Abstract
Background:
Prior studies demonstrate that older adults tend to undergo less surgery for thyroid cancer. Our objective was to utilize a discrete choice experiment to identify factors influencing surgical decision-making for older adults with thyroid cancer.
Methods:
Active and candidate members of the American Association of Endocrine Surgeons were invited to participate in a web-based survey. Multinomial logistic regression (MNL) was utilized to assess patient and surgeon factors associated with treatment choices.
Results:
Complete survey response rate was 25.7%. Most respondents were high volume surgeons (88.5%) at academic centers (76.9%). MNL demonstrated that patient age was the strongest predictor of management. Increasing age and comorbidities were associated with the choice for active surveillance (P=0.000), not performing a lymphadenectomy in patients with nodal metastases (RRR:2.5,95%CI:1.4–4.2,P=0.002 and RRR:1.6,95%CI:1.2–2.1,P=0.004, respectively) and recommending hemithyroidectomy versus total thyroidectomy for a cancer>4cm (RRR:4.4, 95%CI:2.5–7.9,P=0.000 and RRR:3.4,95%CI:2.3–5.1,P=0.000, respectively). Surgeons with≥10 years of experience (RRR:3.3,95%CI:1.1–10.3,P=0.039) favored total thyroidectomy for a cancer<4cm, and non-fellowship trained surgeons (RRR:7.3,95%CI:1.3–42.2,P=0.027) opted for thyroidectomy without lymphadenectomy for lateral neck nodal metastases.
Conclusions:
This study highlights the variation in surgical management of older adults with thyroid cancer and demonstrates the influence of patient age, comorbidities, surgeon experience, and fellowship training on management of this population.
This study highlights the variation in treatment patterns among high volume endocrine surgeons for older adults with thyroid cancer, and demonstrates the impact of patient age, comorbidities, and surgeon characteristics on surgical decision-making by utilizing the discrete choice experiment methodology. The understanding of factors that affect treatment decisions is important for the development of risk indices and guidelines specific to the management of older adults with thyroid cancer in order to standardize consistent treatment.
Introduction:
The population of older adults in the U.S. is on the rise, with an estimated 58% increase in the population aged 65 and older and an 85% increase in the population aged over 85 by the year 20351. The incidence of thyroid cancer is also increasing across all age groups2. While thyroidectomy remains the gold standard treatment of thyroid cancer, several population-based studies suggest that an age-related treatment variation may exist. Analysis of patients with differentiated thyroid cancer (DTC) in the Surveillance, Epidemiology, and End Results (SEER) database from 2004–2015 demonstrated that compared to younger adults, older adults (age ≥65) were more likely to undergo active surveillance and had a significantly worse disease-specific survival. Ho, et al. also found that while active surveillance may be an appropriate and safe option for younger adults with DTC, it was associated with significantly worse disease-specific survival for patients >72 years old3. Additionally, both of these studies found that older adults underwent thyroid lobectomy more frequently compared to the younger cohort4. While these studies demonstrate varying patterns of surgical practice, they do not adequately address the decision-making process that led to these findings.
Older adults pose unique challenges for surgical decision-making, as physicians must weigh the risks of surgery against the likelihood of cancer progression, morbidity, and mortality. On one hand, the majority of thyroid cancers are small, slow-growing, and low-risk, and despite its increasing incidence, mortality from thyroid cancer has remained stable2, 5. Furthermore, recent studies have found that patients pursuing active surveillance for DTC ≤1.5cm have slow or minimal progression of disease, especially in older adults6, 7. On the other hand, other studies have found that increasing patient age has been associated with higher rates of recurrence, worse prognosis, aggressive disease characteristics, lymph node (LN) metastasis, and increased mortality8–10. While single institution studies cite the safety of thyroidectomy in older adults, population-based studies demonstrate an increased rate of complications and hospital length of stay compared to younger adults10–12. Surgeon preference and experience may also affect surgical recommendations in this patient population. It remains unclear how surgeons currently assess older adults and decide upon surgical treatment or lack thereof.
We therefore developed a survey, in the form of a discrete choice experiment (DCE) to understand surgeon preferences for treatment of thyroid cancer in older adults based on patient age, functional status, comorbidities, tumor size, and lymph node metastases13. The primary objective of our study was to identify the patient and surgeon characteristics that are most influential in surgical decision-making for older patients with differentiated thyroid cancer.
Methods and Analysis:
Survey Design
The Discrete Choice Experiment (DCE) method is a unique quantitative technique used in health care research to elicit preferences, priorities, and individual features associated with particular services. While these methods have been primarily used to understand patient preferences, they have also been applied to study clinical decision-making14, 15. DCE presents clinical scenarios and forces respondents to make trade-offs among different treatment options, providing insight into the relative importance of each characteristic, or attribute, that makes up that particular scenario. The range of values that define an attribute are known as levels.
