Abstract
Purpose.
Disparities in access to multidisciplinary cancer consultations (MDCc) persist, and the role of physician relationships remains understudied. This study examined the extent to which multilevel factors, including patient characteristics and patient-sharing network measures reflecting the structure of physician relationships, are associated with MDCc and receipt of stereotactic body radiation therapy (SBRT) versus surgery among early-stage non-small cell lung cancer (NSCLC) patients.
Materials and Methods.
In this cross-sectional study, we analyzed Surveillance, Epidemiology, and End-Result (SEER)-Medicare data for patients diagnosed with stage I-IIA NSCLC in 2016-2017. We assembled patient-sharing networks and identified cancer specialists who were locally unique for their specialty, herein referred to as “linchpins”. The proportion of linchpin cancer specialists for each hospital referral region (HRR) was calculated as a network-based measure of specialist scarcity. We used multilevel multinomial logistic regression to estimate associations between study variables and the receipt of MDCc and multilevel logistic regression to examine the relationship between MDCc and patient’s first treatment.
Results.
Our study included 6,120 patients with stage I-IIA NSCLC, of which 751 (12.3%) received MDCc, 1,729 (28.3%) only consulted a radiation oncologist, 2,010 (32.8%) only consulted a surgeon, and 1,630 (26.6%) had no consultations with either specialist within two months following diagnosis. Compared with patients residing in an HRR with a low proportion of linchpin surgeons, those in an HRR with a high proportion of linchpin surgeons had a 2.99 (95% CI: 1.87-4.78) greater relative risk of exclusively consulting a radiation oncologist (vs. MDCc) and a 2.70 (95% CI: 1.68-4.35) greater relative risk consulting neither specialist (vs. MDCc). Patients who received MDCc were 5.32 (95% CI: 4.27-6.63) times more likely to receive SBRT (vs. surgery).
Conclusions.
Physician networks are associated with receipt of MDCc and treatment, underscoring the potential for leveraging patient-sharing network analysis to improve access to lung cancer care.
INTRODUCTION
A multidisciplinary healthcare team refers to an integrated team approach to healthcare in which health care providers consider all relevant treatment options and collaboratively develop an individual treatment plan for each patient. In oncology, multidisciplinary cancer care involves the provision of care from a team of multidisciplinary medical experts, such as medical oncologists, surgeons, radiation oncologists, and palliative care experts, among others.1 Patients with non-small cell lung cancer (NSCLC) often receive care from multidisciplinary healthcare teams during the course of their treatment. Multidisciplinary lung cancer care is associated with improved care quality, including increased receipt of guideline-concordant treatment, enhanced diagnostic accuracy and completeness of preoperative staging, utilization of advanced surgical techniques, and improved timeliness of care.2–6
Treatment of early-stage (stage I and II) NSCLC relies primarily on surgical resection.7,8 However, for inoperable tumors or high-risk patients with low-performance status and severe comorbidity, stereotactic body radiotherapy (SBRT) is the preferred treatment option.9,10 To date, given the lack of level-one evidence, there remains uncertainty as to the optimal management of medically operable early stage lung cancer patients. Patients receiving SBRT often exhibit poorer health and increased medical complexity, creating challenges in comparing the effectiveness of these two treatments due to patient characteristics associated with the selection of each treatment option. Although some studies with a limited sample size have suggested comparable outcomes between the two treatments, evidence from randomized controlled trials is lacking due to challenges in patient accrual.11–15 Given the difficulty in determining the most appropriate treatment among patients who are not obvious candidates for surgery, there is a growing emphasis on the role of multidisciplinary cancer consultations (MDCc) in treatment planning.16 These consultations foster collaboration among healthcare professionals, enabling them to collectively determine the most appropriate treatment approach for each patient, taking into account their unique needs and circumstances. Recent work showing disparities in receipt of MDCc by race, ethnicity, rurality, and socioeconomic status highlight the need to remove barriers to MDCc to ensure equitable access to high quality cancer care.17,18
We posited that features of patient-sharing physician networks would be useful measures to study the associations between physicians relationships and patient receipt of MDCc and treatment. To characterize these networks, we developed a novel measure that identifies physicians who are locally unique for their specialty, termed “linchpin score”.19 For example, a radiation oncologist’s linchpin score assesses the extent to which their non-radiation oncologist patient-sharing peers share patients with other radiation oncologists. An advantage to our network-based approach for measuring access to specialists is that it does not require partitioning patients and physicians into small areas, but rather reflects all cancer patient-sharing relationships. This may better reflect referral patterns for lung cancer care which often span counties, cities or even states. Patients residing in regions with a high proportion of linchpin oncologists are hypothesized to have greater barriers to specialty access due to less redundancy in patient-sharing ties to specialists. Prior work demonstrated that hospital referral regions with a high proportion of linchpin oncologists tracked with socioeconomic disadvantage and lower rates of radiation therapy utilization.20 Furthermore, patients with lung and colorectal cancer who are treated by linchpin oncologists often have worse survival.21
The objective of this study was to investigate potential disparities in the receipt of MDCc among early-stage NSCLC patients and to examine the extent to which patient-sharing networks impact access to MDCc and treatment. Using Surveillance, Epidemiology, and End-Result (SEER)-Medicare linked data for patients diagnosed with NSCLC in 2016-2017, we assembled a patient-sharing network based on all NSCLC care. We then created a cohort of patients diagnosed with stage I-IIA NSCLC, representing those who are potential candidates for surgery or radiation treatment, to examine multilevel factors associated with receiving MDCc and subsequent treatment. We hypothesized that patients residing in regions characterized by cancer specialist scarcity, defined using our measure of linchpin score, would be less likely to receive multidisciplinary consultations, which could then have implications for the treatments the patients received.
