Diagnosis and treatment of neuropathy postchemotherapy in three health systems [abstract] Abstract uri icon

abstract

  • Background: We examined factors associated with peripheral neuropathy (PN), a common side effect of chemotherapy, and PN treatment. Methods: This observational study used data from Kaiser Permanente in Northern California and Washington and from HealthPartners in Minnesota. We included active (≥1 in-person visit) enrolled members ≥40 years of age with diagnosis of invasive cancer (January 1, 2010, to December 31, 2014) and who initiated select chemotherapies (eg, platinums, taxanes) within 12 months of cancer diagnosis. We excluded patients previously diagnosed or treated for PN. Patients were followed until death, disenrollment, new cancer diagnosis, or study end (December 31, 2016). The primary outcomes were time to clinical diagnosis of PN from start of chemotherapy and time to treatment after PN diagnosis. We calculated cumulative incidence curves accounting for competing risks and used Cox proportional hazards models to identify potential risk factors. Results: We identified 13,058 patients, 23.1% (95% CI: 22.4%–23.8%) of whom received a PN diagnosis within 1 year of chemotherapy. Factors associated with higher rates of PN diagnosis included black race (hazard ratio: 1.23 [95% CI: 1.10–1.38]), diabetes (1.14 [1.05–1.25]), chronic pain (1.25 [1.17–1.33]), nonchemotherapy medications associated with PN (1.15 [1.06–1.24]), comorbidity (1.19 [1.10–1.29]), and concurrent platinum and taxane use compared to platinum alone (1.29 [1.18–1.42]) or taxane alone (1.23 [1.12–1.34]). PN rates also varied by health system and age (lower rates at 40–49 and ≥70 years). One-third of patients with a PN diagnosis (95% CI: 30.9%–33.9%) started new PN treatment within 1 year. Treatment rates were associated with black race (1.24 [1.03–1.50]), chronic pain (1.26 [1.12–1.40]) and recency of cancer diagnosis. Rates also varied across health systems. Conclusion: Type of chemotherapy, comorbidity, race, and ethnicity were associated with higher rates of PN. Yet, variations in PN treatment were not adequately explained by clinical risk factors. Additional studies are needed to begin understanding observed variation in PN diagnosis and treatment and to identify strategies for improving patient care and outcomes.

publication date

  • 2019