Watch the video course for this resource
The Clinical Judgment Model, case studies, and the bowtie — everything the NGN changed and how to prepare. Each lesson ends with a 3-question check, and a final 5-question exam at the end.
- L1
What the NGN actually changed
Same pass/fail, same CAT — new item types built around clinical judgment.
3-question quiz after - L2
The Clinical Judgment Measurement Model
The six-step model that every NGN item is built to measure.
3-question quiz after - L3
Case studies — how to read them
Six items, one case: scan the chart, work top-to-bottom by NCJMM step, don't get lost in noise.
3-question quiz after - L4
The bowtie item type
Two actions, one condition, two monitoring parameters — the single strongest predictor of clinical judgment.
3-question quiz after
Save these to your device, print, or import into your calendar. 100% free — no signup.
What actually changed
The NGN launched in April 2023. Pass/fail scoring, the computerized adaptive engine, and the Client Needs blueprint all stayed the same. What's new is a family of item types built to measure clinical judgment — the reasoning that happens between reading a chart and acting on it — and a scoring model that can award partial credit.
The Clinical Judgment Measurement Model: six cognitive steps
Every NGN item maps to one of NCSBN's six steps. Naming the step you're being asked for is the single fastest way to narrow an answer, because each step has its own kind of correct response.
Recognize cues
The question it asks: Which client data is relevant, abnormal, or urgent?
How it appears: Highlight the findings in the nursing note that require follow-up; matrix of relevant vs not relevant.
Common trap: Scanning for abnormals only. A normal value can be a cue when the client's history makes it unexpected.
Analyze cues
The question it asks: What do those cues mean when linked to the history?
How it appears: Cloze: "The client's findings are most consistent with ___ caused by ___."
Common trap: Naming a diagnosis without connecting it to the cues. NGN wants the linkage, not the label.
Prioritize hypotheses
The question it asks: Which explanation is most likely — and which is most urgent?
How it appears: Rank or select the condition the client is most likely experiencing from several plausible options.
Common trap: Picking the most serious diagnosis rather than the one the data actually supports.
Generate solutions
The question it asks: Which interventions and outcomes fit this hypothesis?
How it appears: Matrix of indicated / non-essential / contraindicated interventions.
Common trap: Choosing textbook-correct actions that don't match this client's condition or setting.
Take action
The question it asks: What will I implement first, and how?
How it appears: Extended multiple response, or drag actions into the order of implementation.
Common trap: Ignoring sequence. Assess-before-act and stop-the-cause usually come before medications.
Evaluate outcomes
The question it asks: Did the client improve, and what happens next?
How it appears: Trend item comparing pre- and post-intervention data; matrix of improved / unchanged / worsened.
Common trap: Judging effectiveness by one value instead of the whole trajectory.
Every NGN item type, with a worked example
1. Unfolding case study (6 items)
A single client scenario with a chart that grows as you progress. Six questions follow — one per cognitive step. Roughly three case studies (18 scored items) appear on a typical exam.
Worked example
Chart: 68-year-old, 2 days post-op hip arthroplasty. T 38.4°C, HR 118, RR 28, BP 92/58, SpO₂ 90% on room air, new confusion, right calf swollen and tender.
Step 1 item: Highlight the findings requiring immediate follow-up → RR 28, SpO₂ 90%, HR 118, BP 92/58, new confusion, calf swelling.
Step 3 item: The findings are most consistent with pulmonary embolism from a post-op DVT.
Step 5 item: First actions — apply oxygen, keep the client on bed rest, notify the provider, prepare for anticoagulation. Not massaging the calf or ambulating the client.
Step 6 item: Effective if SpO₂ rises, respiratory rate falls, mentation clears, and blood pressure stabilizes.
2. Bowtie
A stand-alone item with five drop zones: one condition in the center, two actions on the left, two parameters to monitor on the right. It compresses the entire model into one screen.
Worked example
Client missed dialysis; K⁺ 7.1 mEq/L, peaked T waves, widening QRS, muscle weakness.
Condition: life-threatening hyperkalemia
Actions: IV calcium gluconate (stabilizes the myocardium) · IV insulin with dextrose (shifts potassium intracellularly)
Monitor: cardiac rhythm · serial serum potassium
3. Trend item
Data from several time points. You interpret the trajectory rather than any single value.
Worked example
0800 — BP 122/74, HR 88, GCS 13. 1000 — BP 152/62, HR 62, GCS 10, sluggish right pupil.
Widening pulse pressure with bradycardia and a falling GCS is Cushing's triad: intracranial pressure is rising. Elevate the head of bed to 30° with the head midline, avoid clustered stimulation, and notify the provider.
4. Matrix / grid
Each row is a finding or intervention; each column is a judgment you assign.
Worked example
Client on IV heparin — categorize each result as "report to provider" or "no action":
- Platelets 88,000/µL → report (possible HIT)
- aPTT 65 seconds → no action (therapeutic)
- Black tarry stool → report (GI bleeding)
- Hemoglobin 12.4 g/dL → no action (normal)
5. Drag-and-drop cloze
You complete a clinical sentence from dropdown word banks — the format that most directly tests analyze-cues reasoning.
Worked example
"The client's pH 7.24 and PaCO₂ 62 mmHg with a respiratory rate of 7 indicate uncompensated respiratory acidosis caused by opioid-induced hypoventilation."
6. Extended multiple response
Select-all-that-apply with a defined number of choices and, often, partial credit.
Worked example
Which actions are indicated for a client in septic shock? → Obtain blood cultures before antibiotics · give 30 mL/kg isotonic crystalloid · start broad-spectrum antibiotics within one hour · trend lactate. Not fluid restriction.
7. Highlight in text
You click the words in a chart note that indicate a complication — nothing else on the screen changes.
Worked example
Post-thyroidectomy note: "Client resting, high-pitched inspiratory sound noted, reports difficulty swallowing, dressing dry, hoarse voice." Highlight the stridor and dysphagia — hoarseness is expected after tracheal manipulation.
How NGN scoring differs
- 0/1 (dichotomous): Traditional all-or-nothing. Every part must be right to earn the single point.
- +/- (plus/minus): Each correct selection earns a point and each incorrect selection loses one, with a floor of zero. Selecting everything guarantees a low score.
- Rationale scoring: Credit is awarded per linked pair. In a cloze or bowtie, a correct action attached to the wrong condition earns nothing — the pairing is what's scored.
The adaptive engine still delivers items near your ability level and still stops on the 95% confidence-interval rule, so a partially credited NGN item nudges your estimated ability rather than flipping a switch.
Your NGN prep plan
Work through these in order. Progress saves on this device.
Saved on this device — no account needed.
Common NGN mistakes
- Reading the chart top-to-bottom instead of scanning for abnormals, then re-reading for context
- Shotgunning multiple-response items where wrong selections cancel correct ones
- Answering with textbook priorities instead of the priority for this client, right now
- Naming a condition without checking that the cues actually support it
- Skipping the rationale on items you got right