Translation and validation of the Dutch version of the Fear of Cancer Recurrence Inventory (FCRI-NL)

Highlights

• The FCRI-NL has acceptable psychometric properties.

• The multi-dimensional structure of the FCRI-NL was not convincingly replicated in the current study.

• Caution is recommended with using and interpreting the FCRI-NL total score.

• The FCRI-SF-NL is recommended for screening for fear of cancer recurrence.

Abstract

Objective

The study objectives are to translate the FCRI in Dutch, and to explore the factor structure and the psychometric qualities of the Dutch translation of the Fear of Cancer Recurrence Inventory (FCRI-NL).

Method

The original French-Canadian FCRI had been forward-backward translated into English by the developers, and this method was also used to translate the English version of the FCRI into Dutch. Patients were recruited via patient organizations between July 2011 and October 2013. To replicate the original 7-factor structure of the FCRI, confirmatory factor analysis (CFA) was performed. To examine the psychometric qualities, reliability (Cronbach's alpha), test-retest reliability (intra-class correlations; ICC), and convergent and divergent validity (Spearman's correlations) were calculated.

Results

From 290 cancer patients, 255 (88%) were eligible for analysis (aged 51.0 ± 9.8 years, 88.6% women). CFA showed a reasonable yet suboptimal fit of the hypothesized model to the data. The FCRI-NL has good reliability (Cronbach's α = 0.93 for the total scale and α = 0.75–0.92 for the subscales) and test-retest reliability (ICC = 0.84 for the total scale and ICC = 0.56–0.87 for the subscales). Convergent (r = 0.53–0.66 for the FCRI-NL and r = 0.48–0.57 for the FCRI-SF-NL) and divergent (r = − 0.20–− 0.07 for the FCRI-NL and r = − 0.28–− 0.17 for the FCRI-SF-NL) validity was demonstrated.

Conclusion

The FCRI-NL seems to have sufficient psychometric properties. However, the FCRI-NL total score should be interpreted with caution. The Severity subscale (FCRI-SF-NL) may be a valuable screening tool for fear of cancer recurrence severity in clinical care.

Introduction

Fear of cancer recurrence (FCR) is one of the most reported long-term consequences of surviving cancer [1]. Elevated levels of FCR represent a continuing problem in cancer patients, for ten years or more after diagnosis [2], [3], [4]. In 2015, the International Expert Special Interest Group on FCR (FORwaRdS) redefined FCR as “fear, worry, or concern about cancer returning or progressing”, which is broader than previous definitions and more suitable for all types and stages of cancer [5]. This definition also shows that FCR ranges from normal or healthy levels of concerns about cancer recurrence to clinical levels of FCR [5]. Across different cancer types, 39–97% of cancer survivors reported some degree of FCR, 22–87% reported moderate to high degree of FCR and 0–15% reported a high degree of FCR [6]. A growing body of research recognizes FCR as a multi-dimensional construct, including intrusive thoughts, physical sensations, psychological distress, coping strategies, and functioning impairments [7], [8], [9].

Currently, at least 32 assessment instruments for FCR have been developed [6], [10]. Of these, 11 are subscales of comprehensive measures, 17 are brief questionnaires (2–10 items), and four are longer (multi-dimensional) questionnaires (10 + items) [6], [10]. However, many of these scales have limited generalizability due to limited psychometric data and/or have been used in few studies [6], [10]. Other limitations are that measures are only available in one language and/or are cancer site specific, which impedes comparison across cancer populations [6], [10]. Currently, the Fear of Cancer Recurrence Inventory (FCRI) is one of the best measures available [10].

The FCRI was developed by a committee of experts in psycho-oncology, based on their definition of FCR, on DSM-IV diagnostic criteria of anxiety and somatoform disorders and on a cognitive-behavioral conceptualization of FCR inspired by the model developed by Lee-Jones et al. [8], [11]. The original version of the FCRI was developed in French-Canadian and contains 42 items measuring seven factors (Triggers, Severity, Psychological Distress, Coping Strategies, Functioning Impairments, Insight, and Reassurance) [8]. In the initial article it was stated that a higher score on the total scale indicates higher levels of FCR [8]. The FCRI was validated in a sample of 600 participants who had been treated for breast, prostate, lung, or colorectal cancer, and it demonstrated good reliability and validity (reliability Cronbach's α = 0.95; test-retest reliability r (287) = 0.89, p < 0.001) [8]. The psychometric properties of the English version of the FCRI were similar to those of the original French-Canadian version (reliability Cronbach's α = 0.96; test-retest reliability r (135) = 0.88, p < 0.001; ICC = 0.94, p < 0.001) [7]. Simard and Savard recommend more validation studies in other cultures and languages [8]. Recently, an empirically validated cut-off score was determined for the Severity subscale of the FCRI, making it usable as a short form of the FCRI (FCRI-SF) to screen for clinical levels of FCR [12]. The Severity subscale is strongly associated with the FCRI total score and seems the most accurately representation of FCR severity, the other subscales represent related aspects, such as antecedents (e.g., Triggers), modifiers (e.g., Coping Strategies), or consequences (e.g. Functioning Impairments), which give important clinical information about FCR [12], [13]. There is some disagreement about the interpretation of the FCRI. While the authors of the original version of the FCRI recommend to measure FCR with the total FCRI scale or use the Severity subscale for screening purposes, Costa and colleagues recommend separate interpretation of the subscale scores [8], [14].

FORwaRdS recently reported a lack of translations of FCR measures and stated that we need cross-culturally validated measures [15]. To make the FCRI available in Dutch speaking countries, the FCRI was translated into Dutch (FCRI-NL). We will refer to the Dutch translation of the FCRI-SF as FCRI-SF-NL. The present study describes the translation and validation of the FCRI-NL (i) by performing a confirmatory factor analysis (CFA) to investigate whether the multidimensional seven factor structure of the original FCRI could be confirmed and (ii) by assessing internal consistency, test-retest reliability, and convergent and divergent validity of the FCRI-NL and its subscales. We hypothesized that confirmatory analysis would show a good fit and that the FCRI-NL would demonstrate good psychometric qualities.