Validation of the MIRACLE 2 Score for Prognostication After OOHCA

Validation of the MIRACLE2 Score for Prognostication After Out-of-Hospital Cardiac Arrest

Sunderland. @Nicksunderland5 2023. Interventional Cardiology; 18:e29. doi.org/10.15420/icr.2023.08

Clinical Question

In patients with out-of-hospital cardiac arrest (OOHCA), how does the MIRACLE2 score compare with previously reporting scoring systems, for predicting poor neurological outcome at hospital discharge?

Background

Out-of-hospital cardiac arrest has an ~8% survival to hospital discharge
In those patients admitted to intensive care following an OOHCA ~50% survive, with the leading cause of death being hypoxic brain injury
Current guidelines recommend prognostication after 72 hours following a cardiac arrest
The MIRACLE2 score was developed to provide early prediction of neurological outcomes
This study aimed to externally validate the MIRACLE2 score and compare it with other scoring systems

Design

Retrospective cohort study
Single centre
All admissions screened for cardiac arrest ICD-10 codes
Records were then manually reviewed
No sample size calculation

Setting

Single tertiary cardiac centre, UK, with 24 hour access to coronary angiography cardiac surgery and specialist intensive care services
Dates of data collection: January 2019 – July 202

Population

Inclusion:
- Adult patients
- Out-of-hospital cardiac arrest of presumed cardiac origin who had return of spontaneous circulation
Exclusion:
- Died before hospital arrival
- Evidence of non-cardiac cause of arrest (respiratory arrest, suicide, trauma, drowning, substance overdose)
- Confirmed intra-cerebral haemorrhage
- Baseline neurological disability with Cerebral Performance Category (CPC) 3 or 4
- Co-morbidity leading to life expectancy <6 months
Data from 682 patients analysed
- Exclusions
  - Did not relate to new presentation of OHCA (n=321)
  - Obvious non-cardiac cause for arrest (n=40)
  - Dead on arrival (n=2)
  - Paediatric (n=33)
  - Life expectancy <6 months or significant prior neurological disability (n=11)
  - Did not initially present to study centre ie. transferred in (n=28)
  - No medical information available (n=35)
- After exclusions 219 patients included in validation cohort
- Complete data for calculating MIRACLE2 score available for 77% of patients. Missing data for pupillary response in 18% and initial pH in 6%
Comparing baseline characteristics of patients that had good (CPC 0-2) and poor neurological outcome (CPC 3-5) at hospital discharge
- Age: 60 vs. 71
- Male: 83% vs. 70%
- hs-cT (ng/l): 173 vs. 264
- ST elevation: 39% vs. 23%
- LV EF: 50% vs. 45%
- Witnessed arrest: 88% vs. 81%
- Zero-flow time (min), median (range): 0 (0-7.5) vs. 0 (0-20)
- Low-flow time (min), median (range): 17 (9-26) vs. 32 (20-45)
- Shockable rhythm: 96% vs. 57%
- Changing rhythms: 11% vs. 55%
- Any adrenaline dose: 32% vs. 92%
- Initial pH: 7.20 vs. 7.26
- Reactive pupils: 94% vs. 64
- Lactate 3.9 vs. 6.1mmol/l
- GCS motor score of 1: 57% vs. 93%

Test of Interest

MIRACLE2 score – 7 variables with maximum score of 10 points
- Unwitnessed cardiac arrest (1 point)
- Non-shockable initial rhythm (1point)
- Changing rhythm (any 2 of VF, PEA, or asytole): 1 point
- Any adrenaline dose: 2 points
- No pupil reactivity at ROSC: 1 point
- Initial blood pH <7.2: 1 point
- Age
  - ≤60 years: 0 points
  - 61-80 years: 1 point
  - >80 years: 2 points (in this paper it states that age >80 years gives a score of 2 points. This may be an error as the maximum total score would be 9 if given 2 points for age >80; and in the original derivation study age >80 was given a score of 3 points.)

Control Tests

Cardiac Arrest Hospital Prognosis (CAHP) score
Out-of-Hospital Cardiac Arrest (OHCA) score
Target Temperature Management (TTM) score

Outcome

Primary outcome: poor neurological recovery, defined as CPC 3-5 (severe disability to death) at hospital discharge
Area under receiver operating curve for predicting poor neurological outcome at discharge
- MIRACLE 2: 0.93
- CAHP: 0.92
- OHCA: 0.83
- TTM: 0.94
Good vs poor neurological outcome
- Low MIRACLE 2 score: ≤2, (n=89): 94% vs 6%
- Intermediate MIRACLE 2 score: 3-4 (n=66): 53% vs 47%
- High MIRACLE 2 score: ≥5 (n=64): 5% vs 95%
MIRACLE2 score ≥5 predicted poor neurological outcome with a
- Specificity 97.5% (95% CI 93-99.5%)
- Sensitivity 62.3%
- +ve Likelihood Ratio 25.6 (95% CI 8.3-79)
- -ve Likelihood Raio 0.38
MIRACLE2 score ≥3 predicted poor neurological outcome with a
- Specificity 68.9%
- Sensitivity 94.9% (95% CI 88.4-98.3%)
- +ve Likelihood Ratio 3.05
- -ve Likelihood Ratio 0.07 (95% CI 0.03-0.18)
Secondary outcomes:
Area under curve for predicting poor neurological outcome at 6 months
- MIRACLE 2: 0.93
- CAHP: 0.92
- OHCA: 0.86
- TTM: 0.94

Authors’ Conclusions

“The MIRACLE2 score demonstrates good prognostic performance and is easily applicable to cardiac-origin OHCA presentation at the hospital front door. Prognostic scoring may assist decision-making regarding early angiographic assessment.”

Strengths

Screening of all admissions
Compared with 3 other scoring systems

Weaknesses

Only 64 patients included with MIRACLE2 score ≥5
Single centre
Retrospective
Number of patients had missing data

The Bottom Line

The MIRACLE2 score reported high accuracy for determining neurological prognostication at admission. However due to this being a single centre retrospective validation cohort with only small numbers of patients that had a high MIRACLE 2 score, and >20% of patients having missing data, decisions regarding treatment or escalation should not currently be made solely on the MIRACLE 2 score at admission

External Links

Metadata

Summary author: @davidslessor
Summary date: 11/03/24
Peer-review editor: Jonarthan Thevanayagam

Photo by Milad Fakurian on Unsplash

Cardiac Arrest Cardiology Featured Neurology