Content validation of the Self-perception of Family Health Status scale using the Delphi technique

Objective: To ascertain the content validity of the Self-perception of Family Health Status scale. Method: A validation study of an instrument with an online Delphi panel using the consensus technique. Eighteen experts in the subject were intentionally selected, with a multidisciplinary origin and representing different professional fields. Each of the proposed items was assessed using a five-point scale, and open-ended questions, to modify or propose items. Descriptive analysis was performed of the sample and the items, applying criteria of validation/elimination. Results: The first round had a response rate of 83.3% and validated 75 of the 96 proposed items; the second had a response rate of 80%, and validated the 21 newly created items, concluding the panel of experts. Conclusions: We present an instrument to measure self-perception of family health status, from a nursing perspective. This may be an advance in scientific knowledge, to facilitate the assessment of the state of health of the family unit, enabling detection of alterations, and to facilitate interventions to prevent consequences to the family unit and its members. It can be used in clinical care, research or teaching.


Introduction
The family is considered to be the principal nucleus of socialization of its members.It is the place where values are transmitted, where ideas are learned and adopted, and beliefs and norms of conduct are acquired.
Its members are grouped into subsystems separated by symbolic boundaries, each contributing to family functioning through an exercise of roles and tasks necessary for the whole (1).
With the development of systems theory, adapted to living systems, the concept of family nursing has been defined as nursing care delivered to any system or household (2) .It is necessary to be present and in intimate contact with the family, to discover ways that can fortify, mobilize, and propel it to achieve balance and well-being, going beyond the biological aspect of its members (3) .Nurses need to conduct research to recognize the needs of families, sources of support and social networks available to promote the health of its members (4) .
All of this requires a change in clinical practice, to move from an individual to a group approach, including the social context in which the family group is immersed.To achieve this change, it is necessary to have a conceptual basis, an explanatory model, and an organization capable of implementing this process (5) .At present, several theorists have participated in different concepts of family and nursing care of the family (6) , which is necessary to assume the challenge of developing the concept of family health and its dimensions, from a nursing perspective that allows its valuing.
For the World Health Organization (WHO), family health describes its function as a unit of structure, in relation to the family as the primary social agent, to promote health and well-being.This is related to satisfying the needs of its members, the interactions between the individual, the family and society, the solution to problems, or the ability to adapt to crisis situations.
Family health builds on the internal dynamics of the family, in decision making, in the education and socialization of its members, and in the availability and access to resources necessary for meeting its needs (7) .
It is not static, it is a continuous and dynamic process that is undergoing continuous equilibrium and change, in response to events characteristic of the life cycle and those accidental events or situations that affect them.
Family health is intimately related to the family environment, the integrity of its members, the organization and operation of the unit, the capacity for resilience when faced with problems affecting it, and ways to address those problems (6) .resistance (8) .

One of the difficulties in
From a systemic perspective of the family and a nursing approach that takes the NOC as its reference, Lima Rodriguez and colleagues, in their study of the dimensions of family health, proposed a rating system of family health which contemplated, along with the demographic aspects, the composition, structure, family life cycle and stress agents that may be impacting the family (6) .They considered that these were: the family climate, family integrity, the function of family dynamics, family resistance, and family coping.Furthermore, when there was a family member with special needs or who required professional assistance, it was necessary to add family support, family normalization, and family participation in the care process.
To operationalize this proposal it is necessary to develop a series of instruments that, together with interviews and observation, enable the gathering of data necessary to evaluate the health of the family.Although some instruments exist that are applicable to the study of the family, for example the Family Apgar to evaluate family dynamics, the MOS questionnaire for social support, the Social Readjustment Scale for stressful life agents (9)(10)(11)(12) , several authors emphasize the urgent need for efforts to develop and validate new measurement instruments (12)(13) , adapted to the objectives of the discipline that needs them.This article presents the design and content validation of an instrument intended to assess selfperception of family health status, inspired by the criteria of the Nursing Outcome Classification.This can Lima-Rodríguez JS, Lima-Serrano M, Jiménez-Picón N, Domínguez-Sánchez I.
present an advance in nursing knowledge, to allow us to identify family health situations that may require nursing interventions to improve or prevent the consequences they may have on the family system and also on its members, as in the family that may have an economic risk factors for its members (1,14) .

Method
The study was conducted in two phases: design of the scale and validation of its content.

Phase I. Scale design
This was performed in two stages: a) literature review, b) brainstorming and consensus of a small group of health professionals, teachers, managers and researchers, related to the care of the family, for item selection and preparation of the scale.

1) Definition of dimensions and attributes
Following previous authors, the NOC outcome criteria were used as a reference framework for designing the scale (15) .We drew from outcome criteria proposed for the family health domain and the family wellbeing class, such as the family climate, family integrity, family functioning, family resistance, and family coping.In addition, for families in which one of its members has impaired health or necessitates health care, family support, family normalization, and family participation in the care process were added.
A literature review was conducted to identify the dimensions of family health and the instruments used to measure it from the user perspective.The search was conducted in PubMed, SCOPUS, Cuiden Plus; the keywords used in Spanish and English were health, family, assessment or evaluation, scales, indices, test.Articles were included with any methodology (quantitative, qualitative) and whose objective was clearly related to the research problem, or that made reference to aspects of interest for the work.We reviewed the bibliographies of the located articles.

2) Selection of items and developing the scale
The research group, through brainstorming and consensus, adopted a set of items, which were classified according to the dimension of family health to which they referred (9)(10)(11)(12)16) . Subsquently, items were designed specifically for those attributes not represented, and the first version of the scale was constructed (Figure 1).questionnaires, referred to as phases.The responses to each questionnaire are considered for subsequent reformulation (18) .This method allowed the purification and refinement of the initial scale.

