Are the cardiometabolic complications of schizophrenia still neglected? Barriers to care

Mortality rates of people with schizophrenia, even after accounting for suicide, have been described as excessive and premature.1 Death may occur 10–25 years earlier than in the general population,2,3 and cardiovascular disease (CVD) is the major cause.1 The cardiovascular standardised mortality ratio among people with schizophrenia is twice that in the general population (male, 2.3; female, 2.1),4 and the gap appears to be widening.5

These increased rates may reflect the higher than expected range of cardiometabolic risk (CMR) factors experienced by patients with psychotic disorders. First, the metabolic syndrome, a predictor of early CVD and premature mortality, is more prevalent in people with schizophrenia than in the general population.6,7 Compared with the general population, among people with schizophrenia there is an increased prevalence of obesity (1.5–2 times), diabetes (2 times), dyslipidaemia (5 times), and smoking (2–3 times).3 Such patients have a profile similar to members of the general population who are 10 to 15 years older than they are.6 Second, patients with schizophrenia have more prevalent family histories of diabetes and CVD and, although it remains unclear whether this might be genetically or environmentally determined, schizophrenia itself may confer an increased risk.8 In Australia, a large proportion of patients receiving care in the public psychiatric system are from ethnically diverse backgrounds, and may be at higher risk of developing type 2 diabetes because of their ethnic status.9 Third, antipsychotic medications are known to increase patients’ risk of developing diabetes.10 This suggests that antipsychotic choice is a key step in preventive action.11

In the past 5 years, a number of national and international guidelines have been developed to promote cardiometabolic health in patients with schizophrenia.12 Despite the introduction of such guidelines, there is little evidence to suggest that CMR screening has substantially increased.13-16 A population-based study in the United States showed that between 1998 and 2003 there was a small, clinically insignificant increase in the rate of glucose (but not lipid) screening.13 Further, among patients requiring treatment with a second-generation antipsychotic medication, baseline glucose levels were estimated in 19%, lipids in 6%, and both in 5%.13 Predictors of receiving screening included service location, ethnicity, sex, pre-existing disorder, and which antipsychotic was prescribed. This is in contrast with guidelines, which suggest that all patients with psychosis treated with any antipsychotic medication should be regularly monitored.17 The problem may lie with the fact that in many branches of medicine, including psychiatry, evidence-based guidelines do not appear to be readily disseminated into real-world practice.

Several surveys and audits have assessed psychiatrists’ and mental health teams’ knowledge and practice of their patients’ cardiometabolic health. In a US postal survey, 97% of psychiatrists rated CMR monitoring activity as either a “very serious” (36%) or “serious” (61%) concern when prescribing antipsychotics.18 However, this and other surveys reveal a broad disparity between an awareness of the need for, and the actual performance of, adequate monitoring.18,19 Box 1 shows that although ratings concerning the difficulty of obtaining CMR data are generally low, the proportion of patients not receiving the tests in practice is much higher.

One might expect regular assessment of obesity to be more common. A number of lines of evidence suggest that adiposity is a key to identifying at-risk individuals.3 US researchers found that body mass index (BMI) and waist circumference significantly predict insulin sensitivity in healthy controls and participants with schizophrenia,21 suggesting that mental health workers should consider carrying and using a tape measure as part of their usual clinical interaction with patients. Alarmingly, one US survey revealed that although obtaining waist measurements was “difficult to obtain/unobtainable” for 42% of respondents, this parameter was not measured in 95% of patients.18 A study in the United Kingdom found that 70% of mental heath teams reported having no access to a tape measure.20 This is unfortunate because waist circumference, rather than BMI, may be the best tool for screening, as it is more reliably associated with the prediction of myocardial infarction.22

