Characterised by central obesity, dyslipidaemia, hypertension and insulin resistance,1 metabolic syndrome is a major public health issue that affects a diverse range of population groups.2 Compared to the general population, metabolic syndrome increases an individual’s risk of developing cardiovascular disease by two times within the next 5–10 years and increases the risk of developing type 2 diabetes mellitus five times.2

Rates of morbidity and premature death are elevated for most mental health disorders, including depression, bipolar disorder, addictions and personality disorder.3 People who have a severe mental illness experience worse physical health outcomes than the general population.4 In New Zealand, mental health consumers experience over twice the risk of premature death compared to the general population and people with psychotic illnesses experience a three times higher risk of premature death.5 The life expectancy of a person with a severe mental illness is estimated to be 20 years less than the general population,6 and this mortality gap is reported to be increasing.7 Factors contributing to this disparity include a higher incidence of risk factors for chronic diseases, adverse effects of psychotropic medications, higher rates of unnatural death and reduced access to healthcare.4,8 In mental health consumers experiencing severe mental illnesses, modifiable risk factors contributing to this difference in life expectancy include higher rates of obesity, smoking, alcohol and drug misuse, poor nutrition and lower rates of exercise.9,10 These factors contribute to the increased prevalence of metabolic syndrome and consequently cardiovascular disease and type 2 diabetes mellitus.11,12 The prevalence of metabolic syndrome within this population group is also around two to three times higher than in the general population10 and even higher again in people experiencing schizophrenia.12

Second-generation antipsychotic agents are the mainstay medical treatment for people with serious mental illnesses such as schizophrenia and are also prescribed off label for other disorders in both children and adults.1 Second-generation antipsychotic agents may be preferred over first-generation antipsychotics due to the decreased risk of extrapyramidal side effects and reduced rates of relapse.1 Although the exact mechanisms are unclear, second-generation antipsychotics can contribute to metabolic complications and are associated with weight gain, increased triglyceride levels, decreased high-density-lipoprotein (HDL) levels and impaired glucose metabolism.1,6 It is important to note that independent of antipsychotic medications, severe mental illnesses such as schizophrenia are associated with other risk factors for the development of metabolic syndrome, such as obesity, smoking and lack of exercise.9,10

There is consensus across the literature that consumers prescribed antipsychotic agents should be closely monitored and screened for metabolic syndrome. However, the best model of monitoring and frequency with which it should be taking place is still unclear. In addition, there is a lack of worldwide consensus for the diagnostic criteria of metabolic syndrome, resulting in differences in estimates of prevalence and difficulties in comparing data across populations.2 Internationally, metabolic monitoring practices for mental health consumers prescribed antipsychotic medications have been found to be inadequate.13,14,15 A recent Australian study examining the attitudes of prescribing psychiatrists found that although 80% of psychiatrists surveyed felt that metabolic monitoring was their responsibility, the rates of metabolic screening were below 50%.16 The New Zealand Metabolic Working Group Initiative published guidelines in 2008 for monitoring for metabolic syndrome in mental health service users, especially those taking second-generation antipsychotic agents.17 The 2008 guideline did not recommend who should be clinically responsible for metabolic monitoring or how rates of monitoring should recorded, although it did suggest that this responsibility belongs with the prescriber. There have been no further guidelines released for the New Zealand population that align with the most recent evidence and therefore New Zealand has no national standards or guidance for metabolic screening within mental health and addiction services. A recently published guideline for the treatment of schizophrenia18 provides general recommendations for physical health monitoring but not a specific recommendation about metabolic syndrome. To date there has been no published studies considering the guidelines used and the rates of metabolic screening within New Zealand mental health services.

Considering the worsening physical health outcomes for mental health consumers,7 we sought to investigate policies for monitoring metabolic parameters for people with serious mental disorders treated with second-generation antipsychotic medications throughout New Zealand DHBs.

Methods

Metabolic screening policies and guidelines related to health consumers on second-generation antipsychotics were requested from each of the 20 DHBs. Emails were sent with requests for the policies and available information about the rates of monitoring. Policies were analysed using as an audit tool, a best practice standard developed following a review of the literature. The definition of metabolic syndrome agreed by the researchers was the 2009 ‘harmonised’ definition developed by international cardiac and metabolic health organisations,2 and is the most recent available. The harmonised definition includes five factors: waist circumference, blood pressure, fasting triglycerdes, fasting HDL cholesterol and fasting plasma glucose. Abnormalities in three of the five factors must be present to meet the criteria for metabolic syndrome. Cut-offs have been established for each factor, with the exception of waist circumference, which varies according to ethnicity. Criteria for the diagnosis of metabolic syndrome are shown in Table 1. In addition to the five components of metabolic syndrome, we included frequency of monitoring in the best practice guideline.

