Noncompliance in people living with HIV: accuracy of defining characteristics of the nursing diagnosis

ABSTRACT Objective: to evaluate the accuracy of the defining characteristics of the NANDA International nursing diagnosis, noncompliance, in people with HIV. Method: study of diagnostic accuracy, performed in two stages. In the first stage, 113 people with HIV from a hospital of infectious diseases in the Northeast of Brazil were assessed for identification of clinical indicators of noncompliance. In the second, the defining characteristics were evaluated by six specialist nurses, analyzing the presence or absence of the diagnosis. For accuracy of the clinical indicators, the specificity, sensitivity, predictive values and likelihood ratios were measured. Results: the presence of the noncompliance diagnosis was shown in 69% (n=78) of people with HIV. The most sensitive indicator was, missing of appointments (OR: 28.93, 95% CI: 1.112-2.126, p = 0.002). On the other hand, nonadherence behavior (OR: 15.00, 95% CI: 1.829-3.981, p = 0.001) and failure to meet outcomes (OR: 13.41; 95% CI: 1.272-2.508; P = 0.003) achieved higher specificity. Conclusion: the most accurate defining characteristics were nonadherence behavior, missing of appointments, and failure to meet outcomes. Thus, in the presence of these, the nurse can identify, with greater security, the diagnosis studied.

is defined as the evaluator's judgment as to the degree of relevance, specificity and consistency of the clues (clinical indicators) available for diagnosis (6) .
For analysis of the accuracy of each defining characteristic, mathematical calculations are performed, with emphasis on the following dimensions: sensitivity, specificity, positive and negative predictive values.
Through diagnostic accuracy measures, it is possible to differentiate individuals with and without the ND, from the defining characteristics. However, it is well known that the stimulation of the development of research in this field reduces the nurses' subjective uncertainties and leads to more simplified diagnosis (6) .
One study of accuracy of the diagnosis, impaired gas exchange (code 00030) in children with acute respiratory infection, showed that conducting research demonstrating the validity of the relationships between the clinical manifestations and the ND corroborates the reduction of inaccuracy in choosing the diagnosis, by discriminating the most appropriate ND to situations/ problems presented by the patient and, consequently, achieving positive health outcomes (7) .
This study intended to corroborate the accuracy of clinical evidence for such diagnoses, as well as to contribute to the process of diagnostic inference, increasing the nurse's diagnostic ability, reducing clinical uncertainties, so that actions and nursing interventions can be initiated with scientific support. Therefore, the objective of this study was to evaluate the accuracy of the defining characteristics of the NANDA-I diagnosis, noncompliance, in people with HIV.

