Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 1.
Published in final edited form as: J Surg Res. 2024 May 28;306:588–603. doi: 10.1016/j.jss.2024.04.025

Exploration of Patient Retention in Seeking a Second Opinion; a Retrospective Cohort Study

Alec McCranie a, Anna D Lee a, Laura Leonard a, Sudheer Vemuru a, Victoria Huynh a, Jonathan D’Angelo b, Anne-Lise D’Angelo b, Gretchen Ahrendt a, Christodoulos Kaoutzanis a, Christine Fisher c, Ana Gleisner a, Cheryl Meguid a, Richard Schulick a, Marco Del Chiaro a, Sarah Tevis a
PMCID: PMC11602524  NIHMSID: NIHMS1989712  PMID: 38811259

Abstract

Introduction:

It is common for cancer patients to seek a second opinion for a variety of reasons. Understanding what drives patients to choose to receive treatment with their second opinion provider may uncover opportunities to improve the second opinion process. Therefore, we sought to identify the patient, disease, and treatment characteristics that were associated with second opinion retention rates in patients seeking a second surgical opinion for breast, colon, and pancreatic cancer.

Methods:

We conducted a retrospective cohort study to evaluate patients who sought a second opinion within a large, academic healthcare system for breast, colon, and pancreatic cancer. Electronic medical records were reviewed for second opinions. Patient demographics and characteristics were collected and compared between the retained group and the non-retained groups.

Results:

A total of 237 patients obtained second opinions for breast, colorectal, and pancreatic cancer. Patients that were offered a different treatment plan at their second opinion were more likely to be retained for systemic therapy (P=0.009) for pancreatic cancer and any treatment for colon cancer (P=0.003). Seeing a radiation oncologist (P=0.007) or a plastic surgeon (P=0.02) during the multidisciplinary consultation increased retention rates for breast cancer.

Conclusion:

Surgeons can better identify patients that are more likely to be retained after a second opinion by the individual patient characteristics and treatment factors. Understanding the factors that lead to retention for these three cancer types may help physicians provide the best possible resources for most patients presenting for second opinion evaluations.

Keywords: Second opinion, patient retention, quality improvement, breast cancer, colon cancer, pancreatic cancer

Introduction:

A second opinion is a complete re-evaluation by another physician including a complete review of prior history, testing and treatments and may include reperforming diagnostic imaging and procedures. Almost 80% of patients reported that they would want a second opinion if diagnosed with a serious illness and more than half of U.S. cancer survivors have sought second opinions.1, 2

Prior studies have demonstrated that patients have a variety of reasons for seeking a second opinion.3–5 Most patients seek second opinions for information needs and reassurance.3 However, more than 40% of patients request additional opinions because they are dissatisfied with the communication from their initial provider.4 Upwards of 32% of patients have a change in their treatment plan with half having a major change because of their second opinion.6 Some patients view second opinions as a breach of loyalty and are uncomfortable with the idea of obtaining one, but physicians generally understand a patient’s desire for a second opinion.1, 4, 7 Despite these factors, almost 80% of patients return to their original providers.2, 6

Despite numerous studies on why patients choose to seek second opinions, little has been reported on why patients ultimately choose to receive their care at a certain location following their second opinion. In the few studies that have discussed this subject, there has been high variability between ultimate treatment location ranging from 16% to 65.8% receiving treatment with their second opinion provider.6, 8 Understanding factors that may influence patients’ decisions to obtain treatment from their second opinion provider is important for several reasons. Cost can differ by location, with systemic therapy less costly in community clinic compared to a hospital setting.9 Surgical care varies between centers; the experience of providers at comprehensive cancer centers may allow patients to receive novel or complex operative care that may not be offered at regional centers. Comprehensive cancer centers may offer the availability of specialists who may perform technically complex procedures, experimental treatments and clinical trials which are not frequently offered in the community.10, 11

Understanding what drives patients to choose one treatment provider among multiple consulting providers may be informative in designing health systems that are more efficient at allocating resources. This would allow clinicians and administrators to utilize this information to highlight resources to better inform first or second opinion patients. Therefore, we sought to identify the patient, disease, and treatment characteristics that were associated with second opinion retention rates in patients seeking a second surgical opinion for breast, colon, and pancreatic cancer.

Methods:

Study Design and Setting

We conducted a retrospective cohort study of patients who sought a second opinion for breast, colon, and pancreatic cancer within our healthcare system between July 1, 2020, and October 31, 2021. Our healthcare system is comprised of a large, university-based tertiary referral hospital, cancer center, and numerous affiliated hospitals throughout the state. The cancer center at our institution is a National Cancer Institute (NCI) designated Comprehensive Cancer Center. This study was reviewed and approved by the Institutional Review Board (IRB) prior to any study-related activities.

Participants

To evaluate second opinion retention rates in multiple different populations, we reviewed patient records from the breast, colon, and pancreatic cancer multidisciplinary clinics (MDCs) across our healthcare system. We included all patients who were seen at our institution for a second opinion, which was defined as a patient who had received an initial consultation with an outside oncology provider which included surgical oncology, medical oncology, and/or radiation oncology. Patients were excluded if they had previously seen a physician for an opinion regarding their cancer within our institution or its system affiliates. IRB approved waiver of consent of these patients for review.

Data collection

Medical records were reviewed by auditing each surgeon’s calendar for initial appointment visits. Second opinions were determined by reviewing clinic notes, referrals, and outside records when available. Patient demographics and characteristics were collected including age, body mass index (BMI), smoking and alcohol status, American Society of anesthesiologist (ASA) physical status classification, race, education, employment, marital status, home zip code, insurance type, and medical comorbidities. Factors relating to the patient’s cancer consisted of cancer diagnosis, staging, tumor biology, symptoms present at diagnosis, Eastern Cooperative Oncology Group (ECOG) Performance Status, germline mutation status, prior history of malignancy, and family history of cancer. Initial and second opinion treatment recommendations consisted of provider who made diagnosis, date of diagnosis, initial cancer provider type, treatment recommendation, number of previous opinions, second opinion provider type, second opinion clinic visit type, type of treatment offered, and if a clinical trial was offered. Treatment characteristics included treatment prior to second opinion, treatment offered, and ultimate treatment location. More treatment was defined as offering more invasive therapy, for instance offering surgery when the initial provider offered another round of chemotherapy. Different treatment of similar scope was defined as treatment offered with the same level of intervention, for example both the first opinion and the second opinion provider offered systemic therapy, but different chemotherapeutic agents were recommended. Surgical therapy was defined as treatment offered with the end goal of surgery, for instance neoadjuvant chemotherapy with a planned surgery after. Surgical plans offered at the second opinion may have been a similar surgery as the first opinion or more than the initial provider, for instance offering surgical resection when the initial provider deemed the tumor unresectable. The primary outcome of this study was the retention rate for surgery at our institution and the secondary outcomes were retention rates for all other treatments. Patient disease, and treatment variables were compared to ultimate treatment location for surgery, systemic and radiation therapy, and surveillance.

Bias

To minimize selection bias during initial patient identification, the medical record was filtered to only display initial visits in the clinic, referrals, and outside records. The filter blinded all future follow up visits, procedures, and operations. Only medical record numbers (MRNs) were collected initially and chart review for data collection was performed after all the MRNs were collected.

Statistical analysis

Descriptive statistics were used to summarize characteristics identified during chart review. Differences in patient demographic and clinical factors were tested across cancer sites and by retention status using Pearson chi-squared tests and Fishers Exact tests where 25% or more table cells had expected counts less than 5. A 2-sided significance level of 0.05 was used for all statistical tests. Statistical analysis was completed using SAS software Version 9.4 for Windows ((c) 2013 SAS Institute Inc.).

