Natural language processing diagnosed behavioural disturbance phenotypes in the intensive care unit: characteristics, prevalence, trajectory, treatment, and outcomes.

BACKGROUND: Natural language processing (NLP) may help evaluate the characteristics, prevalence, trajectory, treatment, and outcomes of behavioural disturbance phenotypes in critically ill patients. METHODS: We obtained electronic clinical notes, demographic information, outcomes, and treatment data from three medical-surgical ICUs. Using NLP, we screened for behavioural disturbance phenotypes based on words suggestive of an agitated state, a non-agitated state, or a combination of both. RESULTS: We studied 2931 patients. Of these, 225 (7.7%) were NLP-Dx-BD positive for the agitated phenotype, 544 (18.6%) for the non-agitated phenotype and 667 (22.7%) for the combined phenotype. Patients with these phenotypes carried multiple clinical baseline differences. On time-dependent multivariable analysis to compensate for immortal time bias and after adjustment for key outcome predictors, agitated phenotype patients were more likely to receive antipsychotic medications (odds ratio [OR] 1.84, 1.35-2.51, p < 0.001) compared to non-agitated phenotype patients but not compared to combined phenotype patients (OR 1.27, 0.86-1.89, p = 0.229). Moreover, agitated phenotype patients were more likely to die than other phenotypes patients (OR 1.57, 1.10-2.25, p = 0.012 vs non-agitated phenotype; OR 4.61, 2.14-9.90, p < 0.001 vs. combined phenotype). This association was strongest in patients receiving mechanical ventilation when compared with the combined phenotype (OR 7.03, 2.07-23.79, p = 0.002). A similar increased risk was also seen for patients with the non-agitated phenotype compared with the combined phenotype (OR 6.10, 1.80-20.64, p = 0.004). CONCLUSIONS: NLP-Dx-BD screening enabled identification of three behavioural disturbance phenotypes with different characteristics, prevalence, trajectory, treatment, and outcome. Such phenotype identification appears relevant to prognostication and trial design.

Natural language processing diagnosed behavioral disturbance vs confusion assessment method for the intensive care unit: prevalence, patient characteristics, overlap, and association with treatment and outcome.

PURPOSE: To compare the prevalence, characteristics, drug treatment for delirium, and outcomes of patients with Natural Language Processing (NLP) diagnosed behavioral disturbance (NLP-Dx-BD) vs Confusion Assessment Method for intensive care unit (CAM-ICU) positivity. METHODS: In three combined medical-surgical ICUs, we obtained data on demographics, treatment with antipsychotic medications, and outcomes. We applied NLP to caregiver progress notes to diagnose behavioral disturbance and analyzed simultaneous CAM-ICU. RESULTS: We assessed 2313 patients with a median lowest Richmond Agitation-Sedation Scale (RASS) score of - 2 (- 4.0 to - 1.0) and median highest RASS score of 1 (0 to 1). Overall, 1246 (53.9%) patients were NLP-Dx-BD positive (NLP-Dx-BDpos) and 578 (25%) were CAM-ICU positive (CAM-ICUpos). Among NLP-Dx-BDpos patients, 539 (43.3%) were also CAM-ICUpos. In contrast, among CAM-ICUpos patients, 539 (93.3%) were also NLP-Dx-BDpos. The use of antipsychotic medications was highest in patients in the CAM-ICUpos and NLP-Dx-BDpos group (24.3%) followed by the CAM-ICUneg and NLP-Dx-BDpos group (10.5%). In NLP-Dx-BDneg patients, antipsychotic medication use was lower at 5.1% for CAM-ICUpos and NLP-Dx-BDneg patients and 2.3% for CAM-ICUneg and NLP-Dx-BDneg patients (overall P < 0.001). Regardless of CAM-ICU status, after adjustment and on time-dependent Cox modelling, NLP-Dx-BD was associated with greater antipsychotic medication use. Finally, regardless of CAM-ICU status, NLP-Dx-BDpos patients had longer duration of ICU and hospital stay and greater hospital mortality (all P < 0.001). CONCLUSION: More patients were NLP-Dx-BD positive than CAM-ICU positive. NLP-Dx-BD and CAM-ICU assessment describe partly overlapping populations. However, NLP-Dx-BD identifies more patients likely to receive antipsychotic medications. In the absence of NLP-Dx-BD, treatment with antipsychotic medications is rare.

Natural language processing to assess the epidemiology of delirium-suggestive behavioural disturbances in critically ill patients.

Background: There is no gold standard approach for delirium diagnosis, making the assessment of its epidemiology difficult. Delirium can only be inferred though observation of behavioural disturbance and described with relevant nouns or adjectives. Objective: We aimed to use natural language processing (NLP) and its identification of words descriptive of behavioural disturbance to study the epidemiology of delirium in critically ill patients. Study design: Retrospective study using data collected from the electronic health records of a university-affiliated intensive care unit (ICU) in Melbourne, Australia. Participants: 12 375 patients Intervention: Analysis of electronic progress notes. Identification using NLP of at least one of a list of words describing behavioural disturbance within such notes. Results: We analysed 199 648 progress notes in 12 375 patients. Of these, 5108 patients (41.3%) had NLP-diagnosed behavioural disturbance (NLP-Dx-BD). Compared with those who did not have NLP-Dx-DB, these patients were older, more severely ill, and likely to have medical or unplanned admissions, neurological diagnosis, chronic kidney or liver disease and to receive mechanical ventilation and renal replacement therapy (P < 0.001). The unadjusted hospital mortality for NLP-Dx-BD patients was 14.1% versus 9.6% for patients without NLP-Dx-BD. After adjustment for baseline characteristics and illness severity, NLP-Dx-BD was not associated with increased risk of death (odds ratio [OR], 0.94; 95% CI, 0.80-1.10); a finding robust to multiple sensitivity, subgroups and time of observation subcohort analyses. In mechanically ventilated patients, NLP-Dx-BD was associated with decreased hospital mortality (OR, 0.80; 95% CI, 0.65-0.99) after adjustment for baseline severity of illness and year of admission. Conclusions: NLP enabled rapid assessment of large amounts of data identifying a population of ICU patients with typical high risk characteristics for delirium. Moreover, this technique enabled identification of previously poorly understood associations. Further investigations of this technique appear justified.

