Abstract
A randomized survey of 1257 respondents was used to assess willingness to undergo antibiotic treatment of appendicitis with different quoted risks of treatment failure requiring appendicectomy. Overall, 1045 respondents (83.1 per cent) were willing to try antibiotics. Even at a quoted 60 per cent risk of failure, 75 per cent of respondents were willing to attempt antibiotic treatment.
Dear Editor,
The COVID-19 pandemic has led to increased adoption of non-operative management strategies for acute appendicitis1. Despite mounting evidence for their efficacy and safety2, surgeons may still hesitate to recommend antibiotics owing to concerns about high treatment failure risk and eventual appendicectomy (up to 30 per cent at 1 year)3,4. It is unknown how that risk, and the uncertainty around it, influences patients’ appendicitis decision-making.
A survey was undertaken of American adults recruited via Amazon Mechanical Turk in April 2021. The survey described antibiotics as non-operative appendicitis treatment and the probability of treatment failure within 3 months (need for appendicectomy), framed both negatively (chance of needing surgery) and positively (chance of avoiding surgery). Respondents were randomized to 1 of 14 arms varying in how treatment failure risk was described: seven arms reported risks of 10, 15, 20, 30, 40, 50, and 60 per cent; seven analogous arms reported these point estimates with the addition of a range (+/− 4 per cent). The primary outcome was willingness to try antibiotic treatment and the secondary outcome was perceived accuracy and trust in the information provided, measured on a five-point scale5. American Association for Public Reporting of Opinion Research reporting guidelines were followed.
After quality checks (87 of 1429 removed) and exclusion of those who had appendicitis previously (85 of 1342 removed), the sample consisted of 1257 adults. Sociodemographic characteristics were balanced across survey arms. Few (115 of 1257, 9.1 per cent ) were aware that antibiotics could be used to treat appendicitis before taking the survey. Most respondents (1045 of 1257, 83.1 per cent) were willing to try antibiotic treatment, with higher risks of treatment failure resulting in moderately lower willingness (Table 1). Among those who would try antibiotics, over half (599 of 1045, 57.3 per cent) were willing regardless of the risk of treatment failure, whereas, on average, the remaining respondents were willing to try until the treatment failure risk reached a mean(s.d.) of 53.6(23.0) per cent. Male sex and gender identity, increased perceived accuracy of information, and increased trust in the data were associated with willingness to try antibiotics.
Table 1.
Factors associated with willingness to try non-operative management of appendicitis
| Overall | Not willing to try non-operative management | Willing to try non-operative management | P | |
|---|---|---|---|---|
| (n = 1257) | (n = 212) | (n = 1045) | ||
| Risk of treatment failure (%) | 0.001 | |||
| 10 | 92 | 10 (11) | 82 (89) | |
| 15 | 85 | 11 (13) | 74 (87) | |
| 20 | 86 | 17 (10) | 69 (80) | |
| 30 | 95 | 16 (17) | 79 (83) | |
| 40 | 93 | 20 (22) | 73 (78) | |
| 50 | 88 | 18 (20) | 70 (80) | |
| 60 | 92 | 23 (25) | 69 (75) | |
| 10 (6–14†) | 89 | 7 (8) | 82 (92) | |
| 15 (11–19†) | 94 | 10 (11) | 84 (89) | |
| 20 (16–24†) | 88 | 7 (8) | 81 (92) | |
