Nowadays, there are no doubts regarding the role of the implantable cardioverter defibrillator (ICD) for prevention of sudden cardiac death, especially among patients with ventricular dysfunction and arrhythmogenic genetic diseases.^1-3 Due to its proven efficacy in identifying and correcting potentially lethal ventricular tachyarrhythmias, the number of ICD implantations has increased significantly worldwide, and more than 250,000 procedures are performed every year.⁴

The primary purpose of ICD is to correct potentially fatal ventricular arrhythmias by delivering low- or high-energy therapy. Low-energy therapy, known as antitachycardia pacing or antitachycardia pacing (ATP), is a painless method. High-energy therapy delivers shocks of up to 40 J which, in spite of causing major discomfort, usually occur after the patient has lost consciousness, since they are applied about 15 seconds after the initiation of ventricular fibrillation or fast ventricular tachyarrhythmia. In undesirable situations, such as arrhythmias resistant to overstimulation, or in electrical storm, high-energy discharges can occur in awake patients.^3,5,6

It is estimated that the chances of ICD patients will need appropriate electric shocks for primary prevention of sudden cardiac death varies between 2 and 15% per year.^5-8 On the other hand, when the ICD is used for secondary prevention, the incidence of shock therapies may vary between 35 and 53%, within the first year after implantation.^5-8 Despite the high level of technological sophistication of ICDs, unfortunately, there is the risk that the patient may receive inappropriate shock deliveries as a result of erroneous discrimination between supraventricular and ventricular tachyarrhythmias. On these occasions, the sensation reported is a painful and distressing experience.^9-14

ICD patients live with the expectation that, at any moment, the device will deliver shock therapies to interrupt ventricular arrhythmias resulting from their heart disease. Thus, although they recognize the benefits of the treatment, some patients may present with anxiety, depression, mood disorders, post-traumatic stress disorder, as well as fear that the device will not operate in crucial situations.^9-14 On the other hand, ICD implantation has been reported to provide the patient with a great sense of safety, considering the device’s capacity to interrupt unexpected episodes of potentially fatal ventricular arrhythmias.^10-14

In face of the concern about the deleterious effects of ICD on psychosocial adaptation, an scale was specifically developed to assess the level of anxiety related to the presence of ICD and to the shocks delivered by the device, for use both in clinical practice and in the context of scientific research.^15,16 The Florida Shock Anxiety Scale (FSAS) quickly achieved wide international acceptance, and has been translated and validated in several countries (Netherlands,¹⁷ Denmark,¹⁸Poland,¹⁹ China,²⁰ Norway,²¹ Turkey²²), with consistent results.

The purpose of the present study was to assess the psychometric properties of the Brazilian version of the FSAS for ICD patients.

The FSAS was developed in 2006 in the United States to provide a quantitative measure of ICD shock-related anxiety. The instrument consists of 10 items, with five response options (“not at all”,”rarely”,”some of the time”,”most of the time”,”all the time”), corresponding to a 5-point Likert scale.^15,16

Questions are related to patients’ fear or anxiety caused by the expectation that the device may deliver shock therapies and to the behavioral changes (not engaging in physical exercise or in sexual activity, or not getting angry or upset, for instance) to avoid the occurrence of ICD therapies.

The FSAS total score is determined by the sum of all items, with a maximum score of 50 points. The higher the score, the higher the anxiety level. The items receiving three points or more should be considered the most critical aspects.^17,18The instrument can be self-administered or administered by interview.

The cross-cultural adaptation process of the FSAS followed international guidelines and included five stages: (1) Translation by two independent translators; (2) Synthesis of the translations; (3) Back-translation; (4) Harmonization of the translations by the expert committee; (5) Pretest with the target population; (6) Pretest review and final translation.^23-25The expert committee was composed by professionals of the area of artificial cardiac stimulation and by nurses.

The translation of the original instrument into Portuguese was performed by two independent Brazilian translators, proficient both in the Portuguese and English languages. In this stage, the translations produced by the two independent translators (T₁ and T₂) were reconciled into one version (T₁-₂), after discussion with the expert committee. Working from the final version and blind to the original version, two bilingual teachers carried out the back-translations (RT₁ and RT₂). The aim of this stage was to measure the semantic and idiomatic consistency of the translations produced in the first stage. Finally, a new meeting was held with the expert committee to review all the cross-cultural adaptation process and undertake the harmonization across versions, thus obtaining the pre-final version of the instrument (Figure 1).

