951 DEFINING CHARACTERISTICS , VALIDATED BY SPECIALISTS AND MANIFESTED BY PATIENTS-A STUDY OF THE SEXUAL DYSFUNCTION AND INEFFECTIVE SEXUALITY PATTERN DIAGNOSES

This observational and descriptive study was carried out to validate the nursing diagnoses Sexual Dysfunction and Ineffective Sexuality Pattern and relate the relevance of the defining characteristics validated by experts with their incidence in patients. Validation by experts involved 32 specialists and 20 patients to verify clinical evidence. For the diagnosis Sexual Dysfunction, the experts attributed scores higher than 0.80 to seven defining characteristics; for Ineffective Sexuality Pattern, four characteristics received scores between 0.50 and 0.80. The clinical occurrence of these characteristics for the Sexual Dysfunction diagnosis ranged between 55% and 90% of the sample and, for Ineffective Sexuality Pattern, between 30% and 85% of the sample. The study contributed to the improvement of these diagnoses and to careful consideration on their defining characteristics according to experts and as observed in a given clientele.

This observational and descriptive study was carried out to validate the nursing diagnoses Sexual Dysfunction and Ineffective Sexuality Pattern and relate the relevance of the defining characteristics validated by experts with their incidence in patients.Validation by experts involved 32 specialists and 20 patients to verify clinical evidence.For the diagnosis Sexual Dysfunction, the experts attributed scores higher than 0.80 to seven defining characteristics; for Ineffective Sexuality Pattern, four characteristics received scores between 0.50 and 0.80.The clinical occurrence of these characteristics for the Sexual Dysfunction diagnosis ranged between 55% and 90% of the sample and, for Ineffective Sexuality Pattern, between 30% and 85% of the sample.The study contributed to the improvement of these diagnoses and to careful consideration on their defining characteristics according to experts and as observed in a given clientele.

INTRODUCTION
Nursing diagnoses are an essential tool for nursing practice.For them to be considered valid, however, they need to correspond to the true state of phenomena being measured.If they present the same results at different moments, they will be considered reliable and, if they are equally identified in several similar observations, they will be considered stable.
Under these conditions, nursing diagnoses can contribute to the development of knowledge in clinical practice (1)(2) .
Different validation models are used in studies on this topic.They sometimes validate a diagnosis, other times its components, or yet diagnoses prevalent in a certain clientele (3) .Among several methodological designs, one of the most comprehensive is that of Hoskins (4) , which proposes three phases: analysis of concept -to identify attributes and characteristics of a construct; validation by experts -to verify the representativeness of these characteristics in the studied concept (5-6)   ; and clinical validation -to verify whether the defining characteristics determined in the previous phase are present in a certain population.
Although the set of diagnoses employed in practice is broad, just a few have been considered in clinical validation studies, especially if we think about the different types of patients and scenarios in which these diagnoses can occur (7) .Further research is necessary to reduce uncertainties in research on this topic, like the selection of experts and types of validation adopted, which this study looks at.
Different aspects are considered when assessing professional qualification with a view to expert status (7)(8)(9)(10) , which are: number of years of clinical experience; time of bachelor's degree; experience with research; place of work; areas or sectors of work; experience and type of rationale used.Contradictions have been observed in literature regarding the relation between these variables and accuracy in establishing nursing diagnoses (7)(8)(9)(10) .
This study was motivated by the possibility of verifying the potential influence of the expert's area of work on the validation of diagnoses, as well as by the correspondence between clinical evidence appointed by specialists and the evidence presented by patients with a certain diagnosis.
It is worth highlighting that the diagnoses that describe sexuality (13) have undergone few changes in their essential elements since their elaboration in 1980 (sexual dysfunction) and 1986 (ineffective sexuality pattern) despite studies carried out in our field, including two master's theses.One of them (14) presents clinical validation of these diagnoses proposed in the taxonomy and the other presents a study of the Sexual Dysfunction diagnosis.The conclusion is that changes are needed due to disagreement with what had been proposed in the taxonomy and in the literature for this diagnosis (15)   .In addition to these studies, another research extracted from a doctoral dissertation, involving patients with onco-hematological disorders, looks at the incidence of defining characteristics of the two diagnoses mentioned above (16) .The authors also carried out a conceptual validation of these diagnoses (17)   , based on findings in the literature and the NANDA (13) 2005/2006 version.Suggestions made in that publication (17) , which were adopted by NANDA in its 2007/2008 (18) (18) , published after this report.
This study contributes to strengthen changes in the NANDA (18) taxonomy because it presents clinical