Table 1: Criteria for metabolic syndrome.¥

1.     The health consumer must have at least three of the following risk factors:

• Waist circumference with ethnicity-specific values
  
• European
  
• Female ≥80cm, Male≥94 cm
• South Asians, Chinese, Japanese, Ethnic South and Central Americans
  
• Female ≥80cm, Male ≥90cm
• No current data for other groups including Māori or Pacific Island people. These groups should use European values until more research is performed.
  

2.        Systolic blood pressure of ≥130mmHg or diastolic blood pressure of ≥85mmHg

3.        Fasting triglycerides ≥1.7mmol/L

4.     Fasting HDL-cholesterol <1.03mmol/L for males and <1.29mmol/L for females

5.     Fasting plasma glucose ≥5.6mmol/L

¥Source: Alberti KG, Eckel RH, Grundy SM, et al.3

Recommendations for frequency of monitoring were developed from the literature, considering both practicality and time available to DHB mental health and primary care staff. The frequencies were based on the recently released British Association for Psychopharmacology guidelines19 and an Australian guideline published by Waterreus and Laugharne.13 Additional features incorporated into the best practice guideline were a definition of metabolic syndrome, a statement of interventions for when metabolic syndrome was detected, a statement about collaboration with primary care, identification of the clinician/s responsible for monitoring and a process for auditing the DHB’s rate of monitoring. Policies were read independently by two researchers and discussed to reach consensus about their content.

Results

We received a 100% response rate from the DHBs. Fourteen (70%) of DHB mental health services have a metabolic monitoring policy. Additionally, the policy from Ashburn Clinic (a private psychiatric facility) was also included in the study. Many of the policies were embedded within broader physical health polices which were not specific to metabolic syndrome. There was considerable variability in the size (number of pages) of policies, and their clarity when being implemented by clinicians. There was also considerable variability in the extent of information about metabolic syndrome contained in the policies.

Metabolic syndrome

Five of the 14 policies reviewed included monitoring for each of the five components of metabolic syndrome. Of these five policies, two included a complete set of cut-off values. In most cases where cut-off values were given, they were at variance with those of the harmonised standard. For each of the components of metabolic syndrome, policies had a varied range of recommended frequency of measurement. All 14 policies included measuring waist circumference with two policies using the parameters recommended by the International Diabetes Federation. Seven of the policies included cut-off values for waist circumference although frequency of measurement was different in each policy. All policies included measurement of blood pressure with six including cut-off values. All but one DHB included frequency of blood pressure monitoring. Thirteen of the 14 policies included the measurement of fasting plasma glucose with six policies including parameters for blood glucose levels. Thirteen recommended the measurement of triglycerides with eight specifying parameters. Finally, nine of the 14 policies included HDL-C levels with six specifying cut-off values. Six DHBs included LDL/HDL ratio, as recommended in the 2008 New Zealand guideline but not included in the 2009 harmonised guideline. One DHB policy included the use of a cardiovascular disease software tool PREDICT to measure risk.

Three policies included a definition of metabolic syndrome. Two policies defined metabolic syndrome according to the International Diabetes Federation criteria and one DHB defined metabolic syndrome per the World Health Organization criteria. One policy included a system for the diagnosis of metabolic syndrome (three or more of impaired glucose metabolism, increased HDL-C, increased blood pressure or increased waist measurement). No policy specified what to do if metabolic syndrome was detected, although several policies contained recommendations relating to specific parameters.

Interventions

Eight of the 15 policies included actions that should take place if abnormal results are shown. Of these, five referred to liaison with or referral to a general practitioner. Three policies mentioned considering the use of metformin to moderate the effects of antipsychotic-induced weight gain. All the polices mentioning interventions discussed education surrounding reducing risk factors by promoting healthy lifestyles, although none mentioned specific health promotion interventions. Examples included healthy eating and physical exercise. One DHB also stated that referral to an occupational therapist, dietitian or other community service as an intervention for someone whose cardiovascular disease risk was high.

Collaboration with primary healthcare

Eleven of the 15 policies referred to collaboration with primary healthcare services. Six made reference to mental health workers ensuring that service users were actively encouraged and assisted to enrol with a primary healthcare provider. Many of the policies ensured that the GP was notified of commencement of antipsychotic medication, metabolic status, current interventions, other investigations and recommended follow-up by written or other means.

Clinician responsible for monitoring

The area of clinical responsibility for monitoring showed the greatest variability. Eleven policies made a statement about this aspect, with responsibility variously assigned to the prescriber, the psychiatrist, general practitioners, registered nurses, key workers and case managers. In one case all DHB clinical staff were assigned responsibility for monitoring and in several cases there was shared responsibility. Four policies made no statement about who held clinical responsibility for monitoring.

