Topic 4.8 — Electronic Prescribing & Medicines Management

In health and care, medicines are the single most common intervention and one of the leading causes of avoidable harm, so how you digitize the prescribe–dispense–administer loop is a clinical-safety decision that can both remove old errors and quietly manufacture new ones.

Why this matters in health and care#

Medicines are the most frequent thing that happens to a patient. A single inpatient may have dozens of doses prescribed, dispensed and given in a day, each a small transaction with the potential to heal or to harm. Across a health system the numbers are enormous, and so is the error surface: research commissioned by the Department of Health and Social Care estimated that hundreds of millions of medication errors occur in England each year, the large majority minor but a clinically significant minority causing real harm, and some proving fatal. Getting the right drug, in the right dose, by the right route, to the right patient, at the right time is deceptively hard at scale.

The traditional paper drug chart — the bedside document on which prescribing and administration were recorded — concentrated many of these failure modes. It could be illegible, ambiguous about units or frequency, lost, duplicated, or transcribed wrongly at each handover. It offered no active checking: nothing stopped a tenfold dose, a lethal drug–drug interaction, or a prescription for something the patient was documented as allergic to. Electronic prescribing and medicines administration (ePMA) exists to close those gaps by making prescribing structured, legible, checked and traceable.

But digitizing medicines is not a purely benign act. A badly designed system can introduce its own errors — a wrong default dose, a confusing pick-list, an alert so frequent it is ignored — and can do so at scale and at speed. This is why medicines optimization sits at the intersection of clinical safety (see Topic 1.6 — Clinical Safety & Risk Management), decision support (see Topic 4.7 — Clinical Decision Support & Clinical Informatics) and the electronic patient record (see Topic 4.6 — Electronic Patient Records & Clinical Systems). The prize is large; so is the responsibility to implement it well.

Core concepts#

Electronic prescribing is the computer-based generation, transmission and recording of a prescription, replacing the handwritten order. In the hospital setting it is usually bundled with the recording of administration into ePMA (electronic prescribing and medicines administration), a system that manages the whole inpatient medicines process. ePMA is a specialized form of computerised physician order entry (CPOE) focused on medication, and it is almost always delivered as a module of the electronic patient record rather than a stand-alone product.

The defining ambition of a mature ePMA system is closed-loop medication administration: an unbroken, electronically-linked chain from prescribing, through pharmacy dispensing, to a barcode-verified administration at the bedside. At the point of giving a dose, a nurse scans the patient's wristband and the medicine's barcode, and the system confirms that this drug, dose and route are the ones prescribed for this patient now — a digital enforcement of the traditional "five rights" (right patient, right drug, right dose, right route, right time). The loop only closes when scanning is used reliably; a system capable of closed-loop that staff routinely bypass delivers a fraction of the benefit.

In primary care in England, the equivalent digital plumbing is the NHS Electronic Prescription Service (EPS), which transfers a prescription electronically from a GP or other prescriber to the patient's chosen pharmacy and on to the reimbursement agency, removing the paper FP10. Underpinning all of this is a shared vocabulary: the dictionary of medicines and devices (dm+d), the NHS standard terminology that gives every medicine and device a unique, unambiguous code so that systems, prescribers and pharmacies mean exactly the same thing. Without a common dictionary, interoperability between prescribing, dispensing and the record is guesswork (see Topic 4.3 — Interoperability & Data Standards).

Two clinical processes recur. Medicines reconciliation is the process of building an accurate, complete list of a patient's current medicines and comparing it against what is prescribed at every transfer of care — admission, transfer between wards, and discharge — because these boundaries are where medicines are most often lost, duplicated or wrongly changed. Polypharmacy — the simultaneous use of many medicines, common in older and multi-morbid patients — makes reconciliation and review both harder and more important.

The safety concepts frame everything else. A medical error involving medicines can cause an adverse drug reaction or arise from a drug interaction; reducing such harm is the core of patient safety. Crucially, technology itself can be a cause of harm — iatrogenesis induced by the system, sometimes called e-iatrogenesis — so ePMA must be evaluated for the errors it introduces as well as the errors it removes.