Standard DCE methodology was used to design, implement, and analyze this survey14. After a literature review and key informant interviews with high-volume endocrine surgeons and otolaryngologists, attributes, or characteristics, for the clinical scenarios were chosen (Table 1). Attributes included age, comorbidities, functional status, tumor size, and lateral neck lymph node status. Three overarching clinical scenarios based on extent of disease were created: 1) tumor size <4cm, node negative, 2) tumor size <4cm with lateral neck lymph node metastasis, and 3) tumor size ≥4cm, node negative. Tumor size was included to assess variation in extent of thyroidectomy (hemithyroidectomy versus total thyroidectomy), and presence of lateral neck lymph node metastasis was utilized to assess whether a lateral neck lymph node dissection would be performed. Within each of the three extent of disease scenarios, eight patients with varying levels of attributes for age, comorbidities, and functional status were created for a total of 24 clinical scenarios (Figure 1).
Table 1.
Attributes and Levels in DCE Clinical Scenarios
| Attributes | Levels |
|---|---|
| Chronological Age | <55 |
| 55–64 | |
| 65–79 | |
| ≥80 | |
| Comorbidities | ≤1 Comorbidity |
| 2–3 Comorbidities | |
| ≥4 Comorbidities | |
| Functional Status | Good |
| Poor | |
| Nodal Status | Positive |
| Negative | |
| Tumor Size | <4cm |
| ≥4cm |
Figure 1.
Sample of a portion of the first clinical scenario in survey
A male presents with a 2.5 cm palpable, mobile, right-sided PTC with no comperessive symptoms.No suspicious lymph nodes are seen on ultrasound. He has no family history of thyroid cancer or prior exposure to XRT. What would you recommend for each of the following patients?
In addition to the clinical scenarios, the survey instrument included questions regarding surgeon demographics, practice, and methods utilized to assess older adults’ ability to tolerate surgery. The survey instrument was extensively piloted to test face validity and was iteratively refined. Under IRB approval, survey invitations were electronically distributed using the Qualtrics XM (SAP, WA, USA) interface to all candidate and active members of the American Association of Endocrine Surgeons (AAES) who practice in the United States. The link to the survey was e-mailed with an introduction regarding its purpose and informed consent, followed by three separate reminder e-mails sent 1 week apart.
Statistical Analysis
All statistical analyses were performed using STATA version 15 (StataCorp., TX, USA). Descriptive statistics were compared using Fisher’s exact test or rank sum, as appropriate. Patient attributes were modeled as categorical variables. Five surgeon factors were evaluated including surgeon age, annual thyroid surgery volume (low and high volume were defined as <25 and ≥25 thyroidectomies, respectively16), fellowship training, years in practice, and practice limited to endocrine surgery only.
The association among the treatment choice and patient or surgeon factors for each clinical scenario was analyzed using multinomial logistic regression (MNL) with robust variance estimates, clustered by respondent to account for within-respondent variation. The regression provided relative risk ratios, which reflect the probability of choosing a particular treatment.
Results:
Respondent Characteristics
Overall survey response rate was 128/404 (31.7%), with 104 (25.7%) members completing all portions of the survey (Table 2). Of the 104 that completed the survey, the mean age was 46 years old (range 32–71). The majority of the respondents were male (56.7%), white (75%), fellowship-trained (83.6%) and high volume thyroid surgeons (88.5%). The median number of years in practice was 10 (IQR 5–18.5 years), and 46.2% reported an endocrine surgery only practice. The majority of respondents described their practice setting as academic with teaching (76.9%), followed by community hospital with an academic affiliation (17.3%), and community hospital without teaching (5.8%). Over half (56.3%) reported that 25–50% of the thyroid operations they performed in the past two years were on patients ≥65 years old.
Table 2.
