CPOM Module 17, Section 1: Role of Data and Informatics in Pharmacy Operations
MODULE 17: DATA ANALYTICS, INFORMATICS & DECISION SUPPORT

Section 1: Role of Data and Informatics in Pharmacy Operations

The Digital Skeleton of Modern Pharmacy Practice: Understanding the Systems That Drive Every Dose.

SECTION 17.1

Role of Data and Informatics in Pharmacy Operations

From the flow of data in the EHR to the pharmacy database as the engine of all medication-related activities.

17.1.1 The “Why”: From Anecdote to Evidence

As a pharmacy leader, you will constantly be asked to justify your needs, defend your budget, and demonstrate your value. In the past, it may have been enough to rely on anecdotes and professional intuition. A manager could go to a hospital leadership meeting and say, “My team is overwhelmed, we are busier than ever, and I need to hire more pharmacists.” While likely true, this statement is subjective, impossible to quantify, and ultimately, easy to dismiss.

In the modern healthcare environment, data is the official language of administration. It is the universal currency of justification. An anecdote is an opinion; a well-presented data point is a fact. This section is designed to make you fluent in that language. You will learn to transform the subjective feeling of being “busy” into an objective, undeniable business case. Instead of the statement above, you will be equipped to say:

“In the last quarter, CPOE order volume between 2 PM and 4 PM increased by 22%, driven by a new surgical service line. This has resulted in our average order verification turnaround time during that window increasing from 15 minutes to 38 minutes, a 153% rise. This delay in getting first doses to the floor is a significant patient safety concern and directly correlates with a 12% increase in calls from nursing about missing medications. My analysis indicates that an additional 0.5 pharmacist FTE dedicated to this peak period would reduce turnaround time by an estimated 45% and generate goodwill with our nursing colleagues.”

This is the fundamental shift this module will empower you to make. Pharmacy informatics is not a separate, technical discipline for IT specialists. It is the core operational infrastructure that generates the data you need to manage effectively. Understanding this ecosystem—how a medication order is born in the CPOE, travels through the EHR, is processed by your Pharmacy Information System (PIS), and is finally documented in the MAR—is no longer optional. It is the foundational skill required to lead, optimize, and advocate for your department. Without data, you are just another person with an opinion. With data, you are a strategic leader armed with evidence.

Retail Pharmacist Analogy: From Patient Profile to Department Profile

For your entire career, you have been an expert data analyst on a microscopic scale. A patient, Mrs. Jones, hands you a new prescription for Bactrim. Your brain, in a matter of seconds, performs a complex data synthesis using her patient profile as your database.

You don’t just see a list of drugs; you see an interconnected web of data points:

  • Data Point 1 (Current Med): Warfarin 5 mg daily. Immediate Flag: Potential major drug-drug interaction.
  • Data Point 2 (Lab Value): Most recent INR was 2.5 (perfectly therapeutic). Inference: Her clotting is well-controlled; introducing an interacting antibiotic is high-risk.
  • Data Point 3 (Allergy): Documented allergy to sulfa drugs (“rash”). Immediate Flag: Absolute contraindication.
  • Data Point 4 (Dispense History): She received a course of ciprofloxacin two months ago without issue. Inference: She has a safe, recent alternative for a UTI.

Based on this rapid analysis of Mrs. Jones’s “data profile,” you intervene. You call the doctor, state the contraindication and the interaction risk, and recommend ciprofloxacin as a data-supported alternative. You just performed a high-stakes data-driven intervention.

Now, let’s scale that skill up. As a manager, your “patient” is the entire pharmacy operation, and your “patient profile” is your informatics system. The thinking process is identical:

  • Data Point 1 (Inventory Data): Warfarin tablets have an inventory turnover rate of 2. Immediate Flag: This is extremely slow. We are carrying far too much stock.
  • Data Point 2 (Purchasing Data): The acquisition cost (WAC) for warfarin is low, but we are buying it from a secondary wholesaler at WAC + 5%. Inference: We are losing money on a slow-moving product due to poor purchasing strategy.
  • Data Point 3 (340B Data): Warfarin is eligible for 340B pricing when dispensed to an eligible clinic, but our 340B accumulation rate for it is only 10%. Immediate Flag: We are missing a significant savings opportunity.
  • Data Point 4 (Clinical Data): A report shows that 60% of patients on warfarin have a sub-therapeutic INR. Inference: We have a systemic clinical management problem that requires a pharmacist-led anticoagulation service.

