Twenty-four physicians including 12 surgeons and 12 endocrinologists participated in semi-structured interviews about active surveillance of low-risk thyroid cancer. Participants had a median age of 43 years (range 34–68) (Table 1). The majority were male (71%), white (88%), and in academic practice (88%). Most participants (67%) saw at least 10 patients/year with ≤1cm thyroid cancers. To protect anonymity, quotes are identified as “surgeon” or “endocrinologist.”
TABLE 1. DEMOGRAPHIC DATA
|
|
Surgeons
n (%)
|
Endocrinologists
n (%)
|
Age (median, range)
|
41 (34-67)
|
47 (35-68)
|
Female
|
2 (18)
|
5 (42)
|
Caucasian
|
11 (92)
|
10 (83)
|
Academic practice
|
11 (92)
|
10 (83)
|
Location
|
|
|
East/Northeast
|
4 (36)
|
6 (42)
|
South
|
2 (18)
|
0 (0)
|
Midwest
|
2 (18)
|
4 (33)
|
West
|
2 (18)
|
1 (8)
|
Treat >10 PTMC/year
|
8 (67)
|
8 (67)
|
>50% of practice is thyroid cancer
|
12 (100)
|
6 (50)
|
>25 thyroid surgeries/year
|
12 (100)
|
-
|
Member of the ATA
|
12 (100)
|
12 (100)
|
Read all of 2015 ATA guidelines
|
7 (58)
|
9 (75)
|
Read at least part of 2015 ATA guidelines
|
4 (33)
|
2 (17)
|
PTMC, papillary microcarcinoma (defined as papillary thyroid cancer measuring 1 cm or smaller); ATA, American Thyroid Association.
|
Barriers to Active Surveillance
I. Physician-Related Factors
Several key barriers to active surveillance emerged at the physician-level, including physician attitudes and knowledge.
Physician Attitudes
Outcome Expectancy:The most prominent barriers discussed during interviews were related to participants’ beliefs about anticipated patient- and disease-related outcomes or consequences of active surveillance. Participants frequently expressed concern about the risk of metastasis if cancer remained in place. For example, Surgeon 5 stated, “For my own personal beliefs and biases, I feel that any potential risk for cancer to develop and spread throughout the body is unacceptable.” Participants also anticipated experiencing negative emotions, such as guilt, if an adverse outcome occurred during surveillance. Surgeon 1 described, “I would feel uncomfortable…if it grew or spread then I feel like that would be on my shoulders.”
In addition, almost half of participants expressed concern that adverse patient outcomes could result from loss of follow-up. They described hesitancy to offer surveillance to unreliable patients because of the potential for cancer progression. For example, Surgeon 6 explained, “it would be safer to [remove] half the thyroid if the patient is never going to come back.” Regardless of specialty, participants often viewed surgery as “definitive therapy” (Endocrinologist 10) that provided “certainty” (Endocrinologist 1) and a predictable, expected outcome, whereas surveillance introduced uncertainty.
Some participants also worried that patients would file lawsuits if they experienced an adverse outcome during surveillance. They believed it was important to disclose the potential for poor outcomes prior to initiating surveillance. Surgeon 10 described feeling “legally obligated to say, ‘Waiting can allow that cancer a chance to progress.’”
While many participants described concern about cancer spreading, others acknowledged that metastasis rarely occurs during surveillance. For example, Surgeon 8 stated, “every clinician is afraid they’re gonna be surveilling somebody who explodes with lymph node disease…the truth of the matter is that’s extremely unlikely to ever happen.”
In addition to concern about outcomes, some participants discussed the impact of surveillance on their practice and potential for losing patients and/or referring providers as consequence of not offering thyroidectomy (Figure 2). Even non-surgeons acknowledged patients would seek second opinions and eventually find “a surgeon who would agree to do [surgery]” (Endocrinologist 1).
Agreement:A less prominent barrier to active surveillance was lack of agreement with the guidelines and concern about the concept of watchful waiting. Some participants described being hesitant to recommend surveillance because they are “not treating” (Figure 2) and patients might “feel like we’re not caring for them” (Surgeon 6).
Self-Efficacy: Participants also discussed lacking confidence in performing active surveillance. Those who viewed their own lack of experience as a barrier to discussing and offering the approach tended to describe surveillance in vague, as opposed to detailed, terms (Figure 2). Participants also mentioned their inability to convince patients to undergo surveillance because the option is “hard to sell” (Surgeon 1).
