Scientific Papers

Recruiting women with ductal carcinoma in situ to a randomised controlled trial: lessons from the LORIS study | Trials


The CONSORT flow diagram is shown in Fig. 2. Of the 227 eligible women who had confirmed low-risk DCIS, 183 (81%) accepted randomisation and 44 (19%) declined. Two acceptors could not be randomised due to COVID-related issues, so 181 acceptors were randomised. The CTQ was completed by 175/181 (97%) acceptors and 21/44 (48%) decliners. Telephone interviews were conducted with 101 women, including 92/181 (51%) acceptors, and 8/44(18%) decliners. At the time of interview, four of the women who had been randomised to standard treatment had decided not to go ahead with surgery. One of the four also withdrew her interview data so is not included in that analysis below.

Fig. 2
figure 2

Participants

The median age of the 181 acceptors was 58 years (range 46–88 years) and most were post-menopausal (117/181, 65%, post-menopausal; 19/181, 10%, peri-menopausal; 34/181, 19% pre-menopausal; 11/181, 6% menopausal status not known). The median age of the 92 acceptors interviewed post randomisation was 60 years (range 48–88 years), 62/92 (67%) had a partner and 55/92 (60%) were employed. The median age of the 8 decliners interviewed post randomisation was 55 years (range 48–70 years), 5/8 (63%) had a partner and 5/8 (63%) were employed.

Table 1 summarises scores on the STAI, EQ-5D and the SF36-v2 for acceptors. For STAI-trait, 53/174 (30%) scored 40 or above, indicating high trait anxiety and for STAI-state, 75/175 (43%) scored 40 or above indicating high state anxiety. On the EQ-5D, 61/176 (35%) had full health (a profile of 11111) and no women were in the worst possible state (55555). On the SF36-v2, 141/174 (81%) had a PCS measure which was the same or better than the general population. For the MCS measure, 122/174 (70%) were the same or better than the general population.

Table 1 Baseline measure scores for women who consented to randomisation

Clinical Trials Questionnaire

Table 2 displays the frequency of agreement (‘strongly agree’ or ‘agree to some extent’) with each reason and the frequency of endorsement as the most important reason, according to whether participants accepted or declined to be randomised. Although significantly more decliners were concerned about randomisation (76%) and wanted the doctor to choose their treatment (71%), nearly half of acceptors were also worried about randomisation and 39% wanted the doctor to choose their treatment.

Table 2 Clinical Trials Questionnaire responses

There were nine statements which significantly more acceptors agreed with, including the belief that the trial offered the best treatment available (91% vs. 29%), and wanting to help both the doctor’s research (95% vs. 55%) and others with the same illness (99% vs. 57%). These three reasons were also most frequently chosen as the most important reason for agreeing to be randomised. For decliners, the most frequently cited key reasons were the influence of others, such as family or friends (28%) and the idea of randomisation being worrying (17%). Most women, regardless of whether they agreed to be randomised, felt they could trust the doctor providing treatment (98%) and that s/he had told them what they needed to know about the trial (97%).

Most participants who listed other reasons for agreeing to be randomised indicated they thought the trial offered the best (safest) way of avoiding surgery. Some noted the benefits of extended monitoring. Whilst many LORIS participants found randomisation worrying, two women noted that they saw this as a benefit as they were undecided about whether they wanted surgery or not. Most of the women who declined randomisation and offered an additional reason, made it clear that they definitely wanted surgery. For one woman, the opposite was true—she did not want surgery.

Interview results

Reasons for accepting or declining randomisation

Interviews took place a median of 22 days after randomisation. During the interview, the women were given the opportunity to elaborate further on the reasons they had given on the CTQ for accepting or declining randomisation. The three most frequent reasons for accepting randomisation were the same for this subset of 89 women as the 175 who completed a CTQ. Anticipating future benefit for others was selected by 19/89 (21%) as the most important factor, e.g. “I’ve got 9 grandchildren and 6 of them are girls. What if same thing happens when they grow up and the only option was the operation? I hope that anything I do helps, not just for my family, but all women.” (P24). Helping with the doctor’s research was also selected by 19/89 (21%), e.g. “Well it’s like anything else, if we don’t have individuals giving their time and biopsy samples, we’ll never find out if these dormant cells will turn into invasive cancer. It will help future generations, I think this is what you need. I think there should be more research in general and more money should be spent on research. My husband was worried that I’d get monitoring, but I was happy to join either part of the trial.” (P102). And for another 19/89 (21%), the most important reason for accepting randomisation was the belief that the trial offered the best treatment e.g. “I thought getting yearly rather than 3 yearly mammograms was a good idea. Getting checked more often was the main reason.” (P48).

The four women who withdrew from their treatment allocation after randomisation had all been allocated to standard treatment but did not want surgery. For example, one woman talked about how surgery did not fit in with her busy lifestyle as a self-employed fitness instructor and that she felt it was currently unnecessary. Nearly all the decliners interviewed (7/8) were sure they wanted surgery, generally for peace of mind or because of family considerations, with just one woman declining randomisation because she did not want surgery.

Information provision

For 83/89 (93%) acceptors who were interviewed, the trial had been discussed with more than one HCP, including research nurses (82/89, 92%), surgeons (75/89, 84%), radiologists (11/89, 12%), and an oncologist (1/89, 1%). Most acceptors (83/89, 93%) did not feel that their decision to take part had been influenced by their discussion with an HCP. Similarly, all but one of the decliners had discussed the trial with more than one HCP. Two out of the nine women in this group felt their decision had been influenced at least to some extent by an HCP: one because the surgeon had made it clear that the patient had to be content with both options on offer before entering the trial and one because the surgeon had mentioned surgery before anything else.

