RESUMO
BACKGROUND: Prevention of hospitalisation is an important aspect of type 2 diabetes (T2D) management. AIMS: We retrospectively determined the utility of the Hospital Admission Risk Programme (HARP) diabetes risk calculator (HARP tool) in identifying patients with T2D more likely to have unplanned hospital presentations. METHODS: The HARP tool includes a clinical assessment score (Part A) and a psychosocial and self-management impact score (Part B), and categorises patients into low, medium, high or urgent risk of acute hospitalisation. It was completed for T2D patients attending Royal North Shore Hospital, Sydney, in 2013. RESULTS: Within the cohort of 278 patients (age 65.3 ± 10.5 years; 62.9% male; diabetes duration 10.7 ± 6.6 years), 67.3% were classified as low risk, 32.7% as medium risk and none as high or urgent risk. Following adjustment for confounders, a medium HARP score was associated with a 3.1-fold increased risk of unplanned hospital presentations in the subsequent 12 months (95% confidence interval: 1.35-7.31; P = 0.008). Part A scores were significantly higher for patients that presented to hospital compared to those that did not (14.2 ± 6.8 vs 11.4 ± 5.5; P = 0.034), whereas there was no difference in Part B scores (P = 0.860). CONCLUSIONS: In patients with low and medium HARP scores, clinical features were more predictive of hospital presentations than certain psychosocial or self-management factors in the present cohort. Further studies are required to characterise unplanned hospitalisation in patients with higher HARP scores, or whether additional psychosocial assessments could improve the tool's predictability.