Instrument Validation on Nurses’ Knowledge and Practice in Palliative Care for People with Cutaneous Malignant Tumor Wound

Objective: To present the construction and validation process of an instrument to evaluate the knowledge and practice of nurses in palliative care toward the person with cutaneous malignant tumor wound. Materials and Methods: Methodological, quantitative study with applying the Delphi technique, conducted in two stages. The first was carried out with 30 judges and the second with 17. The analysis used the Lambda 2 Guttmann coefficient, Kappa index, and Content Validity Index (CVI). Results: Of the 112 items of the original instrument, 28 were excluded, given that the percentiles of Lambda 2 Guttmann, Kappa, and CVI had indicators lower than acceptable; thus, the second version of the instrument resulted with 84 items, which presented 100 % acceptance in the Delphi 2 phase. Conclusions: From the evaluation by the judges, a version of the instrument was defined with adequate content validity and concordance indices, which could contribute to the evaluation of knowledge and practice of nurses in palliative care toward the person with cutaneous malignant tumor wound. utilizaram-se o coeficiente Lambda 2 Guttman, Kappa e Índice Validade de Conteúdo (IVC). Resultados: 112 itens do instrumento original, 28 foram excluídos, pois os percentuais de Lambda 2 Guttman, o Kappa e o IVC apresentaram escores aceitável; instrumento itens, aceitação 100 % Delphi 2. Conclusões: a avaliação instrumento paliativos pessoa maligna cutânea.


Introduction
Malignant tumor wounds are skin infiltrates of cancer cells that develop as a result of the growth of a primary skin tumor or metastases from any malignant tumor (1). The prevalence of these lesions is not well documented, but it is estimated from international studies that from 5 % to 15 % of people with cancer develop malignant wounds (2)(3)(4).
Cutaneous malignant tumor wounds are a serious problem in the life of the person with cancer, as they progressively disfigure the body, becoming friable, painful, exudative and fetid. Still, they have psychological and social effects that can interfere in interpersonal relationships, given that the person has feelings of rejection, social isolation, anxiety, sadness, and loneliness (1)(2). In addition, these wounds give people a constant visible reminder of their incurable illness, poor prognosis, and, most often, the inevitable condition of approaching death; this last feeling perceived in their speech during the relational environment of the treatment. Faced with this situation, the individuals need palliative care to control the signs and symptoms of these injuries.
The health care of people with this type of wound requires systematized palliative care, provided by the multidisciplinary team and guided by care protocols, availability of material resources and pharmacological products, articulation among levels of care complexity, as well as effective participation of the person with the wound, relatives, and caregivers (1)(2)(3).
Treatment of cutaneous malignant tumor wounds is complex, given that it requires evaluation of the cancer etiology, the lesion, the person's biopsychosocial and spiritual dimensions, the family's economic condition, control of signs and symptoms, as well as planning and guidance of the care to be performed by the person and the family at home, when curing the disease and wound healing are no longer possible (1)(2)(3)(4). Nurses are the members of the health staff that evaluate and handle the clinical management of wounds and place dressings, so it is up to them to develop skills and abilities that allow them to know and identify individual and/ or social characteristics of people with tumor wounds, besides implementing actions specific to the needs identified. In this sense, placing functional, aesthetic, and comfortable dressings is a challenge for professionals, who can contribute, from the care provided, to a more acceptable presentation of the person in relation to their social image. Therefore, it is important to validate measuring instruments based on real situations that reproduce the routine activities of nurses in professional life, based on scientific studies and which contemplate the systematization of care to homogenize and ensure efficient behaviors in teaching, research, and clinical practice (5). In this perspective, it is worth mentioning a validation study conducted with 78 primary-care nurses in the city of João Pessoa, Paraíba, Brazil, whose results not only contributed to explain how the theoretical and practical knowledge applied to care for people with wounds, based on health education and management, as well as ensuring the theoretical-empirical organization in the proposal of a measure capable of assessing the nurse's knowledge and practice in relation to the person with venous ulcer (6).
Given this scenario, this study sought to present the construction and validation process of an instrument directed to palliative care for people with cutaneous malignant tumor wound.

