Article Text
PDF
Abstract
Objectives This study aims to determine the proportion of initial cardiometabolic assessment and its predicting factors in adults with schizophrenia, bipolar disorder or other related diagnoses for whom a second-generation antipsychotic was prescribed in the hospital setting.
Design Cross-sectional study.
Setting The psychiatry unit of a Canadian tertiary care teaching hospital in Montreal, Canada.
Participants 402 patients with aforementioned disorders who initiated, restarted or switched to one of the following antipsychotics: clozapine, olanzapine, risperidone, paliperidone or quetiapine, between 2013 and 2016.
Primary outcome measures We assessed the proportion of cardiometabolic parameters monitored.
Secondary outcome measures We identified predictors that influence the monitoring of cardiometabolic parameters and we assessed the proportion of adequate interventions following the screening of uncontrolled blood pressure and fasting glucose or glycated haemoglobin (HbA1c) results.
Results Only 37.3% of patients received monitoring for at least three cardiometabolic parameters. Blood pressure was assessed in 99.8% of patients; lipid profile in 24.4%; fasting glucose or HbA1c in 33.3% and weight or body mass index in 97.8% of patients while waist circumference was assessed in 4.5% of patients. For patients with abnormal blood pressure and glycaemic values, 42.3% and 41.2% subsequent interventions were done, respectively. The study highlighted the psychiatric diagnosis (substance induced disorder OR 0.06 95% CI 0.00 to 0.44), the presence of a court-ordered treatment (OR 0.79 95% CI 0.35 to 1.79) and the treating psychiatrist (up to OR 34.0 95% CI 16.2 to 140.7) as predictors of cardiometabolic monitoring.
Conclusions This study reports suboptimal baseline cardiometabolic monitoring of patients taking an antipsychotic in a Canadian hospital. Optimising collaboration within a multidisciplinary team may increase cardiometabolic monitoring.
- antipsychotic agents
- schizophrenia
- bipolar disorder
- drug-related side effects and adverse reactions
- cardiometabolic side effects
- guideline adherence
Data availability statement
No data are available.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
- antipsychotic agents
- schizophrenia
- bipolar disorder
- drug-related side effects and adverse reactions
- cardiometabolic side effects
- guideline adherence
Strengths and limitations of this study
We present real-world evidence regarding adherence to guidelines of cardiometabolic parameters monitoring in patients with schizophrenia, bipolar disorder or other related diagnoses.
We selected five antipsychotics with the highest inherent risk of cardiometabolic adverse events (clozapine, olanzapine, risperidone, paliperidone or quetiapine) for increased clinical relevance.
We identified predictors of cardiometabolic parameters monitoring in a tertiary centre in Canada.
The lasso statistical model cannot determine whether some variables are stronger predictors for cardiometabolic parameters monitoring than others, nor does it allow to infer causality.
Despite clear international guidelines to this effect, there is still a large gap in monitoring of cardiometabolic parameters in patients taking antipsychotics with the highest inherent risk of cardiometabolic adverse events.
Introduction
Second-generation antipsychotics (SGA) are first-line therapy for various psychiatric disorders, such as schizophrenia spectrum and other psychotic disorders, as well as bipolar affective disorder and related disorders.1 2 However, this class of medication is known to contribute to cardiometabolic adverse events (AEs), such as hypertension, weight gain, type 2 diabetes and dyslipidaemia.3 Exposure to antipsychotics may also increase the risk of stroke,4 and acute myocardial infarction,5 6 mainly in the older population. Patients with schizophrenia and bipolar affective disorder are at a higher baseline cardiometabolic risk7–13 and have higher risk of cardiovascular disease and associated death.14 Patients having both diabetes and serious mental illness are at increased mortality by 3–4 fold.15 Long-term use of SGA in individuals with serious mental illness further increases the risk of cardiovascular events, including an increased risk of developing diabetes by about threefold.16–18 Although SGA are known to cause cardiometabolic AE, they have been demonstrated to decrease overall mortality in patients with schizophrenia, including from cardiovascular causes.19 The use of antipsychotic polypharmacy among patients with schizophrenia20 and bipolar disorder21 is common, although the impact of polypharmacy on cardiometabolic AEs remains unclear.22 The increased baseline cardiometabolic risk combined with the cardiometabolic AE of SGA emphasises the importance of cardiometabolic parameter monitoring (CMPM) in these patients.