Results:

There was a total of 3,621 patients identified during the study period. There were 1,230 patients with breast cancer, 1,855, patients with colon cancer, and 299 patients with pancreatic cancer that were excluded because they were either a first opinion or were referred from an affiliated hospital for further treatment. A total of 237 patients obtained second opinions for breast, colorectal, and pancreatic cancer at our institution during the study period. There were 85 patients with breast cancer, 71 patients with colon cancer, and 81 patients with pancreatic cancer. Patient demographics are shown in Table 1.

Table 1,

Patient Characteristics by cancer type

Parameter Total N=237 Breast, N=85 Pancreas, N=81 Colorectal, N=71 N (%)
Demographics – mean (SD)
   Age 56.1 (13.4) 51.4 (13.4) 62.5 (11.6) 54.3 (12.6)
   BMI 25.6 (5.6) 25.6 (5.5) 24.5 (5.9) 26.9 (5.0)
Gender – N (%)
   Male 77 (32.5%) 1 (1.2%) 36 (44.4%) 40 (56.3%)
   Female 160 (67.5%) 84 (98.8%) 45 (55.6%) 31 (43.7%)
Race – N (%)
   Asian 11 (4.6%) 3 (3.5%) 6 (7.4%) 2 (2.8%)
   Black 9 (3.8%) 4 (4.7%) 0 (0%) 5 (7.0%)
   White 192 (81.0%) 74 (87.1%) 65 (80.2%) 53 (74.6%)
   More than one race 9 (3.8%) 2 (2.4%) 5 (6.2%) 2 (2.8%)
Ethnicity – N (%)
   Hispanic 20 (8.6%) 8 (9.4%) 5 (6.3%) 7 (10.3%)
   Not Hispanic 213 (91.4%) 77 (90.6%) 75 (93.8%) 61 (89.7%)
Insurance Status – N (%)
   Medicaid 20 (8.4%) 7 (8.2%) 3 (3.7%) 10 (14.1%)
   Medicare 71 (30%) 18 (21.2%) 33 (40.7%) 20 (28.2%)
   Military Insurance 8 (3.4%) 5 (5.9%) 2 (2.5%) 1 (1.4%)
   Private 133 (56.1%) 55 (64.7%) 39 (48.1%) 39 (54.9%)
   Self-pay 5 (2.1%) 0 (0.0%) 4 (4.9%) 1 (1.4%)
Employment status – N (%)
   Employed 106 (44.7%) 42 (49.4%) 29 (35.8%) 35 (49.3%)
   Unemployed 25 (10.5%) 12 (14.1%) 4 (4.9%) 9 (12.7%)
   Retired 50 (21.1%) 13 (15.3%) 23 (28.4%) 14 (19.7%)
Comorbidities – mean (SD)
   CCI 56.1 (13.4%) 51.4 (13.4%) 62.5 (11.6%) 54.3 (12.6%)
Cancer Staging – N (%)
   Stage 0 (DCIS) 23 (9.7%) 23 (27.1%)
   Stage I 59 (24.9%) 28 (32.9%) 25 (30.9%) 6 (8.5%)
   Stage II 50 (21.1%) 25 (29.4%) 21 (25.9%) 4 (5.6%)
   Stage III 43 (18.1%) 8 (9.4%) 17 (21%) 18 (25.4%)
   Stage IV 62 (26.2%) 1 (1.2%) 18 (22.2%) 43 (60.6%)
Open in a new tab

BMI, Body Mass Index

CCI, Charlson Comorbidity Score

Pancreatic Cancer:

There were 81 patients in the pancreatic cancer cohort that were seen as second opinions. The mean age at diagnosis for these patients was 62.5±11.6 years and 55.6% (45) were female. Most patients identified as white (65, 80.2%) and non-Hispanic (75, 93.8%). Patients with locoregional disease represented 77.8% (63) of the cohort, while 22.2% (18) had stage IV disease. 46.8% (37) of the patients proceeded with a different treatment plan than what was offered at their initial consultation. Of these patients 62.2% (23) were offered more treatment than their original provider. Neoadjuvant therapy was offered to 50.6% (40) and of those 72.5% (29) were retained at the second opinion (Table 2). A treatment recommendation with a surgical component was offered to 79% (49), and of those 36.7% (18) received their operation at the second opinion site. There were no significant patient demographics associated with retention for treatment after second opinion. Locoregional disease (Stage I, II, and III) as compared to metastatic was associated with an increased second opinion retention rate for surgery (P=0.03). Patients that had a higher Charlson Comorbidity Index (CCI) were less likely to be retained at the second opinion site for surgery (3.4±2.3 versus 6±3.4) and all additional oncologic care (3.5±2.3 versus 5.9±3.2). Patients that were offered a different treatment plan at their second opinion were more likely to be retained for systemic therapy (P=0.009).

