An anonymous online survey was conducted among clinicians in psychiatry in Hong Kong and Singapore to understand their attitudes toward clozapine prescription. The respondents included trainees in psychiatry, resident physicians and psychiatry specialists in both regions. In Hong Kong, the survey invitations were sent through the Hong Kong College of Psychiatrists, the only professional organization for psychiatrists in Hong Kong. All trainees in psychiatry have to register with this organization prior to their specialist training. Therefore, the college has the email addresses of all practicing psychiatric trainees and specialists in Hong Kong. In Singapore, the eligible clinicians and their contact information were obtained from the Singapore Medical Council and National Psychiatry Residency Program, and crossed-checked against the email addresses of psychiatric departments in all restructured hospitals and available emails of private clinics and hospitals. At the time of the study, there were a total of 556 and 309 eligible participants, respectively, in Hong Kong and Singapore. Data collection took place between September 2019 and February 2020 in Hong Kong and December 2018 and March 2019 in Singapore.

Participants were invited via email to complete the survey on Questionpro, an online survey platform. The emails explained the purpose of the study and the anonymous and voluntary nature of study participation. Two reminder emails were sent to each participant at separate intervals to encourage their participation in the study.

The survey employed was adapted from Gee et al. (23) in their study of attitudes toward clozapine prescription in practitioners in South London and Maudsley NHS Foundation Trust (23). Participants were asked about their sociodemographic information, the proportion of patients with clozapine under their care, their familiarity with antipsychotic treatment guidelines, their prescribing practices in the treatment of those with TRS, including their threshold for prescribing clozapine as well as the alternatives to clozapine they would use, and their prescription patterns for CRS.

Ethics approval was granted by the relevant ethics review board, i.e., the Institutional Review Board of the University of Hong Kong/Hospital Authority Hong Kong West Cluster in Hong Kong and the National Healthcare Group’s Domain Specific Review Board in Singapore.

The Statistical Package for Social Sciences (SPSS) version 23.0 was used for data analysis. Frequencies and percentages were calculated for categorical variables while mode, median and interquartile range were calculated for ordinal data. Comparisons were performed using corrected chi-square test and Mann-Whitney U test, respectively. Associations were tested using Spearman’s rank correlation. All statistically significant differences were evaluated at the 0.05 level using two-sided tests.

When asked to provide an estimate of the proportion of patients with clozapine under their care, 72.4% of respondents in Hong Kong and 53.2% of those in Singapore provided a response. Clinicians in both regions reported a median of 5.0% of patients on clozapine under their care (Hong Kong range = 0.3–40.0%, IQR = 3.0–10.0%; Singapore range = 0.0–90.0%, IQR = 1.0–10.0%).

In terms of familiarity with the antipsychotic treatment guidelines in schizophrenia, 1.0% of respondents in Hong Kong and none in Singapore were “not at all familiar” and 12.2% of respondents in Singapore and none in Hong Kong were “a little familiar” with the guidelines. A higher proportion of clinicians in Hong Kong reported being “fairly familiar” (51.4%) and “very familiar” (47.6%) with the guidelines compared to clinicians in Singapore (“fairly familiar” 63.5%, “very familiar” 24.4%; χ²_2,261 = 21.94, P < 0.001). The guidelines that the respondents consulted with were similar, including the NICE guidelines, local clinical guidelines and Maudsley Prescribing Guidelines in Psychiatry (in order of increasing familiarity).

Those with more years of practice in psychiatry in Singapore were more familiar with treatment guidelines for TRS (r = 0.34, P < 0.001). The same association was found in respondents in Hong Kong, but did not achieve statistical significance (r = 0.19, P = 0.053).

