Cultural Adaptation and Validation of the Ostomy Adjustment Inventory-23 for Brazil

The creation of a stoma results in an aesthetic change in body image that can have substantial psychological impacts.^1–5 Patients may perceive themselves as being disabled or even mutilated because of the surgery, and they may also feel disgust and/or embarrassment for having to eliminate stool through a device in the abdomen; patients may also experience a fear of not being able to relate sexually, which can lead to reduced libido and depression.⁶ These factors can create a series of obstacles for patients, thus impeding their attempts to adapt and reintegrate into their family, social, and work roles.^7,8 

Measuring the impact of a stoma, with regard to it impairing daily activities and affecting the perception of health and functional status, plays an important role in allocating health care resources and is also useful in assessing the efficacy of nursing and multidisciplinary interventions. Furthermore, optimizing the benefits of effective nursing and multidisciplinary interventions increases patient satisfaction, thus facilitating patients’ health management and, consequently, augmenting the adaptive process.⁹ In general, the successful management of stressful conditions leads to positive adaptation and the restoration of psychological balance.^10–12 Thus, it comes as no surprise that psychosocial adjustment, defined as the management of anxiety and depression as well as social issues, such as daily and social activities,¹³ highly influence health through sociodemographic and psychological factors related to disease.

The psychosocial adjustment of individuals living with an ostomy has been evaluated using the Ostomy Adjustment Scale (OAS),¹⁴ the Ostomates Self-Adjustment Scale (OSAS),¹⁵ and the Ostomy Adjustment Inventory-23 (OAI-23).¹⁶ Developed in 1983, the OAS consists of 34 items. It has good validity indices, sufficient internal consistency (Cronbach’s alpha of 0.87), and a satisfactory test-retest coefficient (intra-class correlation coefficient [ICC] of 0.72).¹⁵ In 2015, OAS was adapted and validated for use in China.¹⁷ 

In 2000, Maekawa¹⁴ validated the OSAS with 509 Japanese individuals living with an ostomy. The OSAS has 30 items, distributed among 6 factors, and presents good internal consistency (Cronbach’s alpha coefficient of 0.81). However, the stability, criterion validity, and predictive utility of this instrument have not been evaluated.¹⁴ The OSAS has also been adapted for use in Norway.¹⁸ 

In 2009, Simmons et al¹⁶ developed the OAI-23, which was specifically designed to assess the social and psychological adaptation of individuals living with fecal and/or urinary stomas in the United Kingdom.¹⁶ The original version of the OAI-23 contains 23 items, has good reliability indices (Cronbach’s alpha of 0.93), and has been adapted and validated for use in Italy¹⁹ and China.²⁰ The aim of the present study was to culturally adapt the original OAI-23 instrument to Brazilian Portuguese and assess the measurement properties of this new version.

Design and ethics approval. This methodological study sought to adapt and validate the OAI-23 for use with Brazilian individuals living with an ostomy. The researchers obtained authorization from the authors of the original instrument as well as from the health centers where the questionnaire was previously applied. In addition, the experimental design was approved by the Research Ethics Committee at the University of São Paulo School of Nursing (process no. 829/2009). The research protocol was also approved by the directors of Specialized Outpatient Centres from Santa Maria city, in Rio Grande do Sul; Campo Grande city, in Mato Grosso do Sul; and Cabo Frio city, in Rio de Janeiro, in addition to the Director of São Paulo Ostomy Association, in São Paulo. 

All participants were informed about the research’s aim and steps; they then gave their verbal consent and also signed an informed consent form. The data collection period was from 2011 to 2012.

Measures

Sociodemographic and clinical characteristics. The researchers developed a paper-and-pencil questionnaire for the collection of sociodemographic data as well as clinical features. Sociodemographic data such as sex, age, marital status, religion, education level, and work status were collected. The authors also collected clinical data (including amount of time since stoma surgery, follow-up time, cause of the ostomy, and comorbidities).