Method
This was a diagnostic accuracy study, with a crosssectional design, performed in two stages. In the first, the identification and evaluation of the NANDA-I clinical indicators of lack of adherence to antiretroviral treatment in people living with HIV were performed. In the second stage, the accuracy of the defining characteristics of the respective diagnosis was developed. The information in question is detailed in Figure 1, below.
In the first step, the population of people with HIV was from the outpatient HIV clinic of an intermunicipal referral hospital for the treatment of infectious diseases in the Northeast of Brazil Specifically in the outpatient clinic where patients are receiving antiretroviral treatment, 306 patients are registered and followed. Of this total, 158 patients did not get medication at the pharmacy and did not attend the appointment in 2014, corresponding to 51.63%. In this sense, this population was considered as the universe for the sample calculation. Thus, the formula for finite populations was used, and the criteria adopted were a 95% confidence level (Z∞ = 1.96) and a sample error of 5%, resulting in a sample size of 113 patients (8) .
It is noteworthy that, in the service, there is a record of patients in the outpatient clinic who, although they attend their scheduled appointments, have low adherence to medications.
Recruitment was by convenience in a consecutive manner, adopting the following inclusion criteria: clinically diagnosed with AIDS, age above 18 years, included in the hospital unit system as a low-adherence patient, using antiretroviral therapy for at least six months, and registered at the hospital outpatient clinic at the time of data collection. The exclusion criterion was presenting with altered physical or emotional conditions during the period of data collection that could interfere with such collection.
The data collection occurred from July 15 to September 30, 2015, by means of a physical examination and patient history guide, prepared according an integrative review conducted in national and international journals (4)(5)(6)(7) , and complimentary books (9) as well as the manual of the Ministry of Health (MS) (10) , subdivided into two parts.   (11) .
After development of the instrument, the content was submitted to a process of normalization and validation by eight nurses, professors and specialists in Systematization of Nursing Care and immunology.
The suggestions were included and it was reassessed again. After that time, 10% of the sample was used for a pretest, so that possible gaps could be identified, with no need for modifications identified. The data collection was performed by two-master's students and one doctoral student, who participated in a ten-hour course, divided assigning a value to the defining characteristic of the ND; Delphi technique (optional); weighted mean of the assigned scores; and, finally, inferential analysis of the data (6) .
The author recommends that expert nurses perform this type of study. However, the difficulties in finding a sample of professionals that meet the criteria proposed by the author are known (6) , and the author recognizes this fact. Thus, the criteria were modified for this study, and described as follows: (1 point After searching for specialists, six were selected, in the area of ND, with clinical practice or experience in the teaching of infectious diseases in nursing. Then, theoretical and practical training was given to the specialists, with the application of realistic situations, so that their responses were measured based on capacity, efficiency, and trend, false positive and negative rates, as described in Table 1.

The practice of training was fundamental because
there is no gold standard for the identification of NDs. In this study, the characteristics observed in patients who showed adherence for taking the medication (no delay in receiving the drug, CD4 lymphocyte count greater than 200 cells/mm 3 , undetectable viral load, absence of opportunistic diseases, and not reporting difficulties in taking medications) and attending appointments.
For the analysis of diagnostic inference ability, an instrument developed by the researchers was used, consisting of 12 items that related to each item to be evaluated, resulting in a mean that was compared with the acceptable, marginal and unacceptable levels. After analyzing the results, only three specialists were able to obtain the acceptable mean (> 0.9) for each item of diagnostic ability assessment. After selecting the specialists, a spreadsheet was representativeness (12)(13) . Thus, the positive and negative likelihood ratio, diagnostic odds ratio (DOR) and prevalence were calculated. DCs that presented statistical significance were tested by means of logistic regression, to verify the conjugated association of the DCs, as revealed in Table 4.

