A study on the psychometric properties of the Arabic-translated connor-davidson resilience scale (CD-RISC-25) and Arabic-translated stimulant relapse risk scale (SRRS) in Emirati patients with substance use disorder

Abstract

Introduction

Substance Use Disorder (SUD) presents a significant global health concern. Resilience plays a crucial role in aiding recovery from mental disorders, and its importance in improving SUD treatment outcomes is increasingly recognized. Accurately predicting relapse risk can lead to personalized care, optimizing recovery outcomes. Therefore, culturally validated tools to measure resilience and relapse risk are essential prerequisites.

Aim

To assess the validity and reliability of the Arabic translations of the Connor and Davidson Resilience Scale (A-CD-RISC-25) and the Stimulant Relapse Risk Scale (A-SRRS). The correlation between A-CD-RISC-25 and A-SRRS has also been explored.

Methods

Translation from English to Arabic involved two bilingual mental health professionals. Adjustments were made during back-translation for consistency. 286 adults with SUD participated through consecutive sampling. Internal consistency was assessed using Cronbach’s alpha, and validity was evaluated through item-to-total correlations, item-item correlations, and Spearman’s correlations. Additional analyses using Mann-Whitney U and Kruskal-Wallis tests examined associations with demographic and clinical variables.

Results

Cronbach’s alpha values for A-CD-RISC-25 and A-SRRS were 0.904 and 0.901, respectively. For the A-CD-RISC-25, item-to-total correlations ranged from 0.19 to 0.62, indicating moderate to strong links with resilience. Item 20 showed a weaker correlation, and its removal increased Cronbach’s Alpha to 0.908. The overall internal consistency was excellent (Cronbach’s Alpha 0.898–0.908). Subscales showed strong consistency for Subscale 1 (0.804–0.827), moderate for Subscales 2 and 3, and weaker for Subscale 5. The A-SRRS showed good reliability, with item-to-total correlations ranging from 0.116 to 0.696, though Items 14 and 17 had negative correlations. Subscales 1–4 were reliable, while Subscale 5 showed poor reliability.

Conclusion

Arabic translations of the CD-RISC-25 and SRRS are valid, consistent, and correlate positively, underlining their combined utility in assessing resilience and predicting relapse in SUD patients in the Arabic population.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13033-026-00705-0.

Introduction

Substance Use Disorder (SUD) represents a significant global health burden, and its treatment necessitates patient-centred care that incorporates bio-psycho-social interventions [44]. As a chronic disorder, SUD is characterized by multiple relapses and a variable disease burden [31, 43]. Both broad societal factors and individual-specific (self) factors have been investigated for their influence on recovery, relapse prevention, and overall management. For instance, societal factors that have been studied encompass family involvement, employment status, housing conditions, societal stigma, support networks, prevailing social norms, the legal framework, and the presence and accessibility of effective relapse prevention programs [6, 19, 39]. On the individual front, explored factors include the type of substance abused, the age at which use began, duration of use, genetic predispositions, co-existing mental health disorders, and particular personality traits [6, 19, 39].

There is a growing body of evidence underscoring the significance of resilience in the face of adversity. In cases of depression, resilience has been shown to mitigate the severity of symptoms and bolster emotional regulation [8, 14, 32]. Resilience isn’t merely an attribute but a process that aids individuals in recovering from adversities [15]. It achieves this by enhancing aspects such as emotional coping, self-efficacy, life satisfaction, self-esteem, and spiritual intelligence [9, 15, 41, 45, 46, 47]. It’s worth noting that these interactions can be influenced by cultural variables. Thus, validated tools like the Connor-Davidson Resilience Scale (CD-RISC25) [11], originally developed in English, might necessitate cross-cultural adaptations. Notable alternatives include the Resilience Scale (RS) and the Scale of Protective Factors (SPF), each offering unique theoretical perspectives and measurement approaches [26]. The CD-RISC25 is widely valued for its strong reliability, validity, and cross-cultural applicability. It effectively predicts mental health outcomes and is sensitive to change, making it a versatile tool for clinical and research use [11, 28, 42].

Factors that constitute resilience include connections to supportive adults, cognitive skills, self-regulation, a positive self-image, among others [12, 17, 29, 30]. These elements interact with adversities to cultivate resilience, and this interaction can vary across cultural contexts [15, 17]. By assessing protective factors against adversity in specific populations, we can build more tailored interventions to enhance resilience.

