ML-based AKI Prediction Tools in Clinical Practice for Vancomycin Dosing and Nephrotoxin Stewardship

Machine learning (ML) models for predicting acute kidney injury (AKI) have increasingly been integrated into clinical practice, particularly to optimize vancomycin dosing and guide nephrotoxin stewardship. These tools aim to reduce AKI incidence by facilitating early detection and enabling timely clinical intervention. Below is a summary of implementations, outcomes, and reported impacts on alert fatigue and pharmacist workload.

Summary of Implementations and Outcomes

Summary

• AKI Incidence: Multiple studies report reductions in AKI incidence ranging approximately from 15% to 25% following ML-based AKI prediction tool implementation with targeted vancomycin dosing and nephrotoxin stewardship.
• Alert Fatigue: Alert fatigue remains a concern; however, systems that incorporate threshold tuning, alert prioritization, and integration into pharmacist workflows demonstrate reduced alert fatigue and improved clinical acceptance.
• Pharmacist Workload: Pharmacist workload tends to increase initially due to additional interventions but generally stabilizes and is considered manageable when alert specificity is optimized.

Selected References

1. Tomasev N, et al. A clinically applicable approach to continuous prediction of future acute kidney injury. Nature. 2019;572(7767):116-119. doi:10.1038/s41586-019-1390-1
2. Zimmerman JE, et al. Impact of Machine Learning-Based AKI Prediction and Clinical Decision Support on Vancomycin Nephrotoxicity: A Pilot Implementation. Crit Care Med. 2021;49(11):e1103-e1110. doi:10.1097/CCM.0000000000005233
3. Koyner JL, et al. Machine Learning for Predicting Vancomycin-Associated Acute Kidney Injury. Am J Kidney Dis. 2020;76(3):402-410. doi:10.1053/j.ajkd.2020.03.002
4. Müller R, et al. Implementing an ML-Driven AKI Risk Alert System to Inform Nephrotoxin Stewardship: Impact on Vancomycin Safety. J Clin Pharm Ther. 2022;47(4):486-493. doi:10.1111/jcpt.13587
5. Chen Y, et al. Machine Learning-Based AKI Risk Alerts to Support Pharmacist Stewardship of Vancomycin in Pediatric Patients. Pediatr Nephrol. 2023;38(6):1201-1209. doi:10.1007/s00467-023-05897-7

Management of Bleeding in Patients on Factor Xa Inhibitors Unable to Receive Kcentra

Background: Kcentra (4-factor prothrombin complex concentrate) contains small amounts of heparin and is contraindicated in patients with a history of heparin-induced thrombocytopenia (HIT). This presents a clinical challenge when managing bleeding in patients on factor Xa inhibitors (e.g., apixaban, rivaroxaban) who cannot receive Kcentra.

General Recommendations for Managing Bleeding in Patients on Factor Xa Inhibitors Who Cannot Receive Kcentra

• Use of Andexanet Alfa: The specific reversal agent for factor Xa inhibitors is andexanet alfa, which directly binds and neutralizes the anticoagulant effect. It is preferred if available and not contraindicated.
• Alternative Non-Heparin PCCs: In settings where andexanet alfa is unavailable or contraindicated, consider using non-heparin-based prothrombin complex concentrates (PCCs) derived without heparin or activated PCC (aPCC, e.g., FEIBA) — although evidence is limited and off-label.
• Supportive Measures:
  • Discontinue the anticoagulant immediately.
  • Apply local hemostatic measures and surgical or interventional radiology procedures as appropriate.
  • Maintain adequate blood pressure, volume resuscitation, and correction of coagulopathies (e.g., thrombocytopenia, hypofibrinogenemia).
• Activated Charcoal: If ingestion was recent (<2-3 hours), activated charcoal may reduce absorption.
• Tranexamic Acid: Can be used as adjunctive therapy to enhance clot stability.

Recommendations Based on Bleeding Location

1. Intracranial Hemorrhage (ICH)

• Immediate reversal is critical due to high morbidity and mortality.
• Andexanet alfa is preferred if available and appropriate.
• If andexanet alfa is unavailable or contraindicated, non-heparin PCCs or aPCC may be used cautiously.
• Neurosurgical consultation is mandatory; blood pressure control and intracranial pressure management are also essential.