Learning Culture & Motivation: While no participants were unwilling to implement active surveillance in their own practice, many expressed doubts about other physicians’ motivation to learn about and incorporate surveillance clinically. Participants believed other physicians are aware of the guidelines, but choose not to follow them or present surveillance as an option (Figure 2). Specifically, some participants thought surgeons were not open to surveillance because “they are going to want to cut” (Endocrinologist 6). Meanwhile, surgeons often described thyroidectomy as a “low-risk” option that “provides good results” (Surgeons 7 and 8).
Physician Knowledge
Awareness & Familiarity: Few participants expressed concern regarding complete lack of awareness of the guidelines supporting active surveillance. However, many believed lack of familiarity with and/or inappropriate application of the guidelines were barriers. Participants described lag time from guideline release to clinical implementation and expressed that education on how to perform surveillance is lacking (Figure 2).
II. Guideline-Related Factors
Barriers to active surveillance that emerged at the guideline-level included lack of evidence, lack of clear intervention goals, and complexity.
Evidence: The most apparent guideline-related barrier was concern about the lack of evidence supporting active surveillance. Many participants believed that data were insufficient and highlighted the lack of prospective trials of patients undergoing surveillance in the US (Figure 3). While some participants discussed trials conducted in Japan as support for active surveillance, several others expressed hesitancy as this data “may not apply to the US population” (Endocrinologist 2).
Intervention Goals: Many participants believed the guidelines lacked clear recommendations and expressed concerns about the goals and duration of active surveillance. Participants described a desire for clear surveillance protocols with more explicit details about patient follow-up, age-specific recommendations, and universal thresholds for nodule progression that prompt further intervention (Figure 3).
Complexity: Participants also discussed that the written guidelines for active surveillance were vague and cited the difficulty and risks of implementing surveillance in their practice without a clear system of patient follow-up in place (Figure 3).
III. External Factors
Barriers to active surveillance on the external-level focused on social and clinical norms, lack of collaboration, organizational constraints, and lack of resources.
Norms: Barriers related to social and clinical norms were prevalent throughout the interviews, particularly concerns over societal attitudes and beliefs about cancer. All participants discussed the impact of “the C word” on patients’ willingness to consider active surveillance (Figure 4). Most described this barrier as resulting from “being in a culture where cancer is bad and needs to get out” (Surgeon 1).
Participants discussed the influence of clinical norms on patient expectations. Surgeons in particular described how once a patient is referred to a surgeon, it is difficult to get the patient to consider non-surgical management (Figure 4). For example, Surgeon 6 stated, “It’ll take a long time for active surveillance to be put into practice from the surgeon’s perspective…because [patients] expect surgery when they come to see you.”
Collaboration: Participants believed lack of collaboration between specialists poses a barrier to active surveillance. They highlighted the potential for patients to receive conflicting advice about treatment options (Figure 4). Some participants discussed how patient expectations set by referring providers can “create lots of conflict” if their own recommendation does not align with that of the referring provider (Surgeon 3).
Organizational Constraints: Participants also discussed that active surveillance may be feasible on the individual provider-level, but most participants saw their ability to implement surveillance as being limited by lack of organizational infrastructure and their larger practice environment (Figure 4). Organization-level implementation would require “institution-specific protocols for monitoring patients” which can be difficult to put in place (Endocrinologist 5).
Resources: Participants additionally identified numerous resource-related barriers to active surveillance. These barriers included the lack of patient access to ongoing care, lack of skilled community ultrasound technologists, and the increased physician workload that would result from surveilling a large cohort of patients (Figure 4).
Facilitators of Active Surveillance
While multiple barriers to active surveillance emerged in the interviews, several facilitators were also identified.
I. Physician-Related Factors
Physician-related facilitators of active surveillance included positive attitudes about outcome expectancy and self-efficacy as well as knowledge, awareness, and familiarity with the guidelines (Figure 2). Some participants supported the open-ended structure of active surveillance and framed surveillance as a “bridge” to future, more definitive treatment. Most participants identified shared decision-making as another facilitator of surveillance (Figure 2). In terms of physician knowledge, participants described the role of physician education in improving guideline awareness and familiarity and saw continuing medical education as a strategy to increase knowledge about how to perform active surveillance (Figure 2).
II. Guideline-related factors
At the guideline-level, clear intervention goals was the only facilitator identified. Nearly all participants described how the goal of the guideline was to avoid surgery and its potential complications, which was a benefit and facilitator of active surveillance (Figure 3).
III. External Factors
Facilitators of active surveillance that emerged at the external-level included social and clinical norms, particularly related to attitudes and beliefs about surgery and the presence of collaboration (Figure 4). Many participants described patients’ fear of surgery as a facilitator, citing how surveillance is often the treatment choice for patients who are more afraid of surgery than the possibility of cancer progression. Most participants also expressed that collaboration and a multidisciplinary team approach facilitates guideline implementation (Figure 4).