Complete USMLE Step 3 E-Book

Azhar ul Haque Sario

• Herausgeber: Azhar Sario Hungary
• Kategorie: Fachliteratur
• Serie: USMLE 2026
• Sprache: Englisch

Seitenzahl: 205

Veröffentlichungsjahr: 2026

Beschreibung

Stop studying outdated medicine; the 2026 USMLE Step 3 demands a new standard of care.

This guide covers the complete exam landscape for the modern resident. It targets Day 1 Foundations. It masters Day 2 Clinical Cases. You get updated Biostatistics. We break down the new 2026 screening ages. You learn the new stroke protocols. We explain the shift in diabetes management. You understand the new "Pre-COPD" definitions. We cover the latest antibiotic stewardship. You explore the "Treat to Target" methods for IBD. We detail the 2026 vaccination schedules. You master the CCS software logic. It includes the new "Gold Standard" for heart failure. You get the specific drug interactions for transplant patients. It covers the new "DoxyPEP" protocols. It explains the "Vector Change" for cardiac arrest. It is comprehensive. It is current. It is designed for the independent physician.

This book provides the competitive edge that older resources miss entirely. Most study guides recycle data from three years ago, leaving you vulnerable to new guideline questions. This text integrates the very latest 2024–2026 updates immediately. It replaces old "step-up" algorithms with the new "top-down" precision medicine approach. You won't just memorize drug lists; you will understand the "why" behind the massive shift to GLP-1s and SGLT2 inhibitors in primary care. It clarifies the confusion around the new "sFlt-1/PlGF ratio" in obstetrics. It turns complex "Grey Zone" clinical scenarios into binary, scoring decisions. This is the difference between barely passing and truly excelling.

Imagine walking into the exam center knowing you aren't just relying on old tricks. You will know why we now treat "Pre-COPD" before the spirometry fails. You will understand why we use DoxyPEP to prevent STIs before they start. This guide walks you through the new era of obesity medicine, explaining why we now treat the biology of weight loss with incretins rather than just willpower. You will navigate the new "Triple Test" for breast masses and understand why we now screen for colon cancer at age 45, not 50.

We strip away the academic fluff and focus on the high-yield decisions you make in the CCS software. You will learn how to handle the "crash" patient in the ER, from the new "Vector Change" defibrillation for refractory VFib to the precise fluids for septic shock. We dive into the new Hepatology, where "NAFLD" is dead, and we now treat "MASLD" with Resmetirom. Even the way you manage a simple headache is updated with the new CGRP antagonists. This is not just a review; it is a clinical update for the future of your practice.

Copyright Disclaimer: This book is independently produced by Azhar ul Haque Sario. It is not affiliated with, endorsed by, or sponsored by the Federation of State Medical Boards (FSMB), the National Board of Medical Examiners (NBME), or the USMLE program. All trademarks are the property of their respective owners and are used herein under nominative fair use for educational purposes only.

Leseprobe

Complete USMLE Step 3: 2026 Exam Study Guide

Azhar ul Haque Sario

Copyright

Copyright © 2026 by Azhar ul Haque Sario Published by Azhar Sario Hungary

All rights reserved.

No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the publisher, except in the case of brief quotations embodied in critical reviews and certain other noncommercial uses permitted by copyright law.

Legal & Disclaimer The information contained in this book is for educational and entertainment purposes only. While every attempt has been made to ensure the accuracy of the information provided, neither the author nor the publisher assumes any responsibility for errors, omissions, or contrary interpretation of the subject matter herein.

[email protected]

ORCID: https://orcid.org/0009-0004-8629-830X

LinkedIn: https://www.linkedin.com/in/azharulhaquesario/

AI Disclaimer: This book is free from AI use. The cover was designed in Canva. No part of this text may be used to train artificial intelligence models without express permission.

Publisher Details Azhar Sario Hungary Author Website: www.linkedin.com/in/azharulhaquesario

Cataloging Data ISBN:

Google Version

978-3-384-80654-3

eBook Version

978-3-384-80655-0

Paperback Version

978-3-384-80656-7

First Edition: 2026 Printed & Distributed by PublishDrive

Copyright Disclaimer: This book is independently produced by Azhar ul Haque Sario. It is not affiliated with, endorsed by, or sponsored by the Federation of State Medical Boards (FSMB), the National Board of Medical Examiners (NBME), or the USMLE program. All trademarks are the property of their respective owners and are used herein under nominative fair use for educational purposes only.