Just as you wouldn’t blindly dispense the Bactrim, you don’t blindly accept your operational status quo. You use the data to diagnose the problems (poor inventory management, missed 340B savings, suboptimal clinical outcomes) and propose an intervention. The tools have changed from a patient profile to a data dashboard, but the analytical skill—your greatest asset—is exactly the same. You already know how to think like a data analyst; this section will give you the language and framework to apply that skill on an enterprise-wide scale.

17.1.2 The Pharmacy Informatics Ecosystem: A Deep Dive into the Flow of Data

To leverage data, you must first understand where it comes from and how it moves. The hospital’s informatics ecosystem is a complex network of interconnected systems, each with a specific purpose, that together create a digital representation of patient care. A single medication order is like a digital traveler, passing through multiple systems, each one stamping its passport and adding valuable data to its journey. As a manager, understanding this journey is critical for troubleshooting problems, identifying inefficiencies, and knowing where to look for the data you need. Let’s follow the life of a single order for “Ceftriaxone 1g IVPB daily”.

The Journey of a Medication Order

Step 1: The Order is Born – Computerized Provider Order Entry (CPOE)

A physician on the medical floor determines a patient needs ceftriaxone for pneumonia. They log into the Electronic Health Record (EHR) and place the order in the CPOE module. This is the genesis of all our data.

Critical Data Generated Here:

  • Ordering Provider: Who wrote the order? (Essential for targeted education).
  • Order Date/Time: When was the clinical decision made? (The start of our turnaround time clock).
  • Order Details (Structured Data): Drug name (selected from a list), Dose (1), Unit (gram), Route (IVPB), Frequency (daily). Using structured, pre-defined fields is VITAL for data analysis.
  • Order Comments (Unstructured Data): “Give after blood cultures are drawn.” This is important clinical information but is very difficult to analyze systematically. A key goal of informatics is to minimize reliance on unstructured free text.
Step 2: The Central Hub – The Electronic Health Record (EHR)

The order doesn’t go directly to the pharmacy. It first exists within the larger EHR (e.g., Epic, Cerner, Meditech). The EHR is the patient’s complete digital chart, and it enriches the medication order with crucial context.

Data Enrichment Occurs Here:

  • Patient Context: The EHR links the order to the patient’s allergies, height, weight, and list of other active medications.
  • Clinical Context: It also links to lab results (e.g., serum creatinine of 2.5 mg/dL), diagnoses (pneumonia), and vital signs. This is where Clinical Decision Support (CDS) alerts first fire (e.g., “Patient has documented penicillin allergy – potential cross-reactivity with ceftriaxone”).
Step 3: The Pharmacy’s Brain – The Pharmacy Information System (PIS)

The order is now routed to the pharmacy’s verification queue within the PIS. This is your department’s command center. The pharmacist reviews the order in the context of all the data from the EHR. When they click “Verify,” a cascade of new data is created.

Critical Data Generated Here:

  • Verifying Pharmacist: Who took clinical responsibility? (Essential for quality assurance and performance feedback).
  • Verification Date/Time: When was the order deemed safe and appropriate? (The end of the “Order to Verify” turnaround time metric).
  • Product Selection: The pharmacist links the generic order (“Ceftriaxone 1g”) to a specific, dispensable product from the drug database (e.g., Ceftriaxone 1g frozen piggyback, NDC 12345-678-90). This links the clinical order to its financial and inventory identity.
  • Dispensing Instructions: The PIS, based on the drug database record, generates a label and may route the dispense request to specific technology (e.g., “Send to IV Room” or “Dispense from Carousel position B-12”).
Step 4: Closing the Loop – Dispensing Technology and the MAR

The verified order is now an actionable task. The dose is prepared. Upon leaving the pharmacy, it becomes available on the nursing Medication Administration Record (MAR). The nurse scans the patient’s wristband, then scans the barcode on the IV bag.

Final Data Generated Here:

  • Dispense Date/Time: When did the dose physically leave the pharmacy? (The end of the “Verify to Dispense” turnaround time metric).
  • Administering Nurse: Who is completing the medication use process?
  • Administration Date/Time: The exact moment the medication was given. This is the ultimate confirmation of therapy and the source for countless quality and billing metrics (e.g., was the first dose of an antibiotic for sepsis given within 1 hour?).
  • BCMA (Barcode Medication Administration) Data: A successful scan confirms the “five rights” (right patient, right drug, right dose, right route, right time). A failed scan generates a “near miss” data point that is invaluable for safety analysis.

