Introduction

Improving physical health outcomes in people with psychotic disorders remains a major concern and challenge [1,2,3,4,5,6]. The 15–20-year mortality gap between people with chronic psychotic disorders and the general population may even be widening [7,8,9,10,11,12]. The majority of the excess deaths are due to partially modifiable physical diseases, such as premature cardiovascular disease (CVD), respiratory disease and infection [13,14,15,16]. In the current pandemic, people with schizophrenia are seven times more likely to contract COVID-19 than the general population and have twice the mortality rate (8.5% vs. 4.7%) [17]. Given the increased risk of COVID-19 associated with obesity and metabolic syndrome [18,19,20,21], which are elevated in both early [22] and chronic [23] psychotic disorders, the disparities are likely to increase further.

A complex interaction between lifestyle factors [24], AP side effects and lower-quality healthcare [25, 26] contributes to the higher cardio-metabolic risk. People with chronic psychotic disorders are at a two to three times higher risk of dying from CVD than the general population [27, 28]. Up to 75% of patients with schizophrenia, compared with approximately 33% in the general population, die of coronary heart disease [16], and approximately 66% will have multiple physical conditions [29, 30]. People with chronic psychotic disorders have increased levels of obesity (50%), metabolic syndrome (30%) [23, 31], glucose intolerance (25%) and type 2 diabetes (T2DM) (10%) [32], together with high levels of smoking [33], poor diet [34, 35] and low physical activity levels [36].

Not only are some cardio-metabolic risk factors evident at the earliest stages of psychosis [22, 37, 38], even with minimal AP exposure [23, 39,40,41,42], persistently high fasting insulin levels during childhood are associated with an increased risk of developing a psychosis at-risk mental state and psychotic disorder [43]. APs, by a combination of central and peripheral actions, for example 5-HT2C and histamine receptor antagonism, together with an inhibition of insulin signalling pathways, greatly increase the risk of metabolic dysregulation, diabetes and CVD [7, 15, 44, 45]. Metabolic risk profiles vary between APs [46,47,48], with olanzapine and clozapine being the worst [49,50,51], whereas aripiprazole, brexpiprazole, cariprazine, lurasidone and ziprasidone have the most benign metabolic profiles [52, 53]. Weight gain can occur within weeks of starting APs [54]. In the CATIE study, 30% of people with schizophrenia treated with olanzapine gained 7% or more of their baseline body weight [55], whereas the EUFEST (European First-Episode Schizophrenia Trial) showed that up to 50% of individuals who were treatment-naive gained over 7% of body weight within 1 year [56].

Guidelines state that cardio-metabolic monitoring should be carried out at initial assessment, at 3 months and then annually in people prescribed APs [57, 58]. The Second National Audit of Schizophrenia (NAS) in 2014 of 5500 people in community mental health services showed minimal improvement in the monitoring of physical health parameters compared with 2012 [59,60,61]. Only 33% of people with schizophrenia had all five key physical health risk factors (smoking, weight, blood glucose, blood lipids, blood pressure) monitored, compared with 29% in 2012 [59]. Monitoring of glucose control increased from 50 to 57%, but when abnormal levels were identified, further support was offered in 36% of cases, which dis-improved compared with 2012 (53%). Lipid monitoring increased from 47 to 57%, and approximately half had their BMI checked in 2014 (52%) and 2012 (51%) [59]. A recent cross-sectional study of routine cardio-metabolic monitoring in patients treated with long-acting injectable (LAI) APs (n=116) showed that less than 45% of medical records had documentary evidence of metabolic monitoring over the previous 6 months [31].

The persistent physical health monitoring and implementation deficits highlighted in the second NAS prompted a movement towards greater integration between primary and secondary care and shared responsibility of physical health monitoring/treatment in psychiatric patients. Additionally, it launched the National Health Services (NHS) Lester Positive Cardio-metabolic Health Resource Tool encompassing the “don’t just screen, intervene” strategy [62, 63]. More recently, the Maudsley prescribing guideline series has launched practice guidelines for Physical Health Conditions in Psychiatry [64].

The most recent data from the Early Intervention in Psychosis (EIP) spotlight audit (2019) in the UK, including 9631 patients, demonstrated a marginal improvement compared with the NAS [59,60,61]. In this EIP audit, 64% had been screened for all seven physical health measures (smoking, alcohol use, substance misuse, BMI, blood pressure, blood glucose, lipids) [65]. The rates of intervention varied between measures, from 66% for elevated blood pressure to 93% for harmful/hazardous use of alcohol [65].

