Probability & Decision Theory

Bayesian Updater

First-Attempt Success Probability

Prior probability (base rate) 82%

Population-level first-attempt PIV success ~75–90%

Vein visibility / palpability Visible & palpable

Strongest single predictor of first-attempt success

BMI / subcutaneous fat Normal

Obesity increases depth, reduces tactile feedback

Prior failed attempts (this encounter) 0

Each failed attempt collapses usable veins and increases vasospasm

History of difficult access None

Documented DIVA / DAWA status

Clinician experience level Experienced RN

Operator skill modifies all other risk factors

—

Posterior P(first-attempt success)

Prior (base rate)—

Vein LR—

BMI LR—

Failed attempts LR—

History LR—

Experience LR—

Posterior—

Interactive Decision Tree

Vascular Access Escalation

Monte Carlo Simulation

500 Simulated Patient Encounters

—

PIV success

—

US-guided

—

Midline placed

—

PICC required

—

CVC required

—

Avg attempts

Multi-Criteria Decision Analysis

Expected Utility by Access Strategy

Adjust outcome weights to match patient context. EU = Σ P(oᵢ) × U(oᵢ) normalized across criteria.

Success weight40%

Speed weight25%

Safety weight20%

Patient comfort15%

Strategy	Success	Speed	Safety	Comfort	EU Score

🎯

The 2-attempt rule

After 2 failed PIV attempts, posterior P(success) drops below 50% for most operators. Escalate to US guidance or senior clinician — evidence supports this threshold.

📐

Bayes at the bedside

Prior probability isn't fixed — it updates with every observation. Bruising, small veins, prior chemo, IV drug use: each is a likelihood ratio that shifts your estimate.

⚖️

Utility is not symmetrical

A pneumothorax from a CVC has far greater negative utility than the inconvenience of ultrasound setup. Decision theory formalizes why we accept worse expected value to reduce variance.

🎲

Independence assumption

Bayesian models assume conditionally independent predictors. In practice, obesity, poor visibility, and difficult history correlate — so naïve Bayes underestimates risk. Calibrate accordingly.

How This Mathematics Reached the Bedside

1960s–1990s

Anesthesiology & Critical Care

Development of central venous catheters and flow-rate optimization created the first formal framework for access strategy as a clinical decision problem.

1980s–2000s

Emergency Medicine & Trauma

The "two large-bore IV" doctrine. Rapid infusion systems. Emergency medicine formalized gauge selection as a physics problem — not just a preference.

~2000s → Now

Ultrasound-Guided Access Revolution

Real-time visualization made the geometry measurable. Diameter, depth, compressibility — suddenly the inputs to every equation in this collection are available at the bedside.

The Big Insight · Why All Five Modules Connect

Vascular access is an optimization problem under uncertainty
constrained by fluid physics and human anatomy.

Every needle stick requires simultaneous reasoning across four mathematically distinct domains. The clinician is running a multi-variable optimization in real time — this collection makes that reasoning explicit, and Intracav's platform makes it computable.

Q = πr⁴ΔP / 8ηL

Flow Rate

Gauge selection constrains throughput before insertion.

τ = 4ηQ / πr³

Damage Risk

Every mL/hr carries a shear signature the endothelium reads.

P(A|B) = P(B|A)P(A)/P(B)

Success Probability

Patient anatomy updates first-attempt odds in real time.

EU = Σ P(oᵢ) · U(oᵢ)

Clinical Urgency

Time pressure reshapes expected utility at every acuity level.

Knowing when toescalate your approach

Knowing when to
escalate your approach