Treatment for SUD often encounters challenges. In one study, 59.9% of participants relapsed after a successful treatment [20]. Factors contributing to relapse encompass negative social experiences, cravings, residence in areas with low socioeconomic status, sleep disturbances, family-related issues, and genetic predispositions [16, 20, 22, 23]; [27, 35]. On the other hand, having insight, social support, and employment have been shown to reduce the risk of relapse [38, 45].

It is hypothesized that predicting the risk of relapse can enhance the delivery of patient-centred care. The predisposing, precipitating, perpetuating, and protective factors of relapse have been examined using structured measures. Tools like the Drug Abuse Screening Test-20 (DAST-20) and the Alcohol Relapse Risk Scale (ARRS) assess relapse risk, but the Stimulant Relapse Risk Scale (SRRS) stands out for its strong validity and multidimensional reliability in evaluating stimulant relapse risk ([18, 33]; Yasukazu [34]). The SRRS is a validated and reliable scale developed in English to measure the risk of relapse associated with stimulants [33]. However, limited studies have explored the applicability of the SRRS to other substances or its translation into other languages.

The United Arab Emirates (UAE) is a cosmopolitan country situated in the lower Gulf region, predominantly inhabited by Muslims. It shares borders with the Kingdom of Saudi Arabia, the State of Qatar, and the Sultanate of Oman. Geographically positioned along international trafficking routes, the UAE is in proximity to countries that produce illicit substances. The UNODC estimates that 25% of global seizures are reported in the Kingdom of Saudi Arabia (KSA). Socially, the UAE community is cosmopolitan, fostering a spirit of diversity. Anthropologically, while embracing modernity, citizens of the UAE still uphold many of the cultural traditions and norms inherent to their tribal heritage. Arabic, the mother tongue of UAE citizens, serves as the official language [36, 37].

The UAE established the National Rehabilitation Centre (NRC) in 2002 to address SUD through a comprehensive, harm reduction-based approach. The NRC has become a key regional player and is recognized as a World Health Organization Collaborating Centre [2].To advance evidence-based practices and enhance treatment options for those with SUD, it’s imperative that researchers and clinicians have access to reliable measures for evaluating the effectiveness of various interventions. Precise assessment methods provide policymakers and healthcare administrators with the insights needed to more effectively allocate resources, as well as to craft targeted prevention and intervention programs based on the prevalence and severity of SUD in the community. There appears to be a lack of similar validation studies for the Stimulant Relapse Risk Scale in Arabic, unlike the Arabic CD-RISC-25, which has been validated in Jordan and Lebanon [3, 7]. However, due to dialectical and SUD context differences, validation in the UAE is necessary [1].

Furthermore, accurate assessment tools frame SUD as a medical condition rather than a moral failing, contributing to a reduction in stigma. This shift in perspective promotes broader access to treatment for those affected. Culturally sensitive and linguistically adapted assessment instruments are crucial in addressing disparities in care access and furthering global health equity. The objective of this study is to examine the Arabic psychometric properties of the SRRS and the CD-RISC-25 among a sample of UAE adults voluntarily seeking treatment for SUD at the NRC. The study’s aim is to validate these two scales for this specific sample, with the ultimate goal of integrating them into the clinical management of SUD.

Study methods

Study design

The translation and validation of the tools were done within a cross-sectional correlational study that gathered data from Arabic speaking participants who are diagnosed and being treated for SUD in the UAE.

Study setting and sample

Questionnaires were administered to all outpatients and inpatients at the NRC from June 1, 2022, to March 31, 2023, who were diagnosed with SUD. Invitations were extended individually and through posters at the OPD entrance. Around 40% of those invited participated, and a list of patient file numbers was kept to avoid duplication due to frequent follow-up appointments. The NRC serves the majority of SUD patients in the UAE, including those who have relapsed and others who have been discharged but continue to cope at home. These patients return regularly to the outpatient department (OPD) for routine check-ups. The centre stands as the primary referral hub for SUD patients throughout the UAE. In both the OPD and inpatient environments, the principal author oversaw participant recruitment and guided the completion of the questionnaires. Due to the lack of data on SUD prevalence among Emiratis, consecutive sampling was used, focusing on individuals at the NRC, the UAE’s primary SUD rehabilitation centre.

