Primary care activities associated with reduced emergency hospital admissions in the last three months of life: a national population-based, retrospective cohort study

Abstract

Objective
To examine the association between primary care activities 12 to 4 months before death and emergency hospital admissions in the last three months of life for adults with advanced illness.

Methods
An English population-based, retrospective cohort study using the Clinical Practice Research Datalink (CPRD Aurum) primary care data, linked with Hospital Episode Statistics (HES) records. Cohort: Adults (≥ 18 years) who died with a life-limiting condition. Independent variables: continuity of primary care, recognition of palliative care needs, referral to specialist palliative care, referral to a district/community nurse, and number, mode and timing of primary care contacts. Dependent variable: ≥3 emergency hospital admissions in the last three months of life. Covariables: individual-level sociodemographic and illness-related characteristics and area and GP-practice-level factors. Statistical analysis: Descriptive statistics and multilevel multivariable logistic regression.

Results
The cohort included 1,290,732 people who died between 2010 and 2023. 68% of people who died were 75 or older, and 50% were women. The most common diagnoses recorded in the primary care records of the people in the last five years of life were cancer (40%), followed by heart disease (33%) and chronic renal disease (30%). We will report the model findings, particularly associations between each primary care activity and the likelihood of experiencing three or more emergency hospital admissions in the last three months of life for the whole cohort and diagnostic groups, in addition to further descriptives and trends for the primary care activities and emergency hospital admissions.

Conclusion
The study findings will be used for generating hypotheses for better understanding the associations between primary care and emergency hospital admissions towards the end of life. The findings will add to the evidence base for improving community-based care for people with life-limiting conditions.