Baseline, trial results and clinical effectiveness

Recruitment to the C-GALL trial

Participants were recruited to the C-GALL trial between August 2016 and November 2019. In total, 436 participants were randomised from 20 centres within the UK (Table 2), 218 to each group. The trajectory of recruitment from all centres is shown in Figure 2. Initial recruitment was limited to four pilot centres (Aberdeen Royal Infirmary, Nottingham City Hospital, Musgrove Park Hospital, Taunton and Royal Free Hospital, London). Roll-out to other centres began in April 2017. Recruitment was slower than originally projected and an 18-month extension to recruitment was requested and approved. A revised recruitment projection was initiated in March 2018 (see long dashed line in Figure 2).

FIGURE 2 Recruitment over time. (Figure not reproduced here; see the original article.)

Participant flow

The Consolidated Standards of Reporting Trials (CONSORT) diagram for the C-GALL trial is shown in Figure 3. There were 2667 patients identified to be potentially eligible for inclusion into the trial, of which 1298 were excluded. The main reasons for patients’ exclusion were that they were ineligible (647/1298, 49.9%). Ineligibility was due to unconfirmed symptomatic gallstones (233/647, 36.0%), clinical diagnosis of symptomatic gallstone disease not confirmed by imaging (144/647, 22.3%) and being medically unfit for surgery (105/647, 16.2%). Of the 1369 eligible patients, 933 were not randomised with 910/933 (97.5%) having a preference. Main preference reasons were participants preferred laparoscopic cholecystectomy (538/910, 59.1%), observation/conservative management (167/910, 18.4%) and did not want to be randomised (91/910, 10.0%). Further details on reasons why patients were excluded and not randomised are shown in Appendix 3, Table 38.

FIGURE 3 Participant flow (CONSORT) diagram. (Figure not reproduced here; see the original article.)

Of the 436 participants randomised, one participant was a post-randomisation exclusion in the observation/conservative management group due to the participant immediately withdrawing from the study due to a previously unstated preference for laparoscopic cholecystectomy. One participant was randomised twice in the laparoscopic cholecystectomy group through error. At 24 months, in the observation/conservative management group, 136 (62.1%) participants responded to participant questionnaires (PQs) and 12 (5.5%) had declined further follow-up. In the laparoscopic cholecystectomy group, 138 (63.6%) participants responded to PQs and 11 (5.1%) had declined further follow-up and one participant had died. Appendix 3, Table 39 shows the response rates for each of the follow-up time points.

Baseline characteristics

The baseline characteristics are shown in Table 3. Overall, the randomised groups were well balanced in baseline characteristics. The mean age of participants was approximately 50 years, over 78% were female and over 85% were white. In the observation/conservative management group, 60.4% had a normal gallbladder wall confirmed by transabdominal ultrasonography or another imaging technique compared with 55.3% in the laparoscopic cholecystectomy group. The mean SF-36 norm-based bodily pain score was 44.5 (SD 11.7) in the observation/conservative management group and 43.3 (SD 11.1) in the laparoscopic cholecystectomy group.

Non-responders to questionnaires tended to be younger (mean 46 vs. 52 years), have worse SF-36 bodily pain (mean 41.8 vs. 46.6) and worse disease-specific measures at baseline compared to responders.

Treatment received

At 18 months, 54 (24.9%) in the observation/conservative management group and 146 (67.3%) in the laparoscopic cholecystectomy group received surgery. Further surgery details up to 18 months are provided in Appendix 3, Table 40. At 24 months, 64 (29.5%) in the observation/conservative management group, and 153 (70.5%) in the laparoscopic cholecystectomy group received surgery with a median time to surgery of 9.0 months (IQR 5.6–15.0) and 4.7 months (IQR 2.6–7.9), respectively. The majority of the surgical operations were elective in both groups, and over 95% were performed laparoscopically. In the observation/conservative management group, surgery was straightforward for 36/64 (56.3%) compared with 94 (61.4%) in the laparoscopic cholecystectomy group. In the observation/conservative management group, 54/64 (84.4%) did not have a normal gallbladder and of these 44 (81.5%) had chronic cholecystitis. In the laparoscopic cholecystectomy group, 143/153 (93.5%) did not have a normal gallbladder and of these 130 (90.9%) had chronic cholecystitis. Further surgery details at 24 months are shown in Table 4.

