Clinical deterioration in cancer patients--the role of the rapid response team*

Deep Learning-Based Early Warning Score for Predicting Clinical Deterioration in General Ward Cancer Patients

BACKGROUND

Cancer patients who are admitted to hospitals are at high risk of short-term deterioration due to treatment-related or cancer-specific complications. A rapid response system (RRS) is initiated when patients who are deteriorating or at risk of deteriorating are identified. This study was conducted to develop a deep learning-based early warning score (EWS) for cancer patients (Can-EWS) using delta values in vital signs.

METHODS

A retrospective cohort study was conducted on all oncology patients who were admitted to the general ward between 2016 and 2020. The data were divided into a training set (January 2016-December 2019) and a held-out test set (January 2020-December 2020). The primary outcome was clinical deterioration, defined as the composite of in-hospital cardiac arrest (IHCA) and unexpected intensive care unit (ICU) transfer.

RESULTS

During the study period, 19,739 cancer patients were admitted to the general wards and eligible for this study. Clinical deterioration occurred in 894 cases. IHCA and unexpected ICU transfer prevalence was 1.77 per 1000 admissions and 43.45 per 1000 admissions, respectively. We developed two models: Can-EWS V1, which used input vectors of the original five input variables, and Can-EWS V2, which used input vectors of 10 variables (including an additional five delta variables). The cross-validation performance of the clinical deterioration for Can-EWS V2 (AUROC, 0.946; 95% confidence interval [CI], 0.943-0.948) was higher than that for MEWS of 5 (AUROC, 0.589; 95% CI, 0.587-0.560; p < 0.001) and Can-EWS V1 (AUROC, 0.927; 95% CI, 0.924-0.931). As a virtual prognostic study, additional validation was performed on held-out test data. The AUROC and 95% CI were 0.588 (95% CI, 0.588-0.589), 0.890 (95% CI, 0.888-0.891), and 0.898 (95% CI, 0.897-0.899), for MEWS of 5, Can-EWS V1, and the deployed model Can-EWS V2, respectively. Can-EWS V2 outperformed other approaches for specificities, positive predictive values, negative predictive values, and the number of false alarms per day at the same sensitivity level on the held-out test data.

CONCLUSIONS

We have developed and validated a deep learning-based EWS for cancer patients using the original values and differences between consecutive measurements of basic vital signs. The Can-EWS has acceptable discriminatory power and sensitivity, with extremely decreased false alarms compared with MEWS.

A practical approach to establishing a critical care outreach service: An expert panel research design

BACKGROUND

For over two decades, nurse-led critical care outreach services have improved the recognition, response, and management of deteriorating patients in general hospital wards, yet variation in terms, design, implementation, and evaluation of such services continue. For those establishing a critical care outreach service, these factors make the literature difficult to interpret and translate to the real-world setting.

AIM

The aim of this study was to provide a practical approach to establishing a critical care outreach service in the hospital setting.

METHOD

An international expert panel of clinicians, managers, and academics with experience in implementing, developing, operationalising, educating, and evaluating critical care outreach services collaborated to synthesise evidence, experience, and clinical judgment to develop a practical approach for those establishing a critical care outreach service. A rapid review of the literature identified publications relevant to the study. A modified Delphi technique was used to achieve expert panel consensus particularly in areas where insufficient published literature or ambiguities existed.

FINDINGS

There were 502 publications sourced from the rapid review, of which 104 were relevant and reviewed. Using the modified Delphi technique, the expert panel identified five key components needed to establish a critical care outreach service: (i) approaches to service delivery, (ii) education and training, (iii) organisational engagement, (iv) clinical governance, and (v) monitoring and evaluation.

CONCLUSION

An expert panel research design successfully synthesised evidence, experience, and clinical judgement to provide a practical approach for those establishing a critical care outreach service. This method of research will likely be valuable in other areas of practice where terms are used interchangeably, and the literature is diverse and lacking a single approach to practice.

Utilization of a rapid response team and associated outcomes in patients with malignancy

Background:

Recent advances in diagnosis and treatment have improved long-term outcomes in cancer patients. As a result, the requirement for a rapid response team (RRT) for cancer patients is also increasing. This study aimed to analyze utilization of an RRT and the associations between related factors and mortality in a population of cancer patients.

Methods:

This retrospective cohort study included hospitalized patients at a single academic medical center in Seoul, Korea, who required RRT activation during a 6-year period from June 2013 to December 2018.

Results:

Overall, 164 of the 457 patients who met the above criteria were cancer patients, and they had a significantly higher Charlson comorbidity score than the non-cancer patients (5.0 vs. 7.0, P<0.001). A significantly larger proportion of cancer patients required intensive care unit transfer (51.8% vs. 41.0%, P=0.032). Cancer patients also had significantly higher in-hospital mortality compared with other patients (39.6% vs. 10.9%, P<0.001). Furthermore, presence of cancer was independently associated with in-hospital mortality (adjusted odds ratio [OR], 2.09; 95% confidence interval [CI], 1.11 to 3.93). Among cancer patients, higher Acute Physiology and Chronic Health Evaluation (APACHE) II at the time of RRT activation was significantly associated with in-hospital mortality regardless of malignancy (adjusted OR, 1.08; 95% CI, 1.01 to 1.15).

Conclusions:

Cancer patients requiring RRT activation have significantly higher rates of inhospital mortality than patients not using RRT. Higher severity score at the time of RRT activation in patients with malignancy was significantly associated with in-hospital mortality.

The Intensive Care Medicine research agenda on critically ill oncology and hematology patients

Over the coming years, accelerating progress against cancer will be associated with an increased number of patients who require life-sustaining therapies for infectious or toxic chemotherapy-related events. Major changes include increased number of cancer patients admitted to the ICU with full-code status or for time-limited trials, increased survival and quality of life in ICU survivors, changing prognostic factors, early ICU admission for optimal monitoring, and use of noninvasive diagnostic and therapeutic strategies. In this review, experts in the management of critically ill cancer patients highlight recent changes in the use and the results of intensive care in patients with malignancies. They seek to put forward a standard of care for the management of these patients and highlight important updates that are required to care for them. The research agenda they suggest includes important studies to be conducted in the next few years to increase our understanding of organ dysfunction in this population and to improve our ability to appropriately use life-saving therapies or select new therapeutic approaches that are likely to improve outcomes. This review aims to provide more guidance for the daily management of patients with cancer, in whom outcomes are constantly improving, as is our global ability to fight against what is becoming the leading cause of mortality in industrialized and non-industrialized countries.