Demographics and Practice Characteristics of Complete Survey Respondents
| Characteristic | (n = 104) Number (%) |
|---|---|
| Mean Age (Range) | 46 (32–71) |
| Gender | |
| Male | 59 (56.7) |
| Fellowship Training | |
| Endocrine Surgery | 70 (67.3) |
| Surgical Oncology | 14 (13.5) |
| Endocrine Surgery and Surgical Oncology | 1 (1.0) |
| Otolaryngology, Head and Neck | 2 (1.9) |
| None | 17 (16.3) |
| Percentage of Practice Limited to Endocrine Surgery | |
| 100% | 48 (46.2) |
| ≥50% | 32 (30.8) |
| ≥25% but <50% | 17 (16.3) |
| <25% | 7 (6.7) |
| Years in Practice | |
| <5 | 23 (22.1) |
| 5–9 | 28 (26.9) |
| 10–19 | 28 (26.9) |
| ≥20 | 25 (24.0) |
| Annual Volume of Thyroid Surgeries | |
| High (>25) | 92 (88.5) |
| Practice Setting | |
| Academic with Teaching | 80 (76.9) |
| Community with Academic Affiliation | 18 (17.3) |
| Community | 6 (5.8) |
Self-Reported Methods to Assess Older Adults
When selecting methods to evaluate a patient’s ability to tolerate surgery (Figure 2), respondents placed most importance on number and type of comorbidities for patients both <65 years of age (92.9%) and ≥65 years (96.1%). For patients ≥65 years old, subjective assessment of functional status (89.2%), and the “eyeball test” or a clinician’s initial intuition of the patient’s health (83.3%), were the next most important methods utilized. For patients <65 years old, the frequency of use of the “eyeball test” (80.6%) was similar, but estimate of functional status (64.1%) was less commonly used for this age group.
Figure 2.
Surgeon-reported method deeed important when assessing patients for surgical maagement of thyroid cancer
The least important methods across all age groups were objective measures of comorbidity and functional status, with only 1.0% of respondents using a comorbidity index, and 3.9% and 7.8% of participants reported the calculation of a frailty index for patients <65 and ≥65 years of age, respectively. Chronological age was considered important by 25.2%, 35.7%, and 64.3% in patients <65, 65–79, and ≥80 years old, respectively. Other methods of assessment frequently utilized were the assessment of patient comorbidities by other physicians and an estimate of life expectancy, with the majority (75.0%) of participants saying that life expectancy <3 years would be an indication to not operate. Of the respondents, 69.5% acknowledged the existence of guidelines from the American College of Surgeons Coalition on Geriatric Surgery for the peri-operative management of older adults, but only 24.4% reported using them17.
Association Between Treatment Decisions, Patient Factors, and Surgeon Characteristics
When presented with three overarching clinical scenarios, MNL demonstrated that age was the strongest predictor of management followed by comorbidities regardless of the tumor size or extent of disease. Poor functional status was significant in univariable MNL, but was omitted in multivariable MNL analysis due to collinearity.
For the first clinical scenario, a 2.5-cm, node-negative PTC confined to the thyroid gland, significant practice variation existed. In a 40-year-old, relatively healthy patient, who served as a reference patient, 64.6% of respondents chose to perform a total thyroidectomy, 35.4% thyroid lobectomy, and none opted for active surveillance, whereas for an 88-year-old patient with ≥2 comorbidities, 11.5% chose total thyroidectomy, 17.3% thyroid lobectomy, and 71.2% active surveillance. However, 16.3% of respondents chose the same treatment for all patients in this scenario regardless of patient presentation, of which most chose hemithyroidectomy. MNL demonstrated that patient age (RRR: 3.2×105, 95% CI: 2.2×105–4.5×105, P= 0.000) and comorbidities (RRR: 5.8, 95% CI: 4.3–7.9, P= 0.000) were significant independent predictors for pursuing active surveillance compared to hemithyroidectomy. When comparing the choice for total versus hemithyroidectomy, patient age ≥65 was associated with a lower likelihood of recommending a total thyroidectomy (RRR: 0.7, 95% CI: 0.4–1.0, P= 0.041), whereas surgeon experience ≥10 years was associated with a higher likelihood (RRR: 3.3, 95% CI: 1.1–10.3, P= 0.039) (Table 3A).
Table 3A.
Node Negative PTC <4cm: Patient and Surgical Determinants of Choice for Total Thyroidectomy or Active Surveillance Compared to Thyroid Lobectomy.
