Mechanical Ventilation, Patient-Ventilator Asynchrony
Conditions
Keywords
asynchrony, Patient-ventilator asynchrony, dyssynchrony, Patient-ventilator interaction, mechanical ventilation, AI, algorithm, classification algorithm, ICU, Artificial intelligence
Brief summary
The goal of this observational study is to unravel the occurence, impact and relations of Patient-Ventilator Aynchrony (PVA) in mechanically ventilated patients. The main questions it aims to answer are: * How often does PVA occur? * What are relations between clinical characteristics and PVA occurence? * What are relations between PVA occurence and patient outcomes? All questions will be assessed using data collected during the whole course of mechanical ventilation. Mechanically ventilated patients' medical data will be re-used. PVAs will be automatically classified on ventilator waveform data, using validated Deep Breath software.
Detailed description
Many ventilated patients show excessive breathing efforts and abnormal, irregular breathing. This patient-ventilator asynchrony (PVA) is associated with serious discomfort, lung injury, sleep disruption and higher mortality. PVA exists in many forms and is reported in 10-90% of patients, but identifying and resolving it is challenging, even for expert clinicians. Hence, PVA prevalence and impact is likely highly underestimated, and the direct causal link with worse outcomes is inconclusive. PVAs should be better dettected, understood and resovled to optimize the individual patient's treatment. In a previous study, the investigators validated an AI-based algorithm capable of reliable PVA detection (Deep Breath software). In this study, the investigators will apply this algorithm to the collected ventilator waveform data (offline processing), in order to reliably assess PVA occurrence, and its relation with clinical outcomes and patient characteristics in current clinical care. Data of minimally 110 patients collected over the whole course of mechanical ventilation will be assessed. Patients will be included in three ICUs to promote generalizability. The primary outcome will be the asynchrony index (in total and per PVA type) over time. Secondary outcomes will include, but are not limited to: clinical characteristics (e.g. respiratory and hemodynamic parameters, sedation), (ICU) mortality, ventilator free days at day 28 and 90, duration of ventilation, weaning success and reintubation rate.
Interventions
OTHERPVA classification software
Patients will receive standard care, without an intervention. Data will be captured as part of standard care and analyzed for PVAs retrospectively, using dedicated offline software.
Sponsors
Erasmus Medical Center
Deep Breath
Catharina Ziekenhuis Eindhoven
Leiden University Medical Center
Study design
Observational model
COHORT
Time perspective
OTHER
Eligibility
Sex/Gender
ALL
Age
18 Years to No maximum
Healthy volunteers
No
Inclusion criteria
* Age \> 18 years old. * Recordings available of ventilator waveforms synchronized with the patient's electronic health record during invasive mechanical ventilation. * Duration of mechanical ventilation of at least 24 hours.
Exclusion criteria
* (Previous) registered objection of patient and/or relatives to re-use clinical data for research purposes * No consent for re-use of data for research
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Asynchrony index over time (aggregated and per PVA type) | 28 days | Measure of how much asynchrony occurs and at what time. This measure will be calculated for all PVA types together, as well as per PVA type. The investigators calculate this over time, to see when asynchrony occurs, as well as in total, to get a global PVA prevalence measure (over the whole duration of ventilation). |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Use of sedatives (cumulative dose and type) | 28 days | — |
| Mechanical ventilation settings | 28 days | — |
| Respiratory parameters | 28 days | — |
| Hemodynamic parameters | 28 days | — |
| Relevant medication | 28 days | e.g. vasoactive agents, delirium related medication, analgesics |
| Use of assist devices | 28 days | e.g. dialysis, pacemaker, ventricular assist device, ECMO |
| Gas exchange parameters | 28 days | e.g. P/F ratio, PaO2, PaCO2, pH, bicarbonate |
| Blood inflammatory biomarkers | 28 days | upon availability in the patient's electronic chart, e.g. CRP, lactate |
| Sedation depth | 28 days | Richmond Agitation-Sedation Scale (RASS-score). This score ranges from -5 to +4, where a more positive score indicates more agitation. |
| Reported delirium | 28 days | Reported delirium, observed via the Delirium Observation Screening (DOS), or the Intensive Care Delirium Screening Checklist (ICDSC), depending on the standard of care of the participating center. The DOS ranges from 0-13, with a score ≥3 indicating delirium. The ICDSC ranges from 0-8, with 0-3 indicating absence of delirium and a score of ≥4 inidicating delirium. |
| Illness severity score (SOFA-score) | 28 days | SOFA-score. This score ranges from 0-24, with increasing scores reflecting more abnormal physiology and biochemistry or an increasing degree of intervention. |
| ICU mortality | 90 days | Mortality during ICU stay |
| Mortality at day 28 | 28 days | Mortality at day 28 |
| Mortality at day 90 | 90 days | Mortality at day 90 |
| Duration of ventilation | 28 days | Duration of ventilation in hours or days |
| Ventilator free days (at day 28) | 28 days | Number of ventilator free days at day 28 |
| Ventilator free days (at day 90) | 90 days | Number of ventilator free days at day 90 |
| Reintubation rate | 28 days | Number of times reintubation occured, reported as % of patients needing reintubation. |
| Weaning success | 28 days | Weaning success rate (%), defined as seven consecutive days without ventilator support |
| ICU length of stay | 28 days | Length of ICU stay, measured in days |
| Complications | 28 days | Complications (e.g. ventilator associated pneumonia and ICU acquired weakness) as reported in the patient file. |
Countries
Netherlands
Contacts
CONTACTAnnemijn Jonkman, PhD
Outcome results
None listed