NANDA-I ® nursing diagnoses in adult critical patients with COVID-19

Objective: To identify possible nursing diagnoses according to the NANDA-International classification present in critically ill adult patients with COVID-19 based on diagnostic clues described in the scientific literature. Method: This is a descriptive study, developed in three stages: literature review and grouping of diagnostic clues identified according to Basic Human Needs; survey of NANDA-International nursing diagnoses based on the correspondence between diagnostic clues described in the literature with title and diagnostic indicators; validation of diagnostic correspondence by expert nurses. An agreement index ≥ 0.80 was used. Results: From the reading of


Introduction
COVID-19 refers to a new infectious disease on the world stage and, for this reason, there are gaps in the definition of its clinical, transmissible and therapeutic aspects. (1) Such gaps make the management of these patients challenging, especially in intensive care unit (ICU) environments, where there is a close relationship between life and death. For the nursing team, this challenge becomes even greater, given a scenario with shortage of human resources and devaluation of their practice. On the other hand, the relevance of nurses' clinical role is recognized in to quickly and accurately identify responses to the health conditions of each individual, considering the complexity of care required. (2) For greater effectiveness of nursing practices, there are systems for classification of nursing diagnoses, outcomes and interventions, useful tools to guide nurses' clinical reasoning. These systems consist of standardized languages used to improve and enable communication about the health status of individuals and, consequently, improve the care provided based on a scientific basis. (3) With regard to classification systems in the context of nursing diagnoses (ND), NANDA-International (NANDA-I ® ) stands out. (4) The survey of ND must be based on diagnostic reasoning that considers the so-called diagnostic clues. These can be understood as patient manifestations that represent signs, traces, signs, indications or characteristics of an ND. (5) Studies on ND in patients with COVID-19 were the target of investigations in the context of community health (6) and hospital with analysis of medical records. (7) Nationally, studies were identified based on clinical manifestations described in guidelines of the Ministry of Health (8) and using ICNP ® to relate diagnoses/outcomes and nursing interventions in patients severe cases affected by COVID-19 and sepsis. (9) It is suggested, therefore, that the survey of ND in critically ill patients with COVID-19 through diagnostic clues will favor a better understanding of the clinical aspects of this disease, which is still not fully elucidated in the literature.
In this context, with a view to improving the quality of nursing care for critically ill patients with COVID-19, it is essential to recognize which priority NDs are. In addition to this, the importance of documenting nursing care is emphasized, based on a standardized language, with a view to monitoring the clinical evolution of patients with COVID-19, for the implementation of evidence-based nursing interventions. Thus, this study aimed to identify possible ND according to the NANDA-I ® classification present in adult critical patients with COVID-19, based on diagnostic clues described in the scientific literature.
review, survey of NANDA-I ® ND from the diagnostic clues identified in the literature, and validation of diagnostic correspondence by expert nurses. To guide the presentation of information, the Standards for Quality Improvement Reporting Excellence (SQUIRE) version 2.0 were considered. (10) In the first stage, a literature review (11) was carried out with the following guiding question: "What are the existing clues for identifying ND in adult critically ill patients with COVID-19?".
The bibliographic survey was carried out in the Virtual Health Library (VHL), in the Cumulative Index of Nursing and Allied Health Literature (CINAHL), in the US National Library of Medicine, National Institutes of Health (PubMed), in Scopus, in the Web of Science, and in Embase. For searches in the databases, the controlled descriptors present in the Health Sciences Descriptors (DeCS) and in the Medical Subject Headings (MeSH), plus the Boolean operators "AND" and "OR" were used, resulting in following search strategy: ("Coronavirus Infections" OR "COVID-19") AND ("SARS Virus" OR "SARS-CoV" OR "CoV-SARS") AND (Signs and Symptoms) AND ("Review" OR "Systematic Review"). It is noteworthy that this step was carried out with the help of a librarian with experience in the field of health sciences.
The inclusion criteria used for sample selection were: review articles (narrative, integrative, systematic, scoping or umbrella) published between the years 2019 to 2020, available in full, that discussed diagnostic clues present in critically ill patients with COVID-19. Considering that the topic addressed is recent in the scientific community, the choice of review articles was chosen, as this design allows obtaining comprehensive information on the clinical manifestations of COVID-19 in the world context. Reviews that targeted children and pregnant women were excluded.
To remove duplicate studies, they were imported from the databases into Endnote ® . Subsequently, the screening of titles and abstracts was performed using Rayyan QCRI ® , independently by two researchers (Doctor 1 and 2) and the divergent cases were evaluated by a third researcher (Doctor 3).
As diagnostic clues, the manifestations of critically ill adult patients with COVID-19 that rep-resent signs, traces, signs, indications or characteristics of a ND were considered. (5) It is noteworthy that the identification of diagnostic clues was also performed by three researchers (Doctors 1, 2 and 3) independently. Then, the selected diagnostic clues were submitted to an approval process (Doctor 4) in order to discuss differences, being constantly checked to confirm the findings. Subsequently, diagnostic clues were grouped according to Basic Human Needs (BNH). (12) The choice for this theoretical framework was based on the fact that the BNH Theory is the most widespread among Brazilian hospital institutions, especially in the context of critical care. (13) In the second stage, the survey of NANDA-I ® ND was carried out, following the precepts of diagnostic reasoning of the Risner Model. (14) For the analysis stage, clinically relevant data extracted from the literature were categorized according to the BNH, which made it possible to identify gaps in information not covered in the bibliographic survey. For the synthesis process, the relevant diagnostic clues were grouped to compose the judgment of a diagnostic hypothesis by comparing the clues with normality standards. (14) Based on the taxonomic structure of NANDA-I, ND related to diagnostic hypotheses were sought. The definition of the diagnostic title and the correspondence of diagnostic clues with NANDA-I diagnostic indicators (related factor/risk factor, defining characteristics, associated conditions and population at risk) supported the selection of NDs using standardized language. It is noteworthy that this step was conducted by two researchers (Doctors 1 and 2) in a Microsoft Excel ® spreadsheet and later discussed with two other researchers (Doctors 3 and 4) to identify differences and propose a version only.
In the third stage, correspondence validation between diagnostic clues and NANDA-I ® ND was carried out by five expert nurses, independently. (15) There is divergence in the literature regarding the adequate number of experts to be included in the validation stage, with recommendations between five and ten participants being highlighted. (15,16) It is noteworthy that, in addition to the quantitative selection of experts, it is important to consider their qualitative selection with regard to training, qualification, availability and expertise in the subject in question. (16) For the composition of an expert committee, selection criteria were considered to ensure participants' knowledge of the subject in question, in order to certify reliability of results. Thus, five nurses were selected (located through the Plataforma Lattes -http://lattes.cnpq.br/) who met the following inclusion criteria: professional experience of at least five years in teaching or care and scientific production with the thematic nursing in critical care and nursing classification systems.
The following were sent to the expert nurses via e-mail: a professional profile characterization form; instrument referring to the survey of ND from the diagnostic clues described in the literature; invitation letter with detailed instructions related to filling out the instrument; file in PDF format with NANDA-I ® NDs; and the informed consent form.
The document sent to the experts for validation was structured by the authors in columnar format. In the first column, BNHs were described; in the second, the diagnostic clues extracted from the review articles that made up each BNH; in the next column, the diagnostic titles as well as the diagnostic indicators (defining characteristics, related and risk factors, associated conditions or population at risk) of NANDA-I ® . Thus, the diagnostic correspondence validation was based on the equivalence between the groups of diagnostic clues with NANDA-I ® title and diagnostic indicators.
Expert nurses were instructed to make their notes and suggestions in a space beside each NANDA-I ® ND. Each expert informed whether or not they agreed with the set of diagnostic clues identified for each BNH and NANDA-I ® ND. In case of disagreement, they were also asked to explain the reasons and possible suggestions for NDs.
Data analysis was performed using the Microsoft Excel ® , version 2016. The agreement index (AI) was calculated for each ND listed [AI=NC/(NC+ND)×100], where NC refers to the number of agreements, and ND, to the number of disagreements. (17) The questionnaire was circulated through the group of experts until the minimum agreement value of 80% was obtained. (15,17) As for expert characterization, a descriptive statistical analysis was performed by calculating the absolute and relative frequency. Approval was obtained by the Institutional Review Board (Opinion 4,114,490) and the recommendations of Resolution 466/2012 (CAAE (Certificado de Apresentação para Apreciação Ética -Certificate of Presentation for Ethical Consideration) 33855620.7.0000.5153) were followed.