All the women who were interviewed remembered receiving and reading at least one of the three written information documents: 89/100 (89%) reported reading the PIL, 91/100 (91%) reported reading PIS-A and 97/100 (97%) reported reading PIS-B. Some women talked about reading everything very carefully, and even going online to find out more about both DCIS and the LORIS study or sharing the information with friends who were HCPs. Others described skimming the leaflets or just reading what they considered to be important. One woman explained how she had been too fearful to read the trial sheets initially but had felt comforted about the whole process after a nurse went through everything verbally. Another individual initially found the information too much but followed a friend’s advice to read each PIS bit by bit.

Further interview questions focussed specifically on PIS-B, which explained randomisation. Most of the women (73/100, 73%) shared this leaflet with someone else, usually family members but also friends and colleagues. Some who chose not to share PIS-B talked about it being a private matter or not wanting to worry anyone. Most found PIS-B either very or somewhat useful (97/100, 97%) and either very or somewhat clear (98/100, 98%). Two participants talked about finding the concept of randomisation very difficult to understand. Some wanted more information, e.g. about extra tissue being taken, the risks and side effects of surgery, personal benefits of joining LORIS, whilst two women thought there was too much information or that it was too technical. Despite PIS-B being generally perceived as useful and clear, 39/89 (44%) of acceptors did not think it had helped them make a decision, most commonly because they had already decided to join the trial.

Most of those interviewed (70/100, 70%) had watched the patient information film and often shared it with a family member, friend or colleague (45/70, 64%). Many who did not watch the film had already made their decision or did not have a DVD player and/or access to the internet. All film viewers found it either very or somewhat clear and useful. The part of the film which was most commonly cited as being particularly useful (37/70, 53%) was the Q&A section. Very few, 4/70 (6%), identified aspects of the film as being unhelpful. Two were concerned about terminology, one woman noted she already knew some of the information from the PISs and one felt some of the questions in the Q&A session were poor. Despite positive opinions from the film viewers, there was a 50:50 split in terms of whether the film helped them decide whether to join the trial; most who did not feel the film had helped with decision making had already decided what to do before viewing it. Nevertheless, many described feeling reassured that they had made the right decision having watched the film, e.g. “I’d already decided to take part at that point. DVD was really important for me to see the doctors’ faces, who are doing the study, Good that the DVD is very short, but really helpful. Made me feel part of something.” (P88)

Three additional questions were included in the final 65 interviews, if relevant. One asked about the timing of the PIL: one woman felt it had been given too early but all the other women who remembered receiving it felt it was about right. Some talked about finding the PIL useful because they had been shocked after receiving their mammogram results but appreciated having written information to take away and digest. The other two questions related to whether the women were aware that their biopsy results could make them ineligible for LORIS and how clear they felt this information was in the PIL. Most of the women who answered these questions knew that their eligibility for LORIS depended on their biopsy results (56/61, 92%) and felt that information explaining this in the PIL was either very or somewhat clear (47/51, 92%). Some highlighted how helpful it had been for an HCP to also explain everything orally.

Sites

Initially, four communication workshops were held, two in London, one in Birmingham and one in Wycombe, with a total of twenty delegates attending. Two further workshops, both in London, held as part of the Recruitment Recovery Plan, were attended by 17 health care professionals. The delegates were a mixture of trial coordinators, research nurses, radiologists, radiographers, and surgeons. Analysis of the pre- and post- workshop questionnaires resulted in a significant improvement in knowledge (t = 4.7, d.f. = 28, p < 0.001) and confidence about discussing clinical trials in general (t = 3.1, d.f. = 24, p = 0.005) and randomisation in particular (t = 2.6, d.f. = 25, p < 0.015).

The response rate to the postal survey about which aspects of the LORIS protocol was perceived as challenging was good from PIs, 34/47 (72%), and site leads, 45/53 (85%), but poor from other members of the recruiting team, 12/197 (6%). Most respondents reported they were well aware of the aims of LORIS (88/91, 97%) and comfortable with these (87/91, 96%). However, only about half stated they themselves would participate in LORIS, if relevant, or encourage a friend or relative, (45/87, 52%), many were unsure (35/87, 40%) and a few would not participate (7/87, 8%).

Most respondents were clear about the study logistics pathway (78/91, 86%) and thought the process of obtaining histopathology slides for central review was streamlined (64/85, 75%). However, the need for a second biopsy to confirm eligibility was sometimes or often thought to deter team members from recruiting patients by nearly half of respondents (37/83, 45%), and sometimes or often thought to act as a deterrent to potential participants by most respondents (68/83, 82%). Another challenge for sites was changes in their key staff since LORIS opened at their site, reported by 38/89 (43%) of respondents.

When asked about the most challenging aspect of recruiting patients (free text item), the most common reason given was the low number of eligible patients (34/81, 42%). Additionally, some respondents (12/81, 15%) noted that many patients were deemed ineligible following central review. The second most common barrier, reported by 18/81 (22%), was that many women had a strong preference for a particular treatment arm. The challenge of trying to explain active surveillance (7/81, 9%), and local logistic issues (6/81, 7%) were also barriers reported by more than one respondent. In the semi-structured phone interviews, held with 20/47 PIs (43%) and 40/53 site leads (75%), the same barriers were identified more frequently. These interviews also provided an opportunity for participants to describe in more detail specific local challenges. Some interviewees suggested that it would be beneficial to educate the general public about DCIS and the LORIS trial in general. There was also an issue noted with the use of the word carcinoma in DCIS and the use of the term ‘cancer’ in the patient information provided by hospitals, in that this could lead to more women having a strong preference for surgery.

Based on the information gained from the survey and interview, refresher SIVs were suggested for six sites and attendance at a communication workshop for 14 sites. Online refresher SIVs were held for all six sites but only five sites sent delegates to a communication workshop.



Source link