Although there are specific guidelines that recommend CMPM when initiating a SGA, it is not routinely done in practice.23 24 A meta-analysis assessed the guideline adherence for inpatients and outpatients treated with antipsychotics and found that blood pressure, triglycerides, cholesterol, plasma glucose and weight were measured in 69.8%, 59.9%, 41.5%, 44.3% and 47.9%, respectively.25 To our knowledge, no study has been published on the baseline CMPM in a Canadian inpatient setting and this study aims to fill this gap.
The primary aim of this study was to measure adherence to practice guidelines for CMPM in adult inpatients with schizophrenia or bipolar disorder who are initiating, restarting, or switching to a SGA during their hospitalisation in a tertiary psychiatric care centre. The secondary objectives were to identify predictors of CMPM and to assess the frequency of adequate interventions following the screening of an abnormal value.
Methods
Patient and public involvement
No patient nor public were involved in the design, conduct, choice of outcome or recruitment to the study.
Population
This study was conducted at the Jewish General Hospital, a Canadian tertiary care teaching hospital that serves a multiethnic community in Montreal, Canada. To be included, patients had to be 18 years and older; hospitalised in the psychiatry unit; with a diagnosis of schizophrenia spectrum and other psychotic disorders or bipolar affective disorder and related disorders. The included diagnoses were selected due to greater inherent cardiometabolic risk in this population7–13 (online supplemental table S1).
Supplemental material
At the time of inclusion, patients needed to be initiating, restarting or switching to one of the following SGA: clozapine, olanzapine, quetiapine, risperidone or paliperidone, with a follow-up period within the time frame of 9 March 2013 to 1 January 2017. A schematic representation of the study period is shown in online supplemental figure S1. The five SGA mentioned above will be referred to as study SGA.
Design
In an observational cross-sectional design, we used the pharmacy dispensing software (GesPharx 8, CGSI TI, Quebec City, Quebec, Canada) to identify patients hospitalised in the psychiatry unit of the Jewish General Hospital, and who received a study SGA during the predefined time frame. Patients were randomly selected in blocks throughout the study period. If a patient was hospitalised more than once during the study time frame, a single randomly selected admissible hospitalisation was included.
Studies show that there is generally a greater cardiometabolic risk associated to some SGA compared with most first-generation antipsychotics. In particular, almost all studies put clozapine and olanzapine as the highest cardiometabolic AE, followed by quetiapine.26–30 Indeed, metabolic syndrome is considered one of the potentially most life threatening common AE of clozapine.31 Depending of the studies, risperidone and paliperidone are next in line for cardiometabolic risk,26–28 while current data suggest minimal increased risk for cholesterol and glucose disturbances for the newer generation, including aripiprazole, lurasidone, ziprasidone, brexpiprazole and cariprazine. However, there are some signal that even those new antipsychotics could cause weight gain and triglycerides increase.32 Therefore, even though the guidelines recommend to measure baseline CMPM with all antipsychotics, to increase external validity and clinical relevance, first generation antipsychotics, aripiprazole, lurasidone, ziprasidone, brexpiprazole and cariprazine were excluded as they were deemed at less metabolic risk as per national guidelines24 and recent network meta-analysis,32 and the two latter were unavailable on the Canadian market during the study time frame. We can also note some emerging data seem to put some doubts on strength of this broad affirmation.33 However, in clinical practice, it is recommended to adhere to these guidelines for all antipsychotics.
All doses of study SGA were included in the study because cardiometabolic AE do not appear to be dose-dependent.34 The study SGA had to be prescribed for chronic use, defined as a duration of at least 7 days. Risperidone and paliperidone were considered to be equivalent, not requiring new baseline CMPM, as paliperidone is the active metabolite of risperidone.35
Outcomes
The primary objective of this study was to assess the proportion at which the recommended cardiometabolic parameters were monitored. Monitoring parameters of blood pressure, lipid profile, fasting glucose or glycated haemoglobin (HbA1c), weight or body mass index (BMI), and waist circumference had to be ordered, measured or charted within the follow-up period of 4 weeks prior to or 4 weeks following the index date (online supplemental figure S1).
Secondary objectives consisted of identifying predictors that influence the monitoring proportion of cardiometabolic parameters and assessing the proportion of adequate intervention following the screening of uncontrolled blood pressure and fasting glucose or HbA1c results. Blood pressure and fasting glucose or HbA1c results were considered abnormal as per the definitions published in Canadian Hypertension guidelines (three values of blood pressure ≥140/90 mm Hg or one value of ≥180/110 mm Hg over the follow-up period) and Canadian Diabetes guidelines (fasting glucose ≥7.0 mmol/L or HbA1c ≥6.5%), respectively.24 36 An adequate intervention following an abnormal result was considered to be any of the following: an initiation of treatment or a change in treatment for hypertension and/or diabetes, a consultation, or a referral for a subsequent follow-up in any outpatient clinic regarding the issue.
