Clinical Compass

5.0 (1 Review)
Methods for lectures, sheets for the ward, and how to read imaging and labs. Pay once.
Location hidden
•
•Created byProfile picturehamza
Profile picture
hamzaProfile picture@hamzak5·9h

Why you keep missing things on X-rays (and it’s not a knowledge problem)


4th year here. For ages I’d glance at an X-ray, spot the obvious thing, and stop looking.

Then the consultant would point at something else on the same film and I’d feel my face go red.

What fixed it wasn’t learning more. It was changing how I look:

1. Same order every single time. Pick a routine and never skip a step, even when it looks obvious. The obvious finding is the one that makes you stop early.

2. Describe before you diagnose. Say what you can see in plain words first. Naming the diagnosis too early is how you miss the second finding.

3. Check the boring corners. Edges, areas behind other structures, anything you’d normally skip. That’s where things hide.

4. Compare with the previous one if there is one. A change tells you more than a single film.

5. Say what you’re unsure about. “I can’t tell if this is X, I’d want a senior to look” beats a confident wrong answer every time.

I put my full routine into a short guide, How to Read Imaging, along with the rest of my packs:

And if the freezing happens when you present on the round, there’s a free 90-second card for that:

What’s the type of scan you find hardest to read? 👇

Profile picture
hamzaProfile picture@hamzak5·13h

Three lines on every set of bloods: findings, impression, next action

Three lines on every set of bloods: findings, impression, next action


You freeze on bloods because you read the arrows out loud. A registrar does not want a tour of the printout. They want three lines: what actually changed, what it means in this patient, and what you will do in the next ten minutes.


This is not the four-move hunt, and it is not a trend checklist. Those tell you what to look at. This is how you write the output so you stop sounding like a list of arrows.


Confirm the patient, the time, and whether this is the first set or a repeat. Treat the patient, not the printout.


What students actually say


"U&Es are a bit off. Potassium is high-ish. Creatinine is up. Sodium is a bit low."


That sentence has no baseline, no meaning, and no job. It is the printout with extra words.


If Line 2 is just Line 1 said politely, you have not interpreted anything.


Line 1: findings (the numbers that earn a place)


Only the results that change the next hour, with the last value if it changes the meaning.


  • Name the test. Name the number. Name yesterday or last month if you have it.

  • Drop the mild arrows that do not change the plan. A urea of 7.2 that has been 7.2 is not a finding you present.

  • If the sample is haemolysed, clotted, or from a drip arm, that is the finding. Stop there.


Bad: "Na down, K up, creat up."

Good: "Creatinine 210, was 90 two days ago. K 5.8, not haemolysed. Na 134, unchanged."


Line 2: impression (one clinical sentence)


This line contains no numbers unless the number is the diagnosis you are naming (for example a potassium you are treating).


It answers: what is going on in this person, not what is highlighted in red.


  • Acute kidney injury, not "creatinine up."

  • Expected for known CKD, not "renal function abnormal."

  • Possible haemolysis of the sample, not "high K," if the patient is well and the ECG is fine.

  • If you do not know, say the two leading options and what would split them. Do not invent a syndrome.


If you cannot say Line 2 without looking at the arrows, you are still on Line 1.


Line 3: next action (a job, an owner, a time)


"Monitor" is not a plan. "Keep an eye on it" is not a plan.


Write one of these, specifically:


  • Repeat which test, at what time, who is waiting.

  • Hold which drug, now, and tell the nurse.

  • ECG, senior review, or do not send them for a walk.

  • No new action on this result, because it did not move, and here is what we are actually treating.


A result without Line 3 is theatre.


Three drills (write them, do not just nod)


1. The creatinine that did not move

Findings: creatinine 180, was 170 last month, K 4.6, urine output normal.

Impression: stable known CKD. This is not an acute kidney injury.

Next action: no new action on the creatinine. Find why they are in hospital instead of staring at a number that did not move.