Results
In the first stage, 289 studies were found in electronic and manual searches. Since they are duplicated, 52 were removed from the listing. After reviewing titles and abstracts, 205 articles were excluded, remaining 32 for full text analysis. Of these, one study was not found and 11 were excluded. Thus, 20 articles were included in the study (Figure 1).
As for the results of the second stage, the 51 diagnostic clues were grouped into 11 psychobiological BNHs and resulted in the identification of 23 diagnostic titles. The BNHs with the highest number of diagnostic clues were: vascular regulation (45.0%), oxygenation (16.0%) and thermal regulation (12.0%). The third stage of the study consisted of the diagnostic correspondence validation between the diagnostic clues and the NANDA-I ® ND performed by expert nurses in three rounds. Most participants were female (80.0%), four with a doctorate in nursing (80.0%) and one with a master's degree in nursing (20.0%). The other characterizations of the experts are presented in table 1.
In the first round, 52.2% (n=12) of the correspondence obtained agreement of 100.0% and 17.4% (n=4) agreement of 80.0%, being, therefore, considered valid by experts. Other seven NANDA-I ® NDs China Narration Clarify whether the SARS-CoV-2 virus can reach the central nervous system (CNS) and induce neuronal damage that leads to acute respiratory distress. (20) Nausea/vomiting; dyspnea; fever; headache; raised intracranial pressure.

