Key Points
Question
What factors are associated with uptake of the payer-led, evidence-based guidelines known as oncology clinical pathways programs?
Findings
This cohort study of claims and administrative data from 17 293 US adults with cancer found that only 64% of patients with newly diagnosed metastatic cancer received on-pathway treatments.
Meaning
These results suggest that despite financial incentives, compliance with payer-led pathways has remained at historically reported low rates.
This cohort study of claims and administrative data from US adults with cancer examines patient and physician factors associated with compliance with a payer-led program to standardize prescription practices and control drug costs.
Abstract
Importance
Payers use oncology clinical pathways programs to increase evidence-based drug prescribing and control drug spending. However, compliance with these programs has been low, which may decrease their efficacy, and factors associated with pathway compliance are unknown.
Objective
To determine extent of pathway compliance and identify factors associated with pathway compliance using characteristics of patients, practices, and the companies that develop cancer treatment pathways.
Design, Setting, and Participants
This cohort study comprised patients with claims and administrative data from a national insurer and a pathways health care professional between July 1, 2018, and October 31, 2021. Adult patients with metastatic breast, lung, colorectal, pancreatic, melanoma, kidney, bladder, gastric, and uterine cancer being treated in the first line were included. Six months of continuous insurance coverage prior to the date of treatment initiation was required for determination of baseline characteristics. Stepwise logistic regression was used to identify factors associated with pathway compliance.
Main Outcomes and Measures
Use of a pathway program–endorsed treatment regimen in the first-line setting for metastatic cancer.
Results
Among 17 293 patients (mean [SD] age, 60.7 [11.2] years; 9183 [53.1%] women; mean [SD] Black patients per census block, 0.10 [0.20]), 11 071 patients (64.0%) were on-pathway, and 6222 (36.0%) were off-pathway. Factors associated with increased pathway compliance were higher health care utilization during the 6-month baseline period (measured in inpatient visits and emergency department visits) (5220 on-pathway inpatient visits [47.2%] vs 2797 off-pathway [45.0%]; emergency department visits, 3304 [27.1%] vs 1503 [24.2%]; adjusted odds ratio [aOR] for inpatient visits, 1.32; 95% CI, 1.22-1.43; P < .001), volume of patients with this insurance provider per physician (mean [SD] visits: on-pathway, 128.0 [258.3] vs off-pathway, 121.8 [161.4]; aOR, 1.12; 95% CI, 1.04-1.20; P = .002), and practice participation in the Oncology Care Model (on-pathway participation, 2601 [23.5%] vs 1305 [21.0%]; aOR, 1.13; 95% CI, 1.04-1.23; P = .004). Higher total medical cost during the 6-month baseline period were associated with decreased pathway compliance (mean [SD] costs: on-pathway, $55 990 [$69 706] vs $65 955 [$74 678]; aOR, 0.86; 95% CI, 0.83-0.88; P < .001). There was heterogeneity in odds of pathway compliance between different malignancies. Pathway compliance rates trended down from the reference year of 2018.
Conclusions and Relevance
In this cohort study, despite generous financial incentives, compliance with payer-led pathways remained at historically reported low rates. Factors such as increasing exposure to the program due to the number of patients impacted and participation in other value-based payment programs, such as the Oncology Care Model, were positively associated with compliance; factors such as the type of cancer and patient complexity may have played a role, but the directionality of potential effects was unclear.
Introduction
Both cost and complexity have increased in the practice of oncology.1 Evidence suggests that cancer drug prescriptions account for the largest portion of spending on cancer care and the greatest variation in practice.2 Given this, there has been growing interest among both payers and clinicians in the use of tools that would promote quality by introducing a degree of standardization to drug prescribing and decrease costs by reducing inappropriate non–evidence-based care.
