MONITOR-IC: Determining Long-term Consequences of ICU Care

Specific research questions:

1. What are the post intensive care symptoms that patients experience during five years after ICU admissions and what is their HRQoL?

2. What are important predictors for the various physical, cognitive and mental long-term outcomes?

3. What is the ratio between HRQoL and healthcare related costs?

4. What are the care and support needs of ICU survivors during five years after ICU admission?

Design:

The MONITOR-IC study is a multicenter prospective controlled cohort study in which long-term outcomes of ICU patients are studied for a period of five year.

The study will be carried out in ICUs of four hospitals in the Netherlands; one academic hospital, one teaching and two non-teaching hospitals. ICU patients will be recruited between July 2016 and July 2021 with a subsequent follow-up for five years.

Inclusion:

For the MONITOR-IC study the investigators estimated to include 12.000 patients. This estimation is based on: 1) the initial ICU admissions in the academic hospital and the three other participating hospitals together (2,500 and 2,200 respectively per year), and 2) an estimated response rate of 60%, which is based on previous conducted ICU studies. In the control cohort, the investigators will include approximately 3000 patients during the next four years.

Patient recruitment:

Patients scheduled for ICU admission after elective surgery, will be recruited at the outpatient clinic (anaesthesiology or cardiac surgery) (Figure 1). Patients with a non-scheduled admission will be recruited at the ICU. Patients will receive information by ICU nurses and intensivists regarding the aim, content and relevance of the study, and will be asked for participation. Informed consent is asked for the questionnaires, data from the patients' individual medical record and data from their health insurance company. In case patients are unable to give consent, their legal representative will be asked.

Methods:

Mixed methods will be used to collect data, including questionnaires, medical data from patient records, cost data from health insurance companies and interviews with ICU survivors and their family members.

All patients, or their relatives in case patients are not able to fill in the questionnaire themselves, will be approached to fill in the self-administered paper based or online questionnaire in total eight times:at ICU admission (T0), at hospital discharge (T1), after 3 months (T2), 12 months (T3), 24 months (T4), 36 months (T5), 48 months (T6) and 60 months after ICU admission (T7). To get insight into the situation before the ICU admission, the baseline questionnaire (T0) is provided when the patients is asked for informed consent. This could be preoperatively for the planned admissions or after admission at the ICU. Then, patients are asked to rate their situation before the ICU admission.

Items in questionnaire:

- Patients' health status and HRQoL = SF-36 and the EQ-5D-5L.

- Patients' level of frailty and vulnerability = Clinical Frailty Score (CFS)

- The level of fatigue = CIS-8, a subscale of the Checklist Individual Strength

- Cognitive impairment = abbreviated 14-item Cognitive Failure Questionnaire (CFQ-14).

- Anxiety and depression = Hospital Anxiety and Depression Scale (HADS)

- Subjective distress, caused by traumatic events = Impact of Event Scale Revised (IES-R)

- Care needs and support from professionals and informal caregiver

- Social consequences = novel question set for changes in family circumstances, socio-economic stability and care requirements.

Medical data Patients' demographics and information regarding their diagnosis and treatment, such as primary conditions, pre-existing co-morbidity, disease severity, sepsis, (re)admission, length of mechanical ventilation, length of ICU stay, delirium (CAM-ICU), pain (CPOT), expected mortality (based on the APACHE II, SAPS II, MPM II and APACHE III-IV models and medication, will be extracted from their medical record and the Dutch National Intensive Care Evaluation (NICE) registry.

Health Insurance data Healthcare use and related costs, covered by the Dutch healthcare insurance, will be retrieved from Vektis; a Dutch organization which collects and manage health insurance claimed data of all health insurance companies in the Netherlands.This data is collected based on the Diagnosis Treatment Combination (DTC); a total set of activities carried out by the hospital and medical specialists. Additionally, data is collected regarding nursing days, visits at the outpatient clinic and emergency department, nursing homes, ambulance transport, consultation with general practitioner, paramedical care (including physiotherapist, occupational therapist, dietician and speech therapist), prescribed medication, mental healthcare and revalidation. The Vektis database contains data from all for healthcare insured citizens and covers 99% of the total Dutch population. Using patient's unique insurance number the investigators are able to merge patient's insurance data with the questionnaire data and medical data from the medical record at patient level.

Care delivered by community nurses and informal caregivers is not included in the Vektis database and will be studied via the questionnaire.

Interviews To get insight into the experiences of ICU survivors during five years after ICU admission and their need for support, face to face semi-structured interviews will be conducted with ICU survivors and their family members. Interviews will take place at the participants' preferred location (home or clinic). Interviews will be conducted until data saturation is reached.

Patients will be purposively sampled based on various experienced outcomes, such as the quality of life, daily functioning, anxiety, depression, and their experienced needs for more information or emotional support. Experienced and trained researchers will conduct the interviews using a topic guide. This guide will be developed using the current literature and experience of the research team and will cover the following subjects: experiences with the ICU admission and follow-up, experienced problems and needs for support. All interviews will be audio recorded and transcribed verbatim.

Analysis Questionnaires, medical record and health insurance data During the data collection, data are checked on a regular basis to identify out-of-range answers, inconsistent responses, and missing data. Data from the questionnaires, medical record and healthcare insurance data will be merged at patient level. Descriptive statistics will be used to describe baseline characteristics and the incidence of long-term outcomes. Regression analysis will be used to determine associations between patient characteristics, treatment, and long-term outcomes. Subgroups will be identified based on their illness and condition (for example sepsis, delirium, co morbidities, ARDS), treatment (for example length of ICU stay, duration of mechanical ventilation, dialysis) and social demographics (age, gender, education, family setting etc).

In order to predict the various physical, cognitive and mental long-term outcomes, multiple prediction models will be developed. Multivariable linear (for continuous outcome variables) and logistic (for dichotomous outcome variables) regression analysis will be performed. Linear and logistic multilevel models will be used to compare long-term outcomes between the study population (cohort) and control cohort group.

To determine the ratio between HRQoL and patient outcomes and the health related costs, quality adjusted life years (QALYs) will be calculated. QALYs are a measure of the state of health of a person or group in which the benefits, in terms of length of life, are adjusted to reflect the quality of life. QALYs are calculated by estimating the years of life remaining for particular treatment and weighting each year with a quality of life score.54 Software Package for the Social Sciences (SPSS) will be used for data analysis.

Interviews For the analysis of the interview data, the constant comparative method will be used. Relevant data will be identified and structured by open, axial and selective coding.

Two researchers will independently code the transcripts to minimize subjectivity in findings. The differences and similarities between the codes will be discussed together, and in case of disagreement, a third researcher will be involved. In the meetings with the team, the codebook will be refined and emerging categories and themes will be discussed.

Data analysis will be supported with the use of Atlas.ti, a qualitative data analysis program.