Long-Term Recovery and Quality of Life in Sepsis Survivors

Sepsis is a leading cause of death and long-term disability worldwide. Survivors often experience persistent physical, psychological and cognitive impairments, described as Post-Sepsis Syndrome in addition to Post-Intensive Care Syndrome (PICS). Despite growing recognition of these complications, the factors influencing long-term outcomes - particularly the interaction between individual characteristics, infection-related factors and centre-level practices - remain poorly understood.

UNCOVER is a prospective, international, multicentre observational cohort study designed to investigate these associations in adult ICU patients with sepsis. Geneva University Hospitals (HUG) acts as the promoter. Participating centres enrol patients consecutively; each centre targets 30 consecutive patients during a six-month enrolment period, and enrolment stops once 600 participants have been included. Additional participating ICUs in Europe, the Americas, Asia-Pacific, the Middle East and Africa will be added to this record as local ethics approvals are obtained.

Data are collected from medical records at ICU admission, on day 7 and at ICU discharge, and at follow-up 28 days, 3 months and 6 months after ICU admission. Collected variables include:

* Centre-level data (collected once per centre): hospital and ICU admission rates, availability of infectious disease consultants and clinical pharmacists, physiotherapist-to-bed ratio, nurse-to-patient and doctor-to-patient ratios, junior/senior physician ratio, proportion of mechanical ventilation beds, existence of a local sepsis bundle, family visiting policy, open/closed ICU, public/private funding, university/non-university status, admissions and patient-days during the study period.
* Admission data: socio-demographics (age, sex, nationality, ethnicity, employment, financial status), APACHE II, SOFA, comorbidities (Charlson Comorbidity Index and individual comorbidities), Clinical Frailty Scale, CAM-ICU, Glasgow Coma Scale, baseline psychiatric comorbidity, code status (full code, DNR, DNI), intubation, vasopressors.
* Day 7 data: SOFA, GCS, delirium (CAM-ICU), cultures obtained and positive cultures with pathogen, resistance and antibiogram details, antibiotic adaptation, code status, intubation, vasopressors.
* ICU discharge data: source of infection, source control measures, time to adequate antibiotic therapy, type and duration of antibiotic therapy, new sepsis episodes during the ICU stay, imaging, mechanical ventilation and its duration, tracheostomy, new organ failures, dialysis, ECMO, delirium duration and treatment, caloric intake and deficit, thiamine and steroid administration, Barthel Index at ICU discharge.
* Environmental data: ICU diary, number of physiotherapy sessions per ICU length of stay, family presence at the bedside.

Because patients with sepsis are in an emergency situation and frequently lack decision-making capacity (altered consciousness, sedation, intubation), immediate inclusion with deferred consent is used, in accordance with articles 30 and 31 of the Swiss Human Research Act: advance directives are consulted, an independent physician not involved in the research safeguards the patient's interests, proxy consent is obtained from the legal representative or next of kin as soon as possible, and consent is requested from the patient post hoc as soon as capacity is recovered.

The primary outcome is health-related quality of life measured with the EQ-5D-5L at 6 months; poor quality of life is defined a priori as an EQ-5D-5L index value below 0.6. Secondary outcomes include 6-month mortality, broader quality of life (SF-12), cognitive function (T-MoCA), post-traumatic stress symptoms (IES-R), functional status (Barthel Index, extended Barthel, IADL), frailty (Clinical Frailty Scale), return to work and socio-economic recovery, rehospitalisation and healthcare utilisation, and patient-reported decision regret and preferences regarding future ICU admission.

Pre-specified secondary analyses will examine the association between long-term patient-centred outcomes and: sepsis-related delirium during the ICU stay; duration of antibiotic therapy (7 days or less versus more than 7 days); time to adequate antibiotic therapy; hospital-acquired versus community-acquired sepsis; immunosuppression; hyperoxia exposure during the ICU stay; and clinical parameters before initiation of renal replacement therapy as predictors of renal recovery and associated long-term quality of life.

The primary analysis uses a hierarchical logistic regression model with centre as a random intercept, with predictors at centre level (country), patient level (Clinical Frailty Scale, Charlson Comorbidity Index, SOFA) and infection level (time to adequate antibiotic therapy, source control achieved). Six-month mortality is analysed with a shared-frailty Cox model with centre as the frailty term. Analyses are performed in R (version 4.3 or higher) and STATA 16.1.

No experimental intervention is administered; all treatments follow local standard of care and remain the responsibility of the attending medical team. Data are coded at source and entered into a secure REDCap registry hosted by the Swiss Society of Intensive Care Medicine (SGI/SSMI).

