Data screening process and results
A total of 1,591 records were extracted from the main databases. Of these, 567 records were removed due to duplication and 986 due to not containing the term “ICU” or lacking guideline and consensus statement content. Four additional records were included following review of the bibliography of the obtained records. A search in the supplementary databases yielded 3 additional guidelines. After excluding records that focused explicitly on medication, as well as all duplicates, 32 research papers were retained for analysis, including 19 guidelines and 11 expert consensus statements (Figure.1).
Basic characteristics
The systematic search and screening process yielded no PICS-specific guidelines; in total, 32 guidelines and consensus statements reporting related content, such as risk factors, monitoring, diagnosis, evaluation, prevention, and treatment of PICS and PICS-F were retrieved (Table.1). The 19 guidelines dated from 2002 to May 2020. After excluding the guidelines that were updated to other included versions, 14 guidelines were eventually retained (13 evidence-based guidelines and one general guideline). Majority of the guidelines were published by organizations from European countries (such as the United Kingdom, Germany, and Spain), North America (mainly the United States), Asian countries (such as China and Japan), and Oceania (mainly Australia). The organizations included the SCCM, the American College of Critical Care Medicine (ACCM), SIGN, Pan-American and Iberian Federation of Societies of Critical Medicine and Intensive Therapy (PAIFSCMIT), NICE, and the Chinese Medical Association (CMA). Four of the guidelines were developed through multinational cooperation [12-15], nine were developed by the corresponding state [16-24], and one was developed by a hospital [25]. In addition, the guidelines were divided into single and bundled theme guidelines. Single guidelines included the prevention, management and non-pharmacological intervention of delirium [15, 16, 21], ICUAW diagnosis [20], rest and sleep improvement [25], rehabilitation of critically ill patients [24], and family-centered care [14]. Bundled guidelines included sedation/analgesia [17], management of PAD [12, 19, 22], and prevention and management of PASDIS [13]. Of these, nine guidelines that covered sedation/analgesia, family-centered care and PAD bundling management [26-31] were updated, indicating good continuity and timeliness. Across all guidelines, the main applicable population was critically ill adult patients. Only a few guidelines described the primary disease of the patient or were applicable to caregivers and family members. The main applicable institutions were ICUs, whilst guidelines applicable for standardized general wards, primary and secondary health care institutions, communities, and homecare were limited. The main targeted users were multidisciplinary professional and technical personnel in ICUs, while guidelines for primary health care professionals and community and family caregivers were insufficient.
The screened 10 expert consensus statements were dated between 2013 and 2019, including six evidence-based consensus statements [32-37]. The consensus statements that were published by academic organizations focused on critical cardiac, respiratory, and neurological care, from China, New Zealand, Canada, the United Kingdom, and Australia independently. Only the consensus statements that focused on the rehabilitation of discharged patients [34], physiotherapy for ICU patients [35], and safety criteria of active mobilization for mechanically ventilated patients [37] were developed through cross-regional cooperation. The themes covered in the consensus statements included the rehabilitation of ICU or hospital discharged patients [34, 38-43], safety criteria for the early mobilization of mechanically ventilated patients [37], sedation/analgesia for ICU patients [33], delirium management [32], and physical therapy [35]. The majority of the consensus statements had individual themes rather than being a bundled theme. Only one consensus statement clearly defined PICS/PICS-F and provided physical therapy guidelines to prevent PICS/PICS-F [34]. Seven consensus statements clearly defined the primary disease stages and the applicable population, including patients with severely critical respiratory, cardiovascular, and neurological systems; and mechanically ventilated patients [33, 36-43]. The remaining three consensus statements cited only critically ill patients or patients who survived severe illness. Only one consensus statement identified its users as ICU physical therapists [42]. All consensus statements were applicable to the ICU, however; but only one was applicable to patients following discharge from the hospital. None were developed for in-patients transferred from the ICU.
