In our recent article, we shared our experience with the use of a peer review strategy within a clinical radiation oncology clinic. Key components of our strategy are that it occurs before patient treatment and that it involves a group consensus approach of not just physicians, but also physicists and dosimetrists.
The argument for a serious approach to peer review, despite the cost, is strong and supported by many professional organizations such as the “Zero Harm” movement by the Joint Commission on Accreditation of Healthcare Organizations and also various radiation oncology-specific accrediting bodies such as the American College of Radiology (ACR), the American Society of Therapeutic Radiation Oncology (ASTRO), and the American College of Radiation Oncology (ACRO).
Preventing errors in radiation oncology – peer review
Clinical peer review is not that different from the review of published literature
In 2010, a series of New York Times articles brought to light the potential seriousness of errors in radiation oncology. A patient died as a result of accurate machine parameters not being used. In hindsight, the error would have been preventable had the treatment been tested on the machine before the first patient treatment, instead of within the first few treatments as was common practice.
Today, all plans of this type are required to be validated on the machine before a patient’s first treatment. This new policy would have prevented this deadly error if it had been common practice sooner. However, ensuring the quality of a patient’s treatment plan is much more than just validating the plan on the machine before treatment.
For ensuring quality of the clinical and technical decisions during patient treatment, many clinical departments have commonly turned to peer review from colleagues. Peer review in the clinical environment is not that different from that used in the review of published literature. Reviewers in both cases make sure that appropriate procedure has been used, the message/strategy is appropriate and clear, and that the decisions made are adequately supported by the already known evidence.
Perhaps the primary difference is that with literature, plagiarism is a concern for reviewers, whereas peer review in the clinical environment would look for treatment plan designs to closely resemble the known standard of care and current evidence.
Increasing the number of trained judgments that a clinical decision is appropriate and optimal has been recognized for its value and is difficult to dispute. The unique approach in our clinic is the utility of a pre-treatment timepoint and group consensus review. Peer review is not infallible, but for the instances when errors are caught, if no treatment has yet been delivered, then the error can be proactively corrected before treatment, rather than only mitigated.
The value of the group-consensus approach is that a higher volume of trained expertise is used during plan review and can include both clinical and technical concerns for the quality of the plan. We are entering an age where human limitations are increasingly evident when each cancer patient requires an average of about 7 GB of data and a lung radiation oncologist would likely have to read on the order of 12 papers per day to stay up to date with evidence-based practices. Adding more expertise to the room can mitigate some of these limitations.
Overcoming peer review barriers
In order to implement our pre-treatment, group consensus peer review approach, we have overcome many practical barriers. The first step is investment into “safety culture” where safety instead of speed is the most important. This requires strong leadership and accountability over time.
Secondly, scheduling may be the next barrier. Our clinic has chosen to set time twice per week which are blocked. Lunch breaks have also been used to prevent undue delay in treatment start.
Third, standardization of practice and procedure is important so that the review of plans can be streamlined. Examples of this might be the order in which plan aspects are evaluated, display color schemes, and even naming of structures.
Fourth, tools can be developed that give quicker snapshots of relevant information about the patient and the treatment plan itself. For instance, we use a plan “scorecard” which flags coverage of prescription dose to target volumes, as well as critical structures doses compared to known, evidence-based tolerance doses.
Finally, ownership of the individuals involved is important. The faculty/staff must be invested in the process and make efforts to see it succeed such as arriving early to prepare and start on time. Though barriers exist, many of them can be overcome.
Peer review is likely to be used for high-quality patient care for many more years
Looking to the future
As we move into the age of artificial intelligence and machine learning, we may find that the nature and look of peer review may change drastically. Indeed, IBM’s Watson provides evidence-based decision support by connecting to major peer-reviewed publication journals to provide treatment recommendations for oncology patients.
Potentially, “scorecards” may become more inclusive of information while at the same time being more efficient and we may find ourselves making decisions based on more patient-specific data than previously considered (i.e. genome sequencing). Regardless, peer review is likely to be used for high-quality patient care for many more years. Despite what technology may be on the horizon tomorrow, clinical practices should move closer towards achieving “zero harm” today.