Early detection of clinical deterioration in patients holds promise for reducing mortality and improving outcomes. However, it remains a challenge in both hospital and outpatient settings.
Stanford Health Care has addressed this challenge by incorporating validated AI and machine learning models into clinical decision support systems. They have also integrated artificial intelligence into clinical workflows and improved the patient experience, including reducing wait times, improving quality of care and enabling critical conversations.
Dr. Shreia Shah is a practicing academic internist, board-certified clinical informatics practitioner, and AI healthcare integration expert at Stanford Health Care.
She will talk about the healthcare system’s efforts in the field of artificial intelligence 2023 HIMSS AI in Healthcare Forum, scheduled for 14-15 December in San Diego, offers a case study titled “How Stanford Health Transformed Patient Care by Combining Compassion with AI-Driven Innovation.”
We spoke with Shah to get a sneak peek of her session and a deeper understanding of how Stanford Health Care is using AI and ML.
P. Why does detecting clinical deterioration in patients remain a challenge?
A. Patients in hospitals are increasingly complex and serious diseases, while less acute care is moving to home, ambulatory care or subacute level of treatment. Within an academic medical center, this is even more profound in patients at high risk of clinical deterioration.
Early signs can be subtle and vary widely among patients. Identifying which patients require the most attention is a needle in a haystack activity. Furthermore, these patients are cared for by multi-person care teams and require the assessment of large amounts of data that change over time.
Teams can experience communication gaps, information overload, and cognitive biases that lead to unexpected clinical deterioration with major consequences such as emergency resuscitation efforts and unplanned ICU transfers. There may also be varying degrees of alignment among team members regarding risk perception.
Standardized care coordination workflows that empower all members of the care team to make decisions about patient care can help overcome these challenges.
P. How did you decide that AI and ML were the way to help with this challenge?
A. We needed to identify patients at increased risk and align the care team around a common, standardized clinical response. We found that the ML model could identify patients with a high probability of future clinical deterioration without additional tasks for our working clinicians.
Predictions would need to be made early enough to allow the care team enough time to respond. Accuracy is always a concern, and clinicians often trust that an AI system won’t tell them something they don’t already know.
In our implementations, the emphasis was not on whether the model predictions were correct. Instead, for any given model-labeled patient, physician and nonphysician team members had to implement a structured collaborative workflow to assess risk and response. So the probabilistic model creates a team trigger.
Our implementation efforts were focused on these priority areas: 1) Designing a system that would integrate the ML model into a complex healthcare system, 2) Building effective teams and processes to enable the collaborative workflow required for successful implementation, and 3) Applying these AI – integrated systems in a way that is both sustainable and scalable for healthcare enterprises.
The focus was on creating a holistic system that not only incorporates advanced technology but also aligns with clinical, operational and strategic needs.
P. What is one example of how incorporating validated AI and ML models into clinical decision support systems has helped Stanford address clinical deterioration?
A. Our clinical deterioration model was validated on our data to ensure model performance; then, signals are integrated into our EHR with full transparency, including contributing factors and supplemented by a mobile alert to the care team.
The ML model is able to update predictions about hospitalized patients every 15 minutes and has been used to act as an objective risk estimator and has helped facilitate alignment and coordination in patient care as System integrated with artificial intelligence.
The model underwent site-specific validation to ensure its effectiveness in predicting clinical worsening events such as unplanned ICU transfers within 6 to 18 hours. This workflow led to a significant increase in multidisciplinary standardized patient assessments and a resulting 20% reduction in clinical worsening events.
The results of the qualitative evaluation identified that the model was useful in aligning mental models and initiating the necessary workflows for patients flagged by the model with consensus among members of the multidisciplinary team. Using a reliable and continuously updated risk signal, we aligned physicians with the rest of the care team to implement a consistent workflow.
Follow Bill’s HIT report on LinkedIn: Bill Siwicki
Email him: bsivicki@himss.org
Healthcare IT News is a HIMSS media publication.
#Stanford #Health #artificial #intelligence #reduce #clinical #deterioration #events
Image Source : www.healthcareitnews.com