Translation and validation of the Amsterdam preoperative anxiety and information scale (APAIS) in Serbia

Abstract Objectives Preoperative anxiety is common and might affect surgical treatment outcomes. The aim was to translate and validate the Serbian version of the Amsterdam Preoperative Anxiety and Information Scale (APAIS). Methods Following translation and initial evaluation, the Serbian version (S‐APAIS) was administered to 385 patients. Internal consistency, construct validity, prognostic criteria validity, and concurrent validity between S‐APAIS and Visual Analogue Scale for Anxiety (VAS‐A) were evaluated. Results Factor analysis revealed two factors: APAIS‐anesthesia (items 1, 2, 3) and APAIS‐procedure (items 4, 5, 6). The whole scale, APAIS‐anesthesia, and APAIS‐procedure subscales showed an adequate level of internal consistency (Cronbach's αs: 0.787, 0.806, and 0.805, respectively). High concurrent validity was observed between APAIS‐anesthesia and VAS‐A (ρ = 0.628, p < .001). A moderate correlation was found between APAIS‐procedure and VAS‐A scale (ρ = 0.537, p < .001). At the cut‐off point of 9, the area under the curve (AUC) of APAIS‐anesthesia was 0.815 (95% CI: 0.77–0.85, p < .001). For the APAIS‐procedure, AUC was 0.772 (95% CI: 0.73–0.81, p < .001) at the cut‐off point of 8. Conclusion The structure of S‐APAIS substantially differs from the original and allows separate measurement of anesthesia‐ and procedure‐related anxieties. S‐APAIS is a comprehensive, valid, and reliable instrument for the measurement of preoperative anxiety.


INTRODUCTION
Anxiety can be defined as an aversive feeling which arises from the anticipation of a potentially unfavorable, risky, or unpleasant event or outcome. It is characterized by negative emotional reactions and intense physical manifestations (Hyde et al., 2019). The concept of preoperative anxiety refers to an unpleasant feeling of worry in a patient undergoing surgical treatment. It is usually related to the perception of the forthcoming operation or anesthesia, pain, hospitalization, and disease itself. It is very common and can be seen in up to 94% of patients prior surgery (Aust et al., 2018). Besides the fact that preoperative anxiety is marked as by far the worst aspect of treatment by surgical patients (Walker et al., 2016), it can significantly affect surgical treatment outcomes (Williams et al., 2013). This justifies the need for routine preoperative assessment of anxiety levels.
Several tools are available for the identification of patients with preoperative anxiety. The Amsterdam Preoperative Anxiety and Information Scale (APAIS) was developed in 1996 by Moerman et al. (1996) and the original Dutch version has been translated and validated into many reliable versions in several languages (Maurice-Szamburski et al., 2013;Mohd et al., 2015;Vergara-Romero et al., 2017;Wu et al., 2020).
Nowadays, APAIS has gained widespread use, since contemporary literature offers strong evidence of its validity and reliability (Aust et al., 2018). Some authors even consider it as a "gold standard" for preoperative anxiety measurement (Eberhart et al., 2020). To date, the Serbian version does not exist. Therefore, the aim of the present study was to perform translation of the APAIS and to evaluate its validity and reliability for the Serbian population.