METHODS
Data source and study population
We used the Surveillance, Epidemiology, and End Results (SEER)-Medicare linked database for patient diagnosed with NSCLC in 2016 and 2017. SEER compiles and publishes cancer incidence and survival data from population-based cancer registries, representing about 48.0% of the U.S. population. The diagnostic confirmation of cancer for SEER patients in the NSCLC cohort indicate that histologically/cytologically confirmed cancers represent >99% of cases and clinically-diagnosed cancers represent <1% of cases. SEER is linked with the Medicare database that contains claims for covered health care services provided to both inpatients and outpatients and includes information on diagnoses and billed services for individuals from the time they become eligible for Medicare until the time of their death. We excluded patients who were <66 or >99 at the time of diagnosis. Exclusion of patients <66 years was to allow each patient a full 12 months of claims in the look-back period. We excluded patients who had a missing or non-US residential ZIP code, as ZIP code was used to assign patients to a hospital referral region. We also excluded patients who or were not continuously enrolled in fee-for-service Medicare in the 12 months prior to diagnosis and until the sooner of death or 12 months following diagnosis because we may be missing relevant encounters (Supplement Figure 1). Our analytic cohort included patients with AJCC stage IA, IB, and IIA cancers (7th edition). Additionally, because we examined the effect of the proportion of linchpin oncologists within a patient’s hospital referral region (HRR) on study outcomes, patients residing in an HRR without any surgeons or radiation oncologists in the network were excluded from the study. The study was reviewed and determined exempt by the institutional review board at Dartmouth College.
Outcome variables
Cancer consultation
The outcomes of interest are the types of cancer consultations that patients in our cohort received within two months following their initial cancer diagnosis. The four types of consultations we examined were: 1) receiving MDCc, defined as consulting both a radiation oncologist and a surgeon, 2) consulting only a radiation oncologist, 3) consulting only a surgeon, and 4) not receiving a consultation from either a radiation oncologist or a surgeon. In identifying encounters with surgeons, we included thoracic surgeons and general surgeons. In the absence of specific guidelines for MDCc, we specified a 2-month time frame for the look-forward window to be consistent with prior relevant work.18 Additionally, we considered a 4-month time frame as a look-forward window in sensitivity analysis.
Treatment received
The second outcome variable for this study is the initial treatment received by stage IA-IB NSCLC patients, which was limited to either SBRT or surgery. Treatments were identified based on CPT and HCPCS codes defined by previous studies (Supplemental Table 1). Patients who did not undergo either SBRT or surgery within 12 months of their diagnosis were excluded from this analysis.
Study variables
Patient-level characteristics
Patient age in years at diagnosis, sex, race, ethnicity, AJCC stage, Census tract level poverty, county-level rurality, and SEER region were obtained from the SEER Cancer File. Charlson comorbidity index and presence of chronic obstructive pulmonary disease (COPD) were calculated using the diagnosis codes associated with the 12 months of claims preceding the month of cancer diagnosis.22 To assess COPD severity, we used the ICD-10 code J44.1 to identify patients who experienced a COPD exacerbation within the 12 months preceding their month of cancer diagnosis. Poverty was measured using the Yost index, which was developed using United States-based quintiles of a composite socioeconomic status score from Census tract-level American Community Survey (ACS) 5-year estimates.23 The analysis also included the receipt of MDCc as a covariate, a binary variable indicating whether a patient received MDCc within 2 months following diagnosis.
Patient-sharing network analysis
We identified the providers who had encounters with all NSCLC patients (stages I-IV) in 3 months prior to and 12 months following their cancer diagnosis in the Medicare Carrier files. From these encounters, providers were connected if they had encounters with common patients to form a “patient-sharing network”. In this network, the relationships, or “edges”, between providers were quantified by the number of shared patients. All providers who had at least one connection with another provider were included in the patient-sharing network. We were inclusive of all NSCLC patients and all providers to capture the full patient-sharing network of NSCLC care. Provider specialty was identified from the Medicare Data on Provider Practice and Specialty file.
Linchpin score and proportion of linchpin oncologists
Linchpin score was calculated for each surgeon and radiation oncologist based on all of their ties in the lung cancer patient-sharing networks. In Figure 1, we demonstrate the linchpin score calculation for radiation oncologist i (), by summing edges with peers who lack ties to other radiation oncologists and dividing by the sum of all shared ties.19 In this example, radiation oncologist i shares patients with four other physicians ( and the values along each edge represent the number of shared patients. Only one of those physicians has an established tie with another radiation oncologist. Linchpin score is calculated by summing the edges that radiation oncologist i has with physicians who are not connected to another radiation oncologist (in this example, and dividing by the sum of all edges, resulting in a score that ranges from 0 to 1. We considered a physician to be a linchpin if the linchpin score was in the top 15% of the distribution of linchpin scores for their specialty based on prior work demonstrating meaningful associations with outcomes using this threshold.20,21
Figure 1.
Illustration of linchpin score calculation. Numbers adjacent to edge lines represent shared patients.
Next, we calculated the proportion of linchpin radiation oncologists and surgeons within each HRR. Physicians were assigned to an HRR based on the plurality of where their claims were located. The proportions of linchpin surgeons and radiation oncologists were classified into low, medium, and high tertiles. Patients were assigned to an HRR based on their residential ZIP code.
Statistical Analyses
We used Pearson Chi-square tests to describe differences in patient characteristics across the types of cancer consultations. We then performed mixed effect multinomial logistic regression with a random intercept for HRR to examine the associations between the patient characteristics and the likelihood of receiving different types of cancer consultations after diagnosis. The outcome in this analysis is the consultation type, which includes four possible outcomes: consulted with both a surgeon and a radiation oncologist (received MDCc), consulted with just a surgeon, consulted with just a radiation oncologist, or consulted with neither. To assess whether the choice or membership in one category is independent of the choice or membership in another category (i.e., the dependent variable), we tested the independence of irrelevant alternatives using Hausman-McFadden test to avoid inconsistent parameter estimates due to non-compliance (Supplemental Table 2).24 The results were reported in terms of the relative risk ratio (RRR) and their corresponding p-values while controlling for all other covariates considered in the study. Statistical significance was determined at a threshold of p-value < 0.01.