Best practices#

1. Treat ePMA as a clinical-safety system, not an IT project. Prescribing software changes clinical behaviour at the bedside, so govern it under your clinical risk-management regime from the first design decision to every configuration change. Apply the clinical safety standards that require a clinical safety case and a nominated Clinical Safety Officer (DCB0129 for the manufacturer, DCB0160 for you as the deploying organization), and keep a live hazard log across the whole medicines pathway (see Topic 1.6 — Clinical Safety & Risk Management). A drug-file change or a new default that would never pass a code review can kill someone; it deserves the same rigour.

2. Design for the closed loop, and measure whether it actually closes. The benefit of ePMA is realized at the bedside when barcode-verified administration is used consistently, not merely available. Instrument the system to show scan-compliance rates by ward and drug, understand why staff bypass scanning — broken scanners, unlabelled medicines, workflow that is slower than the workaround — and fix the causes rather than exhorting people to try harder. A closed loop used 60% of the time leaks harm through the 40% that is unverified.

3. Standardize on dm+d and let it do the interoperability work. Anchor prescribing, dispensing, decision support and the record to the dictionary of medicines and devices so that a medicine means the same thing everywhere it travels. This is what lets a discharge prescription flow cleanly into a GP system and a community pharmacy, and what lets decision support reason reliably about the drug in front of it. Bespoke local drug files that drift from dm+d are a slow-acting interoperability and safety hazard (see Topic 4.3 — Interoperability & Data Standards, and Topic 4.9 — Data Quality & Master Data Management).

4. Tune decision support ruthlessly against alert fatigue. Prescribing decision support — allergy, interaction, dose and duplicate warnings — is the feature most likely to save a life and the feature most likely to be switched off in a clinician's head. Configure alerts by severity so that the rare, dangerous interruption stands out and the routine noise is suppressed or made passive; measure override rates and act on them. The wider design of decision support belongs to Topic 4.7 — Clinical Decision Support & Clinical Informatics; here the discipline is to protect the signal in the medicines domain specifically.

5. Get transfers of care right, because that is where medicines break. Admission, ward transfer and discharge are the points at which a patient's medicines list is most often corrupted, so build reconciliation into the workflow rather than bolting it on. Make it easy to import the primary-care medication record on admission, to reconcile explicitly, and to send a structured, dm+d-coded discharge medicines summary that a GP and community pharmacy can consume without re-keying. Pharmacist-led reconciliation at these boundaries is one of the highest-value uses of scarce pharmacy time.

6. Give controlled drugs the extra controls the law requires. Drugs governed by the Misuse of Drugs Act 1971 and its regulations carry statutory requirements — witnessed administration, running balances, secure storage and audit — that your electronic systems must support, not undermine. Ensure the system enforces witnessing where required, maintains an accurate electronic controlled-drugs register, and supports the accountable officer's oversight and reporting duties. Digitizing controlled drugs badly can create both a safety risk and a regulatory breach (see Topic 1.5 — Regulation & Compliance Landscape).

7. Design pharmacy, robotics and stock as part of the same pathway. Pharmacy automation — dispensing robots, automated cabinets and stock systems — only pays off when integrated with prescribing and administration so that supply follows the electronic order and stock levels reflect real use. Link ePMA to the pharmacy and stock systems so that dispensing, ward-stock replenishment and expiry management are driven by data, not clipboards. Smart infusion pumps that receive the prescribed rate and dose electronically extend the closed loop to intravenous medicines (see Topic 4.4 — Connected Devices, IoT & Clinical Engineering).

8. Actively hunt for the errors your system introduces. Every ePMA deployment creates new failure modes — juxtaposition errors in pick-lists, wrong defaults, misleading displays, alerts that train people to click "override". Run post-implementation safety surveillance: review medication incidents for technology-induced causes, watch for e-iatrogenesis, and feed findings back into configuration and supplier fixes. Report device-related and system-related medication incidents through national channels so systemic flaws are caught beyond your own walls, and evaluate benefits honestly rather than assuming digitization equals safety (see Topic 6.2 — Benefits Realization & Value Management).

