ORLANDO – Accumulating evidence is helping researchers better understand why errors occur during the delivery of cancer treatment and how to prevent them. Findings from a trio of studies were reported at a symposium on quality care sponsored by the American Society of Clinical Oncology.
Identifying causes of incidents in radiation therapy
In the first study, Greg D. Judy, MD, a radiation oncology resident at the University of North Carolina at Chapel Hill, and his colleagues retrospectively reviewed records in their institution’s reporting system, called Good Catch, to identify near-miss incidents (ones that didn’t reach the patient) and safety incidents (ones that did) among patients undergoing radiation therapy from October 2014 through April 2016.
Susan London/Frontline Medical News
Dr. Greg D. Judy
Multivariate analysis showed that patients had a significantly higher risk of near-miss or safety incidents if they had stage T2 disease (odds ratio, 3.3), were being treated for cancer involving the head and neck (5.2), or were receiving image-guided intensity-modulated radiation therapy (3.0) or daily imaging as part of their treatment (7.0), Dr. Judy reported.
“Head and neck site and image-guided IMRT [intensity-modulated radiation therapy] are complex entities: They have multiple steps in both the planning and delivery phase,” he said. “Daily imaging as well. It’s a much more complex process to do daily imaging for setup verification than it is to do once-a-week or even once-every-2-weeks setup verification.”
On the other hand, it was unclear why T2 stage was a risk factor. “We kind of hypothesized that it might be more of the disease site that really drives this, as you can have HPV-positive oropharyngeal cancer which usually has lower T stages and more advanced nodal stages, but even then, that’s a head and neck site and we usually use image-guided IMRT, which are both very complex entities,” he said. The most common root causes for the incidents were issues related to documentation and scheduling (29% each), followed by issues related to communication (22%), technical treatment planning (14%), and technical treatment delivery (6%).
Incidents having a communication root cause were more likely than were others to affect patients (P less than .001), and those having a technical treatment delivery root cause were more likely to have higher severity (P = .005).
“Like some other studies, we found really the key factor was the complexity of the treatment plan and complexity of the overall process that is the real driving factor. This is important to understand because it promotes the idea of developing a more dedicated and robust QA system for complex cases,” said Dr. Judy. “It also highlights the importance of a strong reporting system to support a safety culture, as well as promote the continuous learning improvements within a department.”
The national Radiation Oncology Incident Learning System (RO-ILS) has been developed by the American Society for Radiation Oncology (ASTRO). “This is gaining membership very rapidly, and it’s good because it facilitates cooperative research and also safety standards for our field,” he maintained.
Susan London/Frontline Medical News
Dr. Joseph O. Jacobson
“I’ll argue that they did that for a variety of reasons,” he elaborated. “Strong and effective leadership by Dr. Larry Marks, who’s really created a departmental culture of safety in which people can feel free to speak up. They have this wonderful Good Catch program in place. And they have these simulation review huddles ... where people feel free to talk about what happened yesterday or today that may be relevant moving forward.”
As for the national RO-ILS initiative, “I would look out to the audience and say, why is it that we don’t have such a program in medical oncology?” Dr. Jacobson said. “It’s probably time for us to do this,” he maintained.
Reducing chemotherapy errors in pediatric oncology
In the second study, Brian D. Weiss, MD, associate director of safety and compliance at Cincinnati Children’s Hospital Medical Center, and his colleagues studied the impact of a safety initiative to prevent chemotherapy errors at their large urban pediatric academic center (J Clin Oncol 35, 2017 [suppl 8S; abstract 37]).
“Pediatric chemotherapy protocols are different from adult protocols. We dose based on age or weight or body surface area, and that can change within a protocol. You have to do adjustments every time they gain weight or grow some, which is different than for adult protocols,” he explained. “We have parents administering chemo at home. And the protocols most patients are on are very complicated, but there is no standardized format, so it makes crucial information for dose adjustments difficult to find.”
Susan London/Frontline Medical News
Dr. Brian D. Weiss
The team successively implemented about half a dozen interventions, such as dedicated chemotherapy safety zones where staff were not to be disturbed while checking orders and ear protectors as a visual deterrent to interruptions; a new chemotherapy registered nurse role with a detailed list of responsibilities; an event-reporting system to supplement the center’s error-reporting system and capture events not reaching the patient (near-miss events); and a daily chemotherapy huddle to discuss errors in a nonpunitive setting and review upcoming chemotherapy for readiness.
In the 6 years after the start of the initiative, 105,187 chemotherapy doses were administered at the center and 998 errors occurred, including 250 errors that reached the patient, according to results reported at the symposium and recently published (J Oncol Pract. 2017 April;13:e329-e336).