Table 2,

Pancreatic Cancer Retention Rates

Table 2. Pancreatic cancer patient retention for cancer care
Independent Variable Retention for Systemic and Radiation Therapy (N=49) Retention for Surgery (N=30) Retention for Surveillance (N=80) Retention for First Recommended Treatment (N=79) Retention for Any Treatment (N=78)
Retained N (%) P-value Retained N (%) P-value Retained N (%) P-value Retained N (%) P-value Retained N (%) P-value
Gender
   Female 11 (39.3) 0.128 13 (86.7) 1.0 24 (53.3) 0.026 24 (55.8) 0.606 29 (67.4) 0.349
   Male 4 (19.0) 14 (93.3) 10 (28.6) 18 (50.0) 20 (57.1)
Second Opinion
Distance by Zip code
1 to 13 miles 6 (50.0) 0.443 7 (87.5) 1.0 9 (45.0) 0.719 13 (65.0) 0.303 14 (70.0) 0.667
14 to 31 miles 2 (25.0) 6 (100.0) 7 (50.0) 8 (57.1) 9 (69.2)
32 to 149 miles 2 (25.0) 5 (83.3) 12 (44.4) 10 (38.5) 14 (53.9)
150+ miles 2 (22.2) 9 (90.0) 6 (31.60 11 (57.9) 12 (63.2)
Race
   Asian 2 (66.7) 0.249 3 (100.0) 0.564 4 (66.7) 0.512 5 (83.3) 0.202 5 (83.3) 0.362
   Multiple races 0 (0.0) 1 (100.0) 1 (20.0) 1 (20.0) 2 (40.0)
   Unknown 0 (0.0) 2 (66.7) 2 (40.0) 2 (40.0) 2 (40.0)
   White 13 (32.5) 21 (91.3) 27 (42.2) 34 (54.0) 40 (64.5)
Ethnicity
   Hispanic 0 (0.0) 0.543 1 (50.0) 0.2 2 (40.0) 1.0 1 (20.0) 0.184 2 (40.0) 0.359
   Not Hispanic 15 (32.6) 25 (92.6) 31 (41.9) 40 (54.8) 46 (63.9)
Insurance Type
   Medicaid 1 (50.0) 0.534 0 (.) 0.048 1 (33.3) 0.294 1 (50.0) 0.522 1 (50.0) 0.486
   Medicare 7 (29.2) 9 (100.0) 16 (50.0) 16 (48.5) 20 (62.5)
   Private 6 (28.6) 16 (94.1) 17 (43.6) 22 (57.9) 25 (65.8)
   Self-pay 0 (0.0) 1 (33.3) 0 (0.0) 1 (25.0) 1 (25.0)
   Military Insurance 1 (100.0) 1 (100.0) 0 (0.0) 2 (100.0) 2 (100.0)
Employment Status
   Employed 4 (25.0) 0.391 11 (91.7) 0.508 11 (37.9) 0.479 15 (53.6) 0.853 18 (64.3) 0.983
   Unemployed 1 (100.0) 2 (66.7) 3 (75.0) 3 (75.0) 3 (75.0)
   Retired 4 (25.0) 7 (100.0) 8 (36.4) 11 (47.8) 13 (59.1)
   Unknown 6 (37.5) 7 (87.5) 12 (48.0) 13 (54.2) 15 (62.5)
In Medical Occupation
   Medical field 0 (0.0) 0.435 5 (100.0) 1.0 5 (50.0) 0.854 5 (50.0) 0.875 6 (66.7) 0.8
   Non-medical field 9 (34.6) 15 (88.2) 17 (39.5) 24 (55.8) 28 (65.1)
   Unknown 6 (33.3) 7 (87.5) 12 (44.4) 13 (50.0) 15 (57.7)
Family history of cancer
   Yes 10 (34.5) 0.479 18 (94.7) 0.537 23 (47.9) 0.23 28 (58.3) 0.252 31 (66.0) 0.48
   No 5 (25.0) 9 (81.8) 11 (34.4) 14 (45.2) 18 (58.1)
Family history of pancreatic cancer
   Yes 1 (33.3) 1.0 2 (100.0) 1.0 3 (60.0) 0.646 3 (60.0) 1.0 4 (80.0) 0.646
   No 14 (30.4) 25 (89.3) 31 (41.3) 39 (52.7) 45 (61.6)
Gene Mutation
   Yes 9 (42.9) 0.107 6 (100.0) 1.0 14 (51.9) 0.227 15 (55.6) 0.759 18 (66.7) 0.609
   No 6 (21.4) 21 (87.5) 20 (37.7) 27 (51.9) 31 (60.8)
Prior malignancy
   Yes 6 (50.0) 0.148 4 (100.0) 1.0 7 (43.8) 0.91 10 (62.5) 0.402 10 (66.7) 0.732
   No 9 (24.3) 23 (88.5) 27 (42.2) 32 (50.8) 39 (61.9)
Prior pancreatic malignancy
   Yes 0 (.) NC 0 (.) NC 0 (.) NC 0 (.) NC 0 (.) NC
   No 15 (30.6) 27 (90.0) 34 (42.5) 42 (53.2) 49 (62.8)
Pancreatic cancer stage
   Stage I 4 (30.8) 0.973 10 (90.9) 0.03 11 (45.8) 0.347 14 (58.3) 0.238 15 (65.2) 0.175
   Stage II 3 (30.0) 11 (100.0) 7 (33.3) 14 (66.7) 14 (66.7)
   Stage III 3 (25.0) 5 (100.0) 10 (58.8) 8 (47.1) 13 (76.5)
   Stage IV 5 (35.7) 1 (33.3) 6 (33.3) 6 (35.3) 7 (41.2)
Patient received treatment before 2nd opinion
   Yes 6 (27.3) 0.647 19 (90.5) 1.0 20 (45.5) 0.555 25 (58.1) 0.333 30 (71.4) 0.089
   No 9 (33.3) 8 (88.9) 14 (38.9) 17 (47.2) 19 (52.8)
Prior treatment: Systemic
   Yes 1 (11.1) 0.242 16 (88.9) 1.0 11 (39.3) 0.67 17 (63.0) 0.209 20 (74.1) 0.135
   No 14 (35.0) 11 (91.7) 23 (44.2) 25 (48.1) 29 (56.9)
Prior treatment: Surgery
   Yes 1 (20.0) 1.0 1 (100.0) 1.0 4 (66.7) 0.393 2 (33.3) 0.411 4 (66.7) 1.0
   No 14 (31.8) 26 (89.7) 30 (40.5) 40 (54.8) 45 (62.5)
Prior treatment: Radiation
   Yes 1 (100.0) 0.306 5 (100.0) 1.0 5 (83.3) 0.078 6 (100.0) 0.027 6 (100.0) 0.05
   No 14 (29.2) 22 (88.0) 29 (39.2) 36 (49.3) 43 (59.7)
First opinion provider included a general surgeon
   Yes 3 (60.0) 0.16 2 (100.0) 1.0 4 (57.1) 0.451 5 (71.4) 0.438 5 (71.4) 0.621
   No 12 (27.3) 25 (89.3) 30 (41.1) 37 (51.4) 44 (62.0)
First opinion provider included a subspecialty surgeon
   Yes 2 (40.0) 0.635 1 (100.0) 1.0 3 (42.9) 1.0 3 (50.0) 1.0 4 (66.7) 1.0
   No 13 (29.5) 26 (89.7) 31 (42.5) 39 (53.4) 45 (62.5)
First opinion provider included a medical oncologist
   Yes 11 (28.2) 0.47 21 (87.5) 1.0 25 (39.7) 0.326 32 (50.8) 0.402 38 (61.3) 0.582
   No 4 (40.0) 6 (100.0) 9 (52.9) 10 (62.5) 11 (68.8)
First opinion provider included a radiation oncologist
   Yes 1 (100.0) 0.306 1 (100.0) 1.0 2 (100.0) 0.178 2 (100.0) 0.496 2 (100.0) 0.527
   No 14 (29.2) 26 (89.7) 32 (41.0) 40 (51.9) 47 (61.8)
Prior visit at our institution before second opinion visit
   Yes 3 (27.3) 1.0 8 (80.0) 0.251 12 (54.5) 0.18 11 (52.4) 0.933 14 (66.7) 0.67
   No 12 (31.6) 19 (95.0) 22 (37.9) 31 (53.4) 35 (61.4)
Seeking 3rd opinion or more
   No 13 (31.7) 1.0 20 (90.9) 1.0 28 (44.4) 0.498 33 (52.4) 0.782 40 (64.5) 0.542
   Yes 2 (25.0) 7 (87.5) 6 (35.3) 9 (56.3) 9 (56.3)
Second opinion provider included a Surgery
   Yes 10 (24.4) 0.047 27 (93.1) 0.1 30 (42.3) 1.0 37 (52.9) 1.0 44 (63.8) 0.632
   No 5 (62.5) 0 (0.0) 4 (44.4) 5 (55.6) 5 (55.6)
Second opinion provider included a Radiation Oncologist
   Yes 1 (33.3) 1.0 1 (100.0) 1.0 3 (75.0) 0.307 2 (50.0) 1.0 3 (75.0) 1.0
   No 14 (30.4) 26 (89.7) 31 (40.8) 40 (53.3) 46 (62.2)
First and second opinion treatment plans matched
   Yes 6 (20.7) 0.083 9 (90.0) 1.0 12 (30.0) 0.032 15 (38.5) 0.011 19 (50.0) 0.022
   No 8 (44.4) 17 (89.5) 20 (54.1) 25 (67.6) 28 (75.7)
Open in a new tab

P-values based on Chi-Square tests or Fisher Exact tests (where >=25% of cells having an expected count <5) for independence. NC: Not calculated due to having only a single row in the crosstabulation.

Colon Cancer:

There were 71 patients in the colon cancer cohort that were seen as second opinions at our institution. The mean age at diagnosis for these patients 54.2±12.6 years and 56.3% (40) were male. Most patients identified as white (74.6%, 53) and non-Hispanic (89.7%, 61). 39.4% (28) of patients presented with locoregional disease, while 60.6% (43) of patients had metastatic disease. The initial treatment plan offered at the second opinion differed for 68.7% (50) of patients, with 78% (39) of these patients being offered more treatment. 66.2% (47) were offered surgery, and of those 93.6% (44) were retained at the second opinion site (Table 3). Of the variables collected there were no patient demographic variables significantly associated with retention for treatment after second opinion. Patients that were offered a different treatment plan were more likely to be retained for any treatment (P=0.003). Patients that had not undergone treatment prior to the second opinion were more likely to be retained for oncologic surveillance after their treatment period (P=0.006).