Questions to discuss with your team#

1. Does our medicines pathway close the loop, and where exactly does it leak? This matters because the safety case for ePMA rests on the whole chain — prescribe, dispense, barcode-verified administer — and a break anywhere lets harm through while the dashboard still shows green. Map your real pathway end to end and find the gaps: medicines that arrive unlabelled or unbarcoded, wards where scanning is skipped under pressure, intravenous drugs prepared outside the system, and transfers of care where the electronic thread is cut and paper reappears. Debate the honest scan-compliance figure ward by ward and why staff bypass verification, because the reasons are usually rational responses to bad workflow, not carelessness. Consider the trade-off between enforcing scanning rigidly and preserving the ability to act fast in an emergency. A good answer names the specific leak points, distinguishes "capable of closed loop" from "actually closed", and commits to fixing the workflow causes rather than blaming front-line staff.

2. Is our prescribing decision support saving lives or training clinicians to ignore it? Alerts are simultaneously the most protective and the most abused feature of an ePMA system, and the difference is entirely in how you tune and govern them. Look at your override rates: if clinicians dismiss the overwhelming majority of alerts, the system is generating noise that buries the rare, genuinely dangerous warning, and alarm fatigue has already set in. Discuss who owns alert configuration, how severity tiers are decided, whether pharmacists and prescribers are in the loop, and how you would even know if a critical allergy alert was being reflexively clicked through. Weigh the medico-legal and cultural pull towards leaving every possible alert on "to be safe" against the safety reality that too many alerts is itself unsafe. An honest answer accepts that maximal alerting is not maximal safety, sets a target for the ratio of actioned to overridden alerts, and treats decision-support tuning as continuous clinical work, not a one-off go-live setting.

3. Who owns medicines safety across the whole system, including the boundaries and the errors we introduce? Medicines cross organizational lines — GP to hospital to community pharmacy to care home — and cross professional lines between prescribers, pharmacists and nurses, so ownership tends to fall into the gaps. Debate who is accountable for reconciliation at each transfer of care, for the accuracy of the drug file, for controlled-drugs governance, and specifically for surfacing the errors the technology itself creates. Consider whether your incident reviews ever conclude "the system caused this" or whether e-iatrogenesis is invisible because no one is looking for it. Surface the tension between suppliers who own the software, pharmacy who own the medicines expertise, and digital teams who own the deployment, and ask where the buck actually stops. A strong answer assigns a named clinical owner for medicines safety across the pathway, includes technology-induced error explicitly in safety surveillance, and can point to changes made because of what that surveillance found.

4. What happens to medicines when the system goes down? ePMA concentrates the entire prescribing and administration record into software, so an outage — planned or unplanned, network or supplier — removes the very artefact clinicians need to give the next dose safely. Discuss your business-continuity design honestly: how staff read the current drug chart when the screen is dark, how a dose given on paper during downtime is reconciled back into the record without duplication or omission, and how long your read-only failover or printed contingency chart actually lasts before it is stale. Consider that downtime is most dangerous precisely for high-risk and time-critical medicines — insulin, anticoagulants, controlled drugs — where a missed or repeated dose harms fastest. Weigh the cost and drill discipline of a genuinely rehearsed downtime procedure against the temptation to treat resilience as the supplier's problem. Explore who decides to invoke downtime mode, how controlled-drugs witnessing and registers survive it, and how the organization catches the errors that cluster around the return to normal. A good answer shows the downtime plan has been tested with real clinical staff, not just written, and names who owns the reconciliation after recovery.

5. Does our digital medicines work across the whole pathway, or just inside our own walls? The safety and efficiency case for ePMA is easy to make for the hospital ward, but a patient's medicines journey runs through general practice, community pharmacy, care homes and domiciliary care, and the boundaries are where harm concentrates. Debate whether your structured, dm+d-coded discharge summary is actually consumed cleanly by GP and community-pharmacy systems, or whether someone downstream is re-keying it and reintroducing error. Consider the settings with the least digital capability — a care home receiving a discharged resident, a community pharmacy handling a complex regimen — and whether your design helps or simply exports the problem to them. Surface the equity dimension too: patients who lack digital access, confidence or a stable address can be disadvantaged by EPS nomination, app-based ordering or electronic repeat dispensing, and the duty to reach them does not disappear because the default went digital. An honest answer looks beyond the trust's own KPIs to the experience of the weakest link in the chain, and commits to fixing the transfer that fails a real patient rather than the one that is easiest to measure.

