Section 12.1: The Foundations of Healthcare Data & Analytics

12.1.1 The “Why”: From Clinical Intuition to Data-Driven Validation

As an experienced pharmacist, you are a master of pattern recognition. You operate on a highly-tuned “clinical intuition” built from thousands of patient interactions. You know which patients are at high risk for non-adherence. You know which prescribers are most likely to accept your recommendations. You know that your MTM interventions are preventing hospitalizations. You can feel the rhythm of your workflow, and you know when your team is operating efficiently and when it is stretched thin.

For decades, this clinical intuition was the primary tool of our profession. It was enough to be a good clinician. This is no longer the case.

We have entered a new era of healthcare defined by one central shift: the transition from fee-for-service (where you are paid for the act of dispensing) to value-based care (where you are paid for the outcome of your actions). In this new world, your intuition is not enough. You must be able to prove it. Payers, health systems, and manufacturers no longer pay for effort; they pay for demonstrable value. Your “feeling” that you are improving care must be translated into a hard-number report that shows a 15% increase in adherence and a 10% reduction in associated medical costs.

This module is not about replacing your intuition. It is about amplifying it, validating it, and scaling it with the power of data. Your pharmacy’s dispensing system is not just a tool for filling prescriptions; it is a rich, longitudinal data warehouse. Every prescription you fill, every clinical note you type, every patient you counsel is a data point. When aggregated, these data points tell a powerful story.

Your role as a Certified Advanced Specialty Pharmacist (CASP) is to become a “translator” — an expert who can speak both the language of clinical practice and the language of data. You are the nexus. You are the only person who can see the claims data from the payer, the clinical data from the prescriber (via the EHR), the dispensing and adherence data from your own system, and the qualitative data from the patient themselves. This unique position makes you the most valuable data expert on the healthcare team.

This foundational section is your boot camp. We will deconstruct this new responsibility into five core competencies that build upon each other:

1. Real-World Evidence (RWE) & Data Sources: First, we will map the entire data landscape. You will learn to identify and understand the “raw materials” of healthcare data—from claims and EHRs to your own dispensing logs—and how they are used to generate evidence.
2. KPI Dashboard Development: Next, you will learn how to use this data for internal improvement. We will build effective dashboards to monitor your own Key Performance Indicators (KPIs) for both clinical quality and operational efficiency.
3. Predictive Analytics for Risk Stratification: Once you can measure the past, you will learn to predict the future. We will explore how to use data science to proactively identify high-risk patients before they fail therapy, allowing for targeted, life-saving interventions.
4. Outcomes Reporting for Payers and Manufacturers: Then, you will learn to package your data for external stakeholders. We will master the art of creating reports that prove your clinical and economic value to the entities that pay for your services.
5. Data Visualization and Benchmarking: Finally, you will learn to be a master communicator. We will cover how to use clear, compelling charts and graphs to tell your data’s story and how to benchmark your performance against industry standards to prove you are best-in-class.

This is the future of pharmacy. It is how you will secure your role as an indispensable provider, justify new clinical services, and, most importantly, systematically improve the health of your entire patient population. Let’s begin.

12.1.2 Pharmacist Analogy: The Pharmacy Inventory as a Data Warehouse

12.1.3 Deep Dive: The Landscape of Real-World Data (RWD) and Real-World Evidence (RWE)

This is the core of our new discipline. We must first understand the “raw materials” we work with. The entire field of data analytics is built on this foundation.

The Core Definitions: RWD vs. RWE

These two terms are often used interchangeably, but they are critically different. As an expert, you must know the distinction.

• Real-World Data (RWD): The Raw Material.

  This is the “data” itself. RWD is defined as data relating to patient health status and/or the delivery of health care, collected from a variety of sources outside of traditional, highly-controlled Randomized Controlled Trials (RCTs). It is the “messy” data from the real world. Think of RWD as the raw logs, the dispensing records, the lab values, the claims forms. It’s just a collection of facts.

