Hyponatraemia and mortality in psychiatric patients: protocol for Bayesian causal inference study

IntroductionHyponatriaemia has been associated with mortality in somatic patients, but little is known about the nature of such an association in psychiatric patients. This manuscript details a study to investigate if hyponatraemia be a direct contributor to death in schizophrenic and bipolar patients. It will further gauge the potential role of somatic comorbidity and psychotropic polypharmacy, with emphasis on antipsychotics and antidepressants. MethodsDensity sampled case-control study of schizophrenic and bipolar patients from the Capital Region and Region Zealand in Denmark using prospective data from Danish registries and electronic medical records. Cases deceased between 1 January 2011 and 1 July 2016; controls were found by risk-set sampling, matched by date of birth +/- 180 days. We will use a fully Bayesian, non-linear, conditional logistic regression model to estimate the causal effect of serum sodium level on death, with covariates selected by a causal diagram. Continuous covariates will be modelled with restricted (natural) cubic splines to elicit non-linear relationships. A sample of 10 cases and 50 controls was drawn to ascertain that the proposed study be sound and realistic. ConclusionThis study will expand the body of evidence on the effect of hyponatriaemia on mortality in schizophrenic and bipolar patients and may guide clinical decision-making in terms of both monitoring intensity and rational use of psychotropics.