Background The high off-label use of albumin persuaded the pharmacy to introduce a request form that uses the internationals guidelines to assess whether the use of albumin is appropriate. This has resulted in a clear reduction in the costs.

Purpose To monitor the wards using six sigma methodology (a statistical concept that measures a process in terms of defects); to ensure that all procedures have been followed correctly.

Materials and Methods 696 forms for albumin prescriptions coming from 26 wards (August to December 2011) were analysed using Minitab software, which cheques the frequency of the best correct requests (type 1), partially correct requests (type 2) and incorrect requests (type 3). For each ward the β coefficient was used to connect the relationship between the ward and their requests. The wards were grouped into 4 ranges on confidence intervals for the odds ratio (OR) of a width equal to 0.3 called A,B,C,D and then a final logistic regression analysis was made.

Results The analysis showed that group A was the most efficient in terms of probability of obtaining better results, followed by groups D (OR 0.36), C (OR 0.19) and B (OR 0.09). The total number of requests received was: 43% type 1 (299/696); 26% type 2 (181/696); 31% type 3 (216/696). Group B showed the worst result with 51% type 3 requests (58% of the total requests for type 3). Using six sigma we have achieved a cost saving of about €15,000.

Conclusions The results encourage us to apply this methodology to other fields.

No conflict of interest.