2. The sodium you cannot dump as "Na down"

Findings: Na 122, was 136 two days ago, on sertraline, not obviously dry.

Impression: acute fall. Drug-associated until a senior says otherwise. Do not fluid-restrict on the first look.

Next action: repeat Na, check glucose, hold the SSRI pending senior, do not leave free water at the bedside.


3. The inflammatory numbers that are not a diagnosis

Findings: WCC 14 (was 11 yesterday), CRP 80, new fever overnight, day 3 of co-amoxiclav.

Impression: not improving on current antibiotics. This is treatment failure to flag, not "infection" as a personality trait.

Next action: senior review, cultures if not already sent. Do not add a second antibiotic yourself.


Copy those three headings onto a card. Fill them before you speak.


The three ways Line 2 collapses back into Line 1


  1. You repeat the numbers and call it an impression.

  2. You skip the last result, so a chronic number becomes an emergency, or an emergency becomes "a bit off."

  3. Your plan is "monitor" with no person and no time.


If any of those happened, rewrite the three lines. Do not add a fourth.


How to practise this until it is automatic


  • Cover the panel. Write three lines in 45 seconds. If you over-run, you started listing arrows.

  • If Line 2 contains a number, you have not left findings yet. Rewrite it.

  • After every set of bloods, note the one question the registrar asked. That is the line you skipped.

  • Confirm locally. Never invent findings.


How to Read Labs is the structured method for FBC, U&E, ABG, and the bloods you get asked to interpret. It is $12 on its own, or in the complete set ($33) with How to Study Medicine, Ward Workbook, and How to Read Imaging. Pay once. Study Kit is included free with every purchase.


It does not replace your lectures, your hospital, or a question bank.


Profile picture
hamzaProfile picture@hamzak5·1d

You’re on the round. The consultant looks at you and says “so, tell me about this patient.”

Your mind goes blank. Here’s what I use when that happens:

1. “This is [age, sex] admitted with [one-line reason].”

2. “Overnight, [what changed, or nothing].”

3. “I think [your assessment], so my plan would be [plan].”

That’s it. Three sentences buys you time, makes you sound organised, and gives the consultant something to build on instead of watching you panic.

I made a free pocket card with the full version, so you can screenshot it before your next placement:

And if you want the rest of what I’ve written for students (studying, labs, imaging):

What’s the question you’re most scared of getting asked on a round? 👇

Profile picture
hamzaProfile picture@hamzak5·1d

I used to look at bloods and say “um... they’re abnormal.

For ages my lab presenting was just reading numbers off the screen and hoping the consultant didn’t ask a follow-up.

What fixed it was changing the question. I stopped asking “is it normal?” and started asking:

1. Which way is it going? Scroll back before you say anything. One result is a snapshot, three results are a story.

2. How fast did it move? A slow drift and a sudden jump are two different conversations.

3. Does it match the patient? If the numbers and the person in the bed don’t agree, don’t trust either one until you’ve checked.

4. Does it change the plan? If not, don’t spend your 90 seconds on it.

5. What am I waiting on? Know the one result you’re chasing before the round starts, so you’re not caught out.

That’s honestly 80% of sounding like you know what you’re doing with bloods.

I wrote the full method up as a short guide (How to Read Labs), and the rest of my packs are here:

And if the freezing happens when you present in general, I made a free 90-second card for that:

Profile picture
hamzaProfile picture@hamzak5·1d

CXR emergencies drill: tension, white-out, free air

CXR emergencies drill: tension, white-out, free air


You freeze on a chest X-ray because you start at the interesting blob. A registrar does not want a tour of the bones. They want you to name the three things that kill, or to say they are not there, before you admire anything else.


This is not a full CXR method. It is a 60-second drill. Quality first, then three emergencies, then a one-line plan. Stop if any of them is present.


Confirm the patient, the date, and laterality. Treat the patient, not the image.


Before the drill: is this image usable?


Skip this and you will report a finding the image cannot support.


  • Name and date match the person in front of you.

