- Email: [email protected]
How to decide to undertake a randomized, controlled trial of stent or surgery in colorectal obstruction
How to decide to undertake a randomized, controlled trial of stent or surgery in colorectal obstruction Michael K. L. Suen, MBBS, MS, FRACS,a,c Assad Zahid, BSc(Med), MBBS,a Jane M. Young, MBBS, MPH, PhD, FAFPHM,b Laura Rodwell, BSc(HONS), MPSYCHOL(ORG), MBIOSTAT,b Michael J. Solomon, MB, BCH, BAO(Hons), FRACS, MSc(Clin Epid),a,b and Christopher J. Young, MBBS, MS, FRACS, FACS,a,c Sydney, Australia
Background. Colorectal stents have been available as a management option in obstruction for 23 years, yet there is little randomized evidence of their effectiveness. This study investigated current surgeonrelated barriers to conducting a randomized, controlled trial (RCT) of colorectal stent insertion for obstruction in patients with colorectal cancer. Methods. A binational survey of current members of the Colorectal Surgical Society of Australia and New Zealand was conducted by a mailed questionnaire assessing perceived barriers to adoption of colonic stents and willingness to participate in future multicentre randomized controlled trials, and surgeons’ treatment preferences in 16 hypothetical clinical scenarios. Results. Of 148 eligible surgeons, 96 (65%) responded. Colonic stenting was available to 98% of respondents. In the clinical setting of colorectal obstruction, only 29% (95% CI, 20–39%) of surgeons expressed a willingness to participate in a RCT involving colonic stents in the curative setting. More than 70% of surgeons preferred the use of stents in unfit patients for palliation, and preferred surgery in fit patients with curable disease. In the curative setting, most respondents considered colonic stents not cost effective (90%; 95% CI, 82–94%) and believed that their patients would not prefer stents over surgery (80%; 95% CI, 71–87%). Conclusion. This study highlights the limitation to conducting a future randomized controlled trial to assess the efficacy of colonic stenting, especially in the curative setting, based on surgeon preference, despite the lack of level I evidence. (Surgery 2015;157:1137-41.) From the Department of Colorectal Surgery,a Royal Prince Alfred Hospital, NSW, Australia, the Surgical Outcomes Research Centre (SOuRCe),b University of Sydney, Sydney, Australia, and University of Sydney,c Sydney, Australia
LARGE BOWEL OBSTRUCTION is a surgical emergency where traditional operative management is usually associated with a high morbidity and mortality.1 Colonic stents have provided an alternative treatment option for obstruction since first described by Dohmoto in 1991.2 Despite multiple studies demonstrating the safety and efficacy of stenting to manage malignant large bowel obstruction, randomized, controlled trials (RCTs) to confirm its effectiveness are lacking in the literature.3-6 Accepted for publication January 7, 2015. Reprint requests: Christopher J. Young, MBBS, MS, FRACS, FACS, Department of Colorectal Surgery, Royal Prince Alfred Hospital Medical Centre, 100 Carillon Ave, Newtown, NSW 2042, Australia. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2015.01.022
Individual equipoise, a measure of clinical uncertainty, occurs when an individual clinician is completely undecided, whereas community equipoise applies when there are differing views among the profession as a whole.7 Preferences strongly correlate with participation in surgical RCTs,8 and when >70% of expert clinicians prefer 1 treatment modality, a human trial is considered to be unethical.7 Given colonic stents have been a treatment option in large bowel obstruction for 23 years, it is important to assess surgeons’ preference of treatment modalities in obstruction, and any biases, before further trials are conducted in this area. The aim of this study was to assess the willingness of surgeons toward conducting a future RCT with regard to the use of stenting in colorectal obstruction, and to identify the preference of surgeons toward stenting versus surgery in varying scenarios of colorectal obstruction, specifically SURGERY 1137
1138 Suen et al
considering patient’s age and fitness for a procedure, the degree of colorectal obstruction, and the difference between palliative and potentially curative treatment. METHODS All members of the Colorectal Surgical Society of Australia and New Zealand were surveyed in this study. Each surgeon was mailed a covering letter, a questionnaire and a reply-paid envelope. Two mailed reminders were used to follow up with nonresponders. Development of the questionnaire. The questionnaire was divided into 3 parts. (1) Surgeons were asked whether they would participate in an RCT or a non-RCT involving colonic stents in curative and palliative settings. (2) Surgeons were asked to rate 6 potential barriers to the use of colonic stenting, including issues of patients’ preference, risks associated with the use of colonic stents, evidence available in the literature, cost effectiveness of using colonic stents, surgeon training, and availability of the colonic stent. Respondents were asked to indicate their agreement with each statement on a scale from 1 to 5, with ‘‘1’’ being ‘‘strongly disagree’’ to ‘‘5’’ being ‘‘strongly agree.’’ (3) Surgeons were presented with 16 different clinical scenarios which were developed by varying 4 clinical factors, namely patient age, degree of obstruction, American Society of Anesthesiologists grade and curability of the colorectal cancer. These 4 clinical factors were chosen on the basis of a literature review9,10 and were considered to be the most important factors affecting decision making in this setting. For each scenario, surgeons were asked to rate their level of clinical certainty about which of 2 treatment options---colonic stent and open surgery with primary anastomosis or stoma---was their preferred option. To indicate their preference, surgeons were asked to mark an ‘X’ on the line of a visual analog scale, with the treatment options at either end of the scale. The scale was centered on zero to represent ‘‘completely undecided’’ and marked from ‘‘1’’ to ‘‘5’’ toward each end to represent increasing certainty in each treatment approach.11 Statistical analyses. Characteristics of the responding surgeons were tabulated. Descriptive statistics (proportion and 95% confidence intervals) were calculated. The proportion of surgeons who would participate in a RCT was calculated. Individual equipoise in each scenario in part 1 of the questionnaire was measured by the proportion of surgeons choosing the completely undecided
Surgery June 2015
Table I. Demographic and other characteristics of responding colorectal surgeons (n = 96) Characteristic
n
Mean age (y) Sex, n (%) Male Female Practice location* Capital city Other urban area Rural area Appointment type* Conjoint staff/academic VMO/consultant Staff specialist Salaried university academic Hospital characteristics* Tertiary referral teaching hospital District general hospital Private hospital Availability of colonic stents* Yes No Specialists performing colonic stenting in the respondent’s hospital* Colorectal surgeons only Gastroenterologists only Radiologists only Surgeons and gastroenterologists Surgeons and radiologists All 3 specialties
49
%
87 9
91 9
61 24 5
68 27 5
11 64 11 4
12 71 12 5
60 7 24
67 7 26
89 2
98 2
43 21 2 17 3 3
48 24 2 19 4 4
*Information was incomplete in these sections in 7 patients. VMO, Visiting medical officer.
category.10 To assess the community equipoise, surgeons’ responses were categorized into 1 of 3 groups: (1) preferred treatment with open surgery with stoma or primary anastomosis, (2) completely undecided (ie, indicated zero), and (3) preferred treatment with colonic stent. Based on previous studies,7,10-13 community equipoise is considered lacking in that scenario when >70% of respondents favored 1 treatment option. Analyses were carried out using SPSS statistical software. RESULTS Of 149 active members of the Colorectal Surgical Society of Australia and New Zealand, 1 was no longer resident in Australia or New Zealand and was excluded from the study. Ninety-six members (65%) completed and returned the questionnaire. Personal and professional characteristics of the responding surgeons are shown in Table I. The mean age of participants was 49 years. Colonic stenting was available as a treatment option to 98% of the respondents.
Suen et al 1139
Surgery Volume 157, Number 6
Table II. Willingness to participate in RCTs in different settings Willing to Willing to participate participate non-RCT, RCT, n (%) n (%) Palliative Curable
68 (71) 28 (29)
63 (66) 33 (34)
McNemar test v = 0.52; 1 df; P = .74 v2 = 0.83; 1 df; P = .36 2
RCT, Randomized controlled trial.
RCT participation. Few surgeons (29%; 95% CI, 20–39%) would agree to participate in a trial of colonic stent for patients with potentially curable colorectal cancer. The small increase in the proportion of surgeons willing to participate in non-RCT compared with RCT studies was not significant in either palliative or curative settings (Table II). Barriers. Only a small proportion (18%; 95% CI, 11–27%) of surgeons agreed that the use of colonic stenting is limited by unavailability or inadequate evidence of its efficacy (14%; 95% CI, 8–22%). Slightly more than one-half (52%; 95% CI, 42–62%) thought that surgeons’ proficiency was a barrier to the use of stenting (Table III). Most surgeons (80%; 95% CI, 71–87%) considered that, in the curative setting, patients would not prefer a stent over surgery and even more (90%; 95% CI, 82–94%) believed that stenting was not a costeffective option compared with surgery. A significant proportion of surgeons (68%; 95% CI, 58– 76%) also expressed concerns over the risk of stent-related perforation. Clinical scenarios. Surgeons demonstrated strong treatment preferences, with <20% of respondents in the ‘‘completely undecided’’ group across all 16 scenarios (Table IV). Community equipoise was absent in 2 main clinical situations: (1) patients with palliative colorectal obstruction who were unfit for surgery, because most surgeons would prefer to use a colonic stent, and (2) patients with potentially curable colorectal obstruction who are fit for surgery, for whom most surgeons would opt for immediate surgery. No surgeon was completely undecided about treatment options where the colorectal obstruction was partial, curable, and the patient was fit for surgery, despite their age. Community equipoise was apparent in the ‘‘less extreme’’ clinical scenarios (patients with metastatic cancer who were fit for surgery or patients with curable cancer who were unfit for surgery). The only exception was in the scenario referring to a young patient with partial, curable colorectal obstruction who was unfit for surgery.
