Failure Mode and Effects Analysis looks at a healthcare process before anything goes wrong, mapping out where it is most likely to break down so the highest-risk failure points can be fixed in advance. This guide covers the structure of an FMEA assignment, how severity, occurrence and detection scoring works conceptually, and a worked example applied to medication administration.
Failure Mode and Effects Analysis (FMEA) is a structured, proactive method for examining a process step by step to identify where it is most likely to fail, what the effects of that failure would be, and how likely the failure is to be caught before it reaches a patient. The word "proactive" is the key distinction to hold onto for this kind of assignment. Where a root cause analysis (RCA) is reactive, examining an event that has already happened and working backward to find its cause, FMEA works forward from a process that has not yet failed, asking "where could this break, and how badly, before it actually does?"
Both tools serve patient safety and quality improvement work, and it is common for a course to teach them side by side or even ask you to apply both to the same process, one looking back at a past event and the other looking forward at future risk. If your assignment is specifically about analyzing an event that already occurred, the RCA guide is the better fit; if it asks you to anticipate where a process might fail before it does, you are in the right place.
An FMEA is usually built as a table, walking through a process one step at a time. For each step, the analysis identifies:
Many FMEA formats multiply severity, occurrence and detection ratings together to produce a single number, often called a Risk Priority Number (RPN), used to rank which failure modes deserve attention first. It is worth being direct about one important point here: the exact numeric scale used for severity, occurrence and detection (whether it runs 1 to 5, 1 to 10, or something else, and what each number on the scale specifically means) varies by course, textbook and organization. There is no single universal scale, so rather than inventing specific numbers, use whatever scale your course, textbook or assigned tool provides, and describe the CONCEPT of each factor clearly in your own write-up. What matters for the assignment is that your reasoning about relative severity, likelihood and detectability is sound and consistent across every row of your table, not which specific numbering convention you used.
In an actual healthcare organization, an FMEA is almost never completed by one person working alone, and understanding this helps explain why the method is structured the way it is. It is typically carried out by a small interdisciplinary team that includes the people who actually perform the process being studied, since frontline staff usually know the realistic failure points far better than someone reviewing the process from a policy document. A medication-administration FMEA might involve a bedside nurse, a pharmacist, a nurse educator and sometimes a representative from health information technology if the process involves the electronic health record or a dispensing cabinet. Each person brings a different, genuine vantage point on where the process is fragile.
Most student assignments are completed individually, but it is worth writing your analysis as though you were contributing to that kind of team discussion, considering the process from more than one professional angle rather than only your own. A paper that only considers the bedside nursing perspective on a process that also involves pharmacy, technology and unit leadership tends to miss failure modes that a broader view would catch. If your assignment allows or requires a note on who would realistically be involved in conducting the analysis, naming these roles briefly shows you understand how FMEA actually functions outside the classroom.
The following is a simplified, hypothetical illustration of a few steps in the medication administration process, created to show the FMEA method, not a comprehensive or official analysis of any real facility's process.
| Process Step | Potential Failure Mode | Potential Effect |
|---|---|---|
| Retrieve medication from dispensing cabinet | Wrong medication or strength selected due to look-alike packaging | Patient receives incorrect drug or dose |
| Verify patient identity | Identity check skipped or performed from memory rather than two identifiers | Medication given to the wrong patient |
| Calculate dose (for weight-based or titrated medications) | Calculation error, misplaced decimal point | Overdose or underdose administered |
| Administer medication | Route or timing error (for example, IV push given too quickly) | Adverse reaction or reduced therapeutic effect |
| Document administration | Delayed or omitted documentation | Risk of duplicate dose from another provider unaware it was given |
For each row, a full FMEA would go on to rate severity (an overdose is more severe than a delayed but eventually completed documentation entry), occurrence (how often this specific failure mode actually happens given current practice), and detection (whether an existing safeguard, such as a barcode scan or an independent double-check, is likely to catch the error before it reaches the patient), using whatever scale your course assigns. The steps with the highest combined risk, not necessarily the most severe in isolation, are the ones prioritized for intervention.
To see how this reasoning plays out without inventing a specific numeric scale, consider two of the rows in words rather than numbers. The wrong-strength selection failure mode is high severity (an overdose or underdose can cause real harm), moderate occurrence (look-alike packaging is a known, documented risk factor across many facilities, though existing storage separation reduces how often it actually happens), and moderate detection (a barcode scan at the bedside would likely catch it, but only if that step is consistently performed and not overridden). The delayed-documentation failure mode, by contrast, is lower severity on its own (it does not directly harm the patient), but carries a meaningful secondary risk through duplicate dosing if another provider is unaware a dose was already given, moderate occurrence (documentation delays are common in a busy shift), and relatively low detection (nothing else in the process reliably catches a missed or delayed note until someone happens to check). Comparing the two shows why a lower-severity failure mode can still rank as a meaningful risk once occurrence and detection are considered together, which is the actual purpose of scoring across all three factors rather than severity alone.
Be specific about the failure mode. "Medication error" is not a failure mode, it is the category. "Wrong strength selected from adjacent cabinet bins" is a failure mode, specific enough that a real intervention can be designed around it.