Table 3,

Colon Cancer Retention Rates

Table 3. Colon cancer patient retention for cancer care
Independent Variable Retention for Systemic and Radiation Therapy (N=20) Retention for Surgery (N=47) Retention for Surveillance (N=71) Retention for First Recommended Treatment (N=67) Retention for Any Treatment (N=69)
Retained N (%) P-value Retained N (%) P-value Retained N (%) P-value Retained N (%) P-value Retained N (%) P-value
Gender
   Female 7 (58.3) 0.015 16 (100.0) 0.541 16 (51.6) 0.731 23 (82.1) 0.327 25 (83.3) 0.512
   Male 0 (0.0) 28 (90.3) 19 (47.5) 28 (71.8) 30 (76.9)
Second Opinion
Distance by Zip code
1 to 13 miles 3 (75.0) 0.097 9 (100.0) 0.462 8 (61.5) 0.423 12 (92.3) 0.174 12 (92.3) 0.469
14 to 31 miles 2 (40.0) 9 (90.0) 8 (50.0) 11 (73.3) 11 (73.3)
32 to 149 miles 2 (40.0) 13 (100.0) 11 (57.9) 15 (83.3) 16 (84.2)
150+ miles 0 (0.0) 12 (85.7) 8 (36.4) 12(60.0) 15(71.4)
Race
   Asian 1 (100.0) 0.095 1 (100.0) 1.0 1 (50.0) 0.898 2 (100.0) 0.159 2 (100.0) 0.304
   Black 2 (100.0) 3 (100.0) 2 (40.0) 5 (100.0) 5 (100.0)
   Multiple races 0 (.) 2 (100.0) 1 (50.0) 2 (100.0) 2 (100.0)
   Unknown 1 (16.7) 3 (100.0) 3 (33.3) 4 (44.4) 5 (55.6)
   White 3 (27.3) 35 (92.1) 28 (52.8) 38 (77.6) 41 (80.4)
Ethnicity
   Hispanic 1 (33.3) 1.0 4 (100.0) 1.0 3 (42.9) 0.705 5 (71.4) 0.642 5 (71.4) 0.602
   Not Hispanic 6 (40.0) 39 (92.9) 32 (52.5) 45 (78.9) 49 (83.1)
Insurance Type
   Medicaid 2 (40.0) 0.848 4 (100.0) 1.0 6 (60.0) 0.806 6 (66.7) 0.874 7 (77.8) 0.899
   Medicare 3 (42.9) 11 (91.7) 9 (45.0) 14 (73.7) 15 (75.0)
   Private 2 (25.0) 27 (93.1) 20 (51.3) 29 (78.4) 31 (81.6)
   Self-pay 0 (.) 1 (100.0) 0 (0.0) 1 (100.0) 1 (100.0)
   Military Insurance 0 (.) 1 (100.0) 0 (0.0) 1 (100.0) 1 (100.0)
Employment Status
   Employed 2 (33.3) 0.913 24 (96.0) 0.18 20 (57.1) 0.581 26 (83.9) 0.191 30 (90.9) 0.036
   Unemployed 2 (33.3) 3 (100.0) 3 (33.3) 5 (55.6) 5 (55.6)
   Retired 3 (50.0) 6 (75.0) 6 (42.9) 9 (64.3) 9 (64.3)
   Unknown 0 (0.0) 11 (100.0) 6 (46.2) 11 (84.6) 11 (84.6)
In Medical Occupation
   Medical field 1 (50.0) 0.329 7 (100.0) 0.719 6 (60.0) 8 (88.9) 0.661 9 (90.0) 0.832
   Non-Medical field 6 (42.9) 23 (88.5) 20 (46.5) 29 (72.5) 32 (78.0)
   Unknown 0 (0.0) 14 (100.0) 9 (50.0) 14 (77.8) 14 (77.8)
Family history of cancer
   Yes 5 (35.7) 1.0 24 (92.3) 1.0 23 (56.1) 0.18 29 (72.5) 32 (78.0) 0.678
   No 2 (33.3) 20 (95.2) 12 (40.0) 22 (81.5) 23 (82.1)
Family history of Colon cancer
Yes 2 (66.7) 0.27 3 (100.0) 1.0 4 (66.7) 0.429 5 (83.3) 1.0 5 (83.3) 1.0
No 5 (29.4) 41 (93.2) 31 (47.7) 46 (75.4) 50 (79.4)
Gene Mutation
Yes 2 (33.3) 1.0 10 (100.0) 1.0 9 (56.3) 0.527 12 (75.0) 1.0 12 (75.0) 0.724
No 5 (35.7) 34 (91.9) 26 (47.3) 39 (76.5) 43 (81.1)
Prior malignancy
Yes 2 (40.0) 1.0 10 (100.0) 1.0 7 (46.7) 0.819 12 (80.0) 1.0 13 (86.7) 0.718
No 5 (33.3) 34 (91.9) 28 (50.0) 39 (75.0) 42 (77.8)
Prior colon malignancy
Yes 0 (0.0) 0.521 3 (100.0) 1.0 3 (60.0) 0.674 3 (60.0) 0.586 4 (80.0) 1.0
No 7 (38.9) 41 (93.2) 32 (48.5) 48 (77.4) 51 (79.7)
Colon Cancer Stage
Stage I 0 (.) 0.166 5 (100.0) 0.165 3 (50.0) 0.367 5 (100.0) 0.234 5 (100.0) 0.465
Stage II 0 (0.0) 2 (66.7) 2 (50.0) 2 (50.0) 3 (75.0)
Stage III 3 (75.0) 10 (90.9) 12 (66.7) 13 (86.7) 15 (88.2)
Stage IV 4 (26.7) 27 (96.4) 18 (41.9) 31 (72.1) 32 (74.4)
Treatment prior to second opinion
Yes 3 (20.0) 0.031 40 (93.0) 1.0 26 (42.6) 0.006 43 (74.1) 0.675 45 (76.3) 0.491
No 4 (80.0) 4 (100.0) 9 (90.0) 8 (88.9) 10 (100.0)
Prior treatment: Systemic
Yes 2 (15.4) 0.022 36 (94.7) 0.48 22 (40.7) 0.01 38 (74.5) 0.743 40 (76.9) 0.944
No 5 (71.4) 8 (88.9) 13 (76.5) 13 (81.3) 15 (88.2)
Prior treatment: Surgery
Yes 2 (25.0) 0.642 20 (95.2) 1.0 15 (50.0) 0.919 22 (75.9) 0.966 23 (79.3) 1.0
No 5 (41.7) 24 (92.3) 20 (48.8) 29 (76.3) 32 (80.0)
Prior treatment: Radiation
Yes 0 (0.0) 0.521 6 (85.7) 0.391 6 (54.5) 0.705 6 (66.7) 0.437 8 (80.0) 0.367
No 7 (38.9) 38 (95.0) 29 (48.3) 45 (77.6) 47 (79.7)
First opinion provider included a general surgeon
Yes 2 (33.3) 1.0 5 (100.0) 1.0 6 (50.0) 0.957 7 (63.6) 0.438 8 (66.7) 0.624
No 5 (35.7) 39 (92.9) 29 (49.2) 44 (78.6) 47 (82.5)
First opinion provider included a subspecialty surgeon
Yes 2 (50.0) 0.587 2 (100.0) 1.0 5 (71.4) 0.26 4 (66.7) 0.623 5 (71.4) 0.215
No 5 (31.3) 42 (93.3) 30 (46.9) 47 (77.0) 50 (80.6)
First opinion provider included a medical oncologist
   Yes 4 (30.8) 0.651 41 (93.2) 1.0 29 (48.3) 0.753 45 (78.9) 0.234 48 (82.8) 1.0
   No 3 (42.9) 3 (100.0) 6 (54.5) 6 (60.0) 7 (63.6)
First opinion provider included a radiation oncologist
   Yes 0 (.) NC 2 (100.0) 1.0 2 (100.0) 0.239 2 (100.0) 1.0 2 (100.0)
   No 7 (35.0) 42 (93.3) 33 (47.8) 49 (75.4) 53 (79.1) 0.494
Prior visit at our institution before second opinion visit
   Yes 2 (50.0) 0.587 12 (100.0) 0.56 10 (58.8) 0.368 14 (87.5) 0.32 14 (87.5) 0.443
   No 5 (31.3) 32 (91.4) 25 (46.3) 37 (72.5) 41 (77.4)
Seeking 3rd opinion or more
   No 7 (43.8) 0.249 36 (92.3) 1.0 28 (48.3) 0.717 43 (78.2) 0.46 46 (82.1) 1.0
   Yes 0 (0.0) 8 (100.0) 7 (53.8) 8 (66.7) 9 (69.2)
Second opinion provider included a Surgery
   Yes 6 (31.6) 0.35 43 (93.5) 1.0 34 (49.3) 1.0 49 (75.4) 1.0 53 (79.1) 1.0
   No 1 (100.0) 1 (100.0) 1 (50.0) 2 (100.0) 2 (100.0)
Second opinion provider included a Radiation Oncologist
   Yes 5 (83.3) 0.007 1 (50.0) 0.125 5 (55.6) 0.735 6 (75.0) 1.0 7 (77.8) 0.335
   No 2 (14.3) 43 (95.6) 30 (48.4) 45 (76.3) 48 (80.0)
First and second opinion treatment plans matched
   Yes 3 (23.1) 0.583 6 (100.0) 1.0 6 (28.6) 0.034 9 (47.4) 0.003 11 (55.0) 0.003
   No 2 (40.0) 37 (92.5) 26 (56.5) 39 (86.7) 40 (88.9)
Open in a new tab