17.1.3 The Digital Heart of the Pharmacy: A Masterclass on the Formulary Database

Every system described above is critical, but the single most important data repository under the pharmacy’s direct control is the formulary database (also called the drug file or drug database). This is not simply a list of drugs the hospital uses. It is a highly complex, relational database that functions as the central engine for all medication-related activities in the entire institution. The accuracy, completeness, and integrity of this database directly impact clinical safety, operational efficiency, and financial performance. An error in a single field of a single drug record can lead to catastrophic patient harm or hundreds of thousands of dollars in lost revenue. As a manager, you must have a deep and abiding respect for the formulary database and the informaticists who maintain it.

Let’s dissect the record for a single, common medication—Lisinopril 10mg tablets—to understand the depth of data contained within.

Masterclass Table: Anatomy of a Drug Database Record (Lisinopril 10 mg Tablet)
Category Data Field Example Value Operational & Strategic Importance (The “Why it Matters”)
Identifiers Drug Name Lisinopril 10 mg Oral Tablet The primary human-readable name. Clarity and standardization (e.g., using TALLman lettering like lisiNOPRIL) are key to preventing look-alike/sound-alike errors.
Generic Product Identifier (GPI) / Generic Code Number (GCN) 27110020100110 These are hierarchical codes from third-party vendors (like Medispan or First Databank) that classify the drug by therapeutic class. This is what drives most automated clinical alerts (e.g., duplicate therapy alerts fire when two drugs in the same GPI class are ordered).
National Drug Code (NDC) 00185-0021-01 The 11-digit unique identifier for a specific manufacturer’s package. This is the cornerstone of the entire medication supply chain. It is used for purchasing, inventory management, barcode scanning, and billing. A wrong NDC means you might buy the wrong drug, fail a barcode scan, or submit a fraudulent claim.
Route of Administration Oral A structured field that ensures the medication appears on the correct MAR and allows for route-based alerts (e.g., preventing an oral liquid from being programmed into an IV pump).
Financial Acquisition Cost (WAC/340B) WAC: $0.10/dose
340B: $0.01/dose		 The actual cost to the pharmacy. This data is the foundation of all financial reporting, budgeting, and spend analysis. Inaccurate cost data makes it impossible to manage your budget effectively.
Charge Code / Revenue Code CDM: 1234567
Rev Code: 0250		 Links the drug to the hospital’s Charge Description Master (CDM) for billing. An incorrect or missing charge code means the hospital performs the service but never gets paid. For high-cost drugs, this can result in millions of dollars in lost revenue.
Billing Unit per Dose (“J-Code Units”) 10 For many IV drugs, billing is done in specific units (e.g., “per 10 mg”). If the drug record isn’t configured to translate a clinical dose into the correct number of billing units, claims will be denied. This is a massive source of lost revenue and requires meticulous attention.
Formulary Status Formulary, Preferred Designates whether the drug is approved for routine use. This field drives formulary restriction alerts and therapeutic interchange protocols, guiding prescribers to the most cost-effective options.
Logistical Dispense Location(s) Central Pharmacy Carousel, Med/Surg ADC Profile, ED ADC Matrix Drawer Tells the PIS and automation hardware where the drug is physically stored. Incorrect routing leads to workflow chaos, delays, and technician frustration.
PAR Levels Carousel: 500
Med/Surg ADC: 50		 Minimum and maximum stock quantities for each location. This data drives automated inventory replenishment. Poorly managed PAR levels lead to constant stockouts or excess inventory.
Dispense Unit Tablet Defines the base unit for dispensing (e.g., Tablet, mL, Each). A mismatch here can lead to massive dosing errors (e.g., dispensing 10 mL instead of 10 mg of a liquid).
Clinical Safety Max Dose Alert 40 mg/day A hard-coded safety limit that will fire an alert if a provider orders a dose exceeding the established maximum. This is a critical safety net against overdose.
Associated Lab Test Serum Creatinine (SCr) Links the drug to relevant lab values, enabling advanced alerts like, “Patient’s SCr has doubled; consider dose adjusting lisinopril.”
Order Sentence / Smart Text TTake 1 tablet (10 mg) by mouth once daily. #RENALADJUST Pre-built, structured order strings that make it easy for providers to order correctly. Can include embedded logic or reminders, like a placeholder reminding them to consider renal adjustment.

17.1.4 Data, Information, Intelligence, Wisdom: The DIKW Pyramid in Pharmacy

The informatics ecosystem generates a staggering amount of raw data every second. However, data by itself is useless. It is the raw material, like sand on a beach. To be valuable, it must be processed and refined. The DIKW (Data, Information, Intelligence, Wisdom) Pyramid is a classic model that perfectly describes the process of turning raw data into actionable strategy. As a manager, your goal is to constantly push your team and your systems up this pyramid.