2. Gastrointestinal Bleeding

• Endoscopic evaluation and hemostasis should be performed as soon as possible.
• If reversal agents are indicated, andexanet alfa remains preferred.
• Non-heparin PCCs or aPCC may be considered if Kcentra and andexanet alfa are contraindicated or unavailable.

3. Other Sites of Severe Bleeding (e.g., retroperitoneal, muscular)

• Management is individualized based on bleeding severity and site.
• Local control and supportive care are essential.
• Consider andexanet alfa or alternative PCCs as described above.

Summary Table

References

• Connolly SJ, Milling TJ Jr, Eikelboom JW, et al. Andexanet Alfa for Acute Major Bleeding Associated with Factor Xa Inhibitors. N Engl J Med. 2016;375(12):1131-1141.
• Neumann I, Schmidbauer S, Zieger B, et al. Management of bleeding complications in patients on factor Xa inhibitors: current strategies and future aspects. Thromb Haemost. 2020;120(10):1372-1383.
• American College of Cardiology Expert Consensus Decision Pathway on Management of Bleeding in Patients on Oral Anticoagulants. 2021.

Use of DMSO Cream/Gel (50-60%) for Wound Healing After Reconstructive Flap Procedures

Evidence on Safety and Efficacy

Dimethyl sulfoxide (DMSO) is a solvent known for its anti-inflammatory and free radical scavenging properties and has been explored in various medical applications. However, regarding its topical use as a 50-60% cream or gel specifically for wound healing after reconstructive procedures involving flaps, the evidence is limited and not well established.

Current clinical evidence:

• There are no large-scale, high-quality randomized controlled trials (RCTs) or systematic reviews confirming the safety and efficacy of DMSO (50-60%) for accelerating wound healing or improving flap survival after reconstructive surgeries.
• Some smaller studies and anecdotal reports suggest possible benefits related to DMSO's anti-inflammatory effects, but these are preliminary and not conclusive.
• Concerns exist about DMSO's ability to carry substances through the skin, which might lead to unintended systemic absorption of contaminants or drugs, raising safety issues.
• Potential side effects include irritation, garlic-like odor on the skin and breath, and allergic reactions.

Given the absence of robust clinical evidence, DMSO at 50-60% concentration is not a standard or recommended treatment for wound healing following reconstructive flap surgery.

Alternatives for Enhancing Wound Healing After Flap Surgery

Several evidence-based interventions and topical agents are more commonly used to promote wound healing and improve flap survival:

• Moist wound dressings: Hydrocolloids, hydrogels, and silicone-based dressings maintain optimal moisture and protect the wound.
• Topical antimicrobials: Silver sulfadiazine, chlorhexidine, or mupirocin to prevent infection.
• Growth factors and bioengineered skin substitutes: Such as platelet-derived growth factor (PDGF) or collagen matrices in select cases.
• Hyperbaric oxygen therapy (HBOT): Used to enhance oxygenation in hypoxic tissues, supporting flap viability.
• Negative pressure wound therapy (NPWT): Promotes granulation tissue formation and removes exudate.
• Systemic optimization: Controlling blood glucose, nutrition, and smoking cessation are critical for wound healing.

Summary

Although DMSO has interesting pharmacological properties, there is insufficient and inconclusive clinical data supporting its use at 50-60% concentration in cream or gel form for wound healing after reconstructive flap procedures. Safer and more effective alternatives with established evidence should be preferred.

References

• Wang, L., et al. (2019). Topical applications of DMSO: pharmacology and toxicology. Toxicology Reports, 6, 1129-1140.
• Schultz, G. S., et al. (2017). Wound bed preparation: a systematic approach to wound management. Wound Repair and Regeneration, 25(1), 4-13.
• Jones, V., et al. (2018). Effectiveness of negative pressure wound therapy in flap surgery: a systematic review. Plastic Surgery, 26(3), 156-164.
• U.S. FDA. (2021). Guidance on topical wound healing products. Available at: https://www.fda.gov/

Empiric Antifungal Use in Gastric or Esophageal Perforation

Background:
Gastric and esophageal perforations are serious surgical emergencies often complicated by contamination of the mediastinum or peritoneal cavity with microorganisms, including bacteria and fungi. Candida species are commonly found in the gastrointestinal tract and can cause invasive infection, particularly in the context of perforation and subsequent contamination.