Materials and Methods
This was a methodological study of the construction and content validity and agreement of a two-stage data collection instrument, carried out from January to July 2017. Data collection began in March 2017, at the beginning of the first Delphi evaluation, after approval by the Research Ethics Committee at Universidade Federal da Paraíba, Protocol 031/2017, and the issuance of the Certificate of Presentation for Ethical Appreciation (CAAE, for the term in Portuguese) 64122116.1.0000.5188, and continued until July 2017, at the time of the conclusion of the second Delphi evaluation, according to the norms of Resolution 466/2012 of the National Health Council of the Brazilian Ministry of Health for research involving human beings.

First stage -construction of the instrument
The theoretical procedure of the instrument development process involved a search for representative items of the construct under investigation, through a revision of the Manual of the National Cancer Institute for Treatment and Control of Tumor Wound and Advanced Cancer Pressure Ulcers (7) and the construction of an integrative literature review in scientific articles (8).
From this literature review process, an instrument with 112 items was constructed to evaluate nurses' knowledge and prac-tice in palliative care for people with cutaneous malignant tumor wound. The instrument is composed of five categories: 1) assessment of the lesion and needs of the person with cutaneous malignant tumor wound; 2) basic care performed with the cutaneous malignant tumor wound; 3) specific care performed with the cutaneous malignant tumor wound (this category is subdivided into eight subcategories, related with the control of the signs and symptoms: a) pain control, b) exudates, c) itching, d) necrosis, e) fistula, f) bleeding, g) odor, and h) myasis infestation); 4) records of Nursing actions, and 5) guidelines for hospital discharge.

Selection of validation technique
The Delphi technique is a methodological strategy that seeks to achieve maximum consensus from a group of experts on a particular topic, when unanimous opinion does not exist due to lack of scientific evidence or when there is conflicting information (9). The technique has some advantages: economic viability, possibility to gather the opinion of qualified professionals, participation of judges from different geographic locations, elimination of bias inherent to face-to-face meetings (10). For said reasons, it has been widely used in health research (5,10).

Selection of judges
Content validation assumes a subjective judgment on whether a measure makes sense intuitively and refers to the degree to which an instrument represents a domain or the relevance of its items. The number of judges recommended is from 6 to 20 subjects (11). To select the judges, the snowball-type convenience sampling was used, which allows defining people with characteristics common to the research interest. The first-sample members were asked to nominate others who meet the criteria to compose the survey sample (12).
Fifty-seven professionals were invited, based on their fulfilling the following inclusion criteria: nurses with Brazilian nationality; doctors and masters with expertise in teaching and research, and specialists in dermatology/stomatherapy, with practical experience.
The nurses selected were contacted via e-mail and invited to participate as research judges, through an invitation letter, which contained the objectives and methodology of the study, the instrument to be validated and the Informed Consent Form, which had a 30-day limit to return it.
Regarding the sample calculation, the G*Power 3.1 statistical package was used to verify the significance of the sample for the intended study. This software is intended to calculate statistical power, based on verifying not only the "n" required for the type of research, but also the type of calculation to be performed (13). Thus, a probability of 95 % (p <0.05), magnitude of the sampling effect (r ≥0.25) and hypothetical power pattern (π ≥ 0.80) were considered. Generating the sample calculation based on these criteria, the sample of 30 subjects was sufficient for what was sought, with indicators: r ≥ 1.98; π ≥ 0.98; p < 0.05.
With respect to sample normality, the Shapiro-Wilk (S-W) calculation was performed, which is appropriate when the sample size is smaller than 50 subjects, and does not require the distribution parameters to be specified (14)(15).
The form sent to the judges consisted of two parts: the first directed to the characterization of the study participants, with questions about gender, age, professional qualification, area of work, and place of work; the second part referred to the presentation of the structured instrument with 112 items and the instructions for its completion.