One such approach is to implement clinical pathways programs, a subset of evidence-based guidelines, such as those from the National Comprehensive Cancer Network, that help clarify decisions when multiple treatment options exist, typically along 3 priorities in the following order: efficacy, safety, and cost.3 When efficacy and safety profiles are similar between evidence-based treatment regimens, the pathways program favors lower-priced regimens over higher-priced regimens.4 As pathways take cost into account, the hope is that their use will help manage drug use in a dynamic treatment landscape by nudging clinicians into selecting higher-value choices.3,5
With this value proposition, multiple national payers have developed and introduced oncology clinical pathways programs.6 However, execution is critical to the success of the premise. Payers have historically found that compliance is a key challenge, with studies showing that compliance with pathways can be as low as 50% to 70%.7,8,9 For example, UnitedHealthcare piloted a clinician-led pathway program with 5 volunteer medical oncology groups, and later discovered a 3-fold variation in drug costs between groups.10 Further analysis revealed that compliance with the pathway was less than 50%, and this noncompliance was responsible for the variance. Understanding the factors associated with pathway compliance is therefore a key implementation question.
In this retrospective cohort study, we used claims data from a large insurer and administrative data from a company that develops cancer treatment pathways to identify factors associated with compliance that incorporates a wide range of patient-, physician-, and practice-level factors.
Methods
Program Overview
Claims data were obtained from Elevance Health (formerly Anthem Inc), a national insurer that launched an oncology clinical pathway program, the Cancer Care Quality Program (CCQP), through its subsidiary AIM Specialty Health in 2014.11 An external panel of oncology experts meets on a quarterly basis and reviews, approves, and updates regimens over time to reflect current evidence.4 To incentivize selection of pathway-compliant regimens, Elevance Health pays practices an additional reimbursement of $350 per patient each month. Treatment requests made by oncologists for regimens not included in the pathways program are reviewed and still authorized without delay or requirement for peer-to-peer interaction if they are determined to be medically necessary pursuant to medical policies and clinical guidelines.11
Study Design and Data
We obtained longitudinal medical and pharmacy claims and eligibility files from Elevance Health care plans across the US for information about health care utilization, medical spending, and health plan enrollment. Our data were from 3 types of health plan members: Medicare Advantage (MA), fully insured members, and members of plans through self-insured employers for whom Elevance Health offers administrative services only (ASO).
CCQP files contained data from prior authorization requests, which included information on requested drug regimen, line of therapy, cancer type, cancer stage, body mass index (BMI; calculated as weight in kilograms divided by height in meters squared), Eastern Cooperative Oncology Group (ECOG) performance score, National Provider Identifier, and tax identification number of the chemotherapy-prescribing clinicians.12 We obtained data about prescribing oncologists—clinician sex and years from medical school graduation—from Enclarity’s Provider Data Masterfile (formerly LexisNexis).13 We used data from the American Community Survey for insights on neighborhood-level social determinants of health. We derived Oncology Care Model (OCM) participation data by practices from the Centers for Medicare & Medicaid Services program website.14
This study was conducted by researchers using a limited data set for analysis, which was devoid of individual patient identifiers and complied with all relevant provisions of the Health Insurance Portability and Accountability Act (HIPAA) and the HIPAA Privacy Rule (45 CFR 164.514(e)). Findings are reported using the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline for cohort studies.
Study Population
From CCQP data, we identified adult patients with metastatic cancer of the following cancer types: breast, lung, colorectal, pancreatic, melanoma, kidney, bladder, gastric, and uterine cancer, who were prescribed initial anti-cancer drug therapy between July 1, 2018, and October 31, 2021. These cancer diagnoses represented the most common solid tumor malignant neoplasms among the insurer cohort. We excluded prostate cancer due to preponderance of hormonal therapy prescribed for this indication. Index date was defined as the first request data of the first-line anti-cancer drug therapy after a metastatic diagnosis. Six months of continuous insurance coverage prior to first treatment date were required to assess health care characteristics for patients and clinicians during this baseline period.