For those who had not had surgery by 24 months (Table 5), 15 (6.9%) were on a waiting list and 7 (3.2%) declined any further follow-up from hospital records in the observation/conservative management group; 13 (6.0%) and 5 (2.3%) in the laparoscopic cholecystectomy group.

Primary outcome

Figure 4 shows the mean of SF-36 norm-based bodily pain score. At 18 months, the mean SF-36 norm-based bodily pain score was 49.4 (SD 11.7) in the observation/conservative management group and 50.4 (SD 11.6) in the laparoscopic cholecystectomy group (Table 6). For the primary analysis, the mean AUC over 18 months was 46.8 for both groups with no difference: MD –0.0, 95% CI (–1.7 to 1.7); p-value 0.996.

FIGURE 4 Mean SF-36 norm-based bodily pain score up to 24 months. Higher score indicates better quality of life. Bars represent ± SD. (Figure not reproduced here; see the original article.)

Sensitivity and complete case analyses are shown in Appendix 3, Table 41.

Compliance analysis

Figure 5 shows the mean of SF-36 norm-based bodily pain score by compliance, where compliance was defined as participants receiving their allocated treatment within 24 months (excluding those two participants in the cholecystectomy group where surgery was done as an emergency case). For the compliance analysis of AUC over 18 months, there was no evidence of a difference: MD –0.0, 95% CI (–5.1 to 5.1); p-value 0.997 (see Table 6).

FIGURE 5 Mean SF-36 norm-based bodily pain score up to 24 months by compliance. Higher score indicates better quality of life. Compliance was defined as participants who received their allocated treatment within 24 months (apart from the laparoscopic cholecystectomy group if surgery was an emergency). (Figure not reproduced here; see the original article.)

Subgroup analyses

Figure 6 shows the prespecified subgroup analyses for AUC SF-36 norm-based bodily pain score, gender, age (< 35, 35–64, ≥ 65 years) and ethnicity (white vs. other) at 18 months (primary outcome). Overall, there was no evidence that the treatment effect was moderated by any subgroups. Appendix 3, Figures 18 and 19 show the subgroup analyses for the sensitivity and complete case analyses.

FIGURE 6 Subgroups for observation/conservative management vs. laparoscopic cholecystectomy up to 18 months for AUC SF-36 bodily pain primary analysis. Boxes represent differences in AUC and lines are confidence intervals. (Figure not reproduced here; see the original article.)

Secondary outcomes

Patient-reported quality of life

The mean AUC SF-36 norm-based bodily pain score over 24 months was 47.2 (SD 8.6) in the observation/conservative management group and 47.3 (SD 8.7) in the laparoscopic cholecystectomy group and with no evidence of a difference for the primary analysis: MD –0.1, 95% CI (–1.8 to 1.7); p-value 0.948. Sensitivity and complete case analyses are shown in Appendix 3, Table 42.

For the SF-36 norm-based scores (apart from bodily pain), some small differences were observed as statistically significant at 18 months, but these disappeared at 24 months (Table 7). None of the observed effect sizes were clinically important.

The mean CSQ at 18 months was 21.3 (SD 21.0) for the observation/conservative management group and 15.8 (19.7) for the laparoscopic cholecystectomy group and showed evidence of a difference in favour of laparoscopic cholecystectomy: MD 6.6, 95% CI (1.9 to 11.3); p-value 0.006. At 24 months, again there was evidence of a difference in favour of laparoscopic cholecystectomy: MD 9.0, 95% CI (4.1 to 14.0), p < 0.001. There was a similar pattern for the persistent symptoms score.

