Long-acting injectable antipsychotic (LAI) prescribing trends during COVID-19 restrictions in Canada: a retrospective observational study

In response to the COVID-19 pandemic (severe acute respiratory syndrome coronavirus 2 [SARS-CoV-2] and emerging variants), most of Canada (and the world) adopted quarantine measures. These measures ranged from stay-at-home guidelines, to varying degrees of restricted movement and enforced business closures. COVID-19 has uniquely impacted those more vulnerable in society due to mental health concerns such as schizophrenia and related psychotic disorders, as well as the clinical teams who work to provide stability for these patients. During the pandemic, mental health services had to change in response to increasing levels of confinement and the related demands of being asked to work from home, including restricting face-to-face healthcare visits and opting for more telephone or video platform interviews with patients. The types of restrictions adopted by particular healthcare providers varied widely from province-to-province (and within provinces by health authority). As well, restrictions were regularly revised within regions as case numbers and hospitalizations fluctuated.

However, individuals with a psychotic disorder who were on long-acting injectable antipsychotic (LAI) therapy still needed to receive their medication, and thus service delivery had to pivot to align with the public health requirements imposed by local health authorities. In Canada, LAI injections need to be performed by a trained professional (e.g., psychiatrist, general practitioner, registered nurse, nurse practitioner), and are typically administered within psychiatric departments (inpatient/outpatient), mental health clinics (hospital/community), or while visiting with a general practitioner. While COVID-related restrictions were not uniform across these settings they typically included more home visits for delivery of LAIs, more restricted times to attend clinics to receive LAIs, limits on who (if anyone) could accompany a patient while visiting a clinic, potentially less patient interaction with clinical teams, and more patients receiving their LAI from an unfamiliar provider. Broader challenges faced by individuals with schizophrenia and related psychoses during the COVID-19 pandemic have been recently reviewed [1‐3].

Both the Canadian Psychiatric Association (CPA) and American Psychiatric Association (APA) issued guidelines urging hospitals, clinics, and other facilities to continue treatment with LAIs, as it was their view that LAI treatment should be considered a medically necessary procedure [4, 5]. Whereby, the risks associated with discontinuation of LAI medication for patients (e.g., relapse related to inadequate oral medication adherence) were seen as greater than the potential risk of exposure/spread of coronavirus related to continued clinic/home visits to monitor LAI administration. However, the literature shows there were a mixture of reactions to the restrictions imposed by the pandemic in terms of LAI prescribing practices. Some jurisdictions sided with the recommendations from the CPA and APA, choosing to continue LAI administration [3, 6], or even considered switching to compounds with longer half-life [7‐9] - whereas others took the position that patients should be switched to oral medication due fears that some patients would refuse to come into the clinic, or that patients with lower socioeconomic status (SES) would have reduced access to public transportation, thereby limiting their access to clinics [10]. Some clinics also expressed fears related to limited access to LAI medications due to restrictions, or cancellations of medical exports [11]. It is unclear however what the overall impact of these various approaches has been on LAI prescribing and delivery during the pandemic.

This study aimed to fill gaps in knowledge regarding delivery of mental health care during a pandemic by examining LAI prescribing practices during unique time frames prior to and during the pandemic. While there is some limited research on mental health effects of prolonged quarantine on overall mental health status of populations [12‐14], there does not appear to be any Canadian data on changes in prescribing practices during the pandemic; specifically, changes in LAI prescribing in schizophrenia and related psychoses. We do not know if during the pandemic LAI use was increased or decreased due to the changes in healthcare delivery and public health restrictions. Of particular interest is whether there has been change in the use of the three-monthly LAI formulation – an arguably advantageous method of medication delivery during times of diminished/changed face-to face interaction between patients and their clinical teams. Here, we examined the approach to mental health care during the pandemic in the Canadian context by examining LAI prescribing practices at a national and also at a provincial level to address some variation in COVID-19 restrictions regionally.