• Real-World Evidence (RWE): The Finished Product.

  This is the “insight” you generate from the raw data. RWE is defined as the clinical evidence regarding the usage and potential benefits or risks of a medical product or intervention, derived from the analysis of RWD. Think of RWE as the published report, the clinical conclusion, the answer to a question. (e.g., “Our analysis of RWD shows that patients in our MTM program had a 15% higher adherence rate.”).

In short: You analyze RWD to generate RWE.

The Great Debate: RWE vs. Randomized Controlled Trials (RCTs)

The “gold standard” of clinical evidence has always been the RCT. So why do we even need RWE? Because RCTs and RWE answer two fundamentally different questions: Efficacy vs. Effectiveness.

• Efficacy (The RCT Question): “Can this drug work under perfect, ideal, laboratory-like conditions?” RCTs use highly selective, “clean” patient populations (e.g., no comorbidities, perfect adherence) to prove a drug is safe and effective for FDA approval.
• Effectiveness (The RWE Question): “Does this drug actually work in the real world?” RWE studies use “messy” patient populations (e.g., 80-year-olds with 10 comorbidities, poor adherence, multiple prescribers) to see how the drug performs in your actual patient population.

As a pharmacist, you live in the “effectiveness” world. RWE is your domain. An RCT tells you a new DOAC is effective; RWE tells you if it’s still effective in your patient who also takes carbamazepine and misses 10 doses a month.

Masterclass Table: RWE vs. RCT

Metric	Randomized Controlled Trial (RCT)	Real-World Evidence (RWE) Study
Primary Question	Efficacy: “Can it work?”	Effectiveness: “Does it work?”
Study Design	Experimental, Interventional. Always prospective. Randomization and blinding are key.	Observational. Usually retrospective (using past data), but can be prospective (registries).
Patient Population	Homogeneous. Very strict inclusion/exclusion criteria (e.g., “non-smoking males, 40-50, no comorbidities”).	Heterogeneous. “All comers.” Represents your actual, “messy” patient population (e.g., all ages, all comorbidities, polypharmacy).
Internal Validity	Very High. Randomization minimizes bias and confounding. High confidence that the drug caused the outcome.	Low to Moderate. High risk of bias and confounding (e.g., did the statin save the patient, or were they just more health-conscious?). Requires advanced statistical adjustment.
External Validity (Generalizability)	Very Low. Results may not apply to your 80-year-old female patient with renal failure, because she would have been excluded from the trial.	Very High. The results are directly applicable to your real-world patient population because they were derived from it.
Data Collection	Proactive, manual, and extremely expensive. Case Report Forms (CRFs) capture specific, pre-defined data points.	Retrospective, often automated. Uses data that was already collected for routine clinical care or billing. Relatively inexpensive.
Pharmacist’s Role	As a clinical trial investigator, you ensure protocol adherence.	As a practitioner-analyst, you generate and interpret RWE to guide care for populations.

A Comprehensive Guide to Real-World Data (RWD) Sources

To generate RWE, you must be a master of the raw materials (RWD). These sources are not mutually exclusive; their true power is unlocked when they are linked. But first, you must understand their individual strengths and, more importantly, their critical weaknesses.

1. Claims and Billing Data (The “Financial” Record)

What It Is: This is the data generated every time a healthcare service is paid for. It is the “financial exhaust” of the healthcare system. It includes pharmacy claims (NCPDP standard) and medical claims (CMS-1500 for outpatient, UB-04 for inpatient).

What’s Inside a Pharmacy Claim:

• Identifiers: De-identified Patient ID, Prescriber NPI, Pharmacy NPI.
• Drug Info: National Drug Code (NDC), Quantity Dispensed, Days Supply, Date of Fill, Refill Number.
• Financials: Drug Cost (Ingredient Cost + Dispensing Fee), Copay, Payer Name, BIN/PCN/Group.

