Cardiovascular risk screening of patients with serious mental illness or using antipsychotics in family practice.

Patients with a serious mental illness (SMI) and those using antipsychotics (AP) have an elevated risk for cardiovascular disease (CVD). In the Netherlands, the mental healthcare for these patients is increasingly provided by family practitioners (FPs), following a shift from secondary to primary care. It is therefore essential to better understand the characteristics of this patient group and the (somatic) care provided by their FPs. The aim of this study was to examine the rate of cardiovascular risk (CVR) screening in patients with SMI or those using APs in family practice. We performed a retrospective cohort study of 151,238 patients registered with 24 family practices in the Netherlands. From electronic medical records, we extracted data concerning diagnoses, the measurement of CVR factors, medication, and the frequency of visits over a two-year year period. The primary outcome was the proportion of patients who were screened for CVR factors. We compared three groups: patients with SMI or using AP without diabetes or CVD (SMI/AP only), patients with SMI or using AP and diabetes mellitus (SMI/AP+DM), and patients with SMI or using AP and a history of CVD (SMI/AP+CVD). We explored the factors associated with adequate screening using a multilevel logistic regression.

respectively). A high frequency of FP visits, age, the use of AP, and a diagnosis of chronic obstructive pulmonary disease were associated with a higher screening rate. In addition, we examined the differences between patients with SMI and patients using AP in the absence of a SMI.

Conclusions
CVR screening in patients with SMI or using AP is often inadequate or lacking in Dutch family practices. Acceptable screening rates were found only among SMI/AP patients with diabetes mellitus as a comorbidity.
In the UK, a SMI register has been established (37); however, the monitoring of CVR in patients receiving AP in the absence of SMI remains unaccounted for. The mental healthcare for patients with SMI or those receiving AP (SMI/AP) in the Netherlands, as in the UK, is increasingly provided by FPs following a shift from secondary to primary care (38)(39)(40). This creates an opportunity for the uptake of CVR screening in the customary chronic care programs for these patients, with a financial incentive for the FP. FPs play an important role in CVR screening because their daily tasks include the prevention of CVD in high-risk patients. The care of patients with SMI or those using AP by FPs also introduces the question of who is responsible for CVR screening in relation to AP medication use: The initiator (usually the psychiatrist) or the doctor who prescribes the continuation (usually the FP). It is therefore essential to better understand the (somatic) care provided by FPs for these patients.
The primary aim of our study is to examine the CVR screening practice in patients with SMI or those using AP in family practice, and to identify the patient characteristics associated with the rate of screening. We will describe a) the screening rate in patients with SMI or those using AP without additional comorbidities, and compare this to b) the screening rate in a group of patients with SMI or those using AP who have an additional reason for CVR screening, such as diabetes and/or a known cardiovascular morbidity. The latter shows what can be achieved in primary care in this patient category, despite the above-mentioned barriers, while the former shows what is typically achieved for patients with SMI or those using AP alone.

S t u d y d e s i g n
This study is a retrospective cohort study of patients with SMI or those using AP in Dutch family practice.  (43). The database contains reliable data because nearly everyone in the Netherlands is registered in a family practice, and because FPs classify each visit using the ICPC system. The FP operates as a "gatekeeper" for secondary care, with medical specialists subsequently informing the FP about the diagnosis and treatment of referred patients (44). The electronic records for outpatient psychiatric visits in the Netherlands are separate from the FP system; therefore, visits to a psychiatrist and data concerning CVR collected there were not included in this study.
We collected data for patients who require a yearly assessment of their CVR based on their psychiatric disorder or their use of AP medication or lithium (22). A total of 151,238 people were included in this study, all of whom were registered at one of the 24 involved family practices, which were selected by region and for the availability of data in the FP database between January 2013 and December 2014. We selected patients with (I) schizophrenia, affective psychosis, bipolar disorder, or psychosis not otherwise specified (NOS) with a diagnosis date prior to January 1st, 2013; or (II) at least two prescriptions of AP or (III) a prescription of lithium prescribed for the first time before July 1st, 2013. This date was chosen because we only had access to the prescription records in this defined study period. Patients were excluded if they were (I) younger than 18 years old; (II) diagnosed with dementia; (III) diagnosed with delirium without the presence of a psychotic disorder; (IV) not registered in the selected family practice for more than 12 months in our study period, since FPs usually assess a patient's CVR profile once a year (45); or (V) diagnosed with rheumatoid arthritis, since CVR assessment in this patient category was introduced just before our study period and could possibly have confounded our results (45,46).