To examine the association between patient characteristics, receipt of MDCc, and the first treatment received, we conducted a mixed effect logistic regression with a random intercept for HRR and adjusted for study covarites. Our analyses included a random intercept for HRR to allow for the modeling of variation in the outcome variable that was attributable to geographic differences between the HRRs, which could have an impact on the consultation and treatment decisions of patients. All statistical analyses were performed with R version 4.2.2.
RESULTS
Our study included 6,120 patients with stage I-IIA NSCLC diagnosis in 2016-2017 within 16 SEER regions (Table 1). The majority of cohort patients were White (89.9%), non-Hispanic (96.6%), and resided in an urban area (82.8%). Most patients in the cohort were diagnosed with stage IA NSCLC (59.2%), followed by stage IB (27.5%) and stage IIA (13.3%). Over half of the patients (55.4%) had a diagnosis of COPD, and a substantial proportion of the study cohort had one (27.6%) or two (54.1%) comorbidities. We found that 751 (12.3%) received an MDCc, 1,729 (28.3%) only consulted a radiation oncologist, 2,010 (32.8%) only consulted a surgeon, and 1,630 (26.6%) had no consultations with either specialist within the 2 months following diagnosis. Of those patients with an encounter with a surgeon, 1,601 (58%) saw a thoracic surgeon and 1,160 (42%) saw a general surgeon. In bivariate analyses, we observed significant differences (p < 0.01) for age at diagnosis, race, AJCC cancer stage, presence of COPD with or without exacerbations, number of comorbidities, and SES among patients who received different types of cancer consultations. Sensitivity analysis using a 4-month window is presented in Supplemental Table 3.
Table 1.
Characteristics of early-stage NSCLC patients stratified by type of cancer consultation receipt following diagnosis.
| Received MDCc | Only Radiation Oncologist | Only Surgeon | Neither | P-Value | |
|---|---|---|---|---|---|
| Characteristics, N(%) | (N=751) | (N=1729) | (N=2010) | (N=1630) | |
| Male | 390 (51.9%) | 923 (53.4%) | 1063 (52.9%) | 872 (53.5%) | 0.89 |
| Age at diagnosis (years) | |||||
| 66-69 | 129 (17.2%) | 191 (11.0%) | 481 (23.9%) | 281 (17.2%) | <0.01 |
| 70-74 | 202 (26.9%) | 383 (22.2%) | 654 (32.5%) | 496 (30.4%) | |
| 75-79 | 190 (25.3%) | 390 (22.6%) | 524 (26.1%) | 402 (24.7%) | |
| 80-84 | 143 (19.0%) | 409 (23.7%) | 253 (12.6%) | 259 (15.9%) | |
| 85+ | 87 (11.6%) | 356 (20.6%) | 98 (4.9%) | 192 (11.8%) | |
| Race | |||||
| White | 691 (92.0%) | 1569 (90.7%) | 1817 (90.4%) | 1422 (87.2%) | <0.01 |
| Black | 43 (5.7%) | 113 (6.5%) | 106 (5.3%) | 140 (8.6%) | |
| Other | 17 (2.3%) | 47 (2.7%) | 87 (4.3%) | 68 (4.2%) | |
| Hispanic | 24 (3.2%) | 43 (2.5%) | 66 (3.3%) | 74 (4.5%) | 0.01 |
| Cancer Stage | |||||
| IA | 422 (56.2%) | 1101 (63.7%) | 1144 (56.9%) | 953 (58.5%) | <0.01 |
| IB | 193 (25.7%) | 426 (24.6%) | 614 (30.5%) | 453 (27.8%) | |
| IIA | 136 (18.1%) | 202 (11.7%) | 252 (12.5%) | 224 (13.7%) | |
| COPD | |||||
| No COPD | 298 (39.7%) | 557 (32.2%) | 1030 (51.2%) | 706 (43.3%) | <0.01 |
| COPD | 431 (57.4%) | 1070 (61.9%) | 903 (44.9%) | 737 (45.2%) | |
| COPD w/ exacerbation | 22 (2.9%) | 102 (5.9%) | 77 (3.8%) | 187 (11.5%) | |
| Charlson Comorbidities | |||||
| 0 | 114 (15.2%) | 183 (10.6%) | 498 (24.8%) | 325 (19.9%) | <0.01 |
| 1 | 211 (28.1%) | 448 (25.9%) | 593 (29.5%) | 436 (26.7%) | |
| 2+ | 426 (56.7%) | 1098 (63.5%) | 919 (45.7%) | 869 (53.3%) | |
| Rurality | |||||
| Metropolitan | 634 (84.4%) | 1422 (82.2%) | 1691 (84.1%) | 1322 (81.1%) | 0.06 |
| Non-Metropolitan | 117 (15.6%) | 307 (17.8%) | 319 (15.9%) | 308 (18.9%) | |
| Yost Quantile | |||||
| 1(lowest SES) | 96 (12.8%) | 275 (15.9%) | 254 (12.6%) | 347 (21.3%) | <0.01 |
| 2 | 127 (16.9%) | 341 (19.7%) | 335 (16.7%) | 284 (17.4%) | |
| 3 | 158 (21.0%) | 335 (19.4%) | 372 (18.5%) | 327 (20.1%) | |
| 4 | 173 (23.0%) | 390 (22.6%) | 450 (22.4%) | 329 (20.2%) | |
| 5(highest SES) | 197 (26.2%) | 388 (22.4%) | 599 (29.8%) | 343 (21.0%) |
Note. MDCc, multidisciplinary cancer consultation; COPD, chronic obstructive pulmonary disease; SES, socioeconomic status; Values might not add up to 100% due to rounding.