In general, rates of monitoring tend to be worse for male patients, younger patients (16–44 years) and those with schizophrenia [66]. Several previous studies have demonstrated that various interventions can improve the rates of cardio-metabolic health monitoring in early [67] and chronic psychotic disorders [68,69,70,71,72], including efforts to monitor physical health in people’s homes [73]. The HSE National Model of Care for EIP Services (2018) in line with UK guidelines [57, 58] stipulates that weight, BMI, pulse, blood pressure, physical examination, ECG and routine bloods (including fasting glucose/lipids, HbA1c and prolactin levels) should be completed as part of the initial assessment, at 3 months and then annually or more frequently if indicated [74]. In parallel, a comprehensive assessment of lifestyle factors and corresponding lifestyle advice should be conducted [74].

While there is an abundance of research investigating cardio-metabolic monitoring improvement strategies in people with psychotic disorders [67,68,69,70,71,72,73, 75], to the best of our knowledge, no study has investigated metabolic monitoring including lifestyle assessment in FEP in the Irish context. Given the considerable rates of cardio-metabolic pathology, exacerbated by APs, and the persistent deficiencies in the monitoring of modifiable risk factors across the psychosis disorder spectrum, we aimed to determine the baseline rate of monitoring in AP treated FEP patients and separately, in patients with chronic psychosis treated with LAI APs. We then investigated whether pragmatic interventions could improve monitoring rates in both groups.

Methods

Audit standards

The National Institute for Health and Care Excellence (NICE) guidelines for psychosis and schizophrenia in adults: prevention and management. CG178 [58] and NICE quality standards in relation to treating and managing psychosis (QS80, Quality statement 6; QS102, Quality statement 6) were used.

The guidelines recommend that 100% of FEP individuals should have the following physical health parameters recorded in clinical files during commencement of AP medication.

  • Baseline monitoring

    •  Electrocardiogram (ECG), fasting blood glucose, fasting lipids, prolactin.

  • Physical examination

    • Weight, height, body mass index (BMI), waist circumference, blood pressure (BP), heart rate (HR), abnormal movement assessment.

  • Lifestyle assessment

    • Documented enquiry about (i) smoking history, (ii) alcohol use, (iii) illicit drug misuse, (iv) prescription drug misuse, (v) personal family/medical history and (vi) physical activity levels.

  • Lifestyle advice

    • Documentation of advice given on (i) illicit drug misuse, (ii) prescription drug misuse, (iii) alcohol misuse, (iv) recommended physical exercise.

Modification of guidelines

The severity of psychotic illness has implications for behaviour and can affect a patient’s willingness to engage in monitoring and examination. Therefore, after a consensus meeting, it was decided that all physical health parameters recorded within 2 weeks of commencing an AP were acceptable.

Annual monitoring for those on long-term APs

  • Weight, BMI, blood pressure (BP), fasting glucose, fasting lipids. 

Inclusion criteria/eligibility criteria

FEP

All patients aged (16–65 years) enrolled in the Dublin South City First Episode Psychosis Programme (DSFEP), St. James’s Hospital, prescribed AP medication were included. The initial audit included all individuals enrolled on the Programme over an 18-month period from January 2015 to June 2016. The re-audit included individuals enrolled over a 12-month period from December 2016 to December 2017.

Chronic psychosis

All adult (18–75 years) patients with a diagnosis of a psychotic disorder (schizophrenia, schizoaffective, delusional disorder) receiving long-acting injectable (LAI) APs attending the CAMAC Community Mental Health Service at St. James’s Hospital in October 2016 were included. Patients with a primary diagnosis of bipolar affective disorder and depression were excluded.

Intervention

FEP

Clinical records, medication records and hospital electronic records were reviewed by (PG, TOC, ZL, RMcC, RF, AMcL, LM, MH) using a study-specific data collection tool.

Chronic psychosis

Clinical records were reviewed before and 3 months after (January 2017) the implementation of an awareness raising-educational intervention aimed at improving rates of cardio-metabolic monitoring.

Physical health parameter checklist (FEP)

On completion of the initial audit, recommendations to improve adherence to physical health monitoring guidelines were implemented, including the following: development of an evidence-based physical health parameter checklist in hard copy format for completion by medical staff was incorporated into the general admission template at a service wide level.

Pre-antipsychotic electronic laboratory order set (FEP)

A service-wide pre-AP prescribing blood order set was established. Order sets are a form of clinical decision support system (CDSS) where a limited set of evidence-based tests are proposed for a series of indications, integrated in a computerized clinician order entry. The specific pre-AP profile was incorporated into the hospital electronic system with assistance of the IT department, making it accessible to all clinicians in the service. Clinicians were prompted to enter baseline laboratory tests for those patients with psychosis through a searchable drop-down menu of common indications selected through tick boxes. Selecting one or more of these indications prompted a new window with the appropriate tests for the indications. This facilitated efficient FEP blood ordering on the patient’s initial contact with the service via their electronic record.