  • PA vs AP, erect vs supine. AP and supine magnify the heart and hide fluid levels.

  • Rotation: medial ends of the clavicles equidistant from the spinous processes.

  • Inspiration: about 5–6 anterior ribs. Poor inspiration mimics a big heart and basal white-out.

  • If it is unreadable, say so and repeat it. Do not invent a pneumothorax on a rotated, expiratory AP.


Write “technically limited” if quality is poor. Then still hunt the three emergencies. A bad image can still show tension.


Emergency 1: tension pneumothorax


Look at both edges, not the middle.


  • Visceropleural line. No lung markings beyond it.

  • Mediastinum shifted away from the lucent side.

  • Flattened hemidiaphragm on that side. Distended hemithorax.


If you see shift away plus a lucent side, say it out loud. This is not a radiology meeting. Senior, oxygen, and the local decompression protocol. Do not keep scrolling the rest of the image while the patient is hypotensive.


A visceropleural line without shift is still a pneumothorax. Name it. It is not tension until there is shift and a sick patient. Do not wait for the report to tell a senior.


A skin fold is not a visceropleural line. Lung markings continue beyond a fold. If you are unsure, say you are unsure and get a senior, not a longer stare.


Emergency 2: white-out, and which way the mediastinum moved


A white hemithorax is not one diagnosis. The shift tells you which.


  • Shift away from the white side: large pleural effusion (or rarely a huge mass). The fluid pushes.

  • Shift toward the white side: collapse. The lung pulled.

  • No shift, sick patient: mixed, consolidation, or a supine effusion. Do not force a single label.


Say the side, the shift, and what that means in one sentence.


“Left white-out, trachea pulled left: collapse until proven otherwise. Senior, airway, previous image.”


“Right white-out, trachea pushed right: large effusion. Sit the patient up, senior, do not needle it yourself.”


A white-out you call “consolidation” without checking shift is a miss.


Emergency 3: free air under the diaphragm


On an erect chest X-ray, look under both hemidiaphragms before you look at the lungs.


  • Crescent of lucency under the diaphragm: perforated viscus until a senior says it is not.

  • Check both sides. The right is easier. The left hides under the stomach bubble. Do not call stomach gas a perforation, and do not miss free air because you only looked right.

  • If the image is supine, free air is harder. You may not see a crescent. Say the limitation. Do not reassure yourself.


If you see it, say it out loud. Senior, keep the patient nil by mouth, do not send them for a walk.


Chilaiditi (colon under the diaphragm) exists. If the “air” has haustra, pause. If you are not sure, treat it as free air and get a senior. Confirm locally. Never invent findings.


The 60-second script


Say this in your head, every chest X-ray, before ABCDE.


  1. Identity and quality. Usable or not.

  2. Both edges: visceropleural line? Shift away?

  3. White-out? Which way did the mediastinum move?

  4. Under both hemidiaphragms: free air?


If any is yes: name it, next action, senior. Then stop the drill and treat.


If all are no: say “no tension, no white-out, no free air” and only then do the rest of the image.


A chest X-ray without that sentence is theatre.


How to practise this until it is automatic


  • Ten images. For each, only those four lines. Time-box 60 seconds. Do not describe the bones.

  • After every real CXR on the ward, note the one question the registrar asked. That is the emergency you skipped.

  • Teach the three to a peer out loud. If you cannot teach it, you do not own it yet.


How to Read Imaging is the student method for CXR, ECG, and the imaging you get asked to read. It is $12 on its own, or in the complete set ($33) with How to Study Medicine, Ward Workbook, and How to Read Labs. Pay once. Study Kit is included free with every purchase.


It does not replace your lectures, your hospital, or a question bank.


Profile picture
hamzaProfile picture@hamzak5·1d

Your consultant gives you 90 seconds on the ward round. Here’s how to use them.

