Imposter Syndrome Among Surgeons Is Associated With Intolerance of Uncertainty and Lower Confidence in Problem Solving

Abstract

Background

Feelings of imposter syndrome (inadequacy or incompetence) are common among physicians and are associated with diminished joy in practice. Identification of modifiable factors associated with feelings of imposter syndrome might inform strategies to ameliorate them. To this point, though, no such factors have been identified.

Question/purpose

Are intolerance of uncertainty and confidence in problem-solving skills independently associated with feelings of imposter syndrome after accounting for other factors?

Methods

This survey-based experiment measured the relationship between feelings of imposter syndrome, intolerance of uncertainty, and confidence in problem-solving skills among musculoskeletal specialist surgeons. Approximately 200 surgeons who actively participate in the Science of Variation Group, a collaboration of mainly orthopaedic surgeons specializing in upper extremity illnesses primarily across Europe and North America, were invited to this survey-based experiment. One hundred two surgeons completed questionnaires measuring feelings of imposter syndrome (an adaptation of the Clance Imposter Phenomenon Scale), tolerance of uncertainty (the Intolerance of Uncertainty Scale-12), and confidence in problem-solving skills (the Personal Optimism and Self-Efficacy Optimism questionnaire), as well as basic demographics. The participants were characteristic of other Science of Variation Group experiments: the mean age was 52 ± 5 years, with 89% (91 of 102) being men, most self-reported White race (81% [83 of 102]), largely subspecializing in hand and/or wrist surgery (73% [74 of 102]), and with just over half of the group (54% [55 of 102]) having greater than 11 years of experience. We sought to identify factors associated with greater feelings of imposter syndrome in a multivariable statistical model.

Results

Accounting for potential confounding factors such as years of experience or supervision of trainees in the multivariable linear regression analysis, greater feelings of imposter syndrome were modestly associated with higher intolerance of uncertainty (regression coefficient [β] 0.34 [95% confidence interval (CI) 0.16 to 0.51]; p < 0.01) and with lower confidence in problem-solving skills (β -0.70 [95% CI -1.0 to -0.35]; p < 0.01).

Conclusion

The finding that feelings of imposter syndrome may be modestly to notably associated with modifiable factors, such as difficulty managing uncertainty and lack of confidence in problem-solving, spark coaching opportunities to support and sustain a surgeon’s mindset, which may lead to increased comfort and joy at work.

Clinical Relevance

Beginning with premedical coursework and throughout medical training and continuing medical education, future studies can address the impact of learning and practicing tactics that increase comfort with uncertainty and greater confidence in problem solving on limiting feelings of imposter syndrome.

Introduction

Imposter syndrome was first conceived of in the 1970s as an “impostor phenomenon” among high achievers who are unable to internalize and accept success [9]. This results in feelings of anxiety because of believed inadequacy, incompetence, or fraudulence [9]. In a scoping review by Gottlieb et al. [17], the prevalence of threshold feelings of imposter syndrome among medical students ranged from 22% to 60%, with variation based on the type of measure used (the Clance Imposter Phenomenon Scale [8, 10] or the Harvey Impostor Scale [19, 41]), on the threshold used (on a continuum or categorically positive or negative), and based on the country in which the study was performed; the proportion among residents was 33% to 44%. This is important because survey studies among physicians suggest associations between feelings of imposter syndrome and worse mental health, such as lower self-esteem and greater feelings of emotional exhaustion and burnout [20, 23-25].

The ability to navigate and manage uncertainty and the level of confidence in problem solving are important skills for any surgeon in daily practice and are essential to cultivate to ensure a sense of belonging in the field [12, 28, 38]. Given that these are learnable skills and therefore potentially modifiable factors, we sought to understand their relationship with feelings of imposter syndrome among attending orthopaedic surgeons, since, to our knowledge, this has not been evaluated before.

We performed a cross-sectional survey of an international group of surgeons and asked: Are intolerance of uncertainty and confidence in problem-solving skills independently associated with feelings of imposter syndrome after accounting for other factors?

