Temporal sequencing in multimorbidity using population-scale linked data for 1.7 million individuals with 20-year follow-up

Abstract Multimorbidity is defined as the coexistence of two or more chronic health conditions in an individual. The objective of this study was to examine how diseases in a cluster of physical-mental health multimorbidity with a high all-cause mortality (psychosis, diabetes, and congestive heart failure) develop and coexist over time, and to assess the associated impact of different temporal sequences on mortality. Population-scale, individual-level, anonymised, linked, demographic, administrative and electronic health record data were modelled using multi-state models for 1,675,585 individuals over a 20-year period (2000–2019). Cox regression models were used to estimate baseline hazards for transitions between states, adjusted for gender, age, and area-level deprivation. Our findings suggest that the order of disease acquisition in physical-mental health multimorbidity had an important impact and complex relationship on patient mortality. Individuals developing diabetes, psychosis, and congestive heart failure, in that order, had an increased all-cause mortality rate compared to the development of the same conditions in a different order, resulting in the highest loss in expectation of life of 13 years at age 50 compared to the general population. Congestive heart failure as a single condition and in combination with psychosis had an equally high loss in expectation of life. Identification and therapeutic targets for psychosis and congestive heart failure may be beneficial within 5 years following an initial diagnosis of diabetes. The use of multi-state models offers a flexible framework to assess temporal sequences of diseases and associated patient outcomes, and allows identification of potential risk factors, screening opportunities, and therapeutic targets in multimorbidity.