When asked about their threshold for clozapine initiation, 45.7% of respondents in Hong Kong and 57.7% of respondents in Singapore reported they would prescribe clozapine after the failure of two antipsychotics. One respondent from Singapore reported starting clozapine early (after one antipsychotic trial), while none of the respondents in Hong Kong would do so. The rest would delay the initiation of clozapine – after the failure of three antipsychotics in 39.0% and 31.4% of respondents in Hong Kong and Singapore, respectively, and after four or more antipsychotics in 15.2% and 10.3% of respondents in Hong Kong and Singapore, respectively.

Those who were more familiar with the treatment guidelines were less likely to delay the initiation of clozapine in both Hong Kong (r = –0.21, P = 0.03) and Singapore (r = –0.17, P = 0.03).

As an alternative to clozapine for patients with TRS (see Tables 2, 3), approximately half of the respondents in both regions would ‘often’ or ‘always’ use high dose antipsychotics (55.2% in Hong Kong, 48.1% in Singapore, χ²_1,261 = 1.29, P = 0.26) and long-acting injectable antipsychotics (56.2% in Hong Kong, 50.3% in Singapore, χ²_1,260 = 0.87, P = 0.35). More clinicians in Hong Kong (70.5%) were inclined to ‘often’ or ‘always’ prescribe antipsychotic polypharmacy compared to those in Singapore (47.7%; χ²_1,260 = 13.19, P < 0.001). More clinicians in Hong Kong (49.5%) were also inclined to ‘often’ or ‘always’ prescribe adjuvant mood stabilizers compared to those in Singapore (33.5%; χ²_1,257 = 6.60, P = 0.01). In contrast, electroconvulsive therapy (ECT) was more commonly used by Singapore clinicians (56.2%) than those in Hong Kong (24.8%; χ²_1,260 = 25.06, P < 0.001).

When patients become clozapine-resistant, 20.0% of respondents in Hong Kong and 13.4% of respondents in Singapore may “often” or “always” add a mood stabilizer; 2.9% in Hong Kong and 9.6% in Singapore may “often” or “always” add an antidepressant; 18.1% in Hong Kong and 12.2% in Singapore may “often” or “always” add an antipsychotic (see Tables 2, 3). While the use of adjuvants was low in both regions, clinicians in Singapore were more inclined to add an antidepressant to augment clozapine compared to clinicians in Hong Kong (χ²_1,261 = 4.46, P = 0.04).

Some patients with TRS may refuse clozapine or fail to tolerate it. In such instances, alternatives to clozapine may be considered. In our study, approximately half of the clinicians in both regions supported the use of long-acting injectable antipsychotics and the use of high dose antipsychotics. The former has been recommended in instances where pseudo-resistance due to non-adherence to antipsychotics is suspected (5). The latter is less grounded in evidence; multiple guidelines have cautioned against the use of supratherapeutic doses of antipsychotics given the higher risk of side effects and little evidence of benefits (27, 28).

Augmentation strategies were commonly adopted by clinicians in both regions. Antipsychotic polypharmacy was prescribed by close to three quarters of clinicians in Hong Kong and half in Singapore, in comparison to the range of 15.9–64.7% reported in other studies (5, 18, 19). Studies that demonstrated the superiority of antipsychotic polypharmacy over antipsychotic monotherapy tended to be open-label trials and low-quality trials, with no superiority showed in double-blinded and high-quality trials (13). Nonetheless, a recent large nationwide cohort study found that those on antipsychotic polypharmacy had an approximately 10% lower relative risk of psychiatric rehospitalization than those on antipsychotic monotherapy, with some antipsychotic combinations working better than others (29).

Approximately half of the clinicians in Hong Kong and one-third in Singapore also employed the use of adjuvant mood stabilizers as an alternative to clozapine monotherapy. A meta-analysis that studied the efficacy of 42 pharmacologic cotreatment strategies added to antipsychotic monotherapy in schizophrenia found significant effect sizes favoring the addition of mood stabilizers, such as lithium and lamotrigine (15). However, the addition of mood stabilizers was ultimately not recommended as the evidence was generally informed by small, short-term trials with poorly reported data (15, 30, 31).