We limited our evaluation to index cancer drugs prescribed in the first-line setting for metastatic cancer because evidence supporting the choice of initial regimens is stronger than for later-line regimens. Line of therapy was obtained from prior authorization data submitted by oncologists.
Outcome
The primary outcome was use of a pathway program–endorsed regimen among patients with metastatic cancer treated in the first-line setting. A patient’s pathway status was designated as on-pathway, off-pathway, or unresolved. Unresolved pathway status was present when regimen requests were sufficient for authorization but clinical data provided (such as biomarkers) were inadequate to determine on-pathway or off-pathway status. Patients with unresolved pathway status were excluded from the main analysis.
Factors
Patient characteristics included age (categorized into ages 18 to 44, 45 to 49, 50 to 64, 65 to 74, and 75 plus years), sex, BMI, Deyo-Charlson Comorbidity index (DCI) score, cancer type, performance status as measured on the ECOG scale, insurance type (commercial fully insured, commercial ASO, or MA), census region (Northeast, Midwest, South, West), urban or rural residence (inside or outside of a US Census Bureau metropolitan statistical area), distance of patient to treatment facility, third party prescription coverage, and health care utilization and cost during the baseline period (eg, inpatient visits, emergency department (ED) visits, total medical cost, total pharmacy cost). Demographic covariates were assessed on index date, and comorbidities and patient history were assessed from claims during the 6 months prior to index date. We obtained the following social determinants of health variables, drawn at the census block group level from the American Community Survey—median family income, Black race, Hispanic ethnicity, education level, and socioeconomic status index scores.
Clinician and practice covariates included clinician sex, years since medical school graduation, exposure time to the CCQP pathway program (calculated as months between the time when a clinician first appeared in CCQP data and index date), plan member patient volume per physician (number of distinct patients associated with a clinician during 6 months prior to index date), and participation in the OCM.
Statistical Analysis
We used a logistic regression model to identify factors associated with receipt of pathway-compliant drug regimens. Stepwise backward selection was used to retain covariates based on Akaike information criteria at P < .157.
Continuous variables were dichotomized based on median values for easier interpretation; these variables included median family income, years since physicians’ medical school graduation, months of exposure of physician to the pathways program, and number of plan member patients per physician. Log 2 transformation was performed on total medical cost and total pharmacy cost to remove skewness—this implies that the coefficient should be interpreted as the change in pathway compliance rate when doubling costs during the baseline period. Imputation was not performed given small numbers for data missingness within variables such as geographic region (0.6% to 0.7%), residence in urban area (0.6% to 0.7%), clinician covariates including sex and years since medical school graduation (1.3% to 2.6%), socioeconomic status variables (4.2% to 4.8%) and ECOG status (3.3% to 4.7%). Independent variables were tested for multicollinearity, and excluded if the pairwise correlation coefficient was greater than 0.7. Index year of drug prescribing was included to account for secular trends. We present the results of logistic regression as adjusted odds ratios (aORs).
We conducted a sensitivity analysis in which we grouped patients with unresolved pathway status with those on-pathway, with the assumption that an unresolved status would more likely include on-pathway rather than off-pathway regimens, as the regimen requests were otherwise sufficient for authorization.
Statistical significance was set at P < .05 in 2-sided tests. Analyses were conducted using SAS Enterprise Guide version 7.15 (SAS Institute Inc).
Results
Cohort Characteristics
The cohort consisted of 17 293 patients, of whom 11 071 (64.0%) were treated with pathway-compliant regimens, 6222 (36.0%) were treated with off-pathway regimens (Table).