METHODS
The study was conducted during the period from February 1 to October 1, 2019. Consecutive patients who were scheduled for elective vascular surgical procedures (aortic, carotid, and surgery of peripheral arteries) during the eight-month period were enrolled in the study. Procedures were performed under both general and regional anesthesia. The following patients were excluded from the study: APAIS is a self-reported questionnaire, which can quickly assess the level of preoperative anxiety. The scale contains six questions, grouped into two components: anxiety subscale, which measures anxiety related to anesthesia and surgery (questions 1, 2, 4, and 5) and the need for information subscale (questions 3 and 6), assessing the desire for information regarding anesthesia and surgery. Questions are scored based on Likert's method from 1 ("not at all") to 5 ("extremely").
For the subscale related to the anxiety, the total score ranges from 4 to 20, while for the part of the scale related to the need for information the range is from 2 to 10. A higher score speaks in favor of a higher level of anxiety and a greater need for information. According to the part of the scale related to preoperative anxiety, a patient with a score ≥11 experiences anxiety; according to the part of the scale related to the need for information related to anesthesia and surgery, patients are classified into those who have little or no need for information (score 2-4), those who have an average (score 5-7) and those who have high information requirements (score 8-10) (Moerman et al., 1996).
VAS-A consists of a 100 mm long horizontal line, which is marked with 0 ("not anxious at all") at its left and with 100 ("extremely anxious") at its right end (Abend et al., 2014). The patients were asked to draw a vertical line and thus mark the level of their anxiety at the time of the interview. A line drawn closer to the right end indicates a greater degree of anxiety. Although the cut-off point is not precisely established (Stamenkovic et al., 2018), a VAS-A value > 70 mm correlates with a high level of anxiety (Hernandez et al., 2015).
Permission to translate the APAIS and to validate the Serbian version was obtained from the author of the original scale (N. Moerman). The translation and adaptation were performed according to the internationally accepted methodology for translation and crosscultural adaptation of questionnaires (Sousa & Rojjanasrirat, 2011) and included several steps. First, a "forward translation" (from English to Serbian) was performed by two independent professional translators.
Those two initial translations were merged into one by the members of the expert team (the lead investigator, an epidemiologist, two professional translators, and one clinician). This version was as close as possible to the original APAIS regarding its semantic and conceptual characteristics but at the same time the most appropriate for the Serbian cultural environment. This version was used for the "backward translation," from Serbian to English. Then, the expert team discussed this Internal consistency was assessed using Cronbach's α coefficient. The scale was considered reliable if Cronbach's α coefficient was > 0.7. To examine whether the original APAIS scale maintains its factor structure in the data gathered among the Serbian population, a confirmatory factor analysis (CFA) was performed. An acceptable model fit was based on the following thresholds of fit indices: chi-square-degrees of freedom ratio (χ 2 /DF) lower than 2, the root mean square error of approximation (RMSEA) lower than 0.08, the goodness fit index (GFI) higher than 0.9, the adjusted GFI (AGFI) higher than 0.9, normed fit index (NFI) higher than 0.9, and comparative fit index (CFI) higher than 0.9 (Hu & Bentler, 1999). Subsequently, to determine a viable factor structure, an exploratory factor analysis (EFA) was conducted, using the principal component analysis and Varimax rotation method. The appropriateness of data for EFA was assessed using Kaiser-Meyer-Olkin's measure of sampling adequacy and Bartlett's test of sphericity.

RESULTS
The final Serbian version of the APAIS is shown in   (Table 3) Item-total analysis showed that "Cronbach's α if item deleted" ranged from 0.742 to 0.774 and no correlation value less than 0.30 (Table 4).

DISCUSSION
The main finding of the present study is that the structure of the Serbian version substantially differs from the original APAIS scale. As proposed by Moerman et al. (1996), the original APAIS is characterized by two sub-dimensions: (1) scale related to anxiety, with items 1, 2, 4, and 5, and (2)  were also present in our study sample. This might have led to the lack of basic knowledge and information regarding surgical processes and anesthetic techniques and contributed to the modified S-APAIS structure. Also, unlike surgery, which has always been far more appreciated, anesthesia has always been underestimated and marginalized among the Serbian population, perhaps due to more abstract and unfamiliar perspective. This might explain the strict separation of feelings related to surgery and anesthesia and, consequently, a specific structure of S-

