Bedside decision support for discharge risk assessment, triggered reassessment after sentinel events, and extended prophylaxis reference guidance.
Version 2026-03-30 v3 | Locked manuscript models + bedside simplification | N = 4,856,597
PE discharge AUC 0.811 | PE triggered reassessment AUC 0.892 | DVT discharge AUC 0.814 | VTE discharge AUC 0.811
Risk estimates for CLINICAL DECISION SUPPORT only. NOT a substitute for clinical judgment. All decisions must incorporate individual patient factors, contraindications, bleeding risk, and clinical context.
Data: ACS-NSQIP PUF 2020-2024 (4.86M cases). Outcomes: postoperative PE, postoperative DVT, and composite VTE within 30 days. Model: locked logistic regression models with 2024 temporal validation (not FDA-cleared).
Public-use notice: This public calculator is intended for educational and clinical decision-support use. Do not enter direct patient identifiers such as name, date of birth, medical record number, address, phone number, or free-text notes.
1. Search the operation by CPT or plain-language name. 2. Enter the bedside variables available at discharge. 3. Use the PE estimate as the main decision signal; DVT and VTE are secondary context. If the patient is later readmitted or reoperated, move to the Reassessment tab. Do not enter patient identifiers. Not FDA-cleared.
Every patient receives this discharge risk estimate. If the patient is later readmitted or undergoes reoperation, proceed to the Triggered Reassessment tab for an updated estimate that accounts for these sentinel events and their timing from the index operation.
Switch to "Triggered Reassessment" tab if any post-discharge event occurs:
Recalculates PE risk after sentinel post-discharge events. Readmission timing is the single strongest risk signal (OR up to 21.8). Reoperation timing changes risk further.
Model: Validated AUC = 0.892 in 2024 temporal validation. Uses the same locked baseline/procedure backbone as the discharge PE model plus readmission and reoperation timing.
Public-use notice: This public calculator is intended for educational and clinical decision-support use. Do not enter direct patient identifiers or free-text patient information.
Use this tab only after a new post-discharge sentinel event, especially readmission or reoperation. It reassesses the PE estimate after the clinical course changes and should not replace the original discharge calculation. Do not enter patient identifiers. Not FDA-cleared.
This model is used only after a sentinel event — unplanned readmission or reoperation. It reassesses PE risk using the same patient and procedural variables plus sentinel-event timing. Among patients who experience a sentinel event, 42% cross the 0.5% risk threshold for the first time — patients who would not have been flagged at discharge. Early readmission (POD 0-3) carries an OR of 21.8 for PE, representing substantial risk amplification.
The reassessment model captures the substantial risk amplification from post-discharge sentinel events. Readmission within POD 0-3 carries an OR of 21.8 for PE.
Patients with early readmission + reoperation represent the highest-risk subgroup and should receive extended prophylaxis unless actively bleeding.
Risk reclassification: Compare this reassessed risk to the original discharge estimate. Patients below the 0.5% threshold at discharge who cross above it after a sentinel event represent newly identified high-risk patients who may benefit from intensified VTE prevention.
Based on the calculated risk, extended-duration pharmacologic prophylaxis should be strongly considered. Review the options below, weigh bleeding risk, and individualize the decision.
This tab is intentionally organized for fast clinical use. If the patient is high risk for PE and does not have a major bleeding contraindication, the simplest default path is enoxaparin 40 mg subcutaneously once daily for 28 days total, adjusted for renal function and possibly extreme obesity.
Extended-duration thromboprophylaxis (ETP) refers to pharmacologic VTE prevention continued beyond the inpatient stay, typically 28-35 days total. The strongest evidence supports LMWH (enoxaparin) after major abdominal/pelvic and cancer surgery. DOACs are emerging alternatives with oral convenience but variable trial results. Contemporary data suggest the central question is patient selection: the 2025 CASCADE study reported post-discharge VTE rates of only 0.1%, underscoring the importance of identifying which patients truly benefit from extended prophylaxis rather than treating broadly.
Timing of first prophylactic dose after surgery depends on procedure type, bleeding risk, and neuraxial anesthesia. Below are general guidance ranges — always defer to surgical team assessment of hemostasis.
| Surgery Type | LMWH (First Dose) | DOAC (First Dose) | Notes |
|---|---|---|---|
| Major abdominal/pelvic (colorectal, hepatobiliary, gynecologic oncology) | 6-12 hours postop | Day 1-3 postop (varies by trial) | ACCP: start LMWH 12h postop. Most RCTs began LMWH inpatient then extended. DOACs typically started after hemostasis confirmed. |
| Major orthopedic (TKA, THA, hip fracture) | 12 hours postop | 6-24 hours postop | Strong evidence for 28-35 day prophylaxis. ACCP Grade 1B. Both LMWH and DOACs well-studied. Rivaroxaban/apixaban FDA-approved for TKA/THA. |
| Bariatric surgery | 6-12 hours postop | 24-48 hours postop | Weight-adjusted dosing may be needed. Consider enoxaparin 40mg BID if BMI ≥40. Absorption of oral DOACs may be altered post-bypass — limited pharmacokinetic data. |
| Thoracic / lung resection | 12-24 hours postop | 24-48 hours postop | Higher bleeding risk at chest tube sites. Often hold until chest tube output <200 mL/day. Extended prophylaxis less well-studied in this population. |
| Neurosurgery / spine | 24-48 hours postop | Caution — limited data | Risk of epidural hematoma. Mechanical prophylaxis preferred initially. Pharmacologic start deferred 24-48h; some spine surgeons prefer 72h. Discuss with surgical team. |
| Vascular surgery | 12-24 hours postop | 24-48 hours postop | Depends on graft type and anastomotic integrity. Avoid if active oozing from surgical site. Many patients already on therapeutic anticoagulation for PAD. |
| Cardiac surgery (CABG, valve) | 24-48 hours postop | 24-72 hours postop | High baseline VTE risk but also high bleeding risk (mediastinal, pericardial). Extended prophylaxis rarely studied. Many patients transition to therapeutic anticoagulation for AF or mechanical valves. |
| After neuraxial anesthesia (epidural/spinal) | Hold LMWH ≥4h after catheter removal | Hold DOAC ≥6h after catheter removal | ASRA guidelines: remove epidural catheter ≥12h after last LMWH dose; next dose ≥4h after removal. For DOACs, wait ≥6h after removal before dosing. |
When pharmacologic prophylaxis is contraindicated, mechanical prophylaxis (intermittent pneumatic compression devices) should be used. Reassess daily for resolution of contraindication and convert to pharmacologic prophylaxis when safe.
