Pain is a serious problem that affects millions of children and adolescents around the world, with a considerable impact on quality of life.1,2 The emotional component of pain plays an important role in its regulation and development. The emotional factors involved in pain have been studied extensively, and there is evidence that their modification and treatment can reduce pain, thereby decreasing its impact on daily life and improving quality of life.3–6

Pain-related anxiety refers to the thoughts, feelings, behaviours and physical sensations that accompany the experience and anticipation of pain.7 It manifests through (a) cognitive responses to pain, which can lead to difficulty performing cognitive tasks such as concentrating or paying attention; (b) emotional reactions, the most frequent of which is fear of experiencing sensations associated with pain or its intensification; (c) physiological responses in the form of somatic symptoms, such as increased muscle tension, heart rate or breathing rate; and (d) behavioural responses by which efforts are made to reduce or avoid pain. Anxiety is a psychological factor that, in addition to being present in chronic pain, also plays an essential role in stressful acute situations such as surgeries or painful procedures, in which it has been found to have the potential to affect patient outcomes and recovery.6,8

A growing body of research has not only yielded knowledge about these factors, but also tools for their diagnosis and assessment.9 When it comes to pain-related anxiety, in 2010, Pagé et al.10 developed the Child Pain Anxiety Symptoms Scale (CPASS) by modifying the 20-item Pain Anxiety Symptoms Scale (PASS-20) for adults. This initial psychometric study10 was followed by a second one in a sample of surgical patients, and the scale is now widely used for evaluating pain-related anxiety in children and adolescents.11 In Spain, the 25-item Patient Reported Outcomes Measurement Information System (PROMIS) paediatric self-report scale,12 which includes a specific subscale on anxiety, is one of the scales frequently used to assess paediatric pain-related anxiety.13 However, this subscale has only 4 items, so it does not assess the different factors that may contribute to pain-related anxiety comprehensively. In contrast, the CPASS includes 4 different domains aimed at assessing pain-related anxiety from different perspectives. Although its use is widespread, to our knowledge it has not been translated and adapted into Spanish. A Spanish version could increase and improve our knowledge of pain-related anxiety, and therefore of the role of psychological factors in paediatric and adolescent pain.

The aims of this study were to translate the CPASS to Spanish and to evaluate its psychometric characteristics (construct validity, internal consistency, floor/ceiling effects and convergent validity) in a paediatric sample in Spain.

The study was approved by the Ethics Committee of the universitary hospital 12 de Octubre and adhered to the principles of the Declaration of Helsinki. The article was written according to the Standards for Reporting Diagnostic Accuracy Studies (STARD).14

Between January and March of 2021, 190 potential participants were invited to participate in the study (Supplementary Document III, flow chart of sample selection). A total of 160 children and adolescents were included in the study, of who 49.37% were female (81 male, 79 female), with a mean age of 14.5 years (standard deviation [SD], 2.3). Table 1 presents the anthropometric and sociodemographic characteristics of the children and the parents/guardians that participated in the study.

Exploratory factor analysis

Before performing the exploratory factor analysis, we calculated the Cronbach α coefficient for the whole scale (α = .88) and the adjusted item-total correlations (average item-total correlation, 0.51). No items were removed, given that all contributed substantially to the scale. The KMO test showed that the data were suitable for factor analysis (KMO = .87), and the Bartlett test rejected the identity matrix null hypothesis: χ2 (190) = 1168.40 (P < .001). Based on these results, it was justified to proceed with the EFA. Applying the Kaiser criteria, we kept 4 factors, which accounted for 57.61% of the variance (Fig. 1). However, the preliminary EFA revealed the presence of 2 problematic items. Specifically, item 18 had a factor loading of less than 0.40, and item 19 had a significant and similar factor loading in 2 factors other than the factor initially expected based on the theoretical model (0.41 for the fear construct; 0.40 for the escape/avoidance construct). For these reasons, we decided to eliminate both items from the model. The resulting final 4-factor model (18 items) fit the data well. Table 2 presents the goodness-of-fit indices of the model and the factor loadings of each item.

Figure 1.

Scree plot.

(0.06MB).

Table 2.

Final solution of the exploratory factor analysis.