Factors Associated with Receipt of MDCc
The associations between types of cancer consultation and demographic variables, clinical characteristics, and the proportion of linchpin cancer specialists within the patient’s HRR are presented in Table 2. Compared with patients who received MDCc, patients who only consulted a radiation oncologist were older (85 years or above [vs. 66-69 years] RRR: 3.26; 95% CI: 2.33-4.56; p < 0.01), less likely to be stage IIA compared with stage IA (RRR: 0.54; 95% CI: 0.42-0.69; p < 0.01), and more likely to have COPD (RRR: 1.34, 95% CI: 1.08-1.66; p = 0.01) or COPD with exacerbations (RRR: 2.23, 95% CI: 1.36-3.66, p<0.01) compared with no COPD.
Table 2.
Impact of NSCLC patient characteristics on receipts of consultations
| Only Radiation Oncologist vs Received MDCc | Only Surgeon vs Received MDCc | Neither vs Received MDCc | ||||
|---|---|---|---|---|---|---|
| RRR (95% CI) | P-value | RRR (95% CI) | P-value | RRR (95% CI) | P-value | |
| Female | 1.05 (0.88-1.26) | 0.57 | 0.99 (0.83-1.18) | 0.90 | 1.09 (0.91-1.30) | 0.35 |
| Age (years) | ||||||
| 66-69 | 1.0 (Referent) | 1.0 (Referent) | 1.0 (Referent) | |||
| 70-74 | 1.34 (1.01-1.78) | 0.04 | 0.89 (0.69-1.14) | 0.35 | 1.16 (0.89-1.52) | 0.28 |
| 75-79 | 1.48 (1.11-1.98) | 0.01 | 0.74 (0.57-0.96) | 0.02 | 1.02 (0.77-1.35) | 0.89 |
| 80-84 | 2.09 (1.55-2.83) | <0.01 | 0.46 (0.34-0.61) | <0.01 | 0.83 (0.62-1.12) | 0.23 |
| 85+ | 3.26 (2.33-4.56) | <0.01 | 0.26 (0.18-0.37) | <0.01 | 0.99 (0.71-1.40) | 0.97 |
| Race | ||||||
| White | 1.0 (Referent) | 1.0 (Referent) | 1.0 (Referent) | |||
| Black | 1.15 (0.78-1.68) | 0.48 | 0.85 (0.58-1.25) | 0.40 | 1.23 (0.84-1.79) | 0.29 |
| Other | 1.23 (0.69-2.2) | 0.48 | 1.85 (1.08-3.2) | 0.03 | 1.72 (0.98-2.99) | 0.06 |
| Hispanic | 0.73 (0.43-1.23) | 0.24 | 1.03 (0.63-1.68) | 0.92 | 1.28 (0.79-2.07) | 0.32 |
| Cancer Stage | ||||||
| IA | 1.0 (Referent) | 1.0 (Referent) | 1.0 (Referent) | |||
| IB | 0.79 (0.64-0.97) | 0.02 | 1.21 (0.99-1.48) | 0.06 | 0.95 (0.77-1.17) | 0.60 |
| IIA | 0.54 (0.42-0.69) | <0.01 | 0.68 (0.54-0.87) | <0.01 | 0.65 (0.51-0.84) | <0.01 |
| COPD | ||||||
| No COPD | 1.0 (Referent) | 1.0 (Referent) | 1.0 (Referent) | |||
| COPD | 1.34 (1.08-1.66) | 0.01 | 0.67 (0.54-0.83) | <0.01 | 0.78 (0.63-0.97) | 0.02 |
| COPD w/ Exacerbation | 2.23 (1.36-3.66) | <0.01 | 1.27 (0.76-2.10) | 0.36 | 3.87 (2.4-6.23) | <0.01 |
| Charlson comorbidity | ||||||
| 0 | 1.0 (Referent) | 1.0 (Referent) | 1.0 (Referent) | |||
| 1 | 1.13 (0.82-1.55) | 0.46 | 0.79 (0.59-1.06) | 0.11 | 0.80 (0.59-1.09) | 0.16 |
| 2+ | 1.26 (0.92-1.71) | 0.14 | 0.65 (0.49-0.86) | <0.01 | 0.78 (0.58-1.04) | 0.09 |
| Non-Metro | 0.97 (0.74-1.27) | 0.84 | 1.18 (0.9-1.55) | 0.23 | 0.95 (0.72-1.24) | 0.70 |
| Yost Quintile | ||||||
| 1 (lowest SES) | 1.0 (Referent) | 1.0 (Referent) | 1.0 (Referent) | |||
| 2 | 0.91 (0.66-1.26) | 0.58 | 1.01 (0.73-1.4) | 0.96 | 0.65 (0.47-0.9) | 0.01 |
| 3 | 0.73 (0.53-1.01) | 0.06 | 0.87 (0.63-1.21) | 0.41 | 0.62 (0.45-0.85) | <0.01 |
| 4 | 0.75 (0.54-1.05) | 0.09 | 0.94 (0.68-1.31) | 0.74 | 0.56 (0.40-0.77) | <0.01 |
| 5 (highest SES) | 0.66 (0.48-0.93) | 0.02 | 0.99 (0.71-1.39) | 0.98 | 0.5 (0.36-0.69) | <0.01 |
| Proportion Linchpin Surgeons | ||||||
| Low | 1.0 (Referent) | 1.0 (Referent) | 1.0 (Referent) | |||
| Medium | 1.45 (1.12-1.86) | <0.01 | 1.03 (0.79-1.35) | 0.82 | 1.54 (1.19-1.98) | <0.01 |
| High | 2.95 (1.82-4.78) | <0.01 | 1.33 (0.80-2.21) | 0.28 | 2.63 (1.62-4.28) | <0.01 |
| Proportion Linchpin Radiation Oncologists | ||||||
| Low | 1.0 (Referent) | 1.0 (Referent) | 1.0 (Referent) | |||
| Medium | 0.74 (0.57-0.98) | 0.04 | 1.12 (0.84-1.51) | 0.44 | 0.94 (0.71-1.25) | 0.69 |
| High | 0.79 (0.56-1.13) | 0.20 | 1.23 (0.85-1.79) | 0.27 | 1.09 (0.76-1.55) | 0.65 |
Note. RRR, relative risk ratio; MDCc, multidisciplinary cancer consultation; COPD, chronic obstructive pulmonary disease; SES, socioeconomic status.