Concordance analysis
The judges were asked to evaluate the adequacy regarding the clarity and relevance of the items, which were arranged in tables, followed by two columns: one highlighted the clarity of the items presented; the other, their pertinence. Judges had to indicate, by means of a binomial response to each of the questions, how much they agreed on the category, subcategories, and item specificity. In case of inadequacies in the items, space was reserved for comments, justifications, and suggestions that the judges deemed pertinent.
The suggestions were also considered, which resulted in changes made by the authors, later justified. Agreement of at least 80 % among the judges can serve as decision criterion on the pertinence and/or acceptance of the item referred to theoretically (13). Due to this, a minimum 80 % concordance was considered to validate the instrument.

The second assessment by the judges
A form was sent containing both the original and modified versions of the instrument. In this phase, the judges had to evaluate the original and modified instrument according to their suggestions and analytical treatment, assigning the quality of assessment among the instruments related to palliative care for the person with cutaneous malignant tumor wound.
After evaluating the instrument, the following calculations were applied: Lambda 2 Guttmann as best estimation of reliability of instruments that form categories and which are answered by a sample equal to or less than 30 subjects; Kuder-Richardson's Ρ, which seeks to verify the reliability of the degree of agreement between evaluators in a dichotomous measure; the Kappa coefficient, and the CVI to verify the level of concordance of the judges in relation with the items assessed.
On the characterization of the sample, as to titration, 11 were specialists in dermatology/stomatherapy, nine were masters, and 10 were doctors; as for the area of expertise, 15 worked in hospital care and, of these, nine performed care activities directly with people with malignant cutaneous tumor wounds; 11 were teachers and four were technical advisors. Regarding the workplace, 12 worked in public hospitals, eight in public universities, four were self-employed, two worked in outpatient services for wound treatment linked to municipal health departments, two in private wound care clinics, and one in private hospital and one in private faculty.