Most students ramble because they don’t have a fixed order. Try this skeleton:

  1. Who: age, sex, one-line reason for admission

  2. Day number: where they are in the stay

  3. Overnight: what changed (events, obs, new symptoms)

  4. Key results: only the ones that change management

  5. Your assessment: what you think is going on

  6. Plan: what you’d do next, and what you’re unsure about

Say it out loud, time yourself, and cut anything that doesn’t change the plan.

I’m a 4th year student on the Manchester-Alexandria joint program. I turned this method into a free pocket card you can screenshot or print for the ward.

👉 Get the free 90-second ward present card:

If you want to go deeper, the full Clinical Compass packs are here:

Profile picture
hamzaProfile picture@hamzak5·2d

For most medical students, it’s a total freeze. You dump the whole clerking, talk for 4 minutes, and still don't have a plan.

The fix isn't memorizing more notes—it’s having a 90-second skeleton:

  1. Identity

  2. Problem Now

  3. Background

  4. Impression

  5. Next Action

I put together a free 90-second ward presentation kit ($0) so you never freeze on a round again.

Grab the starter card and study packs on the page:

Profile picture
hamzaProfile picture@hamzak5·2d

Hyperkalaemia: the ECG, the U&E, and the next 10 minutes

Hyperkalaemia: the ECG, the U&E, and the next 10 minutes


You freeze on a high potassium because you treat the number as a fact to report. A registrar treats it as a clock. The question is not “is 6.4 interesting.” The question is whether this patient has an ECG that can arrest, a sample you can trust, and a plan for the next ten minutes.


Do not start with the cause. Do not start with a textbook list of peaked T waves. Start with the patient, the tracing, and whether the sample is real.


Confirm the patient, the time, and whether this is the first potassium or a repeat. Treat the patient, not the printout.


Minute 0: is this sample real?


A haemolysed potassium is not a potassium. An old sample, a drip-arm sample, or a clotted tube is not a potassium.


  • If the patient looks well, the ECG is normal, and the lab flags haemolysis: repeat on an unhaemolysed sample. Do not start calcium on artefact.

  • If the patient is unwell, the ECG is wrong, or you have no reason to distrust the sample: treat while you repeat. Do not wait for a second number if the tracing is already a problem.

  • Pull the last U&E. A jump from 4.2 to 6.8 is not the same as 6.1 that has been 6.1 for a week.


If you cannot say whether the sample is valid, you have not started.


Minute 1: look at the ECG before you look up a protocol


Do not hunt for a pretty peaked T and call it done. Read the tracing for what kills.


In order:


  1. Identity, date, and whether the tracing is readable. Repeat it if it is not.

  2. Rhythm: sinus, slow escape, ventricular rhythm, sine wave.

  3. P waves: flattening or disappearing.

  4. PR and QRS: lengthening QRS is the finding that changes the next minute, not a slightly tall T.

  5. T waves: peaked, broad, or not. Supportive, not the whole diagnosis.

  6. Compare with an old ECG if one exists.


Say the emergency out loud if it is there. Sine wave, QRS widening, ventricular arrhythmia, or a very slow escape: this is not a “repeat later” potassium.


What you do in the next 10 minutes (if it is real)


Follow your hospital’s hyperkalaemia protocol. The sequence below is the student skeleton so you do not stand there holding a printout. Confirm locally. Never invent findings.


  1. Protect the heart if the ECG is unstable. Senior in the room. Calcium per protocol to stabilise the membrane. That does not lower the potassium. It buys time.

  2. Stop what is still pushing it up. Hold ACE inhibitors, ARBs, potassium-sparing diuretics, potassium replacement, and any infusion that contains potassium. Say it out loud so it actually happens.

  3. Shift and remove, per protocol. Insulin and glucose, nebulised salbutamol if that is on your guideline, then the plan that actually clears potassium (that is a senior decision: fluids, dialysis, calcium resonium is not an emergency drug). Recheck glucose if insulin is given.

  4. Repeat the potassium and the ECG. Write the time. Name who is waiting for the result. A treatment without a repeat is theatre.