METHOD
This is a methodological study where the second and third steps of Hoskins' (4) validation approach are highlighted.The concepts and elements of two diagnoses were used according to literature, departing from the conceptual (17) evaluation carried out in the previous phase; operational definitions of each defining characteristic were elaborated.Thirtytwo professionals, selected by snowball sampling, participated in the expert validation, according to criteria proposed in the literature (19) .The experts

Sample
Referees, all with a master's and 68.8% with a doctoral degree, participated in the expert validation.
Two groups were composed: Group 1, 17 nurses with experience in NANDA Nursing Diagnoses and human sexuality, with scores varying from 7 to 12 points (maximum score: 14 points) and a predominant score of nine points; Group 2, 15 nurses with experience in Nursing diagnoses and less experience in Human Sexuality, with scores varying from 6 to 18 points and a predominant score of six points.Fehring's (19) requirements were adopted for the score.
The difference in scores between groups occurred because the adopted methodology (6,19) values the professional with scientific experience and clinical practice in the nursing diagnosis under study, although, in general, it does not disregard those with work experience in nursing diagnoses.
Aiming to identify whether the expert groups presented divergences in the evaluation of relevance of each studied defining characteristic, central tendency (median and mean rank) and variability (interquartile distance) measures were used for each diagnosis; Mann-Whitney's (20) test was employed for inter-rater analysis.
Defining characteristics with scores inferior to 0.50 were considered non-validated.The studied diagnoses were determined from scores attributed by experts to each validated defining characteristic according to literature (4,6,19) recommendations, i.e. the total score of validation (DCV total).
Twenty patients participated in the verification of clinical evidence, 11 males and nine females, single, with a predominant age between 18 and 27 years, and with diagnosed leukemia or lymphoma.Two properly trained nurse observers collected data.
There was agreement between observers in the identification of diagnostic characteristics (presence of diagnosis) in the studied clientele, in all analyses, including for patients who did not comply with the study inclusion criteria.
Regarding the identification of clinical expressions (defining characteristics) of the studied diagnoses, there was a 98% agreement between observers regarding the Sexual Dysfunction diagnosis and 97% for the Ineffective Sexuality Pattern diagnosis.The agreement ratios presented for both nursing diagnoses were considered satisfactory in studies of this nature (4) .