D a t a c o l l e c t i o n
Patients with SMI or those using AP were divided into three groups: (I) patients without another indication for a yearly assessment of CVR, according to the current FP guidelines (45) (SMI/AP-only group); (II) patients with SMI or those using AP and diabetes mellitus (DM), and thus an extra indication for CVR assessment (SMI/AP + DM group); or (III) patients with SMI or those using AP and a history of a CVD (i.e., stroke, angina pectoris, acute myocardial infarction, transient ischemic attack, intermittent claudication, and aortic aneurysm), and therefore an extra indication for CVR assessment (SMI/AP + CVD group). Patients with both DM and CVD at the baseline were included in the SMI/AP + DM group because patients with DM are routinely part of a chronic care program that proactively invites patients for monitoring.
Our primary outcome measure was the screening rate of CVR, defined as the proportion of patients in each subgroup who received screening for CVR factors in the defined study period. The CVR factors selected were those recommended by the Dutch FP guidelines; body mass index (BMI), blood pressure, estimated Glomerular Filtration Rate (eGFR), smoking status, fasting glucose, a lipid spectrum, use of alcohol, and a family history of CVD (45). Considering the observational nature of this study and the screening criteria described in previous studies (30,31), we also included a broader range of assessments (Appendix A1).
Based on current Dutch FP guidelines, we divided the observed screening rate into three levels: adequate, moderate, and insufficient (45). The screening rate was considered 'adequate' when the BMI, smoking status, blood pressure, glucose, and cholesterol/HDL ratio were all recorded at least once during the observation period, since these are the assessments needed to quantify the 10-year CVR of a patient and identify the need for CVR-lowering medication. The screening rate was considered 'moderate' when the assessment included BMI, smoking status, and blood pressure, all of which can be measured without a blood test, while the screening rate was considered 'insufficient' if it did not meet these requirements. A two-year window was chosen to gain insight into the role and awareness of the FP in this matter. FPs usually invite their high-risk patients for screening once a year, so the two-year window ensured that patients who were screened at intervals just over one year because of a delay in their response would be included in the screened cohort.
Moreover, we wanted to identify factors associated with any CVR screening (adequate or moderate). The following factors were studied: age, sex, type of psychiatric disease, use of AP, use of antidepressants, CVR medication (i.e., statins, blood pressure drugs, and aspirin), chronic obstructive pulmonary disease (COPD), abuse of alcohol or drugs, any records of social issues, and the frequency of FP visits. We selected ICPC-codes concerning diseases and social problems (see Appendix A2) and obtained prescription records of antidepressants for this purpose. The ATC codes of AP, lithium, and antidepressants are listed in Appendix A3.
S t a t i s t i c a l a n a l y s e s Descriptive analyses were used to describe the patient characteristics and provide insights into the screening rates in the three different patient groups. The hierarchical structure of the study (patients nested within practices) required us to perform multilevel analyses (random intercept model) taking into account the variability associated with each level of clustering. A multilevel logistic regression analysis was performed to test the differences in screening rates between the three groups.
In addition, for the SMI/AP-only group, we investigated the patient characteristics (Table 1) associated with an adequate or moderate screening rate. First, we included characteristics for the multivariate model that were univariately associated with screening (p < 0.20), after which a backward regression analysis was performed for these characteristics. A p-value < 0.05 was considered to be statistically significant, based on two-sided tests. A further analysis was performed to determine whether the results differed between the patients who were included based on their diagnosis (SMI) and the patients who use AP without a SMI diagnosis (see Addendum 1). All analyses were performed using IBM SPSS statistics 22.0.