  5. Find the driver once the patient is safer. AKI, missed dialysis, drugs, acidosis, haemolysis of a real high K, rhabdomyolysis. The cause does not come before the clock.


If the ECG is normal, the sample is valid, and K is high but not crashing: still hold the drugs, still tell a senior, still repeat. “Watch” is not a plan unless someone is actually watching.


The four questions before you walk away


  1. Was this sample valid?

  2. What did the ECG show, in one line?

  3. Which drugs and fluids did I just stop?

  4. When is the repeat, and who owns it?


If you cannot answer those, you have not finished the potassium.


How to practise this until it is automatic


  • Write sample → ECG → protect / stop / shift → repeat on a card. Fill it before you speak.

  • Time-box: ten minutes to have a senior, an ECG, the drugs held, and a repeat time. If you over-run, you started with the cause.

  • After every high K, note the one question the registrar asked. That is the step you skipped.

  • Confirm locally. Never invent findings.


How to Read Labs is the method for the U&E. How to Read Imaging is the method for the ECG. Together they are in the complete set ($33), or $12 each. Pay once. Study Kit is included free with every purchase.


It does not replace your lectures, your hospital, or a question bank.


Profile picture
hamzaProfile picture@hamzak5·3d

The clerking skeleton: before clinic, one patient, after clinic

The clerking skeleton: before clinic, one patient, after clinic


You freeze on placement because you walk in empty and walk out with a pile of names. A registrar does not want your whole day. They want you to own one patient, in order, and still know what happened to the rest.


Use three sheets. Same order every day. Fill them that day. Do not wait until the train home.


This is not the 90-second present. That is how you speak. This is how you arrive, how you hold one person, and how you leave.


Before clinic (five minutes, before you take a coat off)


Write four lines. If you cannot, you are not ready to be useful.


  1. Where am I, and who is in charge. Ward, bay, consultant, registrar, nurse in charge.

  2. The list. How many patients. Who is new. Who is going home. Who is sick.

  3. The jobs already on the board. Bloods, scans, discharges, a family update. Circle two you can actually finish.

  4. One patient you will own. Not five. One. Name, bed, why they are here.


If you skip this, you will shadow all morning and still have nothing to present.


One-patient sheet (the person you actually own)


One page. One person. Not the entire clerking dumped onto a napkin.


Identity. Name, age, bed, why they are in hospital. One sentence.


Problem now. What has changed since yesterday, or since admission. Observations. The one finding that actually matters. If nothing has changed, write that after you have checked, not before.


Background that changes the plan. Relevant past history, relevant drugs, relevant allergies. Not every tablet. If it does not change the next hour, it does not go here.


What you have already done. Bloods sent? Image requested? Drugs held? Write it so you do not repeat it or forget it.


Impression. One sentence of what you think is going on. Not a differential list you memorised.


Next action. What you want in the next ten minutes. Be specific. Repeat potassium. Senior review. Hold the ACE inhibitor. Call the family.


If the sheet is longer than one side, you copied the notes. You did not own the patient.


Copy the sheet. Fill it that day. Confirm locally. Never invent findings.


After clinic (ten minutes, before you leave the building)


This is the part students skip, which is why the next morning feels like a new job.


  1. Jobs left. What is still open, who needs to know, when it is due.

  2. The one patient. What changed today, what you told the team, what you need to check tomorrow.

  3. One line you missed. The question the registrar asked. That is the heading you skipped.

  4. Tomorrow’s first two jobs. Written, not “I’ll remember.”


If you cannot name tomorrow’s first two jobs, you attended clinic. You did not finish it.


What this is not


This is not a full clerking. It is not an OSCE station pack. It is not pharmacology notes.


A clerking lives in the notes. These sheets are how you walk in, hold one person, and walk out without losing the day.


Do not start with the 90-second present if the sheet is blank. Speak from the sheet. If the sheet is empty, sit down and fill it first.


How to practise this until it is automatic


  • Night before: fill the before-clinic four lines from the list you already have, or from what you were told.