Table III. Perceived barriers to colonic stenting in malignant colorectal obstruction Barrier
Agree (%)
Colonic stent not cost effective in curative setting Patient does not prefer colonic stent in curative setting Would not use colonic stent owing to perforation concerns Surgeon’s proficiency will affect colonic stent usage No role in curative setting Colonic stent is not a definitive solution Patient don’t want to wait in curative setting Colonic stent not cost effective in palliative setting Surgeon has more training in surgery, so less like to use stent Colonic stent complication rate too high Patient does not prefer colonic stent in palliative setting Stoma better palliation than colonic stent Colonic stent not readily available Not enough evidence in literature Not use colonic stent due to surgeons bad experience
90 80 68 52 49 47 45 42 40 29 28 25 18 14 5
DISCUSSION This study is the first to assess preferences among surgeons regarding the use of colonic stenting in the management of large bowel obstruction accompanying colorectal cancer. Despite the lack of RCTs, very few surgeons considered that evidence for the efficacy of colonic stenting was inadequate. This is probably the result of longstanding use of stenting in clinical practice and the number of nonrandomized studies in the literature. Although this is potentially viewed as intuitive, this study is the first to demonstrate this practice. Among colorectal surgeons in Australia and New Zealand, there was a general lack of enthusiasm for participating in future RCTs of colonic stents. There were no differences in the participation rate between RCT and non-RCT in both palliative and curative settings. The randomization process was not the cause of the reduction in enthusiasm. This can probably be explained by surgeons’ firmly established views of the role of colonic stenting. Despite the lack of RCT evidence, there was little individual equipoise among surgeons, which calls into question the feasibility of conducting RCTs on this topic.
1140 Suen et al
Surgery June 2015
Table IV. Surgeons’ treatment preferences in different clinical scenarios Level of clinical certainty Clinical scenarios 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
70 70 50 50 70 70 50 50 70 70 50 50 70 70 50 50
yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo;
partial obstruction; metastatic cancer; ASA 4 complete obstruction; metastatic cancer; ASA partial obstruction; metastatic cancer; ASA 4 complete obstruction; metastatic cancer; ASA partial obstruction; metastatic cancer; ASA 1 complete obstruction; metastatic cancer; ASA partial obstruction; metastatic cancer; ASA 1 complete obstruction; metastatic cancer; ASA partial obstruction; curable cancer; ASA 4 complete obstruction; curable cancer; ASA 4 partial obstruction; curable cancer; ASA 4 complete obstruction; curable cancer; ASA 4 partial obstruction; curable cancer; ASA 1 complete obstruction; curable cancer; ASA 1 partial obstruction; curable cancer; ASA 1 complete obstruction; curable cancer; ASA 1
4 4 1 1
Surgery (%)
Undecided (%)
Stent (%)
Evidence of community equipoise?
8 9 15 12 51 40 60 51 66 41 73 50 96 79 96 87
12 8 10 8 19 13 17 14 15 13 10 11 4 12 4 9
80 82 75 80 30 47 23 35 19 46 17 39 0 9 0 4
No No No No Yes Yes Yes Yes Yes Yes No Yes No No No No
ASA, American Society of Anesthesiologists; yo, years old.
Self-reported participation in hypothetical RCTs has been shown to correlate well with recruitment in subsequent RCTs. For example, a preference study predicted 45% of eligible patients would be recruited into a laparoscopic colon cancer study (ALCAS) and when a log of eligible patients of the ALCAS multicenter RCT was taken, this was the exact recruitment.14,15 The same phenomenon happened within our own institution, where our RCT on colonic stenting in the palliative setting had much slower recruitment rate than we initially anticipated. This study has quite simply and adequately confirmed that surgeons’ treatment preferences for different patient populations for the same condition can be measured and this correlates with their likely participation in future RCTs. Feasibility issues rather than methodologic issues (eg, blinding, standardization of techniques) have been shown to be the major barriers to perform surgical RCTs.16 Patient preferences and lack of surgical equipoise are 2 of the top 3 feasibility barriers to surgical RCTs.16 Preference studies like this are a cost-effective way to screen out trials that will never complete. Given that research resources are limited, preference studies should be performed before establishing large surgical RCTs to avoid the often inevitable outcome of lack of recruitment in many surgical trials.17 This is affirmed with this study and the preferences in which to avoid an RCT in this area.
