Critical illness can alter the life trajectory of patients admitted to the intensive care unit (ICU) and result in a traumatic experience for the family. Therefore, one of the current challenges of intensive care medicine is to promote long-term continuous care beyondthe boundaries of ICU and hospital, through follow-up in post-ICU consultations.1 With adequate prevention, detection, and follow-up, we can minimize the occurrence of symptoms and signs related to the Table 6 (PICS), both in patients (PICS-P) and primary caregivers or family members (PICS-F).2,3

In a systematic review, 60 risk factors were identified for developing PICS-P, approximately half of which were related to the patient and the other half to the ICU.4 Interventions should target modifiable risk factors related to the ICU. The application of the ABCDEF bundle may improve patient-centered care, as well as the experience of patients and their families.5

In COVID-19 patients, multiple risk factors related to PICS6,7 have been identified, in addition to barriers hindering the application of the ABCDEF bundle,8 making this population especially vulnerable to syndrome development. Furthermore, severe acute respiratory illness caused by SARS-CoV-2 infection may cause long-lasting symptoms (“long COVID”), persisting long after hospital discharge and potentially overlapping with PICS, exacerbating its symptoms.9 PICS-F may also have worsened during the pandemic due to distancing from loved ones and difficulties in providing support during bereavement.10 This has led to the recommendation to conduct post-ICU consultations for selected patients, with an initial individualized visit after hospital discharge, to predict, identify, and treat long-term problems related to critical illness.11,12

The main objective of this study is to compare the prevalence and sociodemographic and clinical characteristics of PICS-P and PICS-F in 2 cohorts of patients (COVID vs non-COVID). The secondary objective is to identify risk factors associated with PICS-P.

A total of 265 patients and 209 family members/caregivers were analyzed. Among them, 104 (39.2%) were admitted for COVID-19 (COVID group) and 161 (60.8%) for other etiologies (non-COVID group).

The most important finding of this study is that the prevalence of PICS-P and PICS-F was similar in both cohorts (COVID and non-COVID) at the first follow-up consultation, with rates of 65.2% and 19.1% in both patients and family members, respectively. However, at the 1-year follow-up, a higher percentage of non-COVID patients still exhibited signs or symptoms of PICS-P. Additionally, there were differences in the type of changes exhibited by both cohorts of patients and family members. In patients, the most frequent were physical, especially in the COVID group (42.3% vs 29.2%, P = .028), at the expense of a higher, albeit not significant, number of musculoskeletal and respiratory changes, which, however, were more frequent in non-COVID patients at the 1-year follow-up. Regarding the degree of functional dependence and quality of life, these were worse in the non-COVID patient group.

There is significant variability in the prevalence of PICS-P and PICS-F2 across different studies, and the most common changes depend on the characteristics of the patienets and the evaluation methods used. In COVID-19 patients, this variability is equal to or even greater. Various studies conducted in France,22 the United States,6 and the Netherlands,23 involving COVID-19 patients on mechanical ventilation, have shown very significantly different results. In Spain, a study conducted on ventilated COVID-19 patients found that approximately 3 out of every 4 patients met PICS criteria at the 3-months follow-up.24 One of the main limitations of these studies is the lack of comparison with a non-COVID-19 patient cohort, and the fact that most follow-ups (except the Spanish study) were not conducted in person.

Several authors have reported that the physical, psychological, and cognitive deficits observed in ICU survivors after COVID-19 are comparable to those observed in patients with other diseases.25,26 Additionally, it has been demonstrated that ARDS survivors have persistent functional limitations 1 year after ICU discharge due to muscle atrophy and weakness,27 and that the inability to exercise even at the 5-year follow-up is mainly due to extrapulmonary causes.28–30 Similarly, prolonged mechanical ventilation (MV) is a risk factor for long-term physical changes.24 This would explain the higher percentage of functional changes reported in COVID patients who were primarily included due to MV > 48 h and ARDS.

However, age,31 pre-existing comorbidities,1 nutritional changes, and disease severity4 have been identified as determinants of poor prognosis related to increased functional dependency and worse long-term quality of life, all of which are present in our non-COVID patients. Furthermore, the severity of illness upon ICU admission has also been associated with a higher rate of ICU and hospital readmission, both significantly higher in our non-COVID patient group.

