Welcome back! In our previous lesson, we focused on recognizing the distinct laboratory "fingerprints" of major clinical syndromes like AKI, sepsis, and acute liver failure. You learned to synthesize results from multiple panels to identify a coherent pattern of organ dysfunction.
Today, we address the crucial next question: "Now what?" Once you've identified an abnormality or a syndrome, how do you decide what to do next? This lesson is all about the thought process behind creating a rational plan, moving from passive interpretation to active clinical investigation and management.
Our learning outcome is to formulate a diagnostic or monitoring plan based on initial or evolving laboratory abnormalities. We will explore the principles that guide whether you need to order more tests to find a diagnosis or to monitor a patient's progress. We'll conclude by dissecting a real-world case to see how an expert clinician builds and adapts her plan in real-time.
1. The Core Principles of Planning
Before diving into complex cases, let's establish some fundamental principles for acting on lab results. The best plans are built on a solid foundation of context, trends, and clear clinical reasoning.
Blood Test Interpretation Guide | Labs
The Geeky Medics guide, 'Blood Test Interpretation Guide', offers excellent practical advice that forms the bedrock of good clinical planning. It emphasizes moving beyond isolated numbers to a more holistic view.
Read the sections titled 'Consider the clinical context', 'Identify and act on critical results', 'Consider how the blood test helps inform the patient management', and 'Documentation'. Focus on the distinction between looking at a single result and assessing a trend, and how blood tests are used to either diagnose a problem or monitor a known one.
As the article highlights, your plan will generally fall into one of two categories:
- Diagnostic Plan: This is your strategy to investigate a new or unexplained finding. Your goal is to narrow down the differential diagnosis and arrive at a specific cause. For example, if a patient has new-onset anemia, your diagnostic plan might include ordering a reticulocyte count, iron studies, and vitamin B12/folate levels to determine the cause.
- Monitoring Plan: This is for tracking a known condition or a patient's response to treatment. The goal is to assess progression, detect complications, or see if an intervention is working. For example, for a patient with sepsis, your monitoring plan would involve trending serial lactate levels, CBCs (for platelets), and BMPs (for creatinine) to see if their organ dysfunction is resolving with antibiotics and resuscitation.
2. A Framework for Decision-Making: The Threshold Model
How do you decide when to order another test? Why not test for everything? Conversely, when is it safe to stop testing and just treat? These questions are at the heart of clinical decision-making. The "Threshold Model" provides a powerful theoretical framework for answering them.
The Threshold Model of Clinical Decision-Making (Strong Diagnosis)
The Strong Medicine video, 'The Threshold Model of Clinical Decision-Making', provides an excellent explanation of the mental model that clinicians use—often intuitively—to make decisions under uncertainty.
Watch the sections from 01:46 to 18:30. You don't need to memorize the exact percentages used in the example, but focus on understanding these key concepts: Test Threshold (or Rule-Out Threshold): The probability below which a disease is so unlikely that you won't test for it further. Treatment Threshold: The probability above which you are confident enough to start treatment, even if the diagnosis isn't 100% confirmed (this is 'empiric' treatment). Rule-In Threshold: The probability above which you are confident the patient has the disease and will stop pursuing other diagnoses. The goal of a diagnostic plan is to order tests that will move the probability of a disease across one of these thresholds.
Think of it this way: diagnostic tests are most useful when the diagnosis is uncertain.
- If the probability of a disease is very low (below the test threshold), ordering a test is likely to yield a false positive and lead you astray.
- If the probability is very high (above the treatment threshold), withholding treatment to do more testing may cause harm.
- The "sweet spot" for diagnostic testing is in the middle, in the zone of uncertainty. Your diagnostic plan should consist of tests that give you the most "bang for your buck" in moving the probability out of this zone.
Test your understanding!
A patient presents with symptoms that give them a 15% pre-test probability of having a pulmonary embolism (PE). This is above your "rule-out" threshold but below your "treatment" threshold. You order a D-dimer test, which comes back positive. The positive D-dimer increases the probability of PE to 40%.
- Is the diagnosis confirmed?
- What is the purpose of the next test in your diagnostic plan (a CT Pulmonary Angiogram)?
Show answer
- No, the diagnosis is not confirmed. A 40% probability is still in the "zone of uncertainty."
- The purpose of the CTPA is to definitively move the probability across a threshold. If it's positive, the probability will jump above the "rule-in" threshold (~95%+), confirming the diagnosis. If it's negative, the probability will drop below the "rule-out" threshold, effectively excluding the diagnosis.
3. Case Study: Formulating a Plan in Real-Time
Theory is useful, but seeing the process in action is invaluable. We are now going to watch an expert nephrologist work through a complex, undifferentiated case. As you watch, pay close attention to how each new piece of data—a lab result, an imaging finding—prompts her to formulate or change her diagnostic plan.
This is a masterclass in clinical reasoning. The patient is a 60-year-old male presenting with a non-specific complaint: "I itch all over."
Clinical Decision Making with Dr. Eleanor Lederer
In this video from the UofL Internal Medicine Lecture Series, Dr. Eleanor Lederer is presented with a case and walks the audience through her thought process. This is a perfect, real-world demonstration of our learning outcome.