The study is supported by a EUR 20,000 grant from the European Society of Intensive Care Medicine (ESICM), which has no role in the design, conduct, data collection, analysis, interpretation or decision to publish.

• Care for Sepsis (Other): Participants will receive standard clinical care for sepsis according to local hospital protocols. This study is purely observational: no experimental drugs, procedures, or behavioral interventions are introduced. Data will be collected prospectively on infection characteristics, timing of adequate antibiotic therapy, organ support, and rehabilitation practices during the ICU stay. Follow-up at 28 days, 3 months, and 6 months will assess patient-centered outcomes such as quality of life (EQ-5D-5L), cognitive and mental health status, and return to work. All treatments and decisions remain under the responsibility of the attending medical team.

Primary outcomes

• Health-related Quality of Life [6 months after ICU admission]: Health-related quality of life measured with the EQ-5D-5L. The EQ-5D-5L index value ranges approximately from below 0 to 1, with higher values indicating better health-related quality of life. Poor quality of life is defined a priori as an index value below 0.6.

Secondary outcomes

• Health-related Quality of Life [7 days after ICU admission]: Health-related quality of life measured with the EQ-5D-5L. The EQ-5D-5L index value ranges approximately from below 0 to 1, with higher values indicating better health-related quality of life. Poor quality of life is defined a priori as an index value below 0.6.
• Health-related Quality of Life [28 days after ICU admission]: Health-related quality of life measured with the EQ-5D-5L. The EQ-5D-5L index value ranges approximately from below 0 to 1, with higher values indicating better health-related quality of life. Poor quality of life is defined a priori as an index value below 0.6.
• Health-related Quality of Life [3 months after ICU admission]: Health-related quality of life measured with the EQ-5D-5L. The EQ-5D-5L index value ranges approximately from below 0 to 1, with higher values indicating better health-related quality of life. Poor quality of life is defined a priori as an index value below 0.6.
• 6-Month Mortality [6 months after ICU admission]: All-cause mortality within 6 months of ICU admission, ascertained from medical records or during structured follow-up contact. Includes deaths during the index hospitalisation and after discharge. Cause of death is recorded, including death following withdrawal or withholding of life-sustaining treatment.
• Cognitive Function [6 months after ICU admission]: Cognitive function assessed by the Telephone Montreal Cognitive Assessment (T-MoCA). Lower scores indicate greater cognitive impairment. Cognitive impairment is defined a priori as a T-MoCA score below 12.
• Cognitive Function [3 months after ICU admission]: Cognitive function assessed by the Telephone Montreal Cognitive Assessment (T-MoCA). Lower scores indicate greater cognitive impairment. Cognitive impairment is defined a priori as a T-MoCA score below 12.
• Psychological Distress and Post-Traumatic Stress Symptoms [6 months after ICU admission]: Post-traumatic stress symptoms assessed with the Impact of Event Scale-Revised (IES-R, range 0 to 88). Analysed as a continuous score; clinically significant post-traumatic stress is defined a priori as a score above 33.
• Psychological Distress and Post-Traumatic Stress Symptoms [3 months after ICU admission]: Post-traumatic stress symptoms assessed with the Impact of Event Scale-Revised (IES-R, range 0 to 88). Analysed as a continuous score; clinically significant post-traumatic stress is defined a priori as a score above 33.
• Functional Status [3 months after ICU admission]: Barthel Index recorded at ICU discharge and at day 28, from medical record review or telephone contact if the patient has already been discharged. Lower scores indicate greater functional dependency.
• Functional Status [6 months after ICU admission]: Barthel Index recorded at ICU discharge and at day 28, from medical record review or telephone contact if the patient has already been discharged. Lower scores indicate greater functional dependency.
• Return to Work and Socioeconomic Recovery [6 months after ICU admission]: Self-reported return to paid employment, whether family caregivers resumed work, whether the participant receives nursing support at home, and financial status (no difficulty, new difficulty, or same as before sepsis), recorded during structured telephone follow-up.
• Return to Work and Socioeconomic Recovery [3 months after ICU admission]: Self-reported return to paid employment, whether family caregivers resumed work, whether the participant receives nursing support at home, and financial status (no difficulty, new difficulty, or same as before sepsis), recorded during structured telephone follow-up.
• Rehospitalization and Healthcare Utilization [6 months after ICU admission]: Cumulative number and causes of rehospitalisations and outpatient consultations, collected through medical record review and participant self-report during the follow-up interview.