Auditing procedure

Four of the 15 policies included reference to auditing with intervals of either six months or 12 months specified. Of the four policies with audit procedures, two specified an annual random sample of 10 cases with a detailed description of the audit process.

Discussion

This is the first study to analyse the New Zealand DHB policies on metabolic monitoring for consumers prescribed antipsychotic medication. Fourteen DHBs and Ashburn Clinic have metabolic monitoring policies. The proportion of DHBs having metabolic monitoring policies (70%) is slightly higher than the 55% figure reported in Australia,16 but against that it needs to be noted that only two of the policies were consistent with our literature-based guideline. There is limited consensus on what constitutes a comprehensive metabolic screening policy and thus the calibre of monitoring varies widely across DHBs. When measured against a best practice standard, few policies included all necessary components such as cut-off values, a definition of metabolic syndrome and actions to be taken with consumers meeting criteria for metabolic syndrome. Policies also show a large amount of variation in frequency of monitoring and aspects such as clinical responsibility and collaboration with primary care. No DHBs have auditing procedures that would allow them to determine what percentage of mental health consumers have their metabolic status assessed. These findings are consistent with international research showing low levels of monitoring and lack of consensus on responses to metabolic syndrome15 and on what to monitor, what cut-off values should be used and when monitoring should occur.8

We have proposed a best practice guideline (see Appendix) combining the ‘harmonised’ definition of metabolic syndrome developed by Alberti et al,2 with recommendations for frequency of monitoring and other components based on our review of international literature. Our proposed model does not cover every contingency. People taking antipsychotic agents for the first time or children and adolescents with psychotic disorders are at higher risk of metabolic complications and consequently these groups should be monitored more closely for changes in metabolic status.1 This study only looked at physical health measures in relation to metabolic syndrome; it did not consider the place of metabolic monitoring within the overall physical health of mental health consumers. We also did not consider potential for other adverse effects of clozapine, or problems associated with lithium. Our guideline does not take into account age, smoking status or personal or family health history and is not intended to substitute for cardiovascular risk assessment as recommended in the New Zealand Primary Care Handbook.20 Rather than attempt to establish a model that would cover every variable, we aimed to produce a standard that is both practicable in terms of resources, and within the skillset of mental health and primary care nurses. In addition to recording metabolic parameters, our proposed model would incorporate evidence-based interventions aimed at preventing weight gain and improving cardiovascular health. Non-pharmacological interventions include nutrition and physical exercise education,21,22 and physical exercise alone.23 Metformin has also been suggested as a useful intervention24 for reducing metabolic abnormalities.

A prominent theme in the literature is the lack of consensus about whether responsibility for metabolic monitoring of people with mental illness belongs with mental health services or with primary care.25 Internationally, models of primary care differ significantly in coverage, skill mix and payment models, and any New Zealand guidance must be tailored to our primary care system. Even within New Zealand, there are different models and degrees of integration of primary care and mental health services, and any implementation of the guideline will need to accommodate regional variation and local service models. However, the literature is clear that closer integration of primary care and mental health is needed.26 with a growing volume of research suggesting that more integrated approaches—with professionals working more closely together—has the potential to improve outcomes and reduce costs.27 In the case of metabolic syndrome, the best practice guideline should be seen as applying to consumers wherever they are receiving their healthcare and continuing in times of transition from one healthcare provider to another.

One response to the issue of metabolic syndrome is the development of the role of ‘cardio metabolic nurses specialist’, a role dedicated to metabolic monitoring and interventions to reduce cardiometabolic risk.28 However, there is currently little evidence to demonstrate the effectiveness of this role and it carries the disadvantage that metabolic syndrome may be seen as a specialist area of clinical practice, rather than a component of the role of every clinician working with people taking antipsychotic medication. An English trial of a community nurse-led cardiovascular screening programme for health consumers with severe mental illnesses showed an increase in people being screened and proved superior to other interventions tested.29 A recent Australian study reported increases in metabolic monitoring following the location of a general practitioner clinic attached to a mental health service.30

