The five prescribing errors every cohort makes

Introduction: The importance of safe prescribing

Prescribing is the most common therapeutic intervention in clinical practice, and it is also one of the most error-prone. For students and junior doctors, the transition from learning about drugs to writing prescriptions for real patients is a steep one. The errors made during this period are not random. They cluster into five predictable patterns, each with identifiable causes and practical countermeasures.

This article reviews those five error categories, using clinical examples to illustrate how they arise and how they can be prevented. The aim is not to add to the list of things to memorise, but to give you a working framework for checking your own prescribing before it reaches the patient.

Error 1: Inaccurate dose calculations

Dose calculation errors are the most frequently cited prescribing error in student cohorts, and they are disproportionately responsible for serious harm. The problem is rarely mathematical ability. It is more often a failure to convert units, to use the correct body weight, or to account for the difference between a loading dose and a maintenance dose.

Consider a common scenario: a patient with cellulitis weighing 84 kg is prescribed intravenous flucloxacillin. The usual adult dose is 1 g four times daily, which requires no weight adjustment. A student, however, assumes that because flucloxacillin is weight-dosed in children, it must be weight-dosed in adults. They calculate 25 mg/kg and prescribe 2.1 g, which is not a standard dose and may be above the recommended adult maximum. The prescription is dose-checked by a pharmacist, but if it had been given, the patient would have received double the intended amount with no added benefit.

Two other calculation pitfalls deserve specific attention. The first is the confusion between milligram and microgram, which has caused fatal errors with drugs such as digoxin, where the therapeutic window is narrow. A prescription for 0.25 mg written as 250 does not solve the problem, because the unit is left off and the meaning is ambiguous. The second is using a patient's weight in pounds instead of kilograms. Doses that are weight-based, such as enoxaparin given for thromboprophylaxis, become substantially too high or too low when this conversion is missed.

One of the most reliable prevention strategies is to do the calculation twice, using two different mental routes. If you calculate a dose of 1.5 mg/kg for gentamicin as 120 mg for an 80 kg patient, check this by starting with the total dose and dividing by the weight to see if it gives you 1.5 mg/kg. The two methods should agree. If they do not, do not prescribe until you have found the discrepancy.

Error 2: Missed allergy and interaction checks

Allergy checks are often treated as a formality, something to be copied from the clerking notes without independent verification. This is particularly dangerous in the context of antibiotics, where a documented penicillin allergy may be a genuine hypersensitivity or a label applied years ago to a minor rash.

A typical student error is to prescribe amoxicillin for a patient with a documented penicillin allergy without asking what the reaction was. If the patient describes anaphylaxis, the prescription must be changed. If they describe a mild itch that resolved spontaneously, amoxicillin may still be contraindicated depending on the institution's policy, but the decision needs to be made deliberately, not defaulted. Blindly clicking "no allergy" when the clerking sheet was incomplete is also an error, and it is one that students are prone to because they are working from a single source of information.

Drug interactions are a related but distinct problem. Students are often good at recalling major interactions, such as potassium-sparing diuretics with ACE inhibitors, but they miss interactions where the mechanism is less obvious. A useful example is the interaction between clarithromycin and simvastatin. Clarithromycin inhibits CYP3A4, reducing the metabolism of simvastatin and increasing the risk of myopathy and rhabdomyolysis. If a patient on long-term simvastatin is prescribed clarithromycin for a chest infection, the correct action is to check whether the statin can be temporarily withheld or if a different macrolide is suitable. The interaction may not cause immediate harm, but the consequence is serious if it is overlooked.

Efficient prevention relies on hard stops. The prescription should not be signed until an allergy check has been performed against the active ingredient names, and not just the brand names. The interaction check should be run with a current medication list, and any interactions that are flagged must be acknowledged or acted on.

Error 3: Ambiguous prescriptions and dangerous abbreviations

Ambiguity in prescriptions does not always appear as a spelling mistake or a missing decimal point. It often appears when a prescription relies on assumed knowledge that the prescriber does not have. The most dangerous feature of an ambiguous prescription is that it looks clear to the person who wrote it.

A well-known example is the trailing zero: a prescription for "2.0 mg" can look like "20 mg" if the decimal point is not visible on a printed label or a faxed copy. The accepted rule is simple. Do not use trailing zeros. Write "2 mg", not "2.0 mg". By the same logic, write "0.5 mg", not ".5 mg", because a decimal point without a preceding zero can be missed altogether.