Level 1: Data (The Raw Facts)

This is a collection of discrete, unorganized, and unanalyzed items. It has no context. In pharmacy, this looks like a raw data extract from the PIS.

“10/17/25 09:14, Vancomycin 1.5g IVPB, Patient MRN 98765, Verifying RPh: Smith, Dispense from: IV Room”

Value: Minimal. By itself, this tells you almost nothing of strategic importance.

Level 2: Information (Data with Context)

Information is data that has been organized, grouped, and categorized. It answers the basic “who, what, when, where” questions.

“Yesterday, we dispensed 212 doses of vancomycin. 78 of those doses were for patients on the Orthopedics floor. The average dose was 1.25g.”

Value: Moderate. Now we have a basic understanding of our vancomycin utilization. We can track this information over time to see trends.

Level 3: Intelligence / Knowledge (Information put into Action)

Intelligence (or Knowledge) is the synthesis of information to identify patterns, relationships, and insights. It answers the “how” and “why” questions.

“Our vancomycin utilization on the Orthopedics floor has increased 30% month-over-month since they began performing spinal fusion surgeries. Our current dosing protocol is weight-based, but we see that 40% of initial troughs are sub-therapeutic, leading to re-dosing and extended length of stay. Furthermore, our cost for vancomycin has risen 35%, exceeding our budget.”

Value: High. We have now connected disparate pieces of information (utilization, clinical outcomes, financial impact) into a coherent story that identifies a clear problem.

Level 4: Wisdom (Intelligence Applied to Strategy)

Wisdom is the application of intelligence and experience to make sound judgments and strategic decisions. It answers the “what should we do about it?” question.

“We will propose implementing a pharmacist-driven vancomycin dosing protocol that utilizes AUC/MIC monitoring instead of trough-based dosing for orthopedic surgery patients. This evidence-based approach is projected to improve the rate of therapeutic first troughs by 50%, reduce the average length of therapy by one day, and decrease our total vancomycin spend by 15% per patient, generating an estimated annual savings of $150,000 and improving patient outcomes.”

Value: Highest. This is the ultimate goal. We have transformed a single raw data point into a fully-formed, evidence-based strategic initiative that improves clinical, financial, and operational outcomes.

17.1.5 The Power of Proactive Intervention: Clinical Decision Support (CDS)

One of the most powerful applications of a well-structured informatics ecosystem is Clinical Decision Support (CDS). CDS is any tool or process that provides clinicians (physicians, pharmacists, and nurses) with knowledge and person-specific information, intelligently filtered and presented at appropriate times, to enhance health and health care. In simpler terms, CDS is how we use the computer system to act as a vigilant, automated clinical pharmacist, scaling our expertise across the entire hospital 24/7.

Effective CDS is not about bombarding users with thousands of annoying pop-up alerts. In fact, a major challenge in informatics is “alert fatigue,” where users become desensitized and ignore warnings. The best CDS follows the “Five Rights” framework:

