to say okay six monthly or once a year is fine, less frequent. But if you gave me the specific context of the medication, then I would change my view. Only because I personally don’t think from anecdotal clinical practice that it is 1 out of 100 in this situation, it’s probably more.
General practitioners were less resolute about the frequency of monitoring, happy to follow whichever frequency was recommended by national guidance.
Provisos of acceptability
Healthcare professionals highlighted certain conditions they would want in place to accept and adopt this approach into practice. Patients’ must have had stable results, and there must be consensus and endorsement from national guidance, that is NICE, and clinical specialty bodies.
This needs to be consistent across every indication for this drug, rheumatological, dermatological, hepatological, whatever it is, it needs to be the same.
It was clear that an element of flexibility should be built into the approach. Several wanted guidance on how to interpret the risk score and selecting a suitable frequency, although clinical autonomy in the decision-making process was also important to some with the opportunity to over-ride a recommended monitoring frequency, for example when a person was considered particularly high risk. Some felt it should be able to accommodate patients’ preferences. Flexibility was also wanted in the form of a regular review of patients’ risk scores and ability to change their monitoring frequency if a risk score changed.
Flexibility in it I guess … if you have got a patient who you are concerned about … that we would be able to increase monitoring and then decrease it if it was clinically appropriate.
Some questioned how risk factors that could change in between patient reviews, for example fluctuations in BMI, alcohol intake, or introduction of new medications, would be accounted for.
Whenever any new drugs are added, drugs, alcohol, all these factors, someone gains a load of weight and their BMI changes, loses weight, who’s going to take responsibility for that changing the risk score?
General practitioners stressed that given the time that would be required to determine and review a person’s score, this should not be their responsibility.
Although it looks very easy the form to fill in, it’s actually incredibly time-consuming doing something like that … I think it is going to create, unfortunately, quite a lot of workload because I suspect they will move it into primary care, and it fills me with horror that something else is being moved out.
Potential impact of adopting a risk-stratified monitoring strategy
Healthcare professionals felt there would be great benefit in adopting a risk-stratified approach, providing individualised care to patients and ultimately reducing burden by minimising the number of monitoring appointments.
The more they can just take a tablet and forget it, the better, isn’t it?
It would be great for patients; it saves their time.
Some felt reducing the frequency of monitoring may also provide reassurance to patients about the safety of their medication and reframe a common perception that they are very dangerous. However, HCPs also highlighted that some patients will be ‘wedded to the idea of someone keeping an eye on them every three months’ (HCP009, Consultant). On the other hand, some believed reducing monitoring frequency may result in more patients being complacent or forgetting to attend monitoring.
For HCPs, this approach would result having more time to focus on other important aspects of clinical care. For example, in reducing the time spent prescribing and easing the pressure on appointments, there would be time to see patients earlier in their disease trajectory and provide more medication counselling, advice and support to help patients manage their condition.
Where consultants will prescribe, one of the biggest factors will be that they can use that time to actually see more of the patients who are being referred to them. So those patients get treated faster.
You free up that time and that space, and suddenly the prescription clerk, the prescriber, the GP, the receptionist, the phlebotomist, has more time to do that, and therefore the resources could be better spent elsewhere.
A reduction in the number of blood tests performed would also reduce the costs for the NHS associated with the physical and human resources required.
Cost to the hospital in terms of running the blood test, monitoring time, clinician time, all of those things.
Healthcare professionals highlighted that the prescription process would also have to change as currently it depends upon having a (stable) blood test result each 3 months.
Communicating a change in practice
Healthcare professionals felt while the risk-stratified approach would be well received by most clinicians and updating national guidance would suffice to get most on board, some may require further persuasion. Having a champion who could communicate the new approach and answer questions was suggested as a way to overcome any ambivalence to new monitoring guidelines.
There’s a lot of difficulty letting go of the old ways … You might need the odd champion to go round and speak to people in person.
While some HCPs were happy to adopt the risk-stratified approach as it was described to them during the interviews, others wanted further detail about the research evidence before being able to say whether they would accept it in practice. This included being able to read the research themselves, having an explanation of which factors had been included and ruled out of the calculator and why, and the weight each factor plays in affecting someone’s overall risk score. A couple of HCPs wanted to know whether risk factors they believed were important to consider, had been included in the risk score calculation.
I would want to know their past medical history, so if they’ve had any malignancies at all in the past, that’s a really important question particularly within the last five years. Because that puts the risk up higher. I’d also want to know, in terms of the azathioprine particularly and the methotrexate, their sun exposure because they’ve got a higher risk of developing skin cancers.
It was also suggested that the changes be communicated through various channels. One HCP had concerns about discussing it as a cost-saving exercise, because this did not appear to be prioritising patient safety.
After having emphasised the importance of 3-monthly monitoring to patients, HCPs said it would be important to provide a handout and disseminate the information through patient organisations, which would complement any verbal explanation they could provide. Such information should include an explanation of why the monitoring frequency was changing, why it is safe, while also emphasising the importance of continuing to attend monitoring.