This retrospective observational study examined national and provincial patient-level longitudinal prescribing data from Canadian retail pharmacies from January 2019 through to December 2020, using the IQVIA Canadian Longitudinal Prescription (LRx) database [15]. This database includes anonymized, longitudinal, patient level information on approximately 72% of national pharmacy prescriptions in Canada. The second-generation long-acting injectable antipsychotic (SG-LAI) products captured by this dataset included Paliperidone palmitate (Invega Sustenna, and Invega Trinza), Risperidone (Risperdal Consta), Aripiprazole (Abilify Maintena), and first-generation long-acting injectable antipsychotics (FG-LAI) including Zuclopenthixol (Clopixol depot), Flupentixol (Fluanxol depot), Fluphenazine (Modecate depot), Haloperidol (Haloperidal LA), Pipotiazine (Piportil depot), and all relevant generics. The dataset also captured the oral second-generation antipsychotics (SGAs) Aripiprazole (Abilify), Clozapine (Clozaril), Paliperidone palmitate (Invega), Lurasidone HCl (Latuda), Brexpiprazole (Rexulti), Risperidone (Risperdal), Risperidone (Risperdal M), Asenapine (Saphris), Quetiapine (Seroquel), Ziprasidone (Zeldox), Olanzapine (Zyprexa), and related generics. Oral first-generation antipsychotics (FGA) of interest included Phenothiazines [e.g., Fluphenazine (Modecate)], as well as other oral FGAs [e.g., Loxopine (Loxapac)], and related generic formulations. The data was collected from retail pharmacies across Canada and consequently captures prescription information from across Canada. A known limitation of the chosen database includes the inability to track patients that switch between pharmacies (i.e., travelling patients). As such, patients switching pharmacies appear as unique patients. To address this limitation, patients included in this study must have had at least 30-days of data history with a pharmacy. Furthermore, patient data was excluded if age was unknown, age < 18 years, or if gender was unknown.

The database collects the payer information (including private drug plan, public drug plan, or cash transaction), drug product information (drug identification number, trade name, chemical name, formulation, and dosage), and basic demographic information (age, gender, and geographic location). The database does not include information on whether the LAI medication was administered or what the clinical indication for the prescription was, only whether the prescription was filled. No personally identifying data was included in this dataset. A cell suppression threshold value of 5 was used to further protect the confidentiality of patient data. Consequently, some provincial level data was combined and is presented more generally as Atlantic/Western Canada.

Using a retrospective design, data were collected from January 2019 to December 2020 for individuals 18-years and older. Data was examined by month, and by the following time epochs during the COVID-19 pandemic (see Fig. 1, Study Timeline).

(i) Pre-COVID comparator period (March – May 2019); a three-month period in the previous year aligning with Canada’s initial COVID escalation period.

(ii) COVID awareness/sensitization period (December 2019–February 2020); a period in which the virus was known to be active in the country, news reports were publicizing the emerging pandemic, but few preventative measures were yet being taken.

(iii) primary COVID escalation period (March – May 2020); the virus was active and being locally transmitted in all Canadian provinces, and states of emergency were active in all provinces.

(iv) COVID maintenance period (June – August 2020); a period where initial transmission rates had stabilized, but preventative measures remained active in most communities.

(v) second COVID escalation period (September – November 2020); a period where transmission rates began to increase as the country entered the second wave of the pandemic. Second wave infections peaked during January 2021 [16].

The primary outcome of this retrospective study was the number of LAI new starts (i.e., occurrence of an LAI prescription, with no prior LAI prescription in the past 30-days), during Canada’s COVID-19 response; December 2019 to November 2020. The number of LAI discontinuations (i.e., discontinuation of an LAI prescription with no subsequent LAI refills within 30-days of the medication ending) were also evaluated to examine whether oral antipsychotics may have been resumed in the later COVID epochs. While it is typical for a patient to refill before, or on, the date their current medication supply ends, a 30-day buffer was used to account for patients that may have delayed their refill. Switches between a one-month paliperidone palmitate formulation (PP1; Invega Sustenna, Janssen Pharmaceuticals, Inc.) and a three-month paliperidone palmitate formulation (PP3; Invega Trinza, Janssen Pharmaceuticals, Inc.) were also examined. Invega Trinza was the only LAI three-month formulation approved for use in Canada during the study period. Switches were defined as a new paliperidone palmitate LAI prescription (either PP1, or PP3) that were dispensed within 30-days of the previous paliperidone palmitate product prescription ending. Note that LAI switching in this context only includes switching between LAI formulations of paliperidone palmitate, and not patients who discontinued and then started again.

The secondary aims of this study were to examine the above data with respect to patient demographics (i.e., age, gender), possible provincial differences in prescribing, payer type (i.e., public, private, other), as well as concomitant neuroleptic medications (including SSRIs/SNRIs, tricyclic antidepressants, sedative hypnotics, benzodiazepines, anti-epileptic, and other mood stabilizing medications).

Descriptive statistics were the focus of analysis in this retrospective cohort study. Comparisons between time epochs were employed by chi-squared tests to examine potential differences in the average monthly proportion of LAI prescription new starts, discontinuations, and switches between time epochs. Following the methodology and analysis plan outlined, IQVIA conducted the patient level analyses (validated by the study lead).