Patients who exclusively consulted with a surgeon compared with those who received MDCc were younger (85 years or above [vs. 66-69 years] RRR: 0.26; 95% CI: 0.18-0.37; p < 0.001), less likely to be stage IIA compared with stage IA (RRR: 0.68; 95% CI: 0.54-0.87; p < 0.01) less likely to have COPD (RRR: 0.67; 95% CI: 0.54-0.83; p < 0.01), and less likely to have two or more comorbidities (RRR: 0.65; 95% CI: 0.49-0.86; p < 0.01). Finally, we found that patients’ risk of seeing neither specialist was inversely associated with SES quintile (highest SES quintile [vs lowest] RRR [CI]: 0.5 [0.36-0.69]).
Building on our prior work, we hypothesized that patient-sharing networks with a higher proportion of linchpin radiation oncologists or surgeons would reflect sparse referral networks that could contribute to barriers in cancer consultations during treatment planning. We found that patients residing in HRRs with a medium and high proportion of linchpin surgeons were 1.45 (95% CI: 1.12-1.86; p < 0.01) and 2.95 (95% CI: 1.82-4.78; p < 0.01) times more likely, respectively, of exclusively consulting a radiation oncologist versus receiving MDCc, relative to those residing in HRRs with a low proportion of linchpin surgeons. Similarly, patients in HRRs with medium or high proportion of linchpin surgeons were 1.54 (95% CI: 1.19-1.98; p < 0.001) and 2.63 (95% CI: 1.62-4.28; p < 0.001) times more likely of seeing neither surgeon nor radiation oncologist versus receiving MDCc compared with those residing in HRRs with a low proportion of linchpin surgeons.
Factors associated with receipt of SBRT
The associations between patient characteristics and receipt of SBRT compared to surgery are presented in Table 3. Patients aged 70-74 (odds ratio [OR]: 1.75; 95% CI: 1.41-2.17; p < 0.01), 75-79 (OR: 2.26; 95% CI 1.81-2.82; p < 0.01), 80-84 (OR: 4.61, 95% CI: 3.62-5.88; p < 0.01), and 85+ (OR: 18.85; 95% CI: 13.75-25.84; p < 0.01) at diagnosis were more likely to receive SBRT compared to those aged 66-69 years at diagnosis. Patients with stage IB cancer were half as likely to receive SBRT instead of surgery compared with those with stage IA cancer (OR: 0.50, 95% CI: 0.42-0.58, p < 0.01). Patients with COPD had 1.89 greater odds (95% CI: 1.59-2.24; p < 0.01) of receiving SBRT over surgery compared to those without COPD. Similarly, compared with pateints with no comorbidities, patients with one comorbidity had 1.45 greater odds (95% CI: 1.14-1.85; p < 0.01) of receiving SBRT over surgery and patients with 2 or more comorbidities had 1.92 greater odds (95% CI: 1.52-2.44; p < 0.01) of receiving SBRT over surgery.
Table 3.
Patient characteristics associated with receipt of SBRT vs. surgery
| Adjusted OR (95% CI) | P-value | |
|---|---|---|
| Female Sex | 1.06 (0.92-1.22) | 0.44 |
| Age at diagnosis (years) | ||
| 66-69 | 1.0 (Referent) | |
| 70-74 | 1.75 (1.41-2.17) | <0.01 |
| 75-79 | 2.26 (1.81-2.82) | <0.01 |
| 80-84 | 4.61 (3.62-5.88) | <0.01 |
| 85+ | 18.85 (13.75-25.84) | <0.01 |
| Race | ||
| White | 1.0 (Referent) | |
| Black | 1.42 (1.06-1.92) | <0.01 |
| Other | 0.79 (0.52-1.21) | 0.22 |
| Hispanic | 0.73 (0.48-1.11) | 0.27 |
| Cancer Stage | ||
| IA | 1.0 (Referent) | |
| IB | 0.50 (0.42-0.58) | <0.01 |
| COPD | ||
| No COPD | 1.0 (Referent) | |
| COPD | 1.89 (1.59-2.24) | <0.01 |
| COPD w/ exacerbation | 1.02 (0.68-1.53) | 0.93 |
| Charlson comorbidity | ||
| 0 | 1.0 (Referent) | |
| 1 | 1.45 (1.14-1.85) | <0.01 |
| 2+ | 1.92 (1.52-2.44) | <0.01 |
| Non-Metropolitan | 0.90 (0.71-1.13) | 0.20 |
| Yost Quantile | ||
| 1 (lowest SES) | 1.0 (Referent) | |
| 2 | 1.12 (0.86-1.44) | 0.35 |
| 3 | 0.89 (0.68-1.16) | 0.40 |
| 4 | 0.87 (0.67-1.13) | 0.38 |
| 5 (highest SES) | 0.78 (0.59-1.03) | 0.07 |
| Receive MDCc | 5.78 (4.58-7.30) | <0.01 |
| Proportion Linchpin Surgeons | ||
| Low | 1.0 (Referent) | |
| Medium | 1.17 (0.85-1.60) | 0.40 |
| High | 1.93 (1.19-3.14) | <0.01 |
| Proportion Linchpin Radiation Oncologists | ||
| Low | 1.0 (Referent) | |
| Medium | 0.68 (0.48-0.95) | 0.01 |
| High | 0.66 (0.44-0.98) | 0.02 |
Note. OR, odds ratio; MDCc, multidisciplinary cancer consultation; COPD, chronic obstructive pulmonary disease; SES, socioeconomic status.