Data collection

Connor-davidson resilience scale (CD-RISC25) scale

The Connor-Davidson resilience scale (CD-RISC25) [11] is a validated 25-item self-administered measure designed to quantify resilience in adults with SUD. The CD-RISC-25 includes five subscales: personal competence, high standards, and tenacity (measuring goal-setting and perseverance), trust in one’s instinct and tolerance of negative affect (assessing emotional resilience), positive acceptance of change (evaluating adaptability), control (measuring emotional regulation), and spiritual influence (examining the role of spirituality in coping). Participants rate their responses to each item on a 4-point scale. Each item is scored over a 4-point Likert scale (“not true at all”: 0, “rarely true”: 1, “sometimes true”: 2, “often true”: 3, to “true nearly all of the time”: 4), with scores ranging from 0 to 100, higher scores indicating greater resilience. Cronbach’s alpha for the English version was 0.89 [11].

The stimulant relapse risk scale (SRRS)

The Stimulant Relapse Risk Scale (SRRS) is a validated 35-item scale. The SRRS consists of six subscales: anxiety and intention to use drugs (linking anxiety to relapse risk), emotionality problems (emotional difficulties contributing to relapse), compulsivity for drug use (compulsive urge to use substances), positive expectations (belief in positive outcomes from drug use), lack of negative expectancy (absence of concern for consequences), and insight into mental condition (awareness of mental health and substance use). It is rated on a three-point Likert scale: “strongly disagree and disagree” (1), “neither disagree nor agree” (2), and “agree and strongly agree” (3). with scores ranging from 35 to 105, higher scores suggest an increased risk of relapse. The Cronbach’s alpha for the English version stands at 0.86 [33].

Tools translation

The scales were initially translated from English to Arabic by two bilingual mental health professionals. Any discrepancies in their translations were reconciled to reach a consensus. This unified translation, now termed “version 1,” was then refined by a professional translator. Following this, two other bilingual mental health professionals, unfamiliar with the original text, back-translated version 1 to English, creating “version 2.” The similarity between version two and the original questionnaires were checked to ensure accurate translation. Finally, both versions were reviewed and adjusted for discrepancies by two different mental health professionals and another professional translator.

The comprehensibility and readability of the translated questionnaire were evaluated in a preliminary pilot study with 25 participants. A content validity questionnaire was developed, covering all items intended for the final version. Participants were asked to read each item, respond to it, and provide feedback regarding its clarity, readability, and the appropriateness of the wording used. Items were considered valid if they achieved a score above 3, as per established criteria. Based on feedback from the pilot study, two items from each scale were adapted to better align with the local dialect, ensuring increased clarity. Following these adjustments, the scales were administered to an additional 261 patients with SUD using a consecutive sampling technique over a 9-month period, resulting in a total of 286 participants. The original and validated version of the instrument is presented in Appendix 1.

Ethics statement

Approval for this study was granted by the NRC research ethics committee’s Institutional Review Board (IRB) (Reference number: #1_2022). The study adhered to local regulations and institutional guidelines. Written informed consent for participation was obtained from the principal author.

The analysis of data

Reliability analysis was conducted on the items of both the resilience scale and the relapse risk scales to assess internal consistency using Cronbach’s alpha [13]. For validity, relapse risk scores were compared with physician clinical diagnosis to assess the criterion-based validity of the A-SRRS (i.e. Concurrent validity or criterion-based validity). Items in the translated tools, were also correlated with other items in the scale (item-item correlation) and the total items in the scale (Item to total correlation). Finally, variables under consideration were also associated with participants’ demographic or clinical attributes, such as psychological comorbidities, classification status (e.g., relapse, lapse, abstinent), and employment status. The relationships were assessed using appropriate parametric and non-parametric statistical methods, including Spearman’s correlations, Mann-Whitney, and Kruskal-Wallis tests.

Relationships between the variables in this study were compared to those in similar studies from the literature. These studies utilized either the same tool or tools analogous in nature, examining the same variables. This comparison helped evaluate the validity of the Arabic version of the tool. Statistical analyses were performed using SPSS software, version 24. The data underwent screening, with missing entries excluded from percentage calculations and associated test probabilities. Items that were reversed were identified and subsequently recoded.

Results

Demographics of the participants

The majority of participants were aged between 29 and 34 years (34.6%), followed by those aged 24 to 28 years (26%). A significant proportion of participants were male (95.5%) and lived within a family unit (94.4%). When considering monthly income, 42.0% reported having none, while 26.9% indicated an income range of “10,001–20,000 AED”. In the context of employment, the most substantial group was those unemployed (50.3%), trailed by government employees (40.0%). In terms of educational attainment, over half of the participants (52.8%) had completed only high school, while 18.2% held a bachelor’s degree. Most participants did not report any medical co-morbidities like hypertension, diabetes, hepatitis, or HIV. However, a noTable 36.7% did have a history of psychological co-morbidities. Table 1 provides a comprehensive breakdown of the participants’ demographic characteristics.