Most respondents had clear opinions on the role of stenting in different clinical scenarios. In patients presenting with colorectal obstruction, >70% of surgeons preferred the use of stents for palliation in the unfit patient, whereas surgery was preferred in the fit patient with curable disease. In the palliative setting, surgeons in our study believed that their patients would prefer a stent rather than surgery, and a stent was thought to be a better palliative option than a stoma, whereas in the curative setting, stents were considered to be expensive, not definitive, and having the potential to cause perforation. In 7 of 8 scenarios involving palliation of fit patients or managing curable disease in unfit patients, surgeons were more divided in their preferences for stenting versus surgery. This reflects the complexity of the decision-making process in these situations. The development of future trials in colonic stenting should focus on these clinical situations where community equipoise was apparent and the result from this study can act as a guide to formulate the inclusion criteria. Although this study achieved a good response rate for mailed surveys, it is possible that responders are not representative of all colorectal surgeons’ attitudes. Furthermore, self-reported treatment preferences may differ from actual clinical practice, although the lack of rigorous evidence to provide a standard of care for the management of malignant large bowel obstruction
Surgery Volume 157, Number 6
likely mitigates the potential for social desirability bias in survey responses. The individual surgeon proficiency with stenting was also not obtained and this may have affected the responses. In conclusion, the low interest in research participation, particularly in the curative setting, represents a major barrier to the conduct of future RCTs to strengthen the scientific evidence base for colonic stenting for people with malignant bowel obstruction. Surgeons preferred the use of stents for palliation in unfit patients presenting with obstruction secondary to colorectal cancer whereas they preferred surgery in fit patients with potentially curable disease. There was community equipoise in the context of the palliation of fit patients and treatment of potentially curable but unfit patients.
REFERENCES 1. Seah DW, Ibrahim S, Tay KH. Hartmann procedure: is it still relevant today? ANZ J Surg 2005;75:436-40. 2. Dohmoto M. New method – endoscopic implantation of rectal stent in palliative treatment of malignant stenosis. Endosc Digest 1991;3:1507-12. 3. Khot UP, Lang AW, Murali K, Parker MC. Systematic review of the efficacy and safety of colorectal stents. Br J Surg 2002; 89:1096-102. 4. Sebastian S, Johnston S, Geoghegan T, Torreggiani W, Buckley M. Pooled analysis of the efficacy and safety of self-expanding metal stenting in malignant colorectal obstruction. Am J Gastroenterol 2004;99:2051-7. 5. Watt AM, Faragher IG, Griffin TT, Rieger NA, Maddern GJ. Self-expanding metallic stents for relieving malignant colorectal obstruction. A systematic review. Ann Surg 2007;246: 24-30.
Suen et al 1141
6. Farrell JJ. Preoperative colonic stenting: how, when and why? Curr Opin Gastroenterol 2007;23:544-9. 7. Johnson N, Lilford R, Brazier W. At what level of collective equipoise does a clinical trial become ethical? J Med Ethics 1991;17:30-4. 8. Young J, Solomon M. Improving the evidence-base in surgery: evaluating surgical effectiveness. ANZ J Surg 2003; 73:507-10. 9. Tan K, Sim R. Surgery for obstructed colorectal malignancy in an Asian population: predictors of morbidity and comparison between left-and right-sided cancers. J Gastrointest Surg 2010;14:295-302. 10. Biondo S, Pares D, Frago R, Martı-Rague J, Kreisler E, De Oca J, et al. Large bowel obstruction: predictive factors for postoperative mortality. Dis Colon Rectum 2004;47: 1889-97. 11. Young J, Harrison J, White G, May J, Solomon M. Developing measures of surgeons’ equipoise to assess the feasibility of randomized controlled trials in vascular surgery. Surgery 2004;136:1070-6. 12. Freedman B. Equipoise and the ethics of clinical research. N Engl J Med 1987;317:141-5. 13. Weijer C, Shapiro SH, Cranley Grass K. Clinical equipoise and not the uncertainty principle is the moral underpinning of the randomised controlled trial. BMJ 2000;321: 756-7. 14. Solomon MJ, Pager CK, Keshava A, Findlay M, Butow P, Salkeld GP, et al. What do patients want? Patient preferences and surrogate decision making in the treatment of colorectal cancer. Dis Colon Rectum 2002;46:1351-7. 15. Abraham NS, Hewett P, Young JM, Solomon MJ. Non-entry of eligible patients into the Australasian Laparoscopic Colon Cancer Study. ANZ J Surg 2006;76:825-9. 16. Solomon MJ, McLeod RS. Should we be performing more randomized controlled trials evaluating surgical operations? Surgery 1995 Sep;118:459-67. 17. Taylor KM, Margolese RG, Soskolne CL. Physicians’ reasons for not entering eligible patients in a randomized clinical trial of surgery for breast cancer. N Engl J Med 1984;310: 1363-7.