Table. Characteristics of Patients, Physicians, and Practices Participating in the Study Sample.
| Characteristic | Participants, No. (%) | ||
|---|---|---|---|
| On-pathway (n = 11 071) | Off-pathway (n = 6222) | Total (n = 17 293) | |
| Age, mean (SD), y | 60.8 (11.1) | 60.5 (11.6) | 60.7 (11.2) |
| Age category | |||
| 18-44 y | 761 (6.9) | 544 (8.7) | 1305 (7.6) |
| 45-64 y | 6842 (61.8) | 3755 (60.4) | 10 597 (61.3) |
| 65-74 y | 2181 (19.7) | 1175 (18.9) | 3356 (19.4) |
| ≥75 y | 1287 (11.6) | 748 (12.0) | 2035 (11.8) |
| Sex | |||
| Men | 5312 (48.0) | 2798 (45.0) | 8110 (46.9) |
| Women | 5759 (52.0) | 3424 (55.0) | 9183 (53.1) |
| Insurance | |||
| Commercial fully insured | 2363 (21.3) | 1327 (21.3) | 3690 (21.3) |
| Commercial ASO | 6187 (55.9) | 3472 (55.8) | 9659 (55.9) |
| Medicare Advantage | 2521 (22.8) | 1423 (22.9) | 3944 (22.8) |
| Index year | |||
| 2018 | 2065 (18.7) | 710 (11.4) | 2775 (16.1) |
| 2019 | 3701 (33.4) | 1586 (25.5) | 5287 (30.6) |
| 2020 | 2766 (25.0) | 2189 (35.2) | 4955 (28.7) |
| 2021 | 2539 (22.9) | 1737 (27.9) | 4276 (24.7) |
| Cancer type | |||
| Breast | 1836 (16.6) | 1453 (23.4) | 3289 (19.0) |
| Lung | 2943 (26.6) | 2324 (37.4) | 5267 (30.5) |
| Colorectal | 2791 (25.2) | 1020 (16.4) | 3811 (22.0) |
| Pancreatic | 1476 (13.3) | 204 (3.3) | 1680 (9.7) |
| Melanoma | 811 (7.3) | 406 (6.5) | 1217 (7.0) |
| Kidney | 687 (6.2) | 283 (4.5) | 970 (5.6) |
| Bladder | 221 (2.0) | 221 (3.6) | 442 (2.6) |
| Gastric | 144 (1.3) | 230 (3.7) | 374 (2.2) |
| Uterine | 162 (1.5) | 81 (1.3) | 243 (1.4) |
| Body mass index, mean (SD) | 28.1 (6.8) | 27.9 (6.9) | 28.0 (6.8) |
| Performance status on ECOG scale, mean (SD) | 0.7 (0.7) | 0.7 (0.7) | 0.7 (0.7) |
| Charlson Comorbidity Index Score, mean (SD) | 6.9 (2.1) | 6.8 (2.1) | 6.9 (2.1) |
| Region | |||
| Northeast | 1853 (16.7) | 1077 (17.3) | 2930 (16.9) |
| Midwest | 3841 (34.7) | 2030 (32.6) | 5871 (34.0) |
| South | 3529 (31.9) | 1980 (31.8) | 5509 (31.9) |
| West | 1770 (16.0) | 1091 (17.5) | 2861 (16.5) |
| Urban residence | 8492 (76.6) | 4756 (76.4) | 13 248 (76.6) |
| Use of third-party pharmacy | 3275 (29.6) | 1865 (30.0) | 5140 (29.7) |
| Distance of patient to treatment facility, mean (SD), mi | 34.4 (136.3) | 38.1 (161.6) | 35.7 (145.9) |
| Social determinants of health, mean (SD), No. per census block | |||
| Median family income, $ | 81 903.40 (38 881) | 81 985.74 (38 899) | 81 933 (38 886) |
| Black race | 0.10 (0.18) | 0.10 (0.19) | 0.10 (0.19) |
| Hispanic ethnicity | 0.10 (0.15) | 0.10 (0.15) | 0.10 (0.15) |
| High school or more | 0.89 (0.09) | 0.89 (0.09) | 0.89 (0.09) |
| 4 y of college or more | 0.31 (0.20) | 0.31 (0.20) | 0.31 (0.20) |
| SES index score | 54.63 (5.63) | 54.68 (5.64) | 54.65 (5.63) |
| Inpatient visits | |||
| Any inpatient visits during baseline, No. (%) | 5220 (47.2) | 2797 (45.0) | 8017 (46.4) |