Table 1.

Details the participants demographic characteristics

		Count	N %
Age Group	18–28	113	40%
	29–40	140	49%
	41–51	25	9%
	Above 51	8	3%
Gender	Male	273	96%
	Female	13	5%
Monthly income	No income	120	42%
	Less than 10,000	32	11%
	10,001–20,000	77	27%
	20,001–30,000	30	11%
	30,001–40,000	14	5%
	Above 40,000	13	5%
Marriage status	Single	158	55%
	Married	99	35%
	Divorced	27	9%
	Widow	2	1%
living alone or with family	Alone	16	6%
	With family	270	94%
Employment	Governmental	115	40%
	Private	17	6%
	Charity	3	1%
	Business	7	2%
	Not working	144	50%
Education	School education only	211	74%
	Diploma	15	5%
	Bachelor	52	18%
	Master or PhD	8	3%
Age when start the drug use		M = 19, SD = 7
Age of start of treatment		M = 27, SD = 8
Time since last drug use in days		M = 260, SD = 38.2

The average age for starting substance use was 19 years, with treatment typically beginning at 27 years. This indicates an average duration of 7.75 years between the commencement of substance use and the initiation of treatment. The span of days since the last drug use ranged from 1 to 1825 days. 49% of participants began using drugs before turning 18, and 61% started by or on their 18th birthday.

In terms of substance consumption, 13% of participants used only one substance, 78% used at least two, and 64% used three or more. Additionally, 58% consumed more than four substances, with nearly 51% consuming five or more.

On the resilience scale, participants achieved an average score of 72.92 out of a possible 100 (SD = 16.99). For the relapse risk, the mean score was 59.07 out of a potential 105 (SD = 12.23).

Resilience and relapse risk scale’s reliability testing

The Cronbach’s alpha stood at 0.904, suggesting a strong reliability index. For the subscales, all except for “subscale 5: Lack of negative expectancy” showed satisfactory results, with alpha values ranging from 0.622 to 0.854.

In a similar vein, the relapse risk scale had a Cronbach’s alpha of 0.901, again indicating a strong reliability index. All its subscales, except “subscale 5: Spiritual influences” demonstrated good reliability, with alpha values between 0.66 and 0.83.

Detailed Cronbach’s Alpha values for the subscales are presented in Table 2.

Table 2.

Subscales Cronbach Alpha values

Scale	Subscale	Mean (SD)	Cronbach’s alpha
CD-RISC-25	Subscale 1: Personal competence, high standards and tenacity	24.68 (6.40)	0.830
	Subscale 2: Trust in one’s instinct tolerance	17.96 (5.67)	0.712
	Subscale 3: Positive acceptance for change	14.76 (4.10)	0.690
	Subscale 4: Control	8.67 (2.92)	0.660
	Subscale 5: Spiritual influence	6.85 (1.59)	0.305
	Total	72.92 (16.99)	0.904
SRRS	Subscale 1: Anxiety and intention to use drugs	13.55 (3.52)	0.779
	Subscale 2: Emotionality problems	15.86 (3.66)	0.779
	Subscale 3: Compulsivity for drug use	4.94 (1.64)	0.787
	Subscale 4: Positive expectations	10.19 (3.63)	0.854
	Subscale 5: Lack of negative expectancy	6.57 (1.70)	0.323
	Subscale 6: Insight into mental condition	7.96 (2.28)	0.622
	Total	59.07 (12.23)	0.901

Construct based validity and internal consistency

The item-to-total correlations for A-CD-RISC-25, which range from 0.19 to 0.62, show that most items are moderately to strongly related to the overall resilience construct, though Item 20 exhibits a weaker connection. The scale demonstrates high internal consistency, with Cronbach’s Alpha values between 0.898 and 0.908 if an item is deleted. Notably, removing Item 20 slightly increases the overall reliability to 0.908, suggesting it may not align as closely with the rest of the scale. Overall, the A-CD-RISC-25 shows excellent reliability, with most items contributing positively to the resilience score. For more details see Table 3.

Table 3.

