Prognostic models for outcome prediction following in-hospital cardiac arrest using pre-arrest factors: a systematic review, meta-analysis and critical appraisal

Machine learning algorithm for predicting 30-day mortality in patients receiving rapid response system activation: A retrospective nationwide cohort study

This study investigated the accuracy of a machine learning algorithm for predicting mortality in patients receiving rapid response system (RRS) activation. This retrospective cohort study used data from the In-Hospital Emergency Registry in Japan, which collects nationwide data on patients receiving RRS activation. The missing values in the dataset were replaced using multiple imputations (mode imputation, BayseRidge sklearn. linear model, and K-nearest neighbor model), and the enrolled patients were randomly assigned to the training and test cohorts. We established prediction models for 30-day mortality using the following four types of machine learning classifiers: Light Gradient Boosting Machine (LightGBM), eXtreme Gradient Boosting, random forest, and neural network. Fifty-two variables (patient characteristics, details of RRS activation, reasons for RRS initiation, and hospital capacity) were used to construct the prediction algorithm. The primary outcome was the accuracy of the prediction model for 30-day mortality. Overall, the data from 4,997 patients across 34 hospitals were analyzed. The machine learning algorithms using LightGBM demonstrated the highest predictive value for 30-day mortality (area under the receiver operating characteristic curve, 0.860 [95 % confidence interval, 0.825-0.895]). The SHapley Additive exPlanations summary plot indicated that hospital capacity, site of incidence, code status, and abnormal vital signs within 24 h were important variables in the prediction model for 30-day mortality.

Agonal breathing upon hospital arrival as a prognostic factor in patients experiencing out-of-hospital cardiac arrest

Background

Agonal breathing is a relatively common symptom that follows cardiac arrest when the brainstem function is preserved. Agonal breathing is associated with favorable survival in patients experiencing out-of-hospital cardiac arrest (OHCA). While previous studies focused on agonal breathing observed in the pre-hospital setting for all study subjects, we focused on agonal breathing observed upon hospital arrival. In this multicenter prospective study, we aimed to assess the prognosis of patients exhibiting agonal breathing upon hospital arrival were compared. We hypothesized that agonal breathing at hospital arrival would be associated with favorable neurological outcomes among patients with OHCA.

Methods

The data on incidence of agonal breathing were prospectively collected for all evaluable participants in a multicenter, observational study in Japan (SOS-KANTO [Survey of Survivors after Out-of-Hospital Cardiac Arrest in Kanto Area] 2017 Study). Groups with and without agonal breathing were compared upon hospital arrival. Propensity-score with inverse probability of treatment weighting (IPTW) analysis was performed to adjust for confounding factors. The primary outcome was a favorable neurological outcome (Cerebral Performance Category 1-2) at 1 month.

Results

A total of 6,457 participants out of the 9,909 registered in SOS-KANTO 2017 (in which 42 facilities participated) were selected for the current study. There were 128 patients (2.0%) in the with-agonal breathing group and 6,329 (98.0%) in the withoutagonal breathing group. The primary outcome was 1.1% in the with-agonal breathing group and 0.6% in the without-agonal breathing group (risk difference, 0.55; 95% confidence interval, 0.23-0.87) after IPTW analysis.

Conclusion

In this multicenter prospective study, agonal breathing at hospital arrival was significantly associated with better neurological outcomes and increased survival at 1 month. Thus, agonal breathing at hospital arrival may be a useful prognostic predictor for patients experiencing OHCA.

Intensive care following in-hospital cardiac arrest / periarrest calls-experience from one Scottish hospital

BACKGROUND

In-hospital cardiac arrest/periarrest is a recognised trigger for consideration of admission to the intensive care unit (ICU). We aimed to investigate the rates of ICU admission following in-hospital cardiac arrest/periarrest, evaluate the outcomes of such patients and assess whether anticipatory care planning had taken place prior to the adult resuscitation team being called.

