Veteran Population and VA expenditures for SoftMath Consultants, LLC. www.softmathconsultants.com
What we do Analytics http://www.softmathconsultants.com Predictive Analytics Big Data and Advanced Analytics Healthcare Analytics Text Analytics - Deep Learning Algorithms Real Estate Market Analytics www.softmathconsultants.com 2
People Dr. Swarna Reddy https://www.linkedin.com/in/drswarnareddy Prof. Brian Steele http://web.math.edu/steele/ www.softmathconsultants.com 3
Healthcare Analytics-Synopsis Healthcare analytics refers to data analytic methods applied in the healthcare domain. Healthcare analytics is becoming a prominent data science domain because of the societal and economic burden of disease. because there are significant opportunities to better understand the system through the analysis of data from the system. www.softmathconsultants.com 4
Healthcare Analytics-Synopsis cont.. Healthcare accounts for 17% of the gross domestic product of the United States, much more than any other country in the world.[2] The providers of healthcare are under immense pressure to improve delivery and reduce costs. Reining in costs while maintaining quality of care is critically needed. The system is complex, opaque, and undergoing change, all of which makes it difficult to answer even elementary questions about the system and the people that rely on it. Healthcare analytics has developed in the pursuit of information supporting changes in the delivery of healthcare. www.softmathconsultants.com 5
Importance of the study Type 2, or adult-onset diabetes is of great importance in the U.S. public health arena because it is a serious chronic disease in the U.S. population. It s associated with a spectrum of complications including, but not limited to nerve and kidney damage, cardiovascular disease, retinopathy, skin conditions, and hearing impairment. Diabetes is considered to be a lifelong condition and considered hard to cure. This particular study is about the Veteran Population. www.softmathconsultants.com 6
Annual Healthcare Cost Calculations as a Function of Body Mass Index (BMI) To gain some exposure to healthcare data and analytics, we undertake an analysis of diabetes prevalence and incidence and computed annual healthcare costs as a function of body mass index (BMI). What is Prevalence? Prevalence is the proportion or percent of a population that is affected by a condition or disease For example, if the prevalence of diabetes is.12, then 12 of every 100 individuals in the population is has the condition. www.softmathconsultants.com 7
Incidence What is incidence? Incidence is the rate at which new cases appear among a population at risk. in other words: The incidence of diabetes is the annual rate at which individuals become afflicted with the disease. If the incidence is.004 then 4 new cases will occur per 1000 unaffiliated individuals. www.softmathconsultants.com 8
The data We used publicly available data collected and maintained by Centers for Disease Control and Prevention. The Behavioral Risk Factor Surveillance System (BFRSS), Medical Expenditure Panel Survey (MEPS) www.softmathconsultants.com 9
BFRSS and MEPS The survey asks a sample of U.S. adult residents a large number of questions regarding health and health-related behaviors as well as personal status. In this survey, veterans constitute 12% to 15% of the population in the observed years. The following table shows the example of a question relates to veterans status. Have you ever served on active duty in the United States Armed Forces, either in the regular military or in a National Guard or military reserve unit? Value Value Label Frequency Percentage 1 Yes 36,525 14.18 2 No 220,376 85.55 7 Don t know/not Sure 82 0.03 9 Refused 613 0.24 Blank Missing 7088 www.softmathconsultants.com 10
BFRSS survey The following figure shows the percentage of veteran population (2003-2014) in terms of total population participated in the survey. The left side figure shows the frequency in the veteran population increased over time. The right side figure shows the percentage of total decreased over time. www.softmathconsultants.com 11
Medical Expenditure Panel Survey (MEPS) The Medical Expenditure Panel Survey, which began in 1996, across the United States. MEPS collects data on the specific health services that Americans use, how frequently they use them, the cost of these services, and how they are paid for, as well as data on the cost, scope, and breadth of health insurance held by and available to U.S. workers. To associate demographic profiles from BFRSS, with the demographic profiles of MEPS to associate the costs. We used smoothing distances (knn-algorithm) for this purpose. To gain exposure, we used Prescription Drug Costs of MEPS Consolidated data. We analyzed Total Cost (covered by all possible coverages), Medicaid Coverage, Medicare coverage and Veteran dministration Coverage. www.softmathconsultants.com 12
Mean age distribution over the years Agecateg ories Actual Age 1 18-24 2 25-29 3 30-34 4 35-39 5 40-44 6 45-49 7 50-54 8 55-59 9 60-64 10 65-69 11 70-74 12 75-79 13 >80 www.softmathconsultants.com 13
Mean-BMI distribution over the years www.softmathconsultants.com 14
Cost share distribution among the coverage www.softmathconsultants.com 15
Prescription drug cost distribution between Gov. Agencies The cost distribution in the figures show the distribution of annual prescription drug (pdx) costs for Medicaid (MCD), Medicare (MCR), and Veteran Administration (VA) and total prescription drug costs. Costs have been transformed to log-10 dollars for readability We see the Medicare prescription drug costs are greatest among the three agencies with an approximate median of $1000. Medicare pdx costs per patient are greater than VA costs. www.softmathconsultants.com 16
Prescription drug coverage costs- box plot The box plot shows comparison of the mean prescription drug coverage costs www.softmathconsultants.com 17
Prevalence Incidence We used simple linear regression to compute incidence (rate of change in prevalance). Prevalence is a function of year. We found out the rate of change per year is 0.004387 Here tthe incidence is.004 then 4 new cases will occur per 1000 unaffiliated individuals. www.softmathconsultants.com 18
The rate of change in Prevalence or Incidence The plot shows the diabetes prevalence over the years. Year is plotted on the x-axis prevalence of diabetes is plotted on the y- axis. www.softmathconsultants.com 19
References 1. Algorithms for Data Science., Authors: Steele, Brian, Chandler, John, Reddy, Swarna http://www.springer.com/us/book/9783319457956 2. American Diabetes Association. Statistics about diabetes 2014. www.softmathconsultants.com 20