Introduction

During the COVID-19 pandemic, expeditious dissemination of education on the manifestations, effect, and treatment of COVID was paramount, including in psychiatry. Clozapine is the gold standard for treatment-resistant schizophrenia, and case reports of COVID-19–associated clozapine toxicity evoked serious patient safety concerns in the clinician community.^1-3 Symptoms of clozapine toxicity include sedation; sialorrhea; speech and gait disturbances; and more seriously, hypotension, delirium, and seizures. Toxicity is likely mediated by elevation of clozapine concentrations via significant inhibition of hepatic cytochrome P450 (CYP) isoenzymes, primarily CYP1A2 and, to a lesser degree, CYP3A4/5, CYP2C19, and CYP2D6.⁴ The most commonly prescribed oral COVID-19 treatment, nirmatrelvir/ritonavir, strongly inhibits CYP3A4, leading to early concerns about clozapine toxicity and initial contraindication of coprescription of the 2 agents.⁵ Additionally, COVID-19 infection itself inhibits clozapine metabolism and may amplify the effect of nirmatrelvir/ritonavir. In working with our patients and colleagues, it was clear that this critical information on clozapine toxicity, COVID-19, and nirmatrelvir/ritonavir was not widely known. Therefore, the goal of this quality improvement (QI) initiative was to disseminate needed safety information on clozapine toxicity and these potentially important interactions to increase awareness, improve patient safety, and optimize management of clozapine in the setting of an active COVID-19 infection.

Methods

This project was a multistep, multidisciplinary, structured, clinical QI project focused on providing interdisciplinary health care professionals, patients, and patient caregivers education on the risks of clozapine toxicity and the potential drug-drug interaction between nirmatrelvir/ritonavir and clozapine during the COVID-19 pandemic. Targeted clinicians included physicians, pharmacists, advanced practice providers, and nurses. Patients/caregivers were, in part, targeted to bridge knowledge gaps in which clozapine is not frequently prescribed and psychiatric services are not consistently embedded, for example, in primary care offices, urgent cares, and emergency departments. The project was approved by the University of Pittsburgh Medical Center (UPMC) Quality Improvement Review Committee and, thus, no separate institutional review board approval was needed.

Results

The slide set was emailed to 310 clinicians in our health care system and to 738 in the state hospital system. The 2 CE videos had a combined 329 views by December 7, 2023. The UPMC WBH Schizophrenia Conference had 304 attendees; the Quality Improvement Fair had more than 190. Please see the Table. The CE evaluations from the UPMC WBH Schizophrenia Conference were available for 165 attendees and incorporated 2 key questions. For the first question, “As a result of participating in this CE activity, to what degree were you able to evaluate the medical and nursing management of clozapine toxicity using the illustrative case?”, 97.6% rated their understanding as “very high” or “high” and 2.4% as “moderate” (Figure). For the second question, “As a result of participating in this CE activity, to what degree were you able to understand the significant precautions of using Paxlovid (nirmatrelvir/ritonavir) in clozapine-treated patients who experience mild to moderate COVID-19?”, 95.6% rated it “very high” or “high,” 3.6% as “moderate,” and 0.80% as “low” (Figure). Summary reports were available from 1 CE web-hosted platform. Among the 76 participants who answered the above question 1, 77.6% rated it as “very high” or “high,” 20.7% as “moderate,” and 1.7% as “low,” and for question 2, 75% said “very high” or “high,” 24% said “moderate,” and 1% said “low” (Figure). In comparing question 1 results between the live UPMC WBH Schizophrenia Conference and the recorded CE videos, results were significantly different with “very high” or “high” ratings being more common among conference attendees (Fisher exact test P < .00001). Significant differences also were observed for question 2 with more conference attendees giving “very high” or “high” ratings (Fisher exact test P < .00001). The SMI Adviser clinical tip received more than 843 views by January 2024.

TABLEResults based on local, regional, and national dissemination efforts

View Fullscreen

View Figure

• Download as Powerpoint
• Download Figure

Phone counseling of 149 (86.6%) clozapine clinic patients/caregivers was completed via the project workflow as described. Clozapine clinic patients/caregivers who were not successfully contacted by phone (n = 23, 13.4%) were sent letters. An additional 59 out of a total 66 clozapine patients/caregivers not served by the clozapine clinic, but by other office- or community-based CRS services, received verbal counseling as part of standard care delivery. Clinicians working in these settings chose to provide the project-recommended patient education and use the documentation autotext as part of routine clozapine care delivery, demonstrating successful practice change without additional workflow support. In total, verbal counseling of 208 clozapine patients/caregivers was completed as confirmed by SQL query generated report of autotext usage.

Discussion

Rapid dissemination of education on clozapine toxicity and putative interaction with COVID-19 infection and treatment was needed during the pandemic to help ensure continued safe prescribing of this critical medication. Combined local, regional, and national efforts led to education of more than 2000 health care professionals and more than 200 local patients/caregivers, all of whom helped promote shared decision making and patient safety during a time of strained health care resources. Clinicians rated their understanding of the topics as quite high after engagement with project-generated CE. Interestingly, statistically significant differences in learner evaluations were observed between the virtual schizophrenia conference and the prerecorded CE videos, suggesting that the schizophrenia conference may have more optimally achieved its learning objectives. During the pandemic, opportunities in synchronous and asynchronous medical education grew exponentially with some studies suggesting superiority of synchronous learning,¹¹ including in case-based style approaches as were used in this project.¹² This may at least partially account for differences in perceived learner benefit between modalities. Furthermore, EHR optimization through autotexts and of drug-drug interactions checkers further enhanced education and safety. Given these findings, our efforts, as outlined, may be used as a road map for expedient patient, caregiver, and continuing medical education during pandemics and similar emergencies.

There are a few limitations of this project. Given the need to quickly integrate this information into practice and the challenges of the pandemic, we did not collect pre-post measures of knowledge, nor could we objectively track COVID-19-related changes in clozapine dose that occurred subsequent to knowledge gain. Evidence of practice change was limited to capturing use of the autotext. High rates of patient/caregiver engagement may be a product of the high-intensity, well-resourced clozapine clinic model. However, clozapine-prescribing clinicians and nurses not affiliated with the clozapine clinic did educate patients and caregivers without additional staff support. This supports the principle that the uptake of novel clinical interventions can be best facilitated when perceived as both clinically important and efficient.

Conclusion

This QI project led to the successful, rapid education of health care professionals, patients, and caregivers about the risks of clozapine toxicity associated with COVID-19 infection and potential drug-drug interactions between nirmatrelvir/ritonavir and clozapine. Professional education and EHR workflow change was demonstrated through a model that may be modified to support other practice changes that require rapid adoption.