Methodological quality
The AGREE-II scores of the guidelines were as follows (Figure.2): (1) The mean for scope and purpose was 86.89% (±16.12%). Specifically, 13 guidelines clearly described the purpose, specific clinical problems, and patients targeted; only one guideline did not clearly report its scope and purpose. (2) The mean for stakeholder involvement was 55.86% (±29.17%). Specifically, seven guidelines described the fields of specialization of the experts involved; 10 introduced the applicable population; five specified the views and choices of the target patients. (3) The mean for rigor of development was 76.35% (±15.82%). Specifically, four guidelines provided evidence retrieval steps and screening processes; eight clearly specified the strengths and weaknesses of the evidence selection process; nine explained how the recommendations were formed and the association between the recommendations and evidence; seven presented the health benefits, adverse effects, and risks of the recommendations, and 10 provided the comments given by external experts and future updates plans. (4) The mean for the clarity of presentation was 88.03% (±15.55%). Specifically, the main recommendations of 10 guidelines were clear and identifiable and the remaining nine guidelines were clear and provided multiple options for a given clinical problem. (5) The mean for applicability was 61.79% (±23.16%). Specifically, nine guidelines included likely facilitating and hindering factors during application; eight provided suggestions on ways to apply the recommendations in practice and the support tools required; seven explained the potential resource input issues during application; six specified the standards for monitoring the use of guidelines. (6) The mean for editorial independence was 66.67% (±37.96%), where eight guidelines declared the relationship of interest between the sponsors and recommendations and four provided any conflicts of interest between the editing team members and recommendations made. According to the overall AGREE-II score, 11 guidelines had satisfactory methodological quality (> 60%) and could be recommended. The scope and purpose, rigor of development, and clarity of presentation of the investigated guidelines were of high quality. However, stakeholder involvement (participially the involvement of patients/caregivers), applicability, and editorial independence required further improvement in quality.
The AGREE-II scores of the 10 expert consensus statements (six of which were evidence-based) were as follows (Figure.3): (1) The mean for scope and purpose was 79.47% (±11.80%). The investigated consensus statements all described the purpose and targeted clinical problems and patients. (2) The mean for stakeholder involvement was 16.72% (±16.39%). Only two consensus statements specified the applicable population, one considered the viewpoints and choices provided by the target patients, and all 10 provided the authors’ professional background. (3) The mean for rigor of development was 36.09% (±26.05%). Only two statements provided evidence of the retrieval and screening processes, three explained the strengths and weaknesses of evidence selection processes, and only one elaborated on the relationship between the supporting evidence and the recommendations and its formulation process. In addition, none of the consensus statements were reviewed by external experts prior to publication, and only two provided an update plan. (4) The mean for clarity of presentation was 58.84% (±28.94%). Specifically, six included clear and readable recommendations, and three clearly listed different recommendations for a given clinical problem. (5) The mean for applicability was 32.95% (±20.91%). Only four consensus statements explained the potential facilitating and hindering factors during the application process, three provided suggestions and supporting tools for applying the recommendations, two introduced standards for supervision and auditing, and none of the consensus statement explained the impact of recommendations on potential resource allocation. (6) The mean for editorial independence was 28.79% (±36.54%). Specifically, three consensus statements declared that no conflict of interest existed between the sponsors, editing team members, and the recommendations made. The overall scores revealed that only one consensus statement was methodologically qualified (> 60%) and could be recommended. The quality in terms of scope and purpose of the investigated consensus statements was high. However, the quality of the remaining domains, such as rigor of development, clarity of presentation, editorial independence, and applicability, was low. It was evident that the methodological quality of the consensus statements required improvement, particularly in their rigor of development and applicability. Moreover, six of the consensus statements were evidence-based, indicating that evidence-based decision-making concepts and techniques have begun to be applied to the development of consensus statements.