Contents

Copyright

Section I: Foundations of Independent Practice (Day 1 Focus)

Biostatistics & Medical Literature

Systems-Based Practice, Ethics & Safety

Preventive Medicine & Health Maintenance

Section II: Ambulatory & Chronic Care Management (Day 2 Focus)

Cardiovascular Medicine (Chronic & Acute)

Endocrinology (Metabolic Management)

Respiratory Medicine

Gastrointestinal & Hepatic Disease

Renal & Genitourinary

Musculoskeletal & Rheumatology

Section III: Emergency (CCS Focus) & Special Populations

Emergency Medicine & Toxicology

Infectious Diseases

Neurology & Psychiatry

Hematology & Oncology

Obstetrics & Gynecology

Pediatrics

About Author

Section I: Foundations of Independent Practice (Day 1 Focus)

Biostatistics & Medical Literature

Part 1: The Architecture of Evidence (Study Designs)

The validity of any medical claim rests entirely on the structure of the study used to prove it. You must be able to look at a study abstract and immediately identify its skeleton. Is it strong? Is it weak? Is it applicable to the patient sitting in front of you?

1. The Gold Standard: Randomized Clinical Trials (RCTs)

The Randomized Clinical Trial is the courtroom of medicine. We put a therapy on trial to see if it works.

In 2026, the USMLE places heavy emphasis on the nuance between Explanatory and Pragmatic trials.

Explanatory Trials: These are the "classic" RCTs. They test efficacy under ideal, airtight conditions. The inclusion criteria are strict. The patients are usually young and have only one disease.

The Trade-off: These studies have high internal validity (we know the drug caused the result), but low external validity (the results might not apply to your messy, complicated patients).

Pragmatic Trials: These are gaining massive traction. They test effectiveness in "real-world" conditions. They include elderly patients. They include patients with comorbidities.

Why it matters for Step 3: If a vignette presents a study done on "healthy 25-year-olds" and asks if you should apply the findings to a 70-year-old diabetic, the answer is often "No" due to poor external validity.

The Four Phases of Clinical Trials

You must know the lifecycle of a drug. It is a funnel. Thousands of molecules enter; very few survive.

Phase I (Safety First): We give the drug to a tiny group of healthy volunteers. We are not looking for a cure here. We are looking for toxicity. We want to know: Will this kill you? We also study pharmacokinetics (how the body moves the drug).

Example: A new breakdown of a monoclonal antibody is tested in 20 healthy medical students to see if it causes anaphylaxis.

Phase II (Does it Work?): Now we move to a small group of patients with the disease. We are testing for efficacy and trying to find the right dose.

Example: The drug is given to 100 patients with Rheumatoid Arthritis to see if joint swelling decreases.

Phase III ( The Big Show): This is the large-scale RCT. We compare the new drug against the current "standard of care" (or placebo). This involves thousands of patients across multiple centers. This is the phase that usually gets FDA approval.

Example: Comparison of the new antibody vs. Methotrexate in 3,000 patients to prove it is superior.

Phase IV (Post-Marketing Surveillance): The drug is already on the market. This phase never really ends. It detects rare or long-term side effects that were too scarce to show up in Phase III.

2026 Note: This is high-yield. If a question asks how to detect a 1-in-10,000 side effect, the answer is Phase IV surveillance.

2. Observational Architecture

Sometimes we cannot randomize patients. It would be unethical to force 500 people to smoke cigarettes just to see if they get cancer. For this, we use observational studies.

Cohort Studies (The Movie)

Think of a Cohort Study as a documentary film. We find a group of people (the cohort) and follow them forward in time.

The Setup: We separate them based on exposure. (Group A smokes, Group B does not).

The Goal: We wait to see who develops the outcome (Lung Cancer).

The Math: This allows us to calculate Relative Risk (RR).

Pros/Cons: They are powerful but expensive and take years. They are prone to "loss to follow-up" (patients moving away), which can ruin the data.

Case-Control Studies (The Flashback)

Think of a Case-Control study as a detective looking at a crime scene and working backward.

The Setup: We separate them based on outcome. (Group A has lung cancer, Group B does not).

The Goal: We look backward into their history to look for exposure (Did you smoke?).