| 30 (26–34†) | 85 | 11 (13) | 74 (87) | |
| 40 (36–44†) | 93 | 18 (19) | 75 (81) | |
| 50 (46–54†) | 87 | 19 (22) | 68 (78) | |
| 60 (56–64†) | 90 | 25 (28) | 65 (72) | |
| Age (years)* | 37.37(12.38) | 37.46(12.53) | 37.36(12.35) | 0.910 |
| Sex | 0.001 | |||
| F | 752 | 150 (19.9) | 602 (80.1) | |
| M | 501 | 62 (12.4) | 439 (87.6) | |
| Prefer not to say | 4 | 0 (0) | 4 (100) | |
| Gender identity | 0.028 | |||
| Woman | 735 | 147 (20.0) | 588 (80.0) | |
| Man | 498 | 63 (12.7) | 435 (87.3) | |
| Genderqueer/gender non-conforming | 11 | 2 (18) | 9 (82) | |
| Trans male/trans man | 5 | 0 (0) | 5 (100) | |
| Prefer not to say | 4 | 0 (0) | 4 (100) | |
| Trans female/trans woman | 2 | 0 (0) | 2 (100) | |
| Different identity | 2 | 0 (0) | 2 (100) | |
| Racial identity | 0.748 | |||
| Black | 125 | 24 (19.2) | 101 (80.8) | |
| East Asian | 75 | 11 (15) | 64 (85) | |
| Multiple identities | 56 | 12 (21) | 44 (79) | |
| Other Specified Identity | 29 | 7 (24) | 22 (76) | |
| South Asian | 43 | 7 (16) | 36 (84) | |
| Unknown | 12 | 1 (8) | 11 (92) | |
| White | 917 | 150 (16.4) | 767 (83.6) | |
| Ethnicity | 0.212 | |||
| Hispanic/Latino/Latinx | 105 | 24 (22.9) | 81 (77.1) | |
| Non-Hispanic/Latino/Latinx | 1095 | 181 (16.5) | 914 (83.5) | |
| Prefer not to say | 27 | 2 (7) | 25 (93) | |
| Prefer to write it down | 30 | 5 (17) | 25 (83) | |
| Education level | 0.369 | |||
| Some high school | 4 | 1 (25) | 3 (75) | |
| High school/GED | 112 | 14 (12.5) | 98 (87.5) | |
| Some college | 268 | 55 (20.5) | 213 (79.5) | |
| 2-year college degree | 134 | 24 (17.9) | 110 (82.1) | |
| 4-year college degree | 499 | 76 (15.2) | 423 (84.8) | |
| Graduate degree | 235 | 42 (17.9) | 193 (82.1) | |
| Unknown | 5 | 0 (0) | 5 (100) | |
| Insurance | 0.726 | |||
| Employer-provided | 649 | 111 (17.1) | 538 (82.9) | |
| Private | 179 | 36 (20.1) | 143 (79.9) | |
| Other government | 48 | 6 (12) | 42 (88) | |
| Medicaid | 131 | 21 (16.0) | 110 (84.0) | |
| Medicare | 96 | 17 (18) | 79 (82) | |
| Not insured | 124 | 18 (14.5) | 106 (85.5) | |
| Other | 30 | 3 (10) | 27 (90) | |
| Employment status | 0.881 | |||
| Employed full-time (≥ 40 h/week) | 661 | 107 (16.2) | 554 (83.8) | |
| Employed part-time (< 40 h/week) | 144 | 21 (14.6) | 123 (85.4) | |
| Self-employed | 124 | 22 (17.7) | 102 (82.3) | |
| Retired | 59 | 12 (20) | 47 (80) | |
| Student | 80 | 17 (21) | 63 (79) | |
| Unemployed (looking for work) | 97 | 17 (17) | 80 (83) | |
| Unemployed (not looking for work) | 80 | 13 (16) | 67 (84) | |
| Prefer not to say | 12 | 3 (25) | 9 (75) | |
| Annual household income (euros) | 0.446 | |||
| < 21 076 | 176 | 27 (15.3) | 149 (84.7) | |
| 21 077–42 150 | 326 | 60 (18.4) | 266 (81.6) | |
| 42 151–63 224 | 248 | 45 (18.1) | 203 (81.9) | |
| 63 225–84 299 | 203 | 39 (19.2) | 164 (80.8) | |
| >84 299 | 262 | 34 (13.0) | 228 (87.0) | |
| Prefer not to say | 42 | 7 (17) | 35 (83) | |
| Perceived information accuracy*‡ | 3.40(0.98) | 3.01(1.10) | 3.48(0.94) | < 0.001 |
| Trust in information given*§ | 3.42(1.00) | 2.87(1.09) | 3.53(0.94) | < 0.001 |
Values in parentheses are percentages unless indicated otherwise; *values are mean(s.d.). †Range. ‡Scale from 1 to 5, with 5 being highest perceived accuracy. §Scale from 1 to 5, with 5 being greatest trust. GED = General Educational Development Test, t-test for continuous variables and chi-squared test for categorical variables test.