Complications

At 18 months, 32 (10.1%) participants in the observation/conservative management group had had a complication compared with 44 (25.3%) participants in the laparoscopic cholecystectomy group, with no evidence of a difference between groups: RR 0.72, 95% CI (0.46 to 1.14); p-value 0.17 (further details are shown in Appendix 3, Table 43). At 24 months, there were two additional complications in the laparoscopic cholecystectomy group with no evidence of a difference between groups: RR 0.69, 95% CI (0.44 to 1.09); p-value 0.11. Details of the complications are shown in Table 8. In regard to complications that occurred before laparoscopic cholecystectomy (including in participants who did not have a laparoscopic cholecystectomy), they occurred in 25 (11.5%) participants in the observation/conservative management group and 11 (5.1%) in the laparoscopic cholecystectomy group, with the majority of the complications being either cholecystitis or biliary colic. For intraoperative complications, 9 participants (4.1%) had a complication in the observation/conservative management group and 24 (11.1%) in the laparoscopic cholecystectomy group with the majority being bile/stone spillage from the gallbladder. For postoperative complications, there were 7 (4.1%) and 24 (11.1%) participants, respectively, with bile/stone spillage from gallbladder as the main complication. For postsurgery complications, there were 2 (0.9%) and 3 participants (1.4%), respectively, within 30 days of discharge, and 1 participant (0.5%) after 30 days in both groups. There was one cardiovascular-related death in the laparoscopic cholecystectomy group (myocardial infarction). By treatment received at 18 months, there were 8/234 participants (3.4%) who had a complication in the observation/conservative management group and 68/200 (34.0%) in the laparoscopic cholecystectomy group. At 24 months, there were 7/217 (3.2%) and 71/217 (32.7%) participants who had a complication, respectively (see Appendix 3, Tables 44 and 45 for further details).

Need for further treatment

At 18 months, there were 9/202 (4.5%) participants in the observation/conservative management group who had had a further treatment and 12/203 (5.9%) in the laparoscopic cholecystectomy group with no evidence of a difference: RR 0.75, 95% CI (0.31 to 1.78); p-value 0.509 (further details are shown in Appendix 3, Table 46). At 24 months, there were 10/202 (5.0%) participants in the observation/conservative management group who had further treatment and 16/203 (7.9%) in the laparoscopic cholecystectomy group with no evidence of a difference: RR 0.62, 95% CI (0.28 to 1.38); p-value 0.242 (Table 9). The main treatments were additional pain relief, antibiotics and endoscopic retrograde cholangiopancreatography (ERCP).

Appointments with health professionals, medications prescribed and further investigations

Details of appointments with health professionals, medications prescribed and further investigations related to their gallstones by 18 months are shown in Appendix 3, Table 47. By 24 months (Table 10), 108/202 (53.5%) of participants in the observation/conservative management group and 127/203 (62.6%) in the laparoscopic cholecystectomy group required appointments with healthcare professionals with the majority visiting their GP, or an NHS hospital A&E or referred to outpatient department. Further medication for their gallstones was prescribed to 67/202 (33.2%) participants in the observation/conservative management group and 59/203 (29.1%), respectively, and further investigation was required in 11/202 (5.4%) participants in the observation/conservative management group and 10/203 (4.9%) in the laparoscopic cholecystectomy group.

Impact of coronavirus disease 2019

All randomised participants reached the 3-month follow-up time point before the COVID-19 pandemic began (defined as 11 March 2020 by the WHO).38 In addition, 97/167 (58.1%) in the observation/conservative management group and 102/161 (63.4%) in the laparoscopic cholecystectomy group had completed their 18 months of follow-up before this date (see Appendix 3, Table 48). The AUC difference for the SF-36 norm-based bodily pain score by 18 months was MD –0.1, 95% CI (–1.8 to 1.6); p-value 0.94 in the subset of data pre-COVID-19. There was no evidence that this differed from the intention-to-treat (ITT) estimate.

In summary, the C-GALL group delivered the trial in the UK NHS setting, across 20 secondary care sites, and randomised 434 participants. We found no evidence of differences in bodily pain (primary outcome), QoL, complications or in need for further treatment between the policies at up to 24 months follow-up. There was statistically significant evidence that gallbladder-specific measures of quality of life improved in the cholecystectomy randomised group at 24 months.