What’s Inside a Medical Claim:

• Identifiers: De-identified Patient ID, Provider NPI, Facility ID.
• Encounter Info: Date of Service, Place of Service.
• Billing Codes:
  • ICD-10-CM Codes: International Classification of Diseases. These are the diagnosis codes (e.g., `E11.9`: Type 2 diabetes, `I10`: Hypertension).
  • CPT/HCPCS Codes: Current Procedural Terminology. These are the procedure codes (e.g., `99213`: Office visit, `83036`: HbA1c lab test).
• Financials: Amount Billed, Amount Paid.

Strengths of Claims Data:

• Longitudinal & Comprehensive: This is its superpower. Because it follows the money, you can see every encounter a patient has, even with different doctors in different health systems. You can build a patient journey over years.
• Captures Adherence: Claims data is the only reliable source for calculating adherence metrics (like PDC/MPR) because it represents what the patient actually picked up, not just what was prescribed.
• Captures Cost: It is the source of truth for all pharmacoeconomic analyses (e.g., “Total Cost of Care”).

2. Electronic Health Record (EHR) Data (The “Clinical” Record)

What It Is: This is the data generated from the clinical workflow within a health system (e.g., Epic, Cerner, Allscripts). It is the digital version of the patient’s chart.

What’s Inside an EHR Record:

• Identifiers: Medical Record Number (MRN), Patient Demographics.
• Structured Data (The “Easy” Stuff):
  • Problem List: Diagnoses (ICD-10).
  • Medication List: e-Prescriptions (what was ordered) and Medication Administration Records (MARs) (what was given inpatient).
  • Allergies: Standardized list.
  • Vitals: Blood Pressure, Heart Rate, Weight, Height, BMI, Temp.
  • Lab Results: The actual numeric values (e.g., `HbA1c = 8.1%`, `SCr = 1.4 mg/dL`).
• Unstructured Data (The “Hard” Stuff):
  • Provider Notes, Discharge Summaries, Consult Notes, Radiology Reports. This is free text (e.g., “Patient reports frustration with high copay and is splitting pills.”).

Strengths of EHR Data:

• UNMATCHED CLINICAL RICHNESS: This is its superpower. You have the lab results, the vitals, the diagnoses. You can answer “why” a drug was prescribed and “what happened” as a result.
• Real-Time: The data is generated during the encounter. It is the most up-to-date source of information.
• Captures Prescriber Intent: You see what was ordered, not just what was filled. This is how you find “primary non-adherence” (prescriptions that were ordered but never picked up).

The Pharmacist’s Opportunity: You are the only one who can bridge the gap between the EHR (what was prescribed) and the Claims data (what was picked up).

3. Pharmacy Dispensing & Management System Data (YOUR Data)

What It Is: This is the data from your own pharmacy’s software (e.g., PioneerRx, Rx30, EnterpriseRx, QS/1). It is a powerful hybrid dataset that you create every day.

What’s Inside Your System’s Data:

• Claims/Dispensing Data: You have a copy of every prescription claim you’ve ever processed for that patient.
• Pharmacist Clinical Notes: This is your gold mine. MTM platform notes, adherence call logs, intervention documentation (e.g., “Called Dr. Smith, rec’d switch from Xarelto to Eliquis due to cost”).
• Clinical Service Data: Vaccination records, point-of-care test results (Strep, Flu, A1c), blood pressure readings from your in-store cuff.
• Patient-Reported Data: Information you collect directly (e.g., “patient reports taking med 5/7 days,” “patient reports side effect of cough”).

Strengths of Your Data:

• LINKS INTERVENTION TO OUTCOME: This is the only dataset in the world that connects your specific clinical intervention (e.g., an adherence call) to a dispensing outcome (e.g., the patient refilled the next day).
• Captures the “Why”: Your notes are the richest source of “unstructured” clinical insight.
• It is YOURS: You can structure it, control it, and query it in real-time to manage your own practice.

