MMI 406 Decision Support Systems and Health Care
Key Artifacts:
Syllabus
Diabetes Decision Tree
Group Project: Patient Adaptive Diabetic Clinical Decision System
Syllabus
Diabetes Decision Tree
Group Project: Patient Adaptive Diabetic Clinical Decision System
Reflection
Learning Goals:
· Understand the basic features, benefits and limitations of machine learning and intelligent decision support methods in the healthcare environment.
· Understand the benefits and limitations of medical decision making techniques
· Understand the role of performance measurement in guiding DSS deployment
· Describe current uses of medical decision making and decision support systems in healthcare.
· Understand the use of various decision making and analytic models to solve both structured and unstructured problems
In 1999, To Err is Human, a report by the Institute of Medicine, was published and brought to light up to 98,000 deaths per year resulting from preventable medical errors. As a result, the focus on prevention of medical errors became critical. Among the proposed changes were implementation of Computerized Physician Order Entry (CPOE) and Clinical Decision Support System (CDSS) to improve patient safety.
As a clinical person, I always scoff at the idea of having a computer decide what is best for the patient. For me, no amount of computer generated decision support could replace the subjective and objective information presented by the patients’ right in front of the healthcare provider. This class helped enlighten me on some of the points I was resistant of. To help us learn more about decision support systems, each student leased TreeAge (pronounced “triage”), a decision analysis software. In a nutshell, a clinical decision support system helps identify what the problem is, what treatment options are available and what are the possible outcomes for each treatment options. A decision support system helps clinicians reach a decision in a systematic and thorough manner. Expensive diagnostic tests and therapeutic interventions are used wisely as a result. However, a clinician’s experience and the individual patient’s condition should not be overlooked or ignored.
For our group project, I was lucky enough to be in the group with David Mishler and David Madison. Both of them had extensive technical experience. We choose to work on a decision support system for Diabetes, taking into consideration all the complications associated with it. We used TreeAge to create an episodic and chronic decision trees. We wanted to demonstrate how a decision tool could improve outcomes for complex diseases such as Diabetes. We wanted our tool to have knowledge based information, reasoning mechanism which has the rules or formula and a when combined with the patient data, would come up with the appropriate decision and then communicate this to the clinician by way of a recommendation or alert system. We were however cognizant of the fact that decision making techniques using the decision support system has its limitations. First, the software standardization is still in its infancy. They are unregulated and competitive. Second, the decision support system should not be used in place of a sound clinical judgment. Instead, it should be viewed as a guide in making healthcare decisions.
We also realized how performance measurement affects deployment of DSS. Knowing that 76% or diabetic retinopathy and 56% of diabetic nephropathy could be reduced if patients maintain a close to normal range of blood glucose level, emphasis was placed on making sure abnormal results get relayed to a clinician so a response could be initiated. Because there is significant impact in keeping blood sugars within normal limits, decision support tools targeting Diabetes is very important. This not only standardizes the approach to Diabetes, it also improves the quality of care provided, reduces the risks of errors and reduces utilization of services.
Today, medical decision making and decision support systems in healthcare is necessary to decrease healthcare spending and improve the quality of care. In these tough economic times, providers have to think of ways to eliminate unnecessary testing, reduce inpatient stays and readmission and cut down healthcare costs. By using a decision support system, clinicians could easily rule out a particular condition. Performance measurements are easily obtained when the clinical decision support is in place. This is particularly helpful in a pay-for-performance environment where performance is closely tied to financial incentives.
Clinical Decision Support Systems have been used to assist in various situations. Unstructured problems could be broken down into smaller problems to make it more manageable. As in the case of a patient coming in with Diabetic Ketoacidosis (DKA) with Congestive Heart Failure (CHF) and End Stage Renal Disease(ESRD) as the comorbities, the clinician has the responsibility to take care of the problem at hand (the DKA) and work on the CHF and ESRD once the patient is medically stable or wait until the patient is in an outpatient setting.
At the end of the term, my aversion towards Clinical Decision Support System has dissipated. It is a necessary tool to standardize diagnosis protocol, promote knowledge sharing, provide decision support, and improve the quality of care. CDSS also helps reduce unnecessary tests and control the costs of healthcare. Most importantly, it helps reduce medical errors which no patients should have to be worried about when seeking care from healthcare providers.
