Comparative Perspectives on Diagnostic Error Discussions Between Inpatient and Outpatient Pediatric Providers

Abstract

Diagnostic error remains understudied and underaddressed despite causing significant morbidity and mortality. One barrier to addressing this issue remains provider discomfort. Survey studies have shown significantly more discomfort among providers in discussing diagnostic error compared with other forms of error. Whether the comfort in discussing diagnostic error differs depending on practice setting has not been previously studied. The objective of this study was to assess differences in provider willingness to discuss diagnostic error in the inpatient versus outpatient setting. A multicenter survey was sent out to 3881 providers between May and June 2018. This survey was designed to assess comfort level of discussing diagnostic error and looking at barriers to discussing diagnostic error. Forty-three percent versus 22% of inpatient versus outpatient providers (P = 0.004) were comfortable discussing short-term diagnostic error publicly. Similarly, 76% versus 60% of inpatient versus outpatient providers (P = 0.010) were comfortable discussing short-term diagnostic error privately. A higher percentage of inpatient (64%) compared with outpatient providers (46%) (P = 0.043) were comfortable discussing long-term diagnostic error privately. Forty percent versus 24% of inpatient versus outpatient providers (P = 0.018) were comfortable discussing long-term error publicly. No difference in barriers cited depending on practice setting. Inpatient providers are more comfortable discussing diagnostic error than their outpatient counterparts. More study is needed to determine the etiology of this discrepancy and to develop strategies to increase outpatient provider comfort.

Introduction

Patients experience negative health outcomes resulting from diagnostic errors, an area of patient safety that remains understudied.^1,2 Despite being a well-identified patient safety concern, most major health organizations lack efforts to address diagnostic errors.³ One study of 61 hospitals showed that awareness of diagnostic error was high; however, only 25% cited it as an institutional priority.⁴ Prior investigations of malpractice suits and autopsies show that diagnostic error causes significant morbidity and mortality and is implicated in 17% of hospital adverse events and 10% of patient deaths.^2,5 Researchers studying malpractice claims found that diagnostic error accounts for 46% of payouts in outpatient settings compared with 21% in inpatient settings.⁶ Also, when looking at the morbidity and mortality reflected in the malpractice claims, two-thirds of major injury or death arising from diagnostic error occurred among outpatients.⁶ These findings suggest that learning from and reducing diagnostic error, particularly in the outpatient setting, should be a patient safety priority.

Understanding and reducing diagnostic error requires a culture open to discussing them and efforts are impaired if providers hold negative perceptions regarding the subject.^7,8 Physicians identify diagnostic errors as common. When surveyed 54% of pediatricians report making diagnostic errors as frequently as 1 to 2 times per month.⁹ Despite this frequency reported by pediatricians, there remains a paucity of diagnostic error literature in pediatrics compared with that in adult populations.^10,11 Clinicians also report less comfort discussing diagnostic error compared with other error types.¹² Initial work on diagnostic error focused on the inpatient setting in adults; however, recent research highlights that this is an equally concerning problem in the outpatient setting and a gap in pediatrics.^13,14

While comfort regarding diagnostic error has been studied, to the authors’ knowledge, differences attributable to a provider’s practice setting remain unexplored. The team hypothesizes that there will be a significant difference between inpatient versus outpatient pediatric provider comfort when discussing diagnostic errors. To this end, a secondary analysis of a multicenter survey of inpatient and outpatient pediatric providers was performed to assess comfort level in discussing diagnostic errors. Identifying gaps in comfort level is essential to inform work on addressing barriers to implementation of diagnostic error mitigation strategies.

Methods

The team performed a secondary analysis of data collected in a multicenter survey study of pediatric providers from 4 free-standing children’s hospitals of varying size (190–743 beds) located in the Midwestern (2 hospitals), Southern, and the Western United States.¹³ Physicians, advanced practice providers (APPs) and residents, and fellows in general pediatrics, pediatric subspecialties, child neurology, adolescent gynecology, and child psychiatry were eligible. To facilitate interpretation of results, this secondary analysis was restricted to providers whose practice is primarily inpatient (anesthesia, critical care, hospital medicine, and neonatology) or primarily outpatient (adolescent medicine and gynecology, allergy and immunology, child abuse, dermatology, developmental pediatrics, diagnostic imaging, genetics, general pediatrics, rheumatology, and sports medicine). Institutional Review Boards for 2 institutions exempted the study from review; one determined the study to be quality improvement not requiring review; the fourth Institutional Review Board approved the study.