OAI-23. The OAI-23 is intended to measure the social and psychological adaptation of individuals who underwent ostomy surgery as well as the adjustment necessary for accepting the stoma and improving coping-related behaviors.¹⁶ The original English version of the OAI-23 is based on items from the English-adapted version of the OSAS.¹⁶

In the factorial analysis of the OAI-23, 4 factors account for the total variance of the phenomenon. These factors include acceptance, which refers to the adaptation of the patient to the stoma (factor 1: items 1, 3, 4, 6, 9, 14, 15, 19, and 23); anxiety/preoccupation, which refers to constant thought and worry about the stoma (factor 2: items 12, 13, 17, 20, and 21); social engagement, which refers to the reintegration of patients into their family, social, and work environments (factor 3: items 5, 7, 8, and 11); and anger, which refers to the degree of irritation/frustration expressed by the patient that is caused by the stoma (factor 4: items 2 and 10). Items 16, 18, and 22 had factor loads greater than 0.40 in 2 or more factors; for this reason, they were considered independent and not included in any of the instrument factors.¹⁶

Each item was graded on a 5-point Likert scale (0 to 4). Twelve (12) items (2, 5, 7, 8, 10, 11, 12, 13, 16, 17, 18, and 21) were negatively described and scored inversely. The total score ranged from 0 to 92, with higher scores being indicative of better psychosocial adaptation. A cutoff point has yet to be established.¹⁶

Janis and Field Self-Esteem Scale. The Janis and Field Self-Esteem Scale assesses the level of self-esteem and feelings of social adequacy.^21,22 Developed by Janis and Field²¹ and translated and adapted by Ulhôa for Brazil,²² this scale is composed of 23 items with graded answers ranging from 1 to 5: always (1), frequently (2), sometimes (3), rarely (4), and never (5). Answers are not classified as correct or incorrect. The answers to questions 3, 7, 8, 9, and 10 have inverse values. There is no cutoff grade or rating for the self-esteem level. The total scores are obtained by summing the values from all the items and can range from 23 to 115.²²

Cultural adaptation. The cultural adaptation of the OAI-23 for Brazilians was conducted according to the standards published and revised by Beaton et al,²³ and included some modifications as proposed by the Institut Hospital del Mar d’Investigacions Mèdiques (IMIM).²⁴ 

The translation of the original instrument was performed by 2 independent and qualified translators. One translator’s native language was English, and they were also fluent in Portuguese. The other translator was an expert in English. Both were aware of the objectives of the translation. For this phase, the translations were evaluated using a standardized instrument that contained, in addition to the column for recording the translation, columns for difficulty scores (0 to 10, less to more difficult); comments about these difficulties were also provided.

The committee of expert evaluators consisted of 2 enterostomal therapy nurses, a nurse specializing in quality of life, and a coloproctologist. All evaluators were fluent in English and communicated via telephone and/or e-mail. Their evaluations were aimed at analyzing and comparing the semantic, idiomatic, cultural, and conceptual equivalences of the original and translated versions of this instrument. A minimum level of agreement of 80% was adopted for determining equivalence among the evaluators.²⁵

Following this first phase of translation and evaluation, a pretest was applied to a heterogeneous focus group (n = 11). Inclusion criteria included being 18 years of age or older, having a temporary or permanent stoma of any type and for any reason, and having sufficient physical and mental conditions to be interviewed (eg, no pain, cognitively able to understand the questions). This step, which may also be denoted as cognitive debriefing, was performed with a small sample size according to recommendations found in the literature.²⁶ Patients who met these criteria were called by the president of the Ostomy Association of the State of São Paulo (AOESP) and asked to participate in the study.

The version of the instrument administered during the pretest was then back-translated into English by a different set of translators, both fluent in English and Portuguese and unaware of the study objectives. After completing the translation and pretest phases, and prior to clinical application, back-translations were sent to Dr. Kingsley Simmons, one of the authors of the original instrument, for analysis. 

