top of page

Cardiometabolic Analytics

Data analysis to identify patterns in patient demographics, pharmacotherapy adoption, and healthcare resource utilization (HRU) to optimize clinical care pathways for cardiometabolic populations.

Business question: Which cardiometabolic indications drive the highest post-diagnosis outpatient volume, and what is the standard timeline and rate for patients to initiate first-line pharmacotherapy?

Technology stack: SAS (for cleansing data & ETL), Microsoft Power BI (for dashboarding).

Cardiometabolci_Analytics.jpg
SAS_Cardiometabolic.png
Key Insights

​

  • The 1-year treatment initiation rate is exceptionally high at 96.3%, indicating rapid adoption of pharmacotherapy following a cardiometabolic diagnosis.

  • Healthcare resource utilization (HRU) demonstrates a clear increase in outpatient visits during the 1-year post-index period compared to the pre-index baseline, most notably for Hypertension and Coronary Artery Disease.

  • The Kaplan-Meier time-to-treatment curve shows a steep initial step function, revealing that the vast majority of treated patients receive their first prescription immediately or very shortly after diagnosis.

  • While "Other Regimen" is the most common broad category for initial treatment, Thiazide and CCB (Amlodipine) lead the specific first-line cardiovascular monotherapies.

  • The overall cardiometabolic cohort consists of 573 patients, representing a substantial 49.3% prevalence rate within the broader analytical population.

  • The average age at diagnosis across the cohort is roughly 27 years, with Type 2 Diabetes and Prediabetes patients generally presenting at older ages compared to the Hypertension cohort.

1. Data Source

​

  • Data source is derived from synthetic patient records (Synthea).

  • The dataset is standardized and structured according to the OMOP Common Data Model (CDM) v5.4 architecture.

  • Core clinical domains analyzed include: Patient Demographics, Condition Occurrence (Diagnoses), Drug Exposure (Prescriptions), and Visit Occurrence (Encounters).

2. Data Exploration
 

  • Data was processed via SAS OnDemand scripts and exported as analytical CSV files (fact_cardio_cohort.csv and fact_hru_utilization.csv) containing 573 primary index records.

  • Raw electronic health records were standardized into the OMOP CDM v5.4 architecture by systematically mapping clinical source terminologies (ICD-10, SNOMED, and RxNorm) into standard concept identifiers across condition, drug exposure, visit, and person tables.

  • Through SAS OnDemand workflows, cohort exploratory analyses established a uniform baseline by indexing first-occurrence cardiometabolic diagnoses, harmonizing person identifiers, and computing longitudinal time-to-treatment windows and pre/post encounter frequencies for downstream visual reporting.

3. Data Visualization
 

  • In Power BI, a snowflake schema data model was developed featuring custom DAX calculated measures, categorical slicers, and interactive dashboard filters.

  • The visualizations used include KPI Scorecard cards, a Horizontal Bar Chart, Clustered Column Charts, a Step-Interpolated Line Chart, and a Donut Chart.

  • Each visualization includes customized tooltips to provide precise patient counts, encounter averages, and percentages on hover.

Other projects

bottom of page