Practice-based evidence to evidence-based practice: building the National Radiation Oncology Registry.

The National Radiation Oncology Registry (NROR), sponsored by the Radiation Oncology Institute and the American Society for Radiation Oncology, is designed to collect standardized information on cancer care delivery among patients treated with radiotherapy in the United States and will focus on patients with prostate cancer. Stakeholders were engaged through a forum that emphasized the need for patient-centered outcomes, minimal data burden, and maximal connectivity to existing registries and databases. An electronic infrastructure is under development to provide connectivity across radiation oncology and hospital information systems. The NROR Gateway features automatic abstraction as well as aggregation of treatment and outcome data. The prostate cancer data dictionary provides standardized elements in four domains: facility, physician, patient, and treatment. The pilot phase will consist of clinical centers chosen to provide a representative mix of radiation treatment modalities, facility types, population-based settings, and regional locations. The initial set of radiation practice metrics includes physician board certification and maintenance, ordering of staging scans, active surveillance discussion, dose prescriptions for low-risk/high-risk disease, radiation fields for low-risk/high-risk disease, image-guided radiation therapy use, androgen deprivation therapy use, post-brachytherapy implant computed tomography dosimetry, collection of toxicity assessments, and longitudinal patient follow-up. The NROR pilot study will provide the framework for expansion to a nationwide electronic registry for radiation oncology.

The impact of computerized clinical reminders on physician prescribing behavior: evidence from community oncology practice.

The objective of this study is to examine the impact of a clinical reminder generated by an electronic medical record (EMR) system on physician prescribing behavior in community oncology practice setting. A case-control trial assessing the prescribing rates of erythropoietin by physicians is used. The participants and setting involves a total of 11,644 physician-patient encounters in 2 community oncology practices in the United States during a 21-month period. The intervention is a clinical reminder generated in real time during a physician-patient encounter by an EMR identifying cancer patients with low hemoglobin (Hgb) levels (ie, anemic), that is, patients with Hgb less then 12 g/dL. The main outcome measure is to determine the frequency of erythropoietin prescription by physicians to cancer patients with low Hgb levels. Implementation of a clinical reminder generated by way of an EMR significantly improved the likelihood of low-Hgb patients receiving treatment with erythropoietin. Low-Hgb patients in the experimental clinic during the time that the clinical reminder system was in place were almost twice as likely (ie, adjusted odds ratio = 1.92, P = .008) to have been treated with erythropoietin. The data support the effectiveness of clinical reminders as a way to influence physician prescribing behaviors and potentially improve the quality of patient care. However, we feel that there is a need to investigate the use of reminders in other aspects of cancer care that may be undertreated or when new drugs may be available but are underused.

Antidepressant prescribing in community cancer care.

GOALS: To describe patterns of antidepressant (ADs) prescribing in community oncology practice. PATIENTS AND METHODS: Data were collected using an electronic medical record on all staged breast, colon, and lung cancer patients in three community-based oncology practices. The data were analyzed retrospectively, using descriptive and bivariate analyses and multivariate logistic regression modeling. There were 850 breast, 299 colon, and 473 lung cancer patients identified in this analysis. MAIN RESULTS: Overall, 19.2% of breast, 11% of colon, and 13.7% of lung cancer patients had been prescribed ADs during the 2-year period. The clinic in which cancer treatment was received predicted AD prescribing. The relationship between AD administration and age proved to be nonlinear; the pattern exhibited an "inverted U" shape. Patients with comorbidities and on pain medications were more likely to be administered ADs. Colon cancer patients on pain medications were five times more likely to be administered ADs than those not on pain medications. CONCLUSIONS: While some predictors of AD prescribing appear to be consistent with other studies, such as being on pain medication, there is still a great amount of variability in prescribing patterns across community practices, age groups, and cancer diagnoses. This study demonstrates that prescriptions of ADs seem to be influenced by parameters other than psychopathology. Given the importance of major depression in oncology care, diagnosis of psychiatric disorders and prescription patterns of psychotropics should be part of the routine monitoring and quality management in oncology patient care.