Sample size
Based on the results from a meta-analysis and a previous study, we hypothesised the following CMPM adherence levels: blood pressure ≥60%, lipid profile ≥50%, fasting glucose or HbA1c ≥40%, weight or BMI ≥40% and waist circumference ≥10%.25 37 A sample size calculation was necessary due to limited time and budget. Thus, a sample size of 402 patients was calculated to obtain a precision of ±5% of the expected proportion with a 95% CI (Sample Size Calculators, Clinical and Translational Science Institute, University of California, San Francisco, San Francisco, California, USA).
Statistical analysis
Clopper-Pearson CI was used to calculate 95% CI as some parameters contained few data points. For the first secondary objective, a lasso logistic regression model was used to identify predictors of CMPM.38 39 The monitoring of at least three cardiometabolic parameters was chosen as it was thought to be practically feasible while still assessing at least one indicator of metabolic dysfunction. A prespecified sensitivity analysis was planned to set the threshold of monitoring to 4/5 parameters. Given a large number of predictors, a lasso model was used as it permits to identify a smaller subset of predictors that exhibit the strongest effects. These predictors can be interpreted with confidence to exert an effect on CMPM. Coefficients and OR are presented for descriptive purposes. The parameters analysed by the LASSO model as potential predictors of CMPM are presented in online supplemental table S2. This list is comprised of parameters that have been previously analysed in the literature, including some exploratory parameters judged to be pertinent by the investigators.37 40–42 Statistical analyses were performed using the R statistical software, V.3.5.1.
Results
Levels of adherence to practice guidelines for CMPM
A total of 402 patients were included in the study. The main diagnosis was Schizophrenia spectrum disorder (69.7%) and almost half (49.1%) were prescribed risperidone or paliperidone. More than half of patients had legal confinement during their hospitalisation. Their demographic and clinical characteristics are presented in table 1.
Study population
Figure 1 shows the proportion of patients in whom cardiometabolic parameters were assessed and the expected rates from the literature. While blood pressure monitoring was almost always carried out, screening for lipid and glycaemic parameters was done in approximately a quarter to a third of patients. Weight or BMI were reported in almost all patients’ charts; however, waist circumference was seldom measured. One third of the patients had the majority of baseline cardiometabolic assessment. A complete baseline CMPM profile was found in less than 2% of patient charts. Of the five cardiometabolic parameters, blood pressure and weight or BMI were ordered, measured or charted more often than expected.
Proportion of cardiometabolic parameter monitoring. n=402. Appropriate monitoring=3/5 parameters monitored. The dark grey bar represents expected monitoring proportion compared with the proportion observed in light grey. BMI, body mass index; HbA1c, glycated haemoglobin.
Predictive variables
Of the predefined parameters in online supplemental table S2), risk of aggressivity was excluded from the lassomodel due to a high number of missing data. Four patients were excluded from the lasso regression model due to missing data or because a variable was unique and could not be included. Finally, route of administration for risperidone and paliperidone, including PO or intramuscularly, were combined.
As shown in table 2, CMPM was influenced by the underlying psychiatric condition, whether the patient underwent court-ordered treatment, and the treating psychiatrist. Patients with a substance-induced disorder diagnosis were the least likely to have the majority of CMPM (OR=0.06), followed by bipolar and related disorders (OR=0.93) and other diagnosis (OR=3.09) as compared with the category of schizophrenia and other psychotic disorders. The presence of a court-ordered treatment (OR=0.79) negatively influenced cardiometabolic monitoring. Each treating psychiatrist was compared with an arbitrarily chosen psychiatrist in order to preserve prescriber anonymity; the treating psychiatrist was a key determinant of CMPM, as the smallest OR of 0.80 and the largest OR of 34.
Predictors of 3/5 cardiometabolic parameters monitored**, obtained from lasso regression model
A sensitivity analysis was performed using a threshold score of ≥4/5 cardiometabolic parameters for monitoring. Three variables were shown to be predictors of cardiometabolic monitoring: number of concomitant antipsychotic medications, presence of a court-ordered treatment and treating psychiatrist.