Patients who received MDCc had 5.78 greater odds (95% CI: 4.58-7.30; p < 0.01) of receiving SBRT instead of surgery compared to those who did not receive MDCc. Patients living in areas with a high proportion of linchpin surgeons were found to have significantly higher odds of receiving SBRT compared to those in areas with a low proportion of linchpin surgeons (OR: 1.93; 95% CI: 1.19-3.14; p < 0.01). Patients living in areas with a high proportion of linchpin radiation oncologists were less likely to receive SBRT, but the significance of the effect was borderline statistically significant based on the threshold for statistical signficance we used in this study.
DISCUSSION
Our study of patients diagnosed with early-stage NSCLC in 2016-2017 found there was a significant association between SES, age, cancer stage, comorbidities, and the proportion of linchpin surgeons in the patient’s HRR and the type of cancer consultation received by patients following their cancer diagnosis. Our research aligns with existing literature that has identified disparities in the receipt of MDCc based on patients’ characteristics and their access to high-quality healthcare. While we reported on differences in patient characteristics associated with each consultation type, we also observed that 26.6% of our patients did not have a consultation with a surgeon or a radiation oncologist within the two months following diagnosis, which is the guideline-directed care for early-stage NSCLC cancer. In exploratory analysis of this group of patients, we found that 40% had an encounter with a medical oncologist during this time-frame, suggesting they may have been referred for systemic therapy. However, the higher percentages of Black and Hispanic patients in the group that received no consultations with a surgeon or a radiation oncologist compared with corresponding percentages in the other consultation groups provides additional evidence that racial and ethnic minority groups experience more structural barriers to specialty care referral.25,26 Our study also found that patients residing in areas in the lowest SES quintile were more likely to receive no consultation with a surgeon or radiation oncologist (vs. MDCc) compared to patients residing in higher SES areas. Prior work has shown that patients residing in areas characterized by socioeconomic disadvantage are more likely to present with advanced stage cancer, less likely to undergo cancer-directed surgery, and have worse survival.27–30 Our results add to this literature by highlighting disparities in access to specialty consultations during NSCLC treatment planning and initiation by patient race, ethnicity, and socioeconomic status.
Our findings also expand upon previous research by not only considering patient-level characteristics but also examining the scarcity of cancer specialists within patient-sharing networks and its impact on patients’ cancer care utilization. An advantage to our network-based approach for measuring access to care from a patient-sharing network not constrained by geography is that it may better reflect referral patterns and multidisciplinary teams of providers that practice across locations. Specifically, we found that patients residing in regions with a high proportion of linchpin surgeons had reduced likelihood of receiving MDCc. We also found that these patients were more likely to receive SBRT instead of surgery. This suggests that in areas where lung cancer surgeons are scarce, referrals to and treatment by radiation oncologists may be more predominant for early-stage lung cancer care. We did not see statistically significant associations between the proportion of linchpin radiation oncologists and receipt of consultations or treatment, although we did find a trend suggesting that patients in areas with a high proportion of linchpin radiation oncologists were less likely to receive SBRT. This could be reflective of surgeons typically being the specialist to whom patients are initially referred, such that barriers to MDCc and treatment are most pronounced when access to lung cancer surgeons is limited. Referrals to MDCc during the treatment planning phase can lead to more informed treatment choices, and network-based measures such as linchpin score have the potential to be leveraged to characterize points in a referral network that are particularly vulnerable to specialist scarcity.
Finally, our study found that patients who received MDCc were over 5-times more likely to receive SBRT instead of surgery. This may be due to selective referral of patients to MDCc among those who initially present as less favorable candidates for surgery. On the other hand, it may reflect the significance of the role played by radiation oncologists when guiding treatment decisions for patients with early-stage NSCLC. Future studies, including qualitative studies of physicians, patients, and caregivers, is warranted to further understand reasons driving selective referral of patients to surgical and radiation oncology consultations.
While the SEER-Medicare data is a rich dataset inclusive of patient-level clinical, socioeconomic, and sociodemographic information, there are limitations to this study. First, patients in our study are Medicare beneficiaries and over the age of 66, so our results may not be generalizable to a younger population. This could be addressed in future analysis beyond fee-for-service Medicare data. For instance, privately insured individuals or those enrolled in Medicaid may experience barriers to care due to narrow referral networks, which although beyond the scope of this study, could uncover novel drivers of disparities in treatment access and outcomes for these patient populations. Second, patients diagnosed with NSCLC at later stages were excluded from the analysis because patients with more advanced cancers are presented with different treatment strategies that were beyond the focus of this study. This could lead to the underestimation of disparities found in access to care as prior studies had shown that diagnoses of advanced cancer stage are more prevalent among Black populations and people residing in lower income and education areas.31–33 Third, we were not able to adjust for some clinical risk factors, such as smoking status or supplemental O2 use, which could influence the patient’s likelihood of referral to radiation oncology for SBRT treatment. Fourth, we are unable to determine if a patient was recommended to seek a consultation with a provider and declined, or if they were never referred. Fifth, multidisciplinary meetings such as tumor boards or other nonbillable conferences, which frequently take place prior to initiation of treatment, were not documented in the claims data. These conferences are important aspects of multidisciplinary care coordination that may drive referrals at the local level and could result in underappreciation of multidisciplinary care using our method of defining MDCc.
