Background: Although high-cost (HC) patients make up a small proportion of patients, they account for most health system costs. However, little is known about HC patients with cancer or whether some of their care could potentially be prevented. This analysis sought to characterize HC patients with cancer and quantify the costs of preventable acute care (emergency department visits and inpatient hospitalizations). Methods: This analysis examined a population-based sample of all HC patients in Ontario in 2013. HC patients were defined as those above the 90th percentile of the cost distribution; all other patients were defined as non–high-cost (NHC). Patients with cancer were identified through the Ontario Cancer Registry. Sociodemographic and clinical characteristics were examined and the costs of preventable acute care for both groups by category of visit/condition were estimated using validated algorithms. Results: Compared with NHC patients with cancer (n=369,422), HC patients with cancer (n=187,770) were older (mean age 70 vs 65 years), more likely to live in low-income neighborhoods (19% vs 16%), sicker, and more likely to live in long-term care homes (8% vs 0%). Although most patients from both cohorts tended to be diagnosed with breast, prostate, or colorectal cancer, those with multiple myeloma or pancreatic or liver cancers were overrepresented among the HC group. Moreover, HC patients were more likely to have advanced cancer at diagnosis and be in the initial or terminal phase of treatment compared with NHC patients. Among HC patients with cancer, 9% of spending stemmed from potentially preventable/avoidable acute care, whereas for NHC patients, this spending was approximately 30%. Conclusions: HC patients with cancer are a unique subpopulation. Given the type of care they receive, there seems to be limited scope to prevent acute care spending among this patient group. To reduce costs, other strategies, such as making hospital care more efficient and generating less costly encounters involving chemotherapy, should be explored.
Claire de Oliveira, Joyce Cheng, Kelvin Chan, Craig C. Earle, Murray Krahn and Nicole Mittmann
Amanda Putri Rahmadian, Seanthel Delos Santos, Shruti Parshad, Louis Everest, Matthew C. Cheung and Kelvin K. Chan
Background: Restricted mean survival time (RMST) overcomes limitations of current measures of survival benefits because it directly captures information of the entire area under Kaplan-Meier survival curves. Using RMST difference (absolute survival benefit) and RMST ratio (relative survival benefit), we quantified the magnitude of survival benefits of recent oncology drugs and compared immunotherapies with nonimmunotherapies. Methods: Kaplan-Meier curves were extracted from phase II/III randomized controlled trials used by the FDA for oncology drug approvals from January 2011 through November 2017 with overall survival (OS) or progression-free survival (PFS) as primary endpoints. RMST differences, ratios, and their 95% confidence intervals were meta-analyzed to estimate absolute and relative survival benefits of contemporary oncology drugs and to compare immunotherapies with nonimmunotherapies. Meta-regression was conducted to adjust for potential confounders. Results: Ninety-four trials with a total of 51,639 patients were included. Overall absolute survival benefits (RMST differences) were 1.55 months for OS (95% CI, 1.32–1.77) and 2.99 months for PFS (95% CI, 2.65–3.33). Overall relative survival benefits (RMST ratios) were 1.11 for OS (95% CI, 1.09–1.13) and 1.42 for PFS (95% CI, 1.36–1.48). Immunotherapy absolute PFS benefit was less than that of nonimmunotherapy (1.56 vs 3.23 months), whereas immunotherapy absolute OS benefit was larger than that of nonimmunotherapy by 0.59 months (2.02 vs 1.43 months). Adjusted OS RMST difference was 0.91 months greater for immunotherapy than for nonimmunotherapy after adjusting for confounders. Conclusions: Absolute survival benefits of recent oncology drugs are modest. Survival benefits of immunotherapies are not dramatically superior to those of nonimmunotherapies. Routine reporting and use of RMST may help patients, physicians, and payers make more informed and responsible decisions regarding the care of patients with cancer.
Sierra Cheng, Matthew C. Cheung, Di Maria Jiang, Erica McDonald, Vanessa S. Arciero, Doreen Anuli Ezeife, Amanda Rahmadian, Alexandra Chambers, Kelley-Anne Sabarre, Ambika Parmar and Kelvin K.W. Chan
Background: Clinical benefit scores (CBS) are key elements of the ASCO Value Framework (ASCO-VF) and are weighted based on a hierarchy of efficacy endpoints: hazard ratio for death (HR OS), median overall survival (mOS), HR for disease progression (HR PFS), median progression-free survival (mPFS), and response rate (RR). When HR OS is unavailable, the other endpoints serve as “surrogates” to calculate CBS. CBS are computed from PFS or RR in 39.6% of randomized controlled trials. This study examined whether surrogate-derived CBS offer unbiased scoring compared with HR OS–derived CBS. Methods: Using the ASCO-VF, CBS for advanced disease settings were computed for randomized controlled trials of oncology drug approvals by the FDA, European Medicines Agency, and Health Canada in January 2006 through December 2017. Mean differences of surrogate-derived CBS minus HR OS–derived CBS assessed the tendency of surrogate-derived CBS to overestimate or underestimate clinical benefit. Spearman’s correlation evaluated the association between surrogate- and HR OS–derived CBS. Mean absolute error assessed the average difference between surrogate-derived CBS relative to HR OS–derived CBS. Results: CBS derived from mOS, HR PFS, mPFS, and RR overestimated HR OS–derived CBS in 58%, 68%, 77%, and 55% of pairs and overall by an average of 5.62 (n=90), 6.86 (n=110), 29.81 (n=101), and 3.58 (n=108), respectively. Correlation coefficients were 0.80 (95% CI, 0.70–0.86), 0.38 (0.20–0.53), 0.20 (0.00–0.38), and 0.01 (–0.18 to 0.19) for mOS-, HR PFS–, mPFS-, and RR-derived CBS, respectively, and mean absolute errors were 11.32, 12.34, 40.40, and 18.63, respectively. Conclusions: Based on the ASCO-VF algorithm, HR PFS–, mPFS-, and RR-derived CBS are suboptimal surrogates, because they were shown to be biased and poorly correlated to HR OS–derived CBS. Despite lower weighting than OS in the ASCO-VF algorithm, PFS still overestimated CBS. Simple rescaling of surrogate endpoints may not improve their validity within the ASCO-VF given their poor correlations with HR OS–derived CBS.