Survey Content

Survey development is described in the parent study.¹³ The survey instrument assessed respondents’ comfort discussing different medical errors using a 5-point ordinal scale ranging from not at all comfortable to very comfortable. The survey presented 5 different medical error types representing a spectrum of direct provider responsibility with brief clinical examples for each. For all 5 errors, respondents were instructed to assume that they were the provider caring for the patient when the error occurred. Respondents indicated their comfort in discussing these errors privately with a trusted colleague and publicly in venues such as morbidity and mortality conferences (MMCs). To view the scenarios, please see the survey in its entirety in the Supplemental Material section http://links.lww.com/AJMQ/A102.

Respondents also selected up to the 3 most significant barriers to discussing their diagnostic errors resulting in harm at MMCs from a predefined list. To minimize biasing responses related to provider comfort, survey items assessing perceived barriers to discussing errors followed the scenarios related to provider comfort. Respondents also estimated their frequency of making diagnostic errors regardless of patient harm and the frequency of making diagnostic errors that caused an adverse event.

Survey Administration

Site investigators distributed the survey via their institution’s e-mail and encouraged participation by enlisting the support of local institutional leaders (eg, department chairs, chief medical and safety officers). Surveys were distributed between May 1, 2018 and June 30, 2018. Reminders were sent at 6, 12, and 22 days to maximize participation, accounting for variability in clinical schedules and service responsibilities. The survey remained open for 30 days. Responses were collected electronically using Research Electronic Data Capture, a secure web-based platform for managing online surveys and databases, hosted by the University of Colorado Clinical and Translational Sciences Institute.¹⁵ Participants received no incentive for survey completion. All study sites confirmed that their institutional peer review and quality improvement protections included unit or section-based MMCs.

Statistical Analysis

The authors report descriptive statistics at the respondent level. Responses to ordinal scale items regarding comfort discussing diagnostic errors were condensed into a dichotomous variable of comfortable (ie, pretty and very comfortable) and uncomfortable (ie, somewhat, a little, and not at all comfortable). Analyses accounted for the clustering of respondents within institutions by using Stata’s survey data analysis procedures and poststratification weights. These analytical weights were created using iterative proportional fitting procedures.¹⁶ Within each institution, the marginal distributions of eligible respondents were known for respondent age, gender, provider type, and years in practice. Information for these variables for eligible respondents was obtained from the medical staff offices of each participating institution. One institution would not release information regarding the ages of trainees and APPs and the years in practice for all providers; this applied to 163 (4.2%) and 218 (5.6%) eligible respondents, respectively. Response rates differed significantly across these variables (except gender) in the parent study.¹² The response rate was 22.6% overall varying from 10.3% to 34.9% across the participating institutions. Survey respondents differed significantly from the eligible population supporting the use of poststratification weights to control for nonresponse bias; the weighted response proportions more closely reflect the original eligible population.¹² Therefore, poststratification weights were applied to correct for any nonresponse bias introduced into the team’s estimates.¹⁷ Significance was set at α < 0.05. Statistical comparisons were performed between categorical variables using the F-test for comparison of proportions. All analyses were performed using Stata 14.2 (StataCorp, 2015; Stata Statistical Software: Release 14).

Results

Surveys were sent to 3881 providers, which included 2078 physicians, 1144 APPs, and 659 residents. Of the eligible participants accessing the survey, 838 answered at least 50% of the questions representing the evaluable cohort. The secondary analysis includes responses from 393 participants: 238 (65%) outpatient and 155 (35%) inpatient providers. No significant demographic differences existed between the inpatient and outpatient cohorts apart from gender reflecting the gender distribution in pediatrics (Table 1).

Table 1.

Comparison of Demographic Characteristics Between Inpatient and Outpatient Clinicians.

