Metabolic monitoring for patients on antipsychotic medication: are we failing to provide reasonable standard of physical healthcare?
- Correspondence to Valsamma Eapen
Department of Psychiatry, University of New South Wales, L1 Mental Health centre, Elizabeth Street, Sydney, NSW 2170, Australia;
Implications for practice and research
■ Patients with mental illness are vulnerable to weight gain and metabolic risk induced by antipsychotic medication, therefore, proactive monitoring and management are critical in improving outcomes.
■ Suboptimal metabolic monitoring even after the introduction of monitoring guidelines suggests the need for matching guidelines with appropriate training and quality assurance programmes.
■ There is a need to bridge the research gap through investigating the optimal methods of detecting, preventing and treating metabolic risk in mentally-ill patients; the implementation of effective monitoring methods is crucial.
■ An integrated evidence-based approach linking physical and mental health aspects using a shared care model with general and specialist services would need to be developed to suit the local heath service delivery system.
Previous research suggests that metabolic risks are increased in patients receiving antipsychotic medication and that there is appreciable yield from routine testing.1 However, there is concern that screening programmes are not effective and that implementation of monitoring guidelines is falling short of reasonable standards of care;2 this study examined these issues.
Mitchell et al conducted a systematic review and meta-analysis using the PRISMA (Preferred Reporting Items for Systematic reviews and Meta-analyses) guidelines. Studies were mainly identified through Medline/PubMed/EMBASE abstract databases from inception to May 2011. Data were extracted by one author and checked by a second author. Proportion meta-analysis was performed.
Metabolic monitoring rates were low with only blood pressure and triglycerides reaching above 50% but still suboptimal. There was a modest statistically significant increase in glucose testing following implementation of monitoring guidelines, but most testing procedures still remained inadequate or suboptimal except for weight (75.9%) and blood pressure (75.2%) monitoring which were noted as just adequate.
Although the PRISMA guidelines were used, none of the inclusion criteria were applied for stratifying studies at the analysis stage (eg, past vs current use of antipsychotics, population based vs insurance database) which may have impacted on the screening practices. Guideline Concordant standards are listed a priori but no rationale is provided as to why these standards were chosen. Identifying studies as high or low-risk of bias is critical in systematic reviews, but this was not done using a validated risk of bias assessment tool.
While ‘proportion meta-analysis’ is an appropriate method (estimating an overall proportion from a set of proportions), the main weakness is the failure to ensure that the numerators and denominators in the proportions are comparable across studies. Also metabolic screening rates across settings (population based; inpatient; metabolic clinics in the community) were pooled rather than separately analysed in individual populations.
A flow diagram would have been helpful to detail the number of divisions of data. Presenting median rather than mean age and the range would have been appropriate. The authors refer to different diagnostic groups but it is not clear as to their characteristics (eg, ‘non-schizophrenia group’) and which studies are being referred to. Furthermore, determining whether rates differed at the start or during the course of antipsychotic use did not identify the studies used nor at what time points they were monitored. Regarding the Forest Plot, some studies clearly are outside the combined estimate and careful examination of the study characteristics might have given some clues as to the reasons. With insufficient information on the study population, it is difficult to determine how some of these variables may have impacted on the outcomes. Combining screening estimates across study designs and settings is a major weakness of this study which in turn has affected the validity and generalisability of the results.
There is extensive discussion on the reasons behind low rates of monitoring which is useful in clinical practice. While reporting on the limitations, the issue of co-morbidity was mentioned for the first time but the concurrent use of other medication has been overlooked. This review highlights an important service gap in the physical care of mentally-ill patients on antipsychotic medication. The need for effective metabolic screening and the low uptake of routine monitoring practices in this population is consistent with findings from previous research.