| Total inpatient visits during baseline, mean (SD) | 0.67 (0.89) | 0.64 (0.88) | 0.66 (0.89) |
| ED visits | |||
| Any ED visits during baseline, No. (%) | 3004 (27.1) | 1503 (24.2) | 4507 (26.1) |
| Total ED visits during baseline, mean (SD) | 0.38 (0.8) | 0.34 (0.75) | 0.37 (0.78) |
| Total medical cost during baseline, mean (SD), $ | 55 990 (69 706) | 65 955 (74 678) | 61 213 (75 492) |
| Total pharmacy cost during baseline, mean (SD), $ | 4418 (15 074) | 3729 (13 715) | 4171 (14 604 |
| Physician years since medical school graduation, mean (SD) | 17.2 (12.7) | 17.6 (12.7) | 17.4 (12.7) |
| Physician women | 3276 (29.8) | 1937 (31.3) | 5.213 (30.1) |
| Physician experience with pathways, mean (SD), mo | 49.3 (21.0) | 51.7 (21.5) | 50.2 (21.2) |
| Anthem patient volume, patients per physician during baseline, mean (SD) | 128.0 (258.3) | 121.8 (161.4) | 125.8 (228.3) |
| Participation in Oncology Care Model | 2601 (23.5) | 1305 (21) | 3906 (22.6) |
Abbreviations: ASO, administrative services only; ECOG, Eastern Cooperative Oncology Group; ED, emergency department; SES, socioeconomic status.
On-pathway and off-pathway groups had similar mean (SD) age (60.8 [11.1] years vs 60.5 years [11.6]; P = .08), with fewer women in the on-pathway group (5759 of 11 071 [52.0%] vs 3424 of 6222 [55.0%]; P < .001). The top 3 cancer types in both groups were breast, lung, and colorectal cancer. The majority of patients were from the Midwest (5871 [34.0%]), had commercial ASO plans (9659 [55.9%]), and resided in urban areas (13 248 [76.6%]). Mean (SD) BMI was 28.0 (6.8), mean Deyo-Charlson Comorbidity Index score was 6.9 (2.1), and 1535 patients (8.9%) had an ECOG value of 2 or greater.
Results from Logistic Regression
The odds of pathway compliance decreased over time. Compared with the reference year 2018, in which the pathway compliance rate was 74.4%, the compliance rate was 70% in 2019 (aOR, 0.73; 95% CI, 0.65-0.82), 55.8% in 2020 (aOR, 0.36; 95% CI, 0.32-0.41), and 59.4% in 2021 (aOR, 0.43; 95% CI, 0.38-0.49) (all P < .001) (Figure). We found heterogeneity in the level of compliance between cancer types. For example, compared with the reference lung cancer, gastric and bladder cancer had lower odds of being pathway compliant (gastric cancer: aOR, 0.50; 95% CI, 0.40-0.64; P < .001; bladder cancer: aOR, 0.76; 95% CI, 0.61-0.94; P < .001), while pancreatic cancer had much higher odds of being pathway compliant (aOR, 6.06; 95% CI, 5.14-7.14; P < .001).
Figure. Factors Associated With Pathway Compliance.
Forest plot contains only those factors included after stepwise selection. Results for pancreatic vs lung cancers (aOR, 6.06; 95% CI, 5.14-7.14; P < .001) were removed to improve readability. ASO indicates administrative services only; ED, emergency department; OR, odds ratio.
