Torbic, Heather

Background Over half of ICU survivors experience post intensive care syndrome (PICS). Few PICS clinics were operational in the United States at the onset of the COVID-19 pandemic. Here we describe early experience of such a clinic at a large academic medical center. Our objective was to describe the (a) model of care (b) clinic interventions (c) short-term cognitive, psychiatric, and functional outcomes (d) one-year outcomes of COVID-19 ICU survivors at the Cleveland Clinic post-ICU recovery clinic (PIRC). Methods We conducted a retrospective cohort study through a chart review of all patients seen in PIRC from April 2020 – December 2020. A total of 59 patients completed the visit with 49% being virtual, and 51% being in-person. The cohort was predominantly white (50.8%), and male (57.6%) with a median age of 58.2 years. We screened for cognitive and psychiatric impairments using several different validated questionnaires. Functional impairments were self-reported and detected after assessment by a physical therapist. Quantifiable clinic interventions were referrals, pharmacy medication reconciliation, and vaccine administration. One-year outcomes were health care utilization, and death. Results In our cohort, 36%, 39.6%, and 17% of patients screened positive for cognitive impairment, anxiety and/or depression, and PTSD respectively. 42.3% of patients had lower extremity mobility impairment. 57.6% of patients required at least one referral after clinic. Of the 67.9% of patients who were working and 93.9% driving prior to hospitalization, only 24.6% and 73.2% had returned to those activities, respectively. The most common pharmacy intervention was discontinuation of a medication no longer in use. By one year, 34% of patients had hospital re-admissions with 5% requiring the ICU. There were no deaths. Conclusions Adult survivors of COVID-19 critical illness have impairments in all PICS domains as well as high health care utilization in the first year after hospital discharge. MeSH Terms Post intensive care syndrome, COVID-19 Key Points/Summary Adult survivors of COVID-19 critical illness have impairments in all PICS domains as well as high health care utilization in the first year after hospital discharge. Post-ICU recovery clinics can help survivors with recovery.

Abstract Introduction Delirium occurrence is common and preventive strategies are resource intensive. Screening tools can prioritize patients at risk. Using machine learning, we can capture time and treatment effects that pose a challenge to delirium prediction. We aim to develop a delirium prediction model that can be used as a screening tool. Methods From the eICU Collaborative Research Database (eICU-CRD) and the Medical Information Mart for Intensive Care version III (MIMIC-III) database, patients with one or more Confusion Assessment Method-Intensive Care Unit (CAM-ICU) values and intensive care unit (ICU) length of stay greater than 24 h were included in our study. We validated our model using 21 quantitative clinical parameters and assessed performance across a range of observation and prediction windows, using different thresholds and applied interpretation techniques. We evaluate our models based on stratified repeated cross-validation using 3 algorithms, namely Logistic Regression, Random Forest, and Bidirectional Long Short-Term Memory (BiLSTM). BiLSTM represents an evolution from recurrent neural network-based Long Short-Term Memory, and with a backward input, preserves information from both past and future. Model performance is measured using Area Under Receiver Operating Characteristic, Area Under Precision Recall Curve, Recall, Precision (Positive Predictive Value), and Negative Predictive Value metrics. Results We evaluated our results on 16 546 patients (47% female) and 6294 patients (44% female) from eICU-CRD and MIMIC-III databases, respectively. Performance was best in BiLSTM models where, precision and recall changed from 37.52% (95% confidence interval [CI], 36.00%–39.05%) to 17.45 (95% CI, 15.83%–19.08%) and 86.1% (95% CI, 82.49%–89.71%) to 75.58% (95% CI, 68.33%–82.83%), respectively as prediction window increased from 12 to 96 h. After optimizing for higher recall, precision and recall changed from 26.96% (95% CI, 24.99%–28.94%) to 11.34% (95% CI, 10.71%–11.98%) and 93.73% (95% CI, 93.1%–94.37%) to 92.57% (95% CI, 88.19%–96.95%), respectively. Comparable results were obtained in the MIMIC-III cohort. Conclusions Our model performed comparably to contemporary models using fewer variables. Using techniques like sliding windows, modification of threshold to augment recall and feature ranking for interpretability, we addressed shortcomings of current models.

Background: Variability in sedation may increase the incidence of delirium and mortality, as well as increased intensive care unit (ICU) and hospital lengths of stay (LOS), despite mean Richmond Agitation Sedation Scale (RASS) scores at goal. Coefficient of variation (CV) can be used to represent variability with a higher ratio indicating increased variability. Study Question: Do patients with an increased variability in sedation, as evaluated by CV in RASS, have an increased incidence of delirium? Methods: We conducted a retrospective chart review of adult medical ICU patients requiring mechanical ventilation (MV) for ≥24 hours between January and April 2013. Patients were excluded if intubated at an outside hospital, neuromuscularly blocked, suffering from anoxic brain injury, or had a goal RASS of −4 or −5. Outcomes assessed included the presence of delirium, as evaluated by the Confusion Assessment Method, RASS, CV in RASS, duration of MV, ICU, and hospital LOS, and survival. Results: Of 45 included patients, 32 experienced delirium during their ICU admission and 13 did not. The groups were similar at baseline. There was no difference in mean RASS when comparing the delirium and nondelirium groups (−1.6 ± 1.3 vs. −1.8 ± 0.8, respectively; P = 0.61). Patients with delirium had a greater CV in RASS (0.3 ± 0.135 vs. 0.2 ± 0.105; P = 0.02), a longer MV duration [4 (2–8) vs. 3 (2–3) days; P = 0.045], and a trend toward increased ICU LOS [8 (5–12.25) vs. 4 (3–8) days; P = 0.096], but no difference in hospital LOS [13 (10–25) vs. 18 (9–39) days; P = 0.83] and survival (71.9% vs. 69.2%; P = 1.0). Conclusion: Medical ICU patients with delirium had a higher CV in RASS compared with patients without delirium, suggesting that greater variability in sedation may increase the incidence of delirium. Patients with delirium also had a greater duration of MV and a trend toward longer ICU LOS.