Materials and Methods

Study Design and Participants

We created a cross-sectional, survey-based experiment using an online Health Insurance Portability and Accounting Act–compliant survey tool, SurveyMonkey. The participants were hand and upper extremity surgeon members of the Science of Variation Group, which is an international group of surgeons who collaborate to study variations in diagnosis and treatment with no financial incentives. Approximately 200 upper extremity members actively participate in at least one survey-based experiment annually. These participants include mostly academic orthopaedic surgeons in the United States and Europe at different stages of practice experience. For this study, four email invitations were sent between May 1, 2021, and June 1, 2021. All participants completed the Clance Imposter Phenomenon Scale, the Intolerance of Uncertainty Scale 12-item short form, the Personal Optimism and Self-Efficacy Optimism Scales, and a survey of basic demographics (such as age, gender, race, marital status, and geographic location of practice), in randomized order (simple randomization).

A total of 102 members agreed to participate, about half the number of annual participants, which is typical of most upper extremity Science of Variation Group studies [22, 27, 36, 44]. Because we were interested in associations and not absolute rates, diversity among the sample was more important than the participation proportion. The rates are specific to the sample, but the relationships are likely reproducible in other samples and thus generalizable to other populations of academic surgeons.

Participants’ Baseline Data

The responding participants were characteristic of other Science of Variation Group experiments including age, gender, race, and surgical subspecialty (Table 1). The mean age was 52 ± 5 years, and 89% (91 of 102) of participants were men. Participants self-selected race using the survey instrument from the following categories: African American/Black, Latino/a, Asian/Pacific Islander, White, and Other; 81% (83 of 102) of the participants identified as White. Most subspecialize in hand and/or wrist surgery (73% [74 of 102]). The participants ranged in years of experience, with 54% (55 of 102) having greater than 11 years of experience.

Table 1.

Surgeon characteristics

Variable	Value (n = 102)
Age in years	52 ± 5
Women	11 (11)
Racea
White	81 (83)
Other	19 (19)
Location
United States	64 (65)
Europe	23 (23)
Other	14 (14)
Years of experience
0 to 5	21 (21)
6 to 10	25 (26)
11 to 20	27 (28)
21 or more	26 (27)
Subspecialty
Hand and/or wrist	73 (74)
Shoulder and elbow	23 (23)
Other	5 (5)
Supervising trainees	82 (84)
Married	90 (92)
First-generation physician	69 (70)
Clance Imposter Syndrome (10-item) score	26 ± 7.0
Intolerance of uncertainty scale (12-item) score	27 ± 7.0
Personal optimism and self-efficacy optimism scale
Total	30 (28-32)
Personal optimism	13 (12-15)
Self-efficacy optimism	16 (15-19)

Data presented as mean ± SD, % (n), or median (IQR).

^aRace was self-selected using the survey instrument from the following categories: African American/Black, Latino/a, Asian/Pacific Islander, White, and Other. This variable is dichotomized because of the low prevalence of other selections by survey respondents.

Measurements and Outcome Variables

The primary outcome measure was feelings of imposter syndrome on a continuous scale, measured using an adapted 10-item version of the Clance Imposter Phenomenon Scale [8]. All items are rated on a 5-point Likert scale from 1 (“not at all”) to 5 (“very true”). Scores range from 10 to 50, with higher scores indicating greater feelings of imposter syndrome. Continuous scores can be converted into the categories of mild (20 points or lower), moderate (21 to 30 points), and severe (more than 30 points). Our explanatory variables were the Intolerance of Uncertainty-12 and Personal Optimism and Self-Efficacy Optimism Scale [6, 15]. The Intolerance of Uncertainty-12 is a 12-item measure of discomfort with uncertainty with statements such as “It frustrates me not having all the information I need.” Items are rated on a 5-point Likert scale, ranging from 1 (“not at all characteristic”) to 5 (“entirely characteristic”). Scores range between 12 and 60, with higher scores reflecting greater intolerance of uncertainty. The Personal Optimism and Self-Efficacy Optimism Scale is a nine-item measure of confidence in problem-solving ability. Items are scored on 4-point Likert scales from 0 (“completely incorrect”) to 3 (“completely correct”). Scores range between 0 to 36, with higher scores reflecting greater confidence in problem-solving skills.

Ethical Approval

This survey-based experiment was reviewed and approved by our institutional review board and deemed to be exempt from ongoing review.

Statistical Analysis

We performed descriptive statistics for all participants utilizing Stata 14.0 (StataCorp LLC). Categorical variables are reported as a percentage with numerator and denominator in parentheses, and continuous variables are reported as the mean ± SD or median (IQR). We performed bivariate analyses seeking to identify factors associated with the adapted Clance Imposter Phenomenon Scale score as a continuous score (Table 2). We used t-tests and one-way ANOVA for categorical variables, where appropriate; the Pearson correlations or the Spearman rank-order correlations were used for continuous variables, depending on the distribution. In addition, we sought to identify factors associated with the categorical severity of imposter syndrome. Chi-square tests were used for categorical explanatory variables, and ANOVA and Kruskal-Wallis H tests were used for continuous variables, where appropriate. All variables with p values below 0.10 were moved to a multivariable analysis. Linear regression was used for the continuous score, and multinomial logistic regression was used for the categorical severity, using mild imposter syndrome as a reference category. Regression coefficients (β), ORs, standard errors, 95% confidence intervals (CIs), partial and adjusted r-squared values, and p values are reported. Alpha was set at 0.05.

Table 2.

Bivariate analysis of factors associated with the Clance Imposter Syndrome Assessment score^a

Categorical variables	Mean ± SD	p valueb
Gender		0.53
Men	26 ± 7
Women	27 ± 6
Race		0.76
White	26 ± 7
Other	27 ± 8
Location		0.92
United States	26 ± 7
Europe	26 ± 7
Other	26 ± 7
Years of experience		0.21
0 to 5	28 ± 7
6 to 10	27 ± 7
11 to 20	24 ± 7
21 or more	25 ± 7
Subspecialty		0.62
Hand and/or wrist	26 ± 7
Shoulder and elbow	27 ± 7
Other	26 ± 6
Supervising trainees		0.06
Yes	27 ± 7
No	23 ± 6
Marital status		0.03
Married	26 ± 7
Other	31 ± 8
First-generation physician		0.76
Yes	26 ± 7
No	26 ± 8

Continuous variables	Correlation	p valueb
Age (r)	-0.13	0.19
Intolerance of Uncertainty Scale (r)	0.54	< 0.01
Personal Optimism and Self-Efficacy Optimism Scale (ρ)
Total	-0.58	< 0.01
Personal optimism	-0.52	< 0.01
Self-efficacy optimism	-0.37	< 0.01

^aThe adapted 10-item version of the Clance Imposter Phenomenon Scale is rated on a 5-point Likert scale from 1 to 5. Higher scores indicate greater feelings of imposter syndrome. The Intolerance of Uncertainty-12 has 12 items pertaining to discomfort with uncertainty. Items are rated on a 5-point Likert scale, ranging from 1 to 5. The higher the score the greater intolerance of uncertainty. The Personal Optimism and Self-Efficacy Optimism Scale is a nine-item measure reflecting the survey respondent’s confidence in problem-solving ability. Items are scored on 4-point Likert scales from 0 (“completely incorrect”) to 3 (“completely correct”), with higher scores reflecting greater confidence in problem-solving skills.

^bAll variables with p < 0.10 were moved to a multivariable linear regression analysis.

An a priori sample size calculation determined that a sample of 95 surgeons would generate 80% statistical power, with alpha set at 0.05, for a linear regression with 11 explanatory variables, given that intolerance of uncertainty or problem-solving accounts for 15% of the variation in the imposter syndrome (continuous score) and because the full model accounts for 25% of the variation or more.

Results

Intolerance of Uncertainty and Imposter Syndrome

Accounting for potential confounders such as years of experience and supervision of trainees in the multivariable linear regression analysis, greater feelings of imposter syndrome were modestly associated with greater intolerance of uncertainty (β 0.34 [95% CI 0.16 to 0.51]; p < 0.01) and lower confidence in problem-solving skills (β -0.70 [95% CI -1.0 to -0.35]; p < 0.01) (Table 3). As a sensitivity analysis, we also performed bivariate analysis of categories of levels of imposter syndrome (Supplementary Table 1; http://links.lww.com/CORR/A926). We then performed multinomial logistic regression analysis of categories of levels of imposter syndrome and found similar relationships (Supplementary Table 2; http://links.lww.com/CORR/A927).

Table 3.

Multivariable linear regression analysis of factors associated with the Clance Imposter Syndrome Assessment score^a

Variables	Regression coefficient (95% CI)	Partial r2b	p valuea
Intolerance of Uncertainty Scale	0.34 (0.16 to 0.51)	0.13	< 0.01
Personal Optimism and Self-Efficacy Optimism Scale (total)c	-0.70 (-1.0 to -0.35)	0.14	< 0.01
Supervising trainees
Yes	Reference value
No	0.84 (-2.0 to 3.7)	0.003	0.57
Marital status
Married	Reference value
Other	-3.2 (-6.9 to 0.40)	0.03	0.08

^aThe adjusted r² represents the goodness-of-fit for the model as a whole after adjusting for explanatory variables; in this model, the adjusted r² was 0.40. This suggests that all the terms in this model, taken together, account for about 40% of the variance.

^bThe partial r² represents the proportion of variance explained by each independent variable, such that a partial r² of 0.13 suggests that factor would account for 13% of the variance in the sample.

^cThe personal optimism and self-efficacy optimism subscales were dropped from the final model because of collinearity with the total score.

Discussion

Imposter syndrome is an increasingly recognized barrier to physician enjoyment of daily practice [20, 24]. Amelioration of feelings of imposter syndrome might be aided by identification of associated modifiable psychological factors. We chose two candidate factors, confidence in problem solving and tolerance of uncertainty, because they seem related to imposter syndrome and are modifiable. Among a group of academic musculoskeletal surgeons, we found that intolerance of uncertainty and confidence in problem-solving skills have modest-to-strong independent associations with feelings of imposter syndrome. These findings suggest that strategies to mitigate discomfort with uncertainty and increase confidence in problem solving have the potential to help surgeons ameliorate feelings of imposter syndrome and bolster joy in practice.

Limitations

This study has several limitations. First, the Science of Variation Group is limited in its generalizability, owing to its homogeneity in race, gender, and type of practice. The Science of Variation Group is composed primarily of White men who are academic surgeons primarily in Europe and the United States. We desire greater representation of women, other backgrounds, and other continents; those who wish to participate may do so at https://www.surveymonkey.com/r/SOVGsignup. The sample had sufficient variation in intolerance of uncertainty and confidence in problem-solving skills to make those analyses meaningful. However, a more diverse study population is required to confirm our finding of no relationship between gender or race and feelings of imposter syndrome. Race is a social construct that is most representative of social experiences. Our limited number of non-White participants resulted in the need to pool non-White races; this dilutes potentially important experiences documented in studies in other contexts. The focus of our study on uncertainty and confidence may not capture important aspects of imposter syndrome such as feelings of exclusion in cultures with systematic or institutional barriers and how to address and change these systemic issues or unhelpful scrutiny that may affect professional development or career trajectory over a longitudinal period of time. Along the same lines, our study only examined practicing surgeons, and we were not able to study how training (such as residency) affects tolerance of uncertainty and confidence in problem-solving skills. The number of years in practice may serve as a helpful proxy, but more study is merited to assess how different periods of training throughout practice affects these factors.

Second, the modest-to-strong associations observed could be due to common method variance, which is when correlations occur because of similarities in the measurement methods rather than the content of the tools. It is possible that we measured willingness to disclose feelings of uncertainty, low confidence, and imposter syndrome on multiple Likert scales more so than a true correlation between these factors. On the other hand, our sense is that even a “willingness to disclose” factor can lead us to helpful interventions on the basis that it is possible that attention to any one of these factors is likely to benefit the others. Third, intolerance of uncertainty and low confidence had modest associations with feelings of imposter syndrome, suggesting there are other factors to consider. This study elucidates this potentially important association that we hope will inspire others to consider investigation of methods for limiting feelings of imposter syndrome. Fourth, it should be noted that identifying an association is distant from developing effective interventions. For instance, a mindfulness intervention did not ameliorate symptoms of burnout in pediatric trainees [14]. There may be several steps to identifying effective strategies for limiting imposter syndrome.

Discussion of Key Findings

The observed association between intolerance of uncertainty in general and uncertainty about oneself (feelings of imposter syndrome) seems consistent with other lines of evidence [7, 16, 30, 31]. For instance, feelings of imposter syndrome were associated with feelings of low self-esteem in several studies [8, 13, 39]. The internal misrepresentation of self, which is characteristic of feelings of imposter syndrome and limited confidence in problem solving, may be rooted in a lack of acceptance of one’s limitations as well as the lack of acceptance that aspects of life and health are often more probabilistic than deterministic.

Intolerance of uncertainty is also associated with symptoms of depression and anxiety and diminished job satisfaction, which could be considered related to feelings of imposter syndrome. A study of Australian general practitioner trainees demonstrated that physician intolerance of uncertainty, anxiety about clinical uncertainty, and reluctance to disclose that uncertainty to patients were associated with greater feelings of burnout [11]. A study of Swiss clinicians found that anxiety related to generalized uncertainty was associated with lower rates of work satisfaction [5]. A study assessing resident physicians in the United States found that both higher levels of stress related to uncertainty and low levels of resilience among resident physicians were associated with greater symptoms of burnout and depression [37].

A systematic review identified clinician attitudes toward risk and uncertainty as a source of between-clinician variability and suggested interventions to improve comfort with uncertainty as targets for reducing unwarranted variations in care [43]. In addition, patient and clinician comfort with unavoidable levels of uncertainty may enhance health through improved patient accommodation of pathophysiology and impairment [46]. Interventions for reducing intolerance of uncertainty alleviated maladaptive behaviors and emotional symptoms associated with conditions such as generalized anxiety disorder and multiple sclerosis [1, 4, 21, 40]. The interventions were based on reduction of cognitive biases and the principles of cognitive behavioral therapy: mental exercises that teach people to separate thought from fact and cultivate the healthiest possible inner narrative [3, 26, 29, 35, 40]. The finding that modifying cognitive biases can reduce intolerance of uncertainty in people with anxiety and mood disorders suggests a line of research where interventions to reduce intolerance of uncertainty and lack of confidence in problem solving can be tested to determine whether such interventions reduce feelings of imposter syndrome among physicians and trainees. There is evidence that orthopaedic surgeons can learn to recognize intolerance of uncertainty in patients and their physician colleagues [45, 47]. Patients and clinicians can discuss areas of uncertainty and limitations in the ability to resolve uncertainty.

Feelings of imposter syndrome may be a form of cognitive bias arising from mental shortcuts (heuristics) that can be altered by mindset training focused on increased comfort with uncertainty and increased confidence in problem solving. A randomized controlled study examining the effect of a didactic-based mindfulness program on burnout in pediatric residents found that the program did not reduce burnout [14]. Programs and interventions based on peer feedback and improvement of both technical and emotional-intelligence skills, however, may be better suited to tackle aspects contributing to imposter syndrome and reduce feelings of imposter syndrome [18]. Surgical coaching programs such as the Wisconsin Surgical Coaching Program and the Harvard Surgical Coaching for Operative Performance Enhancement might also help ameliorate feelings of imposter syndrome in part by increasing comfort with uncertainty and confidence in problem solving [32-34]. Both programs, although focused on intraoperative skills, also coach nontechnical skills including cognitive and interpersonal skills. They promote a culture of support and mentorship among peers as well as an internal growth mindset in an individual. Future studies might assess change in feelings of imposter syndrome after entering a coaching program [2].

Conclusion

Our finding that greater feelings of imposter syndrome were modestly to strongly associated with greater intolerance of uncertainty and lower confidence with problem solving helps confirm that feelings of imposter syndrome may relate as much to one’s inner narrative as to one’s circumstances; both merit additional attention. In our opinion, both cognitive behavioral therapy approaches for mindset training (exercises for an optimal inner narrative) combined with peer surgical coaching programs that foster both a growth mindset as well as a culture of inclusion, belonging, and mutual support merit study for their ability to help limit surgeons’ feelings of imposter syndrome and improve joy in practice. Surgeons experiencing feelings of imposter syndrome can consider the degree to which intolerance of uncertainty and limited confidence in problem solving are contributing to such feelings and make specific efforts to ameliorate them.

Group Authors

The Science of Variation Group includes: Miguel Pirela Cruz, Gerald A. Kraan, Philipp Muhl, Richard S. Gilbert, Todd Siff, Anne Spaans, Adam Shafritz, Julie Adams, Anne J. H. Vochteloo, Andreas Platz, Andrew L. Terrono, Todd Bafus, H. Brent Bamberger, Ben Sutker, Bernard F. Hearon, Jonathan Braman, Ryan P. Calfee, Carl Ekholm, Carlos H. Fernandes, Charles Metzger, Chris Bainbridge, Constanza L. Moreno-Serrano, Duffield Ashmead IV, Darren Drosdowech, Desirae M. McKee, Daniel Falcon, Dan Polatsch, David P. Patterson, Camilo Jose Romero Barreto, Mohamed Shafi, Juan M. Patiño, Roger van Riet, Eric Raven, Ellen Satteson, Erik T. Walbeehm, Evan D. Schumer, Ezequiel E. Zaidenberg, Fabio Suarez, Fred O'Brien, Frederik Verstreken, George Pianka, Grant Bayne, Guido Fierro, Thierry G. Guitton, Michael Nancollas, Jeffrey A. Greenberg, Greg P. Watchmaker, Lewis B. Lane, Eric P. Hofmeister, Jack E. Kazanjian, Jacob W. Brubacher, Jacob Gire, Jason D. Tavakolian, James F. Nappi, John M. Erickson, John Taras, Julie Balch Samora, Sanjeev Kakar, Ken Butters, Kendrick E. Lee, Kevin Rumball, Lawrence Weiss, Lars Adolfsson, C. Liam Dwyer, Luis F. Naquira Escobar, Marco van der Pluijm, Marc J. Richard, Maurizio Calcagni, John A. McAuliffe, Michell Ruiz-Suarez, Michael J. Palmer, Michael W. Grafe, Ngozi M. Akabudike, Nathan A. Hoekzema, Martin Richardson, Jose A. Ortiz Jr, Jeff W. Johnson, Prosper Benhaim, Philip Blazar, Peter Jebson, Patrick W. Owens, Ralf Walbeehm, Ramon de Bedout, Russell Shatford, Sergio Rowinski, Richard Wallensten, Craig Rodner, Vani J. Sabesan, Stephen A. Kennedy, Betsy M. Nolan, Steve Kronlage, Sebastiaan Souer, Steven L. Henry, Taco Gosens, Taizoon Baxamusa, F. Thomas D. Kaplan, Thomas Apard, W. Jaap Willems, Warren C. Hammert, Mohammad Waseem.