Item Characteristics, Item-Total Correlations, and Cronbach’s Alpha Values if an Item is Deleted for the Arabic Version of the CD-RISC-25 (N = 286)

Item no.	Mean (SD)	Item-total correlation	Cronbach’s alpha if item deleted
1	2.79 (1.25)	0.47	0.901
2	2.84 (1.36)	0.33	0.904
3	3.61 (0.89)	0.27	0.904
4	3.01 (1.17)	0.54	0.899
5	3.22 (1.11)	0.60	0.898
6	2.62 (1.26)	0.53	0.900
7	2.86 (1.31)	0.59	0.898
8	2.90 (1.24)	0.53	0.900
9	3.24 (1.17)	0.39	0.902
10	3.24 (1.07)	0.46	0.901
11	2.99 (1.13)	0.59	0.899
12	3.07 (1.25)	0.60	0.898
13	3.01 (1.24)	0.52	0.900
14	2.29 (1.48)	0.56	0.899
15	2.73 (1.22)	0.56	0.899
16	2.90 (1.28)	0.62	0.898
17	3.31 (1.09)	0.58	0.899
18	2.91 (1.21)	0.26	0.905
19	2.57 (1.35)	0.59	0.898
20	1.98 (1.50)	0.19	0.908
21	2.98 (1.25)	0.51	0.900
22	2.68 (1.28)	0.60	0.898
23	2.88 (1.26)	0.52	0.900
24	3.15 (1.09)	0.60	0.898
25	3.14 (1.28)	0.52	0.900

Item-to-subscale correlations show varying levels of internal consistency across the A-CD-RISC-25, as detailed in Table 4. Subscale 1 exhibits strong reliability, with item-to-subscale correlations ranging from 0.412 to 0.606 and Cronbach’s Alpha values between 0.804 and 0.827 if an item is deleted. Subscale 2 shows moderate reliability, though Item 20 stands out with a low correlation of 0.230, which increases the Alpha to 0.732 when removed, suggesting it may be less consistent with the subscale. Subscale 3 has moderate correlations (0.352 to 0.545) with Cronbach’s Alpha values between 0.598 and 0.685. Subscale 4 demonstrates fairly good internal consistency with correlations ranging from 0.462 to 0.479. However, Subscale 5, with low correlations of 0.180 for both items, indicates weaker reliability.

Table 4.

Item Characteristics, Item-Total Correlations, and Cronbach’s Alpha Values if an Item is Deleted for the Arabic Version of the CD-RISC-25 (N = 286)

Item No.	Mean (SD)	Item to subscale total correlation	Cronbach’s alpha if item deleted
Subscale 1
10	3.24 (1.07)	0.412	0.827
11	2.99 (1.13)	0.565	0.809
12	3.07 (1.25)	0.583	0.806
16	2.90 (1.28)	0.598	0.804
17	3.31 (1.09)	0.587	0.807
23	2.88 (1.26)	0.531	0.814
24	3.15 (1.09)	0.606	0.804
25	3.14 (1.28)	0.565	0.809
Subscale 2
6	2.62 (1.26)	0.507	0.659
7	2.86 (1.31)	0.398	0.686
14	2.29 (1.48)	0.511	0.655
15	2.73 (1.22)	0.535	0.654
18	2.91 (1.21)	0.298	0.708
19	2.57 (1.35)	0.520	0.654
20	1.98 (1.50)	0.230	0.732
Subscale 3
1	2.79 (1.25)	0.523	0.605
2	2.84 (1.36)	0.352	0.685
4	3.01 (1.17)	0.545	0.598
5	3.22 (1.11)	0.444	0.642
8	2.90 (1.24)	0.382	0.667
Subscale 4
13	3.01 (1.24)	0.472	0.562
21	2.98 (1.25)	0.462	0.576
22	2.68 (1.28)	0.479	0.554
Subscale 5
3	3.61 (0.89)	0.180
9	3.24 (1.17)	0.180

A univariate analysis was conducted to examine the correlations between the total resilience scores and the scores of its subscales. The scores for the total resilience scale showed a positive and moderate correlation with each of its subscales. Table 5 details the correlations between the total resilience score and its subscales.

Table 5.