· Understand the basic features, benefits and limitations of machine learning and intelligent decision support methods in the healthcare environment.
· Understand the benefits and limitations of medical decision making techniques
· Understand the role of performance measurement in guiding DSS deployment
· Describe current uses of medical decision making and decision support systems in healthcare.
· Understand the use of various decision making and analytic models to solve both structured and unstructured problems
In 1999, To Err is Human, a report by the Institute of Medicine, was published and brought to light up to 98,000 deaths per year resulting from preventable medical errors. As a result, the focus on prevention of medical errors became critical. Among the proposed changes were implementation of Computerized Physician Order Entry (CPOE) and Clinical Decision Support System (CDSS) to improve patient safety.
As a clinical person, I always scoff at the idea of having a computer decide what is best for the patient. For me, no amount of computer generated decision support could replace the subjective and objective information presented by the patients’ right in front of the healthcare provider. This class helped enlighten me on some of the points I was resistant of. To help us learn more about decision support systems, each student leased TreeAge (pronounced “triage”), a decision analysis software. In a nutshell, a clinical decision support system helps identify what the problem is, what treatment options are available and what are the possible outcomes for each treatment options. A decision support system helps clinicians reach a decision in a systematic and thorough manner. Expensive diagnostic tests and therapeutic interventions are used wisely as a result. However, a clinician’s experience and the individual patient’s condition should not be overlooked or ignored.
For our group project, I was lucky enough to be in the group with David Mishler and David Madison. Both of them had extensive technical experience. We choose to work on a decision support system for Diabetes, taking into consideration all the complications associated with it. We used TreeAge to create an episodic and chronic decision trees. We wanted to demonstrate how a decision tool could improve outcomes for complex diseases such as Diabetes. We wanted our tool to have knowledge based information, reasoning mechanism which has the rules or formula and a when combined with the patient data, would come up with the appropriate decision and then communicate this to the clinician by way of a recommendation or alert system. We were however cognizant of the fact that decision making techniques using the decision support system has its limitations. First, the software standardization is still in its infancy. They are unregulated and competitive. Second, the decision support system should not be used in place of a sound clinical judgment. Instead, it should be viewed as a guide in making healthcare decisions.
We also realized how performance measurement affects deployment of DSS. Knowing that 76% or diabetic retinopathy and 56% of diabetic nephropathy could be reduced if patients maintain a close to normal range of blood glucose level, emphasis was placed on making sure abnormal results get relayed to a clinician so a response could be initiated. Because there is significant impact in keeping blood sugars within normal limits, decision support tools targeting Diabetes is very important. This not only standardizes the approach to Diabetes, it also improves the quality of care provided, reduces the risks of errors and reduces utilization of services.
Today, medical decision making and decision support systems in healthcare is necessary to decrease healthcare spending and improve the quality of care. In these tough economic times, providers have to think of ways to eliminate unnecessary testing, reduce inpatient stays and readmission and cut down healthcare costs. By using a decision support system, clinicians could easily rule out a particular condition. Performance measurements are easily obtained when the clinical decision support is in place. This is particularly helpful in a pay-for-performance environment where performance is closely tied to financial incentives.
Clinical Decision Support Systems have been used to assist in various situations. Unstructured problems could be broken down into smaller problems to make it more manageable. As in the case of a patient coming in with Diabetic Ketoacidosis (DKA) with Congestive Heart Failure (CHF) and End Stage Renal Disease(ESRD) as the comorbities, the clinician has the responsibility to take care of the problem at hand (the DKA) and work on the CHF and ESRD once the patient is medically stable or wait until the patient is in an outpatient setting.
At the end of the term, my aversion towards Clinical Decision Support System has dissipated. It is a necessary tool to standardize diagnosis protocol, promote knowledge sharing, provide decision support, and improve the quality of care. CDSS also helps reduce unnecessary tests and control the costs of healthcare. Most importantly, it helps reduce medical errors which no patients should have to be worried about when seeking care from healthcare providers.