Weaknesses of Your Data:

• HIGHLY FRAGMENTED: It’s siloed to your pharmacy or chain. You don’t know what the patient filled at a competitor.
• NOT STANDARDIZED: This is the biggest challenge. One pharmacist might type “Pt non-adherent,” another “patient missed refill,” and another “Called pt re: late Rx.” These are impossible to query.

4. Patient-Generated Health Data (PGHD) & Other Sources

This is the new frontier of RWD, and it’s exploding.

• Patient Registries: These are organized, observational studies that collect highly specific data on one disease (e.g., a cystic fibrosis registry). They are excellent for studying disease progression and often include Patient-Reported Outcomes (PROs) (e.g., quality-of-life surveys).
• PGHD: Data generated by the patients themselves. This includes data from:
  • Smart Devices: Apple Watch (EKG, Afib), smart scales (weight).
  • Connected Medical Devices: Continuous Glucose Monitors (CGMs) like Dexcom/Freestyle, smart inhalers (Propeller Health), smart blood pressure cuffs.
  • Apps: PRO surveys, mood trackers, diet/exercise logs.
• Social Determinants of Health (SDoH) Data: This is dataset enrichment. You can buy data (from sources like LexisNexis) that links a patient’s zip code to data on income, education level, transportation access, and “food deserts.” This is highly predictive of health outcomes.

The Holy Grail: Data Linkage. The true power comes when you link these datasets. Imagine a patient record that combines:
1. Claims Data: We see they picked up their insulin (PDC = 90%).
2. EHR Data: Their last HbA1c was 9.5% (a clinical problem).
3. PGHD Data: Their CGM log shows they are adherent, but their glucose is spiking every night after dinner.
4. Pharmacist Data: Your note says, “Patient is adherent but reports confusion about carb-counting for her evening meal.”

Without linkage, you have conflicting data. With linkage, you have a precise, actionable clinical insight. The problem isn’t adherence; it’s diet education. This is the power of advanced data analytics.

12.1.4 Application 1: Key Performance Indicator (KPI) Dashboard Development

Now that you understand your “raw materials” (RWD), your first application is to use them for internal improvement. You cannot manage what you do not measure. A Key Performance Indicator (KPI) dashboard is your practice’s “cockpit” — it gives you a real-time view of your performance.

• Key Performance Indicator (KPI): A measurable value that demonstrates how effectively your pharmacy is achieving key clinical or business objectives.
• Dashboard: A visual display of your most important KPIs, consolidated on a single screen so they can be monitored at a glance.

Your Skill Translation: You already use dashboards every day. Your pharmacy system’s “To-Do List” or “Refill Queue” is a basic operational dashboard. It shows you KPIs like `Prescriptions Waiting for Verification` or `Prescriptions Due Today`. We are simply expanding this concept to include clinical and financial KPIs.

A great pharmacy dashboard is balanced. It measures two key areas: Clinical Quality (Are we improving health?) and Operational Efficiency (Are we doing it well?).