Goodness of fit
χ2 = 111.90, P = .040; CFI = 0.97; TLI = 0.95; SRMR = 0.03; RMSEA = 0.04; 95% CI 0.01–0.06
Item	Fear	Physiological Anxiety	Cognitive	Escape/Avoidance
1. Pienso que si mi dolor es demasiado intenso, nunca mejorará.	0.63*	–	–	–
I think that if my pain hurts too much, it will never get better.
2. Cuando tengo dolor, tengo miedo de que algo horrible pase.	0.61*	–	–	–
When I feel pain, I am afraid that something terrible will happen.
3. Cuando tengo mucho dolor, descanso tan pronto como puedo.	–	–	–	0.49*
I rest right away when my pain hurts too much.
4. Mi cuerpo empieza a temblar cuando hago alguna actividad que empeora mi dolor.	–	0.69*	–	–
My body starts to shake when I am doing an activity that makes my pain worse.
5. No puedo pensar con claridad cuando tengo dolor.	–	–	0.69*	–
I can’t think straight or think clearly when I feel pain.
6. Paro cualquier actividad cuando empiezo a sentir dolor.	–	–	–	0.70*
I will stop any activity when I start feeling pain.
7. Cuando tengo dolor, mi corazón late más rápido.	–	0.65*	–	–
When I feel pain, my heart beats faster.
8. Tan pronto como el dolor empieza, pido a mis padres que me den medicación.	–	–	–	0.54*
As soon as pain begins, I ask my parents for medication.
9. Cuando tengo dolor, pienso que podría estar muy enfermo.	0.62*	–	–	–
When I feel pain I think I might be really sick.
10. Cuando tengo dolor me es muy difícil pensar en cualquier otra cosa.	–	–	0.51*	–
When I feel pain, it is hard for me to think about anything else.
11. Cuando tengo dolor, no hago algunas actividades importantes.	–	–	–	0.45*
I don’t do important activities when I hurt.
12. Cuando tengo dolor, me siento mareado y/o débil.	–	0.52*	–	–
When I feel pain, I feel dizzy or faint.
13. Tener dolor me asusta mucho.	0.53*	–	–	–
Feeling pain is very scary.
14. Cuando tengo dolor, pienso en él todo el rato.	–	–	0.66*	–
When I feel pain, I think about it all the time.
15. Cuando tengo dolor, siento ganas de vomitar.	–	0.59*	–	–
When I feel pain, I feel like I am going to throw up.
16. Cuando me duele mucho, pienso que no seré capaz de volver a moverme.	0.63*	–	–	–
When my pain hurts too much I think I might not be able to move again.
17. Encuentro difícil concentrarme y prestar atención cuando tengo dolor.	–	–	0.47*	–
I find it hard to concentrate and pay attention when I feel pain.
18. Me resulta difícil relajar mi cuerpo después de sentir dolor.	–	–	–	–
I find it hard to relax my body after I feel pain.
19. Me preocupo cuando tengo dolor.	–	–	–	–
I worry when I feel pain.
20. Intento no hacer aquellas actividades que me hacen sentir dolor.	–	–	–	0.55*
I try not to do activities that make me feel pain.

CFI, comparative fit index; CI, confidence interval; RMSEA, root mean square error of approximation; RMSR, standardised root mean square residual; TLI, Tucker-Lewis index.

*

P < .05.

Confirmatory factor analysis

We performed a CFA of the original version and of the scale with the deleted items (see Table 3). The final 18-item version (without items 18 and 19) of the CPASS was the best fit, with all items fitting in the expected hypothetical construct and showing optimal factor loadings (>0.50). In addition, because there was a moderate-high correlation between the 4 CPASS factors (fear, physiological anxiety, cognitive, and escape/avoidance: r = 0.46–0.72; P < .001), we added a higher-order factor for a single unifying construct named pain anxiety. The higher-order 18-item 4-factor solution was a good fit for the data; thus, it was established in the final model: χ2 (131) = 150.56, P = .116; CFI = 0.97; TLI = 0.97; RMSEA = 0.03, 95% CI 0.001–0.05; SRMR = 0.06. Fig. 2 shows the standardised factor loadings of this final higher-order 18-item 4-factor solution.

Table 3.

Model fit in the confirmatory analysis of the CPASS.

Model	χ2 (df)	CFI	TLI	RMSEA	SRMR
20-item	207.50 (164)	0.94	0.93	0.04 (0.02−0.06)	0.06
19-item (without item 18)	180.47 (146)	0.95	0.94	0.04 (0.02−0.057)	0.06
18-item (without items 18 and 19)	144.66 (129)	0.98	0.97	0.03 (0.001−0.05)	0.05

CFI, comparative fit index; df, degrees of freedom; RMSEA, root mean square error of approximation; RMSR, standardised root mean square residual; TLI, Tucker-Lewis index.