Abbreviations such as "OD", "BD" and "TDS" are widely used in the UK, but they are not universal. A prescription written in one hospital that uses "OD" for once daily may be read by a pharmacist trained to interpret "OD" as right eye. The safest approach is to write the full English phrase: once daily, twice daily, three times daily. Similarly, the abbreviation "U" for units is a known source of error because the letter can be mistaken for a zero. Insulin should always be prescribed as "units" in full, and this is a requirement in most health systems.

A third category of ambiguity is the incomplete prescription. The correct dose, route and frequency are specified, but the duration is missing. This is an error in its own right, because it leaves the ward team to guess whether the antibiotic should be continued for five days or seven, and it can lead to either under-treatment or a prolonged course. A complete prescription includes what, how much, how, when, and for how long.

Error 4: Failure to reconcile medications

Medication reconciliation is the process of creating an accurate list of what a patient is actually taking, comparing it with the prescribed list, and resolving any discrepancies. For students and junior doctors, the most common failure is not forgetting to ask about medications, but accepting the patient's first answer without checking the details.

A classic example is the patient who says they do not take any inhalers, but who is later found to have a long-acting beta-agonist and an inhaled corticosteroid in their bag, which they take only when they feel breathless. If the admission drug list is not reconciled against the patient's own supply, a relapse can be attributed to the infection when the real cause is that the patient never received their bronchodilators on the ward.

Another example concerns over-the-counter medications. Anti-inflammatory drugs such as ibuprofen are not always volunteered when patients are asked about their prescriptions. If a patient is prescribed warfarin and no one asks about over-the-counter painkillers, a serious bleeding risk can go unnoticed. Similarly, herbal preparations such as St John's wort are well known to interact with warfarin, amitriptyline and oral contraceptives, yet they are rarely disclosed unless specifically requested.

Where formal reconciliation is possible, it should be used. It is the presence of a structured review, done with the patient or carer, that catches these discrepancies. There is no shortcut that avoids the question, "What else are you taking, including anything you bought yourself or that was prescribed by a different doctor?"

Error 5: Inadequate monitoring and follow-up

The fifth error is often the one that is hardest to see at the point of prescribing, because it concerns what happens after the drug is given. A prescription can be correct in every detail and still be a poor prescription if it does not include a monitoring plan.

Take the example of a patient started on amiodarone for atrial fibrillation. The loading regimen is complex, and the drug has significant side effects. If the prescription does not include baseline thyroid function and liver function tests, and a plan for repeat testing at one month and three months, the patient can develop thyrotoxicosis or hepatotoxicity that is only discovered incidentally. The error was not in the dose but in the absence of a follow-up plan.

Another example is the patient prescribed an ACE inhibitor who already has a borderline potassium level of 5.0 mmol/L. The prescription is written, but no repeat electrolytes are arranged. The next blood test, taken a week later, shows a potassium of 6.1 mmol/L with a rising creatinine. Both the patient and the prescriber should have been protected by a simple instruction to check renal function and electrolytes within seven days of starting the drug.

The prevention strategy is to make monitoring an explicit part of the prescription. When you prescribe a drug with established monitoring requirements, write the monitoring in the same entry: the drug, the dose, and the investigation and timescale. This does not mean you need to list every possible test for every drug. It means that for drugs with a narrow therapeutic index, or with well-documented long-term toxicity, the follow-up plan should be as deliberate as the dose.

Conclusion: Strategies for error prevention

These five errors are avoidable, but they are not avoided by vigilance alone, because vigilance is not consistent. They are avoided by building checks into the task.

There are three practical strategies that work. The first is to always use a structured prescribing template, either electronic or paper. The template should force you to complete every field, including allergy status, weight, monitoring plan and duration. A field that is left blank is a sign that the check has not been done.

The second strategy is to prescribe by the drug's generic name whenever possible. Generic names are less prone to confusion across formularies and less likely to conceal a similar-sounding alternative. Drug name confusion is a cause of error that we have not discussed separately, because it is usually dealt with by prescribing generic names and by the pharmacist's final check, but it still depends on the prescriber writing clearly.

The third strategy is to specifically ask, before signing any prescription, "What is the harm I am trying to prevent here?" If the answer is "the disease", then have you addressed the risk of drug toxicity? If the answer is "the symptom", then have you addressed the risk of long-term effects? Framing the prescription as a balance of risk, rather than a single decision, is the core habit that underlies safe prescribing.

Every cohort makes the same mistakes because the mistakes come from the same gaps in experience. The gap, however, closes quickly when the error is discussed, and the mechanism is understood. For the student who reads this article early in clinical training, the most valuable thing they can do is to seek feedback on every prescription they write. Feedback is the one resource that the electronical prescribing systems do not provide.