Masterclass Table: Designing Your Pharmacy KPI Dashboard

KPI Category	Key Performance Indicator (KPI)	Data Source(s) Needed	The “So What?” (Why It Matters)
Clinical Quality KPIs (Are we improving health?)	Adherence (PDC) for Star Ratings	Pharmacy Dispensing Data or Payer Claims Data	Directly impacts payer reimbursement and patient outcomes. This is your #1 quality metric.
Statin Use in Persons with Diabetes (SUPD)	Dispensing Data + Payer Claims (for diabetes diagnosis)	A key HEDIS/Star Ratings measure. Shows you are practicing evidence-based medicine.
Comprehensive MTM (CMR) Completion Rate	MTM Platform (e.g., OutcomesMTM)	Payer contract compliance and direct revenue generation.
New-to-Therapy (NTT) Discontinuation Rate	Pharmacy Dispensing Data	Measures your effectiveness at counseling and onboarding patients to new meds. High value to manufacturers.
Vaccination Rate (e.g., Flu, Shingrix)	Dispensing/EHR/Registry Data	Public health impact and a primary driver of non-dispensing revenue.
Operational Efficiency KPIs (Are we running well?)	Prescription Turnaround Time (Avg)	Pharmacy System (e.g., time from ‘Received’ to ‘Verified’ and ‘Verified’ to ‘Filled’)	Directly impacts patient satisfaction and workflow stress.
Dispensing Accuracy Rate	Pharmacy System + Quality Assurance Log	Measures patient safety. (e.g., `(Total Fills – Dispensing Errors) / Total Fills`).
Call Center Wait Time / Abandon Rate	Phone System Data (if available)	A major driver of patient frustration. Identifies staffing needs.
Inventory Turn Rate	Inventory/Purchasing Data	Measures your financial health. `(Cost of Goods Sold / Avg. Inventory)`. High turns = good cash flow.
Clinical Interventions per Pharmacist FTE	Pharmacy System (Clinical Notes)	Measures the clinical productivity of your team. (Requires structured note-taking!)

12.1.5 Application 2: Predictive Analytics for Risk Stratification

Once you have mastered measuring your performance (KPIs), the next evolution is to predict the future. This is the most powerful application of data analytics in healthcare. Instead of reacting to problems, you prevent them.

• Risk Stratification: The process of dividing your patient population into “buckets” of low, medium, and high risk.
• Predictive Analytics: The engine that does this. It uses data science and statistical modeling on your RWD to calculate a “risk score” for every patient for a specific negative outcome.

The Goal: To move your pharmacy from a reactive model to a proactive, data-driven model.

Reactive Model (The Past): “Mr. Smith’s lisinopril is 10 days late in the refill queue. I should call him.” (The non-adherence has already happened).

Proactive Model (The Future): “My dashboard shows Mr. Smith has a 92% probability of becoming non-adherent in the next 30 days. I should call him today to find out why.”

How Does a Predictive Model Work? (A Pharmacist’s Guide)

You don’t need to be a data scientist to use a model, but you must understand how it thinks. Your clinical intuition is already a predictive model. When you see a patient with 12 prescriptions, 5 prescribers, a new depression diagnosis, and a high copay, your brain screams “HIGH RISK!” A predictive model simply formalizes, quantifies, and scales that intuition.

Here is the (simplified) 4-step process:

1. 1. Define the Outcome (The “Target”): You must ask a specific question. What, exactly, do you want to predict?
   • Example: “Hospitalization in the next 90 days.”
   • Example: “Becoming non-adherent (PDC < 80%) in the next 6 months."
   • Example: “Discontinuation of a new specialty drug in the first 30 days.”
2. 2. Gather Features (The “Ingredients”): These are the RWD variables you will “feed” the model. The more data sources you have, the more powerful your model.
   • Claims Data Features: Past PDC, # of prescribers, # of pharmacies, # of medications (polypharmacy), # of ER visits in past 6 months.
   • EHR Data Features: Specific diagnoses (e.g., ‘Depression’, ‘CHF’, ‘Substance Use’), last HbA1c value, BMI, smoking status.
   • SDoH Data Features: Zip code-derived income level, transportation access.
   • Pharmacy Data Features: Patient’s preferred language, use of auto-refill, use of med-sync.
3. 3. Build (Train) the Model: A data scientist “trains” a statistical model (like a logistic regression) on 100,000 past patients. The model “learns” the mathematical relationship between each feature and the outcome. It creates an algorithm.
   (Example Algorithm): `Risk Score = [0.3 * (Past PDC)] + [0.5 * (# of Prescribers)] + [0.8 * (Depression Dx)] – [0.2 * (Auto-Refill)] …`
4. 4. Deploy & Score: The model is now “live.” It runs on your current patient population and assigns every single patient a “risk score” (e.g., 0-100). You can now sort your entire population from highest risk to lowest risk.

