According to the results of a study published in the Journal of Oncology Practice, oncologists’ participation in a project involving serial simulations of patients with metastatic lung cancer and early-stage breast cancer was significantly associated with the provision of evidence-based care.
Approaches to reduce variations in clinical practices across oncology providers are being increasingly recognized as a means of improving the quality of care for patients with cancer. Described in this article are the structure and results of a project involving the use of serial patient simulations, known as clinical and performance (CPV) vignettes, designed to evaluate clinical decision making for patients with advanced lung cancer or early-stage breast cancer.
In this study, 103 medical oncologists from a large clinically integrated network with experience treating patients with solid tumor cancers voluntarily agreed to participate in 2 simulated patient platforms administered every 4 months for a period of 1 year between 2016 and 2017. A total of 12 different patient cases (6 cases with early-stage breast cancer and 6 cases with advanced lung cancer, respectively) were randomly assigned to a participant for each round of the study, and required only 6 hours of time to complete.
These CPVs were web-based and assessed oncologists’ decisions in a number of clinical domains including history taking, physical examination, ordering of diagnostic tests, establishing a diagnosis, treatment approaches, and recommendations for follow-up.
Responses were open-ended, did not involve answering multiple choice questions, and were scored (0% to 100%) by trained physician abstractors. Feedback on each case was provided to the study participant in the form “an overall quality score, benchmarked peer performance … quality improvement opportunities, relevant guideline references, and a detailed list, with costs, of unnecessary tests ordered.” In addition, participation in telebroadcast discussions focusing on areas of greatest clinical variation was available to study participants.
At baseline, participating oncologists had an average age of 51.9 years, an average of 18 years in practice, and 90% of them reported seeing more than 40 patients each week. Notably, response data from the first CPV case regarding the work-up of early-stage breast cancer were validated by the observation of a high degree of concordance between real-world clinical practice patterns.
For the study as a whole, responses to the first CPV simulation revealed a high level of variability across participating oncologists; average CPV quality scores were 61.5% overall and 48.7% for the treatment domain.
Notably, however, a comparison of response results for the first 2 and final 2 CPV cases revealed an overall increase in CPV quality score of +10.7% (P <.001) and +12.9% (P <.001) for breast and lung cancer cases, respectively.
Specifically, inclusion of palliative care in the care plan of a simulated patient, a particularly relevant approach for patients with metastatic lung cancer, increased from 39.6% at baseline to 61.6% at the end of the study (P <.001). Nevertheless, palliative care and end-of-life needs were not addressed for approximately 40% of simulated patients who would have been eligible for this approach.
Another study finding demonstrated that adoption of programmed cell death-ligand 1 (PD-L1) therapies in the setting of metastatic lung cancer rose sharply following inclusion of this recommendation in the National Comprehensive Cancer Network (NCCN) guidelines during the course of the study.
“These data show a clear and significant (P <.001) gradient in tumor PD-L1 testing, which demonstrates a quick, but not complete, adoption rate in the 5 months after the release of the guidelines,” the study authors commented.
Despite this positive result, however, responses in favor of tumor PD-L1 testing were 67.3% at the end of the 6th simulated patient case, again demonstrating the need for additional education.
In the setting of asymptomatic early-stage breast cancer, ordering of a metastatic work-up, dropped from 31.5% at baseline to 21.1% by the end of the study, a finding that had considerable cost implications.
Some of the study limitations identified by the study authors included the absence of a control group as well as only limited real-world data for comparison to simulated patient case responses.
In their concluding remarks, the study authors commented that “this study shows that patient simulations are a valid measure of practice patterns in oncology that, when administered serially with individual and group feedback, lead to more evidence-based, cost-effective care.”
Richards JM, Burgon TB, Tamondong-Lachica D, et al. Reducing unwarranted oncology care variation across a clinically integrated network: A collaborative physician engagement strategy [published online October 1, 2019]. J Oncol Pract. doi: 10.1200/JOP.18.00754