A recent systematic review of screening practices for people treated with antipsychotic medication concluded that despite availability of guidelines, rates of screening are low.26 There is clearly a need to translate guidelines into practice. Qualitative research experiences and expectations of physical healthcare show that service users expect clinicians to work proactively with them in this area, especially if their motivation appears low.31 We suggest that in addition to screening there needs to be an emphasis on evidence-based interventions for those at risk. The best practice guideline suggested in this study could be implemented in both mental health and primary care services, the major aim being that all consumers at risk will be monitored, with an identified clinician managing the monitoring process. Consumers moving between services would maintain the same metabolic monitoring record in secondary mental health and primary care. Responsibility for maintaining the record could move between primary care and mental health services. Clinicians would ensure that the necessary tests are ordered, clinical assessments undertaken and interventions provided as indicated. In some cases, for example with consumers having routine cardiovascular assessment, metabolic monitoring might involve extracting laboratory results and information from clinical records rather than arranging additional appointments specifically for the purposes of metabolic monitoring. Maintaining a shared electronic record would help facilitate the sharing of information. Any such model will need to address attitudinal and skill barriers identified in the literature32 but would address the common finding of disagreement over who is responsible for monitoring.25 The model would need to involve collaboration and liaison with all health professionals, including psychiatrists, general practitioners, pharmacists, nurses, key workers and support workers. In some regions the cost of access to primary care is a barrier and initiatives such as the funded visits provided by Tairawhiti DHB is one way of addressing this issue.33 (http://www.tepou.co.nz/news/improving-access-to-primary-care-in-tairawhiti/628)

Need for Māori and Pacific parameters

Although the Alberti et al ‘harmonised’ statement on metabolic syndrome includes waistline parameters for various ethnic groups, it does not provide waistline parameters for Māori and Pacific people. A New Zealand cross sectional study found the prevalence of metabolic syndrome for Māori was 32% and Pacific 39% compared to 16% for the general population.34 Similar differences in obesity were found in the most recent New Zealand Health Survey.35 Future development of our guideline needs to consider whether waistline parameters are needed that are specific to these groups.

National monitoring

During the process of data collection many DHB staff expressed interest in the outcome of the research as a means of providing guidance on metabolic monitoring. We received many requests for results and recommendations for the study to assist DHBs in reviewing their metabolic monitoring policies. Based on these requests, and after considering the many literature reports of low levels of monitoring, it is our recommendation that this guideline is considered as a basis for a nationally standardised monitoring programme which is subject to audit and reporting to the Ministry of Health. Implementation would need to be managed jointly between DHB mental health services and primary care services, with the aim of continuity of monitoring across points of service transition.

Conclusion

The physical health status of people with mental illness is an issue of critical significance that needs to be a priority area for the mental health and primary care sectors. Metabolic monitoring for consumers taking antipsychotic medication offers one strategy of identifying those at increased risk of adverse cardiovascular events and diabetes. Our best practice standard could be developed into a practical tool for DHBs and the primary care sector. An inter-sector model of monitoring could see the implementation of more systematic monitoring of metabolic status for people with mental illness. The five-part model of monitoring, together with recommended actions, and collaboration between the mental health and primary care sectors and a system of measuring and reporting on adherence to the model, has the potential to increase levels of monitoring and to improve the health status of people with mental illness. There is a need for further research on effectiveness of service delivery models.

Appendix

Appendix 1:Metabolic monitoring best practice standard.

Frequency

Parameters

Interventions

Waist circumference

Baseline

Monthly for first 3 months

6 months

9 months

Annually

European:

Female ≥80cm

Male ≥95cm

South Asians, Chinese, Japanese, South and Central Americans:

Female ≥80cm

Male ≥90cm

No current norms for other groups. These groups should use the European values until more research becomes available.

Should be a recommendation or pathway to follow if an abnormality is detected.

Fasting plasma glucose

Baseline

3 months

6 months

9 months

Annually

≥5.6mmol/L

Should be a recommendation or pathway to follow if an abnormality is detected.

Blood pressure

Baseline

3 months

6 months

9 months

Annually

Systolic ≥130mmHg and/or

Diastolic ≥85mmHg

Should be a recommendation or pathway to follow if an abnormality is detected.

Fasting lipids (triglyceride)

Baseline

3 months

6 months

9 months

Annually

≥1.7mmol/L

Should be a recommendation or pathway to follow if an abnormality is detected.

Fasting lipids (HDL-cholesterol)

Baseline

3 months

6 months

9 months

Annually

Female <1.29mmol/L

Male <1.03mmol/L

Should be a recommendation or pathway to follow if an abnormality is detected.

Values outside the normal range for three or more of these parameters indicates metabolic syndrome. The person should receive evidence-based interventions aimed at restoring metabolic health. Consideration should be given to medication review. The health record should include a statement of the consumer’s involvement with their primary care provider. This could include ensuring the mental health consumer is enrolled with a primary care provider or ensuring clear communication with the general practitioner about the commencement of psychotropic medication, monitoring needed and expectations of the role of each health professional in caring for the health consumer. The policy should clearly state who is responsible for monitoring and the roles of each person involved should be clearly described. Lastly, the policy should include a way for DHBs to audit the monitoring process to determine the rate of adherence to the best practice standard for metabolic screening.

‍