Intervention following abnormal blood pressure or glycaemic control
Blood pressure and glycaemic values were deemed abnormal in 104 patients and in 17 patients, respectively. Of the 104 patients with uncontrolled blood pressure, 42.3% (95% CI 32.7% to 52.4%) received a subsequent intervention. Similarly, 41.2% (95% CI 18.4% to 67.1%) of the 17 patients with uncontrolled glycaemic indices received a subsequent intervention.
Discussion
While the non-adherence to guidelines was already suspected by clinicians, this is the first study, to our knowledge, to rigorously assess adherence to practice guidelines of cardiometabolic monitoring in inpatients receiving an SGA in a Canadian tertiary hospital setting. We have found that (1) while some parameters were almost always monitored (eg, blood pressure and weight or BMI), others were less often assessed (eg, lipid profile, fasting glucose or HbA1c, and waist circumference); (2) in the context of a global baseline cardiometabolic assessment, approximately one-third of patients had at least three cardiometabolic parameters measured, and less than 2% received a complete baseline cardiometabolic assessment and (3) the quality of cardiometabolic screening was strongly influenced by the treating psychiatrist, the presence of a court-ordered treatment and the psychiatric diagnosis.
The high proportion of blood pressure and weight or BMI measurement was expected, as it is routine nursing practice in the psychiatry unit of the Jewish General Hospital to obtain these values without requiring a psychiatrist’s intervention. It was also expected that the proportion of patients receiving a complete monitoring would be very low, as it has been reported that a cultural barrier limits the frequency at which waist circumference measurement is performed.37 However, abdominal obesity, which can be measured by waist circumference, is a better predictor of cardiovascular risk than BMI.43 Therefore, waist circumference measurement should be reinforced in addition to weight measurement. Moreover, it is concerning that lipid profile and fasting glucose or HbA1c, major predictors of cardiovascular events, were not monitored more consistently.44
The hospital setting is an important environment to obtain baseline CMPM, as psychiatric patients are often subject to lost to follow-up in an outpatient setting.45 It has been demonstrated that this population receives suboptimal medical screening for other diseases as well.46 The inpatient setting should be considered an opportune timeframe for appropriate CMPM, however, this study highlights that the majority of patients do not get full CMPM.
Predictors and their impact
A court-order treatment and the treating psychiatrist were deemed predictors by the LASSO model for both the primary analysis and the sensitivity analysis.
A court-order treatment was negatively associated to CMPM, suggesting an inverse correlation between a difficult patient-treating team collaboration and an adequate CMPM.
This predictor was evaluated as a surrogate marker for disease severity and patient collaboration. Appropriate biopsychosocial treatment could help to minimise the impact of the severity of the illness, and decrease hostility while increasing collaboration. Baseline CMPM can be obtained later in the admission when patient-treating team collaboration is more optimal.
The treating psychiatrist was a predictor of adequate CMPM, with a wide range of OR between psychiatrists, indicating that different practices could greatly influence the level of monitoring. It also suggests that the published recommendations can be adhered to in practice.
A few variables put in the model were not statistically significant, including some that were more surprising to us. The prevalent use of cardiometabolic medication, which would indicate a more physically impaired population, did not play a role in the assessment. Moreover, the use of SGA with the highest risk of AE (clozapine and olanzapine), as well as polypharmacy, noted in almost half of the patients, were not associated to higher monitoring.
Patient care is a shared responsibility that often involves the expertise of a highly trained multidisciplinary team. While baseline assessment generally falls under the responsibility of the treating psychiatrist who initiates SGA treatment, the lack of resources and time may be a limiting factor. The involvement of medical hospitalists or general practitioners in this type of monitoring has shown promise in improving CMPM.47 In-hospital multidisciplinary collaboration, has become common practice and has been associated with positive patient outcomes.48 In particular, the presence of a nurse practitioner in an acute psychiatric unit has been shown to significantly increase the number of patients with cardiometabolic screening.49
Furthermore, it has been demonstrated that pharmacists favour adherence to prescribing guidelines or CMPM in both hospital and outpatient settings.50–53 One study shows that pharmacists play a key role in the multidisciplinary team in an outpatient metabolic risk screening programme, specifically in patients to whom antipsychotics are prescribed.54 Currently, pharmacists are permitted to prescribe laboratory testing in the context of pharmacotherapy surveillance in some jurisdictions.55 Prescribing lipid levels and fasting glucose and/or HbA1c may significantly help improve adherence rates to guidelines.56
Considering the potential impact pharmacists and nurses can have on CMPM, the implementation of an internal procedure consisting of sharing and/or delegating the monitoring of SGA cardiometabolic parameters to the multidisciplinary team could be a solution to this lingering issue and could increase the quality of patient care.