Evaluation of measurement properties of the adapted version. The Brazilian Portuguese adapted version of the OAI-23, resulting from the first phase of the present study, was applied to patients who had undergone ostomy surgery at 3 outpatient hospitals in Cabo Frio, Rio de Janeiro (RJ); Campo Grande, Mato Grosso do Sul (MS); and Santa Maria (RS), Rio Grande do Sul (RS) as well as patients registered at the AOESP in São Paulo (SP).

For the quantitative composition of the sample, a minimum of 8 participants was considered for each instrument item,²⁷ totaling 184 people. The same inclusion criteria applied during the focus group phase were also considered in the next phase of the experimental approach. 

A convenience sample of patients was invited to participate prior to their outpatient appointment. After signing the informed consent form, data collection occurred through individual interviews in a private office or restricted area of the health care center or location. The researcher recorded the interviews in a paper-and-pencil instrument. To evaluate the stability of the adapted version, all patients were invited to return within 15 days and complete the OAI-23 questionnaire again.

In addition to the data collected from the adapted version of the OAI-23, an instrument to obtain sociodemographic and clinical data as well as the Janis and Field Self-Esteem Scale¹⁵ were also applied.

Data management and analysis. The data were organized and analyzed using the software Statistical Package for the Social Sciences, SPSS1 version 20.0 Descriptive analyses were performed for all variables. For all statistical tests, a significance level of 5% was considered. 

The reliability of the OAI-23 was assessed by evaluating internal consistency and stability using Cronbach’s alpha coefficient and the ICC, respectively. It has been reported previously that Cronbach’s alpha coefficients ≥ 0.70 are indicative of satisfactory internal consistency.²⁸ Although a sample corresponding to 15% of the total was established to analyze the reliability and stability of the test/retest, 81 patients with ostomy (42.4%) responded to the retest. The magnitude of the test/retest agreement level was standardized as follows: < 0.40 (weak), 0.41 to 0.60 (moderate), 0.61 to 0.80 (good), and 0.81 to 1.022 (excellent), with higher coefficients being associated with greater instrument stability. Confirmatory factor analysis (CFA) was performed to determine whether the adapted version of the OAI-23, presented herein, exhibited the factorial structure of the original version when only including the factors of the instrument. For the CFA, the following model fit indices and parameters were considered: goodness of fit index (GFI) < 0.80; root mean square error approximation (RMSEA) ≤ 0.08; adjusted goodness of fit index (AGIF) ≤ 0.90; and chi-square measure < 0.05.27. 

The convergent validity of the adapted version of the OAI-23 was assessed by correlating the total scores on the OAI-23, as well as the scores of each factor with the Janis and Field Self-Esteem Scale scores. This was accomplished by applying the Kolmogorov-Smirnov (normality) and Spearman tests. To visualize and analyze the magnitude of these correlations, the following classification was adopted: < 0.30 (weak), 0.30 to 0.50 (moderate), 0.50 to 0.99 (strong), and 1.00 (perfect).²⁹

The discriminant validity was assessed using backward logistic regression after comparisons were made between the scores on the adapted version of the OAI-23 and age, sex, ostomy time, type/anatomical origin, cause of the stoma, and character of the stoma (temporary vs. permanent).

Cultural adaptation. At the cultural adaptation phase of this study, the committee of expert evaluators reached consensus on the translated title Stoma Adaptation Inventory-23. However, it was decided to maintain the abbreviation OAI-23 to facilitate international comparisons. Regarding the instructions for completing the questionnaire, the authors opted for the following sentences: “The statements below are related to how you feel about your stoma. Please mark an “X” to indicate your opinion. Please try to answer all of the questions.” 

Regarding the responses, the following 5 were suggested: I completely agree, I agree, I disagree, I am not sure, and I strongly disagree. For the 23 inventory questions, the committee reached a consensus on 8 items from the first translator and 15 items from the second translator; that is, agreement between the judges was higher than 80%, with some suggestions considered and modified in the instrument. Examples included replacing the word “surgery” with “operation” in item 1 and replacing “taking care” with “dealing with” in item 20.