Cognitive impairment was also more common in non-COVID patients, due to the older age of these patients, but also to the longer duration of deep sedation and higher incidence of delirium, both factors clearly associated with long-term cognitive impairments.32,33 Deep sedation contributes to increased mortality and worsens clinical outcomes, both in the short- and long-term (delirium and cognitive impairment).34,35

Therefore, although physical changes (primarily musculoskeletal and respiratory) were more frequent in the COVID group, as a consequence of their respiratory disease (more patients with ARDS and mechanical ventilation > 48 h), the greater presence of other predictive factors of poor prognosis in the non-COVID patient group, such as older age, comorbidities, worse nutritional status, and disease severity, influenced and significantly impacted the quality of life and degree of dependency within the first few months of follow-up, as well as the persistence of PICS-P 1 year after hospital discharge. These factors also had an impact on the hospital and ICU readmission rates at 1 year.1 In addition to these non-modifiable factors, other factors external to the patient and preventable in the ICU, such as deep sedation and delirium, contributed to a greater long-term cognitive decline in COVID patients.

PICS-F was present in 19.1% of family members and/or caregivers. The most frequent change was family burden (P = .014), which was significantly higher in non-COVID patients' families (17.6% vs 5.5%; P = .013), followed by anxiety. There is great variability in the prevalence of PICS-F in different studies,36,37 with mental and physical changes being described in up to 60% and 40%, respectively, depending on the variables analyzed and methods used. Similarly, multiple risk factors for PICS-F38 have been identified such as disease severity, communication, female sex, etc. In our series, the incidence was lower, and caregivers who experienced greater burden were those of patients with more dependency and worse quality of life.

The risk factors identified in our patients for developing PICS-P were greater comorbidity, delirium, several reasons for post-ICU follow-up consultation, treatment with midazolam, and being admitted within the early years of the study. The association of pre-existing comorbidity, delirium, and other factors such as age with PICS has been previously discussed. Additionally, the use of midazolam has also been identified as an important factor associated with this syndrome.39 Regarding being admitted in the early years of the study, this relationship could be due to the greater implementation of the ABCDEF bundle throughout the study period, which has been shown to improve the patients' long-term prognosis.1,40 In our ICU, we gradually implemented a program to prevent PICS (optimization and monitoring of analgosedation and delirium, flexible visiting hours, improvement of communication and entertainment through clocks, TVs, augmentative communicators, etc.), culminating in 2020 (during the pandemic) with the addition of a physical therapist and a psychologist into our unit. This may have influenced the admission in the early years of the study as a risk factor for developing PICS and COVID patients presenting with less physical, functional, and cognitive deterioration than non-COVID patients (most of whom were admitted before the pandemic).

This work has strengths and weaknesses. It is a study with face-to-face follow-up, including a relevant number of patients and family members, and compares 2 different cohorts: COVID and non-COVID, which is a strength of the study. Regarding weaknesses, firstly, due to its single-center design, the generalization of results may be compromised to some extent. Secondly, due to follow-up, some of the evolving variables have not yet been measured at the time of this work. Finally, although the inclusion of analyzed variables was certainly comprehensive, we may have overlooked some important variables. Nevertheless, we believe that the study's conclusions remain valid. We propose future research based on a national registry of PICS.

There were no differences in the incidence of PICS in patients and family members of both cohorts (COVID vs non-COVID) at the first follow-up consultation, although it was more frequent in non-COVID patients at 1 year. The main factors associated with developing PICS-P were greater comorbidity, presence of delirium, sedation with midazolam at the ICU stay, as well as being included in the consultation for more than 1 reason and during the early years of the study, when the implementation of the ABCDEF bundle was lower.

No total or partial funding was received to conduct study, nor any type of grant or financial support.

None declared.

Carola Giménez-Esparza Vich: Data mining. Elaboration and drafting of the article.

Beatriz Hurtado Oliver: Data mining. Drafting.

Maria Angeles Relucio Martinez: Data mining. Drafting.

Salomé Sanchez Pino: Data mining.

Cristina Portillo Requena: Data mining.

José David Simón Simón: Data mining.

Isabel María Pérez Gómez: Data mining.

Fernando Mario Andrade Rodado: Data mining.

Fadoua Laghzaoui Harbouli: Data mining.

Fernando Javier Sotos Solano: Data mining.

Carlos Augusto Montenegro Moure: Data mining.

Andrés Carrillo Alcaraz: Statistical analysis and Drafting.