Extended prophylaxis is a tradeoff: VTE reduction versus bleeding increase. The table below summarizes the risk-benefit data from major trials.
| Agent / Trial | VTE Reduction | Major Bleeding Increase | Net Assessment |
|---|---|---|---|
| Enoxaparin (EXCLAIM, medical) | ARR −1.53% (~1 VTE prevented per 65 treated) | +0.51% (~1 major bleed per 196 treated) | Favorable in Level 1 immobility; marginal in Level 2. ~2.5x bleeding increase. |
| Enoxaparin (LMWH ETP, CRC surgery pooled) | OR 0.18 (0.07-0.50) | OR 0.25 (0.03-2.24) at 90d — NS | Strong VTE benefit, no significant bleeding increase. Favorable risk-benefit in CRC/IBD surgery. |
| Rivaroxaban (MAGELLAN, medical) | RR 0.77 (0.62-0.96) | RR 2.5 (1.85-3.25) | Net clinical benefit NOT demonstrated. VTE benefit offset by significant bleeding harm. |
| Rivaroxaban (ETP, CRC surgery pooled) | OR 0.22 (0.08-0.67) | OR 2.68 (0.32-22.35) — NS, wide CI | Promising VTE reduction. Bleeding data underpowered. Use with monitoring. |
| Apixaban (ADOPT, medical) | RR 0.87 (0.62-1.23) — NS | RR 2.58 (1.02-7.24) | Failed efficacy + increased bleeding. NOT recommended for extended prophylaxis. |
| LMWH extended (THA/TKA, meta-analysis) | 80% relative risk reduction for DVT | No significant increase in major bleeding | Gold standard for extended prophylaxis after major joint replacement. ACCP Grade 1B. |
This section does not prescribe treatment. The information below summarizes published guideline and trial data for bedside reference.
All medication selection, timing, dose, and duration decisions remain clinician decisions and must account for bleeding risk, renal and hepatic function, drug interactions, and procedure-specific concerns.
Apply this framework after calculating the patient's PE risk using the Discharge or Triggered Reassessment tab:
Step 1 — Risk tier. If Low (<0.15%), extended prophylaxis is generally not indicated. If Moderate (0.15-0.34%), consider patient-specific factors. If High (≥0.35%) or Very High (≥1.0%), extended prophylaxis is strongly recommended unless contraindicated.
Step 2 — Bleeding assessment. Review absolute and relative contraindications above. If absolute contraindication present, use mechanical prophylaxis only and reassess daily.
Step 3 — Agent selection. First-line: enoxaparin 40 mg SC daily × 28 days (strongest evidence base). If LMWH contraindicated (HIT, refusal of injections): consider rivaroxaban 10 mg PO daily × 28-35 days. Avoid apixaban for extended prophylaxis (negative trial).
Step 4 — Patient education. Counsel on injection technique (if LMWH), signs/symptoms of DVT/PE, when to seek emergency care. Arrange 2-week follow-up or phone check-in.
Step 5 — Reassess at follow-up. If readmission or reoperation occurs, return to the Triggered Reassessment tab for recalculated risk and adjust prophylaxis accordingly.
Discharge tab: bedside PE, DVT, and composite VTE risk at discharge using clinically available variables, work RVU spline terms, and pooled CPT-3 family. Validated AUCs: PE 0.811, DVT 0.814, VTE 0.811.
Triggered Reassessment tab: PE-only triggered reassessment using the same baseline/procedure backbone plus readmission timing and reoperation timing. Validated AUC: 0.892.
This bedside calculator intentionally omits functional status to reduce data-entry burden. The locked manuscript models retained functional status, but its incremental discrimination gain was negligible in ablation testing. Calculator estimates therefore prioritize usability over exact reproduction of the full manuscript model.
These predictions estimate observed postoperative risk under real-world NSQIP-era care patterns, not untreated biologic thrombosis risk. Prophylaxis exposure is incompletely observed in NSQIP, so some patients who appear lower risk may reflect successful preventive care rather than intrinsically low thrombosis risk.
Use only for operative patients similar to those represented in ACS-NSQIP. This tool does not estimate bleeding risk and should support, not replace, clinician judgment about thromboprophylaxis, mobility, malignancy, and post-discharge surveillance.