Notwithstanding these limitations, this work serves as a foundation for future investigations and interventions aimed at improving the receipt of MDCc among disadvantaged populations. Our investigation of a novel patient-sharing network measure found that HRRs characterized by higher proportions of linchpin surgeons was associated with lower likelihood of MDCc and greater odds of receiving SBRT instead of surgery, even after adjusting for several relevant clinical characteristics. As nationwide or regional strategies are developed to address uneven geographic distributions of specialty care, these efforts may benefit from creating referral pattern guidelines that use network analysis to identify areas with sparse referral networks for specialty care. In conclusion, improving equitable receipt of cancer consultations through strengthening multidisciplinary relationships among oncologists may reduce unwarranted variation in treatment decisions and clinical outcomes.
Supplementary Material
[Acknowledgements]
The funder did not play a role in the design of the study; the collection, analysis, and interpretation of the data; the writing of the manuscript; and the decision to submit the manuscript for publication. This study used the linked SEER-Medicare database. The interpretation and reporting of these data are the sole responsibility of the authors. The authors acknowledge the efforts of the National Cancer Institute; Information Management Services (IMS), Inc.; and the Surveillance, Epidemiology, and End Results (SEER) Program tumor registries in the creation of the SEER-Medicare database. The collection of cancer incidence data used in this study was supported by the California Department of Public Health pursuant to California Health and Safety Code Section 103885; Centers for Disease Control and Prevention’s (CDC) National Program of Cancer Registries, under cooperative agreement 1NU58DP007156; the National Cancer Institute’s Surveillance, Epidemiology and End Results Program under contract HHSN261201800032I awarded to the University of California, San Francisco, contract HHSN261201800015I awarded to the University of Southern California, and contract HHSN261201800009I awarded to the Public Health Institute. The ideas and opinions expressed herein are those of the author(s) and do not necessarily reflect the opinions of the State of California, Department of Public Health, the National Cancer Institute, and the Centers for Disease Control and Prevention or their Contractors and Subcontractors.
[Funding Statement]
This work was supported by the National Cancer Institute at the National Institutes of Health (grant number R37CA263936 to E.L.M).
Footnotes
Conflict of Interest: None
[Data Availability Statement for this Work]
The SEER-Medicare linked data underlying this article cannot be shared due to provisions outlined in the Data Use Agreement between the study Principal Investigator (E.L.M.) and the National Cancer Institute. Researchers interested in obtaining these data can submit a project-specific data request to the National Cancer Institute.
References
- 1.ASCO-ESMO Consensus Statement on Quality Cancer Care. J Clin Oncol. 2006;24(21):3498–3499. doi: 10.1200/JCO.2006.07.4021 [DOI] [PubMed] [Google Scholar]
- 2.Ellis PM. The importance of multidisciplinary team management of patients with non-small-cell lung cancer. Curr Oncol Tor Ont. 2012;19(Suppl 1):S7–S15. doi: 10.3747/co.19.1069 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Kowalczyk A, Jassem J. Multidisciplinary team care in advanced lung cancer. Transl Lung Cancer Res. 2020;9(4):1690. doi: 10.21037/tlcr.2019.11.33 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Denton E, Conron M. Improving outcomes in lung cancer: the value of the multidisciplinary health care team. J Multidiscip Healthc. 2016;9:137–144. doi: 10.2147/JMDH.S76762 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pillay B, Wootten AC, Crowe H, et al. The impact of multidisciplinary team meetings on patient assessment, management and outcomes in oncology settings: A systematic review of the literature. Cancer Treat Rev. 2016;42:56–72. doi: 10.1016/j.ctrv.2015.11.007 [DOI] [PubMed] [Google Scholar]
- 6.Prabhu Das I, Baker M, Altice C, Castro KM, Brandys B, Mitchell SA. Outcomes of multidisciplinary treatment planning in US cancer care settings. Cancer. 2018;124(18):3656–3667. doi: 10.1002/cncr.31394 [DOI] [PubMed] [Google Scholar]
- 7.Howington JA, Blum MG, Chang AC, Balekian AA, Murthy SC. Treatment of stage I and II non-small cell lung cancer: Diagnosis and management of lung cancer, 3rd ed: American College of Chest Physicians evidence-based clinical practice guidelines. Chest. 2013;143(5 Suppl):e278S–e313S. doi: 10.1378/chest.12-2359 [DOI] [PubMed] [Google Scholar]
- 8.Raman V, Yang CFJ, Deng JZ, D’Amico TA. Surgical treatment for early stage non-small cell lung cancer. J Thorac Dis. 2018;10(Suppl 7):S898–S904. doi: 10.21037/jtd.2018.01.172 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Turzer M, Brustugun OT, Waldeland E, Helland A. Stereotactic body radiation therapy is effective and safe in patients with early-stage non-small cell lung cancer with low performance status and severe comorbidity. Case Rep Oncol. 2011;4(1):25–34. doi: 10.1159/000324113 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Keane FK, Driscoll E, Bowes C, Durgin B, Khandekar MJ, Willers H. Low Rates of Chest Wall Toxicity When Individualizing the Planning Target Volume Margin in Patients With Early Stage Lung Cancer Treated With Stereotactic Body Radiation Therapy. Pract Radiat Oncol. 2021;11(3):e282–e291. doi: 10.1016/j.prro.2020.10.001 [DOI] [PubMed] [Google Scholar]