Discussion
Considering the instrument in question, it was observed that most of the items had a percentile equal to or above 80 % regard-ing the degree of concordance in clarity, as well as in pertinence (16); for this reason, they remained on the proposed measurement scale; however, items 4, 5, 6, 22, 24, 44, 45, 46, 47, 56, 58, 71, 74, 75, 76, 79, 91, 94, 96, 97, and 98 did not present desired percentages, which were excluded from the respective categories. It is worth mentioning that some of these results are related to the lack of knowledge by the judges about the products and coatings highlighted in these items, whereas the evaluators themselves made it explicit in the space for suggestions and comments; for this reason, it was decided to exclude them.
Also, all items with a percentage equal to or greater than 80 % were significant with p <0.001. The intention in using Guttmann's Lambda 2 (λ) calculation is because some studies (17)(18) found that using Lambda is the best reliability estimate when the instrument assessed is composed of few items that form categories or by a sample equal to or less than 30 subjects (19)(20)(21).
Guttmann's Lambda 2 (λ) in this study revealed (for the set of items on the assessment of clarity and relevance λ2 = 0.84 and 0.81, respectively) higher score than required by the statistical literature (19)(20)(21)(22), which is 0.80, intended for low-participation assessments, as well as instruments that require instructional comments. Thus, the higher the statistically required λ, the better the test is able to differentiate from actual results and measurement errors. Considering these results, the Lambdas presented mean the percentage of the real variation of the measurement, that is, the Lambdas explain values > 80 %.
The appropriateness as to the degree of clarity and relevance of the items of the instrument in question indicated by the judges can be used for their assessment of the phenomenon "the knowledge and practice of nurses in palliative care for people with cutaneous malignant tumor wounds".
In addition, the analysis generated a version with changes in the instrument content, based on the judges' comments and suggestions, namely: item 28 was excluded, but the sentence was directed to item 27, once the judges understood that the information contained in these items complemented each other; items 38, 43, and 49 were directed from Category 3 to 2, once the judges understood that they would be better articulated in basic care; items 42, 81, 83, and 84 were excluded from Categories 3 and 8, because the judges realized that the information was repeated and understood that they should remain only in the basic care performed with the lesion.
The Kappa (K) index, which measures inter-observer agreement and degree of agreement and consistency of the judges regarding the permanence or not of the instrument's items, taking into account the indications of "inadequate" for them, ranges from "minus 1" to "plus 1"; thus, the closer to 1, the better the level of agreement between the observers. As an acceptance criterion, agreement was established ≥ 0.61 among the judges, which presents a good level (23).
The instrument subjected to the first evaluation by the judges contemplated aspects related with palliative care expected for execution by the palliative-care nurse on the person with malignant tumor wound; however, the procedures involved in the care process did not include a detailed description of the technique. This initial configuration influenced the Kappa values and subsidized the instrument reformulations for the second evaluation of the instruments (original and modified); however, some were maintained, given that they are palliative care considered essential and supported by the relevant literature, namely: items 2, 3, and 12, regarding clarity (7-8).
Finally, apparent validity was verified (such considered the theoretical definition of a variable that, in fact, seems to measure the item proposed). Specifically, it refers to what the item, by the type of questions or situations presented, seems to evaluate; for this, the CVI was performed (12,16) related to the quality of agreement and content of the items in the judges' conception for clarity and importance, and noting that this indicator ranged from 0.91 to 1.00.
Although items 2, 3, and 12 did not have an acceptable Kappa value, they were chosen to remain because in item 2, size refers to the measurement of the lesion, i.e., it is an assessment parameter that confers the evolution of the wound. It is recommended to measure length and width and, if possible, to measure depth and/or bulging, given that malignant tumor wounds also have a vegetative aspect (1,(7)(8). Regarding the permanence of item 3, it should be highlighted that staging is another evaluation method for tumor wounds and it is a classification proposed by the manual of the Brazilian Ministry of Health (7) and by the relevant literature (1)(2)(3)8), whose purpose is to classify the wound according to the degree of tissue compromise and predominant signs and symptoms. Item 12, related to the assessment of the border and the periphery skin, was also maintained because the border and the peri-lesion area presented, besides peeling, a common problem found in tumor wounds, maceration, excoriation, fragility, and hyperkeratosis. These damages cause pain and discomfort in people, as well as affecting negatively their quality of life (24). Items 36 and 58 were excluded because they did not have an acceptable Kappa percentage, although they are supported by the relevant literature (1-3, 7-8).
All comments and suggestions from the judges were accepted and the items were reformulated to serve them in their entirety. To leave no doubt, a second evaluation by the judges was requested; for this, the original and modified instruments were sent so that they could analyze and attribute the evaluation quality between the versions. Of the 30 judges who participated in the first instrument evaluation, over 50 % (17) responded to the second version entirely and accepted it. It should be highlighted that loss of participants is expected with each Delphi evaluation, but it is not a limiting factor to analyze the research data (25).
Overall, although these results guarantee the initial quality of the instrument, further studies are important to determine its reliability, either in the more classical or the most complex psychometric perspective, which aims at the quality of the intended theoretical construct, a condition that will permit determining the degree of coherence with which the instrument measures the theoretical attribute under study. Given the aforementioned, the definitive instrument, which follows the compliance by the judges and agrees with the analytical treatment, remained with five categories and 84 items, as seen in Table 2.

Conclusions
In general, this study sought to describe the validation process of a measure that contributes to the evaluation of palliative care for people with cutaneous malignant tumor wound, performed in clinical oncology and palliative care units. From the results observed, it is possible to state that, in this first stage of the analysis of the measure, the instrument proved adequate for what it was intended to measure, given that the statistical indicators revealed percentages equal to or higher than those required by the literature, as well as categories and subcategories related to palliative care.
In this sense, it is believed that the construction of instruments with such theoretical and practical perspective is valid for the health area and, specifically, for Nursing; this condition probably contributes to assessing the quality of care and, thus, direct-ing necessary changes in the process of caring for the Nursing staff, whose main objective is to expand the quality of care and care in the context of the local reality.
Thus, in practice, reduced data loss, refined opinions, and the consensus of judges regarding palliative care for people with cutaneous malignant tumor wound aimed at standardizing the measure and clearer interventionist attitudes, based on "place" in which respondents pointed out the degree of importance of items and categories in the area of nursing, which indicated that the instrument in question is applicable.