Evaluation of end-tidal carbon dioxide as a predictor of return of spontaneous circulation in dogs and cats undergoing cardiopulmonary resuscitation

Abstract

Objective: To determine whether the partial pressure of end-tidal carbon dioxide (PetCO₂ ) could predict return of spontaneous circulation (ROSC) in patients with cardiopulmonary arrest (CPA) undergoing CPR.

Design: Prospective observational study.

Setting: Two private specialty referral hospitals.

Animals: Thirty-five client-owned dogs and cats in CPA in which CPR was performed and pertinent data recorded on a purpose-made form.

Interventions: None.

Measurements and main results: PetCO₂ was recorded at 1-minute intervals during CPR. Hospital, animal, arrest, and outcome variables were also reported in the Utstein style where possible. Twelve animals (7 dogs and 5 cats) achieved ROSC; 4 of these (2 dogs and 2 cats) had sustained ROSC, of which 1 dog was discharged alive. Patients that achieved ROSC had significantly higher initial PetCO₂ (P = 0.0083), peak PetCO₂ (P < 0.0001), average PetCO₂ (P < 0.0001), and ΔPetCO₂ (difference between last and first recorded PetCO₂ ; P = 0.0004) than patients not resuscitated. The PetCO₂ accurately discriminated between ROSC and failure to achieve ROSC at minutes 3, 4, 5, 6, 7, and 8 of CPR with area under the receiver operating characteristic curve of 0.926, 0.967, 0.938, 0.933, 0.956, and 1.00, respectively. The optimal cutoff PetCO₂ was 18 mm Hg (2.4 kPa), with a sensitivity of ≥80% and a specificity of ≥95% at minutes 3, 4, 5, 6, and 8, correctly classifying 91-100% of cases.

Conclusions: The results of this small study support previous recommendations to monitor PetCO₂ during CPR and suggest that PetCO₂ during CPR may be useful for determining the probability of ROSC. Absolute values and trends of PetCO₂ may assist clinicians and owners in making decisions for pets with CPA.

Keywords: BLS; CPR; basic life support; capnometry; cardiopulmonary arrest; monitoring.