CPAP Module 18, Section 5: Data Validation and Error Prevention
MODULE 18: WORKING WITH EMRS & EHRS

Section 5: Data Validation and Error Prevention

Developing Systematic Workflows for Verifying EHR Data Accuracy and Preventing Common, Costly Errors Before Submission.

SECTION 18.5

Data Validation and Error Prevention: The Specialist’s Oath

“Trust, but Verify” as a Core Professional Mandate.

18.5.1 The “Why”: The EHR Is Not an Infallible Source of Truth

Throughout this module, we have treated the EHR as a vast library—a repository of the clinical evidence required for your work. We have learned how to search it, navigate it, and automate documentation within it. Now, we must confront the final, most critical lesson: the library is full of errors. The Electronic Health Record is a dynamic, complex, and deeply human document. It is subject to data entry mistakes, outdated information, copy-and-paste errors, and the simple chaos of a busy clinical environment. Believing that every piece of data in the chart is accurate and up-to-date is the single most dangerous assumption a PA specialist can make.

Submitting a prior authorization based on flawed data is not a neutral act. It has profound negative consequences. At best, it leads to a preventable denial, wasting hours of your time and delaying patient care. At worst, it can perpetuate a clinical error that compromises patient safety. An outdated allergy listing can lead to a harmful prescription. An inaccurate problem list can lead to a misdiagnosis being carried forward for years. An erroneous medication history can lead to dangerous therapeutic duplications.

Therefore, the final layer of your expertise is to cultivate a profound sense of professional skepticism. You must evolve from a simple data extractor to a meticulous data validator. Your role is not just to find the information, but to critically appraise it, to question its provenance, and to systematically verify its accuracy before you incorporate it into your clinical narrative. This section will equip you with the workflows and mental models necessary to become the ultimate safety net for your providers and patients. You will learn to identify the highest-risk data domains—the problem list, the medication history, and the allergy list—and apply systematic verification techniques to each. This is not about adding more work; it’s about performing the right work at the right time to prevent hours of rework and potential harm down the line. It is the specialist’s oath: I will not only find the data, I will ensure the data is true.

Retail Pharmacist Analogy: The “DUR Rejection – Refill Too Soon” Investigation

A patient’s prescription for atorvastatin rejects with a common message: “Refill Too Soon.” The computer screen shows a days’ supply of 30 and a last fill date of 20 days ago.

The Novice Technician’s Approach (Trusting the Data): The technician sees the data on their screen and takes it as absolute truth. They tell the patient, “Sorry, your insurance won’t cover this for another 10 days. It’s too soon.” The patient becomes frustrated, insisting they are out of medication.

The Expert Pharmacist’s Approach (Data Validation): You see the rejection, but you don’t trust the initial data. You immediately begin a systematic validation workflow because you know the data can be wrong.

  • Hypothesis 1: Dose Change. You ask the patient, “Has your doctor recently changed the dose of your atorvastatin?” The patient replies, “Yes, he doubled it last week.”
  • Verification Step 1 (Internal Data): You look at the prescription image or the hard copy. The previous fill was for 40 mg daily. Today’s script is for 80 mg daily. The data entry technician incorrectly processed it as a simple refill instead of a new prescription with a dose change. The computer’s “days’ supply” calculation was based on the old dose and is therefore wrong.
  • Verification Step 2 (External Data): You call the insurance company’s pharmacy help desk. You don’t just ask them to override it. You state, “I’m calling about a refill-too-soon rejection for John Doe’s atorvastatin. I am validating that the patient had a dosage increase from 40 mg to 80 mg on October 10th. Can you please process a dose-change override?”
  • Correction: You reverse and re-bill the prescription correctly, applying the override code. The claim pays instantly.

You did not simply accept the EHR’s (in this case, the pharmacy software’s) initial data. You questioned it, formed a hypothesis, and used a workflow to validate it against the source document (the hard copy) and an external authority (the payer). This is the exact process of data validation for a PA. You must look at the Problem List, Medication History, and Allergy List and constantly ask, “Is this data still correct? What is the original source? Does it make clinical sense?”