USA Systematic
Describe neurological and psychological effects of COVID-19 based on a literature review. (25) Dizziness; convulsion; muscle pain; headache; raised intracranial pressure.

Italy
Systematic Synthesize the available evidence on the main neurological signs and symptoms in patients with COVID-19. (26) Dizziness; agitation; cerebral hemorrhage; convulsion; headache; raised intracranial pressure.
China Narration Report the gastrointestinal manifestations and pathological findings of patients with COVID-19 and discuss the possibility of fecal transmission of the virus. (27) Anorexia; Nausea/vomiting; diarrhea; abdominal pain/discomfort; gastrointestinal bleeding; abdominal pain/discomfort.

Germany Narration
Present the main cutaneous symptoms described in the literature for patients with COVID-19. (28) Acroischemia; cold ends; dry gangrene; skin injuries.

China Systematic
Assess the prevalence of comorbidities in patients with COVID-19 and acute respiratory syndrome. (29) Dyspnea; fatigue; fever.
Taiwan Narration Analyze the evidence on how the digestive system and liver are affected by the SARS virus-CoV-2. (31) Anorexia; nausea/vomiting; diarrhea; abdominal pain/discomfort.

China Narration
Understand the mechanism of viral sepsis caused by COVID-19. (32) Weak peripheral pulse; cold ends; acute renal failure.
Taiwan Narration Not described. (    into Psychobiological BNH, with emphasis on Vascular Regulation and Oxygenation, which consequently culminated in the survey of NDs in the domain "activity/rest" and class "cardiovascular responses /pulmonary". This result may be related to the pathogenesis of COVID-19, in which most severe cases admitted to ICU present comorbidities such as systemic arterial hypertension and respiratory failure. (1) This fact is confirmed in another study, which also shows a predominance of clinical manifestations aimed at cardiac and pulmonary responses. (37) There is a lack in the literature on the relationship between cardiovascular alterations and symptoms of COVID-19. However, studies show a similarity in the pathophysiological mechanism of cardiac injury caused by SARS-CoV-2 with that of pulmonary involvement. It is believed to be related to the virus's affinity for the angiotensin-converting enzyme (ACE) II, (38) highly expressed in the lungs and heart, allowing infection of these organs and dissemination of the virus. (39) Injury to cardiac tissue can also be associated with an exacerbated inflammatory response, which leads to high levels of cytokines and hypoxia, resulting from pulmonary impairment or ischemic injury due to vascular alterations. (40) Therefore, these alterations may favor the presence of diagnostic clues such as weak peripheral pulse, cold extremities, hypotension, hypoxemia, systemic inflammatory response syndrome, acute coagulopathy, ventricular/atrial fibrillation, among others.
Other NDs in the activity/rest domain were specifically associated with oxygenation, such as "Impaired spontaneous ventilation", "Ineffective breathing pattern" and "Dysfunctional ventilatory weaning response". It is known that the main target of the coronavirus pathogen is the respiratory system. It is believed that primary viral replication occurs in the mucosal epithelium of the upper respiratory tract, with greater multiplication in the lower respiratory tract and gastrointestinal mucosa, giving rise to mild viremia. It can also cause acute lung injury with consequent "Impaired gas exchange", aggravating acute respiratory distress syndrome (ARDS) and pulmonary failure. (20) Therefore, diag-(30.4%) showed agreement below 80.0%. For these, experts suggested including diagnostic clues and diagnostic titles that had not been previously listed, therefore, a second round of validation was needed.
The result of the second round showed that 73.9% (n=17) of the ND had 100% agreement, and 26.1% (n=6) had 80% agreement, being therefore considered valid by experts. Also in the second round, the inclusion of three new NDs were suggested.
The result of the third round of validation indicated the inclusion of "Impaired tissue integrity", "Risk for pressure injury" and "Risk for unstable blood glucose", all with 100.0% of agreement among experts. Thus, at the end of the third round, the 51 diagnostic clues resulted in the identification of 26 NANDA-I ® NDs.
The NANDA-I ® domains with the highest number of NDs were activity/rest (n=9); safety/protection (n=7) and nutrition (n=4). It is noteworthy that 23 diagnostic clues (45.1%) corresponded with diagnostic indicators (related and risk factors, defining characteristics, associated conditions and population at risk) present in more than one NANDA-I ® ND. All diagnostic clues corresponded to at least one NANDA-I ® ND. Furthermore, most ND (60.0%) refer to real problems and 40.0% to potential problems. Chart 2 presents the final result of the diagnostic correspondence validation between diagnostic clues and NANDA-I ® ND.

