Physical health examination in outpatients with schizophrenia: the cost effectiveness of laboratory screening tests

Health care workers providing services for people with schizophrenia and other severe mental disorders are advised to pay close attention to the physical well-being of their patients [1]. Patients with schizophrenia are prone to a variety of physical illnesses and risk factors, and have a shortened life-expectancy [2‐4]. Especially cardiovascular diseases and metabolic disturbances, such as metabolic syndrome, type 2 diabetes and obesity, are substantially more common than in the general population [5‐8].

Guidelines on laboratory monitoring in schizophrenia usually recommend fasting lipids and glucose testing annually after the first year of antipsychotic medication use [9]. The aim of the testing is to reveal increased cardiovascular risk, established diabetes, or severe hyperlipidemia, and subsequently offer preventive measures and/or timely treatment for these disturbances [1, 9]. In addition to the aforementioned laboratory tests, weight, blood pressure and cigarette smoking should be assessed at least once a year. Moreover, the British guideline for people with psychosis and schizophrenia recommends comprehensive, annual health checks that focus on common physical health problems [10]. Due to the extensive physical comorbidity [2], the question arises whether additional laboratory tests beyond fasting lipids and glucose that ought to be monitored in the annual physical health check for individuals with schizophrenia.

The need of laboratory screening among psychiatric patients has gained some scientific interest in the 1970s and 1980s [11, 12]. More recent research, conducted especially in the field of the emergency medicine, has mainly focused on the utility of medical clearance, i.e., use of laboratory testing in the differential diagnostics between physical and mental illnesses, once a patient with psychiatric symptoms is treated in the emergency department [13]. In addition, a few studies have examined thyroid function testing among psychiatric patients [14‐16].

To our knowledge, three studies have assessed the need and the cost effectiveness of routine laboratory screening upon patient´s admission to a psychiatric facility [17‐19]. Nowadays the treatment of schizophrenia is primarily provided on an outpatient basis. However, no studies exist on utility and cost effectiveness of routine laboratory screening panels as a part of physical health monitoring in outpatients with schizophrenia.

The aims of this study were to examine: (1) the proportion of test results outside the reference range in a basic laboratory panel and (2) the cost effectiveness of these tests among outpatients with schizophrenia. In addition, we studied whether patients who used clozapine or olanzapine or were obese had more abnormal laboratory results and would need more extensive laboratory screening.

Of 409 patients treated at the clinic, 275 (67.2%) participated in the health examination. The mean age of the participants was 44.9 years (SD 12.6) and 55.1% were men, 190 (68.8%) were diagnosed with schizophrenia, 49 (17.8%) with schizoaffective disorder, and 37 (13.4%) with other schizophrenia spectrum disorders (hereafter referred as schizophrenia). Of the sample, 96.4% used antipsychotic medication. Somatic medication use was common, as only 38 (13.8%) did not use any somatic drugs. Mean number of separate somatic medications was 3.3 (SD 2.69, min 0-max 13). Of the patients, 75 (27.2%) used lipid-lowering medication, 59 (21.4%) antihypertensives, 37 (13.4%) diabetes medication and 34 (12.3%) medication for hypothyroidism. A detailed description of the sample is given in Eskelinen et al. [3].

We evaluated the proportion and the cost-effectiveness of abnormal results in a basic laboratory test panel in a sample of outpatients with schizophrenia. Most participants of our study had a long duration of psychotic disorder and were clinically stable, and the majority of them had medications for physical illnesses and multiple physical health care needs [3].

If the direct cost spent to find one abnormal result (DCSAR) is used as the criterion to select screening laboratory tests, our results support the inclusion of glucose, lipids, sodium, creatinine, GGT, and CBC in the protocol. The DCSAR was below 10 € for all of them, and from 13 to 43% of the patients had abnormal test results in these tests. In an earlier study of newly hospitalized psychiatric patients, of whom 39% had a schizophrenia spectrum disorder, abnormal test results were less common, except for total cholesterol [18]. However, DCSAR values used in the study by Arce-Cordon are not directly comparable with ours, as we used the costs from the laboratory serving in our catchment area.

If the test with the highest number of abnormal results (NNSR) was to be chosen for screening, Vit D ought to be routinely measured. Vit D deficiency has been shown to be more common among individuals with psychosis compared to the general population [20, 21], and it increases the risk of osteoporosis and fractures, common in this patient group [22‐23]. In Finland a systematic Vit D fortification of liquid dairy products and fat spreads was started in 2003, and the dose was doubled in 2010. As a result, the mean Vit D concentration has increased at the population level from 48 (2000) to 65 (2011) [24]. Among the participants of our study conducted between 2009–2013, the mean Vit D level was lower (53). In addition, Vit D deficiency (< 50) was five times as common, compared to the Finnish adult population in 2011: 47% vs. 9% [24]. To date, the evidence does not support screening for Vit D deficiency in community-dwelling, asymptomatic adults [25]. However, patients with schizophrenia could be considered as a high-risk population for whom screening would be indicated. They often have an unhealthy diet and stay indoors, are smokers, have difficulties in moving, and may have extrapyramidal, anticholinergic and hypogonadal side-effects from antipsychotic medication, features that increase the risk of Vit D deficiency, falls and fractures. The DSCAR of Vit D testing was 15€ because the test is fairly expensive compared to the other tests. However, the testing could be reasonable due to straight forwardness of prescribing a Vit D supplementation for a deficient patient.