Correlation between the resilience total and subscales

	Res-Total	Res-1	Res-2	Res-3	Res-4	Res-5
Res-Total	1	0.897**	.864**	.806**	.771**	.502**
Res-1	0.897**	1	0.656**	0.617**	0.687**	0.365**
Res-2	0.864**	0.656**	1	0.627**	0.562**	0.373**
Res-3	0.806**	0.617**	0.627**	1	0.509**	0.379**
Res-4	0.771**	0.687**	0.562**	0.509**	1	0.320**
Res-5	0.502**	0.365**	0.373**	0.379**	0.320**	1

**significant r value, Res: Resilience Subscale x

The analysis of item-to-total correlations for the SRRS reveals a range of relationships between individual items and the overall scale. Most items show moderate to strong correlations, with values ranging from 0.116 to 0.696, indicating that they generally contribute well to the overall measure of relapse risk. Items 14 and 17, however, exhibit negative correlations (-0.181 and − 0.119, respectively), which suggests they may not align well with the other items. Cronbach’s Alpha if an item is deleted ranges from 0.894 to 0.910, indicating that the overall scale’s reliability would only slightly change if any single item were removed. Notably, removing Item 14 would increase the Alpha to 0.910, suggesting that this item might detract from the scale’s internal consistency. Cronbach’s Alpha of 0.917 if both 14 and 17 removed. Overall, the SRRS demonstrates good reliability, though certain items may need further examination. For more details, see Table 6.

Table 6.

Item Characteristics, Item-Total Correlations, and Cronbach’s Alpha Values if an Item is Deleted for the Arabic version of the SRRS (N = 286)

Item No.	Mean (SD)	Item-Total Correlation	Cronbach’s Alpha if Item Deleted
1	2.23 (0.76)	0.480	0.898
2	2.16 (0.86)	0.638	0.895
3	1.69 (0.78)	0.464	0.898
4	1.82 (0.80)	0.163	0.903
5	2.13 (0.78)	0.320	0.901
6	2.33 (0.81)	0.337	0.901
7	2.04 (0.81)	0.566	0.896
8	1.34 (0.60)	0.558	0.897
9	1.43 (0.65)	0.257	0.901
10	1.70 (0.73)	0.247	0.902
11	1.29 (0.61)	0.369	0.900
12	1.22 (0.52)	0.261	0.901
13	1.54 (0.71)	0.578	0.897
14	1.86 (0.85)	-0.181	0.910
15	1.48 (0.64)	0.487	0.898
16	2.08 (0.82)	0.429	0.899
17	1.94 (0.82)	-0.119	0.908
18	1.64 (0.80)	0.564	0.897
19	2.45 (0.78)	0.316	0.901
20	1.73 (0.79)	0.696	0.894
21	1.34 (0.62)	0.116	0.903
22	1.44 (0.69)	0.664	0.895
23	1.77 (0.76)	0.466	0.898
24	1.56 (0.75)	0.666	0.895
25	2.02 (0.81)	0.383	0.900
26	1.84 (0.83)	0.612	0.896
27	1.39 (0.63)	0.601	0.897
28	1.26 (0.53)	0.540	0.898
29	1.73 (0.83)	0.631	0.895
30	1.65 (0.77)	0.633	0.895
31	1.15 (0.45)	0.378	0.900
32	1.87 (0.83)	0.629	0.895
33	1.34 (0.57)	0.641	0.896
34	1.19 (0.50)	0.549	0.898
35	1.44 (0.70)	0.530	0.897

The item-to-subscale correlations for the Stimulant Relapse Risk Scale (SRRS) reveal varying levels of internal consistency across the six subscales. Subscale 1 shows moderate internal consistency, with item-to-subscale correlations ranging from 0.234 to 0.638, and a Cronbach’s Alpha ranging from 0.724 to 0.787 if an item is deleted. Subscale 2 exhibits moderate correlations (0.247 to 0.586) with Cronbach’s Alpha ranging from 0.656 to 0.725. Subscale 3 shows stronger consistency, with correlations between 0.541 and 0.659 and Alpha values from 0.700 to 0.762. Subscale 4 demonstrates the highest internal consistency, with correlations from 0.537 to 0.694 and Alpha values between 0.820 and 0.849. Subscale 5, however, shows poor reliability, with low correlations (0.111 to 0.217) and very low Alpha values (0.198 to 0.322), indicating potential issues with these items. Subscale 6 shows moderate reliability with item-to-subscale correlations from 0.236 to 0.537 and Alpha values between 0.483 and 0.645. Overall, Subscales 1, 2, 3, and 4 demonstrate acceptable reliability, while Subscale 5 may require further revision. For more details, see Table 7.

Table 7.