P-values based on Chi-Square tests or Fisher Exact tests (where >=25% of cells having an expected count <5) for independence. NC: Not calculated due to having only a single row in the crosstabulation.

Breast Cancer:

A total of 85 patients with breast cancer were seen for a second opinion at our institution. Patients with Ductal Carcinoma in Situ (DCIS) represented 27.1% (23) of the cohort, Stage I, II, and III represented 71.8% (61), and stage IV disease represented 1.2% (1). 51.8% (41) of patients were offered a different treatment than what was recommended by the initial provider. More treatment was offered to 58.3% (21) of these patients than their original provider. Upfront surgical therapy was offered to 62.4% (53), and of those 71.7% (38) were retained at the second opinion site. Initial treatment with neoadjuvant therapy was offered to 21.2% (18) of patients and 5.9% (5) were offered adjuvant chemotherapy. Definitive chemotherapy was offered as initial treatment to 4.7% (4) and 3.5% (3) were offered only adjuvant radiation. Of these patients that were offered systemic therapy or radiation, 65.4% (17) were retained at the second opinion site for treatment (Table 4). Patients that identified as non-Hispanic were more likely to be retained at the second opinion site for treatment offered (73.6%, 53, P=0.024) and surveillance (81.8%, 63, P=0.012). Patients that had a visit with a radiation oncologist during their MDC visit were more likely to be retained for treatment (P=0.007) and for surveillance (P=0.0041). Of the 85 patients with breast cancer, 21.2% (18) were seen by a plastic surgeon on the same day as their MDC. Patients seen by a plastic surgeon during their MDC visit were more likely to be retained for the treatment offered (88.2% versus 63.9%; P=0.04) and for surveillance (83.3% versus 55%, P=0.02) at the second opinion site.