18.5.2 The Three Red Zones: High-Risk Areas for EHR Data Corruption

While any data in the EHR can be incorrect, your validation efforts must be focused on the areas with the highest potential for error and the greatest impact on your PA submissions and patient safety. These are the “Red Zones.”

Red Zone 1: The Problem List

Often cluttered with outdated, inactive, or inaccurately coded diagnoses.

Red Zone 2: The Medication History

Riddled with duplications, omissions, and a lack of context for discontinued drugs.

Red Zone 3: The Allergy List

Frequently contains vague entries, confuses intolerances with true allergies, and lacks crucial reaction details.

18.5.3 Masterclass Workflow: Validating the Problem List

The Problem List is the foundation of your PA. An incorrect or non-specific diagnosis code is often the sole reason for a denial. Your job is to ensure the diagnosis you use is active, specific, and, most importantly, supported by recent clinical documentation.

Common Problem List Errors and Their Consequences
Common Error Example PA-Related Consequence
Lack of Specificity The list says “Diabetes” (E11.9) but the PA for Jardiance requires a diagnosis of “Type 2 Diabetes Mellitus with established cardiovascular disease” (E11.51). The claim is denied because the submitted diagnosis code does not meet the payer’s specific criteria for medical necessity.
Outdated/Inactive Problems A patient’s problem list still shows “Active – Atrial Fibrillation” from an episode of post-op Afib five years ago, but they have been in sinus rhythm ever since. You submit a PA for Eliquis based on this outdated diagnosis. The payer reviews the latest cardiology note, sees no mention of current Afib, and denies the request.
“Rule-Out” Diagnoses A provider adds “Multiple Sclerosis” to the problem list while the patient is undergoing a diagnostic workup. The final diagnosis is actually fibromyalgia. The incorrect MS diagnosis remains on the list. A year later, a new provider sees it and orders a high-cost MS medication, triggering a PA you submit based on the wrong diagnosis.
The Problem List Validation Workflow

Before using any diagnosis from the problem list in a PA, perform this rapid three-step verification:

  1. Step 1: Find the Diagnosis on the List. Locate the specific diagnosis that corresponds to the drug’s indication. Pay close attention to the ICD-10 code.
  2. Step 2: Find the Most Recent Specialist Note. Navigate to the most recent progress note from the relevant specialist (e.g., Cardiology, Oncology).
  3. Step 3: Corroborate in the “Assessment and Plan.” Read the “Assessment and Plan” at the end of that note. Does the provider explicitly list the same, specific diagnosis? If the problem list says “Heart Failure, unspecified” but the cardiologist’s A&P says “HFrEF, EF 30%,” you must use the more specific, documented diagnosis in your PA. If the diagnosis isn’t mentioned at all in the most recent note, that is a major red flag that the problem may no longer be active.

The Golden Rule: The clinical narrative in the most recent specialist note is the source of truth. The problem list is merely a summary, and it can be wrong. Always trust the note over the list.

18.5.4 Masterclass Workflow: Validating the Medication History

The medication history is the core of any step-therapy review. An incomplete or inaccurate history makes it impossible to prove formulary failure. Your validation task is twofold: to clean up erroneous data and to enrich the history with the critical context that the EHR often lacks.

Common Medication History Errors and Their Consequences
Common Error Example PA-Related Consequence
Omission of Past Trials A patient tried and failed methotrexate for RA at an outside facility, but it was never added to the “discontinued” list in your EHR. You submit a PA for Humira. The payer’s system sees no record of a prior DMARD trial and instantly denies it for “step therapy required.”
Lack of Discontinuation Reason The EHR shows the patient was on metformin and it was discontinued, but there is no documented reason why. You cannot prove whether the patient failed metformin due to intolerance (a valid reason) or simply stopped taking it (not a valid reason). The payer denies for insufficient information.
“Ghost” Duplicates The list shows active prescriptions for both lisinopril 10mg (from a PCP) and losartan 50mg (from a cardiologist) because the lisinopril was never formally discontinued. This creates clinical confusion and can lead to a payer denying a new agent due to apparent, though not actual, therapeutic duplication.
The Medication History Reconciliation & Enrichment Workflow

For any PA requiring step therapy, you must become a medication history detective. Your goal is to build a complete and defensible story of the patient’s treatment journey.

