Study: Artificial Intelligence reduces the deaths of patients in hospitals

A new study states that an artificial intelligence system could reduce the risk of unexpected deaths by identifying patients in hospital with a significant risk due to health weakening. The study, published by the Canadian Medical Association Magazine, showed the potential of automatic learning to improve the results of patients in hospitals by evaluating the clinical effect, for a system known as Chartwatch, an early warning system based on machine learning, and is designed to predict the patient’s deterioration. The new system enables healthcare teams to respond faster, more efficiently and has been developed and implemented in the General Unit for Internal Medicine in the St. Michael General Hospital in Toronto, Canada. Early signs of clinical decline The system aims to discover early signs of clinical decline at hospital patients, allowing healthcare providers to act before the patient’s condition deteriorates significantly. The decline in hospitals can lead to serious complications, including unplanned entry into the intensive care unit, or even death, and the artificial intelligence system helps identify such a decline in reducing deaths early. Chartwatch uses an advanced automatic learning model, which analyzes data in the actual time of the electronic medical records of the hospital, with the aim of predicting the dangers of the patient’s deterioration, based on a variety of clinical data points. The model constantly integrates data from the Patient’s Electronic Medical Registry, including Vital Signs, Laboratory Results and Other Clinical Information, as it follows how the patient’s condition develops over time, then uses previously risk DEGRES and Changes in these Degrees to Determine Wher Patient improves, or deteriorates, and make predictions that are likely to be the patient at risk of victorious deterioration, such as the need to transfer him to the intensive care unit, or care altastiya, based on his clinical condition. Deaths in hospital wings The study included more than 13,000 patients between the ages of 55 and 80, and the results showed that deaths decreased significantly during the intervention period as it dropped from 2.1% to 1.6%, and among patients determined by the system, deaths dropped from 10.3% to 7.1%. These results indicate that early warning systems can play a decisive role in reducing death in hospital wings, despite the need for more research to improve these technologies. Other sub -specializations units in the hospital, which did not use the system, saw no significant changes in mortality rates. The lead author, dr. Amoul Verma, a clinical research scientist in St. Michael Hospital believes that it is important to carefully use the use of artificial intelligence instruments in medicine to ensure that it is safe and effective, especially in light of its increasing use in the health sector. Artificial intelligence and early warning systems say: “Our results indicate that early warning systems in the early base are promising to reduce unexpected deaths in hospitals.” He explains that the regular communication between the system, the doctors and the nursing staff helped reduce the deaths through real time warnings, e -mails twice a day to nursing teams and daily email messages to the palliative care team. The team has also established a care path for patients who are at risk, while increasing the monitoring by nurses, increasing communication between nurses, doctors and demands to encourage doctors to reconsider patients. The researchers improve the model to promote its predictive accuracy and reduce “false positive”. Dr. Verma says more research is needed to understand the long -term effect to use these systems on the total results in the areas of clinical health care, including death rates, and the patient’s return to hospital within 30 days of exit.