APAIS.
This is the first study that reports the process of translation of the APAIS scale into Serbian and its validation among the Serbian population. Results of the present study indicate that the Serbian version of the APAIS scale possesses good psychometric properties. Namely, S-APAIS shows an adequate level of internal consistency: Cronbach's α coefficient = 0.787. This value is comparable with those reported by other authors (Mohd et al., 2015;Vergara-Romero et al., 2017;Zeleníková et al., 2017). In the original APAIS study (Moerman et al., 1996), Cronbach's α coefficient was calculated for the two subscales separately and not for the whole scale and since the S-APAIS structure differs from the original scale significantly, those values cannot be compared. In addition, two subscales of the present study-APAISanesthesia and APAIS-procedure, are also characterized by a high level of internal consistency (Cronbach's α = 0.806 and 0.805, respectively), suggesting that those anxieties can be measured separately.
Although different authors consider other scales as "gold standard" for the measurement of preoperative anxiety level (Goebel & Mehdorn, 2018;Lemos et al., 2019;Tulloch & Rubin, 2019;Vergara-Romero et al., 2017), VAS-A is as efficient and reliable and can be used for both research and clinical purposes (Hernandez et al., 2015;Homzová & Zeleníková, 2015;Kindler et al., 2000). Results of the present study demonstrate strong and moderate correlations between two S-APAIS subscales-APAIS-anesthesia and APAIS-procedure with VAS-A (ρ = 0.628, p < .001 and ρ = 0.537, p < .001, respectively). This association describes the external validity of S-APAIS, suggesting that it can be effectively used for the assessment of preoperative anxiety.
While data regarding the incidence of preoperative anxiety are inconsistent, a vast majority of authors agree that it is very common.
In fact, a recent cross-sectional study by Aust et al. that included over 3000 subjects showed that only 6% of patients do not feel anxious at all during the perioperative period (Aust et al., 2018). Our results show that preoperative anxiety can be seen in over 40% of patients, which is in concordance with the findings of the other authors (Kuzminskaitė et al., 2019;Nigussie et al., 2014;Woldegerima et al., 2018).
Results of prognostic criteria validity testing show that patients who score more than 9 points on the APAIS-anesthesia subscale have 63% chance of experiencing anesthesia-related anxiety. Patients with a score ≤9 will not be anxious about anesthesia in 83% of cases. Patients will experience anxiety related to the procedure if they score more than 8 points on the APAIS-procedure subscale with a probability of 72% and if this score is ≤8, the chance of them not presenting surgeryrelated anxiety is 69%. These results are approximate, but not identical to the ones in the study by Bakalaki et al. (2017), who reported the same two-factor structure of the modified APAIS scale. Specifically, cut-off points in that study were 6.5 and 8.5 for the APAIS-anesthesia and APAIS-procedure subscales, respectively. These differences might partially be explained by smaller sample size, different gender structure (fewer men), younger subjects, and the fact that patients underwent various surgical procedures in the Greek study. The present study included solely vascular surgical patients, so the percentage of "major surgeries"-defined as more severe ones by Bakalaki et al. (2017), is considerably higher, which might also contribute to differences in cutoff values.
Finally, besides the fact that S-APAIS is characterized by good psychometric properties, results of the initial evaluation pilot study show that this questionnaire is brief, understandable, and practical. Since there was no significant misinterpretation, lack of understanding, or need for further explanation of the questions, this finding also supports the application of S-APAIS in preoperative anxiety assessment in the Serbian population.
Limitations of the present study are reflected in some characteristics of the patients. The sample is not homogeneous in terms of gender structure: there were almost four times fewer females than men.
This could have led to lower anxiety incidence since the female gender is known as a risk factor for preoperative anxiety (Gonçalves et al., 2016;Laufenberg-Feldmann & Kappis, 2013). Also, patients in our study were mostly older, unequally distributed according to surgery type, and operated on in a single center. The effects of surgery type, anesthesia techniques, previous anesthetic/surgery experiences, and other clinical characteristics of patients on anxiety level were not taken into account and will be of interest in our future research. The fact that we did not use the State-Trait Anxiety Inventory (STAI) to assess concurrent validity as most of the previous studies did, may also be considered a weakness of the present study. Still, a higher cut-off point on the VAS-A scale (> 70 mm) for the detection of anxiety cases should compensate for this limitation.
Based on the results of the present study, the Serbian version of APAIS is a valid, reliable, easy to administer, and well-accepted questionnaire that can effectively be used in the assessment of preoperative anxiety. The structure of the Serbian version substantially differs from the original and has two subscales that can clearly distinguish between anxiety related to anesthesia and anxiety related to surgery.
In that manner, the practical value of the S-APAIS scale can be seen: anesthesia-and procedure-related anxieties can be observed and measured separately and selected patients might be addressed individually by the surgeon or anesthetist, thus providing a higher level of care for the patient and lowering the incidence of preoperative anxiety.

ACKNOWLEDGMENTS
We sincerely appreciate Nelly Moerman, PhD MA for allowing us to translate and adapt the original version of the APAIS scale.
We are also thankful to the doctors from the Anesthesia Depart-

CONFLICT OF INTEREST
The authors declare that they have no conflict of interest.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from the corresponding author upon reasonable request.