Awareness-raising educational sessions (FEP and chronic psychosis group)

Educational sessions with members of the MDT were conducted at interdepartmental meetings. These educational meetings focused on the rationale for routine screening and monitoring of physical health parameters, and provided an opportunity to discuss potential service-specific barriers to monitoring (i.e. resources/equipment/clinician time and availability). As such, these sessions provided a forum to problem-solve around these barriers and explore potential solutions. Feedbacks of baseline audit results were presented at MDT meetings, case conferences and audit meetings in the Department of Psychiatry St. James’s Hospital.

Statistical analysis

Fisher’s exact in GraphPad was used to determine differences in the proportion of patient records in compliance with standards pre and post-intervention.

Results

Demographics

See Table 1.

Table 1 Demographics
Full size table

Antipsychotics

Prescribed APs in FEP are shown in Table 2. In the chronic psychosis group, 75.6% (31/41) were prescribed only second-generation antipsychotics (SGA), 14.6% (6/41) were prescribed only first-generation antipsychotic (FGA) and 9.8% (4/41) were prescribed both SGA and FGA.

Table 2 Pre-antipsychotic physical health monitoring in first-episode psychosis
Full size table

First-episode psychosis

In the first cohort, 35 patients enrolled in the DSFEP programme over the 18-month period of initial audit. Two patients were excluded as they were not prescribed APs. In the second cohort, 21 patients enrolled in the DSFEP programme over a 12-month period. One patient was excluded as was not prescribed APs.

Changes in cardio-metabolic parameters in FEP post-intervention

Monitoring of fasting glucose (p=0.047), HbA1c (p=0.005) and prolactin (p=0.001) significantly improved at re-audit (Fig. 1). Conversely, the monitoring of blood pressure (p=0.039) and heart rate (p=0.03) significantly dis-improved. There were no significant changes in fasting lipids (p=0.23), weight (p=1.0), BMI (p=0.54), waist circumference (p=1.0) or ECG monitoring (p=1.0) (Table 2; Fig. 1).

Fig. 1
figure 1

Monitoring of physical health parameters. a There were significant improvements in the monitoring of fasting glucose (p=0.047), HbA1c (p=0.005) and prolactin levels (p=0.001) after the implementation of targeted improvement strategies in FEP. There were no changes in the rate of ECG, weight or BMI monitoring. Blood pressure (p=0.039) and heart rate (p=0.03) monitoring dis-improved (data not shown). b The routine monitoring of blood pressure improved (p=0.053) post-intervention in the chronic psychotic disorders group (HbA1c: glycated haemoglobin, FEP: first-episode psychosis, ECG: electrocardiogram, BMI: body mass index)

Full size image

Changes in lifestyle assessment parameters in FEP post-intervention

There was a significant improvement in the recording of prescription drug misuse (p=0.01) from baseline to re-audit. In contrast, the documentation of illicit drug misuse (p=0.01) and alcohol misuse (p=0.047) advice decreased from baseline to re-audit. There were no other significant lifestyle assessment or advice parameter changes.

Changes in cardio-metabolic parameters in chronic psychosis post-intervention

The routine monitoring of blood pressure improved (p=0.053) post-intervention (Fig. 1b). There were no significant changes in the monitoring of weight, BMI, waist circumference, fasting glucose or fasting lipids post-intervention (Table 2).

Four percent of the sample had diabetes (DM), though none were prescribed diabetic treatment. Hypertension (HTN) was recorded in 9.7%, of which 2.4% were prescribed anti-hypertensive treatment (Table 3). GPs were notified of DM or HTN findings via a letter.

Table 3 Monitoring of physical health parameters in chronic psychosis
Full size table

Cardio-metabolic parameters in FEP compared with chronic psychosis post-intervention

The rates of weight (p=0.03), fasting glucose (p<0.001) and fasting lipids (p<0.001) monitoring were higher in the post-intervention FEP group compared with the chronic psychosis group. There were no differences in BMI, waist circumference or blood pressure (Table 3).