Results
Of the 2247 patients with SMI or those using AP (prevalence in the 24 FP practices = 1.5%), 542 were excluded. Figure 1 shows the flow chart of in-and exclusion of these patients. C V R f a c t o r a s s e s s m e n t   Multivariate multilevel logistic analysis showed that a high frequency of visits, age, AP use, and a diagnosis of COPD were positively associated with an adequate screening rate in the SMI-only group (Table 3). SMI and AP are correlated, and therefore could not be simultaneously included as part of the model. We chose to include the variable with the most significant p-level, which was AP use. Cardiovascular risk screening was considered to have been performed if the assessment included at least BMI, smoking status, and blood pressure. All significant variables identified using a logistic regression analysis (p < 0.05) were included in this backwards stepwise regression procedure. *Reference is ≤ 10 visits FP/year. OR: odds ratio; CI: confidence interval; AP: antipsychotics; COPD: chronic obstructive pulmonary disease; FP: family practice.

Discussion
S u m m a r y The rate of adequate CVR screening by FPs in patients with SMI or those using AP is very low (8.5%). In patients with an additional comorbidity that requires CVR screening, this rate was considerably higher, especially in patients with type 2 diabetes (68.4%). The screening rate increased with age and the number of visits, with AP use and the presence of COPD also being associated with a higher CVR screening rate. Furthermore, it was striking that, in the majority of patients using AP, a diagnosis of SMI was not recorded in their EMR.

S t r e n g t h s a n d l i m i t a t i o n s
The main strength of our study is the size of the study sample and the broad inclusion of patients, based on diagnoses or on prescriptions of AP, which resulted in a realistic overview of the number of psychiatric patients with an increased CVR in primary care. We therefore think the diversity of our study group is representative of primary care patients in the Netherlands, which contributes to the validity and reliability of our findings.
Several limitations should also be mentioned. First, we studied whether FPs screened patients with SMI or those using AP for their CVR; however, the retrospective design used offers limited insight into their motives. Second, we did not have access to patient records in mental health institutions, since we only used the EMRs from FPs. About half of patients with SMI receive (additional) care from such institutions (47); consequently, it is possible that CVR was assessed in mental healthcare institutions, meaning our results underestimate the level of CVR screening received by these patients. (47)Third, it is important to keep in mind that the exclusion of patients who were listed for less than 12 months in a family practice (n = 225, 10% of all patients) could result in a potential selection bias. Patients who switch FPs regularly might be homeless, uninsured, or move frequently, and consequently might not be screened at all. Their absence in our study could have resulted in an overestimation of CVR screening. Fourth, we think the relatively small number of patients reported to abuse of alcohol (5.9%) and drugs (6.2%) is due to a lack of capturing these data in the FP EMRs. A study in the US found the prevalence of substance use to be uniformly high in individuals with SMI (marihuana 43-52% and alcohol 26-30%) (48) The expected inverse relationship between alcohol/drug abuse and adequate CVR screening could therefore neither be proven nor rejected. Lastly, the large group of AP users without a SMI diagnosis may indicate an off-label use of AP; however, some of these patients could be those who did have SMI but whose FP lacked information about the precise psychiatric disease or did not use the correct code. In addition, there are a few onlabel indications for non-psychotic diseases, such as the use of Quetiapine for unipolar therapy-resistant depression. Other studies endorse the possibility of a high prevalence of off-label AP use (21,(49)(50)(51).
C o m p a r i s o n w i t h e x i s t i n g l i t e r a t u r e The screening rate for CVR in patients with SMI and/or those using AP has been evaluated in several studies in different countries, resulting in a wide range of screening rates (28,30,32,35,52,53). This variation can be explained by differences in the study population and methods, providing insights into the most important factors to take into account when considering an intervention. A study among patients with newly prescribed AP use in a US Medicaid program found that 79.6% of those without DM were tested for their glucose levels (non-fasting tests included) and 41.2% were tested for their blood lipids (35

Consent for publication: Not applicable
Availability of data and materials: The datasets used and/or analyzed during the current study is available from the corresponding author on reasonable request.

Competing interests:
The authors declare that they have no competing interest.

Funding:
No funding was provided for this study.