Patients with MA plans had lower odds of being pathway compliant compared with those with ASO plans (aOR, 0.84; 95% CI, 0.75-0.96; P = .008). The presence of inpatient hospitalizations or ED visits during the baseline 6-month period compared with no utilization was associated with higher odds of pathway compliance (inpatient visit in last 6 months: aOR, 1.32; 95% CI, 1.22-1.43; P < .001; ED visit in last 6 months: aOR, 1.21; 95% CI, 1.12-1.31; P < .001) while higher total medical costs in the baseline period was associated with lower odds for pathway compliance (aOR, 0.86; 95% CI, 0.83-0.88; P < .001).
One clinician characteristic that was significantly associated with pathway compliance after adjusting for other factors was the patient volume seen by the prescribing clinician. Prescribing clinicians who treated an above average number of patients were more likely to prescribe pathway-compliant regimens (aOR, 1.12; 95% CI, 1.04-1.20; P = .002). Practices that participated in the Oncology Care Model were positively associated with prescribing pathway-compliant regimens (aOR, 1.13; 95% CI, 1.04-1.23; P = .004).
In the sensitivity analysis, we grouped together 2794 patients with unresolved pathway status with those on-pathway and found that almost all factors remained significant, with the exception of the OCM participation as a factor losing significance.
Discussion
We found that despite participation in an oncology clinical pathway program that included financial incentives to encourage clinicians to prescribe on-pathway treatment regimens, only 64% of patients with newly diagnosed metastatic cancer received on-pathway treatments. This figure is consistent with prior studies.7,8
Variable compliance threatens the internal validity of the use of pathways compliance as an indicator of quality prescribing. While the goal is not to achieve a 100% compliance rate given that a certain portion of pathway deviations are not only expected but likely appropriate, previously published experiences of operationalizing cancer clinical pathways have targeted between 70% to 80% compliance as a threshold for success.15 In order to boost compliance rates, Elevance Health pays practices an additional reimbursement of $350 per patient each month when they adhere to pathways, which is not insignificant. As a benchmark, the additional monthly per-patient payments in the OCM were $160.16 Like other pathway programs, adherence reports are also provided for their clinic customers.
Many factors likely influence pathway compliance, such as patient factors, individual physician’s beliefs and routines, peer effects, etc. In at least 1 notable study looking at uptake of guideline-directed care,17 the most important variance in drug prescribing was associated with physician characteristics, not patient or disease characteristics. Given that many older adults are excluded from clinical trials, geriatric oncology research has shown that value-based assessments often lead to deviations from standards of care.18 Other studies have found that pathway adherence can range from 53% to 70% depending on disease site, and tends to decrease with increasing pathway complexity.7,8 We would be remiss not to mention the impact of the COVID-19 pandemic on treatment modifications and pathway deviations, particularly in the early, prevaccine days of the pandemic. One additional and important possibility is related to clinic workflow: payer-led pathways are not routinely incorporated into the electronic treatment ordering system. This can lead to oncologists not being aware of which treatments are on- or off-pathway at the moment of clinical decision-making, and requires that oncologists asynchronously study pathways for any differences in their plan of care and selection of regimen. Executing on pathways under these circumstances involves creating prior awareness and socialization of the program and the pathway choices, and effective communication is needed between the clinical staff and the administrative staff that enters the regimen and clinical scenario details into the electronic portal.
We found that the rate of pathway compliance decreased over time. This was a surprising finding, as one might expect that program maturity and physician familiarity would tilt the results in the other direction. Hypotheses to explain this finding include an increasing multiplicity of treatment options with time, lag time from regulatory approval of a new regimen to inclusion on pathway, the aforementioned implementation challenges related to workflow, increased biomarker reporting requirements for pathway adjudication that increase risk of missing data, and pathway deviations in the setting of the pandemic, particularly with the disruption of the flow of information in the practices with the displacement of administrative staff.
Two findings are noteworthy to policy makers or program architects seeking to influence clinician prescribing behavior. First, the positive association between plan member patient volume within a physician’s panel and prescribing of on-pathway regimens suggests that increasing familiarity with the pathways program is likely to influence compliance. In its State of Cancer Care in America 2017 report, ASCO reported a 42% increase in the preceding 2 years in practices reporting implementation of a pathway program.19 As such, it remains to be seen if this hypothesis bears out in practice, as physician exposure to pathways programs increases over time.
