NR-569 Week 3 Ranking the Differential: How to Write It

What NR-569 Week 3 asks for

A rapid response is called on a surgical floor for confusion and a heart rate of 128. Six clinicians can name eight possible causes inside a minute. Only two of them can say which one they are ruling out first, and why that one rather than the more likely one. That distinction is the whole content of this stage. Generating a differential is a recall task most graduate students already pass. Ordering a differential is a judgment task, and judgment is what a written assignment can actually capture and grade.

Two axes have to appear in your prose. Likelihood is where the base rates live: what the risk category, the tempo and the setting make probable. Lethality is where the must-not-miss diagnoses live: the ones that are unlikely but will kill the patient inside hours if missed, and that therefore get worked up out of order. A written differential that lists only the probable has failed the acute care version of the task, and one that lists only the catastrophic reads as anxiety rather than reasoning. The order you defend is the evidence that you can hold both.

Expect a deliverable that pairs one or more de-identified presentations with a ranked list and a paragraph of defense per item, sometimes as a table with a discussion around it. Some sections run a parallel discussion post. Draft that elsewhere and paste the finished version, because Canvas posts do not reopen after submission and a ranking you would revise on rereading is exactly the thing peers respond to.

The standing boundary applies here as everywhere in this course. This manual supports the written and preparatory layer only. Clinical hours, encounter logs, patient counts, site paperwork and preceptor evaluations are your own record and are never drafted, reconstructed or estimated with help. The differential you write about must be one you genuinely worked through, with every patient detail de-identified before it reaches the page.

The NR-569 Week 3 method, step by step

Six moves that turn a list into a defended ranking.

1. Generate wide before you generate deep

   Work from a structured prompt rather than memory alone, whether anatomic, physiologic or systems-based. Unstructured recall reliably produces the four diagnoses you saw most recently, which is the mechanism of availability bias operating in plain sight.

2. Split the list into two columns before ranking

   Probable on one side, dangerous on the other, with the overlap marked. The overlap is where your workup starts, and naming it explicitly makes the rest of the ranking almost write itself.

3. Attach discriminating features, not descriptions

   For each candidate write the one finding that would most raise it and the one that would most lower it. A diagnosis with no discriminator attached cannot be moved by data, which means it is decoration on the list rather than a live hypothesis.

4. Rank by consequence-weighted priority and say so

   State the rule you are using in a sentence: what is worked up first is what combines plausibility with time-critical harm. A ranking whose logic is left implicit gets read as arbitrary, and arbitrary is what the reasoning row penalizes.

5. Write the argument against your own leading candidate

   One honest paragraph naming what does not fit. This is the single highest-yield move in the assignment, because premature closure is the documented failure mode of acute diagnostic reasoning and the antidote is legible on the page.

6. Close with what would change the order

   Name the result, the finding or the interval that would rearrange your ranking. A differential that cannot be updated is a conclusion wearing a list's clothing.

A layout and word budget for a ranked differential

Our frame for a written differential built around a single presentation, sized for roughly 1,100 to 1,400 words. It is our own outline rather than anything the university issues, and your week's rubric outranks it wherever the two disagree.

Evidence craft for differential writing

Support likelihood claims with published base rates. Saying a diagnosis is common is an assertion; saying it accounts for a described share of presentations in a cited population is a supported claim. Where a rate exists in the literature, name it with its source and its population, because a base rate from an ambulatory sample does not transfer to a critical care unit.

Use test characteristics rather than adjectives. Sensitivity, specificity and likelihood ratios are the vocabulary that lets a reader check your ruling in and out. Helpful and reassuring are words that survive in conversation and lose points on a graduate paper about diagnosis.

Prefer current specialty guidance for anything time-critical. Where a professional society has published diagnostic guidance for a must-not-miss condition, cite the current edition and give its year in the sentence. Time-critical pathways are revised, and an outdated citation reads as an outdated clinician.

Name the bias you are guarding against, with a source. Anchoring, availability and premature closure are described in diagnostic error literature. Naming the specific bias your ranking is designed to resist, and citing where the concept comes from, converts a self-aware sentence into scored analysis.

Keep the patient unrecognizable and the site unnamed. No exact ages, dates, transfer origins, unusual occupations or facility identifiers. If the presentation is rare enough that removing identifiers would not protect the person, write the assignment around a different encounter.

Five mistakes that cost points in this week's territory

• An alphabetical or system-by-system list. Completeness is not judgment, and a list with no defended order has skipped the graded part of the task.
• Must-not-miss diagnoses missing entirely. In acute care the lethality axis is not optional, and its absence is the fastest way to lose the reasoning row.
• Candidates with no discriminators. If nothing would raise or lower a diagnosis, it is not a hypothesis and it is padding the list.
• Certainty language on a probabilistic task. Clearly, obviously and rule out entirely all overstate what a bedside differential can support this early.
• No counterargument anywhere. A paper that never questions its own leading candidate is demonstrating the exact failure mode the course exists to correct.

Before you submit

• The list is ordered and the ordering rule is stated in a sentence
• Both probable and time-critical candidates appear, with the overlap marked
• Every candidate carries a raising feature and a lowering feature
• Likelihood claims are supported and the source population is named
• At least one paragraph argues against your own leading diagnosis
• Update triggers are specific results or intervals, not general vigilance
• The encounter is de-identified and no site, unit or preceptor is named

Questions students ask about this stage

Keep going