RESULTS
First the results of the expert validation are presented for each studied diagnosis.
Variation in the scores experts attributed to defining characteristics.
It is observed (Table 1) that there was no significant statistical difference between the two groups of experts for the defining characteristics of the Sexual Dysfunction diagnosis (p>α=0.05),although the mean ranks were different in the majority of defining characteristics.Thus, they can be considered statistically homogeneous.It is observed that the scores attributed to defining characteristics of this diagnosis were h i g h e r o r e q u a l t o 0 .8 0 i n s e v e n d e f i n i n g characteristics, five of which were already part of the Sexual Dysfunction diagnosis (13) .The other two are characteristics that emerged from literature.
Thus, we can affirm that nurse experts agree that these are clinical manifestations that characterize the Sexual Dysfunction nursing diagnosis.In fact, one of these characteristics ("perceived alteration insexual excitation") received the highest score, 0.88.
The defining characteristic that obtained the lowest score (0.50) is a characteristic that was fragmented (17)(18) in "change of interest in self" and "change of interest in others", which in the taxonomy (13) was originally "change of interest in self Scores attributed by the two groups were different for three defining characteristics.In the characteristics in which the median values differed, the first and third quartiles show that the characteristics "seeking of confirmation of desirability" and "verbalization of problem" received the highest scores by Group 1 and Group 2, respectively.For the characteristic "inability to achieve desired satisfaction", according to the mean value, the highest scores were attributed by Group 2.
Regarding the Ineffective Sexuality Pattern diagnosis, there was statistical difference in two defining characteristics: "reported difficulties, limitations or changes in sexual behaviors or activities" and "alteration in relationship with significant other" (p < than α=0.05); their mean ranks also differed.Thus, we perceive that both groups of experts are statistically homogeneous for the three defining characteristics (Table 2).It is observed that four characteristics received scores higher than 0.50 and lower than 0.80 and none presented scores higher than or equal to 0.80.These scores showed that nurse experts agree that the characteristic "reported difficulties, limitations, or changes in sexual behaviors or activities" should be kept in this diagnosis, as it is already part of the taxonomy (13) .Difficulties, limitations or changes, as well as sexual behaviors or activities were fragmented in distinct defining characteristics in NANDA's (18) new version.These data also evidence that subjects agree with the presence, in this diagnosis, of the defining characteristics "alteration in relationship with significant other" and "alterations in achieving perceived sex role", which are characteristics transferred from Sexual Dysfunction.
On the other hand, the characteristic "conflicts involving values", also transferred from the Sexual Dysfunction diagnosis, was not accepted by the experts in this diagnosis because it obtained score 0.48.Therefore, this characteristic should undergo another expert validation process.
It is observed, through the mean ranks, that members with less experience in the Sexuality area (Group 2) attributed a higher score to the defining characteristic "reported difficulties, limitations, or changes in sexual behaviors or activities".This might have happened because it is a defining characteristic that was already part of the taxonomy (13) .
However, it is observed that the highest scores for the characteristic "alteration in relationship with significant other" came from the group with more experience in the sexuality area (Group 1).This fact might be related to their deeper knowledge of subjective psychosocial aspects involved in sexuality.
The only defining characteristic in which median values were not the same was "alterations in achieving perceived sex role".In this case, although the variability of answers was equal between the groups, as shown by the quartiles, Group 1 attributed a higher score than Group 2, illustrated by the medians.
In the defining characteristic "conflicts involving values", which obtained a score lower than 0.50, despite equal medians between groups, there was different variability of answers; lower scores were attributed by Group 2.
Therefore, it is verified that the two groups of experts equally identified clinical manifestations that characterize the Ineffective Sexuality Pattern nursing diagnosis in four characteristics.
The total validation score (DCV Total) was 0.80 for the Sexual Dysfunction diagnosis and 0.70 for the Ineffective Sexuality Pattern diagnosis.Thus, there were four characteristics that received scores higher than 0.50 and lower than 0.80 and none higher than or equal to 0.80 in this diagnosis.
Considering the two studied nursing diagnoses, experts of both groups were homogeneous in their answers in 86.7% of the defining characteristics.

Relation between validation by experts and clinical validation
It is important to stress that some variables behave differently when comparing the importance attributed to a given defining characteristic in the expert validation and its frequency observed in patients for the Sexual Dysfunction diagnosis, which was verified in the studied clinical context (15%).
However, despite its low frequency, we stress that this is a relevant characteristic because lack of sexual desire can be described as "change of interest in others".Table 3 -Distribution of the 10 defining characteristics of the Sexual Dysfunction nursing diagnosis, according to scores assigned by specialists in the expert validation and prevalence of incidence in patients with oncohematological diseases, agreement ratio and Weighted interrater reliability ratios.Ribeirão Preto, 2003 The characteristics "change of interest in self" and "verbalization of problem" were considered less frequent characteristics in the validation by experts.This type of characteristic provides evidence of support to the diagnosis, though patients may not present it.Through the reliability coefficient, we observe that the first of these characteristics appeared in only 30% of the sample and the second in 45%.These data reinforce findings from the previous phase.
In this study, defining characteristics nurse experts believe to be present in patients with the Sexual Dysfunction diagnosis were observed in between 55 and 90% of the sample.Regarding agreement ratios for this diagnosis, it is observed (Table 3) that nurses making the diagnosis do not agree in only three characteristics.