The Math: You cannot calculate risk here because you started with the disease. You calculate the Odds Ratio (OR).

The Niche: These are the only option for very rare diseases.

Example: If you want to study a cancer that only affects 1 in 1,000,000 people, you can't do a cohort study (you'd need millions of participants). You essentially have to go find the few people who already have it.

3. The Meta-Analysis & Forest Plots

This sits at the top of the evidence pyramid. It pools data from many small studies to create one giant, mathematically powerful result.

Reading the Forest Plot: You do not need a calculator for this. You need visual literacy.

The Line of Null Effect: This is the vertical line in the middle.

For Ratios (OR, RR), the line is at 1.0.

For Differences, the line is at 0.

The Boxes: Each box is a single study. The horizontal line running through the box is the Confidence Interval.

The Rule: If the horizontal line touches the vertical Null Line, that specific study is not statistically significant.

The Diamond: This is the pooled result (the bottom of the chart). If the diamond touches the line, the meta-analysis found no significant difference.

The Danger of Heterogeneity (I2): Just because you pooled studies doesn't mean they fit well together. If one study involved children and another involved elderly patients, pooling them is bad science.

Look for the I2 statistic.

If I2>50, the studies are too different (heterogeneous). The meta-analysis might be invalid.

Part 2: The Language of Risk (Measures of Association)

When you counsel a patient, you don't say "maybe." You talk in risks. Understanding these calculations is vital for the Biostats blocks of Step 3.

1. Relative Risk (RR) vs. Odds Ratio (OR)

Relative Risk (RR): Used in Cohort studies and RCTs.

RR=Risk in UnexposedRisk in Exposed

RR > 1: Exposure increases risk (e.g., Smoking causes cancer).

RR < 1: Exposure decreases risk (e.g., The vaccine is protective).

Odds Ratio (OR): Used in Case-Control studies.

OR=Odds of exposure in controlsOdds of exposure in cases

Interpretation: "Patients with cancer had 5 times higher odds of being smokers than healthy patients."

The Rare Disease Assumption: This is a favorite concept for test writers.

Normally, OR and RR are different numbers.

However: If the disease is very rare (prevalence < 10%), the OR ≈ RR.

Why it matters: This allows us to estimate risk even when we are forced to use a retrospective Case-Control study design.

2. Hazard Ratios (HR) & Survival Analysis

In oncology and chronic disease (Heart Failure), we care about time. It's not just "did the patient die?" but "how long did they survive?"

The Hazard Ratio (HR): Think of this as Relative Risk, but at any specific second in time.

Kaplan-Meier Curves: These are the "stair-step" graphs you see in oncology journals.

The Y-axis: % of patients surviving.

The X-axis: Time.

The Visual Test: Look at the curves for Drug A and Drug B. Do they separate? Great. Do the Confidence Intervals (often shown as shading around the lines) overlap?

The Rule: If the CI shading overlaps, there is no statistical difference, even if the lines look different.

Part 3: Clinical Significance vs. Statistical Significance

A study can have a p-value of 0.0001 (highly significant) but be clinically useless. If a drug lowers blood pressure by 0.1 mmHg, it is statistically significant but clinically irrelevant.

1. Number Needed to Treat (NNT)

This is the most "human" metric we have. It translates abstract percentages into bodies.

NNT=AbsoluteRiskReduction(ARR)1

Step 1: Calculate ARR. (Risk in Placebo Group - Risk in Drug Group).

Step 2: Divide 1 by ARR.

Interpretation: "I need to treat 20 patients with this drug to prevent 1 death."

Low NNT is Good: We want to treat fewer people to get a save. An NNT of 1 is perfect (everyone gets cured).

2. Number Needed to Harm (NNH)

This works the same way, but for side effects.

NNH=AbsoluteRiskIncrease(ARI)1

Example: If the drug causes a rash in more people than placebo.

High NNH is Good: We want the NNH to be very high. We want to treat thousands of people before we accidentally harm one.

3. The Power of Confidence Intervals (CI)

In 2026, we care more about CIs than p-values. A p-value is a simple "Yes/No" switch. A CI tells you the precision.

The Logic of the 95% CI: If we repeated this study 100 times, the true result would fall inside this range 95 times.

Rule of Thumb for Significance:

For Ratios (RR, OR, HR): If the CI includes 1.0, the result is NOT significant.