  • On the ward: one-patient sheet for the person you own, updated when something changes, not at 5pm from memory.

  • Before you leave: after-clinic four lines. If you over-run ten minutes, you wrote a diary instead of jobs.

  • Next morning, read yesterday’s after-clinic before you take a coat off. That is your before-clinic.


The Ward Workbook is the daily sheets for this: before clinic, one-patient sheet, 90-second present, after clinic, clerking skeleton. It is $9 on its own, or in the complete set ($33) with How to Study Medicine, How to Read Imaging, and How to Read Labs. Pay once. Study Kit is included free with every purchase.


It does not replace your lectures, your hospital, or a question bank.


Profile picture
hamzaProfile picture@hamzak5·4d

Turn a 100-slide lecture into something you can use on the ward

Turn a 100-slide lecture into something you can use on the ward


You are not behind because you did not highlight enough. You are behind because the lecture ended and you still cannot answer three questions: what is this, how do I recognise it, and what do I do in the next ten minutes.


A 100-slide deck is not a study plan. It is a pile. If you try to keep all of it, you will keep none of it when a registrar asks.


Run this intake the same day as the lecture. Same day. Not “this weekend.”


Before you open the slides again


Write the session title and one sentence of why a ward doctor would care.


“Community-acquired pneumonia: recognise it, risk-stratify it, start treatment without waiting for a perfect culture.”


If you cannot write that sentence, you sat through content. You did not take a lecture.


Pass 1: strip it to the skeleton (15 minutes)


Do not reread. Do not Anki yet. Walk the slides once and keep only four kinds of thing:


  1. The definition that changes a label. What makes this this, and not the lookalike.

  2. The bedside findings. History, exam, observations. What you would actually see.

  3. The first tests. The ones that change the next hour, not the research workup.

  4. The first actions. Drugs, fluids, oxygen, hold a tablet, call someone.


Everything else is storage. Mechanisms you cannot use at the bedside, historical asides, ten extra differentials you will not name on a round. Park them. You can come back. You will not, and that is fine.


Write those four headings on one page. If a slide does not feed one of them, it does not go on the page.


Pass 2: one patient, not ten facts (10 minutes)


Invent or borrow one patient who would have been in that lecture.


Age, why they are here, the one finding that matters, the first test, the first action.


Example after a heart-failure lecture: “72-year-old, overnight orthopnoea, raised JVP, wet bases. CXR and BNP later. Sit up, oxygen to target, senior review, hold the fluids you were about to chart.”


If you cannot put the lecture into one patient, you still have a list. Lists die on the ward. Patients do not.


Pass 3: the three questions you must be able to answer tomorrow


Cover the page. Say them out loud.


  1. How would this walk into clinic or the take?

  2. What would I look at first on the patient, the bloods, or the image?

  3. What would I do in the next ten minutes, and what would I not do?


If you stall on any of them, that heading is the only thing you restudy. Not the whole deck.


What this is not


This is not a summary of every slide. It is not a question bank. It is not “make 80 cards tonight.”


Cards come after the skeleton, and only for the bits that are easy to mix up: criteria, first-line drug, a number your hospital actually uses. Confirm locally. Never invent findings.


If you spend two hours colouring the slides and cannot present one patient, you studied the wrong object.


How to practise this until it is automatic


  • Same night as the lecture. Fifteen minutes strip, ten minutes one patient, then the three questions.

  • Next morning, before you leave, answer the three questions again with the page closed.

  • After a ward case that matches the lecture, add one line: what you missed. That line is worth more than the leftover slides.

  • One lecture, one page. If the page is longer than one side, you kept storage.


How to Study Medicine is the methods course for this: lecture intake, memory, and exams. It is $12 on its own, or in the complete set ($33) with Ward Workbook, How to Read Imaging, and How to Read Labs. Pay once. Study Kit is included free with every purchase.


It does not replace your lectures, your hospital, or a question bank.