| Characteristic | Total Thyroidectomy vs. Thyroid Lobectomy (Ref) | Active Surveillance vs. Thyroid Lobectomy (Ref) | ||||
|---|---|---|---|---|---|---|
| RRR | 95% CI | P Value | RRR | 95% CI | P value | |
| Patient Attributes | ||||||
| Age | ||||||
| <55 Years | Ref | Ref | ||||
| 55–64 Years | 1.1 | 0.9–1.3 | 0.35 | 27.4* | 19.9–37.5 | 0.000 |
| Age ≥65 | 0.7* | 0.4–1.0 | 0.041 | 1.2×104* | 8.0×103–1.7×104 | 0.000 |
| 65–79 Years | 1.0 | 0.8–1.2 | 0.92 | 39.0* | 30.3–50.2 | 0.000 |
| ≥80 Years | 0.6* | 0.4–0.9 | 0.03 | 296.2* | 222.8–393.8 | 0.000 |
| Comorbidities | ||||||
| ≤ 1 | Ref | Ref | ||||
| ≥2 Comorbidities | 0.9 | 0.7–1.1 | 0.286 | 5.8* | 4.3–7.9 | 0.000 |
| 2–3 | 0.9 | 0.7–1.1 | 0.375 | 1.8* | 1.4–2.3 | 0.000 |
| ≥4 | 0.9 | 0.7–1.2 | 0.626 | 3.6* | 2.6–5.0 | 0.000 |
| Surgeon Characteristics | ||||||
| Age | 1.0 | 0.9–1.0 | 0.179 | 1.0 | 0.9–1.0 | 0.365 |
| Years in Practice | ||||||
| < 10 | Ref | Ref | ||||
| ≥10 | 3.3* | 1.1–10.3 | 0.039 | 2.4 | 0.9–5.6 | 0.079 |
| Fellowship Training | ||||||
| Yes | Ref | Ref | ||||
| No | 1.5 | 0.5–4.8 | 0.462 | 0.5 | 0.2–1.3 | 0.140 |
| Endocrine Surgery Only Practice | ||||||
| Yes | Ref | Ref | ||||
| No | 1.3 | 0.6–2.7 | 0.513 | 0.8 | 0.4–1.5 | 0.421 |
| Annual Volume | ||||||
| Low | Ref | Ref | ||||
| High | 0.3 | 0.04–1.5 | 0.131 | 0.9 | 0.3–2.3 | 0.746 |
Ref= Reference
RRR= Relative Risk Ratio
CI= Confidence Interval
Functional status omitted due to co-linearity with co-morbidities
All P values set for significance <0.05
For the second clinical scenario, a 2.5-cm PTC confined to the thyroid gland with lateral neck lymph node metastases, significant practice variation was again observed. For the reference 40-year-old patient, 88.1% of respondents chose to perform a total thyroidectomy with central and lateral neck lymph node dissection (LND), 5.9% hemithyroidectomy with central and lateral LND, 6.0% hemi or total thyroidectomy without lateral neck LND, and 0% opted for active surveillance, whereas for an 88-year-old patient with ≥2 comorbidities, 27.7% chose to perform a total thyroidectomy with central and lateral neck LND, 7.9% hemithyroidectomy with central and lateral neck LND, 10.0% hemi or total thyroidectomy without lateral neck LND, and 54.5% opted for active surveillance. Interestingly, 26.7% of respondents chose to perform a total thyroidectomy with central and lateral neck LND for all patients in this scenario. When comparing the choice for active surveillance to total thyroidectomy with central and lateral LND, MNL demonstrated that increasing patient age (RRR: 8.9×105, 95% CI: 5.9×105–1.5×106, P= 0.000) and ≥ 2 comorbidities (RRR: 9.9, 95% CI: 6.6–14.9, P= 0.000) were associated with a higher likelihood of active surveillance. Surgical practice dedicated to endocrine surgery only was associated with a lower likelihood of pursuing active surveillance (RRR: 0.4, 95% CI: 0.2–0.9; P= 0.037). Regarding management of the lateral neck lymph node metastases, factors associated with surgical management without a lateral neck LND included patient age (RRR:2.5, 95% CI: 1.4–4.7, P= 0.002) and comorbidities (RRR:1.6, 95% CI: 1.2–2.1, P= 0.004). Non fellowship trained surgeons were also more likely to forgo lateral neck LND (RRR: 7.3, 95% CI: 1.3–42.2 P= 0.027), as were high volume surgeons (RRR: 1.6×107, 95% CI: 6.7×106 −3.8×107, P= 0.000) (Table 3B).
Table 3B.
Node Positive PTC <4cm: Patient and Surgical Determinants of Choice for Management of Lateral Neck Lymph Node Metastasis and Choice for Active Surveillance compared to Total Thyroidectomy with Central & Lateral Neck Dissection.
