Impact of quality assurance and feedback on radiotherapy prescribing practices: a randomized controlled trial

Purpose/Objective(s)

Clinical Audit and feedback (CA&F) is a demonstrated successful knowledge translation (KT) method that positively impacts change in practice in numerous clinical scenarios. However, there has been low implementation of CA&F in oncology, and specifically in radiotherapy (RT) prescribing practices. We sought to identify the impact of a simple CA&F intervention on prescribing practices across six BC Cancer Centers by investigating four clinical scenarios: (1) women <50 years old with invasive breast cancer post partial mastectomy who receive surgical cavity RT boost vs not; (2) patients with 1-3 positive lymph nodes receiving regional RT vs not; (3) women >70 years old with T1N0 breast cancer omitting whole breast RT vs not; (4) all women who received local or locoregional radiotherapy who are prescribed 16 versus 25 fractions of RT.

Materials/Methods

Thirty-five radiation oncologists were randomized to receive CA&F intervention vs not and stratified by BC Cancer center. Patient data was obtained retrospectively for patients treated before and after intervention. The study measured one of two possible prescriptions per clinical scenario, with prescription 1 containing the most evidence in favor of its use. A logistic fixed effect regression model was used to measure odds ratio of prescription 1 before and after intervention. This model captured the difference in prescription use within and between control and intervention groups pre and post CA&F, while adjusting for the number of patients seen by each physician. Time as a potential confounding variable that could impact prescription rates from pre to post intervention was also adjusted for.

Results

A total of 5215 unique cases were identified. The change in odds of using prescription 1 after CA&F in cohorts 1, 2, 3 and 4 was 0.45 (p=0.12), 0.59 (p=0.34), 2.23 × 10⁻⁹ (p=0.98, unstable model), and 0.77 (p=0.11) respectively.

Conclusion

There was no significant change in odds of utilizing the more evidence-based prescription across any of the four clinical scenarios, suggesting our simple CA&F intervention was not effective in changing prescribing practices. Further, radiation oncologists may have been optimally informed of prescribing patterns prior to intervention. A more complex CA&F intervention may be necessary in future quality assurance efforts to achieve a greater impact on radiotherapy prescribing practices.