DNP 805 Select a specific clinical problem and post a clinical question that could potentially be answered using data mining
DNP 805 Select a specific clinical problem and post a clinical question that could potentially be answered using data mining
According to Alexander et al. (2019), data mining refers to analyzing large sets of data to identify valuable and understandable patterns. Such patterns can aid in forecasting trends and help with improving product safety and usability, and patient experience, and have proven to be effective in medicine and the healthcare industry. Electronic health records have tremendously improved data collection and has contributed to data mining to prevent and reduce medical errors. A question that may be answered through data mining is as follows: Does telemedicine help reduce the number of hospital readmissions for patients with congestive heart failure (CHF)? According to Reddy and Borlaug (2019), CHF is a common cause of hospitalization that accounts for almost $30 billion of expenditure in the United States. Over five million individuals are affected by CHF and studies show that there has been an increase in readmission rates for those who were hospitalized related to the disease (Garcia, 2017). Data mining techniques that may be used include tracking patterns, association, and prediction. A technique that I would not consider using is clustering analysis.
Alexander, S., Frith, K., & Hoy, H. (2019). Applied clinical informatics for nurses (2nd ed.). Jones & Bartlett Learning.
Reddy, Y. N. V., & Borlaug, B. A. (2019). Readmissions in heart failure: It’s more than just the medicine. Mayo
Clinic Proceedings, 94(10),
The work to reduce hospital readmissions is going to be ongoing for a long time because of the complexity of CHF, this is an issue with all hospitals. To reduce the number of preventable readmissions, the Centers for Medicare & Medicaid Services (CMS) initiated the Hospital Readmissions Reduction Program (HRRP) in 2012. Also, they realized that only 30% of all the patients with CHF had a scheduled follow-up appointment with the PCP or cardiologist on discharge and that of all those who were discharged, only 37% kept their follow-up appointments and, about 41% were lost to follow-up visits. The hospitals started an intervention program to reduce the readmission rate by making sure that all the CHF patients had follow up appointments and they started a weekly or biweekly phone call to the patients, telemonitoring, and home visits. At the end, about 60% of the CHF patients when discharged had a scheduled follow-up appointment with a PCP or cardiologist within two weeks. At the end of the intervention about 56% of all discharged patients kept their follow-up appointments and they were able to reduce the 30-day readmission rates for CHF patients to 14%, which was a 50% reduction from the previous rates. These interventions have proven to help reduce the readmission rates of CHF patients as well as by having an adequate number of nursing staff to help with the education, optimizing of medical therapy and carrying out the interventions (Nair, Lak, Hasan, Gunasekaran, Babar, & Gopalakrishna, 2020).
Nair, R., Lak, H., Hasan, S., Gunasekaran, D., Babar, A., & Gopalakrishna, K. V. (2020). Reducing all-cause 30-day hospital readmissions for patients presenting with acute heart failure exacerbations: A quality improvement initiative. Cureus. https://doi.org/10.7759/cureus.7420
It can be extremely difficult to obtain a patients medical and psychiatric history if the patient or family is unable to provide these details. Having a database of patient information including any Emergency Department visits, Psychiatric services, any outpatient clinic or primary care visit and current medications or discharge plans or appointments that the patient was to attend. This is especially difficult with patients who suffer from serious mental illness and require emergency services due to unsafe behavior or thoughts. The person may not be well enough to provide a medication list or provide historical data.
Having a database of medical and psychiatric history at every providers fingertips allows the most efficient decisions possible for the patient. This removes barriers for treatment and can accelerate wellness for the patient. “Health information exchange or HIE connects the electronic health record (EHR) systems of providers and clinicians allowing them to securely share patient information and better coordinate care. Health Current is Arizona’s health information exchange, connecting over 900 Arizona organizations, from first responders, hospitals, labs, community behavioral health and physical health providers to post-acute care and hospice providers.” (Healthcurrent, 2022).
Healthcurrent. What is HIE?. https://healthcurrent.org/hie/what-is-hie/.2022
Click here to ORDER an A++ paper from our Verified MASTERS and DOCTORATE WRITERS DNP 805 Select a specific clinical problem and post a clinical question that could potentially be answered using data mining:
The Covid 19 pandemic has left nursing across all disciplines forever changed. For some this has resulted in burn out and nursing leaving the profession. This in parallel to the nursing shortage has a potential for a negative impact in being able to provide care for those in need. The demand will continue to rise as the baby boomers continue to age in greater numbers than historically seen. What interventions are impactful in improving decreasing nursing turnover among nurses?
Data mining is looking at relationships and correlations to aims to predict outcomes. This is not a typical problem that you would think as being something that can be work on with data mining, but there are some opportunities using a principle component analysis. Data mining can reveal if there is a relationship between preventable nursing turnover and nurse salaries. Examples of unpreventable nursing turnover is retirement, relocation, death, and involuntary termination. Other variables to look at related to preventable nursing turn over would be leapfrog rating, CMS Stars, Magnet status, mandated patient ratios, workplace violence incidents, employee injuries, and union hospitals. The ability to data mine these items in comparison preventable nursing turnover will help guide what is most important to nursing to then have targeted interventions to decrease this turnover and keep nurses in the profession. One study did find a correlation with workplace violence and turnover in two large teaching hospitals (Yeh et al., 2020).
Once it is identified what seems to be the most important components that keep nurses in their roles will allow for focusing on those things to improve and then market that when recruiting nurses into the organization. With the shortage it is important to retain the nurses that you have and creatively market new ones in. This includes taking more new graduate nurses than historically taken.
Yeh, T.-F., Chang, Y.-C., Feng, W.-H., Sclerosis, M., & Yang, C.-C. (2020). Effect of Workplace Violence on Turnover Intention: The Mediating Roles of Job Control, Psychological Demands, and Social Support. Inquiry : A Journal of Medical Care Organization, Provision and Financing, 57, 46958020969313. https://doi-org.lopes.idm.oclc.org/10.1177/0046958020969313
A specific clinical problem is unplanned hospital readmissions. Hospital readmission costs an estimated $25 billion dollars annually and compromises patient safety (Olson, et al., 2016). A clinical question that could potentially be answered using data mining is “Which patients are at the highest risk for hospital readmissions”. The data mining technique that appears most appropriate to analyze this clinical question would be the data relationship of classes which uses data to organize patients into groups to predict outcomes (Grand Canyon University, 2014). The Clinical Question is: What can be done to reduce the risks of re-hospitalization? The data to be analyzed for relationship trends and to answer this question might include:
1. Patients re-hospitalized within 30 days – to identify the patients in question.
2. Age – to look for age groups at risk for re-hospitalization.
3. Gender – to see if being a male or female is a possible predictor of re-hospitalization.
4. Diagnoses – to identify diagnosis groups that increase the risk of re-hospitalization.
5. Complications – to identify any complications during hospitalization that may increase the risk of re-hospitalization.
6. Length of stay – to identify if the length of hospital stay trend may pose a risk of readmission.
By identifying factors common among patients who experience re-hospitalization, the hope would be that patients could be identified as high risk upon admission to the hospital and care could be planned to prevent the identified trends and reduce the risk of re-admission. By decreasing the readmission rate, hospitals could save a significant amount of healthcare dollars and improve patient safety. However, in long-term care settings, re-hospitalization is a big challenge and threat. In the bid to reduce the re-hospitalization rate, patients’ emergency issues are sometimes managed with inadequate infrastructure, and nurses will eventually transfer the patient at the last minute.
Grand Canyon University. (2014). DNP-805 Lecture 5.
Olson, C. H., Dey, S., Kumar, V., Monsen, K. A., & Westra, B. L. (2016). Clustering of elderly patient subgroups to identify medication-related readmission risks. International Journal of Medical Informatics, 8543-52.
In evidence-based practice (EBP), the clinical question is generated before the literature search. The question serves as a guide for the search to find a targeted, empirically based answer, followed by integration of the external and internal evidence (clinical expertise with patient values and preferences) The clinical question subsequently guides the research, planning, implementation, and evaluation of a practice (Bermudez, 2021). The clinical question helps to narrow down the literature search results; therefore, the results of the search depend on a well-written question (Melnyk & Fineout-Overholt, 2019).
Data mining is a process that identifies valid and useful patterns in data as defined in research by Fayyad. Data mining is also sometimes used interchangeably with data analytics to describe also a process of discovering knowledge by using large databases though there are some small differences between the two (Alexander, Hoy, & Frith, 2019). The algorithms used in analyzing data in data mining are a set of mathematical instructions that is used in combinations for constructing predictive models which makes predictions about data with known results while descriptive models are used to explore data and identify the patterns or their relationships. An algorithm is a set of instructions that is used to perform a task (Alexander, Hoy, & Frith, 2019). Examples of predictive algorithms are decision tree, artificial neural networks, instance-based learning classifiers, support vector machine modeling, Bayesian modeling, classification, regression, time series analysis, and prediction and the forms of descriptive algorithms was summarization, clustering rules and association rules, sequence discovery (Alexander, Hoy, & Frith, 2019), (Anwar Lashari, Ibrahim, Senan, & Taujuddin, 2018).
The clinical problem of chronic heart failure within healthcare and the clinical question of whether to continue to use standardized medication treatments that may not be working is a question that needs to be answered. The guidelines have recommended standardized drug treatments for heart failure but there are still challenges with making the right clinical decision for the medications because of the complicated clinical presentations of heart failure. The decision trees used will be non-mutually exclusive and will have several leaf nodes and recommendations which are constructed via knowledge rules summarized from the HF clinical guidelines using the Apriori algorithm tool to mine the frequent patterns for transaction data (Bai, Yao, Jiang, Bian, Zhou, Sun, Hu, Sun, Xie, & He, 2022).
The other predictive algorithms that I will not use would be the Bayesian modelling which is used to estimate the conditional probability of given data points which belongs to a particular class. Decisions of classifications are made when collective probabilities are used with prior knowledge. This system is great with disease outbreaks and pandemics (Alexander, Hoy, & Frith, 2019).
Anwar Lashari, S., Ibrahim, R., Senan, N., & Taujuddin, N. S. (2018). Application of data mining techniques for medical data classification: A review. MATEC Web of Conferences, 150, 06003. https://doi.org/10.1051/matecconf/201815006003
Alexander, S., Hoy, H., & Frith, K. (2019). Applied clinical informatics for nurses (2nd ed.). Jones & Bartlett Learning.
Bai, Y., Yao, H., Jiang, X., Bian, S., Zhou, J., Sun, X., Hu, G., Sun, L., Xie, G., & He, K. (2022). Construction of a non-mutually exclusive decision tree for medication recommendation of chronic heart failure. Frontiers in Pharmacology, 12. https://doi.org/10.3389/fphar.2021.758573
Bermudez, N. (2021). Formulating Well-Written Clinical Practice Questions and Research Questions. Nursing & Health Sciences Research Journal, 4(1), 70-82. https://doi.org/10.55481/2578-3750.1113
Melnyk, B. M., & Fineout-Overholt, E. (2019). Evidence-based practice in nursing & healthcare: A guide to best practice. Fourth edition. Lippincott Williams & Wilkins.
I was reading up on your post and the reference to know more about the IPFQR. Inpatient Psychiatric Facilities (IPFs) that participate in the IPFQR Program report data related to inpatient psychiatric quality of care measures to CMS. The IPFQR Program is a pay-for-reporting program because IPFs that participate in the program and successfully meet all requirements receive the full annual payment update (APU). As required by the Social Security Act, participating IPFs must report these measures or receive a two-percentage point reduction to their APU. It appears to be similar to the fee for service that CMS also developed for the physicians. CMS issued a Medicare Physician Fee Schedule (PFS) final rule which will update the payment policies, payment rates and update other provisions for services. This has been in effect from January 2022. The rule is CMS-1751-F, a final rule revision to the payment policies under the Medicare Physician Fee Schedule Quality Payment Program was displayed on November 2, 2021 and published on November 19, 2021. The changes were to the physician fee schedule (PFS); Medicare Part B payment policies to ensure that payment systems are updated to reflect changes in medical practice, relative value of services, and changes in the statute; Medicare Shared Savings Program requirements; updates to the Quality Payment Program; Medicare coverage of opioid treatment programs and others.
Centers for Medicare & Medicaid (CMS). (2021, November 19). CMS-1751-F. Centers for Medicare & Medicaid Services | CMS. https://www.cms.gov/medicaremedicare-fee-service-paymentphysicianfeeschedpfs-federal-regulation-notices/cms-1751-f
Data mining is defined as the practice of examining a large pre-existing database in order to generate new information. Technology has a large potential to help organizations to focus on the most significant information in stored data warehouses. Data mining tools and techniques will predict future trends by making the business more proactive, and better knowledge-driven decisions. Data mining techniques could answer some questions that are related to the business which traditionally were too time-consuming to resolve (Zentut, 2018).
Heart disease is considered as one of the major causes of death throughout the world. It cannot be easily predicted by the medical practitioners as it is a difficult task which demands expertise and higher knowledge for prediction. Medical diagnosis is considered as a significant yet intricate task that needs to be carried out precisely and efficiently. The automation of the same would be highly beneficial. Clinical decisions are often made based on doctor’s intuition and experience rather than on the knowledge rich data hidden in the database. This practice leads to unwanted biases, errors and excessive medical costs which affects the quality of service provided to patients. Data mining have the potential to generate a knowledge-rich environment which can help to significantly improve the quality of clinical decisions (Sayad AT, Halkarnikar PP. (2014).
Data mining can be a very expensive process. For example, companies must hire additional employees and technology specialists to ensure that the data mining is done correctly. Many businesses must invest in advanced data mining software, which can also be expensive. The costs of data mining generally outweigh the benefits for most small businesses because they do not produce enough valuable insights.
Sayad AT, Halkarnikar PP. (2014). Diagnosis of heart disease using neural network approach. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5863635/ Int J Adv Sci Eng Technol. 2014;2:88–92.
Zentut, “Data Mining Techniques,” Zentut.com, 2018, http://www.zentut.com/data-mining/data-mining-techniques/.