Figure 1

The pretest version was administered in a convenience sample of 20 ICD patients, aged between 18 and 80 years, All patients were recruited during an outpatient cardiovascular clinic appointment, and were at least 6 months postimplant. The pretest was conducted to identify and correct possible translation problems. Following self-completion of the instrument, a clarifying interview was held to verify the existence of irrelevant or hardly understandable items, as well as to measure the understanding of each item of the instrument. It was established that the translation would be reviewed or reformulated if less than 80% of the participants were able to understand the items.

The final Portuguese version of the FSAS was administered to a convenience sample of 151 participants with the following characteristics: (1) adults, aged between 18 and 80 years, of both sexes and with any education level; (2) ICD implanted for more than 6 months; (3) capable of understanding and answering the questionnaire used in the study; (4) having agreed to participate in the study by signing the informed consent form. We did not include in the study patients presenting at least one of the following situations: (1) indication for cardiac transplantation; (2) ongoing pregnancy; (3) malignant neoplasia.

Patients were selected consecutively, during outpatient care or by visits to the inpatient unit of our institution. Individuals who met the eligibility criteria were invited to answer the FSAS questionnaire. At the same time, demographic, clinical and ICD data were collected by using electronic case report forms developed in REDCap²⁶ (Research Electronic Data Capture) hosted at the hospital’s server.

Through descriptive analysis of the instrument items, it was possible to identify that normality of distribution was violated, indicating, therefore, the need for polychoric correlation, instead of Pearson’s correlation coefficient.

The means of the instrument items ranged from 1.5 to 2.9. The FSAS average score was 22.8 ± 11.1, with a median of 20 points and variation of 10 to 50 points. There was no impact of extreme values on the mean (Table 3).

Table 3

Item	Average	SD	Inferior threshold	Superior threshold	5 % trimmed average	Median	Range	IQR	Asymmetry	Kurtosis	KS	Sig.
1	2.95	1.86	2.66	3.25	2.95	3.00	4.00	4.00	0.12	-4.80	0.29	0.01
2	2.46	1.72	2.19	2.74	2.40	1.00	4.00	4.00	2.84	-3.67	0.33	0.01
3	2.26	1.69	1,99	2,53	2,18	1,00	4,00	3,00	4,06	-2,90	0,37	0,01
4	2,47	1,69	2,20	2,74	2,41	1,00	4,00	3,00	2,63	-3,69	0,33	0,01
5	2,43	1,62	2,17	2,69	2,37	2,00	4.00	3.00	2.97	-3.21	0.30	0.01
6	1.54	1.25	1.34	1.74	1.38	1.00	4.00	0.00	10.81	7.67	0.48	0.01
7	2.36	1.68	2.09	2.63	2.29	1.00	4.00	3.00	3.34	-3.30	0.34	0.01
8	2.74	1.72	2.47	3.02	2.71	3.00	4.00	4.00	1.30	-4.14	0.28	0.01
9	2.07	1.59	1.81	2.32	1.96	1.00	4.00	2.00	5.37	-1.56	0.39	0.01
10	1.54	1.24	1.34	1.74	1.37	1.00	4.00	0.00	10.87	7.84	0.49	0.01

The values obtained from the Kaiser-Meyer-Olkin index (KMO= 0.88), the Bartlett’s sphericity test (X²= 565.5, df= 45; p<0.001) and the matrix determinant (0.0206 (p<0.0001)) revealed a significant correlation between the items, which confirmed the adequacy of the EFA.

The parallel analysis indicated the existence of only one dimension for the instrument. Moreover, this item set can explain 64.4% of latent variable (above the values recommended in the literature).^29-31 The eigenvalue criteria also indicated only one dimension, with a an eigenvalue of of 6.08. The fact that the instrument was unidimensional waived requirements for methods of matrix factor rotation. Unidimensionality indicated the use of the normal-ogive graded response IRT model, which is more adequate for a unidimensional polytomous model.³¹

Table 4 presents the factor loads, which ranged from 0.67 to 0.86, representing excellent levels of adherence of the items to the latent variable, greater than the minimum criterion of 0.50, with no evidence of multicollinearity. The unidimensionality ruled out the possibility of cross-loading. Communalities varied between 0.46 and 0.74, with all the items above the threshold of 0.40. For item discrimination (a), the values ranged from 0.91 to 1.71, also indicating good adherence to the latent variable and corroborating the data obtained from factor loading.