  • The Right Information: Evidence-based, relevant to the clinical task.
  • To the Right People: Don’t show a surgical scheduling alert to a pharmacist.
  • In the Right Format: A hard-stop alert for a deadly interaction, but just a helpful suggestion for a formulary alternative.
  • Through the Right Channel: An in-line suggestion in an order set is less intrusive than a pop-up window.
  • At the Right Time in the Workflow: Present the information when the decision is being made, not after the fact.
Masterclass Table: Types of Clinical Decision Support in Pharmacy
CDS Type How it Works Pharmacist Manager’s Role & Oversight
Allergy Alerts The system checks if a new medication order belongs to the same drug class as a documented patient allergy. This is the most basic but most important form of CDS. Ensure your allergy classes are mapped correctly in the formulary database. A miscategorized drug will not fire the alert. Periodically review overrides to see if clinicians are ignoring valid warnings.
Drug-Interaction Alerts When a new drug is ordered, the system checks all other active medications for potential interactions based on a third-party knowledge base (e.g., First Databank). Combat Alert Fatigue. Work with your informatics team to customize and tier the alerts. A deadly interaction (Warfarin + Bactrim) should be a “hard stop” requiring a reason for override. A minor interaction (Ibuprofen + Lisinopril) might be suppressed or just a passive notification. Your clinical judgment is key to making these alerts meaningful.
Dose Range Checking Each drug in the formulary has minimum and maximum dose limits defined. The system will alert the provider if they order a dose outside this safe range (e.g., ordering “Dabigatran 1500 mg” instead of “150 mg”). Regularly review and update these limits, especially for new drugs or changing guidelines. Implement weight-based and age-based dose checking for pediatrics, which is far more complex and critical. Ensure your team has a process for building and maintaining these rules.
Renal & Hepatic Dosing Guidance Advanced CDS can automatically pull the latest serum creatinine, calculate an estimated CrCl using a formula like Cockcroft-Gault $$(CrCl = \frac{(140 – Age) \times \text{Weight (kg)}}{72 \times SCr} \times 0.85 \text{ if female})$$ and then suggest a renally appropriate dose for drugs like enoxaparin or gabapentin. This is a massive safety win but requires meticulous setup. You must ensure the rules are based on current, evidence-based guidelines and are tested rigorously. An incorrect rule is more dangerous than no rule at all.
Duplicate Therapy Alerts Using the GPI or GCN codes, the system can detect when two drugs from the same therapeutic class are ordered (e.g., Lisinopril and Losartan, or Ibuprofen and Ketorolac). This is highly effective at catching errors, especially during medication reconciliation at admission. Your role is to ensure that every drug in your formulary is correctly mapped to a therapeutic class code.
Guideline-Based Order Sets This is a proactive form of CDS. Instead of waiting for a wrong order, an order set for a condition like Sepsis or ACS guides the provider to place a complete set of correct, evidence-based orders from the start. As a manager, you must champion the use and maintenance of order sets. Work with medical staff committees (e.g., P&T Committee) to develop, approve, and implement order sets for common, high-risk conditions. This is one of the most effective ways to standardize care and improve outcomes.

17.1.6 Garbage In, Garbage Out (GIGO): Your Role as the Steward of Data Integrity

Your ability to analyze data, build a business case, or even ensure basic patient safety is entirely dependent on the quality of the data in your systems. The principle of “Garbage In, Garbage Out” (GIGO) is the immutable law of informatics. If the data entered into the system is inaccurate, incomplete, or inconsistent, then any report, analysis, or alert based on that data will also be flawed, potentially leading to disastrous conclusions.

As a pharmacy operations manager, you are a primary steward of data integrity for the entire hospital. While you may not be building the database yourself, you oversee the people and processes that generate the majority of the medication-related data. Ensuring they are trained properly and that your core data (the formulary) is meticulously maintained is a non-delegable leadership responsibility.

Real-World Consequences of “Garbage In”
  • Clinical Harm: A pharmacist accidentally links the formulary record for oral Vincristine solution to the “IV” route field. The system now allows a provider to order it intravenously. A nurse, trusting the system, administers it. Result: A fatal, irreversible neurotoxic event.
  • Financial Loss: The NDC for a new, high-cost oncology drug is entered incorrectly into the formulary database. Every time the drug is dispensed, the barcode fails to scan at the bedside (causing nursing frustration and delays) and the claim submitted to the payer is automatically rejected because the NDC doesn’t match the billing code. Result: A month of administrations worth $500,000 is never reimbursed.
  • Operational Chaos: The PAR levels for propofol in the operating room ADCs are not updated to reflect an increase in surgical volume. The system doesn’t generate automated replenishment orders. Result: Anesthesiologists experience critical stockouts mid-procedure, creating patient safety risks and generating furious calls to the pharmacy.
Your Core Responsibilities for Data Integrity
  • Champion Meticulous Formulary Management: Support your informatics pharmacists. Give them the time and resources to maintain the drug database with the same level of focus and rigor as a pharmacist compounding a sterile product. It is that critical.
  • Insist on Standardization: Advocate for the use of structured data fields and discourage the use of free-text comments for critical information. Work with medical leadership to promote the use of standardized order sets over individual, ad-hoc orders.
  • Invest in Training: Ensure your staff—pharmacists and technicians—are properly trained on how to use the PIS. They must understand that selecting the correct product, entering the correct dispense quantity, and documenting interventions correctly are not just clerical tasks; they are acts of data creation that have downstream consequences.
  • Establish Auditing and Oversight: You don’t need to be the expert, but you need to ask the expert the right questions. Regularly meet with your informatics team to review key data integrity metrics. Ask questions like: “How many unmatched NDCs are in our system?” “What is our rate of barcode scan failures at the bedside?” “Can you show me a report of overridden duplicate therapy alerts?” Your attention to these details signals their importance to the entire department.