METHODS

Analysis of all referrals to the ICU page-holder within our district general hospital is between 1st November 2018 and 31st May 2019. From this, the frequency of adult resuscitation team calls was determined. Case notes were then reviewed to determine details of the events, patient outcomes and the use of anticipatory care planning tools on wards.

RESULTS

Of the 506 referrals to the ICU page-holder, 141 (27.9%) were adult resuscitation team calls (114 periarrests and 27 cardiac arrests). Twelve patients were excluded due to health records being unavailable. Admission rates to ICU were low - 17.4% for cardiac arrests (4/23 patients), 5.7% (6/106) following periarrest. The primary reason for not admitting to ICU was patients being "too well" at the time of review (78/129 - 60.5%). Prior to adult resuscitation team call, treatment escalation plans had been completed in 27.9% (36/129) with Do Not Attempt Cardiopulmonary Resuscitation (DNACPR) forms present in 15.5% of cases (20/129). Four cardiac arrest calls were made in the presence of a valid DNACPR form, frequently due to a lack of awareness of the patient's resuscitation status.

CONCLUSIONS

This study highlights the significant workload for the ICU page-holder brought about by adult resuscitation team calls. There is a low admission rate from these calls, and, at the time of resuscitation team call, anticipatory planning is frequently either incomplete or poorly communicated. Addressing these issues requires a collaborative approach between ICU and non-ICU physicians and highlights the need for larger studies to develop scoring systems to aid objective admission decision-making.

The value of a machine learning algorithm to predict adverse short-term outcome during resuscitation of patients with in-hospital cardiac arrest: a retrospective study

Background and importance Guidelines recommend that hospital emergency teams locally validate criteria for termination of cardiopulmonary resuscitation in patients with in-hospital cardiac arrest (IHCA). Objective To determine the value of a machine learning algorithm to predict failure to achieve return of spontaneous circulation (ROSC) and unfavourable functional outcome from IHCA using only data readily available at emergency team arrival. Design Retrospective cohort study. Setting and participants Adults who experienced an IHCA were attended to by the emergency team. Outcome measures and analysis Demographic and clinical data typically available at the arrival of the emergency team were extracted from the institutional IHCA database. In addition, outcome data including the Cerebral Performance Category (CPC) score count at hospital discharge were collected. A model selection procedure for random forests with a hyperparameter search was employed to develop two classification algorithms to predict failure to achieve ROSC and unfavourable (CPC 3-5) functional outcomes. Main results Six hundred thirty patients were included, of which 390 failed to achieve ROSC (61.9%). The final classification model to predict failure to achieve ROSC had an area under the receiver operating characteristic curve of 0.9 [95% confidence interval (CI), 0.89-0.9], a balanced accuracy of 0.77 (95% CI, 0.75-0.79), an F1-score of 0.78 (95% CI, 0.76-0.79), a positive predictive value of 0.88 (0.86-0.91), a negative predictive value of 0.61 (0.6-0.63), a sensitivity of 0.69 (0.66-0.72), and a specificity of 0.84 (0.8-0.88). Five hundred fifty-nine subjects experienced an unfavourable outcome (88.7%). The final classification model to predict unfavourable functional outcomes from IHCA at hospital discharge had an area under the receiver operating characteristic curve of 0.93 (95% CI, 0.92-0.93), a balanced accuracy of 0.59 (95% CI, 0.57-0.61), an F1-score of 0.94 (95% CI, 0.94-0.95), a positive predictive value of 0.91 (0.9-0.91), a negative predictive value of 0.57 (0.48-0.66), a sensitivity of 0.98 (0.97-0.99), and a specificity of 0.2 (0.16-0.24). Conclusion Using data readily available at emergency team arrival, machine learning algorithms had a high predictive power to forecast failure to achieve ROSC and unfavourable functional outcomes from IHCA while cardiopulmonary resuscitation was still ongoing; however, the positive predictive value of both models was not high enough to allow for early termination of resuscitation efforts.