At the 22-month mark, the rate of chemotherapy errors reaching the patient had fallen from 3.8 in 1,000 doses at baseline to 1.9 in 1,000 doses. The reduction has since persisted for more than 4 years and translates to an estimated 155 fewer predicted errors reaching the patient because of the initiative.
“The errors that reached the patient were more often administration and dispensing errors,” Dr. Weiss said. “About two-thirds of those errors that didn’t reach the patient – because they got caught by the pharmacists and the nurses – were prescription errors.
“Our chemotherapy huddle has certainly increased our reporting of errors. We also now use it for patients on clinical trials ... any patient getting PK [pharmacokinetics] or PD [pharmacodynamics] sampled within the next 24 hours is reviewed at that meeting. And our missed samples have gone down significantly,” he noted. “It’s allowed us to manage our bed space better because now everybody knows who’s definitely coming in the next day and who’s maybe coming in the next day.”
“We are a large urban academic pediatric medical center. Some of these things may seem difficult to translate [to smaller facilities], but I’m not sure they are,” concluded Dr. Weiss.
Dr. Jacobson, the discussant, noted that the initiative was in keeping with this health system’s longstanding “obsessive” focus on patient safety and commended its rigorousness in, for example, setting clear goals, focusing on key drivers [processes] needed for change, and selecting a good outcome metric.
“This is very successful project,” he said. The success can be attributed to “strong and effective organization and leadership, building a culture of safety at Cincinnati Children’s Hospital, and an important predefined measurement program and methodology.”
Building chemotherapy regimens more accurately
In the third study, a team led by Andrea Crespo, BSc, BScPhm, BCOP, an oncology pharmacist and member of the Systemic Treatment Team at Cancer Care Ontario, Toronto, studied errors introduced when chemotherapy regimens were moved from publications into orders used by centers in the province (J Clin Oncol. 35, 2017 [suppl 8S; abstract 51]).
Nearly all outpatient intravenous systemic treatment visits in Ontario are supported by a Systemic Treatment Computerized Prescriber Order Entry (ST-CPOE) system, she noted. Such systems can reduce error rates but are not foolproof.
Susan London/Frontline Medical News
Andrea Crespo
She and her colleagues asked all Ontario treatment centers to review their active chemotherapy regimens. Data were analyzed to determine whether the regimens were built as intended with respect to their component drugs and doses, leading to identification of any unintentional discrepancies with the original regimen.
A total of 33 centers performed the review, and the median number of regimens reviewed was 375 per center, Ms. Crespo reported.
Unintentional discrepancies in regimens were found at 27% of centers. The total number reported was 369 discrepancies, with a range from 2 to 198 per center.
All of the nine centers where discrepancies were found participated in the provincial ST-CPOE system, and most had for at least 20 years. Furthermore, eight of them used a team of at least two pharmacists and one oncologist to build their regimens. “So you can see that discrepancies occurred despite a fairly rigorous regimen-build process and many years of experience with the system,” she said.
Of the 369 total discrepancies, 41% were related to alignment with the Systemic Treatment Quality-Based Program regimen, and 32% were regimens flagged to be inactivated because of outdated information, new standards, or lack of use.
A detailed analysis of the remaining 27%, or 101 unintentional discrepancies, showed that the majority were due to missing information (35.6%) or missing drugs (13.9%), incorrect doses (10.9%), and incorrect or missing schedules (10.9%). Potential to cause harm was mild for 55%, moderate for 28%, and none for 17%.
“Corrective action has been taken to address the discrepancies identified,” said Ms. Crespo.
Only 6% of the 33 centers reported having an established regimen review and maintenance process in place before the study, but all now have such a process. In addition, some centers that did not find any regimen discrepancies nonetheless reported adding quality improvement activities, such as changes in the ways regimens were built and documented, and revising regimen names to facilitate accurate selection.
In discussing the study, Dr. Jacobson noted the low proportion of centers having an established process at baseline to ensure appropriate regimen maintenance and updates. “You might want to think to yourselves, the medical oncologists in the group, whether your center has such a process in place,” he proposed.
It is not yet known whether the project has met its goal of improving the quality and accuracy of oncology regimens in Ontario, he maintained. “We are going to have to invite [Ms. Crespo] back in a year or two to see whether that turns out to be true.” On the other hand, “clearly what they have achieved was the ability to measure the variance between what was intended and what was actually built.”
Chief among the reasons for success, again, “was a strong and effective leadership and organizational structure, not at the department level or hospital level, but across the entire province through Cancer Care Ontario,” Dr. Jacobson said. “It’s clear that they have a focus on quality and patient safety, and this measurement program that they have put in place turned out to be useful.”
Dr. Judy, Dr. Weiss, and Ms. Crespo disclosed that they had no relevant conflicts of interest.