Objective There is interest in identifying reliable prognostic biomarkers in schizophrenia and related disorders. Serum inflammatory markers, such as white cell count and C-reactive protein (CRP), have been shown to be elevated during psychotic episodes; however, their pathogenic role is uncertain. There is limited data relating to their variability in clinical practice and relationship to clinical outcome. We have sought to investigate whether routine clinical case records contain the necessary data to further understand the relationship between serum inflammatory markers and prognosis.

Method This is a retrospective case note review of patients admitted to an inner city female acute psychiatry ward. Cases were identified by reviewing electronic ward round records. Patients included had a diagnosis of non-affective, non-drug induced psychosis (schizophrenia, acute and transient psychotic disorder, persistent delusional disorder, schizotypal disorder and nonorganic psychosis) and had received a routine admission blood test. Exclusion criteria included pregnancy, significant recreational drug use prior to admission, clinical evidence of infection or history of inflammatory or haematological disease.

Results A total of 20 patients met the inclusion criteria between April 2015 and October 2016. Mean age was 43 years (SD=15) and the most common ethnicities were White British (23%), Mixed Ethnicity (23%), and Caribbean (23%). The majority of cases were detained under the Mental Health Act (68%) and had previously been treated with antipsychotic medication (94%). Mean admission duration was 38 days (SD=30) and average time from admission to routine admission blood test was 4 days (SD=3 days). Admission duration was moderately positively correlated with white cell count (r=0.41, n=20, p=0.07), platelet count (r=0.40, n=20, p=0.08) and albumin (r=0.42, n=20, p=0.07). Admission duration was weakly correlated with neutrophil count (r=0.27, n=20, p=0.24) and CRP (r=−0.22, n=20, p=0.35). When patients split according to those above and below the median admission length, patients with longer admissions had significantly higher platelet count (p<0.05) only.

Conclusion This small retrospective review suggests that in routine clinical case practice information is collected which could be used to explore the role of inflammatory markers as prognostic biomarkers in patients with schizophrenia and related disorders. Despite our extremely small sample we have found a positive correlation between platelet count and admission length. However this needs to be considered in the presence of multiple confounders. Use of electronic patient databases may be helpful in extending the sample to formally establish any prognostic relationships.