Electroconvulsive therapy (ECT) as an alternative to clozapine was adopted more strongly by clinicians in Singapore than those in Hong Kong. This is in line with the common use of ECT for schizophrenia in Singapore. The Institute of Mental Health is the sole tertiary psychiatric institute in Singapore and sees the majority of patients who are severely psychotic or aggressive in the country (32). It is also the largest center for ECT in Singapore with half of all the performed ECT sessions indicated for schizophrenia (33). Overall, one-third of all ECT sessions in Singapore were indicated for schizophrenia (33). This approach is backed by moderate quality evidence that suggest that ECT augmentation to of a non-clozapine antipsychotic monotherapy is superior to antipsychotic monotherapy alone in the treatment of TRS (34–36).

In our study, only 10–20% of clinicians would opt to use combination antipsychotics or adjuvant “mood stabilizers” with clozapine. An even smaller number would consider adjuvant antidepressants. In contrast, 76.3% of psychiatrists in a survey in India would combine clozapine with another antipsychotic, with amisulpride being one of the three most preferred agents, followed by risperidone, aripiprazole and haloperidol (19). In a Danish study, when treating resistant-positive symptoms, 39.8% would increase the clozapine dose, 29.6% of respondents would add an atypical antipsychotic, 23.9% would add a conventional antipsychotic, and 4.6% would add a mood stabilizer; when negative symptoms are resistant, 9.5% would increase the clozapine dose, 77.0% would add an atypical antipsychotic, 8.1% an antidepressant and 2.6% a mood stabilizer (18).

In existing literature, the recommendations on augmentation strategies for clozapine non-response have varied across different guidelines (27). This is perhaps a reflection of the lack of robust evidence to support one strategy over the other. The TRRIP Working Group (37) offered the following recommendations on the basis of international expert consensus: Combining clozapine with an antipsychotic, specifically amisulpride or aripiprazole, was suggested for patients who have persistent positive or mixed symptoms or persistent aggression. The combination of clozapine and aripiprazole was found in the earlier-mentioned nationwide cohort study on antipsychotic polypharmacy versus monotherapy to have the best outcome in reducing the risk of rehospitalization, in comparison to clozapine monotherapy as well as other antipsychotic combinations (29). The TRRIP Working Group also recommended the use of adjuvant mood stabilizers (namely, lithium, and lamotrigine) in patients with persistent aggression and/or suicidal ideation, and adjuvant antidepressants in patients with persistent negative symptoms or suicidal ideation (37). While the use of ECT in CRS was not studied in our survey, ECT shows promise and was mentioned by the TRRIP Working Group as a possible treatment for those with persistent positive symptoms, mixed symptoms or suicidal ideation (36–38).

This is the first survey that described the insights on experience, familiarity and prescribing practices in the treatment of patients with TRS and CRS in clinicians from Hong Kong and Singapore. The response rates of 19% and 50% in Singapore in Hong Kong were comparable to the average response rates in previous survey studies from the same regions (39, 40). There was a good completion of the survey questions by the respondents, with missing responses in 1.3% and 0.3% of the total responses in Hong Kong and Singapore, respectively. The similar sociodemographic characteristics of respondents in both regions also allowed for a more meaningful comparison of the survey responses. However, the small sample size for the Hong Kong participants might have led to false negatives in some of the results. The management approach of TRS in this study was limited to the use of pharmacotherapy and ECT, and did not include the use of psychological interventions such as cognitive behavioral therapy for psychosis. Psychological interventions are gaining ground as complementary treatment in psychosis, though the evidence-base for their use remains limited and more research in this field is still needed (12, 41–43). Other limitations of this study include the possibility of recall bias and social desirability bias due to the self-reporting nature of the survey. Selection bias may also be present as the respondents were more likely to be clinicians who were experienced in the use of clozapine and those who were concerned about its underutilization.