Prognostic performance of qSOFA in oncology patients admitted to the emergency department with suspected infection.

AIM: We aimed to test the performance of the quick Sequential Organ Failure Assessment score (qSOFA) in predicting the outcomes of oncology patients admitted to the emergency department (ED) with suspected infection. METHODS: Retrospective cohort analysis of all oncology patients presenting to the ED of a tertiary hospital with suspected infection from 1 December 2014 to 1 June 2017. Patients were identified by cross-linkage of ED and Oncology electronic health records. The primary outcome was in-hospital mortality and/or ICU stay ≥ 3 days. RESULTS: A total of 1655 patients were included in this study--1267 (76.6%) with solid tumor and 388 (23.4%) with hematological malignancies. At presentation, 495 patients had chemotherapy, and 140 had radiotherapy within the preceding 6 months. Four hundred patients received chemotherapy and/or radiotherapy in the previous 4 weeks. Overall, 371 (22.4%) patients had qSOFA ≥ 2. Such patients had a higher likelihood of respiratory infections compared to patients with a qSOFA < 2 (43.9% vs 29%) and were more likely to be admitted to ICU or require mechanical ventilation. In-hospital mortality or in-hospital mortality and/or ICU stay ≥ 3 days were 17.3% and 21%, for qSOFA ≥ 2 patients versus 4.7% and 6.9% for qSOFA < 2 patients (P < .001). qSOFA ≥ 2 had a negative predictive value of 95% for in-hospital mortality and 93% for in-hospital mortality or ICU stay ≥ 3 days. CONCLUSION: Among oncology patients presenting to the ED with suspected infection, a qSOFA ≥ 2 is associated with a threefold risk of hospital mortality/prolonged ICU stay. Its absence helps identify low-risk patients.

Using language descriptors to recognise delirium: a survey of clinicians and medical coders to identify delirium-suggestive words.

OBJECTIVE: To develop a library of delirium-suggestive words. DESIGN: Cross-sectional survey. SETTING: Single tertiary referral hospital. PARTICIPANTS: Medical, nursing and allied health staff and medical coders. MAIN OUTCOME MEASURES: Frequency of graded response on a 5-point Likert scale to individual delirium-suggestive words. RESULTS: Two-hundred and three complete responses were received from 227 survey respondents; the majority were medical and nursing staff (42.4% and 43.8% respectively), followed by allied health practitioners and medical coders (10.3% and 3.4%). Words that were "very likely" to suggest delirium were "confused/ confusion", "delirious", "disoriented/disorientation" and "fluctuating conscious state". Differences in word selection were noted based on occupational background, prior knowledge of delirium, and experience in caring for intensive care unit patients. Distractor words included in the survey were rated as "unlikely" or "very unlikely" by respondents as expected. Textual responses identified several other descriptors of delirium-suggestive words. CONCLUSION: A comprehensive repertoire of delirium-suggestive words was validated using a multidisciplinary survey and new words suggested by respondents were added. The use of natural language processing algorithms may allow for earlier detection of delirium using our delirium library and be deployed for real-time decision making and clinical care.

Differential clinical characteristics, management and outcome of delirium among ward compared with intensive care unit patients.

BACKGROUND: Delirium is common in hospitalised patients but its epidemiology remains poorly characterised. AIMS: To test the hypothesis that patient demographics, clinical phenotype, management and outcomes of patient with delirium in hospital ward patients differ from intensive care unit (ICU) patients. METHODS: Retrospective cohort of patients admitted to an Australian university-affiliated hospital between March 2013 and April 2017 and coded for delirium at discharge using the International Classification of Diseases System, 10th revision, criteria. RESULTS: Among 61 032 hospitalised patients, 2864 (4.7%) were coded for delirium. From these, we studied a random sample of 100 ward patients and 100 ICU patients. Ward patients were older (median age: 84 vs 65 years; P < 0.0001), more likely to have dementia (38% vs 2% for ICU patients; P < 0.0001) and less likely to have had surgery (24% vs 62%; P < 0.0001). Of ward patients, 74% had hypoactive delirium, while 64% of ICU patients had agitated delirium (P < 0.0001). Persistent delirium at hospital discharge was more common among ward patients (66% vs 17%, P < 0.0001). On multivariable analysis, age and dementia predicted persistent delirium, while surgery predicted recovery. CONCLUSIONS: Delirium in ward patients is profoundly different from delirium in ICU patients. It has a dominant hypoactive clinical phenotype, is preceded by dementia and is less likely to recover at hospital discharge. Therefore, delirium prevention, detection and goals of care should be adapted to the environment in which it occurs.