Putting the Model into Practice: Targeted Interventions

This risk list is now your new workflow. Instead of just working your refill queue, you are working a “risk queue.” This allows you to allocate your most expensive resource—your time—effectively.

Masterclass Table: Risk-Stratified Intervention Model

Risk Tier	Risk Score	Example Patient Profile (Predicting Non-Adherence)	The Proactive, Targeted Intervention
High Risk	71 – 100	45 y/o, new to specialty drug (e.g., Humira), 10+ total meds, 4 prescribers, diagnosis of depression, past PDC of 65%, high copay.	High-Touch, Pharmacist-Led: Proactive call before first fill, benefits investigation, copay card enrollment, side-effect counseling at day 7, monthly adherence check-in calls.
Medium Risk	31 – 70	55 y/o, stable on 3 meds, but has 2 prescribers and PDC just dropped to 82%. Model flags them as “at risk” of falling.	Tech-Enabled: Targeted adherence call from a pharmacy technician, offer enrollment in med-sync program, send automated text reminders.
Low Risk	0 – 30	65 y/o, stable on lisinopril for 10 years, 1 prescriber, PDC of 98%, on auto-refill.	Standard/Automated Care: No manual intervention needed. Let the automated systems (auto-refill, text alerts) handle this patient.

12.1.6 Application 3: Outcomes Reporting for Payers and Manufacturers

You have used RWD to build KPIs (measuring the past) and predictive models (predicting the future). Now, you must complete the cycle by packaging your results to prove your value to external stakeholders. This is how you get paid for your clinical services. This is how you move from a “cost center” (the cost of dispensing) to a “value center” (a generator of savings and quality).

Your Skill Translation: This is no different from preparing your meticulous records for a State Board of Pharmacy inspection, a PBM audit, or a DEA audit. In those cases, you are gathering, organizing, and presenting specific data (dispensing logs, invoices) to prove compliance. For outcomes reporting, you are gathering, organizing, and presenting specific data (adherence rates, clinical notes) to prove value.

Speaking Their Language: What Each Stakeholder Wants

You cannot send the same report to a payer and a manufacturer. They have completely different goals, and you must tailor your message to what they care about.

1. Reporting to Payers (Insurers, PBMs, Health Plans)

• What They Care About: Two things: Total Cost of Care (Medical + Rx) and Quality Metrics (HEDIS / Star Ratings). They are asking: “Did your pharmacy service save me money by preventing ER visits, and did you help me hit my 5-Star quality bonuses?”
• Data You Need: Claims data is king here. You need pharmacy claims (to calculate PDC) and medical claims (to see hospital/ER costs). You also need your own pharmacy data to prove it was your intervention that caused the change.
• Your Report Should Look Like This:

  Example Payer Outcomes Report (Executive Summary)

  “For Q4, we enrolled 150 of your highest-risk diabetes patients (identified by our predictive model) into our Pharmacist-Led Diabetes Management Program. We compared their outcomes to a matched control group of 150 similar patients.”

  • Clinical Quality: Our intervention group’s average PDC for statins increased from 78% to 91% (a 5-Star measure), while the control group’s PDC remained flat at 79%.
  • Economic Value: Our intervention group had a 22% reduction in diabetes-related ER visits and a 15% reduction in all-cause hospitalizations compared to the control group.
  • The ROI: This resulted in a net medical cost savings of $28,500. The cost of our pharmacist service was $15,000. This represents a 1.9-to-1 Return on Investment (ROI) for your health plan.“

2. Reporting to Manufacturers (Pharma)

• What They Care About: One thing: Adherence & Persistency for their specific, brand-name drug. They are launching a new, expensive specialty drug and are terrified of “patient drop-off” in the first 90 days due to cost or side effects.
• Data You Need: Your internal pharmacy dispensing data (for their specific NDCs) and your clinical intervention notes.
• Your Report Should Look Like This:

  Example Manufacturer Outcomes Report (Executive Summary)

  “For Q4, 85 of your patients were prescribed your new biologic, [Drug Name]. We enrolled 100% of them in our ‘New-to-Therapy (NTT)’ Onboarding Program.”

