Applicability of somatic monitoring instructions in clinical practice guidelines on antipsychotic drug use

The objective of this study was to assess the consistency in CPGs of somatic monitoring instructions for antipsychotic-induced adverse events and whether these instructions are applicable in daily clinical practice for adult and (frail) elderly populations.

We included both national and international CPGs for somatic monitoring instructions for adults (18–65 years old) and (frail) elderly patients using antipsychotic drugs. We defined frailty as “a condition in which the individual is in a vulnerable state at increased risk of adverse health outcomes and/or dying when exposed to a stressor” [17].

A search for national and international CPGs was performed by exploring the electronic database PubMed and a general search engine (Google).

Search terms were “antipsychotics”, “schizophrenia”, “bipolar disorders”, “dementia”, and “guidelines”. Criteria for the selection of CPGs for assessment were that the CPG 1) was clearly defined as a guideline, 2) was written in Dutch or English, and 3) included a section on antipsychotic treatment.

The search resulted in 114 CPGs of which 21 were not written in Dutch or English. CPGs were selected that were developed by major international societies with published CPGs on schizophrenia, bipolar disorders and dementia. The preselection procedure was conducted by consensus by HM and JB (Fig. 1).

In order to assess the consistency of somatic monitoring instructions in different CPGs, each selected CPG was examined by authors EOH and JB for somatic monitoring instructions related to antipsychotic drug use. All available somatic monitoring instructions were collected from the CPGs. Somatic monitoring instructions of additional CPGs were collected whenever a referral was made within a selected CPG.

Somatic monitoring instructions were defined as instructions to measure a parameter prior to and during the use of antipsychotic drugs.

We defined a monitoring parameter as a parameter that can be measured as an anthropometrical parameter (e.g. height, weight, blood pressure, or waist circumference), a laboratory parameter (e.g. blood glucose, HbA_1C, cholesterol, or prolactin) or a clinical monitoring parameter for indicating adverse effects (e.g. movement disorders).

Somatic monitoring instructions were classified as either a baseline monitoring instruction prior to the start of an antipsychotic drug or a monitoring instruction during maintenance therapy. When monitoring timing was not specified, the monitoring instruction was classified as both a baseline and a maintenance therapy monitoring instruction.

Somatic monitoring instructions were allocated either to mandatory or to recommended instructions based on how the monitoring instruction was formulated in the CPG.

Somatic monitoring instructions in CPGs were allocated to (frail) elderly psychiatric populations when these instructions were mentioned either in guideline sections specifically for (frail) elderly populations or in more general terms that every patient with the particular severe mental illness had to be monitored regardless of age.

The applicability of somatic monitoring instructions was assessed using the SIM score [14‐16]. The SIM score tool rates somatic monitoring instructions on the following six items: 1) what to monitor, 2) when to start monitoring, 3) when to stop monitoring, 4) how frequently to monitor, 5) critical values of the parameter, and 6) how to respond. Each item yields either a score of 0 (monitoring instruction is not clearly described) or a score of 1 (monitoring instruction is clearly described), resulting in a total SIM score varying from 0 to 6. Somatic monitoring instructions were considered applicable if they yielded an acknowledged minimal total score of 3 [16].

The authors believe that an important explanation for poor adherence to monitoring guidelines is the lack of information regarding which healthcare professional is responsible for the interpretation and treatment of aberrant monitoring outcomes. In order to address this issue, we focused on an additional item apart from the SIM score: who is responsible for treatment of aberrant monitoring outcomes. We did not add this item to the total SIM score because doing so would reduce the comparability of our results with those of other studies comparing CPGs with the SIM score.

The classification and assignment of the SIM scores of all somatic monitoring instructions was carried out by two authors independently (EOH and JB). Discrepancies were discussed with a third author (AR) until consensus was reached.

Outcome measures are the following: 1) an overview of recommended somatic monitoring instructions related to antipsychotic drug use per selected CPG; 2) the percentage of clearly described somatic monitoring instructions per selected CPG, outlined to each of the six SIM items and the additional item; and 3) the percentage of applicable somatic monitoring instructions per selected CPG.

For each CPG, the number of somatic monitoring instructions were presented along with the SIM score items, corresponding mean/median SIM scores, and the percentage of somatic monitoring instructions that were applicable (SIM score ≥ 3).

Although somatic monitoring instructions indicate which parameter should be monitored and how often, they lack information regarding the presentation of critical values, how to respond to aberrant outcome values, and whose responsibility it is to monitor and to intervene based on aberrant outcome values. The lack of information regarding whose responsibility it is to initiate monitoring may contribute to the low monitoring rates among psychiatric populations [38], as psychiatrists and primary health care professionals express different opinions regarding who is responsible for the monitoring and treatment of, for example, metabolic effects due to psychotropic drugs [39‐41]. In addition, the lack of clarity in how to respond in terms of interventions and who is responsible to intervene may be a contributing factor in the inadequate treatment of psychiatric patients in the case of, among other things, metabolic abnormalities [38, 42].

