Abstract
BACKGROUND: Lipohypertrophy (LH) is a major complication of subcutaneous insulin treatment brought about by multiple overlapping injections and/or needle reuse. It is responsible for unacceptable glucose oscillations due to a high rate of hypoglycaemic episodes and rebound glucose spikes. Skin ultrasound scans (USS), the gold standard for its detection, is too expensive for screening purposes. AIMS: To define a structured method allowing health professionals (HPs) to identify LH lesions as inexpensively and correctly as possible. METHODS: Out of 129 insulin-treated people with diabetes identified by USS as having LH lesions, only 40 agreed to participate in the study (24 females, age 54 ± 15 years, daily insulin dosage 57 ± 12 IU). Each was blindly examined by four well trained and four non-trained HPs according to a standard method involving repeated well codified maneuvers. RESULTS: A specific training allowed inexperienced HPs to acquire high diagnostic accuracy in identifying LH lesions independent of site, size, shape, and even BMI. This kind of training also allowed to reach a 97 % consistency rate among HPs as compared to USS, while the lack of training was associated with a wide variability and inconsistency of identification results. CONCLUSIONS: Diabetes teams should follow systematically the simple procedure reported in this paper for the diagnosis of LH and try to get it further implemented and progressively refined in large scale studies. This would have a major impact on patient education in terms of (1) correct injection technique and (2) ability to identify lesions early enough to prevent poor metabolic outcome.