Discussion

This study investigated the rates of compliance with cardio-metabolic monitoring as per NICE physical health guidelines in antipsychotic (AP) treated patients with early and chronic psychotic disorders [58]. Furthermore, we determined whether the implementation of targeted improvement strategies could advance the baseline monitoring rates. We found that baseline cardio-metabolic monitoring was suboptimal in both first-episode psychosis (FEP) and chronic psychosis. In the FEP cohort, the introduction of a physical health parameter checklist and pre-AP prescribing electronic laboratory order set significantly improved the rate of blood glucose, HbA1c and prolactin monitoring (Fig. 1a). Separately, in the group with chronic psychotic disorders treated with LAIs, our psychoeducational and awareness-raising intervention significantly improved blood pressure monitoring (Fig. 1b).

While there were improvements in the monitoring of some cardio-metabolic parameters, particularly related to blood markers in FEP, overall the rates in both FEP and chronic psychosis were considerably less than the 100% stipulated by the NICE guidelines [58]. Our post-intervention rates in the FEP group, while below the 100% guidelines, were approximately in line with the most recent data from the EIP spotlight audit (2018–2019) in the UK [65], highlighting the challenges of adhering to cardio-metabolic monitoring guidelines across different health care systems.

In contrast to our high rates of smoking, alcohol and illicit drug assessment in the FEP group, the rates of diet and exercise assessment were very poor (Table 2). Notably, physical exercise, a key component in reducing cardio-metabolic risk factors in general, was discussed only in 6% at baseline and 15% at re-audit, whereas diet was not assessed. This may indicate a lack of knowledge of evidence based guidelines among health professionals caring for those with FEP or, potentially suggest a professional mindset that those with psychosis were unlikely to engage in positive lifestyle behaviours. Regardless, it represents a missed opportunity for lifestyle modification, all the more important in the early stages of psychotic disorders [38]. Ongoing education of healthcare professionals and patients is necessary to improve incorporation of lifestyle assessment and advice in daily clinical practice in mental health settings. The inclusion of a self-screening checklist, in addition to the standardized AP information leaflet distributed to all patients starting on APs, may also improve rates.

A variety of “lifestyle interventions” aimed at adjusting dietary and/or physical activity behaviours, to benefit the physical health of people with psychotic disorders have been developed in recent years [76]. The majority are based on behaviour change psychological theory and require enquiry by healthcare professionals into an individual’s current lifestyle to establish a behaviour baseline to guide individualized advice and interventions. A pilot study of a physiotherapy-led motivational programme to increase physical activity and improve cardio-metabolic parameters in people with major mental illness in St. James’s Hospital showed promising results [4], but the consistent integration of standardized, effective and implementable physical health improvement interventions for people with psychotic disorders remain a challenge [2, 77]. Positive effects can be delivered, especially if delivered early, but may only last for the duration of the intervention [78]. Motivational issues are problematic, as highlighted by a 30% non-attendance rate in a naturalistic cohort study of an exercise physiology service in an FEP programme [2] and a 34% non-attendance rate for metabolic screening in patients prescribed LAI APs [31]. We await with interest the results of a trial investigating whether the introduction of more complex interventions such as a dedicated physical health nurse into an FEP programme will improve physical health outcomes [79].

The multi-level approach to mitigating the cardio-metabolic risk in psychotic disorders also includes personalized AP treatment strategies. APs form a vital adjunct to multidisciplinary psychosocial interventions, including cognitive behavioural therapy and family interventions in early onset psychosis [80, 81]. Consequently, an AP minimization strategy [82] based on individual preferences, avoidance of polypharmacy and high doses is essential [83,84,85]. Interestingly, the Australian early psychosis guidelines switched olanzapine to second-line treatment in 2014 [86]. This reduced the rate of olanzapine as a first line agent from 25 to 20% at the Early Psychosis Prevention and Intervention Centre (EPPIC) [86]. A recently published retrospective cohort study of EIP services in South County Dublin showed that 80% of EIP patients were initially prescribed olanzapine [87]. The HSE EIP model of care suggests olanzapine as a second-line agent, but advise that this “should be balanced by the relatively lower risk of extra-pyramidal side effects” [74]. In accordance, our rate of olanzapine prescribing decreased by 10% from baseline to 45% at re-audit, conversely aripiprazole doubled to 15% (Table 2).

The rates of monitoring in the FEP group were better than the chronic psychotic disorders group. Specifically, rates of weight, fasting lipids and fasting glucose monitoring were significantly higher in the post-intervention FEP group compared with the post-intervention chronic psychosis group (Table 3). While both groups incorporated awareness-raising educational sessions, only the FEP group had a specific electronic laboratory order set and physical health parameter checklist, which could account for the better rates. However, we acknowledge that a 3-month period between baseline and re-audit in the chronic psychosis group may have been insufficient. Nonetheless, our post-intervention monitoring rates in chronic psychosis, similar to the Lydon et al. study [31] were all below 45%.