Table 4,

Breast Cancer Retention Rates

Table 4. Breast cancer patient retention for cancer care
Independent Variable Retention for Retention for Retention for Retention for Retention for
Systemic and Radiation Therapy (N=26) Surgery (N=53) Surveillance (N=85) First Recommended Treatment (N=79) Any Treatment (N=85)
Retained P-value Retained P-value Retained P-value Retained P-value Retained P-value
N (%) N (%) N (%) N (%) N (%)
Gender
   Female 17 (65.4) NC 37 (71.2) 1.0 65 (77.4) 1.0 54 (69.2) 1.0 68 (81) 1.0
   Male 0 (.) 1 (100) 1 (100.0) 1 (100.0) 1 (100)
Second Opinion
Distance by Zip code
   1–13 miles 4 (57.1) 0.710 13 (76.5) 0.376 19 (76.0) 0.179 17 (70.8) 0.463 20 (80.0) 0.275
   14–31 miles 5 (55.6) 7 (58.3) 16 (66.7) 12 (57.1) 17 (70.83)
   32–149 miles 3 (100.0) 9 (64.3) 17 (94.4) 12 (70.6) 17 (94.4)
   150+ miles 4 (66.7) 9 (90.0) 13 (76.5) 13 (81.3) 14 (82.4)
Race
   Asian 0 (.) 0.365 3 (100.0) 0.293 3 (100.0) 0.067 3 (100.0) 0.399 3 (100) 0.305
   Black 3 (100.0) 1 (100.0) 4 (100.0) 4 (100.0) 4 (100)
   White 13 (59.1) 33 (71.7) 58 (78.4) 46 (67.6) 1 (50)
   Multiple races 1 (100.0) 0 (0.0) 1 (50.0) 1 (50.0) 1 (50)
   Unknown 0 (.) 1 (50.0) 0 (0.0) 1 (50.0) 60 (81.1)
Ethnicity
   Hispanic 1 (25) 0.104 1 (33.3) 0.190 3 (37.5) 0.012 2 (28.6) 0.024 4 (50) 0.038
   Not Hispanic 16 (72.7) 37 (74) 63 (81.8) 53 (73.6) 65 (84.4)
Insurance Type
   Medicaid 2 (100) 0.3197 4 (80) 1.0 6 (85.7) 0.75 6 (85.7) 0.505 6 (85.7) 0.696
   Medicare 1 (25) 9 (69.2) 13 (72.2) 10 (58.8) 14 (77.8)
   Private 12 (66.7) 24 (70.6) 42 (76.4) 36 (69.2) 44 (80)
   Military Insurance 2 (100) 1 (100) 5 (100.0) 3 (100.0) 5 (100)
Employment Status
   Employed 9 (75) 0.567 19 (73.1) 0.834 34 (81.0) 0.213 28 (73.7) 0.703 36 (85.7) 0.221
   Unemployed 3 (60) 4 (57.1) 9 (75.0) 7 (58.3) 9 (75)
   Retired 3 (75) 6 (75) 12 (92.3) 9 (75.0) 12 (92.3)
   Unknown 2 (40) 9 (75) 11 (61.1) 11 (64.7) 12 (66.7)
In Medical Occupation
   Medical field 3 (75) 0.534 5 (71.4) 0.793 9 (75.0) 0.055 8 (72.7) 0.965 9 (75) 0.146
   Non-medical field 10 (71.4) 22 (68.8) 43 (86.0) 32 (69.6) 44 (88)
   Unknown 4 (50) 11 (78.6) 14 (60.9) 15 (68.2) 16 (69.6)
Family history of cancer
   Yes 11 (61.1) 0.667 17 (70.8) 0.89 9 33 (71.7) 0.156 28 (66.7) 0.543 34 (73.9) 0.06 3
   No 6 (75) 21 (72.4) 33 (84.6) 27 (73.0) 35 (89.7)
Family history of breast cancer
   Yes 5 (62.5) 1.0 2 (50) 0.56 8 9 (69.2) 0.475 7 (58.3) 0.496 9 (69.2) 0.231
   No 12 (66.7) 36 (73.5) 57 (79.2) 48 (71.6) 60 (83.3)
Gene Mutation
   Yes 6 (66.7) 1.0 7 (53.8) 0.1 55 20 (83.3) 0.43 13 (59.1) 0.206 20 (83.3) 1.0
   No 11 (64.7) 31 (77.5) 46 (75.4) 42 (73.7) 49 (80.3)
Prior malignancy
   Yes 5 (62.5) 1.0 6 (75) 1.0 14 (82.4) 0.752 11 (68.8) 1.0 14 (82.4) 1.0
   No 12 (66.7) 32 (71.1) 52 (76.5) 44 (69.8) 55 (80.9)
Prior breast malignancy
   Yes 2 (66.7) 1.0 2 (66.7) 1.0 6 (85.7) 1.0 4 (66.7) 1.0 6 (85.7) 1.0
   No 15 (65.2) 36 (72) 60 (76.9) 51 (69.9) 63 (80.8)
Breast cancer stage
   Stage 0 (DCIS) 4 (80) 0.851 8 (50) 0.051 17 (73.9) 0.535 12 (57.1) 0.164 17 (73.9) 0.555
   Stage I 3 (60) 17 (85) 24 (85.7) 20 (80.0) 25 (89.3)
   Stage II 8 (66.7) 11 (84.6) 17 (68.0) 19 (76.0) 19 (76)
   Stage III 1 (33.3) 2 (50) 7 (87.5) 3 (42.9) 7 (87.5)
   Stage IV 1 (100) 0 (.) 1 (100.0) 1 (100.0) 1 (100)
Treatment prior to second opinion
   Yes 8 (66.7) 1.0 9 (75) 1.0 20 (76.9) 0.956 17 (70.8) 0.888 22 (84.6) 0.497
   No 8 (61.5) 28 (71.8) 42 (76.4) 36 (69.2) 43 (78.2)
Prior treatment: Systemic
   Yes 2 (66.7) 1.0 6 (85.7) 0.658 8 (80.0) 1.0 8 (80.0) 0.715 10 (100) 0.1 96
   No 15 (65.2) 32 (69.6) 58 (77.3) 47 (68.1 59 (78.7)
Prior treatment: Surgery
   Yes 6 (75) 0.667 3 (60) 0.614 12 (80.0) 1.0 9 (69.2) 1.0 12 (80) 1.0
   No 11 (61.1) 35 (72.9) 54 (77.1) 46 (69.7) 57 (81.4)
Prior treatment: Radiation .
   Yes 0 (.) NC 0 (.) NC 0 (.) NC 0 (.) NC 0 (.) NC
   No 17 (65.4) 38 (71.7) 66 (77.6) 55 (69.6) 69 (81.2)
First opinion provider included a general surgeon
   Yes 8 (80) 0.399 15 (78.9) 0.381 26 (81.3) 0.536 23 (79.3) 0.154 28 (87.5) 0.247
   No 9 (56.3) 23 (67.6) 40 (75.5) 32 (64.0) 41 (77.4)
First opinion provider included a subspecialty surgeon
Yes 5 (62.5) 1.0 11 (64.7) 0.520 20 (71.4) 0.335 16 (64.0) 0.46 20 (71.4) 0.107
No 12 (66.7) 27 (75) 39 (72.2) 49 (86)
First opinion provider included a medical oncologist
Yes 5 (50) 0.234 11 (73.3) 1.0 19 (70.4) 0.272 16 (64.0) 0.46 0 20 (74.1) 0.253
No 12 (75) 27 (71.1) 47 (81.0) 39 (72.2) 49 (84.5)
First opinion provider included a radiation oncologist ‘
Yes 0 (.) NC 2 (100) 1.0 2 (100.0) 1.0 2 (100.0) 1.0 2 (100) 1.0
No 17 (65.4) 36 (70.6) 64 (77.1) 53 (68.8) 67 (80.7)
Prior visit at our institution before second opinion visit
Yes 7 (58.3) 0.683 20 (76.9) 0.407 32 (76.2) 0.75 27 (71.1) 0.79 34 (81) 0.958
No 10 (71.4) 18 (66.7) 34 (79.1) 28 (68.3) 35 (81.4)
Seeking 3rd opinion or more
Yes 9 (64.3) 1.0 7 (77.8) 1.0 17 (73.9) 0.615 15 (71.4) 0.833 50 (80.6) 1.0
No 8 (66.7) 31 (70.5) 49 (79.0) 40 (69.0) 19 (82.6)
Second opinion provider included a Surgery
Yes 14 (63.6) 1 36 (78.3) 0.015 59 (79.7) 0.255 50 (73.5) 0.08 61 (82.4) 0.4 26
No 3 (75) 2 (28.6) 7 (63.6) 5 (45.5) 8 (72.7)
Second opinion provider included a Radiation Oncologist
Yes 12 (66.7) 1.0 24 (82.8) 0.049 45 (88.2) 0.004 36 (76.6) 0.102 46 (90.2) 0.009
No 5 (62.5) 14 (58.3) 21 (61.8) 19 (59.4) 23 (67.6)
Second opinion provider included a Plastic Surgeon
Yes 1 (100.0) 0.411 14 (87.5) 1.0 15 (83.3) 0.020 15 (88.2) 0.042 16 (88.9) 0.257
No 38 (40.4) 95 (83.3) 120 (55.0) 133 (63.9) 157 (73.4)
First and second opinion treatment plans matched
Yes 5 (50) 0.417 24 (70.6) 1.0 31 (70.5) 0.178 29 (65.9) 0.712 34 (77.3) 0.500
No 9 (69.2) 12 (70.6) 30 (83.3) 21 (70.0) 30 (83.3)
Open in a new tab

P-values based on Chi-Square tests or Fisher Exact tests (where >=25% of cells having an expected count <5) for independence. NC: Not calculated due to having only a single row in the crosstabulation.

Comparison of Breast, Colon, and Pancreatic Cancer

Across all cancers examined, a different treatment was offered to 55.3% (131) of the patients. Patients with pancreatic cancer were more likely to have Medicare as compared to breast and colorectal. (40.7% versus 21.2% versus 28.2%, P=0.011). Surgical therapy was offered to 54.9% (130) of patients at the second opinion appointment and of these ultimately 83.8% (109) of patients were retained (Table 5). Breast cancer was less likely to be retained for surgery compared to colon and pancreatic (71.7% versus 90% versus 93.6%, P=0.009). Systemic therapy was offered to 40.1% (95) of patients and 41.1% (39) of those patients were retained for treatment. Overall, 62.4% (148) of patients were retained for the treatment recommended by the second opinion. Patients that had pancreatic cancer were less likely to be retained for the treatment recommended compared to breast and colorectal (53.2% versus 69.6% versus 76.1%, P=0.01)(Figure 2).

Table 5,

Second opinion characteristics

Parameter Breast, N=85 N (%) Pancreas, N=81 N (%) Colorectal, N=71 N (%)
Three of more opinions 23 (27.1%) 17 (21%) 13 (18.3%)
Radiation oncologist at second opinion 51 (60.0%) 4 (4.9%) 9 (12.7%)
Treatment recommendations matched first opinion 44 (55%) 41 (52.6%) 21 (31.3%)
More treatment offered at second opinion 21 (58.3%) 23 (62.2%) 47 (66.2%)
Retained for surgery 38 (71.7%) 27 (90%) 44 (93.6%)
Retained for systemic or radiation therapy 17 (65.4%) 15 (30.6%) 7 (35.0%)
Retained for surveillance 66 (77.6%) 34 (42.5%) 35 (49.3%)
Retained for any treatment 69 (81.2%) 49 (62.8%) 55 (79.7%)
Open in a new tab

Figure 2.

Figure 2.

Open in a new tab

Rates of retention for care after second opinion visit by cancer type.

Discussion:

Studying three cancer groups allowed us to both compare the characteristics associated with second opinion retention rates between cancer types and within each cancer cohort.