Second, practice participation in Centers for Medicare & Medicaid Services’s flagship value-based payment model for cancer—the OCM—was also modestly associated with higher likelihood of pathway compliance. The OCM is a total cost-of-care model, and as such, it is focused on the outcome of cost. Pathway programs focus on the process, or the actual selecting of drug regimens. The synergy between these 2 approaches should be of interest to policy makers looking to iterate on the next value-based care model.
We were surprised by the lack of association between ECOG status and the Deyo-Charlson Comorbidity Index on pathway compliance. Given the frequent intersection between complexity and high health care utilization, we studied patients’ health care utilization (dichotomized to any use or no use) and cost during the baseline period.20 While health care utilization during the baseline period was positively associated with compliance, higher total medical costs in the baseline period led to lower odds of pathway compliance. These results are discordant; we would note though that both are imperfect measures of complexity. In future studies, we would consider creating distinct classes of health care utilization along a gradation in order to describe profiles of health complexity and utilization with more nuance.
Limitations
Limitations of this study include the fact that this was a retrospective analysis. There was likely confounding by indication in the odds of whether a patient receives pathway-endorsed drug regimens or not. We were not able to distinguish the reasons for off-pathway prescribing, such as when there is a delay between when evidence is generated and inclusion of that regimen into the pathways program (which would not be indicative of poor-quality care). Administrative data lack clinical details, which could have informed some of the counterintuitive findings such as the association between clinical complexity and likelihood of pathway compliance. We did not have 100% capture rate of on- or off-pathway status, with unresolved cases due to missing data. While we have information about intended treatment from prior authorizations, we do not have information on whether patients actually started their prescribed regimens. We tried to adjust for selection bias by limiting our analysis by stage and line of therapy, and by adjusting for patient characteristics including age, sex, and comorbidity score. Program execution was only alluded to in covariates such as patient volume and payer type, but was likely a critical factor in the success of this program.
Conclusion
This cohort study on factors associated with compliance with a national payer-led oncology clinical pathway program yielded several expected and unexpected results. Despite significant financial incentives, compliance with the pathways program remained at previously reported rates, and furthermore, compliance appeared to trend down over the years. There is synergy between this program and other value-based payment programs such as the OCM. The contribution of patient complexity toward physician treatment decisions remains poorly understood. A forthcoming companion study will examine associations of pathway compliance with several cost and quality outcome measures. These data will be useful to architects of value-based payment models seeking best practices.
Data Sharing Statement
References
- 1.Mariotto AB, Yabroff KR, Shao Y, Feuer EJ, Brown ML. Projections of the cost of cancer care in the United States: 2010-2020. J Natl Cancer Inst. 2011;103(2):117-128. doi: 10.1093/jnci/djq495 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Clough JD, Patel K, Riley GF, Rajkumar R, Conway PH, Bach PB. Wide variation in payments for Medicare beneficiary oncology services suggests room for practice-level improvement. Health Aff (Millwood). 2015;34(4):601-608. doi: 10.1377/hlthaff.2014.0964 [DOI] [PubMed] [Google Scholar]
