Development of a practically usable prediction model for quality of life of ICU survivors: A sub-analysis of the MONITOR-IC prospective cohort study

Development and validation of a machine learning-based model for post-sepsis frailty

Background

The development of post-sepsis frailty is a common and significant problem, but it is a challenge to predict.

Methods

Data for deep learning were extracted from a national multicentre prospective observational cohort of patients with sepsis in Korea between September 2019 and December 2021. The primary outcome was frailty at survival discharge, defined as a clinical frailty score on the Clinical Frailty Scale ≥5. We developed a deep learning model for predicting frailty after sepsis by 10 variables routinely collected at the recognition of sepsis. With cross-validation, we trained and tuned six machine learning models, including four conventional and two neural network models. Moreover, we computed the importance of each predictor variable in the model. We measured the performance of these models using a temporal validation data set.

Results

A total of 8518 patients were included in the analysis; 5463 (64.1%) were frail, and 3055 (35.9%) were non-frail at discharge. The Extreme Gradient Boosting (XGB) achieved the highest area under the receiver operating characteristic curve (AUC) (0.8175) and accuracy (0.7414). To confirm the generalisation performance of artificial intelligence in predicting frailty at discharge, we conducted external validation with the COVID-19 data set. The XGB still showed a good performance with an AUC of 0.7668. The machine learning model could predict frailty despite the disparity in data distribution.

Conclusion

The machine learning-based model developed for predicting frailty after sepsis achieved high performance with limited baseline clinical parameters.

Changes in quality of life 1 year after intensive care: a multicenter prospective cohort of ICU survivors

BACKGROUND

With survival rates of critical illness increasing, quality of life measures are becoming an important outcome of ICU treatment. Therefore, to study the impact of critical illness on quality of life, we explored quality of life before and 1 year after ICU admission in different subgroups of ICU survivors.

METHODS

Data from an ongoing prospective multicenter cohort study, the MONITOR-IC, were used. Patients admitted to the ICU in one of eleven participating hospitals between July 2016 and June 2021 were included. Outcome was defined as change in quality of life, measured using the EuroQol five-dimensional (EQ-5D-5L) questionnaire, and calculated by subtracting the EQ-5D-5L score 1 day before hospital admission from the EQ-5D-5L score 1 year post-ICU. Based on the minimal clinically important difference, a change in quality of life was defined as a change in EQ-5D-5L score of ≥ 0.08. Subgroups of patients were based on admission diagnosis.

RESULTS

A total of 3913 (50.6%) included patients completed both baseline and follow-up questionnaires. 1 year post-ICU, patients admitted after a cerebrovascular accident, intracerebral hemorrhage, or (neuro)trauma, on average experienced a significant decrease in quality of life. Conversely, 11 other subgroups of ICU survivors reported improvements in quality of life. The largest average increase in quality of life was seen in patients admitted due to respiratory disease (mean 0.17, SD 0.38), whereas the largest average decrease was observed in trauma patients (mean -0.13, SD 0.28). However, in each of the studied 22 subgroups there were survivors who reported a significant increase in QoL and survivors who reported a significant decrease in QoL.

CONCLUSIONS

 This large prospective multicenter cohort study demonstrated the diversity in long-term quality of life between, and even within, subgroups of ICU survivors. These findings emphasize the need for personalized information and post-ICU care.

TRIAL REGISTRATION

The MONITOR-IC study was registered at ClinicalTrials.gov: NCT03246334 on August 2nd 2017.

Machine learning to predict poor school performance in paediatric survivors of intensive care: a population-based cohort study

PURPOSE

Whilst survival in paediatric critical care has improved, clinicians lack tools capable of predicting long-term outcomes. We developed a machine learning model to predict poor school outcomes in children surviving intensive care unit (ICU).

METHODS

Population-based study of children < 16 years requiring ICU admission in Queensland, Australia, between 1997 and 2019. Failure to meet the National Minimum Standard (NMS) in the National Assessment Program-Literacy and Numeracy (NAPLAN) assessment during primary and secondary school was the primary outcome. Routine ICU information was used to train machine learning classifiers. Models were trained, validated and tested using stratified nested cross-validation.

