Alcohol and hospital readmission (1)

Editor – I read the article by Shalchi et al with great interest (Clin Med October 2009 pp 426–30). Having recently conducted an audit on hospital readmission in patients with alcoholic liver disease I am surprised that alcohol-related problems are not mentioned in the article.

I audited admissions of patients coded to have alcoholic liver disease in one Glasgow hospital during one year (September 2006 to August 2007). Of 124 patients admitted with the diagnosis, 22 died during the initial admission and 102 were discharged alive and followed up for one year. Seventy-six patients (75%) were readmitted at least once after discharge, about 50% within two months (Fig 1). The average number of readmissions was 3.2 per patient (virtually all emergency admissions). The average duration of the admissions and readmissions was 12 days, accounting for 3,887 days in hospital. Significantly, 28 of the patients discharged with alcoholic liver disease died within one year, most (24) during a readmission. Thus 40% of those studied died within the period analysed and the burden for the hospital was enormous. Even though the epidemiology of alcohol-related problems is worse in Scotland than the rest of the UK and Western Europe, I do not think that this problem is specific to Glasgow.¹

Fig 1.

Readmission-free survival of patients with alcoholic liver disease after being discharged from hospital.

I therefore believe that certainly alcoholic liver disease, and in my experience also other alcohol-related problems, are an important predictor for readmission and worth investigating further (they are lacking in Table 1 and Fig 4 of Salchi et al's article). Alcoholic patients present a significant proportion of admissions to wards with a variety of medical problems including many of the ones mentioned in the authors' Table 1. Alcoholic patients are stigmatised, difficult to deal with and difficult to treat. Therefore, if readmission in alcoholic patients was regarded as avoidable, most are probably in the categories ‘inadequate therapy’ and ‘poor discharge planning’ of Shalchi et al's Fig 4.