Blood pressure measurements were deemed difficult to obtain in 17%–18% of patients, and yet not measured in 45%–65%.18,20 In the UK, about 40% of respondents had issues with the reliability or retrievability of test results,20 somewhat higher than the quarter that suggested difficulty in obtaining these data in the US.18 With respect to baseline and ongoing blood tests for glucose and lipids, there is remarkable concordance between studies in the rates of non-assessment — 60%–65% of patients do not have regular glucose monitoring and 70%–75% do not have regular lipid monitoring (Box 1).18-20 Similar figures were found in an outpatient study that contrasted patients with psychotic disorders with those with HIV.23 In this study, over 90% of patients with HIV had chart recordings for CMR factors (smoking, lipids, blood pressure, BMI, glucose). Patients with psychosis, on the other hand, had 28%–70% of tests for these risk factors charted.23

What are the barriers that deny psychotic patients these same standards of care? A considerable number of barriers have been identified that prevent patients with mental illness from receiving adequate physical care. Box 2 outlines barriers associated with the health care system and health care professionals. Box 3 outlines barriers related to patients and their illnesses.

The problems lie not only with screening and monitoring, but also in taking action once CMR factors are identified. In assertive community teams, where services should be integrated, high rates of non-treatment were identified even when risk factors had been detected — 38% of patients with diabetes, 63% of those with dyslipidaemia, and 52% of those with hypertension were not receiving treatment for these conditions.20 It should be stressed that this study also indicated that for every case of diabetes, hyperlipidaemia, and hypertension recorded, the anticipated rates of disease incidence, based on previous studies, were expected to be 2, 7 and 4 times higher, respectively. These figures support previous findings of undertreatment of metabolic comorbidities in the CATIE (Clinical Antipsychotic Trials in Intervention Effectiveness) study — 30% of patients with diabetes, 88% of those with dyslipidaemia, and 62% of those with hypertension.14

Mental health clinicians may transfer a degree of therapeutic “nihilism” regarding schizophrenia outcomes to such patients’ physical health. However, many barriers outlined in Box 2 and Box 3 can be remedied by clinicians taking a more proactive role. Physical examinations are performed less often than necessary, even though structured physical assessments of patients with schizophrenia are effective in revealing physical illness.27 Many medications used in general practice (statins, hypoglycaemics, antihypertensives) are infrequently prescribed in mental health settings despite their effectiveness. For CVD, removing or reducing risk factors improves outcomes over a period of months to several years. Modest decrements can have large clinical effects — for example, a CVD risk reduction of 30% may be achieved by reducing cholesterol by 10%.1 The lack of coordination between primary and secondary care is a notable barrier that may perpetuate undertreatment.

A number of changes to services, training, and (in many countries) funding are needed to improve patient outcomes.30,34 Central to the problem is poor integration between the general and specialist mental health sectors. Although it is beyond the scope of this article to discuss policy issues in depth, based on the obstacles to integrated care, Horvitz-Lennon and colleagues have proposed a number of policy recommendations to address the need for improved integrated health care for those with major psychiatric illnesses (Box 4).30

There is a dearth of research into models to improve medical care despite the premature mortality, long morbidity and the barriers outlined above. A range of models should be provided and adapted for selective use, reflecting variations in local micro-medical culture and its specific barriers.36

Finally, screening and management for CMR factors should be undertaken at a younger age as the relative risks may be higher for patients in this group.37 Ultimately, the greatest potential for decreasing the premature mortality from CVD in patients with psychoses is primary prevention in combination with a reduction in the misuse or underuse of evidence-based general medical procedures.15,34

Compared with the general population, patients with persistent psychotic disorders, such as schizophrenia, experience significant physical health problems, leading to decreased quality of life and premature mortality. This occurs despite a growing awareness of the need to screen for common metabolic risks. The disparity between “knowing” and “doing” may be attributed to a number of health care barriers that should be urgently addressed in this high-risk, undertreated population. The solution, at least in a preliminary form, may lie in the implementation of a variety of strategies that reinforce the need for truly integrated health care between the primary and secondary health sectors — mandated by government, and subject to stringent accreditation assessments.