• Decision regret regarding the ICU stay (Decision Regret Scale) [6 months after ICU admission]: Regret regarding the ICU stay assessed with the five-item Decision Regret Scale, administered during structured telephone follow-up. Item responses are summed and linearly transformed to a total score ranging from 0 (no regret) to 100 (maximum regret); higher scores indicate greater regret. The continuous score is the registered outcome measure. For dichotomised analyses, moderate-to-strong regret is defined a priori as a score above 25, with a score of 0 indicating no regret and 1 to 25 indicating mild regret. Where a participating centre administers a validated language version with a different reported threshold, that threshold will be applied in pre-specified sensitivity analyses.
• Freedom from renal replacement therapy [6 months after ICU admission]: Percentage of participants who are alive and free from renal replacement therapy 6 months after ICU admission, among those who received renal replacement therapy for sepsis-associated acute kidney injury during the ICU stay. Freedom from renal replacement therapy is defined as no renal replacement therapy of any modality during the 30 days preceding the 6-month assessment, ascertained from medical record review and structured telephone follow-up. Participants who died before the 6-month assessment are counted as not having recovered renal function.
• Physical health-related quality of life (Short Form-12 Physical Component Summary) [3 months after ICU admission]: Physical Component Summary (PCS) score of the Short Form-12 Health Survey (SF-12), administered during structured telephone follow-up. The PCS is a norm-based score calibrated to a general-population mean of 50 and standard deviation of 10, ranging from 0 (worst physical health) to 100 (best physical health); higher scores indicate better physical health. The continuous score is the registered outcome measure. For dichotomised analyses, impaired physical health is defined a priori as a PCS score more than one standard deviation below the general-population norm (score below 40); where a country-specific general-population norm is available for a participating centre, that norm will be used in pre-specified sensitivity analyses.
• Delirium during the ICU stay [6 months after ICU admission]: Presence of delirium assessed with the CAM-ICU during the ICU stay, with duration of delirium and treatments administered.
• Mental health-related quality of life (Short Form-12 Mental Component Summary) [3 months after ICU admission]: Mental Component Summary (MCS) score of the Short Form-12 Health Survey (SF-12), administered during structured telephone follow-up. The MCS is a norm-based score calibrated to a general-population mean of 50 and standard deviation of 10, ranging from 0 (worst mental health) to 100 (best mental health); higher scores indicate better mental health. The continuous score is the registered outcome measure. For dichotomised analyses, impaired mental health is defined a priori as an MCS score more than one standard deviation below the general-population norm (score below 40); where a country-specific general-population norm is available for a participating centre, that norm will be used in pre-specified sensitivity analyses.
• Mental health-related quality of life (Short Form-12 Mental Component Summary) [6 months after ICU admission]: Mental Component Summary (MCS) score of the Short Form-12 Health Survey (SF-12), administered during structured telephone follow-up. The MCS is a norm-based score calibrated to a general-population mean of 50 and standard deviation of 10, ranging from 0 (worst mental health) to 100 (best mental health); higher scores indicate better mental health. The continuous score is the registered outcome measure. For dichotomised analyses, impaired mental health is defined a priori as an MCS score more than one standard deviation below the general-population norm (score below 40); where a country-specific general-population norm is available for a participating centre, that norm will be used in pre-specified sensitivity analyses.
• Physical health-related quality of life (Short Form-12 Physical Component Summary) [6 months after ICU admission]: Physical Component Summary (PCS) score of the Short Form-12 Health Survey (SF-12), administered during structured telephone follow-up. The PCS is a norm-based score calibrated to a general-population mean of 50 and standard deviation of 10, ranging from 0 (worst physical health) to 100 (best physical health); higher scores indicate better physical health. The continuous score is the registered outcome measure. For dichotomised analyses, impaired physical health is defined a priori as a PCS score more than one standard deviation below the general-population norm (score below 40); where a country-specific general-population norm is available for a participating centre, that norm will be used in pre-specified sensitivity analyses.

Sex: All, ages: 18 Years to any

Inclusion Criteria:

* Adults aged 18 years or older
* Sepsis (community- or hospital-acquired) diagnosed on ICU admission, defined as either: (a) a diagnosis of sepsis made by the treating clinician at ICU admission AND fulfilment of Sepsis-3 criteria; or (b) a diagnosis of sepsis made by the treating clinician at ICU admission alone, where Sepsis-3 criteria are not routinely available according to local protocol

Exclusion Criteria:

* Refusal to participate, by the patient or the legal representative
* Inability to ensure follow-up (e.g. overseas residence, no fixed home address)
* Sepsis acquired in the ICU
* Patient admitted for palliative care

• Niccolo' Buetti, MD, Study chair, Geneva University Hospitals (HUG)