| Characteristic | −Lymphadenectomy vs. +Lymphadenectomy (Ref) | Active Surveillance vs. Total Thyroidectomy with Central & Lateral Neck Dissection (Ref) | ||||
|---|---|---|---|---|---|---|
| RRR | 95% CI | P value | RRR | 95% CI | P value | |
| Patient Attributes | ||||||
| Age | ||||||
| <55 Years | Ref | Ref | ||||
| 55–64 Years | 1.0 | 0.8–1.2 | 0.991 | 29.3* | 20.9–41.2 | 0.000 |
| Age ≥65 | 2.5* | 1.5–4.4 | 0.001 | 3.0×104* | 2.1×104–4.5×104 | 0.000 |
| 65–79 Years | 1.2 | 0.9–1.5 | 0.144 | 72.7* | 56.9–92.9 | 0.000 |
| ≥80 Years | 2.1* | 1.4–3.3 | 0.000 | 419.1* | 299.8–585.9 | 0.000 |
| Comorbidities | ||||||
| 0–1 | Ref | Ref | ||||
| ≥2 Comorbidities | 1.6* | 1.2–2.1 | 0.004 | 9.9* | 6.6–14.9 | 0.000 |
| 2–3 | 1.2 | 0.9–1.6 | 0.264 | 2.9* | 2.1–4.1 | 0.000 |
| ≥4 | 1.3 | 1.0–1.9 | 0.093 | 3.4* | 2.5–4.7 | 0.000 |
| Surgeon Characteristics | ||||||
| Age | 1.0 | 0.9–1.1 | 0.866 | 1.0 | 0.9–1.0 | 0.386 |
| Years in Practice | ||||||
| <10 | Ref | Ref | ||||
| ≥10 | 0.3 | 0.1–1.8 | 0.195 | 3.0 | 1.0–9.0 | 0.052 |
| Fellowship Training | ||||||
| Yes | Ref | Ref | ||||
| No | 7.3* | 1.3–42.2 | 0.027 | 0.7 | 0.2–2.1 | 0.489 |
| Endocrine Surgery Only Practice | ||||||
| Yes | Ref | Ref | ||||
| No | 0.3 | 0.1–1.0 | 0.053 | 0.4* | 0.2–0.9 | 0.037 |
| Annual Volume | ||||||
| Low | Ref | Ref | ||||
| High | 1.6×107* | 6.7×106–3.8×107 | 0.000 | 1.4 | 0.3–6.5 | 0.634 |
Ref= Reference
RRR= Relative Risk Ratio
CI= Confidence Interval
Functional status omitted due to co-linearity with co-morbidities
All P values set for significance <0.05
For the third clinical scenario, a 4.1-cm, node-negative PTC confined to the thyroid gland, variation in treatment choice was seen across age groups. In the 40-year-old reference patient, 92.1% of respondents chose to perform a total thyroidectomy, 40.6% with prophylactic central LND, 8.0% chose hemithyroidectomy, 5.0% with prophylactic central LND, and 0% opted for active surveillance, compared to an 88-year-old patient with ≥2 comorbidities, 35.7% opted for a total thyroidectomy, 11.9% with prophylactic central LND, 25.7% chose hemithyroidectomy, 6.9% with prophylactic central LND, and 38.6% opted for active surveillance. However, 24.8% of respondents chose the same treatment for all patients, of which, 88.8% chose a total thyroidectomy, with 40.7% choosing concomitant central LND. In comparing the choice of hemithyroidectomy to total thyroidectomy, multivariable MNL demonstrated that patient age (RRR: 4.4, 95% CI: 2.5–7.9, P= 0.000) and comorbidities (RRR: 3.4, 95% CI: 2.3–5.1, P= 0.000) were significant predictors. Choice for active surveillance was also more likely with increasing patient age (RRR: 4.1×105, 95% CI: 2.7×105– 6.3×105, P= 0.000) and comorbidities (RRR: 24.5, 95% CI:10.4–57.8, P= 0.000) (Table 3C).
Table 3C.
Node Negative PTC >4 cm: Patient and Surgical Determinants of Choice for Thyroid Lobectomy or Active Surveillance Compared to Total Thyroidectomy.
