Lack of knowledge/training
The survey results showed that 14% (
n = 18) of responders were not familiar with the term RBM. Of the participants who did not conduct a risk assessment in the most recent clinical trial that they worked on (
n = 35), 17% felt that they did not have the expertise to perform a risk assessment (Table
2). Over 80% (
n = 114) of survey responders categorised barriers to implementing centralised monitoring. Almost two thirds of these participants (62%,
n = 71) identified lack of education as a very important barrier (Table
3). The interview data confirmed that several participants had not used RBM in past trials because they were not familiar with this type of monitoring and many did not know that RBM would be introduced in the new ICH-GCP guidelines:
Table 2
Reasons why survey responders did or did not conduct a risk assessment prior to developing the monitoring plan
To improve patient safety | 43 (86%) | Question not relevant, developing monitoring plan is a sponsor duty | 15 (43%) |
To improve data accuracy | 32 (64%) | It is not a GCP requirement | 7 (20%) |
To fulfil GCP requirements | 29 (58%) | Do not have the expertise to perform a risk assessment | 6 (17%) |
To determine a schedule for on-site monitoring visits | 21 (42%) | It is too time consuming | 6 (17%) |
To fulfil HPRA/IMB requirements | 20 (40%) | It will not improve patient safety | 2 (6%) |
To reduce monitoring costs | 8 (16%) | It is too expensive | 1 (3%) |
Table 3
Perceived problems associated with the implementation of centralised monitoring (n = 114)
Lack of education and training in centralised monitoring | 71 (62%) | 36 (31%) | 8 (7%) |
Cost associated with centralised monitoring | 45 (40%) | 54 (48%) | 13 (12%) |
Information Technology (IT) demands of centralised monitoring | 53 (46%) | 53 (46%) | 9 (8%) |
Workload associated with centralised monitoring | 47 (41%) | 48 (16%) | 18 (16%) |
‘Well, just from talking to yourself, I have to admit, prior to that I hadn’t heard about this, so I wasn’t aware that the GCP was going to be changing’ (Study physician-1).
Several interviewees, who had not conducted RBM in past trials, felt that they did not have sufficient RBM training to confidently perform RBM in future trials:
‘It would come down to the practical aspects on how is risk defined …what information are people using to make that judgement. How is it actually implemented? But ultimately I’d have to understand that before I could say I was happy to do it’ (PI-1).
They did not feel able to classify clinical trial risks and to translate these risks into monitoring activity. Similarly, some interviewees who had conducted RBM in past trials still felt ill-equipped to perform RBM in their future trials:
‘We would say we have conducted a type of risk-based monitoring, but it’s getting to the actual nitty-gritty of exactly what fields you’re going to look at and exactly what parameters are in those fields. I would say that I’d be still a bit unsure of that’ (Nurse-1).
Survey responders reported having limited experience of using centralised monitoring for essential monitoring activity such as assessing protocol compliance, inspecting informed consent and recording pharmacovigilance information. Lack of education was the main reason that survey participants did not perform centralised monitoring (Table
3). A small number of participants from the five CRF/Cs (
n = 17) reported having a Standard Operating Procedure (SOP) for centralised monitoring in their CRF/C. However, over a third of participants (
n = 48) were unsure if such a SOP existed in their CRF/C. Analysis of the qualitative interviews showed that some study nurses and monitors did not know how centralised monitoring could replace on-site monitoring. One participant felt that on-site monitoring offered better governance of junior clinical trial staff. This participant felt that centralised monitoring would result in monitors having less oversight of clinical trial activity:
‘As sponsor, all of the monitoring is on-site, and that’s for two reasons…, because it’s our first time working with a lot of these investigators and we’re not sure of their experience in running regulated trials, we want to make sure that they understand what’s required and what they need to do in terms of quality’ (Monitor-2).
