November 09 2016 | 0 Comments | 244 reads Average Rating: 3
Leveraging Predictive Analytics to Make Sure Diseases Don’t Have a Fighting Chance
Currently caring for patients with chronic illness consumes over half of the health care dollar. Not surprisingly, then, many healthcare providers leverage analytics to determine how to best help patients who have chronic conditions – ultimately trying to keep them out of the hospital. However, healthcare organizations might also want to consider leveraging analytics to target the next level of patients who are bubbling under the surface, the ones who aren’t suffering yet but will be if they continue down the road they’re on.
Indeed, many industry leaders are advocating for healthcare organizations to move in this direction.
“Our traditional model, I think our meme for the way we practice healthcare is actually not healthcare, it has been sick care. We wait until often we get sick, we have the heart attack, the stroke, the cancer that pops up. Where things are sort of heading, I believe, is from waiting for disease to happen to wellness and prevention. In fifty years from now, we will have the equivalent of almost an “On-Star” for the body that’s going to be taking information from our environment, from our diet, from our social networks, from our genomics. Health will become something that is truly infused and integrated into our life,” said Daniel Kraft, MD, during his video presentation entitled Medicine 2064 with Dr. Daniel Kraft.1
With visionaries such as Dr. Kraft and others imagining the future, predictive analytics is emerging as a tool that can help shift from focusing on treating patients who have already been diagnosed with certain conditions – such as diabetes, chronic obstructive pulmonary disease (COPD), heart disease and asthma – to preventing people from developing these conditions in the first place.
For example, predictive analytics could help providers prevent patients from developing type 2 diabetes by analyzing data in past health assessments, previous medical claims, lab results and existing electronic medical records to anticipate who is at risk for the disease.2 Providers could then work with these patients on lifestyle modifications such as healthy eating and exercise to help avert or delay the onset of type 2 diabetes. Without such intervention, prediabetes is likely to progress to diabetes within 10 years for many patients, according to the Mayo Clinic.3
Predictive analytics could also help to identify undiagnosed diabetic patients – which could have a significant impact on population health. According to the American Diabetes Association 29 million Americans have diabetes – and of those more than 8 million are undiagnosed.4 With predictive analytics, healthcare organizations can segment patients by risk and compliance and identify those who are undiagnosed. The use of zip code data could then help to identify geographic pockets of undiagnosed patients while patient persona data could help move toward a better understanding of why these patients are at risk – and what interventions or therapies are most likely to improve their health.
Understanding the gaps in care and non-compliance is critical. For example, an analytic study conducted by SCIO on behalf of a Florida-based healthcare organization identified pockets of high risk patients who were not filling or refilling medications. When qualitative analytics came into play, however, SCIO was able to help the organization uncover why this gap existed. The analysis showed that patients had to take two left turns to get to the pharmacy and that elderly patients in Florida do not like to take left turns. Thus, through this analysis, healthcare leaders could start to identify interventions that might help to address this compliance roadblock.
Can you think of other ways that qualitative analysis might help to move predictive analytics efforts forward?
1. Medicine 2064 with Dr. Daniel Kraft. Accessed at: https://www.youtube.com/watch?v=iOgt85cPU8Q
2. Beveridge, R. Using predictive modeling to prevent diabetes. Accessed at: http://managedhealthcareexecutive.modernmedicine.com/managed-healthcare-executive/news/using-predictive-modeling-prevent-diabetes
3. Mayo Clinic. Prediabetes. Accessed at: http://www.mayoclinic.org/diseases-conditions/prediabetes/basics/definition/con-20024420
4. American Diabetes Association. Statistics About Diabetes. Accessed at: http://www.diabetes.org/diabetes-basics/statistics/