| Characteristic | Thyroid Lobectomy vs. Total Thyroidectomy (Ref) | Active Surveillance vs. Total Thyroidectomy (Ref) | ||||
|---|---|---|---|---|---|---|
| RRR | 95% CI | P value | RRR | 95% CI | P value | |
| Patient Attributes | ||||||
| Patient Age | ||||||
| <55 Years | Ref | Ref | ||||
| 55–64 Years | 0.8 | 0.6–1.1 | 0.147 | 10.5* | 5.6–19.9 | 0.000 |
| Age ≥65 | 5.4* | 3.1–9.7 | 0.000 | 3.8×104* | 2.2×104–6.8×104 | 0.000 |
| 65–79 Years | 1.2 | 1.0–1.6 | 0.090 | 32.8* | 22.2–48.5 | 0.000 |
| ≥80 Years | 4.3* | 2.7–6.9 | 0.000 | 1191.3* | 623.0–2278.3 | 0.000 |
| Comorbidities | ||||||
| 0–1 | Ref | Ref | ||||
| ≥2 Comorbidities | 3.4* | 2.3–5.1 | 0.000 | 24.5* | 10.4–57.8 | 0.000 |
| 2–3 | 1.6 | 1.3–1.9 | 0.144 | 4.2* | 2.8–6.2 | 0.000 |
| ≥4 | 2.0* | 1.6–2.7 | 0.000 | 6.7* | 3.8–11.9 | 0.000 |
| Surgeon Characteristics | ||||||
| Age | 1.0 | 0.9–1.1 | 0.884 | 1.0 | 0.9–1.1 | 0.882 |
| Years in Practice | ||||||
| <10 | Ref | Ref | ||||
| ≥10 | 1.0 | 0.3–3.3 | 0.961 | 1.7 | 0.5–5.5 | 0.407 |
| Fellowship Training | ||||||
| Yes | Ref | Ref | ||||
| No | 1.4 | 10.5–4.3 | 0.535 | 0.8 | 0.2–2.7 | 0.683 |
| Endocrine Surgery Only Practice | ||||||
| Yes | Ref | Ref | ||||
| No | 1.5 | 0.7–3.3 | 0.300 | 0.7 | 0.3–1.7 | 0.468 |
| Annual Volume | ||||||
| Low | Ref | Ref | ||||
| High | 0.6 | 0.2–2.2 | 0.462 | 2.2 | 0.3–15.3 | 0.430 |
Ref= Reference
RRR= Relative Risk Ratio
CI= Confidence Interval
Functional status omitted due to co-linearity with co-morbidities
All P values set for significance <0.05
Discussion:
This is the first study to utilize the discrete choice experiment (DCE) methodology to quantify the impact of key clinical and surgeon factors on the choice of therapy for older adults with thyroid cancer and to shed light on reasons for the existence of practice variation. Increasing patient age, followed by number of comorbidities, was the strongest predictor of management of thyroid cancer in older adults, and increased the likelihood of active surveillance across all scenarios. Extent of surgery was also associated with surgeon characteristics including ≥10 years of experience, surgical volume, fellowship training, and amount of practice dedicated to endocrine surgery. Finally, the preferences elicited from the scenarios identified factors deemed important for surgical decision-making, which differed from surgeon-reported factors.
Variation in choice for treatment of thyroid cancer has been well-documented and reflected in the clinical practice guidelines by the American Thyroid Association (ATA)18, 19. Even though the majority of the participants in our study were high volume, fellowship-trained endocrine surgeons practicing at academic institutions with at least 10 years of experience, significant practice variation existed in the choice of treatment for thyroid cancer, especially regarding extent of surgery. Factors that influenced extent of thyroidectomy included tumor size, presence of nodal metastases, patient age and comorbidities, in addition to surgeon characteristics. Surgeon experience ≥10 years was associated with a 3.3-fold higher likelihood in a tumor <4cm. Although many recent studies support the safety and adequacy of treatment with lobectomy for DTC10, 20, 21, other studies demonstrate a high incidence of multifocality and completion lobectomies22, 23. This inconsistency in the literature may be reflected in our results.
Factors associated with the choice for active surveillance included increasing patient age and comorbidities across all patient scenarios. There is growing evidence supporting active surveillance for PTC <1.5cm, with some studies additionally citing a lower risk of growth with increasing age6, 7, 24. However, a recent SEER-based study by Ho, et al. showed that disease specific mortality significantly worsened with patient age >72 years, and also for patients with tumors measuring ≥2cm that were not resected3. Furthermore, although lymph node metastases have been associated with higher rates of recurrence and worse disease-specific survival18, patient age and comorbidities were associated with opting for non-surgical management in patients with lateral neck lymph node involvement. It is not known if there is a patient life expectancy or other objective measure that corresponds with a survival benefit from surgery. This was exhibited in the wide range of answers reported when asked what duration of life expectancy would cause respondents to not recommend surgery. Treatment variation is not inherently a negative finding, and variation is expected in the setting of ongoing research supporting the safety of more conservative treatment strategies in select cases. However, it is important to understand the patient and surgeon factors that lead to deviation from evidence based practice, as nonadherence to guidelines has been shown to compromise survival25.
Not only does the DCE highlight factors important in decision-making, but this method has also been proven to determine preferences better than physician self-reflection alone14. Respondents ranked subjective assessment of comorbidities and functional status as the most important methods to assess surgical fitness, yet objective measurements of these factors, comorbidity and frailty indices, respectively, were used the least frequently. Additionally, less than half reported that chronological age was essential in this decision, despite the influence of age demonstrated by MNL. The subjective and wide range of pre-operative assessments coupled with treatment variation suggests that incorporating objective assessments specific to older adults with thyroid cancer may help to standardize care.