Discussion
The clinical validation process of the NDs, performed by obtaining accuracy in the DCs, is a practice that corroborates the process of differentiation of the presence and absence of diagnoses, as well as attenuating the biases of the clinical inferences of the nurse, as the interpretation of the clinical evidence is subjective (6)(7) .
In the study, the noncompliance diagnosis is defined as the "behavior of the person and/or caregiver that fails to coincide with a health-promoting or therapeutic plan agreed on by the person (and/or family, and/or community) and health care professional. In the presence of an agreed--upon, health-promoting or therapeutic plan, the person's or caregiver behavior is fully or partly nonadherent and may lead to clinically ineffective or partially effective outcomes" (9) . This diagnosis is presented with the DCs, development-related complication, nonadherence behavior, exacerbation of symptoms, failure to meet outcomes and missing of appointments (9) .
Among these, about 67.3% of the patients with a noncompliance diagnosis had the DC, missing of appointments; that is, they did not attend scheduled appointments, did not take doses of indicated medications, did not take medication at the correct time and did not attend the health service to receive the medication (14) .
In this sense, the non-attendance at the care site may be aimed at not accepting the new health condition and a disbelief in the medication regimen (14) .
The complexity of treatment requires the person with HIV to have previous guidelines, because some medications need to be taken with food, others while fasting, or in temporal sequences combined with other medications, which requires the patient's organization and commitment to treatment (15) .
Studies demonstrate that the adherence to antiretroviral therapy can also be influenced by the reduction of barriers to accessing health services; use of health technologies, such as welcoming, bonding, accountability, and trust in the service; and, improvement of the interpersonal relationships with health professionals (1)(2)(3)(4) .
The role of the professional who provides care is important, especially his/her ability for dialogue and negotiation. Guidance on the disease, the importance of adherence, medications and the correct way of using them, the side effects of the treatment, the actions to be taken, and when these effects occur, are a priority in the treatment and should involve all the professionals in the health care services (2)(3)(4) .
In the present study, in addition to the respective factors of abstention, the missing of appointments was also related to forgetfulness. In a study showing that forgetting was the main cause for missing appointments, health professionals implemented telephone followup, considerably reducing the number of missed appointments (15) . Thus, nurses can implement active searches and therapeutic activities, such as conversation wheels and community therapies (2,(16)(17) .
The second DC, development-related complication, was identified in the noncompliance diagnosis, which had as indicators the signs and symptoms presented by the evolution of the disease and its complication, most often due to failure to adhere to treatment (9)(10)(11)(12)(13)(14)(15)(16)(17) .
Researchers point out that one of the motives claimed by people with HIV for non-adherence is that they are not aware of the importance of antiretroviral medications, suppression of viral load, increasing the CD4 and CD8 (18) , and attenuation of the developmentrelated complications and opportunistic infections, which are linked to abandonment or irregularity in the continuity of treatment (18) . Therefore, the nurse can implement actions that corroborate the demystification of medications regarding their use and side effects (2,(16)(17) .
In order to help patients to improve their adherence to antiretroviral treatment, many strategies It was pointed out in a study that showed that patients with a CD4 count lower than 200 cells/mm 3 presented a higher risk of noncompliance (25) .
It is important to highlight that the analysis of these clinical indicators, in an isolated way, seeks to better understand the specificities that interfere in the formulation of the diagnosis, as well as the nurse's knowledge (3)(4)(5)(6) , such as the fact that antiretroviral therapy (ART) delays the development of the disease and reduces the exacerbation of symptoms, by means of viral suppression and immune system restoration. Thus, it is important to discuss and explain the results of the laboratory tests to the patient during the appointments, so that he understands and visualizes his progress, recognizing the advantages of the treatment, thus strengthening the motivation for adherence (16)(17)(18)(19)(20)(21)(22)(23)(24)(25) .
Given this, it is understood that noncompliance is, in turn, one of the main problems of people with HIV, as it is a chronic disease, which, by itself, represents one more difficulty for following the recommendations. Finally, it is important to note that, despite the relevance of accuracy studies, these are recent and remain scarce in the nursing area (6) . In addition, the lack of studies that also analyze accuracy in NDs related greater chances of targeting and efficacy (12) .
One study investigating the accuracy of the DCs of the diagnosis of ineffective family therapeutic health management (code 00080) in diabetic patients receiving care at a family health center, showed the need to work on adherence from a family perspective, especially considering the complexity of treatment for people with chronic disease (13) .
As a limitation of the research, we highlight the difficulty in characterizing the conceptual and operational definitions of the DCs related to the diagnosis studied, which were not clearly defined in the literature: an obstacle that may impair the reliability of the data collected, regarding the measurement. In this way, it is also proposed to investigate, in future research, the validation of the conceptual and operational definitions of the DCs of the noncompliance diagnosis.

Conclusion
The present data demonstrate that the DC with higher sensitivity was missing of appointments due Silva RAR, Costa MMN, Souza Neto VL, Silva BCO, Costa CS, Andrade IFC. to the noncompliance diagnosis. As for the higher specificity, nonadherence behavior and failure to meet outcomes were highlighted. In addition, the indicators, nonadherence behavior, missing of appointments and failure to meet outcomes were identified in the logistic regression model as the most accurate DCs for identifying the diagnosis in people with HIV.
The DCs studied will help the nurse in inferring the nursing diagnosis of noncompliance in people with HIV, with greater accuracy and safety, and thus enable them to establish nursing outcomes and interventions with greater chances of targeting and effectiveness.