Analysis of ineffective breathing pattern and impaired spontaneous ventilation of adults with oxygen therapy 1

ABSTRACT Objective: to analyze the manifestation of the defining characteristics of the nursing diagnoses of ineffective breathing pattern and impaired spontaneous ventilation, of the NANDA International and the defining characteristics identified in the literature for the concept of “ventilation” in adult patients hospitalized in an intensive care unit with use of oxygen therapy. Method: clinical diagnostic validation study, conducted with 626 patients in intensive care using oxygen therapy, in three different modalities. Multiple correspondence analysis was used to verify the discriminative capacity of the defining characteristics and latent class analysis to determine the diagnostic accuracy of them, based on the severity level defined by the ventilatory mode used. Results: in the multiple correspondence analysis, it was demonstrated that the majority of the defining characteristics presented low discriminative capacity and low percentage of explained variance for the two dimensions (diagnoses). Latent class models, separately adjusted for the two diagnoses, presented a worse fit, with sharing of some defining characteristics. Models adjusted by level of severity (ventilation mode) presented better fit and structure of the component defining characteristics. Conclusion: clinical evidence obtained in the present study seems to demonstrate that the set of defining characteristics of the two nursing diagnoses studied fit better in a single construct.


Introduction
The NANDA International Taxonomy II (NANDA-I) is the globally accepted classification of Nursing Diagnoses (NDs), composed of 13 domains, 47 classes and 234 diagnoses. The fourth domain, called "Activity/rest", is defined as the "production, conservation, expenditure or balance of energy resources" and contains the class "Cardiovascular/pulmonary responses", which is defined as "cardiopulmonary mechanisms that support activity/ rest". This class includes the Ineffective Breathing Pattern (IBP) and Impaired Spontaneous Ventilation (ISV) NDs (1) .
The IBP ND was introduced into the classification in 1980 and was revised three times in the years 1996, 1998 and 2010. It is defined as "inspiration and/or expiration that does not provide adequate ventilation" and has 16 Defining Characteristics (DCs). The ISV ND, in turn, was introduced into the NANDA-I in 1992 and has never been revised. It is defined as "decreased energy reserves resulting in an inability to maintain independent breathing that is adequate to support life" and has 11 DCs. The difficulties in accurately establishing each of these NDs come from both their closely related definitions and their similar or shared DCs.
In the clinical practice, the ISV ND is often attributed to patients admitted to an Intensive Care Unit (ICU) who are dependent on invasive mechanical ventilation (IMV) (2) . However, this ND is mainly used due to its title, because, when the DCs and related factors are analyzed, both the IBP and the ISV ND could be attributed to the patient on IMV. It is also appropriate to examine the definitions of each of the NDs. The IBP ND presents the words "adequate ventilation" in its definition, while the ISV ND presents the words "adequate breathing". Therefore, key concepts of IBP ND title are present in the definition of the ISV ND and vice versa. In the analysis of the central concept of IBP and ISV, both probably refer to the concept of "ventilation" and not to the concept of "breathing" present in the IBP title. The ND available for patients with problems in the processes of respiration/exchange of gases in cell membranes is Impaired Gas Exchange (IGE), defined as "excess or deficit in oxygenation and/ or carbon dioxide elimination at the alveolar-capillary membrane" (1,3) Therefore it is possible that IBP and ISV respond for the same diagnostic concept, being in fact different levels of severity of the same ND. Thus, the aim of this study was to analyze how the DCs of the IBP and ISV NDs and the DCs found in the literature for the key concept "ventilation" are manifested in adult patients hospitalized in ICU using oxygen therapy.

