ABSTRACT
BACKGROUND: We surveyed the quality of risk stratification politics and monitored the rate of entries to our company-wide protocol for venous thrombembolism (VTE) prophylaxis in order to identify safety concerns. PATIENTS AND METHODS: Audit in 464 medical and surgical patients to evaluate quality of VTE prophylaxis. RESULTS: Patients were classified as low 146 (31 %), medium 101 (22 %), and high risk cases 217 (47 %). Of these 262 (56.5 %) were treated according to their risk status and in accordance with our protocol, while 9 more patients were treated according to their risk status but off-protocol. Overtreatment was identified in 73 (15.7 %), undertreatment in 120 (25,9 %) of all patients. The rate of incorrect prophylaxis was significantly different between the risk categories, with more patients of the high-risk group receiving inadequate medical prophylaxis (data not shown; p = 0.038). Renal function was analyzed in 392 (84.5 %) patients. In those patients with known renal function 26 (6.6 %) received improper medical prophylaxis. If cases were added in whom prophylaxis was started without previous creatinine control, renal function was not correctly taken into account in 49 (10.6 %) of all patients. Moreover, deterioration of renal function was not excluded within one week in 78 patients (16.8 %) and blood count was not re-checked in 45 (9.7 %) of all patients after one week. There were more overtreatments in surgical (n = 53/278) and more undertreatments in medical patients (n = 54/186) (p = 0.04). Surgeons neglected renal function and blood controls significantly more often than medical doctors (p-values for both < 0.05). CONCLUSIONS: We found a low adherence with our protocol and substantial over- and undertreatment in VTE prophylaxis. Besides, we identified disregarding of renal function and safety laboratory examinations as additional safety concerns. To identify safety problems associated with medical VTE prophylaxis and "hot spots" quality management-audits proved to be valuable instruments.