Background. Colorectal stents have been available as a management option in obstruction for 23 years, yet there is little randomized evidence of their effectiveness. This study investigated current surgeonrelated barriers to conducting a randomized, controlled trial (RCT) of colorectal stent insertion for obstruction in patients with colorectal cancer. Methods. A binational survey of current members of the Colorectal Surgical Society of Australia and New Zealand was conducted by a mailed questionnaire assessing perceived barriers to adoption of colonic stents and willingness to participate in future multicentre randomized controlled trials, and surgeons’ treatment preferences in 16 hypothetical clinical scenarios. Results. Of 148 eligible surgeons, 96 (65%) responded. Colonic stenting was available to 98% of respondents. In the clinical setting of colorectal obstruction, only 29% (95% CI, 20–39%) of surgeons expressed a willingness to participate in a RCT involving colonic stents in the curative setting. More than 70% of surgeons preferred the use of stents in unfit patients for palliation, and preferred surgery in fit patients with curable disease. In the curative setting, most respondents considered colonic stents not cost effective (90%; 95% CI, 82–94%) and believed that their patients would not prefer stents over surgery (80%; 95% CI, 71–87%). Conclusion. This study highlights the limitation to conducting a future randomized controlled trial to assess the efficacy of colonic stenting, especially in the curative setting, based on surgeon preference, despite the lack of level I evidence. (Surgery 2015;157:1137-41.) From the Department of Colorectal Surgery,a Royal Prince Alfred Hospital, NSW, Australia, the Surgical Outcomes Research Centre (SOuRCe),b University of Sydney, Sydney, Australia, and University of Sydney,c Sydney, Australia
LARGE BOWEL OBSTRUCTION is a surgical emergency where traditional operative management is usually associated with a high morbidity and mortality.1 Colonic stents have provided an alternative treatment option for obstruction since first described by Dohmoto in 1991.2 Despite multiple studies demonstrating the safety and efficacy of stenting to manage malignant large bowel obstruction, randomized, controlled trials (RCTs) to confirm its effectiveness are lacking in the literature.3-6 Accepted for publication January 7, 2015. Reprint requests: Christopher J. Young, MBBS, MS, FRACS, FACS, Department of Colorectal Surgery, Royal Prince Alfred Hospital Medical Centre, 100 Carillon Ave, Newtown, NSW 2042, Australia. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2015 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.surg.2015.01.022
Individual equipoise, a measure of clinical uncertainty, occurs when an individual clinician is completely undecided, whereas community equipoise applies when there are differing views among the profession as a whole.7 Preferences strongly correlate with participation in surgical RCTs,8 and when >70% of expert clinicians prefer 1 treatment modality, a human trial is considered to be unethical.7 Given colonic stents have been a treatment option in large bowel obstruction for 23 years, it is important to assess surgeons’ preference of treatment modalities in obstruction, and any biases, before further trials are conducted in this area. The aim of this study was to assess the willingness of surgeons toward conducting a future RCT with regard to the use of stenting in colorectal obstruction, and to identify the preference of surgeons toward stenting versus surgery in varying scenarios of colorectal obstruction, specifically SURGERY 1137
1138 Suen et al
considering patient’s age and fitness for a procedure, the degree of colorectal obstruction, and the difference between palliative and potentially curative treatment. METHODS All members of the Colorectal Surgical Society of Australia and New Zealand were surveyed in this study. Each surgeon was mailed a covering letter, a questionnaire and a reply-paid envelope. Two mailed reminders were used to follow up with nonresponders. Development of the questionnaire. The questionnaire was divided into 3 parts. (1) Surgeons were asked whether they would participate in an RCT or a non-RCT involving colonic stents in curative and palliative settings. (2) Surgeons were asked to rate 6 potential barriers to the use of colonic stenting, including issues of patients’ preference, risks associated with the use of colonic stents, evidence available in the literature, cost effectiveness of using colonic stents, surgeon training, and availability of the colonic stent. Respondents were asked to indicate their agreement with each statement on a scale from 1 to 5, with ‘‘1’’ being ‘‘strongly disagree’’ to ‘‘5’’ being ‘‘strongly agree.’’ (3) Surgeons were presented with 16 different clinical scenarios which were developed by varying 4 clinical factors, namely patient age, degree of obstruction, American Society of Anesthesiologists grade and curability of the colorectal cancer. These 4 clinical factors were chosen on the basis of a literature review9,10 and were considered to be the most important factors affecting decision making in this setting. For each scenario, surgeons were asked to rate their level of clinical certainty about which of 2 treatment options---colonic stent and open surgery with primary anastomosis or stoma---was their preferred option. To indicate their preference, surgeons were asked to mark an ‘X’ on the line of a visual analog scale, with the treatment options at either end of the scale. The scale was centered on zero to represent ‘‘completely undecided’’ and marked from ‘‘1’’ to ‘‘5’’ toward each end to represent increasing certainty in each treatment approach.11 Statistical analyses. Characteristics of the responding surgeons were tabulated. Descriptive statistics (proportion and 95% confidence intervals) were calculated. The proportion of surgeons who would participate in a RCT was calculated. Individual equipoise in each scenario in part 1 of the questionnaire was measured by the proportion of surgeons choosing the completely undecided
Surgery June 2015
Table I. Demographic and other characteristics of responding colorectal surgeons (n = 96) Characteristic
n
Mean age (y) Sex, n (%) Male Female Practice location* Capital city Other urban area Rural area Appointment type* Conjoint staff/academic VMO/consultant Staff specialist Salaried university academic Hospital characteristics* Tertiary referral teaching hospital District general hospital Private hospital Availability of colonic stents* Yes No Specialists performing colonic stenting in the respondent’s hospital* Colorectal surgeons only Gastroenterologists only Radiologists only Surgeons and gastroenterologists Surgeons and radiologists All 3 specialties
49
%
87 9
91 9
61 24 5
68 27 5
11 64 11 4
12 71 12 5
60 7 24
67 7 26
89 2
98 2
43 21 2 17 3 3
48 24 2 19 4 4
*Information was incomplete in these sections in 7 patients. VMO, Visiting medical officer.
category.10 To assess the community equipoise, surgeons’ responses were categorized into 1 of 3 groups: (1) preferred treatment with open surgery with stoma or primary anastomosis, (2) completely undecided (ie, indicated zero), and (3) preferred treatment with colonic stent. Based on previous studies,7,10-13 community equipoise is considered lacking in that scenario when >70% of respondents favored 1 treatment option. Analyses were carried out using SPSS statistical software. RESULTS Of 149 active members of the Colorectal Surgical Society of Australia and New Zealand, 1 was no longer resident in Australia or New Zealand and was excluded from the study. Ninety-six members (65%) completed and returned the questionnaire. Personal and professional characteristics of the responding surgeons are shown in Table I. The mean age of participants was 49 years. Colonic stenting was available as a treatment option to 98% of the respondents.
Suen et al 1139
Surgery Volume 157, Number 6
Table II. Willingness to participate in RCTs in different settings Willing to Willing to participate participate non-RCT, RCT, n (%) n (%) Palliative Curable
68 (71) 28 (29)
63 (66) 33 (34)
McNemar test v = 0.52; 1 df; P = .74 v2 = 0.83; 1 df; P = .36 2
RCT, Randomized controlled trial.
RCT participation. Few surgeons (29%; 95% CI, 20–39%) would agree to participate in a trial of colonic stent for patients with potentially curable colorectal cancer. The small increase in the proportion of surgeons willing to participate in non-RCT compared with RCT studies was not significant in either palliative or curative settings (Table II). Barriers. Only a small proportion (18%; 95% CI, 11–27%) of surgeons agreed that the use of colonic stenting is limited by unavailability or inadequate evidence of its efficacy (14%; 95% CI, 8–22%). Slightly more than one-half (52%; 95% CI, 42–62%) thought that surgeons’ proficiency was a barrier to the use of stenting (Table III). Most surgeons (80%; 95% CI, 71–87%) considered that, in the curative setting, patients would not prefer a stent over surgery and even more (90%; 95% CI, 82–94%) believed that stenting was not a costeffective option compared with surgery. A significant proportion of surgeons (68%; 95% CI, 58– 76%) also expressed concerns over the risk of stent-related perforation. Clinical scenarios. Surgeons demonstrated strong treatment preferences, with <20% of respondents in the ‘‘completely undecided’’ group across all 16 scenarios (Table IV). Community equipoise was absent in 2 main clinical situations: (1) patients with palliative colorectal obstruction who were unfit for surgery, because most surgeons would prefer to use a colonic stent, and (2) patients with potentially curable colorectal obstruction who are fit for surgery, for whom most surgeons would opt for immediate surgery. No surgeon was completely undecided about treatment options where the colorectal obstruction was partial, curable, and the patient was fit for surgery, despite their age. Community equipoise was apparent in the ‘‘less extreme’’ clinical scenarios (patients with metastatic cancer who were fit for surgery or patients with curable cancer who were unfit for surgery). The only exception was in the scenario referring to a young patient with partial, curable colorectal obstruction who was unfit for surgery.