Method
This was a cross-sectional, methodological study, for clinical validation between the IBP and ISV NDs.
Clinical validation was used to test whether the clinical data (defining characteristics) supported the existence of two distinct NDs, IBP and ISV, or only one ND with the key concept "ventilation", consisting of different levels of severity. The variable "ventilatory mode" was used to determine the severity levels of the assumed ND with "ventilation" as the key concept. Thus, patients undergoing oxygen therapy using face masks or nasal cannulars, that is, Spontaneous Ventilation (SV), were considered as the lowest severity level, patients since there is no uniform definition of prolonged intubation in the literature, ranging from 7 and 21 days (4)(5) . The exclusion criteria established were: 1) to have neurological and/or muscular disease that may cause a change in the clinical presentation of the DCs of the NDs under study and 2) to be using neuromuscular blockers and/or moderate to profound sedation according to the Richmond Sedation and Agitation Scale (RASS) (6) . Subjects were included in the study by consecutive sampling. In studies using the Latent Class Analysis (LCA) technique, the minimum sample size presented is at least 20 observations (individuals) for each item (DC) to be analyzed (7) . Thus, the minimum sample size for 25 DCs equates to 500 patients, with the sample in this study consisting of 626 individuals.
The data collection instrument was developed containing the characterization of the patient, ventilatory data and a table with four columns. The first column contained the DCs found in the analysis of the "ventilation" concept previously performed (8) , as well as the NANDA-I DCs for the IBP and ISV NDs (1) . The second column contained the respective operational definitions of each DC, constructed based on the clinical experience of the researcher, of the specialists and in the literature.
The last two columns presented the options "yes" and "no", regarding the presence of that DC in the physical www.eerp.usp.br/rlae 3 Seganfredo DH, Beltrão BA, Silva VM, Lopes MVO, Castro SMJ, Almeida MA. examination of the patient. For example, the "decrease in inspiratory pressure" DC was operationally defined as inspiratory pressure greater than -90 cmH2O, with the "decrease in expiratory pressure" DC defined as expiratory pressure lower than +100 cmH2O, measured by means of a manometer (9) . However, in the pilot test, it was evidenced that this examination depended on the cooperation and level of consciousness of the patient, which was impaired in the critically ill patients. Thus, 100% consensus was reached among the specialists in assigning these DCs as present in patients requiring positive pressure ventilatory support (IMV and NIMV) and not present in SV patients. Also the "alterations in tidal volume" DC was evaluated when changes were noticed in the depth of the ventilatory movement in SV patients, that is, the increase or decrease in its amplitude. In patients with IMV or NIMV, current volumes greater than or less than 4-8 ml/kg (9) were considered.
Data collection was carried out, where the study was performed, by the researcher and eight ICU care for the IBP and ISV NDs, with the intention of performing the differential diagnosis (7) . The LCA with random effects was used to calculate the sensitivity and specificity of the DCs of the ND in question (7) . For defining the DCs to be included in these models, sensitivity and/or specificity values were defined with higher confidence intervals, the lower band of which was above 0.5 (50%). The significance level of 5% was adopted for all analyses.  proportions (see 95% CI in Table 1). of the two dimensions, only the "orthopnea", "altered blood gases", and "altered ventilation/perfusion ratio"

Results
DCs presented consistently higher measures for the same dimension. Another four DCs ("hypoxia", "use of accessory muscles to breathe", "increased restlessness" and "decrease in cooperation") showed measures of discrimination with alternation of high values between the two dimensions, in the analysis of the three levels of ventilatory support. In addition, three DCs ("increased metabolic rate", "increase in heart rate" and "alterations in respiratory rate") presented high values for one dimension, in the patient subsamples, and high values for the other, when considering the total sample. Finally, the percentage of explained variance was less than 30%, with the internal consistency values being lower than 0.7 for the second dimension, in all samples. Table 4 presents the measures of diagnostic accuracy obtained by LCA, from the DCs of the IBP and ISV NDs.
The models adjusted separately for IBP and ISV, and including only the DCs described for each ND, together with the DCs identified in the review, did not present a good fit, bordering the level of significance adopted. More DCs were included in the IBP model (13) compared to the DCs included in the ISV model (5).