Watch the case presentation and Dr. Lederer's step-by-step analysis from 03:17 to 17:59. For each new piece of information, pause and consider what your next step would be, then see what she does. Initial History (3:17-6:17): She gathers history and forms a very broad initial differential for pruritus. Initial Labs (6:17-10:51): The BMP reveals significant acute kidney injury (AKI). Notice how she immediately asks for a key test to investigate the AKI. What is it? Urinalysis (10:51-13:32): The urinalysis reveals significant proteinuria. This allows her to synthesize the findings and form a new, more specific hypothesis: nephrotic syndrome. Renal Ultrasound (15:29-17:59): The ultrasound reveals enlarged kidneys, which is an unexpected finding. This forces her to dramatically expand her differential diagnosis and, by extension, her diagnostic plan.
Let's break down Dr. Lederer's evolving plan:
- Initial Finding: Acute Kidney Injury (Creatinine 9.2).
- Her Plan: "I'd like to see a urinalysis." This is the first step in any AKI workup, to look for clues to the cause (protein, blood, casts).
- New Finding: Urinalysis shows 3+ protein.
- Her Plan: She synthesizes this with the low calcium (implying low albumin) and hypothesizes nephrotic syndrome. Her implicit plan is to confirm this with a 24-hour urine protein or urine protein:creatinine ratio.
- New Finding: Renal ultrasound shows enlarged kidneys.
- Her Plan: This finding changes everything. She recognizes that AKI with nephrotic syndrome and enlarged kidneys has a specific differential diagnosis. She explicitly lists it:
- Multiple Myeloma
- Amyloidosis
- Sarcoidosis
- HIV nephropathy
- Renal vein thrombosis
- Her diagnostic plan is now to order tests to investigate these possibilities (e.g., SPEP/UPEP for myeloma, HIV test, serologies for autoimmune disease).
- Her Plan: This finding changes everything. She recognizes that AKI with nephrotic syndrome and enlarged kidneys has a specific differential diagnosis. She explicitly lists it:
- Final Step: She ultimately concludes that despite all the non-invasive tests, the picture is still unclear.
- Her Final Plan: "I would have biopsied this guy." A kidney biopsy is the ultimate step in the diagnostic plan to get a definitive tissue diagnosis when imaging and serology are inconclusive.
This case perfectly illustrates the iterative nature of diagnostic planning. It's a cycle: Hypothesize -> Test -> Refine Hypothesis -> Test Again.
4. From Diagnosis to Monitoring
The case above focused on a diagnostic plan. Once a diagnosis is made, or a patient is being treated for a known condition, your focus shifts to a monitoring plan. Here, the goal is to use serial labs to track the patient's trajectory.
Many monitoring plans follow established algorithms or protocols. A classic example is the diagnostic algorithm for abnormal liver tests.
Here are some examples of common monitoring plans you will formulate:
-
Monitoring Heart Failure Exacerbation: After admitting a patient and starting diuretics, your monitoring plan would include:
- Daily BMP: To track creatinine (to watch for diuretic-induced AKI) and potassium/magnesium (which are lost with loop diuretics).
- Daily Weights: The most reliable indicator of fluid removal.
- Trending BNP: A falling BNP suggests the patient is responding to therapy.
-
Monitoring Diabetic Ketoacidosis (DKA): A patient on an insulin drip for DKA requires a very strict monitoring plan:
- Hourly blood glucose checks: To titrate the insulin drip.
- BMP every 2-4 hours: To monitor the anion gap (which should be closing), bicarbonate (which should be rising), and potassium (which will fall as insulin drives it into cells).
-
Monitoring Anticoagulation:
- For unfractionated heparin: Serial aPTT or anti-Xa levels every 6 hours until two consecutive therapeutic results, then daily.
- For warfarin: Daily INR until therapeutic and stable.
The key to a good monitoring plan is knowing what to expect. You must anticipate the potential complications of the disease and its treatment and order labs to watch for them proactively.
Conclusion
Today we've moved from simply recognizing abnormal labs to building rational plans based on them. You've learned the principles that guide when to test, what to test for, and how to use serial labs to monitor a patient's clinical course.
Key Takeaways:
- A diagnostic plan aims to resolve uncertainty by ordering tests to move the probability of a disease across a rule-in or rule-out threshold.
- A monitoring plan uses serial measurements to track a known disease, response to treatment, or to watch for anticipated complications.
- Expert clinical reasoning, as demonstrated by Dr. Lederer, involves an iterative cycle of forming a hypothesis based on available data, ordering a specific test to probe that hypothesis, and then refining the plan based on the new result.
- Standardized algorithms can guide planning for common problems, but complex cases require dynamic, hypothesis-driven investigation.
Preview of the Next Lesson:
We've learned how to formulate a plan. But what happens when a lab result is so dangerously abnormal that it requires immediate, protocol-driven action before any further diagnostic planning can occur? In our next lesson, we will focus on how to prioritize urgent clinical actions based on critical laboratory values, tackling medical emergencies like severe hyperkalemia, profound hypoglycemia, and critical INR elevation.