BACKGROUND: Computerized clinical decision support systems have been claimed to reduce prescription errors and improve patient care. They may play an important role in the care of hospitalized patients with diabetes. AIM: To collate evidence for the use of clinical decision support systems in improving the care of hospitalized patients with diabetes in a non-critical care setting and to assess their effectiveness. METHODS: We searched four databases from 1980 to 2010 without language restrictions. All types of studies other than case reports were included. Data extraction and quality assessment were carried out based on the Centre for Review and Dissemination guidance. A narrative synthesis was conducted. RESULTS: Fourteen studies met the inclusion criteria, including two cluster randomized controlled trials, eight before-and-after studies and four other descriptive studies. Generally, the quality of the studies was not very high. Nine out of 10 studies reported reduction in mean blood glucose or similar measures (patient-day-weighted mean blood glucose) during inpatient stay. The reduction using computerized physician order entry system in patient-day-weighted mean blood glucose ranged from 0.6 to 0.8 mmol/l (10.8-15.6 mg/dl). Other beneficial effects during inpatient stay included reduced use of sliding scale insulin and greater use of basal-bolus insulin regimen. Only one study found a significant increase in hypoglycaemic events. CONCLUSIONS: Clinical decision support systems have been used, often as part of a complex programme, to improve the care of hospitalized patients with diabetes. There is some evidence that they may have a beneficial effect, but this needs further confirmation.
BACKGROUND: Computerized clinical decision support systems have been claimed to reduce prescription errors and improve patient care. They may play an important role in the care of hospitalized patients with diabetes. AIM: To collate evidence for the use of clinical decision support systems in improving the care of hospitalized patients with diabetes in a non-critical care setting and to assess their effectiveness. METHODS: We searched four databases from 1980 to 2010 without language restrictions. All types of studies other than case reports were included. Data extraction and quality assessment were carried out based on the Centre for Review and Dissemination guidance. A narrative synthesis was conducted. RESULTS: Fourteen studies met the inclusion criteria, including two cluster randomized controlled trials, eight before-and-after studies and four other descriptive studies. Generally, the quality of the studies was not very high. Nine out of 10 studies reported reduction in mean blood glucose or similar measures (patient-day-weighted mean blood glucose) during inpatient stay. The reduction using computerized physician order entry system in patient-day-weighted mean blood glucose ranged from 0.6 to 0.8 mmol/l (10.8-15.6 mg/dl). Other beneficial effects during inpatient stay included reduced use of sliding scale insulin and greater use of basal-bolus insulin regimen. Only one study found a significant increase in hypoglycaemic events. CONCLUSIONS: Clinical decision support systems have been used, often as part of a complex programme, to improve the care of hospitalized patients with diabetes. There is some evidence that they may have a beneficial effect, but this needs further confirmation.
Authors: Katharina Maria Neubauer; Lukas Schaupp; Johannes Plank; Thomas Augustin; Selma Isabella Mautner; Bernd Tschapeller; Thomas Rudolf Pieber Journal: J Diabetes Sci Technol Date: 2013-03-01
Authors: Katharina M Neubauer; Julia K Mader; Bernhard Höll; Felix Aberer; Klaus Donsa; Thomas Augustin; Lukas Schaupp; Stephan Spat; Peter Beck; Friedrich M Fruhwald; Christian Schnedl; Alexander R Rosenkranz; David B Lumenta; Lars-Peter Kamolz; Johannes Plank; Thomas R Pieber Journal: Diabetes Technol Ther Date: 2015-06-05 Impact factor: 6.118
Authors: Nouf Sahal Alharbi; Nada Alsubki; Simon Jones; Kamlesh Khunti; Neil Munro; Simon de Lusignan Journal: J Med Internet Res Date: 2016-11-25 Impact factor: 5.428