6. Are we designing the workflow so the safe path is the easy path, or bolting technology onto old habits? ePMA changes what clinicians physically do at the bedside and the desk, and adoption fails — or worse, generates workarounds — when the digital workflow is slower or clumsier than the paper it replaces. Discuss whether prescribing, verifying and administering have been redesigned around how clinicians actually work, or whether the system simply digitized the old form and left staff to absorb the friction. Consider the concrete failure points: a login that takes too long at a crash call, a scanning step that competes with a queue of unwell patients, a default that is convenient but wrong. Weigh the pull towards configuring for the supplier's demo path against the reality that front-line staff will optimize for getting the round done, and that every workaround is data about a design flaw. Involve prescribers, nurses and pharmacists in configuration and watch them use it under real pressure, not in training. A strong answer treats usability as a safety property, can name workflow changes made because staff struggled, and measures adoption depth rather than counting go-live as done.

In practice: a health & care example#

Brackenmoor NHS Foundation Trust runs a busy acute hospital that has used a paper drug chart for decades. After two serious incidents — a tenfold insulin overdose from a misplaced decimal point, and a missed allergy that caused an anaphylactic reaction — the board commissions an ePMA system as a module of its existing electronic patient record. The Chief Pharmacist and the Chief Clinical Information Officer co-sponsor the programme, and a Clinical Safety Officer is named from day one.

The team resists the temptation to treat go-live as the finish line. They build the drug file on dm+d rather than a bespoke local list, so that discharge prescriptions will flow cleanly to GPs and community pharmacies. They configure decision support in tiers: a small number of hard-stop alerts for the most dangerous interactions and allergies, and passive, non-interruptive prompts for lower-risk situations, agreed by a group of prescribers and pharmacists rather than left at the supplier's defaults. A live hazard log tracks risks across the whole pathway under DCB0160.

Closed-loop administration is phased in ward by ward. Early data shows scan-compliance stuck around 70% on two wards; instead of issuing a reminder, the team investigates and finds a shortage of working barcode scanners and a batch of ward-stock medicines arriving without scannable labels. Fixing the scanners and the labelling lifts compliance above 95%. Controlled drugs are brought onto an electronic register with enforced witnessing, and the accountable officer gains oversight she never had on paper. Smart pumps are linked so intravenous infusion rates are pushed from the prescription rather than keyed in by hand.

Six months in, prescribing is legible and dose errors from illegibility have effectively vanished. But the pharmacy team notices a new pattern in the incident data: several near-misses where a prescriber picked the wrong medicine from a pick-list because two drugs sat adjacently with similar names. Rather than dismiss it, they treat it as e-iatrogenesis, reconfigure the list with tall-man lettering and separation, and report the design flaw to the supplier. The lesson the board takes is the honest one: ePMA removed a large class of old errors and introduced a smaller class of new ones, and the safety gain came from actively looking for both.

Sector lenses#

Startup#

A digital-health startup building a prescribing or medicines-optimization product is on the manufacturer side of the safety line and must act like it. That means conforming to DCB0129, producing a clinical safety case and hazard log, and building on dm+d and recognized interoperability standards from the first sprint rather than retrofitting them after a pilot. The commercial reality is that no NHS trust will connect a prescribing tool to its record without evidence of clinical safety, information governance and a credible support commitment, so those are features, not overheads. A small team's edge is a focused, well-evidenced solution to a real medicines problem — reconciliation, or high-risk-drug monitoring — that plugs into the EPR others already run, not a rival to it.