While surgical decision-making is ultimately a discussion between the patient and surgeon, the implementation of an objective surgical risk assessment tool would help to facilitate informed shared decisions without subjective biases. Extension of discrete choice experiment surveys to other physicians involved in the care of older adults with thyroid cancer, such as a broader group of surgeons and endocrinologists, would aid in further characterization of the treatment variation seen in large, population-based studies. This would then yield a platform for possible intervention.
There are several limitations of this study. First, stated preferences for choice of therapy in thyroid cancer was assessed in an idealized context as opposed to understanding actual practice patterns. Real-world scenarios would additionally include information regarding referral information, details of past medical history, unique imaging findings, or patient/family desires, which alter surgical decision-making. Referring physicians also play a significant role and may impact the choice of therapy. The response rate for completion of the survey was 25.7% for a total of 104 respondents, which is low but consistent with other DCE surveys, and non-response bias may have influenced our results13, 14. Furthermore, all participants were endocrine surgeons, mostly general surgery-trained, and the majority had high volume practices in an academic setting with greater than 10 years of experience, which may limit the generalizability of these results as the majority of thyroid surgery is performed by low volume surgeons16.
In conclusion, this study highlights the variation in treatment patterns for older adults with DTC and the impact of chronological age on the surgical decision-making process. While most survey respondents did not perceive chronological age as an important factor for surgical decision-making, MNL analysis found that patient age is the most significant predictor of treatment choices, followed by patient comorbidities, while holding tumor size and extent of disease constant. Surgeon characteristics such as fellowship training, surgical volume, and years of experience also influenced the surgical recommendations delivered. Greater than 10 years of surgical experience and fellowship training were associated with a higher likelihood of performing greater extent of surgery in a patient with lateral neck lymph node metastases. As clinical practice guidelines for DTC continue to evolve, it is important to understand the factors that affect adherence to or deviation from recommended therapies in order to standardize consistent, high-quality treatment. This work encourages the incorporation of risk indices dedicated to the treatment of older adults with thyroid cancer, as well as guidelines specific to this population to optimize treatment strategies and outcomes.
Figure 3.
Treatment Variation for adults with Thyroid Cancer in 3 Overarching Clinical Scenarios,Stratified by Patient
Figure 3A. Treatment Choice for Various Patients with a 2.5cm, Node-Negative PTC
Figure 3B.
Treatment Choice for Various Patients with a 2.5cm PTC with Lateral Neck Lymph Node Metastasis
Figure 3C.
Treatment Choice for Various Patients with a 4.1cm, Node-Negative PTC
Acknowledgements:
Funding/Support: This work was supported by Dr. Mathur’s grant, NIH K23 AG053429, funded by the NIH.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
COI/Disclosure: The authors have no conflicts of interest or other acknowledgements.
References:
- 1.Bureau USC. National Population Projection Tables. 2017.
- 2.Lim H, Devesa SS, Sosa JA, Check D, Kitahara CM. Trends in Thyroid Cancer Incidence and Mortality in the United States, 1974–2013. JAMA. 2017;317:1338–48. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Ho AS, Luu M, Zalt C, Morris LGT, Chen I, Melany M, et al. Mortality Risk of Nonoperative Papillary Thyroid Carcinoma: A Corollary for Active Surveillance. Thyroid. 2019;29:1409–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Shi LY, Liu J, Yu LJ, Lei YM, Leng SX, Zhang HY. Clinic-pathologic Features and Prognostic Analysis of Thyroid Cancer in the Older Adult: A SEER Based Study. J Cancer. 2018;9:2744–50. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Ito Y, Miyauchi A, Kihara M, Higashiyama T, Kobayashi K, Miya A. Patient age is significantly related to the progression of papillary microcarcinoma of the thyroid under observation. Thyroid. 2014;24:27–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ito Y, Miyauchi A, Kudo T, Higashiyama T, Masuoka H, Kihara M, et al. Kinetic Analysis of Growth Activity in Enlarging Papillary Thyroid Microcarcinomas. Thyroid. 2019;29:1765–73. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Tuttle RM, Fagin JA, Minkowitz G, Wong RJ, Roman B, Patel S, et al. Natural History and Tumor Volume Kinetics of Papillary Thyroid Cancers During Active Surveillance. JAMA Otolaryngol Head Neck Surg. 2017;143:1015–20. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Lerch H, Schober O, Kuwert T, Saur HB. Survival of differentiated thyroid carcinoma studied in 500 patients. J Clin Oncol. 1997;15:2067–75. [DOI] [PubMed] [Google Scholar]