RESULTS

13,957 childhood ICU survivors with 37,200 corresponding NAPLAN tests after a median follow-up duration of 6 years were included. 14.7%, 17%, 15.6% and 16.6% failed to meet NMS in school grades 3, 5, 7 and 9. The model demonstrated an Area Under the Receiver Operating Characteristic curve (AUROC) of 0.8 (standard deviation SD, 0.01), with 51% specificity to reach 85% sensitivity [relative Area Under the Precision Recall Curve (rel-AUPRC) 3.42, SD 0.06]. Socio-economic status, illness severity, and neurological, congenital, and genetic disorders contributed most to the predictions. In children with no comorbidities admitted between 2009 and 2019, the model achieved a AUROC of 0.77 (SD 0.03) and a rel-AUPRC of 3.31 (SD 0.42).

CONCLUSIONS

A machine learning model using data available at time of ICU discharge predicted failure to meet minimum educational requirements at school age. Implementation of this prediction tool could assist in prioritizing patients for follow-up and targeting of rehabilitative measures.

Development and External Validation of a Prediction Model for Quality of Life of ICU Survivors: A Subanalysis of the MONITOR-IC Prospective Cohort Study

OBJECTIVES

To develop and externally validate a prediction model for ICU survivors' change in quality of life 1 year after ICU admission that can support ICU physicians in preparing patients for life after ICU and managing their expectations.

DESIGN

Data from a prospective multicenter cohort study (MONITOR-IC) were used.

SETTING

Seven hospitals in the Netherlands.

PATIENTS

ICU survivors greater than or equal to 16 years old.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Outcome was defined as change in quality of life, measured using the EuroQol 5D questionnaire. The developed model was based on data from an academic hospital, using multivariable linear regression analysis. To assist usability, variables were selected using the least absolute shrinkage and selection operator method. External validation was executed using data of six nonacademic hospitals. Of 1,804 patients included in analysis, 1,057 patients (58.6%) were admitted to the academic hospital, and 747 patients (41.4%) were admitted to a nonacademic hospital. Forty-nine variables were entered into a linear regression model, resulting in an explained variance ( R2 ) of 56.6%. Only three variables, baseline quality of life, admission type, and Glasgow Coma Scale, were selected for the final model ( R2 = 52.5%). External validation showed good predictive power ( R2 = 53.2%).

CONCLUSIONS

This study developed and externally validated a prediction model for change in quality of life 1 year after ICU admission. Due to the small number of predictors, the model is appealing for use in clinical practice, where it can be implemented to prepare patients for life after ICU. The next step is to evaluate the impact of this prediction model on outcomes and experiences of patients.

A novel nomogram to stratify quality of life among advanced cancer patients with spinal metastatic disease after examining demographics, dietary habits, therapeutic interventions, and mental health status

Background

It would be very helpful to stratify patients and direct patient selection if risk factors for quality of life were identified in a particular population. Nonetheless, it is still challenging to forecast the health-related quality of life among individuals with spinal metastases. The goal of this study was to stratify patient’s populations for whom the assessment of quality of life should be encouraged by developing and validating a nomogram to predict the quality of life among advanced cancer patients with spine metastases.

Methods

This study prospectively analyzed 208 advanced cancer patients with spine metastases, and collected their general characteristics, food preferences, addictions, comorbidities, therapeutic strategies, and mental health status. The functional assessment of cancer therapy-general (FACT-G) and hospital anxiety and depression scale (HADS) were used to assess quality of life and mental health, respectively. The complete cohort of patients was randomly divided into two groups: a training set and a validation set. Patients from the training set were conducted to train and develop a nomogram, while patients in the validation set were performed to internally validate the nomogram. The nomogram contained significant variables discovered using the least absolute shrinkage and selection operator (LASSO) approach in conjunction with 10-fold cross-validation. The nomogram’s predictive ability was assessed utilizing discrimination, calibration, and clinical usefulness. Internal validation was also completed using the bootstrap method after applying 500 iterations of procedures. A web calculator was also developed to promote clinical practice.