Definitely like a good leaflet or a handout that we could send to the patients would be really good, explaining the reasons why we’ve decided to change the frequency.
Discussion
Findings show that while patients and HCPs find monitoring of immune-suppressing medication reassuring, many do not feel ongoing 3-monthly monitoring is necessary considering how infrequently side effects arise once on a stable medication dose. The current monitoring strategy was considered burdensome for both patients and HCPs, and a superfluous use of NHS resources, with non-compliance to monitoring blood tests and false positives risking unnecessary pausing of prescriptions and flares of inflammatory conditions. A risk-stratified approach was acceptable to all although the greater the gap in testing, the less acceptable it became. Applying a tapered approach to reducing monitoring frequency, so long as patients continue to have stable blood test results, was preferred by most patients and HCPs. Accepting this strategy into practice was also dependent upon building in clinician autonomy and patient involvement when determining appropriate monitoring frequencies for individuals, taking a flexible approach to monitoring based upon individual clinical need, and endorsement and consensus from national bodies. HCPs were concerned that adopting a risk-stratified approach using the risk calculators would increase their workload as the current Excel-based calculators that they reviewed required manual data entry and was cumbersome. This suggests that the proposed solution should have a large element of automation of the process by which risk scores are calculated.
There is a need to manage the burden of treatment on patients with long-term conditions, which can be high and varied due to time taken up by medical appointments, impact on daily life and the mental load of health management.386 Further, there is a move away from a one-size-fits-all approach in medicine towards tailoring treatments to individuals and this study demonstrates appetite from patients and HCPs to apply a more individualised approach to monitoring of their treatment. Another aspect of such care is involving patients in the decision-making around their treatments. We initially proposed to undertake logistic regression to ascertain predictors of non-adherence to monitoring. However, due to a small sample size and the fact that non-adherence to monitoring was low, we chose not to undertake logistic regression.
The findings of this work show that it would be acceptable to bring about a change in practice, which will have implications for policy-makers as for it to be adopted universally, a change in national guidelines will be required. This was also an important condition of HCP acceptability. As immune-suppressing medications are used across multiple specialties, efforts should be targeted towards changing the overarching monitoring guidelines by NICE, the Medicines and Healthcare products Regulatory Agency and drug manufacturing authorisation holders, as well as specialty guideline writing groups. The findings of this study will shape recommendations going forwards to these organisations, notably the frequency of monitoring, mode of introduction and communication strategy.
The health economic exercise suggested that 6-monthly, annual or biannual monitoring frequencies would all be safe and cost-effective. In the interviews, there was variability among patients and HCPs in which of these were acceptable. All patients were comfortable with reducing the frequency of their testing to every 6 months; however, some HCPs were hesitant to reduce monitoring beyond 3-monthly for a risk score of 3 or 4. This difference between patients and HCPs is likely because the latter have responsibility for patient safety. With variability in which frequencies were deemed acceptable, as well as preference by some patients and HCPs to have agency and autonomy in the decision-making process, monitoring recommendations may need to be given for risk score ranges. Furthermore, all HCPs were happy to move to annual monitoring for anti-TNFs, suggesting a different recommendation could be made for biologic medications.
Strengths of this study include national recruitment with findings reflecting the experiences of those receiving and providing medical care from different hospital trusts where monitoring practices may differ. There was a broad eligibility criterion for patients with different inflammatory conditions, on different medication types, and different HCP roles representing each disease area, enhancing transferability of the findings. Participants were informed that the interviewer was not involved in the clinical care of patients, which minimised the risk of response bias. Data collection and analysis were conducted concurrently, which allowed for identification of areas requiring further exploration in the following interviews. With no changes to the codebook upon analysis of the final patient and HCP interviews, we can be confident that we reached a sufficient level of data saturation for the findings to be clinically meaningful and transferable. Involvement of a second coder and clinical investigator enhanced the rigour of analysis. Although this was a self-selecting group, our purposive sampling ensured a mix of patients were interviewed with different levels of adherence to their recommended monitoring frequency, and less adherent patients were forthcoming in discussing reasons for non-compliance. Participants in the interview study were at varying risk of AEs, ranging from very low to high.
Limitations include a lack of patient participants with very-high risk score, that is an annual risk score of 3 and above who may feel more cautious about adopting a reduced frequency of monitoring. This is not a serious limitation as there were very few patients with annual risk score of 3 or more (15 or more over 5 years) in WP2. Similarly, there were no participants older than 70 years in age. There were also no patient participants of non-White ethnicity whose views may differ to those who took part. Despite our best attempts, we could not recruit any non-White participants for the interview studies. Additionally, these findings should not be extrapolated to other drugs used to manage inflammatory conditions, for example, calcineurin inhibitors that were not included in this study.
A risk-stratified approach to monitoring was acceptable to patients and health professionals. Recommendations towards adopting such a strategy in clinical practice should consider the preferred gaps in testing, tapering, clinician and patient autonomy in deciding appropriate monitoring frequencies, and flexibility to change monitoring frequency according to clinical need. A change in clinical practice will require a carefully considered communication strategy to ensure it will be trusted and adopted across the board.