The short answer: prescription data entry is not a fixed six-field form. It is the work of turning an incoming order into the correct patient record, medication and package, directions, quantity and days supply, refill authority, prescriber record, and billable claim. The exact data varies by prescription type, state, payer, and pharmacy system. The sections below are recurring work areas, not an exhaustive schema.
From outside the pharmacy, "data entry" sounds like typing. Inside it, it's the step where an order is matched to the right patient, interpreted, made legally and clinically usable, connected to the product on the shelf, and prepared for billing and pharmacist review. A pharmacy management system may expose dozens of data elements, prompts, notes, and workflow states around that work. Even the federal rule for controlled-substance prescriptions lists patient, drug, strength, dosage form, quantity, directions, prescriber, date, signature, and registration information rather than a six-box form. The practical map below groups the work into recurring decision areas without pretending every prescription or PMS looks the same.
The same breadth appears in the 2026 ASHP pharmacy technician model curriculum, which treats order entry and processing as a combination of patient and prescriber information, calculations, third-party coverage, claim handling, documentation, prioritization, safety checks, and preparation for pharmacist verification.
What is the difference between a prescription, a pharmacy record, and a claim?
The prescription is the prescriber's order. The pharmacy record is the operational record created as staff interpret that order, select the dispensed product, calculate dispensing values, document clarifications, and prepare it for verification. The claim is a separate billing transaction sent to the payer. The three are linked, but they are not interchangeable and they do not contain the same data.
What patient and prescription information must be matched first?
Entry starts by matching the order to the correct patient and preserving its source context. Staff may need to confirm the patient's name, date of birth, address, allergies, existing profile, prescription date, origin, priority, and any notes that change how the order should be interpreted. These are separate data elements and checks, not one "patient" field.
Requirements vary by prescription type and jurisdiction. For controlled substances, for example, 21 CFR 1306.05 requires the patient's full name and address, the issue date and signature, medication details, directions, and practitioner information. Electronic prescriptions can carry still more context, including notes that must be reconciled with the structured directions rather than ignored.
What does the drug field need to capture?
The source prescription identifies the medication, strength, and dosage form. The pharmacy then selects the product and package it will actually dispense, including the specific NDC, subject to substitution rules, stock, and payer requirements. Pick the wrong one and the claim rejects; worse, it pays and it's wrong.
The NDC (National Drug Code) is why this field is trickier than it looks. It's a three-segment code: a labeler code assigned by the FDA to the manufacturer or repackager, then a product code and a package code the labeler assigns itself. The product segment pins down strength, dosage form, and formulation; the package segment pins down the bottle. That's why one molecule has dozens of NDCs: lisinopril 10 mg exists once as a drug and many times as an NDC, one per manufacturer per package size. FDA publishes the code in 10 digits (4-4-2, 5-3-2, or 5-4-1); claims bill in 11 (5-4-2), with a leading zero padded into whichever segment is short. Tools handle the padding, but the tech still has to pick the NDC that matches what's actually on the shelf. Bill the old bottle's NDC after the wholesaler switches you to a different generic maker, or bill a code the labeler has since retired, and the claim comes back rejected for an invalid or obsolete product code.
Riding along with the drug is the DAW (dispense as written) code, which tells the payer who decided between brand and generic. The three an owner sees constantly: DAW 0, no product selection indicated (the default, generic fine); DAW 1, substitution not allowed by the prescriber (brand medically necessary, and the plan prices it accordingly); DAW 2, substitution allowed but the patient requested the brand (which often shifts the brand-generic price difference onto the patient). Enter DAW 1 when the prescriber never wrote it and you've misbilled; audits look for exactly that.
How should a pharmacy enter SIG directions?
A pharmacy should translate SIG shorthand into a clean, unambiguous instruction on the label. Directions may arrive as 1 tab po qd, ii gtts ou bid, or 1 tab qhs prn. Most are routine. The ones that aren't are exactly the ones you don't want a rushed tech guessing on.
Here's a real shape of the problem. A prednisone script arrives with the sig 1 tab po qod: one tablet by mouth every other day. Handwritten as q.o.d., the periods and a cramped "o" are an established misread for q.i.d., four times daily. One reading is half a tablet a day on average; the other is four. Same ink, eight-fold difference in steroid exposure, and the label prints whichever one the tech typed. This is why ISMP's List of Error-Prone Abbreviations, Symbols, and Dose Designations exists: QD and QOD are on it, "U" for units is on it (10 U read as 100 units, a ten-fold insulin overdose), trailing zeros are on it (5.0 mg read as 50 mg). The best-known catastrophic version is weekly methotrexate entered as daily.
