Variation in non-recommended services for cancer patients suggest physician-driven factors

1. A retrospective study of non-recommended services in patients 66 years or older with cancer demonstrated significant inter-physician variation in utilization.

2. Variation in use of non-recommended services were more likely if the physician had ordered the services for previous patients, suggesting physician-driven rather than patient-driven factors as the cause of the variation.

Evidence Rating Level: 2 (Good)       

Study Rundown: Overuse of health care services despite evidence against their utility can lead to increased cost and potential harm. The Choosing Wisely Campaign has sought to highlight commonly misused services in order to promote discussion between physician and patient to reduce unnecessary medical tests. Understanding the decision making process that leads to the use of non-recommended services can help target further interventions to reduce waste. The following study sought to evaluate if non-recommended services in cancer care were consistently overused by certain physicians and if physician-driven rather than patient-driven factors explained the variation.

The study used the Surveillance, Epidemiology, and End Results (SEER) and Medicare databases to explore how use of 5 non-recommended services (as outlined in Choosing Wisely Campaign-produced lists from relevant specialty groups). There was unexplained variation not due to chance for each of the studied services. Use for each service was shown to be more likely if the physician had ordered the service for a prior patient. The results of the study demonstrated consistency across various services and with prior studies, lending further credibility. However, as this was a large database study and the difference identified was small, the clinical relevance is difficult to quantitate.

Click to read the study, published today in JAMA Internal Medicine

Relevant Reading: Post-Treatment Surveillance in Locoregional Breast Cancer: Guideline Adherence and Patterns in Use of Non-Recommended Testing

In-Depth [retrospective cohort]: This study used linked SEER and Medicare databases to evaluate patients aged 66-years and older diagnosed with new cancer between 2004 and 2011. Patients were excluded if they were enrolled in a Health Maintenance Organization (HMO), had incomplete Medicare coverage, or presented with a cancer diagnosis at death. Physicians were excluded if they only had 1 patient in the database. The services examined were derived from the American Society of Clinical Oncology, and American Society for Radiation Oncology proposed lists of non-recommended services as part of the Choosing Wisely Campaign. Services included: staging imaging for early prostate/breast cancer, surveillance imaging for asymptomatic patients after treatment for low-risk breast cancer, intensity-modulated radiation therapy (IMRT) for whole breast radiotherapy, and extended (>10 fractions) fractionation schemes for palliation of bone metastases.

In random effects analysis, the intraclass correlation for each service was 0.04 to 0.59 with p < 0.001 for each investigated service suggesting non-random variance between physicians. The adjusted odds ratios that a physician would order a service if they had previously done so for a past patient was significant for each service: 1.12 (95%CI 1.07-1.18) for breast cancer surveillance screening, 1.48 (95%CI 1.26-1.75) for extended fractionation schemes, 3.02 (95%CI 2.88-3.17) for staging imaging in early-stage breast cancer, 3.90 (95%CI 3.70-4.10) for staging early-stage prostate cancer, and 24.91 (95%CI 22.86-27.15) for IMRT in breast cancer therapy. Finally, greater perceived inappropriateness for a service (e.g., imaging for stage 0/I compared to stage II breast cancer) ordered for a prior patient predicted increased odds of ordering the service for a future patient.

Image: PD

©2016 2 Minute Medicine, Inc. All rights reserved. No works may be reproduced without expressed written consent from 2 Minute Medicine, Inc. Inquire about licensing here. No article should be construed as medical advice and is not intended as such by the authors or by 2 Minute Medicine, Inc.