Table III. Perceived barriers to colonic stenting in malignant colorectal obstruction Barrier
Agree (%)
Colonic stent not cost effective in curative setting Patient does not prefer colonic stent in curative setting Would not use colonic stent owing to perforation concerns Surgeon’s proficiency will affect colonic stent usage No role in curative setting Colonic stent is not a definitive solution Patient don’t want to wait in curative setting Colonic stent not cost effective in palliative setting Surgeon has more training in surgery, so less like to use stent Colonic stent complication rate too high Patient does not prefer colonic stent in palliative setting Stoma better palliation than colonic stent Colonic stent not readily available Not enough evidence in literature Not use colonic stent due to surgeons bad experience
90 80 68 52 49 47 45 42 40 29 28 25 18 14 5
DISCUSSION This study is the first to assess preferences among surgeons regarding the use of colonic stenting in the management of large bowel obstruction accompanying colorectal cancer. Despite the lack of RCTs, very few surgeons considered that evidence for the efficacy of colonic stenting was inadequate. This is probably the result of longstanding use of stenting in clinical practice and the number of nonrandomized studies in the literature. Although this is potentially viewed as intuitive, this study is the first to demonstrate this practice. Among colorectal surgeons in Australia and New Zealand, there was a general lack of enthusiasm for participating in future RCTs of colonic stents. There were no differences in the participation rate between RCT and non-RCT in both palliative and curative settings. The randomization process was not the cause of the reduction in enthusiasm. This can probably be explained by surgeons’ firmly established views of the role of colonic stenting. Despite the lack of RCT evidence, there was little individual equipoise among surgeons, which calls into question the feasibility of conducting RCTs on this topic.
1140 Suen et al
Surgery June 2015
Table IV. Surgeons’ treatment preferences in different clinical scenarios Level of clinical certainty Clinical scenarios 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
70 70 50 50 70 70 50 50 70 70 50 50 70 70 50 50
yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo; yo;
partial obstruction; metastatic cancer; ASA 4 complete obstruction; metastatic cancer; ASA partial obstruction; metastatic cancer; ASA 4 complete obstruction; metastatic cancer; ASA partial obstruction; metastatic cancer; ASA 1 complete obstruction; metastatic cancer; ASA partial obstruction; metastatic cancer; ASA 1 complete obstruction; metastatic cancer; ASA partial obstruction; curable cancer; ASA 4 complete obstruction; curable cancer; ASA 4 partial obstruction; curable cancer; ASA 4 complete obstruction; curable cancer; ASA 4 partial obstruction; curable cancer; ASA 1 complete obstruction; curable cancer; ASA 1 partial obstruction; curable cancer; ASA 1 complete obstruction; curable cancer; ASA 1
4 4 1 1
Surgery (%)
Undecided (%)
Stent (%)
Evidence of community equipoise?
8 9 15 12 51 40 60 51 66 41 73 50 96 79 96 87
12 8 10 8 19 13 17 14 15 13 10 11 4 12 4 9
80 82 75 80 30 47 23 35 19 46 17 39 0 9 0 4
No No No No Yes Yes Yes Yes Yes Yes No Yes No No No No
ASA, American Society of Anesthesiologists; yo, years old.
Self-reported participation in hypothetical RCTs has been shown to correlate well with recruitment in subsequent RCTs. For example, a preference study predicted 45% of eligible patients would be recruited into a laparoscopic colon cancer study (ALCAS) and when a log of eligible patients of the ALCAS multicenter RCT was taken, this was the exact recruitment.14,15 The same phenomenon happened within our own institution, where our RCT on colonic stenting in the palliative setting had much slower recruitment rate than we initially anticipated. This study has quite simply and adequately confirmed that surgeons’ treatment preferences for different patient populations for the same condition can be measured and this correlates with their likely participation in future RCTs. Feasibility issues rather than methodologic issues (eg, blinding, standardization of techniques) have been shown to be the major barriers to perform surgical RCTs.16 Patient preferences and lack of surgical equipoise are 2 of the top 3 feasibility barriers to surgical RCTs.16 Preference studies like this are a cost-effective way to screen out trials that will never complete. Given that research resources are limited, preference studies should be performed before establishing large surgical RCTs to avoid the often inevitable outcome of lack of recruitment in many surgical trials.17 This is affirmed with this study and the preferences in which to avoid an RCT in this area.
