Drug utilization evaluation of parenteral proton pump inhibitors

Current evidence supports parenteral infusion of proton pump inhibitors (PPI) after endoscopic treatment of bleeding peptic ulcers and such treatment seems reasonable where there is active bleeding or visible vessel on endoscopy. Parenteral boluses of PPI can be used in patients nil by mouth who cannot tolerate oral therapy. We sought to examine the appropriateness of parenteral PPI use. Drug utilisation evaluation was performed on 94 patients admitted to a 500 bed metropolitan hospital. 39 patients received continuous parenteral infusion of omeprazole (8 mg/ h) over a mean of 60 ± 29 h. 55 patients had parenteral boluses (40 mg bd) of omeprazole over a mean of 5 ± 4 days. Indications for PPI infusion (n = 39) were: major haemorrhage requiring transfusion (23), minor haemorrhage (8), dyspepsia (4) and others (4). 31 of the 39 patients on PPI infusion had upper gastrointestinal (GI) endoscopy. PPI infusion was commenced prior to endoscopy in 26 (84%) patients. 13 patients (33%) had active bleeding or visible non bleeding vessels at endoscopy. Only 11 patients (28%) had endoscopically treated peptic ulcers. Indications for parenteral PPI boluses (n = 55) included patients nil by mouth unable to take maintenance PPI orally (21), minor haemorrhage (8), peptic ulcer prophylaxis in seriously unwell (6), major haemorrhage (4), dyspepsia (2), postoprative period following peptic ulcer surgery (2) and others (12). Endoscopy was performed in 10 patients (18%) with only 1 endoscopically treated peptic ulcer. Our data suggest significant inappropriate use of parenteral PPI, which may be used for minor GI bleeding and dyspepsia and are typically commenced prior to endoscopy. These findings can explain the costly hospital expenditure on PPI.

Using measurement data in a TSPSM project

One of the challenges for software engineering is collecting meaningful data from industrial projects. Software process improvement depends on measurement to provide baseline status and confirming evidence of the effect of process changes. Without data, any conclusions rely on intuition and guessing. The Team Software ProcessSM (TSPSM) provides a powerful framework for data collection and analysis, in addition to its primary goal as a basis for highly effective software development. In this paper, we describe the experiences of, and benefits realized by, a team using the TSP for the first time. By reviewing how this particular team collected and used data, we show features of the TSP that make it a powerful foundation for software process improvement.

A formal object-oriented approach to defining consistency constraints for UML models

We discuss how integrity consistency constraints between different UML models can be precisely defined at a language level. In doing so, we introduce a formal object-oriented metamodeling approach. In the approach, integrity consistency constraints between UML models are defined in terms of invariants of the UML model elements used to define the models at the language-level. Adopting a formal approach, constraints are formally defined using Object-Z. We demonstrate how integrity consistency constraints for UML models can be precisely defined at the language-level and once completed, the formal description of the consistency constraints will be a precise reference of checking consistency of UML models as well as for tool development.

A fuzzy approach to texture segmentation

The texture segmentation techniques are diversified by the existence of several approaches. In this paper, we propose fuzzy features for the segmentation of texture image. For this purpose, a membership function is constructed to represent the effect of the neighboring pixels on the current pixel in a window. Using these membership function values, we find a feature by weighted average method for the current pixel. This is repeated for all pixels in the window treating each time one pixel as the current pixel. Using these fuzzy based features, we derive three descriptors such as maximum, entropy, and energy for each window. To segment the texture image, the modified mountain clustering that is unsupervised and fuzzy c-means clustering have been used. The performance of the proposed features is compared with that of fractal features.