Item Characteristics, Item-Total Correlations, and Cronbach’s Alpha Values if an Item is Deleted for the Arabic version of the SRRS (N = 286)

Item No.	Mean (SD)	Item to subscale Total Correlation	Cronbach’s Alpha if Item Deleted
Subscale 1
1	2.23 (0.76)	0.466	0.758
2	2.16 (0.86)	0.638	0.724
6	2.33 (0.81)	0.374	0.776
12	1.22 (0.52)	0.234	0.787
22	1.44 (0.69)	0.559	0.742
27	1.39 (0.63)	0.531	0.748
33	1.34 (0.57)	0.596	0.741
35	1.44 (0.70)	0.487	0.754
Subscale 2
3	1.69 (0.78)	0.341	0.709
5	2.13 (0.78)	0.353	0.706
7	2.04 (0.81)	0.586	0.656
10	1.70 (0.73)	0.247	0.725
16	2.08 (0.82)	0.522	0.670
19	2.45 (0.78)	0.402	0.696
23	1.77 (0.76)	0.450	0.687
25	2.02 (0.81)	0.402	0.697
Subscale 3
8	1.34 (0.60)	0.567	0.757
28	1.26 (0.53)	0.659	0.700
31	1.15 (0.45)	0.541	0.762
34	1.19 (0.50)	0.633	0.715
Subscale 4
18	1.64 (0.80)	0.537	0.849
20	1.73 (0.79)	0.639	0.830
24	1.56 (0.75)	0.659	0.827
29	1.73 (0.83)	0.694	0.820
30	1.65 (0.77)	0.678	0.823
32	1.87 (0.83)	0.643	0.830
Subscale 5
9	1.43 (0.65)	0.171	0.263
14	1.86 (0.85)	0.177	0.256
17	1.94 (0.82)	0.217	0.198
21	1.34 (0.62)	0.111	0.322
Subscale 6
4	1.82 (0.80)	0.236	0.645
11	1.29 (0.61)	0.245	0.624
13	1.54 (0.71)	0.537	0.483
15	1.48 (0.64)	0.480	0.523
26	1.84 (0.83)	0.424	0.543

The total score for relapse risk and scores for its subscales, with the exception of subscale 5 (Lack of Negative Expectancy), showed positive correlations that ranged from moderate to high. Table 8 presents the correlations between the total relapse risk score and its subscales.

Table 8.

Correlation between the relapse risk scale total and subscales ** Correlation is significant at the 0.01 level (2-tailed).

	RR-Total	RR-1	RR-2	RR-3	RR-4	RR-5	RR-6
RR-Total	1	.888**	.778**	.692**	.885**	.095	.767**
RR-1	0.888**	1	0.575**	0.654**	0.747**	− 0.027	0.658**
RR-2	0.778**	0.575**	1	0.426**	0.626**	− 0.071	0.428**
RR-3	0.692**	0.654**	0.426**	1	0.543**	− 0.049	0.474**
RR-4	0.885**	0.747**	0.626**	0.543**	1	− 0.057	0.651**
RR-5	0.095	− 0.027	− 0.071	− 0.049	− 0.057	1	0.043
RR-6	0.767**	0.658**	0.428**	0.474**	0.651**	0.043	1

Criterion based validity for A-SRRS

Comparing the relapse risk scores of patients clinically classified as abstinent, relapsed, or lapsed regarding drug use provides a valuable indication of the scale’s validity. Accordingly, the relapse risk scores of participants within these categories were assessed and compared. A notable distinction was observed in their relapse risk scores (Sig < 0.001). Specifically, participants who had experienced lapses or relapses scored significantly higher than those classified as newly admitted or fully abstinent as indicated in Table 9.

Table 9.

Kruskal-Wallis test results for relapse risk across relapse status groups * Statistically significant at p < 0.05.

Test	Variable	Count	H Statistic	Z or degrees of freedom	Mean rank	Significance (p-value)
Kruskal-wallis test: relapse risk comparisons	Relapse status	New: 75, Relapse: 91, Lapse: 18, Abstinent: 102	26.881	3	New: 125.41, Relapse: 177.91, Lapse: 163.19, Abstinent: 122.62	< 0.001*

