Consumer segmentation and time interval between types of hospital admission: a clinical linkage database study

Abstract

Background: Healthcare policies target unplanned hospital admissions and 30-day re-admission as key measures of efficiency, but do not focus on factors that influence trajectories of different types of admissions in the same patient over time. Objectives: To investigate the influence of consumer segmentation and patient factors on the time intervals between different types of hospital admission. Research design, subjects and measures: A cohort design was applied to an anonymised linkage database for adults aged 40 years and over (N = 58 857). Measures included Mosaic segmentation, multimorbidity defined on six chronic condition registers and hospital admissions over a 27-month time period. Results: The shortest mean time intervals between two consecutive planned admissions were: 90 years and over (160 days (95% confidence interval (CI): 146-175)), Mosaic groups 'Twilight subsistence' (171 days (164-179)) or 'Welfare borderline' and 'Municipal dependency' (177 days (172-182)) compared to the reference Mosaic groups (186 days (180-193)), and multimorbidity count of four or more (137 days (130-145)). Mosaic group 'Twilight subsistence' (rate ratio (RR) 1.22 (95% CI: 1.08-1.36)) or 'Welfare borderline' and 'Municipal dependency' RR 1.20 (1.10-1.31) were significantly associated with higher rate to an unplanned admission following a planned event. However, associations between patient factors and unplanned admissions were diminished by adjustment for planned admissions. Conclusion: Specific consumer segmentation and patient factors were associated with shorter time intervals between different types of admissions. The findings support innovation in public health approaches to prevent by a focus on long-term trajectories of hospital admissions, which include planned activity.