  1. Step 1: Review the EHR’s Discontinued List. This is your starting point. Identify any of the required formulary alternatives that appear on this list.
  2. Step 2: Cross-Reference with External Fill History. Use integrated tools like Surescripts or state PDMPs to see a more complete picture. You may discover fills for required drugs from years ago that are not documented in your institution’s EHR.
  3. Step 3: Hunt for the “Why.” This is the most critical step. For each discontinued formulary drug, you must find the reason it was stopped. Use your advanced EHR search skills ("metformin" W/10 (intolerance OR nausea OR diarrhea)) to find the note where the discontinuation was discussed.
  4. Step 4: Call for Confirmation. If you cannot find the reason in the chart, you must pick up the phone. Call the patient’s retail pharmacy (“Can you check your records for why lisinopril was stopped in May?”) or the previous provider’s office.
  5. Step 5: Document Your Findings. Create a “Medication Reconciliation” or “PA Support” note in the patient’s chart. In it, clearly list each required drug, the dates of trial, the reason for failure, and a citation for your evidence (e.g., “Per Dr. Johnson’s note on 6/1/2025,” “Per phone call with CVS Pharmacy on 10/15/2025”). This creates a permanent, validated record for all future PAs.

18.5.5 Masterclass Workflow: Validating the Allergy List

The allergy list is a critical safety document that is frequently corrupted with low-quality data. For a PA specialist, it can be both a source of justification (proving a patient cannot take a formulary drug due to an allergy) and a source of denials (if the documented “allergy” is not clinically valid). Your job is to be the ultimate arbiter of the allergy list’s accuracy.

Common Allergy List Errors and Their Consequences
Common Error Example PA-Related Consequence
Intolerance vs. Allergy A patient has “Codeine” on their allergy list with the reaction listed as “makes me sleepy.” You submit a PA for a non-formulary pain medication, citing a codeine allergy. The clinical reviewer sees the reaction is a known side effect, not a true allergy, and denies the PA.
Vague or Missing Reaction The list shows “Lisinopril” with the reaction field blank or listed as “unknown.” This is useless for a PA. A reaction of “hives/angioedema” would justify a switch to an ARB. A reaction of “cough” might also. A blank field provides no justification at all.
Incorrect Drug Class Labeling A patient reports a rash to amoxicillin. It is entered as a “Penicillin Allergy.” The patient may or may not be truly allergic to the entire beta-lactam class. A payer may deny a formulary cephalosporin, citing the penicillin class allergy. A detailed history might clarify the patient has since tolerated cephalosporins, allowing you to use a cheaper formulary option.
The Allergy Validation & Clarification Workflow

Never take an allergy entry at face value. Before using an allergy to justify a PA, you must clarify and document it with precision.

  1. Step 1: Review the Documented Reaction. Look at the specifics. Is the reaction a known side effect/intolerance (nausea, headache, drowsiness) or a sign of a true IgE-mediated or severe reaction (hives, rash, angioedema, shortness of breath, SJS)?
  2. Step 2: Interview the Patient (if possible). A direct conversation is the best source of truth. Ask open-ended questions: “I see you have an allergy to lisinopril listed. Can you tell me exactly what happened when you took it?” “When did this happen?” “Did you have to go to the hospital?”
  3. Step 3: Search the Chart for the Original Report. Look for the ED note or office visit note from when the reaction was first documented. This often contains a much more detailed description of the event.
  4. Step 4: Update the Allergy Entry with Specifics. Your most important task is to improve the quality of the EHR data. Work with the provider or use your own documentation protocols to update the allergy entry. Change “Lisinopril – Unknown” to “Lisinopril – Dry Cough.” Change “Codeine – Allergy” to “Codeine – Intolerance (drowsiness).” This clarifies the record for everyone and strengthens your future PA submissions.

Your goal is to transform every vague allergy entry into a precise clinical statement that can be confidently used for both clinical decision-making and payer justification.