The pretest for the Brazilian Portuguese adapted version of the OIA-23 included 11 patients aged 57 to 81 years (mean, 69.3) with a predominance of females (n = 8; 72.7%). Education levels ranged from none (1; 9.0%) to completed college (4; 36.4%). The types of ostomies included ileostomies (6; 54.5%) and colostomies (5; 45.4%). All stomas were permanent, and time since the ostomy procedure ranged from 3 to 45 years (mean, 19.3). 

During this phase, the focus group suggested that at the end of the instructions the word “currently” should be added to promote the temporal sense of the research, demonstrating that the assessment of the adaptation of the patient at the time of the interview. The group also suggested that the text font size be reduced so that the entire instrument fit on 1 page. This modification made it easier to visualize the response categories and reduced the risk of the participants not completing the back of the form. In item 2, “I have trouble touching or looking at my stoma,” it was originally difficult to understand the sentence because of its negative connotation; it was subsequently restructured in a more positive way, while attempting to simultaneously maintain its negative character, which was an important procedure in the statistical analysis of the data. Item 8 generated doubts regarding the meaning of the word “reluctant,” suggesting its change to the expression “I feel insecure.” Finally, item 21 had a clear sentence structure, but the word “anxious” generated different understandings, and its modification to “worried” was suggested.

After the semantic assessment and reaching consensus on the changes to the instrument as suggested by the focus group, the back-translated English language version was submitted to and ratified as equivalent to the original OAI-23 version by Dr. Kingsley.

Evaluation of measurement properties. For the validation of measured properties of the adapted version of the OAI-23, 191 Brazilian individuals living with an ostomy were enrolled to participate in the study; 74 (38.7%) were from Santa Maria (RS), 30 (15.7%) from Cabo Frio (RJ), 30 (15.7%) from Campo Grande (MS), and 57 (29.9%) from São Paulo (SP). There was a predominance of females (109; 57.1%), and a mean age of 58.9 years (SD = 14.7). Clinically, the average time of having the stoma was 7.2 years (SD = 8.5). The majority of patients underwent surgical colostomies (140; 74.1%), and most stomas were permanent (145; 75.9%). Cancer was the main cause for ostomy and subsequent stoma (116; 61.4%). All characteristics of the sample are described in Table 1.

The total score of the Brazilian Portuguese adapted version of the OAI-23 was 62 (SD = 12.1). Factors 1 through 4 had average scores of 28.7 (SD = 4.5), 10.6 (SD = 4.0), 9.4 (SD = 3.2), and 5.2 (SD = 2.0), respectively. The Janis and Field Self-Esteem Scale had a mean score of 79.1 (SD = 12.1). When assessing the adapted version of the OAI-23 for reliability, Cronbach’s alpha coefficients of 0.65, 0.67, 0.61, and 0.59 were determined for factor 1 (acceptance), factor 2 (anxiety and concern), factor 3 (social engagement), and factor 4 (anger), respectively. This indicates that each factor of this version of the instrument had adequate internal consistency, especially regarding factors 1 and 2. Notably, the Cronbach’s alpha coefficient for the total score of the adapted version of the OAI-23 was 0.85, which is also indicative of satisfactory internal consistency. When an item was excluded, the Cronbach’s alpha coefficients ranged from 0.83 to 0.85, showing that the exclusion of some items had little impact on the overall value. The magnitude of correlation of each item with the total score is provided in Table 2 and shows that there are various strengths of correlation present. For example, item 14 showed the weakest level (r = 0.07) and item 21 displayed the strongest (r = 0.65). Regarding stability, the adapted version of the OAI-23 had a statistically significant ICC value of 0.903 (95% confidence interval [CI], 0.850–0.938; P < .001).

Figure 1 shows that chi-square and RMSEA are within the reference values, while the other measurements are close to the ideal values. Additionally, the factorial load for each item ranged between 0.25 and 0.88, with the lowest load attributed to item 3; the factorial loads of factors 1 to 4 ranged from 0.43 to 0.96. 