One in ten of the sample had TSH above the reference range (at highest 7.1), indicative of subclinical hypothyroidism. Among psychiatric patients, reversible and irrelevant abnormalities in thyroid function tests are frequent and often due to other factors such as acute psychiatric state, starvation or substance use, rather than to a thyroid problem per se [14]. Lachman and Garnier do not recommend TSH testing upon admission to psychiatric hospital due to low clinical utility and high costs [16]. In our analysis, TSH had the second highest DCSAR of all tests (33 €) and eight patients had to be screened to get an abnormal result. Interestingly, elevated TSH was three times more common in obese than in non-obese patients. The relationship between thyroid function and obesity has been reported to be bidirectional: hypothyroidism is associated with weight gain, but obesity also influences thyroid function [26]. The association between elevated TSH and obesity is mainly regarded as an adaptation, and the appropriate therapy is adjustment of energy balance and body weight [27]. Based on the current evidence, an annual screening of TSH in obese patients with schizophrenia is not necessary, but the possibility of hypothyroidism should be kept in mind in patients with rapid weight gain.

Elevated GGT was strongly associated with obesity in our sample. This finding, along with a notable proportion of obese patients with high triglycerides and low HDL cholesterol would fit with the features of non-alcoholic fatty liver disease (NAFLD) [28]. Obesity is the major risk factor for NAFLD, and a subtype of NAFLD (NASH, non-alcoholic steatohepatitis) may progress to liver cirrhosis, cancer and extrahepatic diseases [29, 30]. In a study assessing first episode, drug-naïve psychosis patients, one-fourth with normal NAFLD index at the baseline had an index indicative of NAFLD at 3-year follow-up from the initiation of antipsychotic medication [31]. Several risk factors of liver disease are common in patients with schizophrenia: obesity, medications involving hepatic metabolic pathways, higher prevalence of hepatitis C and alcohol use disorder. Therefore, we recommend routine monitoring of a liver function test among patients with schizophrenia.

Clozapine, along with olanzapine, is metabolically the most hazardous antipsychotic drug [32], yet the most effective one [33]. Our findings are in line with the previous research: patients using clozapine hade more often triglycerides and glucose above the reference range. It should be noted that hypertriglyceridemia can be dangerous. Drug-induced acute pancreatitis may rarely be secondary to antipsychotic use-induced hypertriglyceridemia (usually triglyceride level has to rise above 10) [34]. In our sample, thrombocytes were more often below the reference range among clozapine users. However, clozapine-induced thrombocytopenia is often mild and transient and seldom causes clozapine discontinuation [35].

As a secondary analysis, we excluded patients who used clozapine and compared patients who used olanzapine to the other patients. There were no differences in lipids or glucose levels. However, olanzapine is often switched to another antipsychotic if significant weight gain or metabolic abnormalities emerge [36], and therefore current users may be those who were less prone to these side-effects. Regarding elevated creatinine in patients using olanzapine, a recent study found that second-generation antipsychotic use was associated with elevated risk of chronic kidney disease, and the highest risk was found for clozapine use, followed by olanzapine [37]. As there were only 41 patients who used olanzapine in our study sample, this finding warrants further investigation in a larger study sample.

In the 1980s study of 205 stable psychiatric outpatients, with mixture of psychiatric diagnoses, 43% had high fasting glucose, 14% anemia and 5% hyponatremia [11]. In our study almost 40% had high glucose, less than 5% low hemoglobin and 20% hyponatremia. In the study by Beresford [11], anemia was especially prevalent among the aged participants, whereas our study group consisted mainly of middle-aged patients. Higher prevalence of hyponatremia may be due to common use (32% of the sample) of antidepressive medication, mainly selective serotonin reuptake inhibitors, which may cause hyponatremia [32]. However, primary polydipsia is also relatively common in patients with schizophrenia and may contribute to the high prevalence of hyponatremia as well [38].

According to a recent review, a routine laboratory screening is not recommended for patients admitted in the emergency department for a psychiatric reason [13]. In the outpatient care, the rationale for laboratory screening is somewhat different. Patients often develop metabolic disturbances during the treatment with antipsychotics, and the use of other psychotropics (e.g., lithium, valproic acid, antidepressants) may require laboratory monitoring as well [32]. In addition, concomitant use of psychiatric and somatic medications may cause need for laboratory controls (e.g., selective serotonin reuptake inhibitors and diuretic medication may increase the risk of hyponatremia).

According to our results, glucose, lipids, sodium, creatinine, liver function test, CBC, and preferably Vit D ought to be screened annually among outpatients with schizophrenia. There is no need for more extensive routine laboratory monitoring among clozapine users or obese patients, with the exception of TSH if the weight gain has been rapid. Most importantly, clinically significant abnormal laboratory test results should be evaluated and followed by relevant interventions.