- 3.Daly B, Zon RT, Page RD, et al. Oncology clinical pathways: charting the landscape of pathway providers. J Oncol Pract. 2018;14(3):e194-e200. doi: 10.1200/JOP.17.00033 [DOI] [PubMed] [Google Scholar]
- 4.Carelon Medical Benefits . Cancer Treatment Pathways. Reviewed by Oncologists. AIM Specialty Health Accessed Aug 8, 2022. https://providers.carelonmedicalbenefitsmanagement.com/aim_workbook/
- 5.Hoverman JR, Cartwright TH, Patt DA, et al. Pathways, outcomes, and costs in colon cancer: retrospective evaluations in two distinct databases. J Oncol Pract. 2011;7(3)(suppl):52s-59s. doi: 10.1200/JOP.2011.000318 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Polite B, Ward JC, Cox JV, et al. A pathway through the bundle jungle. J Oncol Pract. 2016;12(6):504-509. doi: 10.1200/JOP.2015.008789 [DOI] [PubMed] [Google Scholar]
- 7.Rotenstein LS, Kerman AO, Killoran J, et al. Impact of a clinical pathway tool on appropriate palliative radiation therapy for bone metastases. Pract Radiat Oncol. 2018;8(4):266-274. doi: 10.1016/j.prro.2017.12.001 [DOI] [PubMed] [Google Scholar]
- 8.Kubal T, Letson DG, Chiappori AA, et al. Longitudinal cohort study to determine effectiveness of a novel simulated case and feedback system to improve clinical pathway adherence in breast, lung and GI cancers. BMJ Open. 2016;6(9):e012312. doi: 10.1136/bmjopen-2016-012312 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Yu PP. Oncology clinical pathways: a form of governance? J Oncol Pract. 2018;14(3):144-146. doi: 10.1200/JOP.17.00084 [DOI] [PubMed] [Google Scholar]
- 10.Newcomer LN. Innovative payment models and measurement for cancer therapy. J Oncol Pract. 2014;10(3):187-189. doi: 10.1200/JOP.2014.001378 [DOI] [PubMed] [Google Scholar]
- 11.Carelon Medical Benefits . Cancer Treatment Pathways. Updated November 1, 2022. Accessed Aug 8, 2022. https://aimproviders.com/wp-content/uploads/2019/11/AIM_Workbook.pdf
- 12.Carelon Medical Benefits . Cancer Care Quality Program. Accessed September 9, 2022. https://providers.carelonmedicalbenefitsmanagement.com/medoncology-anthem/
- 13.LexisNexis Risk Solutions . Provider Data MasterFile. Accessed November 1, 2022. https://risk.lexisnexis.com/products/provider-data-masterfile
- 14.Center for Medicare & Medicaid Innovation . Oncology Care Model. Accessed September 7, 2022. https://innovation.cms.gov/innovation-models/oncology-care
- 15.Feinberg BA, Lang J, Grzegorczyk J, et al. Implementation of cancer clinical care pathways: a successful model of collaboration between payers and providers. J Oncol Pract. 2012;8(3)(suppl):e38s-e43s. doi: 10.1200/JOP.2012.000564 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Osarogiagbon RU, Mullangi S, Schrag D. Medicare spending, utilization, and quality in the oncology care model. JAMA. 2021;326(18):1805-1806. doi: 10.1001/jama.2021.18765 [DOI] [PubMed] [Google Scholar]
- 17.Ford LG, Hunter CP, Diehr P, Frelick RW, Yates J. Effects of patient management guidelines on physician practice patterns: the Community Hospital Oncology Program experience. J Clin Oncol. 1987;5(3):504-511. doi: 10.1200/JCO.1987.5.3.504 [DOI] [PubMed] [Google Scholar]
- 18.Ng ZX, Zheng H, Chen MZ, Soon YY, Ho F. Comprehensive geriatric assessment guided treatment versus usual care for older adults aged 60 years and above with cancer. Cochrane Database Syst Rev. 2021;10(CD014875). doi: 10.1002/14651858 [DOI] [Google Scholar]
- 19.American Society of Clinical Oncology . The State of Cancer Care in America, 2016: a report by the American Society of Clinical Oncology. J Oncol Pract. 2016;12(4):339-383. doi: 10.1200/JOP.2015.010462 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pourat N, Chen X, Tsugawa Y, et al. Intersection of complexity and high utilization among health center patients aged 18 to 64 years. Am J Manag Care. 2022;28(2):66-72. doi: 10.37765/ajmc.2022.88751 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Sharing Statement
