People with psychosis may consult a GP on a number of occasions before a formal diagnosis is made and antipsychotic medication prescribed (which is often only after confirmation of diagnosis in specialist care). A delay in diagnosis may increase the risk of poor long-term outcomes. Identifying common patterns of symptoms and comorbidities presented before diagnosis may help patients be identified earlier and improve long-term outcomes.
We will assess if patients with newly diagnosed psychosis have a history of consultation for symptoms potentially related to psychosis, and determine whether there are common patterns and combinations of symptoms and other illnesses presented to primary care prior to diagnosis.
The study uses a national electronic health record database (Clinical Practice Research Datalink). We will compare 12,000 patients newly diagnosed with psychosis to people without such a diagnosis on previously recorded symptoms (including anxiety, depression, hallucinations) and other illnesses. We will determine time from first recorded symptom to diagnosis and to treatment of psychosis. We will use novel statistical methods to identify common patterns (phenotypes) of health care use, symptoms and comorbidity in primary care prior to diagnosis and assess variation by gender, age, ethnicity, and type of diagnosis (schizophrenia, bipolar affective disorder).
Long duration of untreated psychosis (DUP), the period between onset of psychotic symptoms and treatment, is associated with poor outcomes. Initiatives to shorten DUP have largely been unsuccessful. Identification of the patterns and nature of symptoms in the prodromal period prior to the diagnosis may help to identify patients earlier thus helping to reduce DUP.
The objectives are to determine, in patients with first episode psychosis (FEP): (i) prevalence and duration of prior consultation in primary care for symptoms related to FEP, (ii) common patterns of symptoms presented prior to diagnosis, and (iii) extent of pre-existing chronic comorbidity. This is a case-control study comparing prior patterns and duration of symptoms and chronic comorbidity in approximately 12,000 patients with FEP to 48,000 controls.
We will use conditional logistic regression to compare prevalence of each symptom and comorbidity between cases and controls. In cases, we will determine the median (interquartile range) duration of DUP, the interval between first consultation for a potential psychotic symptom and date of prescription of antipsychotics lasting for more than a month. We also will use latent class analysis to determine common patterns of symptoms presented to primary care before diagnosis in cases.
The specific objectives are, in patients newly diagnosed with psychosis, to determine:
1. The prevalence and duration of prior consultation in primary care for symptoms related to FEP;
2. The extent of pre-existing chronic comorbidity;
3. Common patterns of symptoms and comorbidities presented prior to diagnosis.
Ying Chen - Chief Investigator - Keele University
Ying Chen - Corresponding Applicant - Keele University
Athula Sumathipala - Collaborator - Keele University
Carolyn Chew-Graham - Collaborator - Keele University
David Shiers - Collaborator - University of York
John Edwards - Collaborator - Keele University
Kelvin Jordan - Collaborator - Keele University
Saeed farooq - Collaborator - Keele University
Patient Level Index of Multiple Deprivation