GuiDeCode
Calcs › ⚕ Critical Care / ICU

GO-FAR Score for In-Hospital CPR Outcomes

Predicts survival with favorable neurological outcome after in-hospital cardiac arrest treated with CPR

Calculator

Loading calculator…

Why use it

Survival rates after in-hospital CPR are far lower than commonly perceived by patients and families (median survival ~20%). GO-FAR provides an individualized, evidence-based estimate of CPR survival probability, enabling more informed advance care planning and goals-of-care discussions. It is particularly useful for identifying patients for whom CPR has a very low probability of benefit, facilitating discussions about DNAR/DNR orders, palliative care, and alternative goals of care.

About

The GO-FAR (Good Outcome Following Attempted Resuscitation) Score was developed by Ebell and colleagues to predict the likelihood of survival to discharge with favorable neurological outcome following in-hospital cardiac arrest (IHCA) treated with CPR. Published in JAMA Internal Medicine in 2013, GO-FAR was derived from the Get With The Guidelines-Resuscitation (GWTG-R) registry including over 51,000 patients. The tool uses 13 pre-arrest clinical variables that can be assessed at any time before the cardiac arrest event. GO-FAR scores range from -15 to +28, with higher scores predicting worse outcomes. Four risk categories are defined, with survival rates ranging from approximately 27% (Very Good prognosis group) to less than 1% (Very Poor prognosis group). GO-FAR is used primarily as a communication and shared decision-making tool in goals-of-care discussions about CPR preferences.

When to use

  • Goals-of-care conversations about CPR with patients or family members on hospital admission
  • Identifying patients for whom the probability of CPR success is very low
  • Informing advance directive discussions in seriously ill inpatients
  • Multidisciplinary team discussions about appropriate resuscitation status
  • ICU or oncology ward care planning discussions
  • Not to be used to unilaterally withhold CPR — only as a decision-support and communication tool

Formula

GO-FAR Score (sum of applicable items): Age: 70–74 years: +1 75–79 years: +2 80–84 years: +5 ≥85 years: +6 Pre-arrest condition: Non-cardiac admitting diagnosis: +7 Metastatic cancer or stage IV solid organ malignancy: +7 Hepatic insufficiency (cirrhosis or bilirubin >2 mg/dL): +7 Septicemia (infection + ≥2 SIRS criteria before arrest): +7 Medical non-cardiac ICU admission: +7 Hypotension or hypoperfusion (24h before arrest, not treated): +5 Renal insufficiency (creatinine >1.5 mg/dL, not dialysis): +4 Respiratory insufficiency (SpO2 <90% or pO2 <60 before arrest): +4 Pneumonia as admitting diagnosis: +4 No neurological impairment at baseline: -15 Acute stroke as admitting diagnosis: +4 Hematological malignancy: +4 Score Interpretation: ≤ -5: Very Good (~27% survival with favorable neuro outcome) -4 to 0: Good (~12%) 1 to 5: Average (~6%) 6 to 13: Poor (~3%) ≥ 14: Very Poor (<1%)

Pearls & pitfalls

  • GO-FAR predicts survival WITH GOOD NEUROLOGICAL OUTCOME — not just return of spontaneous circulation (ROSC). Overall ROSC rates are higher but mean little without functional recovery.
  • The score was derived in the Get With The Guidelines-Resuscitation database — a US registry; generalizability to other settings (non-US, community hospitals) may vary.
  • GO-FAR is a PRE-ARREST tool — it is used BEFORE cardiac arrest occurs to guide advance care planning, not to triage resuscitation after arrest begins.
  • Even patients in the "Very Good" category (Score ≤-5) have only a ~27% survival rate — all CPR discussions should convey realistic expectations.
  • GO-FAR should NEVER be used to unilaterally withhold resuscitation — it is a communication and shared decision-making tool, not a mandate.
  • Age alone is a strong predictor — patients ≥70 years score higher (worse prognosis) than younger patients with similar clinical profiles.
  • Non-shockable rhythms (PEA/asystole) have substantially worse outcomes than ventricular fibrillation/pulseless VT — rhythm at arrest is not in GO-FAR but is critical.
  • Metastatic cancer adds the highest single point weight in GO-FAR (+11), reflecting the dramatically poor prognosis of IHCA in this population.

Next steps

Very Good (Score ≤-5): ~27% survival with favorable neurological outcome. Discuss realistic expectations — CPR may be beneficial. Good (Score -4 to 0): ~12% survival. Review CPR preferences with balanced discussion of risks and benefits. Average (Score 1–5): ~6% survival. Emphasize low survival probability. Explore patient values and goals. Initiate palliative care if not already involved. Poor (Score 6–13): ~3% survival. High probability of CPR futility. Comprehensive goals-of-care discussion. Consider DNAR. Very Poor (Score ≥14): <1% survival. CPR very unlikely to result in meaningful recovery. Palliative care consultation strongly recommended. DNAR discussion essential.

Evidence

Ebell MH et al. The GO-FAR score to predict survival from in-hospital cardiac arrest. JAMA Intern Med. 2013;173(20):1872-1878. PMID:24018585. Derivation cohort: 51,240 patients (GWTG-R registry). Validation cohort: 25,644 patients. C-statistic 0.74. External validation: Minhas MA et al. Am J Med 2016. Survival rates: Very Good 27%, Good 12%, Average 6%, Poor 3%, Very Poor <1%.

References

  1. Ebell MH et al. The GO-FAR score to predict survival from in-hospital cardiac arrest. JAMA Intern Med. 2013;173(20):1872-1878. — link
  2. Minhas MA et al. External validation of the GO-FAR score to predict survival from in-hospital cardiac arrest. Am J Med. 2016;129(4):408-415. — link
  3. Girotra S et al. Trends in Survival after In-Hospital Cardiac Arrest. N Engl J Med. 2012;367:1912-1920. — link