Pancreatic Cancer

In patients with pancreatic cancer, the retention rates were affected by cancer stage, CCI, and treatment plan offered. Cancer stage was associated with retention at the referred facility in patients with pancreatic cancer. Patients with non-metastatic disease were more likely to be retained at the second opinion facility for both surgery (P=0.05) and the remainder of their care (P=0.018). Stage IV pancreatic adenocarcinoma patients, however, were less likely to be retained. The lack of retention of these patients may be related to the fact that less surgical intervention was offered for these patients and the treatments offered were typically similar to their initial providers treatment plan.

Comorbidity scores were also correlated with retention rates. Higher CCI scores in our pancreatic cancer cohort were associated with lower rates of retention for all treatments. Patients with more comorbidities may not tolerate treatments as well as their peers and may be at higher risk for treatment-related complications.12 Metastatic disease increases CCI score compared to locoregional disease, which may also be contributing these patients’ decisions.13

A change in treatment plan affected retention rates. Those pancreatic cancer patients who were given a different treatment plan were more likely to stay with the secondary physician for systemic therapy. Prior studies by Mellink et al. and Payne et al. showed that 68% of patients hoped for a difference between their first and second opinions and their diagnosis.14, 15 Mellink et al. later found that 32% of patients seeking a second opinion had a discrepancy between the different opinions they received.6 This study also found that half of those with a discrepant plan led to major changes in treatment.6 By comparison, in our study, a new treatment plan was offered to 46.8% pancreatic cancer patients. Of those patients offered a new systemic treatment option, 58% of them were retained by the second opinion physician. When the same opinion was given, only 23.3% of patients were retained. With no change in treatment plan, patients may prefer to pursue treatment closer to home. There are also potential financial considerations. Most patients were referred from the community setting and systemic therapy is typically less expensive in the community setting.9

Previous studies have shown that age is a factor in a patient being referred for or seeking a second opinion. De Roo et al. found that those seeking second opinions were of younger age.16 Groß et al. demonstrated that patients were more likely to be told about a second opinion option if they were between the ages of 18 and 75.17 The median age in our population was 59.8 (range, 18–89) years old. Benbassett determined that doctors are less inclined to inform older patients about the possibility of seeking a second opinion.18 However, our study did not show any difference in the age at diagnosis and the retention rates of patients with pancreatic cancer.

Colon Cancer

Colorectal cancer patient retention rate at their second opinion was associated with the treatment plan offered. At the second opinion, colorectal patients were more likely than those with pancreatic or breast cancer to be offered a different treatment plan. If offered surgery, most colorectal cancer patients were retained for surgery (93.6%), whereas a much smaller percentage (35%) were retained for systemic treatment when offered this treatment modality. This might be explained by the more advanced, and potentially curative, surgical options offered at an academic center such as resection of metastatic lesions to the liver and the financial benefits of systemic therapy in the community.9, 19 There is also a logistical benefit of being closer to the patients’ home.

Stage of colorectal cancer was not associated with retention rate in our analysis. However, a prior study by DeRoo et al. found that patients with colorectal cancer seeking second-opinion were more likely to have stage IV disease or recurrent disease, with 35% of them transitioning their care to a colorectal cancer MDC.16 Of those second opinion patients retained at DeRoo’s institution, 89% had either Stage IV or recurrent disease.16 This is in contrast to our results; stage did not play a part in retention of second opinion colorectal cancer patients, but stage IV pancreatic cancer was associated with lower second opinion retention. There exists a fundamental difference in the expectations with surgery between these cancers. Curative treatment is still pursued for stage IV colorectal cancer, however, for pancreatic cancer, surgery is typically only pursued for palliative purposes.20, 21 Similar retention rates were found in the colorectal (90%) and pancreatic cancer (93.6%) groups when surgery was offered by the second opinion.

Breast Cancer

Characteristics that impacted second opinion retention rates of breast cancer patients included treatment providers involved in the MDC and ethnicity. Plastic surgery and radiation oncology involvement in the MDC were significant components in the retention of patients for surgery. The availability of plastic surgery consult on the same day as the second opinion was a positive factor in retaining breast cancer patients by a second opinion physician. Our study found that 88.2% of patients were retained for the treatment offered at our institution if a plastic surgeon was involved at the same time as the second opinion MDC appointment. These patients may initiate a second opinion with the forethought of plastic surgeon involvement and request for appointments on the same day, making retention much higher in this population. Reconstruction may be an important factor for breast cancer patients and consultation with a plastic surgeon would allow them to make a more informed decision. It is important for this population to receive appropriate education about the effects of their wellbeing post-operatively. This may also be because patients can schedule surgery in a timely fashion. This would be especially pertinent for those that traveled farther distances. Patients may not have heard in depth discussions about their reconstructive options at their initial appointment and likely interpret the involvement of a plastic surgeon as adding to the completeness of this education, leading to the retention for surgery at higher rates. A confounding factor, however, is that our team is the only one in our state with an academic plastic surgery institution and one of few that offers more complex autologous reconstruction including deep inferior epigastric perforator flaps, which may limit generalizability to other breast surgical oncology practices.

More patients were also retained for any treatment and surveillance when they saw a radiation oncologist in the MDC. This might be explained by poor access to radiation oncology in the community setting.22 Considering the high cost of radiation oncology equipment and staff salaries, this treatment may not be readily available in communities with fewer resources.22 Women living far from a community center were found to be more likely to be nonadherent with radiation oncology therapy.23 A second opinion which allows access to an academic center with a radiology oncologist lends the possibility of catering to an otherwise underserved population.

Ethnicity also played a role in retention of breast cancer patients. Non-Hispanic patients were more likely to be retained for any treatment and for surveillance in this study. Importantly, there may be differences in how frequently patients of different ethnicities choose to pursue second opinions. In three prior studies, two found that non-white breast cancer patients were less likely to seek second opinions, whereas one found that Hispanics sought second opinions more often than their white counterparts.2, 24 There has been a call to improve cultural and linguistic communication and education, in hopes of decreasing cancer risk and improving patient knowledge among Hispanic women.24 This approach may help avoid health care inequalities and may also lead to an increase in retention rates in this population.

Limitations include the retrospective nature of this study. Medical histories reviewed were not always explicit in categorizing or identifying patients as second opinions. This likely led to a smaller sample size as not all second opinions might have been identified during the study period. While the total number of patients evaluated made for a good sample size, each individual cancer cohort sample size was small. There may also be influences from individual surgeons. The second opinion providers may have better communication with their patients. Recency bias may also play a psychological role in influencing patient decisions. Future studies should investigate this phenomenon and consider individual variables such as provider communication styles and patient psychological variables. Further, as our institution is the only NCI cancer center in a multi-state area, our findings may be less generalizable to other institutions.

Conclusion

Second opinion evaluations have useful implications for patient care. They are an important way to educate patients about their disease process and their treatment options. This study’s information allows for delivery of pertinent resources to individual patients. Surgeons can better identify patients that are more likely to stay or not by looking at the characteristics of each individual patient. Patients with stage IV pancreatic cancer are less likely to stay if there is no new treatment offered and focusing these visits on education and expectations may be more useful. Patients with colon cancer are more likely to stay with the second opinion physician when a different treatment plan is offered. Patients with breast cancer were more likely to be retained if they had seen a plastic surgeon on the same day as their MDC visit. Hospitals with the resources to do so should consider the addition of a plastic surgeon into their breast cancer MDC to retain more patients. Understanding the treatment factors that lead to retention for these three cancer types may allow first or second opinion physicians and administrators to provide the best possible resources for patients presenting for first or second opinion evaluations.

Figure 1.

Figure 1.

Open in a new tab

STROBE Diagram

Funding/Support:

This work was supported in part by the University of Colorado Cancer Center’s Shared Resource funded by NCI grant P30CA046934, Grant 2020141 from the Doris Duke Charitable Foundation and University of Colorado School of Medicine, and the NCI Paul Calabresi K12 award.