Most respondents had clear opinions on the role of stenting in different clinical scenarios. In patients presenting with colorectal obstruction, >70% of surgeons preferred the use of stents for palliation in the unfit patient, whereas surgery was preferred in the fit patient with curable disease. In the palliative setting, surgeons in our study believed that their patients would prefer a stent rather than surgery, and a stent was thought to be a better palliative option than a stoma, whereas in the curative setting, stents were considered to be expensive, not definitive, and having the potential to cause perforation. In 7 of 8 scenarios involving palliation of fit patients or managing curable disease in unfit patients, surgeons were more divided in their preferences for stenting versus surgery. This reflects the complexity of the decision-making process in these situations. The development of future trials in colonic stenting should focus on these clinical situations where community equipoise was apparent and the result from this study can act as a guide to formulate the inclusion criteria. Although this study achieved a good response rate for mailed surveys, it is possible that responders are not representative of all colorectal surgeons’ attitudes. Furthermore, self-reported treatment preferences may differ from actual clinical practice, although the lack of rigorous evidence to provide a standard of care for the management of malignant large bowel obstruction
Surgery Volume 157, Number 6
likely mitigates the potential for social desirability bias in survey responses. The individual surgeon proficiency with stenting was also not obtained and this may have affected the responses. In conclusion, the low interest in research participation, particularly in the curative setting, represents a major barrier to the conduct of future RCTs to strengthen the scientific evidence base for colonic stenting for people with malignant bowel obstruction. Surgeons preferred the use of stents for palliation in unfit patients presenting with obstruction secondary to colorectal cancer whereas they preferred surgery in fit patients with potentially curable disease. There was community equipoise in the context of the palliation of fit patients and treatment of potentially curable but unfit patients.
REFERENCES 1. Seah DW, Ibrahim S, Tay KH. Hartmann procedure: is it still relevant today? ANZ J Surg 2005;75:436-40. 2. Dohmoto M. New method – endoscopic implantation of rectal stent in palliative treatment of malignant stenosis. Endosc Digest 1991;3:1507-12. 3. Khot UP, Lang AW, Murali K, Parker MC. Systematic review of the efficacy and safety of colorectal stents. Br J Surg 2002; 89:1096-102. 4. Sebastian S, Johnston S, Geoghegan T, Torreggiani W, Buckley M. Pooled analysis of the efficacy and safety of self-expanding metal stenting in malignant colorectal obstruction. Am J Gastroenterol 2004;99:2051-7. 5. Watt AM, Faragher IG, Griffin TT, Rieger NA, Maddern GJ. Self-expanding metallic stents for relieving malignant colorectal obstruction. A systematic review. Ann Surg 2007;246: 24-30.
Suen et al 1141
6. Farrell JJ. Preoperative colonic stenting: how, when and why? Curr Opin Gastroenterol 2007;23:544-9. 7. Johnson N, Lilford R, Brazier W. At what level of collective equipoise does a clinical trial become ethical? J Med Ethics 1991;17:30-4. 8. Young J, Solomon M. Improving the evidence-base in surgery: evaluating surgical effectiveness. ANZ J Surg 2003; 73:507-10. 9. Tan K, Sim R. Surgery for obstructed colorectal malignancy in an Asian population: predictors of morbidity and comparison between left-and right-sided cancers. J Gastrointest Surg 2010;14:295-302. 10. Biondo S, Pares D, Frago R, Martı-Rague J, Kreisler E, De Oca J, et al. Large bowel obstruction: predictive factors for postoperative mortality. Dis Colon Rectum 2004;47: 1889-97. 11. Young J, Harrison J, White G, May J, Solomon M. Developing measures of surgeons’ equipoise to assess the feasibility of randomized controlled trials in vascular surgery. Surgery 2004;136:1070-6. 12. Freedman B. Equipoise and the ethics of clinical research. N Engl J Med 1987;317:141-5. 13. Weijer C, Shapiro SH, Cranley Grass K. Clinical equipoise and not the uncertainty principle is the moral underpinning of the randomised controlled trial. BMJ 2000;321: 756-7. 14. Solomon MJ, Pager CK, Keshava A, Findlay M, Butow P, Salkeld GP, et al. What do patients want? Patient preferences and surrogate decision making in the treatment of colorectal cancer. Dis Colon Rectum 2002;46:1351-7. 15. Abraham NS, Hewett P, Young JM, Solomon MJ. Non-entry of eligible patients into the Australasian Laparoscopic Colon Cancer Study. ANZ J Surg 2006;76:825-9. 16. Solomon MJ, McLeod RS. Should we be performing more randomized controlled trials evaluating surgical operations? Surgery 1995 Sep;118:459-67. 17. Taylor KM, Margolese RG, Soskolne CL. Physicians’ reasons for not entering eligible patients in a randomized clinical trial of surgery for breast cancer. N Engl J Med 1984;310: 1363-7.