STHF compared 200 discharge summaries from before the introduction of speech recognition with 200 discharge summaries created when using Nuance SpeechMagic. Per Urdahl, head of the medical department said: ‘We used an eight point scale to quantify document quality. It turned out that, in terms of content and layout, the new SR documents were rated as being as good as or better than the transcribed documents. The quality of discharge summaries improved on average by 9% across the seven departments that participated in the analysis. The greatest improvements have been achieved in neurology, pulmonology and cardiology (see chart).’
Physicians can access the speech system to feed information into their EMR, with more than 450 workstations across all departments except psychiatry powered by speech recognition. The hospital is waiting for the psychiatric vocabulary to become available in the Norwegian language to close the last gap in the speech infrastructure. ‘We expect this to happen within a year. If so, we will extend the existing installation to the psychiatric departments immediately,’ said Per Urdahl.
Immediate discharge reports
Due to speech-based data capturing, document turnaround times have decreased significantly. ‘In every single month after the introduction of SR we have managed to deliver 90% of all medical reports to the referring physicians within seven days,’ he added. This is well above the national target of 80% set by the Norwegian ministry of health. At the hospital about 70% of the documents are handed directly to patients before they are discharged. At the same time, documents are electronically transferred to their referring physicians.
‘We generate savings of an equivalent of 900,000 euros per year,’ said Per Urdahl. Half the money saved goes to the respective department on top of the annual budget. ‘Speech recognition is the only IT solution I’ve introduced in 20 years that has led to measurable cost savings,’ he added.
Norway, added Marcel Wassink, Vice President of Nuance Healthcare in EMEA, is a good example of how healthcare will go in the future. ‘Information input and reporting using speech recognition has become standard there, and is even used in medical universities. I’m convinced that, in future, every hospital will have a speech recognition infrastructure at the scale of STHF.’