Prevalence and patterns of hospital use for people with frequent alcohol-related hospital admissions, compared to non-alcohol and non-frequent admissions: a cohort study using routine administrative hospital data.

AIMS: This study compared prevalence and hospital use among individuals frequently admitted to hospital in England with wholly attributable alcohol-related diagnoses (WAAD), known as alcohol-related frequent attenders (ARFAs), with those of non-alcohol frequent attenders (NAFAs), non-frequent alcohol attenders (ARNFAs) and non-alcohol non-frequent attenders (NANFAs). DESIGN: Cross-sectional and longitudinal analyses of 5 years of England's Hospital Episode Statistics (HES). SETTING: Hospital inpatients in England, UK, 2011-16. PARTICIPANTS: Two cohorts (2011/12 = 489 580/7 654 944 patients and 2015/16 = 490 384/7 660 108 patients) were selected from all adult patients aged ≥ 18 years, treated in English hospitals between 1 April 2011 and 31 March 2016. Patients were categorized as having alcohol-related admissions if diagnoses included a WAAD (ICD-10 classification, WHO, 2016) and frequent admissions if they had more than three hospital admissions during a single HES year. MEASUREMENTS: Prevalence of ARFA, number of admissions (spells), occupied bed-days (OBDs), average length of stay (ALOS) and total admission costs over 5 years were compared among ARFAs, ARNFAs, NAFAs and NANFAs. FINDINGS: On average, 0.7% of people admitted to hospital per annum in England 2011-15 were ARFAs and more than a quarter of all frequent attenders (for all causes) to hospitals had a wholly attributable alcohol diagnosis on admission. ARFAs had longer ALOS than the other patient groups [5.55 days versus ARNFA 4.7, NAFA 3.39 and NANFA 2.57 days, F = 1088.37 (3, 488 570, P < 0.001)] in the 2015/16 index year; but fewer spells than NAFAs [5.38 ARFAs versus 5.98 NAFAs, F = 20 536.25 (3, 490 380) P < 0.001]. The ARFA cohort reduced in size (from 51 934 ARFAs to 20 548) in the course of 5 years. ARFAs had the highest average total cost of admissions per person over 5 years at £38 189. CONCLUSIONS: People with repeated admissions for alcohol-related problems in England appear to be a high-cost, high-need, complex group of patients that makes up more than a quarter of the country's alcohol admissions