The proportion of individuals willing to try antibiotics was generally higher when ranges were provided alongside the point estimate for treatment failure risks (Table 1). Perceived accuracy (mean(s.d.) score 3.3(1.0) versus 3.5(1.0); P < 0.001) and trust in the information (mean score 3.3(1.0) versus 3.5(1.0); P = 0.001) was also greater in arms including ranges.
Increasing evidence of the efficacy and safety of non-operative treatment for uncomplicated acute appendicitis has led to the recognition that this treatment decision is value- and preference- dependent, and must be made jointly with patients3. Although surgeons commonly report a threshold for risk of treatment failure that makes it too high to be worth trying4, this study found that over 70 per cent of patients were willing to try antibiotics for even a 40 per cent chance of avoiding surgery. This suggests a disconnect between the ways clinicians and patients conceptualize risk and benefit. An additional observation is that surgeons may affect patients’ perceived accuracy of and trust in treatment information by providing uncertainty information. Providing ranges increased trust and accuracy perceptions, which were in turn associated with greater willingness to try antibiotics.
Study limitations include use of a sample that may differ from patients experiencing appendicitis, and a focus on one aspect of appendicitis treatment, which the authors felt was the most likely to have a misalignment between patient and surgeon values. Nonetheless, these findings provide what is to the authors’ knowledge the first evidence to date about a clinically salient dynamic—individuals’ willingness to try treatments that surgeons may consider too high risk. This is a critical area for future work and strong shared decision-making between surgeons and patients.
Acknowledgements
This study was not preregistered in an independent institutional registry. J.M.L. reports personal fees from Kaiser Permanente Washington Health Research Institute, and honoraria from Wolters Kluwer, the journal Clinical Pathways, and the American College of Physicians, all outside of this submitted work .
Funding
J.E.R. is supported by a training grant from the National Institute of Diabetes and Digestive and Kidney Diseases of the National Institutes of Health (T32DK070555). This work was funded in part by a generous gift from Marty and Linda Ellison.
Disclosure. The authors declare no conflict of interest.
Contributor Information
J E Rosen, Surgical Outcomes Research Center, Department of Surgery, University of Washington, Seattle, Washington, USA; Decision Science Group, Seattle, Washington, USA.
N Agrawal, Decision Science Group, Seattle, Washington, USA; Foster School of Business, University of Washington, Seattle, Washington, USA.
D R Flum, Surgical Outcomes Research Center, Department of Surgery, University of Washington, Seattle, Washington, USA; Decision Science Group, Seattle, Washington, USA.
J M Liao, Decision Science Group, Seattle, Washington, USA; Department of Medicine, University of Washington, Seattle, Washington, USA.
References
- 1. Ielpo B, Podda M, Pellino G, Pata F, Caruso R, Gravante G et al. Global attitudes in the management of acute appendicitis during COVID‐19 pandemic: ACIE Appy Study. Br J Surg 2020;108:717–726. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.CODA Collaborative, Flum DR, Davidson GH, Monsell SE, Shapiro NI, Odom SR et al. A randomized trial comparing antibiotics with appendectomy for appendicitis. New Engl J Med 2020;383:1907–1919. [DOI] [PubMed] [Google Scholar]
- 3. Sallinen V, Akl EA, You JJ, Agarwal A, Shoucair S, Vandvik PO et al. Meta‐analysis of antibiotics versus appendicectomy for non‐perforated acute appendicitis. Br J Surg 2016;103:656–667. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4. O’Leary DP, Walsh SM, Bolger J, Baban C, Humphreys H, O’Grady S et al. A randomised clinical trial evaluating the efficacy and quality of life of antibiotic only treatment of acute uncomplicated appendicitis: results of the COMMA trial. Ann Surg 2021;274:240–247. [DOI] [PubMed] [Google Scholar]
- 5. Han PKJ, Klein WMP, Lehman T, Killam B, Massett H, Freedman AN. Communication of uncertainty regarding individualized cancer risk estimates. Med Decis Making 2011;31:354–366. [DOI] [PMC free article] [PubMed] [Google Scholar]