Risken för akut återinläggning kan förutsägas. Tidigare vårdkonsumtionsmönster och vissa diagnoser starkt predicerande
Sammanfattning
Data from 2009 show that 1 % of the population in the County of Stockholm account for approximately 33 % of the total health care spending. These patients are on average admitted to hospital 4 times per year, or more, and accumulate direct health care costs per person of about 39 000 annually. Importantly, there is a significant turnover in this group as on a yearly basis 80 % are replaced by new individuals. Within Stockholm County Council, a number of projects are being developed where additional resources are directed towards this highly vulnerable patient group. Recognizing the rapid turnover of patients within this group, we developed a logistic regression prediction model based on routinely collected data from hospital health care data systems with the aim of enabling rapid and reliable identification of risk patients at an early stage. In addition to the main analyses, we also conducted analyses in subgroups of patients with heart failure, chronic obstructive pulmonary disease and those aged 65 or above. A total of 139 007 patients with more than 280 000 hospitalizations were identified and included in the analyses. As in previous studies, the strongest risk factors for readmission were the number of previous hospitalizations and certain chronic diagnoses (e.g. chronic liver failure, congestive heart failure and renal failure). Results from subgroup analyses were similar as in the overall analyses. Despite limited access to detailed clinical data, prediction results were satisfactory.