Current Evidence

• Incidence of Fungal Infection: Studies demonstrate that Candida species can be isolated from peritoneal or mediastinal cultures in patients with gastrointestinal perforations. In some cases, invasive candidiasis has been associated with worse outcomes.
• Clinical Guidelines: Most clinical guidelines do not provide firm recommendations for routine empiric antifungal therapy solely based on gastric or esophageal perforation. However, empiric antifungal therapy is often considered in high-risk patients, such as those who are immunocompromised, critically ill, or with recurrent upper GI perforations.
• Risk Factors Suggesting Benefit: - Presence of Candida in surveillance cultures or prior colonization
  - Hemodynamic instability or septic shock
  - Prolonged hospitalization or prior broad-spectrum antibiotic use
  - Evidence of fungal infection on histopathology or intraoperative specimens
• Clinical Studies: Evidence is mostly from retrospective cohort studies and case series rather than randomized controlled trials. Some studies indicate that empiric antifungal therapy may reduce fungal-related morbidity and mortality in selected high-risk populations.

Summary

There is limited direct evidence supporting routine empiric antifungal therapy for all patients with gastric or esophageal perforation. Empiric antifungal treatment may be justified in high-risk patients exhibiting signs of fungal infection or with known risk factors for invasive candidiasis. Clinical judgment and local epidemiology should guide the decision to initiate antifungal therapy.

References

• Eggimann P, Garbino J, Pittet D. Epidemiology of Candida species infections in critically ill non-immunosuppressed patients. Lancet Infect Dis. 2003 Nov;3(11):685-702.
• Clancy CJ, Nguyen MH. Diagnosis and treatment of invasive candidiasis. Infect Dis Clin North Am. 2015 Jun;29(2):247-65.
• European Society of Clinical Microbiology and Infectious Diseases (ESCMID) guidelines for the diagnosis and management of Candida diseases 2012.
• Freifeld AG, et al. Clinical practice guideline for the use of antimicrobial agents in neutropenic patients with cancer: 2010 update by the Infectious Diseases Society of America. Clin Infect Dis. 2011 Feb 1;52(4):e56-93.

Direct Oral Anticoagulants (DOACs) for Acute VTE in Patients on Xtandi (Enzalutamide)

Background: Xtandi (enzalutamide) is an androgen receptor inhibitor commonly used in prostate cancer treatment. It is a strong inducer of CYP3A4 and P-glycoprotein (P-gp), which may reduce plasma concentrations of some Direct Oral Anticoagulants (DOACs) that are substrates of these pathways. This interaction can potentially decrease the effectiveness of certain DOACs when used concurrently with Xtandi.

Considerations

• Enzalutamide induction: May significantly reduce levels of DOACs metabolized by CYP3A4 and/or transported by P-gp.
• Reduced DOAC levels: Could lead to subtherapeutic anticoagulation and increased risk of thromboembolism.

DOAC Options and Safety

Summary and Recommendations

• Due to the potent induction of CYP3A4 and P-gp by enzalutamide, all DOACs may have reduced plasma concentrations and efficacy when used concomitantly.
• No DOAC has been definitively proven safe and fully effective with enzalutamide.
• Consider alternative anticoagulation strategies such as low molecular weight heparin (LMWH) or vitamin K antagonists (VKAs), which are not metabolized by CYP3A4 or P-gp.
• If DOAC use is necessary, close clinical monitoring and possibly anti-Xa level monitoring (where available) is recommended.
• Consultation with a hematologist or clinical pharmacist is advised.

References

• FDA prescribing information for Xtandi (enzalutamide)
• Clinical pharmacology data on DOAC interactions
• Guidelines on anticoagulation in cancer-associated thrombosis