Share your process, rubric and scoring scale and get a model FMEA that reasons through every step clearly.
Medication administration is a common example because the process steps are easy to describe, but FMEA is applied across many other areas of nursing and healthcare operations, and your assignment may well ask you to analyze one of these instead. Recognizing the pattern across different processes makes it easier to adapt the method to whatever process you are actually assigned.
| Process Area | Example Failure Mode to Consider |
|---|---|
| Fall prevention rounding | A fall-risk alert not carried forward when a patient transfers between units |
| Central line insertion and maintenance | A dressing change step skipped or performed without full sterile technique |
| Patient handoff or shift report | A critical lab value or pending order mentioned verbally but not documented in the handoff tool |
| Discharge planning | A medication reconciliation discrepancy not caught before the patient leaves |
| Blood product administration | The independent two-person verification step performed by only one person under time pressure |
Notice that each of these, like the medication administration example, breaks down into discrete steps with a specific point of vulnerability, rather than a single vague "something could go wrong somewhere." Whatever process your assignment names, start by listing its steps as concretely as you can before trying to identify failure modes, since a vague or overly broad step (such as "provide patient care" rather than "verify patient identity before administration") makes it nearly impossible to identify a precise, actionable failure mode later. If your course lets you choose the process yourself, picking one you have directly observed, on placement, in simulation, or described in detail in a textbook case, will almost always produce a more specific and more convincing analysis than a process you only know in the abstract.
The point of an FMEA is not the table itself but what it leads to. Once failure modes are ranked, whether by a calculated Risk Priority Number or by qualitative judgment using your assigned scale, the highest-risk items are prioritized for intervention first. A strong FMEA assignment closes with a short action plan section that names the top-ranked failure mode(s) and proposes a specific, realistic change to reduce either how often the failure could occur or how likely it is to be caught before reaching a patient, for example redesigning cabinet storage to separate look-alike medications, or adding a mandatory barcode scan step that did not exist before.
A genuinely useful recommendation targets one of the three underlying factors directly. To reduce occurrence, you might propose a physical or procedural change that makes the failure mode less likely to happen at all, such as separating look-alike medication bins. To improve detection, you might propose adding a safeguard that catches the failure before it reaches the patient, such as a mandatory independent double-check for high-alert medications. Reducing severity is often harder to influence directly through a process change, since the potential harm of an error is usually a fact about the medication or procedure itself rather than something the workflow can change, so most realistic student recommendations focus on occurrence and detection rather than severity.
Because FMEA scoring is judgment-based rather than derived from a fixed formula with objective inputs, a strong paper makes its reasoning visible rather than presenting scores as if they were self-evidently correct. For each row where the score might not be obvious to a reader, add a brief note explaining why you rated it the way you did. "Occurrence rated moderate because look-alike packaging is a recognized industry-wide risk, though this facility's existing bin separation reduces frequency" tells the reader far more than a bare number with no explanation, and it is usually this reasoning, not the numbers themselves, that a grader is actually assessing.
This matters even more when you are working from a hypothetical or simplified process rather than direct access to real incident data, which is the normal situation for a student assignment. Be explicit that your occurrence and detection ratings are reasoned estimates based on general patient-safety knowledge and the logic of the process, not figures pulled from an actual facility's incident log, unless your assignment specifically supplies you with real data to work from. Being transparent about this distinction is a mark of good academic practice, not a weakness in the analysis.
Whatever scale your course, textbook or assigned tool specifies. Scales genuinely differ between programs and organizations, so check your assignment instructions or ask your instructor rather than assuming a particular numbering system is universal.
Only if your assignment specifically asks for one. Some courses want the full severity times occurrence times detection calculation for every row; others are satisfied with a clearly reasoned qualitative ranking. Follow your rubric.
Some assignments do combine them, using an FMEA to identify a high-risk step and then treating a related past event with an RCA, or the reverse. If your assignment calls for both, our root cause analysis guide covers the reactive half of that pairing.
Detailed enough that no meaningful step is skipped. A process with five genuinely distinct steps analyzed thoroughly is stronger than a ten-step map where several steps are really the same action split apart for length.
Only if your instructions ask for one. Even when it is not required, briefly noting which roles would realistically be involved in the process you are analyzing (nursing, pharmacy, technology, unit leadership) can strengthen your discussion by showing you understand FMEA is normally a team activity, not a solitary desk exercise.
You can, but it is harder to identify realistic, specific failure modes for a process you have not observed. Where possible, choose a process you have seen in clinical, in simulation, or described in detail in your coursework, since specificity is what separates a strong FMEA from a generic one.
A strong FMEA looks forward, is specific about what could actually go wrong at each step, and ends with a real recommendation rather than just a ranked list. It treats scoring as a transparent, reasoned judgment rather than a set of numbers dropped in without explanation, and it recognizes that in real practice this kind of analysis is a team effort drawing on more than one professional perspective. For the reactive counterpart to this technique, see the root cause analysis guide, and for a broader quality improvement project built around these findings, see the quality improvement guide.
If you would like a model FMEA for your specific process, request an instant quote at /order. Delivered work comes with 14 days of free revisions, and you can read refund terms on the money-back guarantee page. Use any model paper in line with your institution's academic-integrity policy.