For the diagnosis Ineffective Sexuality
Pattern, it is observed that (Table 4), from five observed defining characteristics, the nurses making the diagnosis presented disagreement about the incidence in three cases among.This might have occurred because these are diagnoses that present characteristics focused on psychosocial aspects involved in sexuality.Thus, they are related to information difficult to express by interviewees and also to be analyzed.
In the characteristic "conflicts involving values", it is observed that nurse experts attributed scores below 0.50 and in fact, this characteristic appeared in only 10% of the sample with a reliability coefficient of 0.10.Nonetheless, this defining characteristic should undergo a new process of investigation with another group of subjects with questions directed to sexual beliefs and superstitions.

DISCUSSION
Scholars have attempted to identify the potential influence of experts' experience on the diagnostic validation process (7)(8)(9)(10) .This influence, if present, can be perceived in the variation of scores experts attribute during validation (11)(12) .This study included two groups of experts, one of them with more expertise in the sexuality area, to participate in the validation process of the studied nursing diagnoses.
The similar medians found for defining characteristics between the two groups of nurse experts (70% for Sexual Dysfunction and 80% for Ineffective Sexuality Pattern) show that, even if the nurses do not directly work in the Human Sexuality area, they pay attention to these aspects.It is worth noting that, considering both studied diagnoses, experts from the two groups were homogeneous in their responses in 86.7% of the defining characteristics.The small difference between groups might be associated to the fact that the area of Human Sexuality is a complex subject when compared to other human basic needs and is seen as an attribution of medical competence, that is, it is an area only partially incorporated into nursing professionals' (16) practice.
The total DCV scores of 0.80 for Sexual Dysfunction and 0.70 for Ineffective Sexuality Pattern are sufficient to accept both diagnoses as validated by experts, according to the adopted model (4,6,19) .
Another aspect of interest in validation studies is related to the observation of the presence, in a certain clientele, of those defining characteristics experts appoint as relevant for a diagnosis.
In this perspective, the prevalence in patients of defining characteristics experts consider of high frequency deserves consideration.According to literature, these characteristics generally appear in the diagnosis, which did not always occur in the studied sample.On the other hand, there was large variation in the prevalence scores of defining characteristics experts considered less frequent.We highlight, for instance, the high occurrence of the defining characteristics "reported difficulties, limitations in sexual behaviors or activities", stressing that this fact might have occurred due to the marked presence of fear of acquiring an infectious disease, as a result of frequent cases of neutropenia in this type of clientele.
The results of the remaining defining characteristics can be subject to the sample size or to the nature and complexity of the studied diagnoses.
Even if some variables behaved differently in the two evaluations, considering the total scores obtained in the verification of prevalence of clinical evidence -0.68% for the Sexual Dysfunction and 0.76 for Ineffective Sexuality Pattern -it can be concluded that the diagnoses reached the scores (4)(5)(6)19) recommended in clinical validation studies.
Although satisfactory results were achieved in the two validations, it is worth to expand the sample size in future clinical validation studies, so that generalizations can be made.