For Means/Differences: If the CI includes 0, the result is NOT significant.

Width Matters (Precision):

Narrow CI: (e.g., RR 1.5, CI 1.4 – 1.6). This is a precise study. Likely a huge sample size. We trust this number.

Wide CI: (e.g., RR 1.5, CI 1.1 – 3.0). This is "statistically significant" (it doesn't cross 1.0), but it is sloppy. The true risk could be 1.1 (tiny) or 3.0 (huge). This usually indicates a small sample size.

Conclusion & Next Steps

Understanding study design is not about memorizing definitions. It is about understanding the flaws in the data.

If it’s a Case-Control, worry about recall bias.

If it’s a Cohort, worry about loss to follow-up.

If it’s an RCT, check if the patients look like your patients (external validity).

This "literacy of evidence" is what separates a technician from a physician.

Part I: Diagnostic Testing and Probability

The core of clinical medicine is not certainty; it is probability management. Every time you order a test, you are attempting to shift the probability of a disease from "maybe" to "treat" or "dismiss."

The Fixed Characteristics: Sensitivity and Specificity

Think of a diagnostic test as a net.

Sensitivity (True Positive Rate) is about the net’s ability to catch fish. If a test is 100% sensitive, it catches every fish in the water. It misses nothing. If a patient has the disease, a sensitive test will be positive.

The Clinical Utility: Because it catches everything, a negative result is powerful. If the net comes up empty, you can be sure there are no fish. This is SnNout: Sensitive tests, when Negative, rule Out disease.

The Trap: High sensitivity often comes with "noise." It catches the fish, but it also catches old boots, tires, and seaweed (False Positives). You use these tests for screening life-threatening conditions where missing a diagnosis is unacceptable (e.g., ELISA for HIV).

Specificity (True Negative Rate) is about the net’s ability to exclude the garbage. A highly specific test ignores the boots and tires. It only signals when it finds a real fish.

The Clinical Utility: If a highly specific test says "Positive," you can trust it. It doesn't cry wolf. This is SpPin: Specific tests, when Positive, rule In disease.

The Trap: To be this precise, it might let some actual fish swim through the holes. You use these for confirmation (e.g., Western Blot/Differentiation assay for HIV).

The Reality of Practice: PPV and NPV

Here is the nuance for Step 3. Sensitivity and Specificity are laboratory values. They tell you how the test performs in a vacuum. But you don’t work in a vacuum; you work in a clinic with a specific population.

You care about the patient in front of you. The patient asks, "Doctor, my test is positive. Do I have cancer?" They are asking for the Positive Predictive Value (PPV).

The Prevalence Effect: This is the most critical concept for the exam. PPV and Negative Predictive Value (NPV) are not fixed. They float based on Prevalence (how common the disease is in your specific crowd).

Let’s look at an example: Imagine a test for "Condition X" has 99% Sensitivity and 90% Specificity. Sounds great, right?

Scenario A (High Prevalence): You use this test in a specialty clinic where 50% of people have Condition X. If the test is positive, the PPV is incredibly high. You can trust it.

Scenario B (Low Prevalence): You use this same test on the general public (asymptomatic screening), where only 1% of people have Condition X. Because the disease is rare, the vast majority of positive results will be False Positives.

Deep Dive Insight: This is why we do not screen for everything. If you order a "total body scan" or tumor markers (like CEA or CA-125) on a healthy person, you will find abnormalities. But in a low-prevalence setting, those positives are likely statistical noise. You then have to chase those false positives with invasive biopsies and anxiety-inducing procedures. Step 3 rewards you for not ordering a test if the pre-test probability (prevalence) is too low.

The Power Tool: Likelihood Ratios (LR)

If Sensitivity and PPV feel like they are fighting each other, Likelihood Ratios are the peacekeepers. LRs are the most robust metric because they do not change with prevalence. They are purely about the strength of the test.

Positive Likelihood Ratio (LR+): How much more likely is a positive test found in a sick person compared to a healthy one?

Formula: Sensitivity / (1 - Specificity)

Rule of Thumb: An LR+ > 10 is a game-changer. It massively shifts your probability. An LR of 2 or 3 is only a slight nudge.

Negative Likelihood Ratio (LR-): How much more likely is a negative test found in a sick person?

Formula: (1 - Sensitivity) / Specificity

Rule of Thumb: An LR- < 0.1 is excellent. It essentially kills the diagnosis.

