USMLE Step 3

Patient Management Skills

Minimum score to pass	184/75
# of questions	480 multiple-choice and 9 clinical simulations (all computerized)
Duration	16 hours ( 2 days)
Additional information	Most states require passing scores on Step1, Step2CK and Step 2CS before taking Step3. Sixteen state borads allow IMGs to take USMLE Step3 before beginning their residency program.

Test Overview

The USMLE* Step 3 is the third in the series of three USMLE exams that candidates must pass in order to qualify for medical licensure in the United States. It is a 2-day computerized examination with approximately 480 multiple-choice items administered over the first day and a half of testing. These items deal with clinical situations typically encountered by a generalist physician. The exam concludes with a half-day consisting of approximately 9 open-ended computer-based case simulations (CCS).

The Step 3 exam focuses primarily on patient management (45-55% of the questions). Other "physician tasks" tested include history and physical (8-12%), lab and diagnostic studies (8-12%), diagnosis (8-12%), prognosis (8-12%), and pathophysiology (8-12%). In addition to "physician tasks," each question also falls into one of three clinical encounter frames: initial workup (20-30%), continued care (55-65%), and emergency care (10-20%).

Multiple Choice Sections

The multiple-choice part of the test is divided into blocks of 25-50 items. You will have 30-60 minutes to complete each block. Detailed clinical situations are provided for each item and are frequently accompanied by pictorial material and laboratory data. Unlike Steps 1 and 2, in which questions in a block do not relate to each other in any way, the questions in a Step 3 block all belong to the same clinical setting. Items in each block represent one of four settings in which you could encounter a patient:

1. Satellite health center
2. Physician's office
3. Hospital
4. Emergency department

At the beginning of each block, you are told which clinical setting you are dealing with and are given information about what sort of resources are available in that setting.

Computer-based Case Simulations (CCS)

Computer-based case simulations (CCS) are a new testing format on the Step 3 exam and comprise the last 3 hours and 45 minutes of testing. Multiple patient cases are used to assess your ability to evaluate history and physical examination information, order diagnostic tests, select initial therapies, and manage the patient as he/she responds to your clinical interventions. The CCS format assesses your clinical judgment in a more "realistic" way than multiple-choice questions do.
The Step 3 patient simulations are delivered using NBME's PrimumTM Software, which allows simulation of physician/patient encounters in a variety of settings. You manage the case without prompting, typing in a variety of diagnostic and treatment options. The software recognizes thousands of possible entries that you can type into the order sheet and is programmed with results for any intervention or test you choose.

As you move the case clock forward, you receive the results of the tests you ordered and the patient's status changes based on the underlying medical problem and in response to the specific management decisions made. Acute cases may need to be managed in a short period of case time, whereas patients with chronic problems will require management over months of simulated time.

Question Formats

There are three question formats used in the multiple-choice sections of Step 3:

Single Best Answer

These items, which are the most commonly used, typically have four or five answer choices. You will be asked to select the one best response. Some negatively phrased items are used. In these items, the negative word or phrase is capitalized for emphasis (e.g., LEAST, NOT, EXCEPT).

Multiple Item Sets

In multiple item sets there is a vignette followed by several questions. The items are designed to be answered independently of one another and focus on different aspects of the case presentation.

Case Clusters

In case clusters, as in multiple item sets, several questions all relate to an initial vignette. However, in case clusters, new information is added as the case unfolds over time from item to item. It is imperative that you answer the questions in the order given since the case changes from one item to the next; you need to answer each item based only on the information revealed about the case up to that point.

Scoring

Scoring of the multiple-choice sections is based on the total number of questions answered correctly. Scoring of the CCS is based on the actions taken, their sequencing and timing, and whether the patient was subjected to unnecessary testing or therapy or placed at serious risk as a result of your action or failure to act.

The USMLE Step 3 is scored and reported on two-digit and three-digit scales. Your score report will also include a pass/fail designation, as well as a graphical performance profile. The new minimum passing score for Step 3 is 184. It will be applied to Step 3 examinations for which the first day of testing is on or after April 1, 2004. This is based on a standardized scale where the mean is defined as 200 and the standard deviation is 20 points. To achieve a passing score, you typically need to answer between 55% and 65% of the questions correctly (the percentage required to pass can vary based on difficulty of your individual test "form"). It is anticipated that, at least initially, performance on the CCS cases will count for LESS than its percentage of testing time (i.e., less than 25%).

With the new computer exam, you should receive your score within 2-4 weeks of your test date. This did not occur for the first group of Step 3 test-takers (Nov '99-Spring '00), as the USMLE needed to accumulate sufficient data before they could release the first set of scores. Those candidates had to wait as long as 6 months to receive their results. Also, please keep in mind that Step 3 is administered through the individual state medical boards and that score reporting procedures can vary by state.