The entry task isn't just decoding; it's committing. A SIG that could go two ways should stop the line and generate a prescriber call, not a best guess. The difference between a good data-entry operation and a fast one is which ambiguities it refuses to resolve on its own.
How are quantity and days supply calculated?
Quantity and days supply are calculated together from the SIG, the package size, and the plan's rules. In our experience watching entry queues, this pair generates more rework than any other field. Not because the math is hard, but because the three numbers (quantity, days supply, and the SIG) have to agree, and plenty of drugs don't come in units that divide evenly.
The hard cases are the ones every tech knows:
- Insulin. Prescribed in units, packaged in milliliters, billed by the milliliter. U-100 insulin is 100 units per mL, so a box of five 3 mL pens holds 1,500 units. At 40 units a day that's 37.5 days. The pharmacy still has to apply the payer's requirements and its documented procedure when an integer days-supply value is required. Enter 30 out of habit and the refill-too-soon clock runs early; enter the quantity as "5" (pens) instead of "15" (mL) and the claim prices absurdly or rejects.
- Eye drops. A 2.5 mL bottle of latanoprost, one drop each eye at bedtime. Two drops a day, but how many drops are in a bottle? The common working convention is about 20 drops per mL, which makes that bottle roughly a 25-day supply, and plans differ on the assumption they'll accept. Enter 30 and some plans reject the math outright.
- Inhalers. An albuterol inhaler with 200 metered actuations, two puffs every four to six hours as needed. Days supply on a PRN inhaler is calculated from maximum use (up to eight puffs a day puts a 200-puff canister at 25 days) even though most patients will stretch it far longer.
Then there's the plan overlay. Quantity limits cap how much the plan covers per fill or per month regardless of what the prescriber wrote. And the 90-day script is its own trap: the prescriber writes 90 with three refills, the plan covers only 30 at retail, and the claim comes back rejected (plan limitations exceeded). The pharmacy must determine whether a covered partial quantity is appropriate, whether clarification or authorization is needed, and how to preserve the authorized total under applicable law and procedure. Getting the pair right the first time is the difference between a clean fill and a reject-fix-resubmit cycle that ties up a tech and delays the patient.
How many refills can a prescription have?
The allowed refill count depends on the drug's schedule, the prescription date, state law, and what the prescriber authorized. The federal floor is set by the DEA's rules in 21 CFR part 1306. Schedule II: no refills, ever. 21 CFR 1306.12 prohibits them outright, so a C-II "refill" means a new prescription. Schedules III and IV: 21 CFR 1306.22 caps them at five refills, and none of them (first fill included) more than six months after the written date. Both limits bind: five refills authorized in month one still die at month six. Schedule V prescriptions aren't held to the five-and-six rule federally, but states often tighten all of this (shorter windows, quantity caps on controls, one-year expiration on everything else) so the count you enter is the stricter of federal, state, and what the prescriber actually wrote.
The judgment calls hide in the soft cases: "refill × 1 year" on a Schedule IV prescription cannot be dispensed beyond the applicable limits and may require clarification. A PRN refill line still expires. An entered refill count that exceeds the legal maximum is the kind of thing that looks fine for months and then becomes an audit finding.
Which prescriber identifiers does a pharmacy need?
A pharmacy needs the right provider matched to the identifiers required for that script. Two numbers do different jobs here. The NPI (National Provider Identifier) is the 10-digit number CMS issues to identify a health care provider in electronic transactions. The DEA registration (two letters and seven digits, with a checksum built into the last digit) is what authorizes controlled-substance prescribing, and it only belongs on claims that need it.
The wrinkles are organizational. A clinic's e-scripts may arrive under the practice's group NPI when the payer wants the individual prescriber's. Nurse practitioners and physician assistants prescribe under their own NPIs, but in some states and for some payers a controlled script needs the supervising physician's information attached too, and an e-script doesn't always carry it. A prescriber who has moved practices may transmit under a stale address or an expired DEA registration.
Mismatches here are the kind of thing a board auditor circles: a C-II billed under a DEA number that doesn't validate, an NPI that belongs to the practice rather than the prescriber, a mid-level's script missing information the state or payer requires. Any of these issues can stop or delay dispensing, particularly for a controlled substance, and unresolved mismatches can also surface later in an audit.
How do insurance and claim data fit into prescription entry?
Insurance is not a field on the prescription itself. It belongs to the patient profile and the separate pharmacy claim transaction that follows prescription entry. The BIN, PCN, and Group fields route a pharmacy claim, while the member ID and person code identify the covered patient. The BIN (a six-digit bank identification number, formally the IIN) gets it to the right processor; the PCN (processor control number) picks the right plan or line of business inside that processor; the Group narrows it to the employer or benefit group. Add the member ID and a person code (typically 01 for the cardholder, 02 for a spouse, higher for dependents) and the claim finds its plan. Any one of them wrong and it doesn't, which is why a new insurance card is a data-entry event, not a filing one.