Figure 2.

Structural equation modeling of the final Spanish version of the Child Pain Anxiety Symptoms Scale (CPASS).

(0.23MB).

The objective of this study was to translate the CPASS into Spanish and evaluate its psychometric properties. The results showed that the Spanish CPASS is a valid and reliable scale that can be used in children and adolescents.

First, the results of the EFA indicated problems with 2 items. Specifically, item 18 (“I find it hard to relax my body after I feel pain”) had a low factor loading, so we excluded it from the final model. A possible reason is that although the construct was explained to the patients, they may not have understood it correctly or clearly identified this sensation. Children may have had problems identifying this sensation because their anxiety may have made it difficult to relax, especially when in pain,29 and because they may express pain anxiety in the form of fearful or avoidant behaviours, as described in previous studies.10,16 As for item 19 (“I worry when I feel pain”), we found that it loaded similarly in 2 dimensions (fear and escape/avoidance) instead of on the cognitive dimension, as reported in the psychometric studies of the original scale10 and the Catalan version.16 Based on the findings of previous studies10,16 and our own, it seems that children tend to construe worry as akin to fear rather than as a cognitive response, probably because they respond to fear with avoidant behaviour, as previously described in the literatury.30,31 Although the difference between both factors was small (fear = 0.41, escape/avoidance = 0.40), we also chose to discard this item from the final model. The CFA corroborated that the 18-item, 4-factor model was the one that offered the best fit. The 4-factor model is the most widespread in the literature in both paediatric and adult populations, and it is also the one that has shown improvement in assessing pain-related anxiety.10,11,16,32,33

The internal consistency of the Spanish CPASS was acceptable (Cronbach α, 0.88), in agreement with the findings of previous studies that analysed the psychometric properties of the scale in other populations (original version10: Cronbach α, 0.90; Catalan version16: Cronbach α, 0.87; and study in a sample of surgical patients11: Cronbach α, 0.91). When it came to the 4 dimensions of the Spanish CPASS, we found adequate values (internal consistency ≥ 0.72). Previous versions also exhibited an acceptable internal consistency for the dimensions, except for the physiological anxiety and escape/avoidance constructs, for which the obtained Cronbach α values were near but inferior to 0.70 (physiological anxiety, α = 0.6810,11,16; escape/avoidance, α = 0.67).11,16 These slight differences could be explained by the better fit of the model with the elimination of items 18 and 19. Nevertheless, the data provide strong evidence that the Spanish CPASS is a reliable instrument that can be used in the paediatric and adolescent populations. The Spanish CPASS also showed an adequate discriminatory capacity since no floor or ceiling effects were identified.

In the convergent validity analysis, the CPASS is strongly correlated to pain catastrophising. Specifically, in our study, the cognitive dimension was strongly correlated to pain-related catastrophising. The relationship between pain-related anxiety and catastrophising can be explained as anticipatory cognitions and psychological feedback constructs.7,10 On the other hand, pain-related anxiety was moderately associated with HRQoL and with the pain interference reported by children. Pain-related anxiety probably amplifies the impact of pain5 because it produces discomfort in children, reducing their quality of life6 and limiting their activities.11,34 Similarly, based on the observed weak correlation between anxiety and pain intensity, the impact of pain could increase or produce high levels of anxiety related to pain.10 As expected, however, anxiety had a greater impact on pain-related behaviours than on pain intensity. This is probably explained by the fact that anxiety, as an emotion/sensation closely related to fear and restlessness, promotes avoidant behaviours (eg, limiting the level of activity).6,35,36 Our results corroborate the relationship between fear and anxiety, given that the fear dimension was the one with the strongest correlation to pain-related variables. Lastly, we did not find an association between pain-related anxiety and the number of visits or of previous surgeries. In the sample under study, 65% of the participants had not undergone a previous surgery, a very high percentage, which would indicate that the development of anxiety related to pain would not be just an innate personality trait and could be mediated by the learning process of pain or by parental influences (for example, parental anxiety can be transmitted to children).37,38

In conclusion, the Spanish CPASS was a valid and reliable tool to assess pain anxiety in a clinical sample of children and adolescents.

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

The authors have no conflicts of interest to declare.

CB conceived and designed the study, designed the data collection instruments, collected data, carried out the analyses and drafted and revised the manuscript.

MC and LDUV conceived and designed the study, coordinated and supervised data collection, and critically reviewed and revised the manuscript, contributing important intellectual content.

GF critically reviewed and revised the manuscript, contributing important intellectual content.