Study Population
Through purposive sampling, 18 experts were selected.Given the difficulty of characterizing the sample, and to define the criteria for inclusion in validation studies, due to the lack of consensus on these (19) , we chose to have experts of a multidisciplinary origin (nurses, physicians and psychologists) belonging to different professional fields (healthcare, manager, teacher, researcher).To determine their inclusion, their knowledge and professional experience in relation to the topic of the instrument under validation were previously evaluated.

Procedure
Employing the web application "e-survey", information about the study objectives and the notions and theoretical considerations adopted for each of the dimensions of the scale was sent to participants.They were invited to participate voluntarily, ensuring the confidentiality and privacy of their input.After consent, the instrument was sent with instructions for completion.The study was approved by the Committee on Research Ethics of the Universidad de Sevilla, which reported favorably on the fulfillment of the requirements for human experimentation.
The degree of adequacy of each item must be established, according to a score of 1 to 5 (1 = minor adequacy, 5 = higher adequacy).Initially 96 items were proposed, grouped into eight subscales, in functions of the dimensions considered family health.They were also offered the opportunity to suggest new items or propose amendments to those proposed.
Following the analysis of the data obtained in each round, the items were validated and eliminated according to the criteria: Validate: Mean ≥ 3.5 and Median ≥ 3 and high ratings (4-5) ≥ 80% and / or standard deviation ≤ 0.90.Eliminated: Mean <3.5 and Median <3.The eliminated items were reviewed and modified according to proposals obtained, and moved through to the next round (20).
After each round, as a measure of feedback, a report was sent to each expert describing the item analysis, results obtained, and items validated, eliminated, or that should be assessed again.96.If we do not agree with the professional we tell him In order to measure responses a Likert scale was chosen, because it allows a subject to respond, giving a degree of intensity to the statement of the item.Regarding the number of items necessary as a starting point, there is no rule, but obviously a higher the number of items will make it easier to make a good final selection (17) .

Type of Design
We conducted a validation study of an instrument through the use of an on-line Delphi panel, with expert judges, with a consensus technique.This technique attempts to obtain consensus of expert opinions on a subject through a series of structured Lima-Rodríguez JS, Lima-Serrano M, Jiménez-Picón N, Domínguez-Sánchez I.

Data analysis
A descriptive analysis of the sociodemographic and study variables was conducted, supported by SPSS 18.0.

Phase I. Questionnaire design
The scale was constructed to measure selfperception of family health status, initially consisting of 96 items grouped into eight subscales, one for each dimension of family health.These were: family environment, family integrity, family functioning, family resistence, family coping, family support, family normalization, and family involvement in the care process.For each dimension, 12 items were established.
Three possible answers were identified: almost never (AN), at times (AT), almost always (AA).

Phase 2 Validation of the content of the scale
The questionnaire was administered to 18 experts, and it took a total of two rounds to reach the necessary level of consensus.
There was an initial response rate of 83.3% (15 experts), and 80% (12 experts) in the second round.
It took a total of two rounds to reach the necessary level of consensus.In the first, 75 items were validated (78.1%) and 21 items were eliminated (21.9%).Taking into account the contributions of the experts, modifications were made and these 21 items were subjected to another round of assessment, which were finally validated in their entirety, closing the panel of experts for the content validation (Tables 1 to 4).

Discussion
Several authors agree on the need to design and validate new measurement instruments when no others exist that measure what we really intend to measure, as is the case of the scale presented here (12)(13) .
Nursing outcome criteria describe the state, behaviors, reactions and feelings of the patient and, although they initially were developed to measure these responses to care administered (19) , they can become evaluation criteria that allow us to determine status, in the case at hand, related to family health.
The Delphi technique is widely used for the validation of scales and questionnaires by expert opinion (12,(20)(21)(22)(23) , making explicit criteria important for the selection thereof and ensuring that such experts are appropriate for the subject matter of the study to validate the instrument, thus avoiding bias (24) .Other measures proposed to avoid bias are standardizing the definitions of each dimension of the scale, and its principal subcomponents, to make them available to the expert group to unify knowledge on the field of study and to be critical when evaluating and scoring the items, with positive repercussion on the validity of the instrument (25) .
The performance of the technique on-line promoted speed in completion and facilitated the opportunity for participation of international experts.Also, the sample size used in the validation process was consistent with the proposed number in previous studies to stabilize the responses of each of the items.A response rate determining family health status is due to limitations that exist in the delimitation of the concept and the lack of consensus as to which indicators are needed to measure it, creating conceptual and methodological difficulties at the time of its evaluation.The Nursing Outcomes Classification (NOC) contains a set of indicators that can provide utility for overcoming these difficulties.The family health domain and family well-being class, contain the outcomes that describe the state of family health and social competence of the family as a unit in the overall family environment, including indicators of coping with family problems, family support during treatment, social climate of the family, health status of the family, family functioning, family integrity, standardization of the family, family participation in health care, and professional and family www.eerp.usp.br/rlaeRev. Latino-Am.Enfermagem 2013 Mar.-Apr.;21(2):595-603.
characterize the sample of experts: age, gender, place of residence, level of schooling and area of professional dedication.

Figure 1 -
Figure 1 -Initial scale items sent to experts

Table 1 -
Items validated for Family Social Environment and Family Integrity

Table 2 -
Estimated items for Family Functioning and Family Resistance

Table 4 -
Family standardization and family involvement in the care process