Elsevier

Journal of Cancer Policy

Volume 12, June 2017, Pages 16-20
Journal of Cancer Policy

Original Research
High-cost cancer imaging: Opportunities for utilization management

https://doi.org/10.1016/j.jcpo.2016.12.004Get rights and content

Highlights

  • We examined the effect of patient and provider characteristics on 104 oncologists’ imaging behaviors over 15 years at an academic cancer center.

  • Imaging utilization variations were not based solely upon patient factors and treatment intensity, but also upon provider ordering preferences.

  • Total imaging utilization was higher among patients with multiple ordering oncologists.

  • Imaging utilization tended to increase as patients’ temporal proximity to death decreased, declining slightly in the final 3 months of life.

  • Our findings suggest improved coordination of care practices may reduce duplicative imaging from multiple ordering oncologists.

Abstract

Objectives

To retrospectively evaluate utilization of high-cost cancer imaging to clarify the extent to which variations in provider preferences drive imaging utilization.

Study design

At a United States academic cancer center, 4,605 patients were identified with 29,740 oncologist ordered tomographic imaging studies. Patients’ dates of death ranged from January 2000 through December 2014. Imaging was restricted to CT, MR, and PET/PET-CT. Outcome variables were total imaging per patient and total imaging per patient by a single oncologist. The number of ordering oncologists per patient, patients receiving imaging in the final year of life, and patients receiving imaging ordered by a high-ordering oncologist were the predictors of interest.

Methods

Zero-truncated negative binomial regressions were used to model collective and individual oncologist per patient imaging counts, with the exposure period defined as the number of days from diagnosis to death.

Results

Patients with imaging ordered by one of the high-ordering medical oncologists predicted nearly a two-fold increase in total images from diagnosis to death (IRR, 1.93; 95% CI, 1.67–2.23). Increasing numbers of providers (2, 3, 4+ ordering oncologists) were associated with greater collective per patient imaging (IRRs 1.65, 2.19, 2.33). Mean imaging intensity increased in a linear manner as temporal proximity to death decreased, from 12 months pre-mortem to death, and imaging in the final year of life was associated with greater per patient imaging (IRR, 0.25; 95% CI, 0.23–0.27).

Conclusion

These findings suggest heterogeneous provider ordering preferences and lapses in care coordination are drivers of high-cost cancer imaging utilization.

Introduction

The costs of cancer care are escalating, both domestically and internationally, calling into question the long-term sustainability of health systems and health care practices in high-income countries [1], [2]. The escalating costs of diagnostic technologies have continued to exceed increases in other cancer care expenditures, drawing greater attention to utilization patterns of high-cost tomographic imaging [3], [4], [5]. The extent to which benefits are derived from greater rates of imaging utilization has not been studied extensively in oncology, but studies have generally been unable to confirm benefits from increased tomographic imaging and associated treatments during the end of life [6], [7]. Given that one-third of end-of-life cancer care expenditures concentrate in the last month due to escalating treatment aggressiveness [8], exploration of imaging intensity within this period may reveal opportunities to improve healthcare value through the elimination of imaging that does not lead to improved outcomes.

Section snippets

Patients and methods

We conducted a retrospective analysis of cancer patients’ tomographic imaging utilization from diagnosis to death at an academic comprehensive cancer center in California to explore variations in oncologist ordering. Collective and individual oncologist imaging rates per patient were the outcomes of interest. Collection of data for this study was approved by the institutional review board and did not require patient consent.

Descriptive statistics

The patient population was comprised of 4,605 patients with 29,740 tomographic imaging studies ordered by oncologists. CT accounted for 67.5% of tomographic imaging (n = 20,083); MR accounted for 19.5% (n = 5,782); and PET accounted for 13% (n = 3,875). Patient demographics and cancer characteristics may be found in Table 1. Table 2 provides frequencies of surgery, radiation, and chemotherapy treatments. All patients had at least one of the three treatment types. Mean tomographic imaging intensity

Discussion

In this 15-year retrospective analysis of imaging utilization among 104 oncologists, we found indications that oncologists’ ordering patterns were highly heterogeneous. There was substantial variation in imaging utilization both between and within oncologic specialties. High-ordering oncologists were found to be substantial drivers of imaging utilization, with 58% more per patient imaging than their peers. Greater total collective per-patient imaging in larger combinations of ordering

Conclusions

Variability in tomographic imaging utilization is not based solely upon patient factors and treatment intensity, but is dependent on provider ordering preferences. An inverse relationship between increasing imaging intensity as temporal proximity to death decreases implies decision support around end-of-life treatment efficacy is lacking.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Acknowledgements

We would like to thank Ann Griffin and Joseph McGuire, of the UCSF Helen Diller Family Comprehensive Cancer Center Cancer Registry; Christopher Jovais, of the UCSF Radiology Information Systems; and Dana Ludwig, of the UCSF Academic Research Systems.

References (20)

There are more references available in the full text version of this article.

Cited by (0)

View full text