Homepage > Research abstracts > Prospective Implementation and evaluation of a computerized decision support system for management of the care of patients at Hartzfeld Geriatric Hospital’s Long Term Care ward: Evaluation of the system’s effect on the compliance of the care providers to the clinical protocol, and assessment of the option for changing the policy for managing patients through an empowerment of the clinical team
Prospective Implementation and evaluation of a computerized decision support system for management of the care of patients at Hartzfeld Geriatric Hospital’s Long Term Care ward: Evaluation of the system’s effect on the compliance of the care providers to the clinical protocol, and assessment of the option for changing the policy for managing patients through an empowerment of the clinical team
Researchers: Yuval Shahar1, Ayelet Goldstein2
- Ben Gurion University of the Negev
- Hadassah Academic College, Jerusalem
Background: Medical errors lead to morbidity and mortality. Automated support for application of evidence-based Clinical Guidelines (GLs) can enhance GL adherence, improve patient outcomes, and reduce costs.
Objectives: Supporting real-time automated GL-based care requires creating a formalized, machine-comprehensible GL representation, and an episodic decision-support system for provision of intermittent treatment advice, accommodating the non-continuous nature of care delivery, including partial actions or partially met objectives.
Method: We developed an episodic algorithm and implemented a system (e-Picard) that performs retrospective quality assessment using fuzzy logic and recommends which actions are [still] relevant based on the patient's record.
Findings: Initial evaluation scores of the e-Picard system were promising, with a mean 94% correctness and 90% completeness based on 50 random pressure ulcer or diabetes patients. Errors were mainly due to knowledge specification, algorithmic issues, and missing data. Post-corrections, scores improved to 100% correctness and a mean 97% completeness, with missing data still affecting completeness.
These results validate the system's capability to assess guideline adherence and provide quality recommendations. A retrospective evaluation involving 1000 patients demonstrated that with the potential use of e-Picard, the mean compliance score increased significantly across different time frames: from 68% to 88% in a weekly review, 68% to 91% over a 3-day period, and 66% to 97% with daily assessments. Thus, using the system might lead to a 21% to 31% potential improvement in compliance.
These results validate the system's capability to assess guideline adherence and provide quality recommendations. A retrospective evaluation involving 1000 patients demonstrated that with the potential use of e-Picard, the mean compliance score increased significantly across different time frames: from 68% to 88% in a weekly review, 68% to 91% over a 3-day period, and 66% to 97% with daily assessments. Thus, using the system might lead to a 21% to 31% potential improvement in compliance.
Conclusions: We have demonstrated the feasibility of e-Picard to provide realistic medical decision-making support for noncontinuous, intermittent consultations
Recommendations: The operation of an episodic decision support system, which will run in real time and support on demand the decision-making of the nursing and medical team, can potentially significantly improve the team's adherence to guidelines, reduce medical errors, lower treatment costs, reduce deterioration rate, and potentially reduce morbidity and mortality.
Research number: R/284/2020
Research end date: 09/2024