Then the plan answers. Rejects come back as NCPDP codes: the National Council for Prescription Drug Programs maintains the standard list, gated behind membership, but state Medicaid programs publish theirs, and a handful account for most of what an owner sees in a given week:
| Code | Reject | What it usually means at the counter |
|---|---|---|
| 75 | Prior Authorization Required | The plan wants the prescriber to justify the drug before it pays. Starts a fax-and-wait cycle measured in days. |
| 79 | Refill Too Soon | Not enough of the last fill's days supply has elapsed by the plan's threshold. Often traces back to a wrong days supply on the previous entry. |
| 76 | Plan Limitations Exceeded | Quantity limit, days-supply cap, or the 90-day-at-retail problem. Fixed by re-entering to what the plan covers. |
| 70 | Product/Service Not Covered | This NDC isn't on the formulary: sometimes the drug, sometimes just this manufacturer's version of it. |
| 65 | Patient Is Not Covered | Termed coverage or a wrong member ID. The card in the file is no longer the card that pays. |
CMS reflects this separation in the standards it adopts: NCPDP SCRIPT carries prescriptions and prescription-related messages, while the NCPDP Telecommunication Standard D.0 governs retail pharmacy claims. A paid response confirms that the billing transaction passed the payer's rules. It does not prove that every clinical or prescription data element was entered correctly.
Where do prescription entry errors get caught?
Prescription entry errors can be caught during claim adjudication, drug utilization review, or pharmacist verification, but no single layer catches every error. A paid claim is not proof of a correct entry because each layer catches a different slice.
Adjudication catches billing-relevant errors: dead NDCs, quantities that violate plan limits, days supply that trips a refill clock. It's blind to everything else: a wrong SIG bills identically to a right one.
DUR (drug utilization review, both the PMS's own screening and the payer's) catches interaction and dose-range problems: the new script against the patient's profile, doses outside plausible bounds. It catches a decimal error that produces an absurd dose; it doesn't catch one that produces a plausible wrong dose.
Pharmacist verification is the layer that actually looks: entered fields against the original image, product against label. It's also the layer staffed by the busiest person in the building, checking work at the rate the queue demands.
What slips through all three is well studied. The largest direct-observation study of US community pharmacy dispensing, Flynn, Barker, and Carnahan, across 50 pharmacies in six cities, found an overall accuracy rate of 98.3%: 77 errors among 4,481 prescriptions, about four errors a day for a pharmacy filling 250 scripts, with 6.5% of those errors judged clinically important. The authors' national extrapolation was roughly 51.5 million errors a year. The study counted dispensing errors of every cause, not entry mistakes alone, but data entry can introduce wrong label directions and wrong-product selections that later checks need to catch.
That's the honest case for checking entry twice.
Why is prescription data entry a bottleneck?
Prescription data entry becomes a bottleneck because every downstream step waits for it, and the work repeats across the entire queue. A single prescription can be complex; even routine decisions become relentless across a hundred. A store that clears its data-entry queue by lunch has a good afternoon; one that doesn't spends the day behind. That's why it's the first thing worth automating. Automating it well means checking each data element and relationship a second time and flagging what's genuinely ambiguous, instead of committing a guess.
That's what PillPilot's agents do inside your system: complete the routine entry work, verify what was entered, run the claim, and park anything uncertain on a flag with the reason attached: the qod that could be qid, the days supply that won't survive adjudication. See how it works, read why we run on top of your PMS instead of integrating with it, or, if the reject table above is your daily reality, see where we're headed on claims and denials.
Sources
- 21 CFR 1306.05: Manner of issuance of controlled-substance prescriptions
- CMS: E-Prescribing Standards and Requirements
- CMS: NCPDP D.0 standard for retail pharmacy claims
- ASHP Pharmacy Technician Accreditation Model Curriculum (2026)
- 21 CFR 1306.22: Refilling of prescriptions (eCFR)
- 21 CFR 1306.12: Refilling prescriptions; issuance of multiple prescriptions (Cornell LII)
- FDA National Drug Code Directory: NDC format and segments (openFDA)
- ISMP List of Error-Prone Abbreviations, Symbols, and Dose Designations
- NCPDP reject codes: Medi-Cal Rx Appendix D (published subset of the NCPDP standard)
- DAW/Product Selection Code definitions: eMedNY (NCPDP field 408-D8)
- Flynn EA, Barker KN, Carnahan BJ. National observational study of prescription dispensing accuracy and safety in 50 pharmacies. J Am Pharm Assoc. 2003.