As shown in Table 2, the convergent construct validity results identified positive correlations of moderate magnitude between factor 2 (r = 0.512; P < .001) and factor 3 (r = 0.407; P < .001) scores on the adapted version of the OAI-23 and the Janis and Field Self-Esteem Scale. A positive but weak correlation was also detected for factor 1 (r = 0.216; P = .003) and factor 4 (r = 0.289; P < .001).

Linear regression analyses showed that the Brazilian Portuguese adapted version of the OAI-23 discriminated individuals with stomas only in terms of age, with older individuals tending to exhibit a better adaptive response when compared to their younger counterparts (constant = 54.072; age coefficient = 0.136).

There are few standardized instruments for the evaluation of psychosocial variables, including adaptation, in individuals living with an ostomy. In addition to the OAS developed by Olbrisch¹⁵ in the 1980s and the aforementioned OSAS,¹⁵ Simmons et al¹⁶ decided that a new instrument was needed because none of the instruments available at that time were adapted and validated for Brazilians. Recognizing the relevance for assessing the adjustment of persons living with a stoma and its important association with health-related quality of life, the current study aimed to culturally adapt and validate the OAI-23 for Brazilian culture and obtained very satisfactory results.

During the cultural adaptation phase, it was found that the instrument questions were easy to understand; however, there were some concerns with those participants who who had received less formal education. To address these concerns, semantic adjustments aimed at facilitating the comprehension of the question were made. For example, original item 2 (“I don’t like to touch or see my stoma”) had a negative overtone due to lack of comprehension,  and so, the new statement (“I have trouble touching or looking at my stoma”) gained a positive structuring in the Brazilian Portuguese language. 

In this study, the adapted version of the OAI-23 had a Cronbach’s alpha coefficient of 0.85 for the total score and around 0.60 for each of the 4 factors. In the original study,¹⁶ Cronbach’s alpha coefficient was higher for the total score (0.93) and ranged from 0.74 to 0.87 for the factors. In that study, only factor 4 presented a Cronbach’s alpha coefficient below 0.65, revealing a psychometric performance similar to that found in the present study, although always with higher values. The low internal consistency may be related to sample homogeneity (ie, samples with greater diversity discriminate more reliably), fewer number of items, items with little or no relationship to the factor, and very confusing and/or obvious items.³⁰ Indeed, it is plausible that the more heterogeneous sample and the reduced number of items account for the low Cronbach’s alpha coefficient for factor 4 in the present study. It is also worth mentioning that a recent Italian cultural adaptation and validation study with the OAI-23 reported Cronbach’s alpha coefficients of 0.91 for the total score and values between 0.87 to 0.93 for the factors.¹⁹

With regard to the stability of the original instrument,¹⁶ the authors reapplied it to 84 patients 15 days after the first application and obtained an ICC of 0.83, which indicates good stability. Similarly, we also observed excellent stability with the the Brazilian Portuguese adapted version of the OAI-23 (ICC = 0.903; P < .001). 

Using exploratory factor analysis (EFA), the original version of the OAI-23 presented 4 factors.¹⁶ The Italian version of the OAI-23 presents satisfactory indices of fit to the model (RMSEA = 0.089; CFI = 0.901)¹⁹ when compared with the version adapted for Brazil (RMSEA = 0.052; GFI = 0.885; AGFI = 0.847). Nevertheless, EFA confirmed that the adjustment construct in the Brazilian Portuguese adapted version fits the original one.