Lydon et al. also highlighted the discrepancy between the higher rates of metabolic monitoring in those prescribed clozapine (>90%) compared with those in the LAI group [31]. This shows that a systematized approach such as dedicated clozapine monitoring clinics can increase metabolic monitoring. Indeed, there is some evidence to suggest that a dedicated metabolic clinic can improve monitoring rates in non-clozapine AP-treated patients [71]. It is noteworthy that despite the significantly higher levels of monitoring in the clozapine group, the rate of metabolic syndrome was 8% higher (at 39%) than the LAI group [31], which is most likely due to the less favourable metabolic profile of clozapine.

Prior research suggests that the prevalence of T2DM among individuals with schizophrenia is approximately 10% [32, 88]. The threshold for diagnosis of diabetes according to clinical notes or glycaemic testing in our group of people with chronic psychotic disorders was lower at 4%, which is double the rate of undiagnosed T2DM in the Irish adult population (1.8%) [89]. In terms of HTN, our 10% rate is approximately in line with the rate of clinically diagnosed HTN in the Irish population (13%) [90]. In our sample, none of the patients identified as having diabetes were on diabetic treatment (Table 3), whereas 2.4% of those identified as HTN were prescribed anti-hypertensive treatment. These findings clearly indicate a deficit in the model of shared care between primary and secondary care, which is imperative for improved outcomes for patients across the psychotic disorder spectrum [91, 92].

Undoubtedly, this could be improved by greater integration of connected IT systems between primary and secondary care [93]. At a local level, occurring post completion of this study, St. James’s Hospital introduced a hospital-wide electronic patient record (EPR) system. It would be interesting in future research to investigate whether this hospital wide EPR system has influenced the rates of cardio-metabolic monitoring and associated utilization rates of the electronic laboratory order sets. Although mental health services are often the main source of metabolic monitoring for patients, it is possible that some patients had metabolic monitoring carried out by their own GP’s. This data was not captured in our study, and this limitation of our study further re-enforces the importance of implementing a shared electronic system between primary and secondary care.

COVID-19 will likely make cardio-metabolic monitoring even more challenging, but simultaneously it may expedite the integration of enhanced IT solutions. For example, an estimated 80% of people with psychotic disorders own a mobile phone and a majority report a willingness to use this modality for self-support and enhanced contact with mental health services [94]. There are preliminary indicators that the tele-medicine approach can be useful for exercise and weight management in people with psychotic disorders [95,96,97,98,99,100], and the extension into personalized home-based physical health monitoring systems, potentially assisted by MDT members may also be on the horizon.

In summary and in keeping with previous research, our study highlights the challenges of improving cardio-metabolic health monitoring practises for people with both early- and chronic-stage psychotic disorders. This study identified suboptimal rates of cardio-metabolic monitoring and lifestyle factor assessment, notably diet and exercise in AP-treated FEP patients. Moreover, it showed that the implementation of a targeted awareness-raising educational intervention and introduction of a physical health checklist and AP electronic laboratory order set resulted in significant improvements in the monitoring of blood glucose, HbA1c and prolactin levels in FEP. Separately, a targeted awareness-raising educational intervention improved the routine monitoring of blood pressure in LAI-treated patients with chronic psychotic disorders. While awareness-raising activities may play a limited role, our study indicates that standardized systems for monitoring are more effective. In line with previous studies, our post-intervention rates of monitoring were well below the 100% guideline requirement.

Conclusions

Targeted improvement strategies resulted in significant improvements in a limited number of cardio-metabolic monitoring parameters in early and chronic psychosis. However, monitoring remained suboptimal in both groups. Improving cardio-metabolic health across the psychotic disorder spectrum requires ongoing awareness-raising and promotion of physical health among MDT members and patients, the availability of effective physical health interventions and enhanced collaboration between primary and secondary care, all facilitated by standardized systems and enhanced utilization of IT platforms. This is all the more important in the era of COVID-19.

Limitations

Direct comparisons between the FEP and the chronic psychotic disorder groups are limited as the interventions were different and occurred over different time scales. Not all post-intervention measures were captured (i.e. diet and exercise). In the group with chronic psychosis, the 3-month period between baseline and re-audit may have been insufficient to fully capture potential improvements. In chronic psychosis, the rate of cardio-metabolic monitoring could be different between those who are prescribed oral APs compared with LAI APs. The chronic psychosis group in our study only contained patients prescribed LAI APs and therefore may not be representative of the broader chronic psychosis group. In addition, the type of LAI AP was not recorded. This study focussed solely on monitoring rates and did not investigate rates of support or interventions when patients screened positively. We did not specifically seek metabolic monitoring records from GPs.