NWCDC 2018 Pharmacy Analytics in Workers Comp Cliff Belliveau VP Business Intelligence
What is Visual Analytics? Transforming data into insights that tell a story 2
A little history Jacquard loom Punch Cards Textile Pattern 3
Transforming data into information 1820 2018 Punch Cards Loom Textile Pattern Data Data Process Visual Information 4
Visual Analytics Data Visualization Modern equivalent of visual communication "An ideal visualization should not only communicate clearly, but stimulate viewer engagement and attention Over 90% of the worlds data has been created in the past 2 years The human brain processes images 60K faster than text 90% of the information transmitted to the brain is visual - Fernanda Viegas and Martin M. Wattenberg, IBM 5
Top Trends in Analytics 1. Data Quality Management (DQM) 2. Data Discovery 3. AI, Machine Learning, Predictive & Prescriptive Analytics 4. Connected Clouds 5. Data Governance and Trust 6. Security, Digital Ethics and Privacy 7. Collaborative BI Machine learning patents grew at a 34% between 2013 and 2017 Third-fastest growing category of all patents granted Spending on AI/ML will grow from $12B in 2017 to $57.6B by 2021 Source: International Data Corporation (IDC) Healthcare s AI market will reach $6.6 billion by 2021, from $600 million in 2014 Source: Accenture 6
Workers Comp Data 7
Organizing data to answer questions quickly Groceries The Pantry Data Warehouses are pantries for data and information 8
Revealing Outliers in a Patient Population Unveiling hidden metrics 9
The ranking dilemma Which patient(s) has the oldest claims and the highest opioid utilization?
Comparing multiple values visually Which patient(s) has the oldest claims and the highest opioid utilization? Answer: Patient C Displays large populations of data Useful for seeing correlations Fixed time period 11
Varying perspectives yield different stories 12
Segmenting population reveals more insights Categorization of data point within the population can lead to deeper understanding: Geography Demographics Market Jurisdiction 13
Measuring Pharmacy Utilization Analyzing patterns to better target patients for intervention 14
Common utilization patterns What We Measure Benzodiazepine Use Compound Use Concurrent Use of Benzodiazepines and Opioids Concurrent Use of Benzodiazepines, Opioids and Soma Concurrent Use of Buprenorphine and Opioids High Dose Acetaminophen More Than One Opioid Type Multiple Opioid Prescribers Multiple Pharmacies Naloxone Use New Start Long-Acting Opioid Opioid Use Sedative Hypnotic Use Skeletal Muscle Relaxant Use Therapeutic Duplication Daily Morphine Equivalent Dose Trans mucosal Immediate Release Fentanyl How We Measure Days Supply Rx Count Concurrent Day Supply Concurrent Day Supply Concurrent Day Supply mg per Day CPE Count Unique Prescribers Unique Pharmacies Rx Count Rx Count Days Supply Days Supply Days Supply CPE Count Aggregate MED Days Supply 15
Difficult to measure utilization patterns from detail
Measuring and trending patient s utilization 17
Assigning safety risk scores to patient s utilization 18
Measuring Intervention Effectiveness Understand the impact of clinical interventions to utilization
Typical interventions in WC pharmacy Pharmacy/Provider alerts Utilization review Physician consultation Patient education Clinical review Prescriber Injured Worker Pharmacist Claims Examiner Email Text Message Letter Voice Call FAX 20
Visualizing the effect of interventions on utilization
Using safety risk score to measure intervention effectiveness
Machine Learning / AI Where we are and what the future holds 23
Machine learning & predictive analytics
Applying predictive models to utilization and safety risk scores
Wrap up Share data Build a pantry Transform data Use visual analytics Think big Experiment Become data-driven
Thank You! analytics@mymatrixx.com 27