For example, Bruins et al. present persistently low treatment rates for metabolic abnormalities among patients with psychotic disorders, despite reasonable annual monitoring rates during a period of 3 years. More than half of the patients for whom pharmacotherapy for metabolic disorders was recommended did not receive any treatment [42]. Although the authors did not investigate the reason for these alarmingly low treatment rates, a contributing factor could be the lack of information in available guidelines for interventions in the case of aberrant outcome values.

The study by Bruins et al. was performed in the Netherlands. The Dutch guidelines contain neither information on interventions in the case of aberrant outcome values nor information on responsibility for treating metabolic abnormalities. Possibly, adding mandatory statements that manage outcomes of monitoring as part of integrative care should be addressed.

The number of instructions and the recommended frequency of monitoring differs significantly between CPGs. Although somatic monitoring among patients with severe mental illness is recognized internationally, some CPGs do not specifically recommend somatic monitoring during the use of antipsychotic drugs. For example, the APA guideline for the treatment of bipolar disorders and the NICE guideline for the treatment of dementia provide no instructions regarding somatic monitoring at all. In addition, CPGs that do present somatic monitoring instructions differ in the amount of instruction. This indicates a lack of consensus on what to monitor prior to and during antipsychotic drug use. This lack of consistency among guidelines may have two causes.

First, the reason for the lack of consistency among guidelines might be the scope of the guideline. Guidelines in which treatment with antipsychotic drugs is only (a minor) part of the treatment protocol probably provide less information on monitoring. This suggestion seems correct in view of the lower rate of somatic monitoring instructions in guidelines for dementia.

A second reason for the lack of consistency among guidelines might be the scarcity of evidence for the positive effects of structural monitoring and interventions in the case of aberrant outcome values. Recommendations in CPGs are preferably evidence-based. For example, (frail) elderly are often excluded from research resulting in hardly any evidence for monitoring in these patients. The lack of evidence results in recommendations for this population that are mostly empirical or based on consensus and therefore guideline committees will make different considerations for which somatic monitoring instructions should be part of a guideline [43‐45].

The results imply that there is no uniformity among CPGs regarding the content and number of somatic monitoring instructions prior to and during the use of antipsychotic drugs among adult and (frail) elderly populations. To standardize somatic monitoring instructions in CPGs, there is a need for consensus statements. In addition, more research is necessary to investigate the associations between the monitoring, prevention, and treatment of antipsychotic-induced adverse effects. In lack of evidence, consensus statements can be established by using, for example, the Delphi method. This method is used to reach consensus by sending questionnaires about a specified topic to a panel of experts. The response to these questionnaires is aggregated and send to the other experts. Experts can adjust their answers in another round based upon the answers of the group. The intended outcome after several rounds is a consensus about the topic supported by all experts. A good example of a consensus statement established with the Delphi method was published recently regarding the monitoring of lithium [46, 47].

This study shows that CPGs, including those for patients with dementia, hardly recommend specific somatic monitoring instructions for (frail) elderly populations using antipsychotic drugs. This might be explained by the lack of research regarding antipsychotic-induced adverse effects in (frail) elderly populations.

However, a systematic review and meta-analysis by Schneider-Thoma et al. indicated a higher risk of second generation antipsychotic-induced somatic serious adverse events in older populations compared to adult populations [48]. Because antipsychotic drug treatment in (frail) elderly populations should be safe, it seems reasonable to monitor according to the monitoring recommendations developed for adult patients. However, discussions may arise as to what to monitor and how frequently to monitor in elderly populations, especially when the frail and vulnerable elderly with short life expectancy are taken into account. For example, it can be argued that the measurement of blood glucose is still necessary in the (frail) elderly to prevent short-term complications, whereas the monitoring of blood lipids seems less necessary, given its association with longer-term complications in a population with a short life expectancy [49, 50].

The lack of specific somatic monitoring instructions for the (frail) elderly using antipsychotic drugs may affect medication safety. For example, Ndukwe et al. show that glucose monitoring was suboptimal in (frail) elderly patients using antipsychotic drugs, putting these patients at risk for hyperglycemia [9]. To address these issues, it is advisable to incorporate a specific section in current and future guidelines regarding monitoring instructions for relevant subpopulations, including the (frail) elderly, according to the SIM definitions.

A few limitations should be mentioned. Although the SIM score tool addresses the applicability of both the somatic monitoring instructions that are present and, therefore, the applicability of the clinical practice guideline, it does not assess the clinical relevance of the presented somatic monitoring instructions. Furthermore, in order to analyze a manageable amount of CPGs we used an arbitrary search strategy. We have based our results upon the selection of CPGs from major societies (APA, NICE and RANZCP) from three continents, one global society (WFSPBP) and CPGs made in The Netherlands. We do not know whether these results are generalizable to, for example, smaller societies and CPGs in other languages than English or Dutch. However, despite this limitation, we analyzed 16 CPGs and therefore we believe that our results provide valuable information to other international societies to critically reflect upon their CPGs.