Small business#

An established small provider — a GP practice, a community pharmacy, a care home or a single-site clinic — carries the same medicines-safety and controlled-drugs duties as a large trust but with a fraction of the digital capability to meet them. Here the practical task is not procuring a bespoke ePMA but using national plumbing well: EPS for electronic prescriptions, dm+d-aligned records in the practice or pharmacy system, and disciplined medicines reconciliation whenever a patient arrives from or returns to hospital. Cash and staff time are tight, there is rarely a dedicated IT or clinical-safety officer, and the supplier's default configuration tends to become the de facto policy, so it pays to know which settings matter — allergy and interaction alerts, controlled-drugs registers, repeat-dispensing rules — and to check them deliberately. The highest-value move is getting the boundary right: consuming a discharge medicines summary without re-keying, flagging high-risk drugs for review, and having a simple, rehearsed plan for when the system or the network is down. Small providers are also the weak link others rely on, so being reachable and accurate at the transfer of care protects patients well beyond the practice walls.

Enterprise#

A large NHS trust or hospital group typically implements ePMA as a module of a major EPR, and its challenge is depth of adoption across thousands of staff and dozens of specialties. The realistic goal is a genuinely closed loop on the wards where it matters most, disciplined alert governance, and reconciliation embedded at every transfer of care — not a flawless configuration everywhere on day one. Scarce pharmacist time is best aimed at the highest-risk boundaries and drugs, supported by the system rather than consumed by it. At this scale the biggest gains come from a few pathway-wide disciplines — dm+d, tiered alerts, scan compliance, controlled-drugs governance — done consistently, and from treating configuration as continuous clinical safety work.

Government#

National bodies shape the medicines environment every provider works in. NHS England and the NHS Business Services Authority maintain dm+d and run the Electronic Prescription Service, making a common medicines vocabulary and electronic transfer the default rather than something each trust negotiates alone. The MHRA regulates medicines and the devices that deliver them and runs post-market safety reporting; NICE and the Royal Pharmaceutical Society set the expectations for medicines optimization; the Specialist Pharmacy Service publishes practical guidance. Government's leverage is to make safe defaults — shared terminology, interoperable transfers of care, national safety reporting — the path of least resistance, and to hold the equity duty so that digital medicines services do not disadvantage patients who lack access or confidence (see Topic 9.1 — Working with Local & National Governments).

Common failure modes#

• The loop that never closes. A system capable of barcode-verified administration that staff routinely bypass delivers a fraction of the safety benefit while implying the full one. Measure scan compliance and fix the workflow causes of workarounds.
• Alert fatigue by design. Turning on every possible alert "to be safe" buries the rare critical warning under routine noise and trains clinicians to override reflexively. Tier alerts by severity and manage override rates as a live metric.
• A bespoke drug file that drifts from dm+d. Local medicine lists that diverge from the national dictionary quietly break interoperability and decision support and reintroduce ambiguity. Anchor to dm+d and govern changes.
• Reconciliation as an afterthought. Leaving medicines reconciliation off the workflow means the admission and discharge boundaries keep corrupting the medicines list. Build it in and resource pharmacist-led reconciliation at transfers of care.
• Ignoring e-iatrogenesis. If incident reviews never conclude "the system caused this", the technology-induced errors are invisible, not absent. Actively surveil for pick-list, default and display errors and feed them back to configuration and the supplier.
• Controlled drugs digitized carelessly. A system that fails to enforce witnessing or maintain an accurate register is both a safety risk and a regulatory breach. Design controlled-drugs handling to the statutory requirements.
• Assuming digital equals safe. Declaring benefits at go-live without measuring them lets regressions hide. Evaluate the errors removed and the errors introduced honestly (see Topic 6.2 — Benefits Realization & Value Management).

Maturity model#

Checklist#

• ePMA is governed as a clinical-safety system with a Clinical Safety Officer, a clinical safety case and a live hazard log across the medicines pathway (DCB0129/0160).
• Scan-compliance for barcode-verified administration is measured by ward and drug, and the causes of bypassing are fixed.
• Prescribing, dispensing, decision support and the record are anchored to dm+d, and local drug-file changes are governed.
• Prescribing alerts are tiered by severity, and override rates are monitored and acted on to prevent alert fatigue.
• Medicines reconciliation is embedded at admission, transfer and discharge, with structured, dm+d-coded discharge summaries.
• Controlled drugs are handled to statutory requirements, with enforced witnessing, an accurate electronic register and accountable-officer oversight.
• Pharmacy automation, ward stock and smart infusion pumps are integrated with prescribing and administration.
• Incident data is reviewed specifically for technology-induced (e-iatrogenesis) errors, and findings feed configuration and supplier fixes.
• System- and device-related medication incidents are reported through national safety channels.
• Benefits and harms of ePMA are evaluated honestly after go-live, not assumed.