- 11.Mayne NR, Lin BK, Darling AJ, et al. Stereotactic Body Radiotherapy Versus Delayed Surgery for Early-stage Non-small-cell Lung Cancer. Ann Surg. 2020;272(6):925–929. doi: 10.1097/SLA.0000000000004363 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Timmerman R, Paulus R, Galvin J, et al. Stereotactic body radiation therapy for inoperable early stage lung cancer. JAMA. 2010;303(11):1070–1076. doi: 10.1001/jama.2010.261 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Matsuo Y, Chen F, Hamaji M, et al. Comparison of long-term survival outcomes between stereotactic body radiotherapy and sublobar resection for stage I non-small-cell lung cancer in patients at high risk for lobectomy: A propensity score matching analysis. Eur J Cancer. 2014;50(17):2932–2938. doi: 10.1016/j.ejca.2014.09.006 [DOI] [PubMed] [Google Scholar]
- 14.Dong B, Wang J, Xu Y, et al. Comparison of the Efficacy of Stereotactic Body Radiotherapy versus Surgical Treatment for Early-Stage Non-Small Cell Lung Cancer after Propensity Score Matching. Transl Oncol. 2019;12(8):1032–1037. doi: 10.1016/j.tranon.2019.04.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Subramanian MP, Meyers BF. Surgical Resection Versus Stereotactic Body Radiation Therapy for Stage I NSCLC: Can Randomized Trials Provide the Solution? Cancers. 2018;10(9):310. doi: 10.3390/cancers10090310 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Arvold ND. Involvement of Radiation Oncologists in Multidisciplinary Decision-making for Early-Stage Lung Cancer. JAMA Netw Open. 2022;5(3):e224849. doi: 10.1001/jamanetworkopen.2022.4849 [DOI] [PubMed] [Google Scholar]
- 17.Quyyumi FF, Wright JD, Accordino MK, et al. Factors Associated with Multidisciplinary Consultations in Patients with Early Stage Breast Cancer. Cancer Invest. 2019;37(6):233–241. doi: 10.1080/07357907.2019.1624766 [DOI] [PubMed] [Google Scholar]
- 18.Sanchez JI, Doose M, Zeruto C, et al. Multilevel factors associated with inequities in multidisciplinary cancer consultation. Health Serv Res. 2022;57 Suppl 2(Suppl 2):222–234. doi: 10.1111/1475-6773.13996 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Nemesure MD, Schwedhelm TM, Sacerdote S, O’Malley AJ, Rozema LR, Moen EL. A measure of local uniqueness to identify linchpins in a social network with node attributes. Appl Netw Sci. 2021;6:56. doi: 10.1007/s41109-021-00400-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Moen EL, Brooks GA, O’Malley AJ, Schaefer A, Carlos HA, Onega T. Use of a Novel Network-Based Linchpin Score to Characterize Accessibility to the Oncology Physician Workforce in the United States. JAMA Netw Open. 2022;5(12):e2245995. doi: 10.1001/jamanetworkopen.2022.45995 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Moen EL, Schmidt RO, Onega T, Brooks GA, O’Malley AJ. Association between a network-based physician linchpin score and cancer patient mortality: A SEER-Medicare analysis. J Natl Cancer Inst. Published online September 7, 2023:djad180. doi: 10.1093/jnci/djad180 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.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):373–383. doi: 10.1016/0021-9681(87)90171-8 [DOI] [PubMed] [Google Scholar]
- 23.Yost K, Perkins C, Cohen R, Morris C, Wright W. Socioeconomic status and breast cancer incidence in California for different race/ethnic groups. Cancer Causes Control CCC. 2001;12(8):703–711. doi: 10.1023/a:1011240019516 [DOI] [PubMed] [Google Scholar]
- 24.Hausman J, McFadden D. Specification Tests for the Multinomial Logit Model. Econometrica. 1984;52(5):1219–1240. doi: 10.2307/1910997 [DOI] [Google Scholar]
- 25.Simpson DR, Martínez ME, Gupta S, et al. Racial disparity in consultation, treatment, and the impact on survival in metastatic colorectal cancer. J Natl Cancer Inst. 2013;105(23):1814–1820. doi: 10.1093/jnci/djt318 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Yearby R. Racial Disparities in Health Status and Access to Healthcare: The Continuation of Inequality in the United States Due to Structural Racism. Am J Econ Sociol. 2018;77(3-4):1113–1152. doi: 10.1111/ajes.12230 [DOI] [Google Scholar]
- 27.Caballo B, Dey S, Prabhu P, Seal B, Chu P, Kim L. The Effects of Socioeconomic Status on the Quality and Accessibility of Healthcare Services. Published online May 3, 2021. doi: 10.5281/ZENODO.4740684 [DOI] [Google Scholar]
- 28.Fiscella K, Williams DR. Health disparities based on socioeconomic inequities: implications for urban health care. Acad Med J Assoc Am Med Coll. 2004;79(12):1139–1147. doi: 10.1097/00001888-200412000-00004 [DOI] [PubMed] [Google Scholar]
- 29.McMaughan DJ, Oloruntoba O, Smith ML. Socioeconomic Status and Access to Healthcare: Interrelated Drivers for Healthy Aging. Front Public Health. 2020;8:231. doi: 10.3389/fpubh.2020.00231 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Wang N, Cao F, Liu F, et al. The effect of socioeconomic status on health-care delay and treatment of esophageal cancer. J Transl Med. 2015;13:241. doi: 10.1186/s12967-015-0579-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Papageorge MV, Woods AP, de Geus SWL, et al. The Persistence of Poverty and its Impact on Cancer Diagnosis, Treatment and Survival. Ann Surg. Published online July 6, 2022. doi: 10.1097/SLA.0000000000005455 [DOI] [PubMed] [Google Scholar]
- 32.Soneji S, Tanner NT, Silvestri GA, Lathan CS, Black W. Racial and Ethnic Disparities in Early-Stage Lung Cancer Survival. Chest. 2017;152(3):587–597. doi: 10.1016/j.chest.2017.03.059 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Gupta A, Omeogu CH, Islam JY, Joshi AR, Akinyemiju TF. Association of area-level socioeconomic status and non-small cell lung cancer stage by race/ethnicity and health care-level factors: Analysis of the National Cancer Database. Cancer. 2022;128(16):3099–3108. doi: 10.1002/cncr.34327 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
The SEER-Medicare linked data underlying this article cannot be shared due to provisions outlined in the Data Use Agreement between the study Principal Investigator (E.L.M.) and the National Cancer Institute. Researchers interested in obtaining these data can submit a project-specific data request to the National Cancer Institute.