Characteristic (n = number who provided any response)	Overall n (%)	Inpatient n (%)	Outpatient n (%)	P-value
Age, years (n = 390)
<30	33 (13.3)	2 (5.1)	31 (17.8)	0.334
30–39	147 (39.9)	63 (43.4)	84 (37.9)
40–49	103 (27.2)	44 (30.7)	59 (25.2)
50–59	61 (9.7)	28 (9.5)	33 (9.8)
≥60	46 (10.0)	16 (11.3)	30 (9.3)
Gendera (n = 385)
Female	271 (68.4)	104 (63.3)	167 (71.3)	0.010
Male	114 (31.5)	48 (36.7)	66 (28.7)
Clinician type (n = 387)
Advanced practice nurse/physician assistant	77 (18.2)	49 (31.8)	28 (10.7)	0.114
Physician	239 (61.4)	94 (59.2)	145 (62.6)
Trainee	71 (20.4)	10 (9.1)	61 (26.7)
Years in practice (n = 391)
Still in training	59 (19.8)	12 (11.7)	47 (24.2)	0.297
<1 year	18 (6.2)	5 (6.0)	13 (6.3)
1–2 years	15 (4.8)	1 (0.5)	14 (7.2)
3–5 years	51 (15.8)	26 (18.4)	25 (14.4)
6–10 years	79 (17.9)	38 (22.4)	41 (15.5)
≥11 years	169 (35.5)	72 (41.1)	97 (32.5)
Last experience with diagnostic error resulting in harmb (n = 366)
<1 year	90 (25.9)	31 (21.4)	59 (28.4)	0.387
1–5 years ago	146 (39.5)	63 (44.9)	83 (36.6)
>5 years ago	71 (16.6)	24 (17.8)	47 (16.0)
Never	59 (18.0)	26 (16.0)	33 (19.1)

^aOne nonbinary and 6 preferred not to answer.

^bInformation only available for respondents; 44 preferred not to answer.

Inpatient and outpatient providers shared similar perceptions on the frequency with which they commit diagnostic errors. Committing diagnostic errors that harmed a patient at least once was endorsed frequently. The proportion of providers committing an error at least 1 to 2 times per year were similar between the inpatient and outpatient providers (Figure 1A; P = 0.577). Regarding errors causing harm, responses for how frequently these events occur was similar for providers in both care settings (Figure 1B; P = 0.181).

Figure 1.

Provider self report of diagnostic errors by practice area. A and B, Clinicians’ self-report of the frequency of committing diagnostic errors that did not result in patient harm (P = 0.577): Clinicians’ self-report of the frequency of committing diagnostic error that did resulted in patient harm (P = 0.181).

When surveying comfort level discussing error, respondents who answered pretty or very comfortable were combined into 1 category for analysis for comparison with all other respondents. For all scenarios presented, both provider groups consistently indicated more comfort discussing errors privately than publicly with no significant difference between groups. Both groups also reported being more comfortable discussing nondiagnostic errors (ie, medication error, patient identification error, or laboratory error) (Figure 2). A significantly greater proportion of inpatient providers reported feeling comfortable discussing errors both privately and publicly compared with outpatient providers in 7 of 10 scenarios presented. Specifically, regarding diagnostic errors, 43.0% of inpatient providers reported feeling comfortable with discussing short-term diagnostic error in public compared with 21.8% of outpatient providers (P = 0.004). A similar discrepancy was seen when they were asked about discussing publicly. This difference existed when discussing short-term errors privately and short-term diagnostic errors with a trusted colleague. All providers responded feeling less comfort discussing long-term diagnostic error with a statistical difference persisting between the inpatient and outpatient groups. No significant difference were found between outpatient and inpatient providers with respect to how frequently diagnostic errors are discussed in MMCs; 53.1% of inpatient providers report that they do it at least sometimes compared with 36.9% of outpatient providers (P = 0.322).

Figure 2.

Providers self-reported comfort level with discussing different types of errors in public vs private setting.

Outpatient and inpatient providers did not differ in responses regarding barriers to discussing diagnostic error publicly. Both groups cited feeling like a bad provider, concerns that decision-making or knowledge base is being judged, and reputation being at stake as the most common barriers (Figure 3).

Figure 3.

Providers self-report to barriers to discussing providers own diagnostic errors in public forums (ie, morbidity and mortality conference).

Multivariate analysis with logistical regression were performed to identify any characteristics that may contribute to differences in comfort when discussing diagnostic error (Tables 2 and 3). Male and older providers were found to be generally more comfortable discussing diagnostic error regardless of whether in public or private. Trainees are more uncomfortable speaking about diagnostic errors in private compared with faculty.

Table 2.

Odds Ratios and 95% Confidence Intervals of Multivariate Logistic Analyses Predicting Comfort Discussing Short- and Long-Term Diagnostic Errors Based on Practice Location and Controlling for Gender, Age, Training, and Timing of Last Diagnostic Error Resulting in Patient Harm.

Dependent variable	Discussing short-term diagnostic error				Discussing long-term diagnostic error
	In private		In public		In private		In public
	N = 358		N = 359		N = 359		N = 359
	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI
Outpatient (vs inpatient)	0.53	0.28-1.01	0.33a	0.16-0.66	0.43a	0.27-0.70	0.39a	0.19-0.81
Male (vs female)	1.83a	1.05-3.27	3.60a	1.81-7.16	1.56	0.98-2.48	2.12	0.95-4.72
Age	0.99	0.81-1.20	1.00	0.75-1.32	1.01	0.88-1.16	1.01	0.77-1.33
APP	3.11a	1.41-6.86	0.85	0.04-19.38	1.58	0.72-3.46	0.56	0.03-9.31
MD (vs trainee)	2.48a	1.10-5.58	0.66	0.04-11.28	1.86a	1.12-3.11	1.05	0.08-13.11
Last diagnostic error resulting in patient harm
≥1 but <5 years	0.40a	0.23-0.70	0.83	0.35-1.96	0.46a	0.27-0.78	0.80	0.25-2.55
≥5 years	0.45	0.18-1.09	0.82	0.43-1.57	0.84	0.30-2.39	0.84a	0.73-0.96
Never (vs <1 year)	0.37a	0.21-0.64	1.80	0.83-3.91	0.65b	0.52-0.81	1.61	0.99-2.64

^aP < 0.05.

^bP < 0.01.

Abbreviations: APP, advanced practice provider; CI, confidence interval.

Table 3.

Odds Ratios and 95% Confidence Intervals of Multivariate Logistic Analyses Predicting Comfort Discussing Laboratory, Patient ID, and Medication Errors Based on Practice Location and Controlling for Gender, Age, Training, and Timing of Last Diagnostic Error Resulting in Patient Harm.

Dependent variable	Discussing laboratory error				Discussing patient ID error				Discussing medication error
	In private		In public		In private		In public		In private		In public
	N = 359		N = 359		N = 359		N = 359		N = 356		N = 357
	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95% CI	Odds ratio	95%-CI	Odds ratio	95%-CI
Outpatient (vs inpatient)	0.45	0.04-5.07	0.22a	0.07-0.68	0.35	0.08-1.51	0.28a	0.13-0.60	0.56b	0.42-0.75	0.40	0.14-1.11
Male (vs female)	0.84	0.29-2.48	2.25	0.36-13.93	1.99	0.33-11.96	2.51	0.53-11.86	2.97b	1.72-5.12	2.96	0.81-10.80
Age	1.09	0.86-1.37	1.23	0.76-1.97	1.26	0.93-1.70	1.21a	1.06-1.38	1.04	0.83-1.31	1.24a	1.03-1.50
APP	0.43	0.06-3.24	0.32a	0.11-0.96	1.57	0.12-20.05	0.39	0.09-1.75	0.75	0.20-2.84	0.37	0.04-3.31
MD (vs trainee)	0.53	0.08-3.59	0.33a	0.11-0.99	1.79	0.47-6.82	0.77	0.14-4.22	1.00	0.21-4.72	0.79	0.10-6.32
Last diagnostic error resulting in patient harm
≥1 but <5 years	0.28	0.03-2.87	0.59	0.11-3.01	0.51	0.14-1.84	0.72	0.41-1.26	0.79	0.20-3.05	0.55	0.11-2.73
≥5 years	0.19	0.02-2.25	0.70	0.15-3.29	0.71	0.29-1.72	0.39b	0.29-0.54	0.95	0.26-3.51	0.33a	0.16-0.67
Never (vs <1 year)	0.19	0.02-1.66	1.95	0.86-4.43	0.36a	0.14-0.95	1.54	0.67-3.54	0.62	0.10-3.93	1.09	0.43-2.77

^aP < 0.05.

^bP < 0.01.

Abbreviations: APP, advanced practice provider; CI, confidence interval.

Discussion

This study demonstrates that outpatient pediatric providers report less comfort discussing medical errors, and specifically diagnostic errors, than their inpatient counterparts both publicly and privately. This difference persisted when adjusting for other contributing factors including gender, provider training, recency of involvement with diagnostic error, and attitudes about the importance of discussing error for system improvement. Although outpatient providers appear to discuss diagnostic errors less frequently than their inpatient peers, this difference was not statistically significant. While these data cannot determine whether frequency of discussion is an independent predictor of comfort, other research suggests that increasing the frequency of such conversations supports general patient safety culture.¹⁸ As both inpatient and outpatient providers report similar barriers to disclosing their own diagnostic errors publicly, the discrepancy in comfort suggests the etiology of this difference may lie in the operational and cultural differences between current inpatient and outpatient pediatric practice. The operational differences between these care arenas may include access to inpatient records and plans; access to subspecialists; the communication infrastructure with other members of the care team; and fewer patient safety practices in outpatient settings including debriefing sessions such as MMCs. The combination of these factors may lead to the discomfort differences observed in this study.

Hospitalist physicians were first introduced in the 1990s, a role that has expanded and altered the traditional role of the primary care physicians for inpatients, which has fundamentally changed the processes of care delivery introducing the need for inpatient to outpatient hand offs.¹⁹ Inpatient providers when previously surveyed believed that the current model has increased the burden of care coordination on all providers.¹⁹ In an interview study of providers in Sweden, physicians indicated that when their role on the care team was not clear, they felt uncertainty or that they were being ignored, which negatively impacted their engagement in patient safety discussion and improvement activities.²⁰ An operational difference between inpatient and outpatient providers is the proximity in time and space to members of the diagnostic team.¹⁹ Distance of ambulatory providers from hospital-based members of the diagnostic team such as laboratory services, subspecialists, pathologists, and radiologists may create ambiguity in their perceived role in a patient’s diagnosis.²¹ In a previous survey study, primary care providers identified failures of communication between generalists and specialists as the most frequent breakdown in diagnostic safety.²² In the hospital, unscheduled impromptu conversations arise easily and organically compared with the effort clinic providers must expend to have equivalent conversations. Although the electronic health record has been promoted as a potential solution to communicate information more effectively about patients moving through primary, secondary, and tertiary care settings, in practice it has not had that desired outcome and, if anything, has created more information silos and barriers between providers.^23,24 By increasing communication across professional boundaries, care teams can improve patient safety, diagnostic safety, and comfort level when discussing challenges.²⁰ Through collective intelligence—the concept that multiple providers are superior at finding solutions to complex problems—diagnostic accuracy will be improved, suggesting that including outpatient providers in the care team during hospitalization can both decrease the incidence and increase discussion of diagnostic error.²⁵

In addition to physical and temporal distance, the infrastructure for patient safety work in the outpatient realm is less robust.^26–28 Initial patient safety efforts in the outpatient setting focused on medication and prescribing safety; little published literature exists about diagnostic safety tools and reporting mechanisms for the outpatient setting.²⁶ Specifically, error reporting systems in the ambulatory care setting have lagged behind those in the inpatient arena.^27,28 Consequently, ambulatory care frequently lacks an infrastructure to detect diagnostic delays in common outpatient scenarios such as hypertension work up, abnormal laboratory value follow up, and depression screening.²⁹

There is a great deal of literature discussing the enhancement of patient safety efforts in care arenas where there is a supportive environment that extends from front line care delivery through leadership.^30,31 However, despite the shift within health care organizations, patient safety culture has not yet pervaded societal norms regarding the provider’s role in medical error.³² The system may predispose to error, but when it comes to malpractice, the individual is more often held accountable.³² While respondents to this study did not identify malpractice risk as a major barrier to discussing diagnostic errors to enhance patient safety, they did cite several aspects of peer perception as significant. This mirrors previous studies where physicians view a competent physician as infallible and flawless.²⁰ The notion that providers lead the care team with ultimate responsibility for patient outcomes is woven in the curriculum from the outset of medical education.^20,32 This ingrained focus on the individual clinician including blame for cognitive errors precludes many efforts to address the systemic gaps that create conditions susceptible to cognitive errors. Therefore, robust patient safety culture requires embracing a concept of collective accountability to create a frameshift.³² Patient implementation of system safety practices including reporting systems, safety huddles, and MMCs has lagged in outpatient compared with inpatient care.^26–28 Perhaps the paucity of exposure to top-down patient safety culture reporting and improvement practices does not challenge the ingrained notion of early training and negatively contributes to outpatient providers willingness to discuss diagnostic safety as seen in the results of this study.

One available option to close the communication gap and create patient safety infrastructure leverages MMCs to ultimately increase comfort level for outpatient providers when it comes to discussing diagnostic error. Studies have shown that up to 21% of cases discussed at MMC involve a diagnostic error and 77% of MMCs uncovered errors not found with other detection methods.^33–35 Although conventional electronic reporting systems tend to identify issues with minute-to-minute care using discrete information within the electronic health record (eg, patient identification errors, medication errors, and other hospital-acquired conditions), long-term problems in the trajectory of care related to diagnosis are only uncovered through thorough case discussion.³³ As MMCs are nearly exclusive to the inpatient setting, outpatient providers lack exposure to regular, often uncomfortable yet normative public conversations about diagnostic errors resulting in harm despite similar experiences in the reported frequency of these errors occurring. One study evaluating the implementation of MMCs, which had not previously been part of routine clinical operations, showed that only 10.6% of physicians reported being comfortable disclosing negative patient outcomes citing barriers including confidentiality concerns, fear of legal repercussions, job security, and damage to professional relationships similar to providers in this study.³⁶ However, repeated participation reduced this discomfort. After a 2.5-year MMC initiative, one institution found staff comfort level in discussing mistakes increased from 13% to 60%.³⁷ While both subspecialists and primary care physicians express interest in receiving feedback about diagnostic delays, tension exists between valuing learning opportunities and wanting to avoid difficult discussions that can decrease job satisfaction.³⁸ Creating the infrastructure for MMCs with both inpatient and outpatient providers represented, in a safe environment supported by expert facilitation, could potentially address some of the operational and cultural challenges for addressing diagnostic errors in the ambulatory setting.

Ultimately, while frontline providers are central to patient safety, without the support at an institutional and management level, engagement with patient safety quickly erodes. Hospitals who facilitate communication openness, nonpunitive response to error, and organization engagement have less patient safety events.³¹ Considering the benefits of MMC and the recent exponential increase in the use of virtual conferencing technology necessitated by the COVID-19 pandemic, a new opportunity exists to promote collaboration in multidisciplinary conferences among providers who historically interact asynchronously. This expanding technology can serve to bridge the distance between inpatient and outpatient infrastructure to successfully complete inclusive MMCs and increase the comfort level of outpatient providers in discussing diagnostic error. Academic institutions have access to a wealth of technology, patient safety resources, and specialists. Therefore, a duty exists to deploy those resources to benefit clinicians in all care settings where resources are less rich. In addition to open discussion between peers through MMCs, large health care institutions also have a duty to provide disclosure coaches and peer support to normalize medical error discussion and provide psychological safety when doing so.³² Ultimately by distributing the resources of MMCs, peer support, and disclosure coaches, it can spread the just culture and close the gap seen regarding diagnostic error discussions between inpatient and ambulatory providers.

This study was limited by a relatively low response rate compared with 2 of 3 previous studies examining diagnostic error among pediatricians (16%–54%).^9,39,40 To correct for this poststratification weight was used to correct for nonresponse bias, and this study included many more participants than those prior investigations. For the purposes of this study, an arbitrary stratification of inpatient and outpatient providers was created based on experiences at the authors’ institutions. Realistically, some providers split their time between both environments and may have perceptions intermediate between the study groups. The team did not have sufficient power to analyze a third intermediate group, which represents an opportunity for future investigation. Also inherently this work studied the comfort level in discussing diagnostic errors, but did not address the nature of the conversation in how they were being discussed, which could be an important area of study in future work. Lastly, this study only included university-affiliated pediatric faculty. Pediatric providers in private practice may have different levels of comfort around diagnostic error discussions than those affiliated with medical centers.

Conclusions

Outpatient providers are significantly less comfortable discussing diagnostic errors than their inpatient counterparts despite similar barriers underlying their reluctance to discuss them. Future works should aim to elucidate the etiology of this discrepancy and develop targeted strategies to decrease provider discomfort surrounding discussing diagnostic errors in the inpatient and outpatient settings.

Acknowledgments

The authors wish to thank the Colorado Clinical and Translational Sciences Institute, which supported the use of the Research Electronic Data Capture (REDCap) database.

Conflicts of Interest

The authors have no conflicts of interest to disclose.