Discussion
Considering the aims of this study, namely, to analyze how the DCs of the IBP and ISV NDs are manifested, as well as those contained in the literature for the key concept "ventilation" in adult patients hospitalized in an ICU, using oxygen therapy, it was found that "decreased inspiratory pressure" and "decreased expiratory pressure" presented the highest frequency, in both (100%) the IMV and NIMV subsamples of patients. This fact is attributed to the need for the patients of both of the severity subsamples to receive oxygen therapy with positive pressure in the airways.
However, in the NIMV therapy, the patients need to use their muscles to ventilate, this being a facilitator of the respiratory work and not a therapy to substitute the ventilatory musculature (10)(11) . Thus, this may explain why the "decrease in inspiratory pressure", "decrease in expiratory pressure", "increase in heart rate", "alterations in respiratory rate", "increased metabolic rate", "use of accessory muscles to breathe", "altered arterial blood gases" and "alterations in tidal volume" provide the necessary support for oxygenation, while the body recovers from a serious injury (12) .
Comparing the DCs that obtained better measures of diagnostic accuracy in the different severity subscales, it is evident that there are DCs common to more than one subsample. The DCs "altered arterial blood gases", "alterations in respiratory rate", "use of three-point position" and "decrease in cooperation" obtained high sensitivity values for the NIMV and IMV patient subsamples and may be associated with later states of injury to the respiratory system, in which the mechanisms of compensation of the organism are not sufficient to compensate for the imbalance of blood gases and, consequently, of the pH (9,13) . Thus, in these subsamples, altered states of consciousness can be evidenced due to the decrease in PO2 and consequent hypoxia, leading patients to present "decrease in cooperation". However, individuals who do not require positive pressure ventilation therapies usually present better clinical conditions and, respectively, lower levels of severity, since their ventilatory muscles are able to overcome the pressure demand necessary for inspiratory and expiratory ventilation movements. In these patients, the "alterations in respiratory rate", "dyspnea" and "orthopnea" DCs may be the first clinical indications that there is progressing ventilatory dysfunction (3) . In the MCA, it was shown that the division of the DCs into two dimensions, that is, into two nursing diagnoses, is relatively inconsistent, with a low percentage of explained variance. Thus, the data indicates that to consider that these DCs represent the IBP and ISV NDs is inadequate, both for the total sample and for the subsamples related to the types of ventilatory support.
To confirm these findings, latency class models On the other hand, the models for the NDs, adjusted separately, showed worse fit. Also, some of the DCs found in the final models, in particular for ISV, did not compose the ND described in the NANDA-I and were included in the models because they were identified in the review of the "ventilation" concept. Therefore, the data obtained from the LCA corroborate what was found in the MCA, that is, there is evidence that the set of DCs studied contemplates a single ND, with three clinical spectra associated with the type of ventilatory support. To ratify this finding, in the latent class model adjusted separately for the two NDs, and including the DCs described for each ND along with the DCs identified in the concept analysis, the DCs that appeared in the ISV were "cyanosis of the skin, lips or extremities", "apprehensiveness", "altered arterial blood gases", "hypoxia" and "decrease in SaO2". However, it is highlighted in the literature that these DCs may indicate the presence of the IGE ND (3) . These DCs demonstrate impairment of pulmonary gas exchange function, not pulmonary ventilation processes. With this, they are observed later, when the mechanisms of compensation of the respiratory system are exhausted. Hypoxia may be preceded by signs of physiological compensation for respiratory stress, including "use of accessory muscles to breathe" and "alterations in respiratory rate". After this stage of compensation, the DCs "apprehensiveness", "altered arterial blood gases", "hypoxia" and "decrease in SaO2" can be evidenced, characterizing the TGP ND (3,13) .
The DCs that presented higher sensitivity values for the SV subsample, in the latent class model with better fit were "decrease in inspiratory pressure", "decrease in expiratory pressure", "fatigue", and "decrease in SaO2". In one study, among the main DCs evidenced to predict the IBP ND in children with acute respiratory infection, there were "use of accessory muscles to breathe" and "dyspnea", in which high sensitivity (88.84% and 86.78%) and specificity (99.53% and 86.18%) values were observed (14) . In the present study, cooperation", "increase in restlessness", "alterations in respiratory rate", "altered arterial blood gases" and "hypoxia". In another study, "abnormal arterial gases" were present in 82.8% of adult patients on IMV with the IBP ND, "abnormal respiratory rate" in 77.6% and "hypoxemia" in 62.1%, demonstrating agreement with the findings of the present study, insofar as they are important DCs for patients with ventilatory dysfunction supported with IMV (15) . Measures of diagnostic accuracy were not calculated, and ISV was not studied because the author understood that this ND was not related to mechanically ventilated patients.
The "tachycardia" CD (44.8%) presented a frequency similar to that reported in the present study ("increase in heart rate"), both for the total sample of patients (46.6%) and for the subsample of patients on IMV (48.4%). The DCs "decrease in inspiratory pressure" and "decrease in expiratory pressure" were the most frequent in the present study for the subsample of patients on IMV (100%), however, in the study cited only "decrease in inspiratory pressure" (24.1%%) was presented by the patients, occupying the fifth position in relation to the IBP ND. On the other hand, the "decreased expiratory pressure" CD was not among the most frequent (15) .
As a limitation in this study, it should be highlighted that all patients included in the sample already had oxygen therapy instituted, in addition to the therapeutic support available in the tertiary level ICU, in which the study was performed. This fact may have mitigated the presentation of the DCs studied or even suppressed them, due to the possible compensation of the cause of the ventilatory impairment.

Conclusion
Clinical evidence, obtained in the present study, seems to demonstrate that the set of DCs of the two NDs studied fit best in a single construct. One of the possibilities would be the incorporation of ISV DCs into the IBP ND, which constituted part of the latent class model with better fit, i.e. "increase in restlessness", "changes in tidal volume", "decreased SaO2", "decreased cooperation", "Altered arterial blood gases", "apprehensiveness" and "hypoxia". Other DCs,