Long-term Mortality Trends in Severe Mental Illnesses and COVID-19 Impact

Vocabulary flashcards defining epidemiological terms, research platforms, statistical models, and clinical concepts from the study on severe mental illness mortality trends.

Severe Mental Illness (SMI)

A classification of chronic mental health conditions that, in this study, encompasses schizophrenia-spectrum disorders (ICD-10 F20–F29) and bipolar affective disorders (ICD-10 F30.0 and F31.0).

South London & Maudsley Trust (SLaM Trust)

One of Europe's largest secondary mental healthcare providers, responsible for delivering secondary mental healthcare to a catchment population of 1.3 million people in southeast London.

Clinical Record Interactive Search (CRIS)

An ethically approved data platform established in 2007 that enables researchers to access de-identified electronic mental health care records from the South London & Maudsley Trust.

Standardized Mortality Ratio (SMR)

A statistical ratio comparing the observed number of deaths in a study sample to the expected number of deaths derived from a reference population's age and sex structure.

Index of Multiple Deprivation (IMD)

A UK measure of relative small-area deprivation evaluating seven domains: income, employment, education, health, crime, housing, and living environment.

Lexis Expansion

A statistical technique in survival analysis used to split person-time into discrete, time-varying intervals for variables such as age, duration since diagnosis, or calendar periods.

Sub-Hazard Ratio (sHR)

A relative risk metric obtained from Fine and Gray competing risks regression models, representing the rate of a specific event while taking competing risk events into account.

Diagnostic Overshadowing

The clinical phenomenon where physical symptoms presented by a patient with mental illness are misattributed to their psychiatric condition, delaying physical healthcare.

Syndemic

A framework describing the socially patterned clustering of two or more co-occurring disease conditions that interact and aggravate adverse health outcomes in marginalized populations.

Natural Language Processing (NLP)

Text-mining algorithms applied to unstructured narrative clinical notes in electronic health records to extract diagnoses and patient details not captured in structured fields.

Competing Risks Model

A survival analysis regression model used when individuals are subject to multiple potential outcomes, where the occurrence of one event precludes or alters the probability of another.