Results

Advance cancer patients with spinal metastases had an extremely low quality of life, as indicated by the average FACT-G score of just 60.32 ± 20.41. According to the LASSO and 10-fold cross-validation, Eastern Cooperative Oncology Group (ECOG) score, having an uncompleted life goal, preference for eating vegetables, chemotherapy, anxiety status, and depression status were selected as nomogram predictors. In the training set, the area under the receiver operating characteristic curve (AUROC) was 0.90 (95% CI: 0.84–0.96), while in the validation set, it was 0.85 (95% CI: 0.78–0.93). They were 0.50 (95% CI: 0.41–0.58) and 0.44 (95% CI: 0.33–0.56), respectively, for the discrimination slopes. The nomogram had favorable capacity to calibrate and was clinically useful, according to the calibration curve and decision curve analysis. When compared to patients in the low-risk group, patients in the high-risk group were above four times more likely to experience a poor quality of life (82.18% vs. 21.50%, P < 0.001). In comparison to patients in the low-risk group, patients in the high-risk group also exhibited significant higher levels of anxiety and depression. The webpage for the web calculator was https://starshiny.shinyapps.io/DynNomapp-lys/.

Conclusions

This study suggests a nomogram that can be applied as a practical clinical tool to forecast and categorize the quality of life among patients with spine metastases. Additionally, patients with poor quality of life experience more severe anxiety and depression. Effective interventions should be carried out as soon as possible, especially for patients in the high-risk group, to improve their quality of life and mental health condition.

Optimizing an existing prediction model for quality of life one-year post-intensive care unit: An exploratory analysis

BACKGROUND

This study aimed to improve the PREPARE model, an existing linear regression prediction model for long-term quality of life (QoL) of intensive care unit (ICU) survivors by incorporating additional ICU data from patients' electronic health record (EHR) and bedside monitors.

METHODS

The 1308 adult ICU patients, aged ≥16, admitted between July 2016 and January 2019 were included. Several regression-based machine learning models were fitted on a combination of patient-reported data and expert-selected EHR variables and bedside monitor data to predict change in QoL 1 year after ICU admission. Predictive performance was compared to a five-feature linear regression prediction model using only 24-hour data (R²  = 0.54, mean square error (MSE) = 0.031, mean absolute error (MAE) = 0.128).

RESULTS

The 67.9% of the included ICU survivors was male and the median age was 65.0 [IQR: 57.0-71.0]. Median length of stay (LOS) was 1 day [IQR 1.0-2.0]. The incorporation of the additional data pertaining to the entire ICU stay did not improve the predictive performance of the original linear regression model. The best performing machine learning model used seven features (R²  = 0.52, MSE = 0.032, MAE = 0.125). Pre-ICU QoL, the presence of a cerebro vascular accident (CVA) upon admission and the highest temperature measured during the ICU stay were the most important contributors to predictive performance. Pre-ICU QoL's contribution to predictive performance far exceeded that of the other predictors.

CONCLUSION

Pre-ICU QoL was by far the most important predictor for change in QoL 1 year after ICU admission. The incorporation of the numerous additional features pertaining to the entire ICU stay did not improve predictive performance although the patients' LOS was relatively short.

MiCare study, an evaluation of structured, multidisciplinary and personalised post-ICU care on physical and psychological functioning, and quality of life of former ICU patients: a study protocol of a stepped-wedge cluster randomised controlled trial

INTRODUCTION

Over 70% of the intensive care unit (ICU) survivors suffer from long-lasting physical, mental and cognitive problems after hospital discharge. Post-ICU care is recommended by international guidelines, but evidence for cost-effectiveness lacks. The aim of this study is to evaluate the clinical effectiveness and cost-effectiveness of structured, multidisciplinary and personalised post-ICU care versus usual care on physical and psychological functioning and health-related quality of life (HRQoL) of ICU survivors, 1- and 2-year post-ICU discharge.