Convergent construct validity analysis of the original OAI-23 was performed by comparing it with the Acceptance of Illness Scale (AIS) scale.¹⁶ The AIS assesses the extent to which patients with chronic illnesses accept disease-related limitations without experiencing negative results. Currently, there is no available instrument for assessing the self-concept of individuals living with an ostomy in Brazil. It has been proposed that the development of the self-concept is antecedent to the development of self-esteem because one must first form an opinion about oneself, systematically assessing abilities and inadequacies, to determine the degree of satisfaction or self-esteem.³¹ In this context, the authors assessed the convergent validity of the Brazilian Portuguese adapted version of the OAI-23 by comparing it with the Janis and Field Self-Esteem Scale.¹⁵ Similar to the positive correlation (r = 0.723; P < .001) detected between the original version of the OAI-23 and the AIS, a positive correlation was also detected when comparing the Brazilian Portuguese adapted version of the OAI-23 and the Janis and Field Self-Esteem Scale,²² but it ranged from weak to moderate in magnitude (from 0.216 to 0.663).

For discriminant validity, the Italian OAI-23 adaptation and validation study¹⁹ used Pearson’s correlation test and found that the scale discriminated individuals according to sex and age, with women adapting less well than men (r = -0.125; P < .05) and older people adapting better than younger individuals living with an ostomy (r = 0.268; P < .01). Similarly, in the present study we were able to discriminate the patients according to age (constant = 54.072; age coefficient = 0.136) using linear regression.

Our results demonstrate the reliability and validity of the culturally adapted Brazilian Portuguese version of the OAI-23, representing the first and uniquely adapted and validated instrument for assessing the psychosocial adjustment of Brazilian individuals living with an ostomy. Being able to assess patient adjustment contributes to holistic and systematic care, especially in a population such as people with stomas. Our study also highlights the role of sample diversity because individuals from 4 Brazilian states were included, effectively increasing the cultural heterogeneity. 

Although considered culturally adapted and validated, some other psychometric properties of the OAI-23 should be analyzed, mainly responsiveness. In addition, the literature also recommends that at least 1 psychometric measure (eg, internal consistency, reliability) be evaluated according to the different target populations even if it was already previously culturally adapted and validated.^32–34

The primary limitation of this study was the use of a convenience sampling method. However, according to the authors’ review, the literature pertaining to sample randomization for this type of study is scarce and lacking.

This methodological study of cultural adaptation and validation of the OAI-23 for Brazilians concluded that the adapted version was reliable and valid for use in that country. The instrument displayed adequate internal consistency and stability and also fit the original model for reliability; content validity of convergent and discriminant constructs was also confirmed. To facilitate comparisons with international versions of the instrument, it was decided to keep the English acronym of OAI-23. Based on these results, the authors recommend that the culturally adapted version of the OAI-23 be applied in clinical practice throughout Brazil. This instrument will not only be useful to researchers and clinicians to assess the adjustment of people with ostomy, but it will also be of assistance when proposing strategies to improve stoma acceptance and support effective coping-related behaviors.

The authors thank Dr. Miako Kimura, Herica Duarte Dias, and Maria Elizete Nunes da Silva for their collaboration during the development of this study. The authors also sincerely thank Dr. Kingsley L. Simmons and Dr. Atsuko Maekawa for granting permission to adapt and validate the OAI-23.

Dr. Santos is a full professor in the Medical-Surgical Nursing Department, School of Nursing of the University of São Paulo, São Paulo, Brazil. Ms. Nascentes is a visiting professor in Specialization in Stomatherapy Nursing, State University of Rio de Janeiro, Rio de Janeiro, Brazil. Dr. Freitas is a professor in the Postgraduate Program in Nursing, Guarulhos University, Guarulhos, Brazil. Ms. Winckler de Oliveira is an enterostomal therapy nurse, Campo Grande, Brazil. Dr. Villela de Castro is a professor in the Graduation Program at the AC Camargo Cancer Center, São Paulo, Brazil. Address all correspondence to: Vera Lúcia Conceição de Gouveia Santos, RN, CETN, PhD, School of Nursing of the University of São Paulo, Av. Dr. Enéas de Carvalho Aguiar, 419 – Jardim Paulista - 05403-000 - São Paulo – SP, Brazil; tel: +55 11 3061-7544 / + 55 11 30617546; email: veras@usp.br