Moreover, less than half of patients with abnormal blood pressure or glycaemic values had a subsequent intervention done. This highlights the importance of the multidisciplinary approach for patient treatment, in addition to baseline monitoring.
Strengths and limitations
The population chosen was clinically oriented, leading to relevant results. Indeed, a more strenuous follow-up in this population would be expected considering that these patients have an elevated inherent cardiometabolic risk due to their underlying disease state in addition to their treatment with higher risk antipsychotics.7–13 26 27 Rigorous data done over a prolonged period of almost 4 years reflects the real-world clinical practice.
The lasso statistical model is useful to identify predictors among a large array of variables, but it cannot determine whether some variables are stronger predictors for CMPM than others, nor does it allow to infer causality.
Furthermore, some patients leaving the psychiatry unit without medical consent may have prevented the planning of appropriate follow-ups and posed an obstacle to baseline CMPM. These patients could have been excluded from this study to limit confusion bias. Moreover, the study design could not take into account the psychiatrist’s clinical judgement in cases where it was justified not to treat an abnormal value on the basis of specific patient characteristics or clinical context. Therefore, the study may overestimate the lack of intervention following abnormal values, which was noted to be less than 50% of the time. In addition, interventions were not evaluated for lipid profile as multiple clinical approaches could be justified for this heterogeneous patient population.
Future directions
The care of patients with serious mental disorder is complex, especially in the treatment integration of multiple comorbidities. Future research addressing the advantages of a multidisciplinary approach to the cardiometabolic follow-up is warranted, with the exploration of automated monitoring. A consideration of an ‘opt-out’ admission prescription for cardiometabolic monitoring rather than an ‘opt-in’ might increase the compliance to guidelines, with the caveat of possible overprescription of those tests. Furthermore, there should be an assessment of the clinical outcomes of patients having regular monitoring and the treatment of SGA-induced cardiometabolic disorders.
Conclusions
This study highlights that despite access to patients in a controlled environment amenable to cardiometabolic screening, global baseline assessment of key cardiometabolic parameters remains a clinical challenge. This is a shared responsibility between the inpatient and outpatient setting, which might benefit from multidisciplinary approaches involving the collaboration of general practitioners, nurses, pharmacists and other healthcare professionals, as well as automation and baseline admission protocols.
Data availability statement
No data are available.
Ethics statements
Patient consent for publication
Ethics approval
This study was approved by the Research Ethics Committee at the Jewish General Hospital (CR17-061). To ensure confidentiality, data were denominalised and coded.
Acknowledgments
We are grateful to Steve Ferreira Guerra for his support with statistical analyses, and to Sylvia Duong and Julie Roy for their involvement in the design of the study.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
ML and VD-B are joint senior authors.
Twitter @Mtl_PlateletLab
Contributors All authors had contributed to the conception of the study, analysis and interpretation of data for the work. JF, EC, J-FP and AR did the acquisition of the data and first draft of the manuscript. DW, CR, SR, ML and VD-B revised it critically. All authors approved the final version of the work and are in agreement to be accountable for all aspects of the work. VD-B accepts full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish
Funding The Faculté de Pharmacie de l’Université de Montréal and the Jewish General Hospital funded this study. ML was supported by the FRQS Junior 1 Research Scholarship and is a Canada Research Chair in Platelets as biomarkers and vectors; and is supported by research grants from the Canadian Institutes for Health Research, the Canada Foundation for Innovation, Diabetes Quebec, Heart and Stroke Foundation of Canada and by the Fondation de l’Institut de Cardiologie de Montréal. Soham Rej received salary support from the FRQS Clinician-Scientist Junior 1 award.
Disclaimer They had no role in the study design, collection, analysis or interpretation of the data, in the writing of the report nor in the submission of the paper for publication. Other authors report other source of funding, independent of the work presented in this paper.
Competing interests ML has received speaker fees from Bayer; has participated in industry-funded trials from Idorsia and Phase3Bio; has served on advisory boards for Servier and Orimed Pharma; and has received in-kind and financial support for investigator-initiated grants from Leo Pharma, Roche Diagnostics, Aggredyne and Fujimori Kogyo, for unrelated work. VD-B has received fees for consulting by HLS therapeutics and Otsuka for unrelated work, and speaker fees by HLS therapeutics, Ostuka and Jansen for unrelated work. SR has received an investigator-initiated research grant from Satellite Healthcare The remaining authors declare that there is no conflict of interest.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.