METHODS AND ANALYSIS

The MONITOR-IC post-ICU care study (MiCare study) is a multicentre stepped-wedge randomised controlled trial conducted in five hospitals. Adult patients at high risk for critical illness-associated morbidity post-ICU will be selected and receive post-ICU care, including an invitation to the post-ICU clinic 3 months after ICU discharge. A personalised long-term recovery plan tailored to patients' reported outcome measures will be made. 770 (intervention) and 1480 (control) patients will be included. Outcomes are 1- and 2-year HRQoL (EuroQol Instrument (EQ-5D-5L)), physical (fatigue and new physical problems), mental (anxiety, depression and post-traumatic stress disorder), and cognitive symptoms and cost-effectiveness. Medical data will be retrieved from patient records and cost data from health insurance companies.

ETHICS AND DISSEMINATION

Due to the lack of evidence, Dutch healthcare insurers do not reimburse post-ICU care. Therefore, evaluation of cost-effectiveness and integration in guidelines supports the evidence. Participation of several societies for physicians, nurses, paramedics, and patients and relatives in the project team increases the support for implementation of the intervention in clinical practice. Patients and relatives will be informed by the patient associations, hospitals and professional associations. Informing healthcare insurers about this project's results is important for the consideration for inclusion of post-ICU care in Dutch standard health insurance. The study is approved by the Radboud University Medical Centre research ethics committee (2021-13125).

TRIAL REGISTRATION NUMBER

NCT05066984.

Development and validation of early prediction models for new-onset functional impairment at hospital discharge of ICU admission

PURPOSE

We aimed to develop and validate models for predicting new-onset functional impairment after intensive care unit (ICU) admission with predictors routinely collected within 2 days of admission.

METHODS

In this multi-center retrospective cohort study of acute care hospitals in Japan, we identified adult patients who were admitted to the ICU with independent activities of daily living before hospitalization and survived for at least 2 days from April 2014 to October 2020. The primary outcome was functional impairment defined as Barthel Index ≤ 60 at hospital discharge. In the internal validation dataset (April 2014 to March 2019), using routinely collected 94 candidate predictors within 2 days of ICU admission, we trained and tuned the six conventional and machine-learning models with repeated random sub-sampling cross-validation. We computed the variable importance of each predictor to the models. In the temporal validation dataset (April 2019 to October 2020), we measured the performance of these models.

RESULTS

We identified 19,846 eligible patients. Functional impairment at discharge was developed in 33% of patients (n = 6488/19,846). In the temporal validation dataset, all six models showed good discrimination ability with areas under the curve above 0.86, and the differences among the six models were negligible. Variable importance revealed newly detected early predictors, including worsened neurologic conditions and catabolism biomarkers such as decreased serum albumin and increased blood urea nitrogen.

CONCLUSION

We successfully developed early prediction models of new-onset functional impairment after ICU admission that achieved high performance using only data routinely collected within 2 days of ICU admission.

Using long-term predicted Quality of Life in ICU clinical practice to prepare patients for life post-ICU: A feasibility study

PURPOSE

To examine the feasibility of using the PREdicting PAtients' long-term outcome for Recovery (PREPARE) prediction model for Quality of Life (QoL) 1 year after ICU admission in ICU practice to prepare expected ICU survivors and their relatives for life post-ICU.

MATERIALS AND METHODS

Between June 2020 and February 2021, the predicted change in QoL after 1 year was discussed in 25 family conferences in the ICU. 13 physicians, 10 nurses and 19 patients and/or family members were interviewed to evaluate intervention feasibility in ICU practice. Interviews were analysed qualitatively using thematic coding.

RESULTS

Patients' median age was 68.0 years, five patients (20.0%) were female and seven patients (28.0%) died during ICU stay. Generally, study participants thought the intervention, which clarified the concept of QoL through visualization and served as a reminder to discuss QoL and expectations for life post-ICU, had merit. However, some participants, especially physicians, thought the prediction model needed more data on more severely ill ICU patients to curb uncertainty.

CONCLUSIONS

Using predicted QoL scores in ICU practice to prepare patients and family members for life after ICU discharge is feasible. After optimising the model and implementation strategy, its effectiveness can be evaluated in a larger trial.