  • Access & Onboarding: Our benefits investigation team successfully resolved 32 prior authorization and/or affordability issues, preventing access delays.
  • Persistency: Our 90-day discontinuation rate for [Drug Name] was only 8%, compared to the 25% national average RWD benchmark for this drug class.
  • Intervention Data: Our pharmacists conducted 112 proactive counseling calls, with 65% of interactions focused on side effect management (e.g., injection site reactions), which is the primary driver of non-persistence.

12.1.7 Application 4: Data Visualization and Benchmarking

You have done the analysis and calculated the ROI. Your final task is to present your findings. A 10,000-row Excel spreadsheet is not insightful; it is a headache. A clear, simple graph is a powerful tool of persuasion. Data visualization is the art and science of communicating complex information clearly, effectively, and persuasively.

Your Skill Translation: You are already a data visualization expert. When you draw a simple graph for a patient showing their HbA1c trend (9.0% -> 8.2% -> 7.5%) to motivate them, you are using data visualization. When you show a prescriber a patient’s BP log, you are using a table (a form of data viz) to make a clinical point. This is the same skill, just applied to a different audience.

Choosing the Right Chart for the Right Question

The #1 mistake in data visualization is choosing the wrong type of chart. The chart you choose depends on the question you are trying to answer.

Masterclass Table: The Pharmacist’s Data Viz Playbook

Chart Type	Icon	What It’s Used For	Pharmacy Example (The Question It Answers)
Line Chart		Showing a trend over time.	“How has our adherence rate for DOACs changed over the last 12 months?”
Bar Chart (Vertical)		Comparing categories or groups.	“What is the average turnaround time for each of our 5 pharmacy locations?”
Bar Chart (Horizontal)		Comparing categories with long labels.	“What are the Top 10 most common clinical interventions our pharmacists perform?”
Pie Chart / Donut Chart		Showing parts of a whole (sum = 100%). Use sparingly.	“What percentage of our new prescriptions come from each local medical group?”
Histogram		Showing the distribution of a single variable.	“What is the distribution of HbA1c values for our 500 diabetes patients? Are they clustered around 7.0 or 9.0?”
Scatter Plot		Showing the relationship between two different variables.	“Is there a relationship between a patient’s age (X-axis) and their number of missed refills (Y-axis)?”
Heat Map		Showing intensity/density in a grid or map.	“Which zip codes in our delivery area have the highest density of non-adherent patients?”

Benchmarking: The Power of Context

A number without context is meaningless. If you tell your boss, “Our statin adherence rate is 88%,” their first question will be: “Is that good?”

Benchmarking is the process of comparing your performance (your KPI) against a standard to provide context and demonstrate excellence. Data visualization is how you show this comparison.

Types of Benchmarks You MUST Use:

• 1. Historical Benchmarking: Comparing yourself to… yourself.
  • Example: A line chart showing your adherence rate was 82% last year and is 88% this year.
  • The Story: “We are improving.”
• 2. Internal Benchmarking: Comparing yourself to your peers.
  • Example: A bar chart showing Pharmacy A (88%), Pharmacy B (84%), and Pharmacy C (91%).
  • The Story: “Pharmacy C is a top performer, and we should learn from their best practices.”
• 3. Industry / Payer Benchmarking: Comparing yourself to the “gold standard.”
  • Example: A bar chart showing Your Pharmacy (88%) next to the “CMS 5-Star Goal” (90%).
  • The Story: “We are a high-performing pharmacy, but we are 2 points away from the top-tier quality bonus.”