Fagan’s Nomogram & Bayesian Reasoning: You don't need to memorize the nomogram visual, but you must master the logic.

Start with Pre-test Probability: This is your gut feeling based on history (e.g., "This 60-year-old smoker with hemoptysis has a 50% chance of cancer").

Apply the Test (LR): You do a CT scan.

ROC Curves: The Trade-Off

No test is perfect. You can tune a test to be more sensitive, but you will lose specificity (and vice versa). The Receiver Operating Characteristic (ROC) curve visualizes this tug-of-war.

The Curve: The Y-axis is Sensitivity. The X-axis is (1-Specificity).

The Perfect Test: Touches the top-left corner (100% Sens, 100% Spec).

The Useless Test: A diagonal line up the middle (a coin flip).

Area Under the Curve (AUC): The bigger the area under the line, the better the test.

Step 3 Application: You may be asked to choose a "cut-off point."

If the disease is fatal if missed (e.g., Ebola), you pick a cut-off with high sensitivity (shift left on the curve). You accept false positives to save lives.

If the treatment is toxic (e.g., Chemotherapy), you pick a cut-off with high specificity. You want to be sure before you poison someone.

Part II: Bias, Confounding, and the "Truth"

In the literature questions on Step 3, you are the detective. The researchers will claim they found a cure. Your job is to find the flaw in their design.

Selection Bias: Who is in the study?

Selection bias happens when the people you studied are not like the people you treat.

Berkson’s Bias: This is the "Hospital Bias."

Example: A study done only on inpatients finds a link between coffee and pancreatic cancer. Why? Because the control group (patients in the hospital for other reasons like ulcers) might avoid coffee due to stomach pain. The general population drinks coffee. The study population was skewed because they were all sick enough to be hospitalized.

Healthy Worker Effect: People who have jobs are generally healthier than people who are unemployed or disabled. Comparing mortality rates of factory workers to the general population will always make the factory look safer than it is.

Information Bias: The Memory Trap

This is about the quality of the data.

Recall Bias: The classic Step 3 trap in Case-Control studies.

Scenario: You ask mothers of children with birth defects, "Did you drink tap water during pregnancy?" You ask mothers of healthy kids the same question.

The Flaw: The mother of the sick child has spent months agonizing over what went wrong. She remembers every sip of water. The mother of the healthy child forgot. It looks like tap water causes defects, but really, it’s just that one group remembered better.

Confounding: The Third Variable

Confounding is the illusion of a relationship. It is when Factor A looks like it causes Disease B, but actually, Factor C is pulling the strings.

Example: A study shows carrying a lighter causes lung cancer.

The Reality: Carrying a lighter is linked to Smoking (The Confounder). Smoking causes the cancer. The lighter is innocent.

How to fix it:

Randomization: The gold standard. It distributes confounders (even the ones we don't know about) equally between groups.

Matching: If you have a 50-year-old smoker in the case group, find a 50-year-old smoker for the control group.

Stratification: Analyze the data in subgroups (smokers vs. non-smokers) to see if the effect persists.

Part III: Screening Biases

Preventive medicine is a huge part of Step 3. You must know when screening works and when it just looks like it works.

Lead-Time Bias: The Illusion of Time

Screening detects disease earlier. That is the point. But detecting it early doesn't always mean living longer.

Analogy: Imagine two runners start a race at 8:00 AM and finish at 9:00 AM.

Runner A is seen at the finish line (Diagnosis at death).

Runner B is spotted at the halfway mark (Screening diagnosis).

Both die at 9:00 AM.

Screening makes it look like Runner B "survived" longer with the disease (from 8:30 to 9:00), but the total lifespan didn't change. You just knew about the disease for longer. This is lead-time bias.

Length-Time Bias: The Illusion of Severity

Screening tests are like taking a snapshot of a highway.

You are more likely to catch the slow cars (slow-growing, indolent tumors) because they stay on the highway longer.

The fast cars (aggressive, rapidly fatal tumors) zoom by between snapshots.

The Result: Screening programs tend to find "good" cancers that might not have killed the patient anyway. This artificially inflates survival statistics, making the screening look more effective than it is.

Part IV: Deconstructing Drug Ads

The Drug Ad questions are unique to Step 3. They simulate the 5 minutes you have between patients when a pharma rep hands you a pamphlet. You need to be cynical and structured.