Table 4

Item	Factor loading	Communalities (h2)	Item description (a)
1	0.76	0.58	1.17
2	0.77	0.60	1.22
3	0.76	0.59	1.19
4	0.81	0.65	1.37
5	0.68	0.46	0.93
6	0.71	0.50	1.00
7	0.67	0.46	0.91
8	0.73	0.53	1.05
9	0.86	0.74	1.71
10	0.74	0.55	1.11

The CFA revealed good adjustment to the unidimensional model, with values similar to those recommended by the literature: Robust Mean and Variance-Adjusted Chi Square X²/ df (35) = 40.40; p < 0.243; NNFI= 0.997; CFI= 0,997; GFI= 0.986; AGFI= 0.982. The residual indicators were at good levels (RMSEA= 0.032; RMSR= 0.077), showing little difference between the original matrix and the matrix generated from factor loadings.³¹

Satisfactory reliability evidence was provided by the FSAS-Br scale, with a Cronbach’s alpha coefficient of 0.92, a McDonald’s Omega coefficient of 0.92 and GLB of 0.98.

In the present study, we described the translation and cross-cultural adaptation process of a brief scale designed to provide a quantitative measure of ICD shock-related anxiety, folllowing international methodological standards.^23-25 The final translation of the FSAS into Brazilian Portuguese (FSAS-Br) presented conceptual, semantic, cultural and measurement equivalences compared to the original items in English.^15,16

Efforts were made to include patients with different sociodemographic profiles and various types of underlying heart diseases to ensure heterogeneous representation, aiming at providing the best calibration of the items. Thus, patients with different ICD types (ventricular, atrioventricular or associated with cardiac resynchronization therapy) were included, as well as patients with indications for primary or secondary prophylaxis of sudden cardiac death. Notwithstanding, the most common kinds of heart disease among these patients’ profiles have also been contemplated, with expressive prevalence of Chagas Disease, ischemic and hypertrophic cardiomyopathy.

In the international scenario, the FSAS scale has been widely used in different scenarios, since it presents good sensitivity to identify the level of ICD shock-related anxiety and requires reduced time for completion.^15-22 Thus, it is important to highlight that the FSAS was not designed to assess relevant aspects of adaptation to the device and its real impact on quality of life, which makes it necessary to use other instruments to complement the assessment of these patients. In this sense, the authors who had created the FSAS developed another instrument, the Florida Patient Acceptance Survey (FPAS),³² which aims at assessing the psychosocial adjustment of ICD patients. The results of the cross-cultural adaptation and validation process of the FPAS into Portuguese will be published in due course.

Evidence of validity of an instrument has been recommended by the scientific community as a way to check whether the instrument actually and accurately measures the latent variable of interest. In addition, it is important to analyze whether the instrument factor structure is adequately represented by its dimensionality, that is, the number of dimensions that make up the instrument of assessment.^27-31 In the original publication of the FSAS, the authors claim that the instrument was bidimensional, presenting two dimensions: Consequence (composed of 7 items) and Trigger (composed of 3 items).¹⁵This model was not reproducible to the Brazilian version, because all analyses performed in this study supported the FSAS-Br scale unidimensionality. Revisiting the study by Kuhl et at.,¹⁵ it is important to highlight that the sample was constituted by only 72 participants, which may have had an impact on the results of the psychometric analyses.

Afterwards, the psychometric properties of the FSAS were evaluated, with a sample of 443 participants.¹⁶ The CFA showed that the two previously identified dimensions were highly related to a second-order factor (“Shock anxiety”). In other words, the two dimensions identified previously could have been better explained by their association to a common factor, namely the “shock-related anxiety” dimension. Due to these results, the authors recommended that the total scale score may be more clinically useful, instead of subdividing it into the two dimensios described before. These results corroborate the factor structure identified in our study.

Reliability assessment of the FSAS-Br scale revealed the accuracy of the Brazilian version, which was confirmed by adequate values of Cronbach’s alpha, McDonald’s Omega and GLB. The adoption of these three indications aimed to increase the accuracy of interpretation, since the Cronbach’s coefficient alpha is affected by the nature of data distribution and by sample size. Besides, its values may be increased by extensive scales, parallel and/or redundant elements or limited coverage of the construct under analysis, decreasing the reliability of the measurement.³³

In general, the results observed in the present study showed that the instrument is reliable and valid for application in Brazil, meeting the quality requirements for patient-reported outcome measurements.

The FSAS-Br instrument presented consistent validity and reliability evidence and, therefore, its use can be recommended for the ICD population in Brazil, both in clinical practice and in scientific research.