Measurement and Data Mining
The purpose of this assignment is to develop and execute a measurement plan and data mining analysis for a quality improvement problem. This assignment addresses course objective 3, where students should be able to:
Demonstrate the conceptual ability and technical skills to develop and execute an evaluation plan involving data extraction from practice information systems and databases.
In previous courses and experiences, youve likely used the PICOT format to create measurable clinical questions to improve quality through evidence based practice. This assignment builds on that knowledge and experience to use information technologies to manage knowledge and information for scholarly advanced nursing practice. The activities and thinking you will be doing as a part of this assignment are very similar to how you might approach portions of your DNP scholarly project, as well as your future work in practice and in positions of leadership to evaluate and implement EBP guidelines, but also to assess practice for quality improvement and generating PBE.
In Practice-Based Evidence, real world practice must be documented and measured to better inform evidence and improve quality. One of the key challenges relates to measurement of concepts and outcomes of practice.
Identify a problem: Identify a clinical or health systems problem (either related to your focus area for your blog or to your practice or just something that youre interested in) related to one of the six dimensions of healthcare quality (safe, timely, effective, efficient, equitable, and patient-centered). Examples could include: effective pain management in ICU patients, proactive access to palliative care, patient-centered mental health integration in primary care, prevention of patient falls, efficient patient education during transitions in care, etc.
Define the problem: Try to choose a problem that can be summarized in a few words or a short phrase. Conduct a brief literature search for the problem youve chosen and based on what you find create a brief (2-3 sentences) definition for the problem. For example, how would the literature briefly define effective pain management in ICU patients (vs. what is ineffective pain management in that population). What are the outcomes of interest for your problem? In order to measure improvement for a problem, you have to be able to clearly define the problem and the critical outcomes that you are interested in.
Select measures for your problem: If your ultimate goal is to improve your identified problem, you need to select appropriate measures based on your problem definition and critical markers or outcomes. In other words, think about how you would measure whether an intervention or changes that you might implement to a healthcare system actually improved the problem. Review existing databases of measures (as well as any other sources you can identify) such as http://www.qualitymeasures.ahrq.gov/ or http://www.hospitalcompare.hhs.gov. Search for any measures you can find for your problem as defined. For example, are there any measures of effectiveness of care? Are any of them specific to effectiveness of pain management? Or effectiveness of care in the ICU? Discuss your searching process and any challenges you experienced in identifying or evaluating measures. Were there measures for your problem? Were there multiple measures? How would you select one to use to evaluate changes in outcomes for your problem? If you cant find any existing measures for your problem, how would you go about identifying a measure or critical outcome to evaluate your problem? Can you directly measure outcomes related to your problem? Or, do you need to identify proxy measures? What would those be?
Data collection: Briefly discuss how informatics relates to measurement for quality improvement. How would you gather data for your identified measures? What data sources or tools would you use to collect, store, and later analyze data related to your measures? What challenges might you anticipate with ensuring high quality data for your measures?
Data mining analysis: Finally, assume that your organization can collect the data needed for your measures. How would you use data mining to explore patterns in the data? For example, how would you evaluate whether there are differences in effectiveness of pain management in the ICU for patients who can communicate vs. those that cannot communicate? Or for patients with a specific diagnosis? Describe specifically what question you might ask and how you would use data structures within health IT systems in a data mining approach (e.g., you might integrate some discussion of ICD codes or specific data fields from an EHR that could be used to create a report to understand patterns or differences in outcomes or measures for your problem).
Upload a brief report (~3 pages using APA format for citations, you can use tables or figures if they help to more clearly and/or concisely convey your points) addressing the points above and post to the assignment.