- 9.Orosco RK, Hussain T, Brumund KT, Oh DK, Chang DC, Bouvet M. Analysis of age and disease status as predictors of thyroid cancer-specific mortality using the Surveillance, Epidemiology, and End Results database. Thyroid. 2015;25:125–32. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Park HS, Roman SA, Sosa JA. Treatment patterns of aging Americans with differentiated thyroid cancer. Cancer. 2010;116:20–30. [DOI] [PubMed] [Google Scholar]
- 11.Sahli ZT, Zhou S, Najjar O, Onasanya O, Segev D, Massie A, et al. Octogenarians have worse clinical outcomes after thyroidectomy. Am J Surg. 2018;216:1171–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Seybt MW, Khichi S, Terris DJ. Geriatric thyroidectomy: safety of thyroid surgery in an aging population. Arch Otolaryngol Head Neck Surg. 2009;135:1041–4. [DOI] [PubMed] [Google Scholar]
- 13.Hauber AB, Gonzalez JM, Groothuis-Oudshoorn CG, Prior T, Marshall DA, Cunningham C, et al. Statistical Methods for the Analysis of Discrete Choice Experiments: A Report of the ISPOR Conjoint Analysis Good Research Practices Task Force. Value Health. 2016;19:300–15. [DOI] [PubMed] [Google Scholar]
- 14.Nathan H, Bridges JF, Schulick RD, Cameron AM, Hirose K, Edil BH, et al. Understanding surgical decision making in early hepatocellular carcinoma. J Clin Oncol. 2011;29:619–25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Timmermans DR, Gooszen AW, Geelkerken RH, Tollenaar RA, Gooszen HG. Analysis of the variety in surgeons’ decision strategies for the management of left colonic emergencies. Med Care. 1997;35:701–13. [DOI] [PubMed] [Google Scholar]
- 16.Adam MA, Thomas S, Youngwirth L, Hyslop T, Reed SD, Scheri RP, et al. Is There a Minimum Number of Thyroidectomies a Surgeon Should Perform to Optimize Patient Outcomes? Ann Surg. 2017;265:402–7. [DOI] [PubMed] [Google Scholar]
- 17.Mohanty S, Rosenthal RA, Russell MM, Neuman MD, Ko CY, Esnaola NF. Optimal Perioperative Management of the Geriatric Patient: A Best Practices Guideline from the American College of Surgeons NSQIP and the American Geriatrics Society. J Am Coll Surg. 2016;222:930–47. [DOI] [PubMed] [Google Scholar]
- 18.Haugen BR, Alexander EK, Bible KC, Doherty GM, Mandel SJ, Nikiforov YE, et al. 2015 American Thyroid Association Management Guidelines for Adult Patients with Thyroid Nodules and Differentiated Thyroid Cancer: The American Thyroid Association Guidelines Task Force on Thyroid Nodules and Differentiated Thyroid Cancer. Thyroid. 2016;26:1–133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Haymart MR, Banerjee M, Yang D, Stewart AK, Sisson JC, Koenig RJ, et al. Variation in the management of thyroid cancer. J Clin Endocrinol Metab. 2013;98:2001–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Nixon IJ, Ganly I, Patel SG, Palmer FL, Whitcher MM, Tuttle RM, et al. Thyroid lobectomy for treatment of well differentiated intrathyroid malignancy. Surgery. 2012;151:571–9. [DOI] [PubMed] [Google Scholar]
- 21.Harries V, Wang LY, McGill M, Xu B, Tuttle RM, Wong RJ, et al. Should multifocality be an indication for completion thyroidectomy in papillary thyroid carcinoma? Surgery. 2020;167:10–7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Dhir M, McCoy KL, Ohori NP, Adkisson CD, LeBeau SO, Carty SE, et al. Correct extent of thyroidectomy is poorly predicted preoperatively by the guidelines of the American Thyroid Association for low and intermediate risk thyroid cancers. Surgery. 2018;163:81–7. [DOI] [PubMed] [Google Scholar]
- 23.Bilimoria KY, Bentrem DJ, Ko CY, Stewart AK, Winchester DP, Talamonti MS, et al. Extent of surgery affects survival for papillary thyroid cancer. Ann Surg. 2007;246:375–81; discussion 81–4. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Miyauchi A, Kudo T, Ito Y, Oda H, Yamamoto M, Sasai H, et al. Natural history of papillary thyroid microcarcinoma: Kinetic analyses on tumor volume during active surveillance and before presentation. Surgery. 2019;165:25–30. [DOI] [PubMed] [Google Scholar]
- 25.Adam MA, Goffredo P, Youngwirth L, Scheri RP, Roman SA, Sosa JA. Same thyroid cancer, different national practice guidelines: When discordant American Thyroid Association and National Comprehensive Cancer Network surgery recommendations are associated with compromised patient outcome. Surgery. 2016;159:41–50. [DOI] [PubMed] [Google Scholar]