The Critique Checklist:

Check the Axes:

Did they truncate the Y-axis? Starting a graph at 50% instead of 0% makes a tiny difference look massive.

Is the scale linear or logarithmic?

Surrogate vs. Clinical Endpoints:

Surrogate: "Drug X lowers LDL cholesterol." (Nice, but does it save lives?)

Clinical: "Drug X reduces the rate of heart attacks." (This is what matters).

Trap: A drug might lower cholesterol (surrogate) but increase mortality due to side effects. Always look for "All-Cause Mortality."

Absolute vs. Relative Risk Reduction (ARR vs. RRR):

The Marketing Trick: "Drug X reduces risk by 50%!" (RRR).

The Reality: If the risk went from 2% to 1%, the Absolute Reduction is only 1%.

You need to calculate the Number Needed to Treat (NNT).

If NNT is 100, you have to treat 100 people (and expose them to side effects) to save one.

Confidence Intervals (CI):

Look at the bars. If the 95% CI for a Relative Risk crosses 1.0, the study is not statistically significant. It doesn't matter how pretty the bar chart is; the result is void.

Conclusion

In 2026, the challenge of Step 3 is not just knowing the definition of Sensitivity. It is understanding that a test with 99% sensitivity is dangerous in the wrong population. It is realizing that a "survival benefit" in a screening study might just be Lead-Time bias. It is looking at a glossy drug ad and calculating the NNT in your head before prescribing.

Unsupervised practice requires you to be a skeptic. You must protect your patients not just from disease, but from bad data, unnecessary testing, and false hope. Master these biostatistical principles, and you master the logic of modern medicine.

Systems-Based Practice, Ethics & Safety

Module 1: Patient Safety

The Philosophy of Error: Systems vs. Individuals

Historically, medicine punished the individual for making a mistake. The "shame and blame" culture is obsolete. The 2026 standard dictates that human error is inevitable. Therefore, safety lies in designing systems that catch these errors before they reach the patient.

We analyze errors through two primary lenses: Retrospective (looking back) and Prospective (looking forward).

1. Root Cause Analysis (RCA)

The Retrospective Approach When a "Sentinel Event" occurs—an error resulting in death, permanent harm, or severe temporary harm—we stop. We do not ask, "Who is to blame?" We ask, "Why did the system allow this?"

RCA is a structured method used to identify the underlying factors of an adverse event. It is reactive. You perform an RCA after the tragedy.

The Swiss Cheese Model To understand RCA, visualize a stack of Swiss cheese slices.

Shutterstock

In this model:

The Holes: These represent latent weaknesses in the system (e.g., understaffing, a broken alarm, a confusing drug label).

The Hazard: A potential error trying to pass through.

The Accident: This only occurs when the holes in every slice align perfectly, allowing the hazard to reach the patient.

Clinical Example: A patient receives a lethal overdose of insulin.

Blame Approach: Fire the nurse who administered it.

RCA Approach: Why was the vial accessible? Why did the computer allow that dose? Why was the nurse fatigued (double shift)?

Solution: We find that the "holes" aligned—fatigue + similar packaging + software bypass. The fix is a system change (barcoding), not firing the nurse.

2. Failure Mode and Effects Analysis (FMEA)

The Prospective Approach FMEA is proactive. It is used before a new process is implemented. It asks, "What could go wrong?"

Imagine your hospital is buying new infusion pumps. You don't wait for a patient to get hurt. You assemble a team to simulate usage. You identify "failure modes" (ways it could break or be misused) and assign them a risk score.

Steps of FMEA:

Identify the process: New chemotherapy protocol.

List failure modes: Doctor writes wrong dose; Pharmacist mixes wrong concentration; Nurse sets wrong rate.

Prioritize: Which failure is most likely and most dangerous?

Redesign: Force the computer to reject non-standard doses.

3. Transitions of Care: The Danger Zones

The highest risk for error occurs when a patient moves. This includes admission, transfer between units (ICU to Floor), and discharge.

Medication Reconciliation This is the mandated process of comparing the patient's home medication list against admission, transfer, and discharge orders.

Common Error: A patient on Warfarin at home is admitted for pneumonia. The admitting resident forgets to order Warfarin. The patient develops a PE.

The Fix: "Med Rec" catches this omission by forcing a comparison at every step.