Key sources#

• NHS England / NHS Business Services Authority — Dictionary of medicines and devices (dm+d) and the Electronic Prescription Service (EPS).
• Royal Pharmaceutical Society — Medicines Optimisation: helping patients make the most of medicines and professional standards for hospital pharmacy and ePMA.
• National Institute for Health and Care Excellence (NICE) — Medicines optimisation (NG5) and related quality standards.
• MHRA — medicines and medical-device regulation and the Yellow Card scheme for adverse drug reactions and device incidents.
• World Health Organization — Medication Without Harm, the third Global Patient Safety Challenge.
• Specialist Pharmacy Service (SPS) — practical guidance on ePMA, medicines reconciliation and transfers of care.
• Care Quality Commission — guidance for controlled-drugs accountable officers under the Controlled Drugs (Supervision of Management and Use) Regulations 2013.
• NHS England clinical safety standards — DCB0129 (manufacture) and DCB0160 (deployment) for health IT systems.

References#

1. Electronic prescribing — Wikipedia — https://en.wikipedia.org/wiki/Electronic_prescribing
2. Computerized physician order entry — Wikipedia — https://en.wikipedia.org/wiki/Computerized_physician_order_entry
3. Medication — Wikipedia — https://en.wikipedia.org/wiki/Medication
4. Electronic health record — Wikipedia — https://en.wikipedia.org/wiki/Electronic_health_record
5. Barcode — Wikipedia — https://en.wikipedia.org/wiki/Barcode
6. NHS Electronic Prescription Service — Wikipedia — https://en.wikipedia.org/wiki/NHS_Electronic_Prescription_Service
7. Medication reconciliation — Wikipedia — https://en.wikipedia.org/wiki/Medication_reconciliation
8. Polypharmacy — Wikipedia — https://en.wikipedia.org/wiki/Polypharmacy
9. Medical error — Wikipedia — https://en.wikipedia.org/wiki/Medical_error
10. Adverse drug reaction — Wikipedia — https://en.wikipedia.org/wiki/Adverse_drug_reaction
11. Drug interaction — Wikipedia — https://en.wikipedia.org/wiki/Drug_interaction
12. Patient safety — Wikipedia — https://en.wikipedia.org/wiki/Patient_safety
13. Iatrogenesis — Wikipedia — https://en.wikipedia.org/wiki/Iatrogenesis
14. Alarm fatigue — Wikipedia — https://en.wikipedia.org/wiki/Alarm_fatigue
15. Misuse of Drugs Act 1971 — Wikipedia — https://en.wikipedia.org/wiki/Misuse_of_Drugs_Act_1971
16. Pharmacy automation — Wikipedia — https://en.wikipedia.org/wiki/Pharmacy_automation
17. NHS Business Services Authority — Dictionary of medicines and devices (dm+d) — NHSBSA — https://www.nhsbsa.nhs.uk/pharmacies-gp-practices-and-appliance-contractors/dictionary-medicines-and-devices-dmd
18. NHS England — Electronic Prescription Service (EPS) — NHS England — https://digital.nhs.uk/services/electronic-prescription-service
19. National Institute for Health and Care Excellence — Medicines optimisation (NG5) — NICE — https://www.nice.org.uk/guidance/ng5
20. Royal Pharmaceutical Society — Medicines Optimisation — RPS — https://www.rpharms.com/recognition/all-our-campaigns/medicines-optimisation
21. World Health Organization — Medication Without Harm — WHO — https://www.who.int/initiatives/medication-without-harm
22. Specialist Pharmacy Service — Guidance on medicines optimisation and ePMA — SPS — https://www.sps.nhs.uk/