Discussion
From the analysis of the diagnostic clues identified in the studies, it was found that all were grouped nostic clues such as dyspnea; tachypnea; tachycardia; agitation; hypoxemia; fatigue and confusion are identified in critically ill patients with COVID-19.
As for "Impaired physical mobility", whose diagnostic clues were the manifestations of ataxia and muscle atrophy, (19,21,22) it is known that given the clinical profile of critically ill patients, they tend to remain restricted to bed. This is a limitation imposed by the pathology itself and the treatments needed to recover from pulmonary impairment, in addition to the use of sedatives and vasoactive drugs. (18) The identification of this ND is essential, as it is associated with planning interventions capable of preventing other possible NDs resulting from motor impairment, such as "Impaired skin integrity" and "Impaired tissue integrity".
Regarding the NDs in the safety/protection domain, such as "Risk for bleeding" and "Risk for venous thromboembolism", the literature established acute coagulopathy as the main diagnostic clue. (18) Coagulation disorders in patients with COVID-19 are associated with increased levels of D-dimer and fibrinogen, in addition to lymphopenia and thrombocytopenia. It is suggested that endothelial dysfunction also plays an important role that contributes to increased thrombin and blocking fibrinolysis, which leads to hyper coagulopathy. (41) Therefore, it is essential that nurses monitor hemoglobin/hematocrit levels and coagulation tests, including prothrombin time, partial thromboplastin, fibrinogen, among other factors. (41) Gastrointestinal manifestations can affect between 3% and 79% of patients with COVID-19, being more common in severe cases. (42,43) Among the possible ND, we can mention "Diarrhea" and "Nausea". It is noteworthy that, in many cases, in the absence of respiratory symptoms, diarrhea can be the first symptom before the disease is diagnosed. (28) The mechanisms involved in the development of gastrointestinal symptoms are still unknown, but the cause is probably the epithelial damage caused by the virus. (28) As for "Imbalanced nutrition: less than body requirements", the main diagnostic clues described in the literature were ageusia/dysgeusia and anosmia/hyposmia. (19,21,22) It is known that the loss or decrease in smell and taste are characteristic chemosensory changes in COVID-19, even in the absence of nasal congestion. Although the mechanism associated with these symptoms is not completely elucidated, in relation to taste, the virus can bind to sialic acid receptors, increasing the degradation of glycoproteins that transport taste molecules. Impaired olfactory perception, in turn, is associated with direct damage from early neural death induced by exacerbated release of inflammatory factors. (44) The identification of "Acute pain", especially when dealing with critically ill patients, is important to attenuate subsequent physical and psychological complications. Specifically in the context of ICU, patients unable to communicate, undergoing sedation, invasive mechanical ventilation or with acute confusion are at increased risk for untreated pain. Physiological parameters, such as heart and respiratory rate and blood pressure, could be used to assess pain; however, they are nonspecific elements and are highly vulnerable to the effects of drugs. (45) Thus, it is reinforced that, although the diagnostic clues identified in this study are based on self-report of pain, it is important that nurses have other methods of pain assessment, including objective indicators that can be verified without verbal communication. (45) It is noteworthy that the studies included in this review did not find diagnostic clues associated with psychosocial and spiritual aspects. However, it should not be overlooked that patients with COVID-19 admitted to ICU are susceptible to problems related to communication, social isolation, anxiety, fear and spiritual suffering, and, therefore, they require the identification of NDs. (46) Among the limitations presented in this study, the inclusion of some medical diagnoses, such as pulmonary edema, cardiac insufficiency, acute renal failure, systemic inflammatory response syndrome, as diagnostic clues. The choice to maintain these medical diagnoses for the survey of ND was based on the fact that they were described by the studies included in the review as important clinical manifestations in critically ill patients with COVID-19. Furthermore, it is noteworthy that these are associated conditions described by NANDA-I ® .
Another limitation of this study is based on the survey of diagnostic clues based on secondary data from review studies. The choice for such a strategy is based on the fact that it is a newly discovered infectious disease, whose clinical aspects are being elucidated. However, the diagnostic clues listed in the studies allowed to reflect the reality experienced in clinical practice in different countries, which minimizes this limitation.

Conclusion
This study allowed us to identify diagnostic clues present in critically ill adult patients with COVID-19 and to verify their equivalence with 26 NANDA-I ® diagnostic titles. In clinical practice, the identified ND may support the construction of instruments for collecting nursing data for patients with COVID-19 hospitalized in ICU, in addition to favoring the cre-ation of software to support the nursing process recording. Studies on classifications with standardized languages in Brazil, including NANDA-I ® , still focus primarily on the development of terminology; however, it is necessary to advance, including its documentation in electronic records so that the data can be analyzed, evidencing nursing care in practice.