CONCLUSIONS
validation and expert validation of defining characteristics related to these two diagnoses and encourages careful consideration of such validation processes, taking into account the background of participant experts.The study objectives were: carry out an expert validation of the Sexual Dysfunction and Ineffective Sexuality Pattern nursing diagnoses; evaluate the presence of variation in the scores attributed to defining characteristics of the studied diagnoses, by two groups of experts with different educational backgrounds, and relate the relevance of defining characteristics validated by experts and their incidence in patients.
the relevance of each diagnostic category for the corresponding diagnosis, based on their label and operational definition, using a five-point Likert scale.An instrument was developed to collect clinical evidence in patients.Its face and content validation were carried out containing all proposed defining characteristics.This phase was developed at a medical clinic unit -hematology sector of a hospital in the interior of São Paulo, Brazil.Besides the author, two clinical nurses with experience in the Nursing Process and Hematological Nursing participated in data collection.After data collection, the nurses independently performed analysis and synthesis of data and identified whether the nursing diagnoses under study were present or not.All ethical procedures for research in human beings were complied with for the different research phases (Process No. 0270/2002 EERP/USP).
and others".Therefore, this characteristic should undergo another validation process.The two groups attributed similar scores to the seven defining characteristics.According to the experts' median values, the defining characteristics that coincided (70%) for the two groups were: "change of interest in self, change of interest in others, perceived limitations imposed by the disease and/or therapy, actual limitations imposed by the disease and/ or therapy, alteration in achieving sexual satisfaction, perceived alteration in sexual excitation and perceived lack of sexual desire".Equal interquartile distances occurred in the groups of nurse experts in six of the seven characteristics with the same median values in both groups.
Two groups of specialists participated in the validation by experts to analyze defining characteristics of the Sexual Dysfunction and Ineffective Sexuality Pattern diagnoses.For the first diagnosis, the groups were statistically homogeneous and agreement was reached in 70% of the defining characteristics when medians were considered, and attributed scores were higher or equal to 0.80 in seven defining characteristics.For the Ineffective Sexuality Pattern, the groups were statistically homogenous in Rev Latino-am Enfermagem 2008 novembro-dezembro; 16(6):951-8 www.eerp.usp.br/rlaeDefining characteristics, validated by specialists… Melo AS, Carvalho EC, Haas VJ. three characteristics, where there was agreement in 80% of the defining characteristics when medians were considered, while the highest scores attributed to four characteristics were higher than 0.50 and lower than 0.80.The comparative analysis between validation by experts and the verification of prevalence of clinical evidence in 20 patients showed that the defining characteristics the experts believed to be present in patients with the Sexual Dysfunction diagnosis were observed in between 55 and 90% of the sample of patients with onco-hematological diseases.For the Ineffective Sexuality Pattern diagnosis, defining characteristics were observed in between 30 and 85% of the sample.The study limitations are related to the size of samples, experts and patients, as well as to the complexity of the studied diagnoses of sexuality.It is relevant to point out that there was difficulty in finding experts with clinical and scientific practice regarding the studied nursing diagnoses.In terms of patients, a larger and more diversified sample is necessary.Further studies with these two nursing diagnoses in other specific populations are essential to verify the incidence of defining characteristics found in this study, in terms of patients with cancer and other pathologies that affect sexuality.The contribution of this study to clinical practice consists in providing more precision in the identification of these two nursing diagnoses and reiterating that human sexuality should be part of the nursing care process.This study also encourages careful consideration of reliability in data collection and the use of experts in research on the area of nursing diagnosis validation.

Table 1 -
Distribution of 10 defining characteristics of the Sexual Dysfunction nursing diagnosis, according to scores, p-value, median, mean rank and interquartile distance obtained by two groups of nurse specialists in expert validation.Ribeirão Preto, 2003 Defining characteristics, validated by specialists… Melo AS, Carvalho EC, Haas VJ.

Table 2 -
Distribution of five defining characteristics of the nursing diagnosis Ineffective Sexuality Pattern, according to scores, p-value, median, mean rank and interquantile distance obtained by the two groups of nurse experts in the expert validation.Ribeirão Preto, 2003

Table 4 -
Distribution of the five defining characteristics of the nursing diagnosis Ineffective Sexuality Pattern according to scores assigned by specialists in the validation by experts and prevalence in patients with oncohematological diseases, agreement ratio and weighted interrater reliability ratios.Ribeirão Preto, 2003