Increased cost caused by greater Information Technology (IT) demands
Almost half the survey responders identified IT demands (46%,
n = 53) and cost (40%,
n = 45) as problems associated with the implementation of centralised monitoring in past and future clinical trials (Table
3). The interview data revealed that this perception was related to higher costs associated with EDC systems. Some interviewees felt that centralised monitoring would be costly to run as they would have to store trial data on an EDC system:
‘As sponsor, all of the monitoring is on-site…, because we don’t have electronic data capture in any of these studies because they’re not commercial studies – they’re usually grant funded, or just the PI – so there’s very little money, you’re using paper CRF’.(Monitor-1).
This was a particular concern for trialists working on smaller trials. They felt that they would not have sufficient budget to support an EDC system and were only resourced to conduct on-site monitoring:
‘Some of the eCRFs, let’s say that company that we had, you could be talking nearly half a million, a million to get it up and running, and what small study has that if you’re talking about an oncology study which has maybe 10 patients coming into it? An eCRF is not going to be worth the set-up costs. So they’ll stick to the paper’ (Monitor-3).
Increased work load
Survey findings showed that perceived work load was the main reason why responders did not conduct a risk assessment prior to developing the monitoring plan for their most recent trial (Table
2). Forty-one percent of survey responders (
n = 114) thought that increased workload was a barrier associated with the implementation of centralised monitoring (Table
3). Interviewees, who had previously conducted centralised monitoring, felt that it resulted in more administration work for trial sites as they had to support trial monitors by scanning and uploading site documents to EDC systems:
‘I noticed one of the girls downstairs was saying in the last couple of weeks… this company kept saying, “We still don’t have the CV,” and she’d sent it three times to them. So you need to have good people at the other side doing the monitoring and stuff like that. It’s just if it’s maybe stuff from the trial master file that they’re not here checking and consent forms and that. That probably might add some work’ (Nurse-3).
Some interview participants felt that sponsors would use RBM as an excuse to perform less on-site monitoring and more remote monitoring. These participants felt that a reduction in on-site visits would results in a trial monitors spending less time on site checking trial documentation such as patient Consent Forms. These participants felt that study nurses may be expected to do extra administration tasks to support trial monitors perform remote monitoring:
‘They have these centralised systems now where everything is stored centrally and it’s, like, “Logon and you’ll find the latest version of your protocol”. So you have to complete training for that system, you have to logon every time the new protocol is available or whatever. The onus is on the site to print it off. The onus is on the site to do everything and it’s just more and more it’s on the site, and we are not paid adequately for everything that we are being requested to do. It’s our admin staff as well. It’s like they’re just working for the pharma companies. There’s just a huge amount of resources, and it’s not accounted for’ (Nurse-4).
Lack of verification
The survey found that 27% (
n = 35) of responders did not conduct a risk assessment prior to developing the monitoring plan for their most recent clinical trial. Some participants did not conduct a risk assessment because they felt that it was not a GCP requirement and would not improve patient safety (Table
2). However, these participants did use an informal process to determine what level of on-site monitoring was required for their clinical trial. Also, 21% of survey responders (
n = 28) reported previous RBM experience.
Interview analysis showed that participants perceived a lack of scientific evidence supporting RBM and saw this as a potential barrier to its implementation in their future clinical trials. Many felt that sufficient proof did not exist to confirm that RBM was at least as effective and efficient as the 100% SDV on-site monitoring process that they currently used:
‘So it’s just our experience that the more frequent the monitoring the better. I have a negative attitude towards already a negative perception of the risk-based monitoring because 100% source data verification is what I would prefer’ (Nurse-1).
Some interviewees believed that RBM would lead to a greater reliance on centralised monitoring and a move away from on-site monitoring:
‘I know the new ICH-GCP guidelines are more into the technology, and I know that’s the way we’re going and things like that. At the end of the day, I don’t think it fully replaces the on-site’ (Nurse-2).
Many felt that the merits of centralised monitoring had yet to be proven and so were not comfortable conducting RBM in future trials if it meant fewer on-site visits:
‘I suppose the fact that things are going more electronic and it is more EDC-based. It’s the management of stuff that cannot be converted into EDC and how that’s going to be verified and how that’s going to be monitored’ (Monitor-1).