Thank you to Christopher Quinn, BA, MS for help with statistical support.

Footnotes

All contributing authors meet the qualification for authorship. Alec McCranie supplied substantial contributions to the design and idea of the project and wrote the majority of the manuscript. Laura Leonard, Sudheer Vemuru, Victoria Huynh, and Sarah Tevis supplied a substantial contribution to the original idea of the project and design as well as contributed heavily to the drafting and editing of the manuscript. Jonathan D’Angelo, Gretchen Ahrendt, Christodoulos Kaoutzanis, Anne-Lise D’Angelo, Christine Fisher, Karina, Ana Gleisner, Cheryl Meguid, Richard Schulick, and Marco Del Chiaro helped substantially with editing and drafting of the manuscript as well as proof reading and approving the manuscript.

Conflict of Interest Statement:

Marco Del Chiaro’s conflicts of interest:

I have been awarded an industry grant (Haemonetics, Inc) to conduct a multi-center study to evaluate the prognostic implications of TEG in pancreas cancer.

I am a co-PI of a sponsored Boston Scientific study on the use of intra-operative pancreatoscopy in IPMN’s patients.

All other authors have no related conflicts of interest to declare.

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.

References

  • 1.Okamoto S, Kawahara K, Okawa A, Tanaka Y. Values and risks of second opinion in Japan’s universal health-care system. Health Expectations. 2015;18(5):826–38. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 2.Hillen MA, Medendorp NM, Daams JG, Smets EMA. Patient-Driven Second Opinions in Oncology: A Systematic Review. Oncologist. 2017. Oct;22(10):1197–211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 3.Tattersall MHN, Dear RF, Jansen J, Shepherd HL, J Devine R, G Horvath L, et al. Second opinions in oncology: the experiences of patients attending the Sydney Cancer Centre. Medical Journal of Australia. 2009;191(4):209–12. [DOI] [PubMed] [Google Scholar]
  • 4.Philip J, Gold M, Schwarz M, Komesaroff P. Second medical opinions: the views of oncology patients and their physicians. Supportive Care in Cancer. 2010. 2010/09/01;18(9):1199–205. [DOI] [PubMed] [Google Scholar]
  • 5.Cecon N, Hillen MA, Pfaff H, Dresen A, Groß SE. Why do newly diagnosed breast cancer patients seek a second opinion? - Second opinion seeking and its association with the physician-patient relationship. Patient Educ Couns. 2019. May;102(5):998–1004. [DOI] [PubMed] [Google Scholar]
  • 6.Mellink WA, Henzen-Logmans SC, Bongaerts AH, Ooijen BV, Rodenburg CJ, Wiggers TH. Discrepancy between second and first opinion in surgical oncological patients. Eur J Surg Oncol. 2006. Feb;32(1):108–12. [DOI] [PubMed] [Google Scholar]
  • 7.Greenfield G, Pliskin JS, Feder-Bubis P, Wientroub S, Davidovitch N. Patient–physician relationships in second opinion encounters – The physicians’ perspective. Social Science & Medicine. 2012. 2012/10/01/;75(7):1202–12. [DOI] [PubMed] [Google Scholar]
  • 8.Clauson J, Hsieh YC, Acharya S, Rademaker AW, Morrow M. Results of the Lynn Sage Second-Opinion Program for local therapy in patients with breast carcinoma. Changes in management and determinants of where care is delivered. Cancer. 2002. Feb 15;94(4):889–94. [DOI] [PubMed] [Google Scholar]
  • 9.Hayes J, Hoverman RJ, Brow ME, Dilbeck DC, Verrilli DK, Garey J, et al. Cost differential by site of service for cancer patients receiving chemotherapy. Am J Manag Care. 2015. Mar 1;21(3):e189–96. [PubMed] [Google Scholar]
  • 10.Sag AA, Selcukbiricik F, Mandel NM. Evidence-based medical oncology and interventional radiology paradigms for liver-dominant colorectal cancer metastases. World J Gastroenterol. 2016. Mar 21;22(11):3127–49. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 11.Unger JM, Vaidya R, Hershman DL, Minasian LM, Fleury ME. Systematic Review and Meta-Analysis of the Magnitude of Structural, Clinical, and Physician and Patient Barriers to Cancer Clinical Trial Participation. J Natl Cancer Inst. 2019. Mar 1;111(3):245–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 12.Sarfati D, Koczwara B, Jackson C. The impact of comorbidity on cancer and its treatment. CA Cancer J Clin. 2016. Jul;66(4):337–50. [DOI] [PubMed] [Google Scholar]
  • 13.Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis. 1987;40(5):373–83. [DOI] [PubMed] [Google Scholar]
  • 14.Mellink WA, Dulmen AM, Wiggers T, Spreeuwenberg PM, Eggermont AM, Bensing JM. Cancer patients seeking a second surgical opinion: results of a study on motives, needs, and expectations. J Clin Oncol. 2003. Apr 15;21(8):1492–7. [DOI] [PubMed] [Google Scholar]
  • 15.Payne VL, Singh H, Meyer AN, Levy L, Harrison D, Graber ML. Patient-initiated second opinions: systematic review of characteristics and impact on diagnosis, treatment, and satisfaction. Mayo Clin Proc. 2014. May;89(5):687–96. [DOI] [PubMed] [Google Scholar]
  • 16.De Roo AC, Morris AM, Vu JV, Schuman AD, Abbott KL, Kandagatla P, et al. Characteristics of Patients Seeking Second Opinions at a Multidisciplinary Colorectal Cancer Clinic. Dis Colon Rectum. 2020. Jun;63(6):788–95. [DOI] [PubMed] [Google Scholar]
  • 17.Groß SE, Hillen MA, Pfaff H, Scholten N. Second opinion in medical encounters − A study among breast cancer patients. Patient Education and Counseling. 2017;100(11):1990–5. [DOI] [PubMed] [Google Scholar]
  • 18.Benbassat J. Obtaining a second opinion is a neglected source of health care inequalities. Isr J Health Policy Res. 2019. Jan 16;8(1):12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 19.Pawlik TM, Choti MA. Surgical therapy for colorectal metastases to the liver. J Gastrointest Surg. 2007. Aug;11(8):1057–77. [DOI] [PubMed] [Google Scholar]
  • 20.Adson MA, van Heerden JA, Adson MH, Wagner JS, Ilstrup DM. Resection of Hepatic Metastases From Colorectal Cancer. Archives of Surgery. 1984;119(6):647–51. [DOI] [PubMed] [Google Scholar]
  • 21.Leonard GD, Brenner B, Kemeny NE. Neoadjuvant Chemotherapy Before Liver Resection for Patients With Unresectable Liver Metastases From Colorectal Carcinoma. Journal of Clinical Oncology. 2005;23(9):2038–48. [DOI] [PubMed] [Google Scholar]
  • 22.Mattes MD, Suneja G, Haffty BG, Takita C, Katz MS, Ohri N, et al. Overcoming Barriers to Radiation Oncology Access in Low-Resource Settings in the United States. Adv Radiat Oncol. 2021. Nov-Dec;6(6):100802. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 23.Sapienza LG, Thomas JJ, Echeverria AE, Sharma S, Hamstra DA, Hall TR, et al. Predictors of follow-up non-compliance after definitive radiotherapy for locally advanced cervical cancer at a community cancer center. Gynecol Oncol Rep. 2022. Dec;44:101091. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 24.Parikh AR, Kaplan CP, Burke NJ, Livaudais-Toman J, Hwang ES, Karliner LS. Ductal carcinoma in situ: knowledge of associated risks and prognosis among Latina and non-Latina white women. Breast Cancer Res Treat. 2013. Sep;141(2):261–8. [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES