Predictive Analytics in HealthCare

of one

Predictive Analytics Case Studies

Three case studies that speak on predictive analysis initiatives brought on by health care systems and health-related institutions to improve patient safety were found. In the first case study, a systematic review of the whole US clinical system was presented. The second study looked at predictive analytics improving the safety of patients with genetic disease, specifically osteoporosis and cardiovascular disease patients in Los Angeles. The third study examined the impact of predictive analytics based on continuous cardiorespiratory monitoring in a surgical and trauma intensive care unit in Charlottesville.

CASE STUDY #1: A Systematic Review on Healthcare Analytics: Application and Theoretical Perspective of Data Mining

  • Data mining and predictive analytics have been used in this study to address provide a systematic view of development in the healthcare field in the US and possible future directions the healthcare system could take based on analytics in clinical and administrative decision-making.
  • The study uses human-generated data pulled from Electronic Medical Record in clinical care. According to the review of the system as of now, prediction models are better used for predicting commonplace events and practices than rare ones.
  • Moreover, in order for predictive analytics to work within the clinical system in the US, data mining needs to be used in combination with expert opinion from each domain of medicine, such as oncologist for cancer study, or internist for gastrointeral studies.

CASE STUDY #2: Predictive Analytics to Determine the Potential Occurrence of Genetic Disease and their Correlation: Osteoporosis and Cardiovascular Disease

  • In this case study, the predictive analytics model was designed, developed, and used to determine "the risk of manifesting osteoporosis in later life using big data processing."
  • The model that was used leveraged the novel genetic pleiotropic information of over 3,500 different patients. However, the information overall tested was based on the population of all patients coming in with osteoporosis and cardiovascular problems in the overall LA area.
  • Additionally, mutations associated with osteoporosis and cardiovascular disease were also included in the analysis.
  • The results from this case study showed that predictive analytics can be used to draw correlation "between a person's regional background and the frequency of occurrence of the 35 Single Nucleotide Polymorphisms associated with osteoporosis and/or cardiovascular disease (CVD)."
  • The models used machine learning algorithms, mainly Logistic Regression, Adaboost, and KNN.

CASE STUDY #3: Impact of predictive analytics based on continuous cardiorespiratory monitoring in a surgical and trauma intensive care unit

  • Predictive analytics monitoring when it comes to the use patient data in this case study for Charlotteville was used to provide continuous risk estimation of deterioration when it comes to the care of individual patients in the intensive care units.
  • The risk estimates were based on "analysis of continuous cardiorespiratory monitoring" and more than 4,275 individual patient records were looked at within a 7-month time period. The overall population in the study included the whole area of Charlottesville with any surgical and trauma intensive care unit experience within that 7 month period.
  • By using predictive analytics, the case study was able to determine "cases of septic shock, emergency intubation, hemorrhage, and death to compare rates per patient care pre- and post-implementation."


In finding case studies that detail successes that predictive analysis had when it comes to improving the safety of patients, we mostly did straightforward searches for case studies on the topic. We ensured that these results were from healthcare-related websites, such as databases for scientific research, including Research Gate, Spriger, and US National Library of Medicine. In our search, we were finding many case studies by singular hospitals, but to find the 'biggest' patient safety lessons, we decided to use case studies of initiatives and programs that had a wider impact based on the health system's or hospital's size or influence. We also used the metric of the number of people that were involved and that would be affected. With this definition, the first case study we chose was about the largest health system the country — the US healthcare system as a whole. This study was chosen as it involved multiple hospitals, health systems, clinics, etc. The second study looked into the population of Los Angeles, one of the largest cities in the US. The third case study was chosen because it involved monitoring ICE units in Charlottesville, a city with over 150,000 people living in it.