Reporting quality
The mean for the overall RIGHT score of the 14 guidelines (35 items) was 58.16% (±30.31%). Specifically, 23 items were found to have a reporting quality of greater than or equal to 50%. Items with higher reporting quality included 1a, 1b, 4, 7a, and 13a (100.0%), while items with lower reporting quality included 18a and 18b (0%). For quality by domain, basic information, guideline background, evidence, recommendations, and other information had higher reporting ratings, while insufficient information was provided in terms of rationale/explanation for recommendations, reviews and quality assurance, funding source(s), and the role(s) of the funding parties (Table 2).
The mean overall RIGHT score for the 10 consensus statements [22, 32, 33, 35-43] was 36.57% (±35.33%). Specifically, 23 items were found to have a reporting quality rating of greater than or equal to 50%, 5 items had a reporting quality of 100% (1a, 1b, 1c, 4, and 7a), while 10 items had a reporting quality of 0% (3, 8b, 10a, 10b, 14a, 14b, 18a, 18b, 19b, and 20). For domain-focused quality of reporting, basic information, background of the guidelines, recommendations, and other information had higher reporting ratings. However, the quality of reporting for health problems, evidence sources and evaluations, rationale/explanation for the recommendations, reviews and quality assurance, funding sources and declarations, management of interest, and funder’s access to the consensus statements was insufficient (Table 2).
Content features of the guidelines/consensus statements
This study included 14 clinical practice guidelines (13 evidence-based) [12-25] and 10 translational consensus statements (6 evidence-based) [22, 32, 33, 35-43]. Based on definitions framework of the PICS/PICS-F, the analysis revealed that only one consensus statement clearly defined PICS/PICS-F, while the remaining documents only described one category of clinical problems normally attributed to PICS/PICS-F. Regarding physical dysfunctions, there was more focus on ICU-AW, acquired neuromuscular diseases, sleep disorders, and pain. Regarding psychological and mental disorders, the majority of guidelines and consensus statements focused on agitation, anxiety, and depression; there was insufficient focus on the PTSD of ICU patients and their family members. Regarding cognitive and behavioral disorders, the focus was on delirium; however, limited attention was given to persistent cognitive deficits and difficulties returning to the family, society, and workplace (Table.3).
An analysis based on the “patient-clinical problem-intervention-target” framework revealed that the guidelines and consensus statements covered patients that suffered cardiac and severe respiratory illnesses and acute respiratory distress syndrome (ARDS), patients that were mechanically ventilated, treated by extracorporeal membrane oxygenation (ECMO), or suffered from brain trauma, as well as ICU survivors and the caregivers of these patients. However, many of the guidelines mentioned critically ill patients without providing information on the primary diseases. Secondly, regarding clinical problem, only one consensus statement provided a definition of PICS/PICS-F. In addition, the majority of the guidelines and consensus statements provided a clear introduction of ICUAW, pain, anxiety, sleep disorders, depression, agitation, and delirium. Some only described the rehabilitation of critically ill patients without clear provisions of specific clinical problems. Thirdly, regarding interventions, most of the guidelines and consensus statements clearly illustrated measures such as ICUAW diagnosis, physical therapy, early mobilization and safety criteria, sleep management, sedation/analgesia, anxiety/depression management, risk, monitoring, prevention and management of delirium, and family-centered care. Several guidelines/consensus statements employed bundling strategies to tackle the problems and design interventions. Some guidelines only described the rehabilitation of critically ill patients but did not clarify the specific rehabilitation interventions utilized. Finally, regarding management target, the majority of the guidelines and consensus statements focused on solving sleep problems, pain (sedation/analgesia), anxiety/depression, and delirium during ICU hospitalization; only one guideline focused on physical therapy to prevent the occurrence of PICS/PICS-F, and no guidelines or consensus statement discussed the patients/caregivers/family members’ quality of life or patients’ difficulty in returning to the family, workplace, and society. Moreover, it was found that the guidelines on the rest and sleep management ICU patients developed by Elliott & McKinley[25], which were formulated using an evidence-based strategy based on internal data collected from their own hospital, could effectively improve the guideline’s clinical translational effects (Table.3).