Discussion

In Arabic culture, developing culturally appropriate assessment tools for substance use and mental health is crucial due to a lack of suitable resources. Reliable methods for assessing risk factors and treatment outcomes can improve diagnosis and tailored interventions. This study systematically translated the A-CD-RISC-25 and SRRS scales, following guidelines from Cha, Kim, & Erlen [10]. The translation process included professional bilingual translation, expert input, back translation, and team discussions to ensure cultural relevance, with the final version tested in a pilot study [4, 10].The resulting Cronbach’s alpha coefficient of the scales showed to be commendable, mirroring the internal consistency metrics found in prior research. In two comparable studies, where the A-CD-RISC25 was translated into Korean and Persian languages, both reported similarly high levels of reliability and validity across diverse samples [5, 21]. The reliability of the Arabic CD-RISC used in this study is comparable to findings from previous studies that translated and validated the CD-RISC into Arabic [7]. Similarly, the SRRS has been translated into both Chinese and Tagalog, with both versions demonstrating high levels of reliability and validity [18, 25]. This high level of internal consistency further supports the robustness of these scales when adapted for use in different cultural contexts. However, this is the first study to translate and validate the SRRS in Arabic.

In the A-CD-RISC-25 scale, items 3, 18, and 20 showed poor internal consistency, likely due to cultural and contextual factors. The minimum acceptable item-to-total correlation is typically 0.30 [40]. Item 3, “Sometimes fate or God can help,” may have been misinterpreted by participants in the UAE It is likely that, due to the presence of strong religious beliefs in an Islamic country, implying that God’s help is occasional may contradict participants’ beliefs (i.e., that God is always present and supportive), potentially leading to inconsistent responses. Items 18, “Make unpopular or difficult decisions,” and 20, “Have to act on a hunch,” part of the subscale on trusting instincts and handling stress, may have been too broad or interpreted differently, with some viewing these actions as impulsive rather than resilient.

In the A-SRRS scale, Items 14 and 17 exhibited negative correlations, while Items 9, 10, and 12 showed poor internal consistency. These results were consistent with previous research, including the original study on the development of the SRRS [25, 33]. Additionally, item 9 (“I feel easier than before”), item 12 (“Thinking about my family, I can no longer use the drug”), item 14 (“I am afraid of hallucinations due to drug use”), and item 17 (“I would not be able to control myself if I use the drug”) were reverse-scored items, intended to reduce response bias but potentially confusing. In UAE culture, traits associated with masculinity, such as self-control, are highly valued. Behaviours perceived as lacking control or “feeling easier” might be misinterpreted as weakness or drug effects. Hallucinations could be wrongly associated with mental illness. Such cultural stigma and misinterpretation feelings that deviate from these expectations can lead to unreliable responses in assessments. Items 9, 14, and 17 are part of the lack of negative expectancy subscale. There is limited evidence on how negative expectations influence substance use behaviour [24]. Items 10 and 12 are related to the emotionality problems subscale and the anxiety and intention to use drug subscale, respectively. Both emotional problems and anxiety are known to increase the risk of relapse.

Regarding the secondary objective of this study, which aimed to examine resilience and relapse levels among patients diagnosed with SUD those classified by physicians as relapsed or lapsed exhibited higher relapse risk. This finding also indicates that the SRRS has predictive ability in identifying individuals at higher risk of relapse. These correlation patterns support our hypotheses and provide strong evidence for the validity of these scales among UAE participants.

Data on resilience and relapse risk scores across different countries are generally limited and often outdated. Comparing these scores is complex due to cultural, social, and economic differences, as well as the varying prevalence of SUDs and the availability of treatment and care facilities. These factors can significantly influence average resilience and relapse risk scores across nations.

Countries with higher resilience and lower relapse risk scores may benefit from comprehensive healthcare systems, accessible mental health services, stronger support networks, or effective substance use disorder prevention programs. In contrast, countries with lower resilience and higher relapse risk scores might face challenges in these areas. It is crucial to evaluate these scores within the specific context of each country. In the UAE and neighbouring GCC and Arabic countries, where stigma and limited services are common, these scores may indicate the need for strategies to enhance resilience and reduce relapse risk among users.

Conclusion

To improve outcomes for individuals with SUDs in Arabic-speaking populations, it is essential to integrate culturally and linguistically adapted assessment tools into clinical and research settings. The Arabic versions of the CD-RISC and SRRS demonstrated acceptable reliability and can be practically applied to screen resilience levels and relapse risk, enabling more accurate diagnosis, treatment planning, and monitoring.

Healthcare providers can utilize these tools to identify patients at higher risk of relapse and tailor interventions accordingly. Moreover, implementing resilience-building programs and relapse prevention strategies should be prioritized within treatment centers. Policymakers and healthcare institutions are encouraged to invest in developing culturally sensitive services and capacity-building among professionals to better address the needs of individuals with SUDs. Future work should also explore how these tools perform across diverse Arabic-speaking populations and in longitudinal studies.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

IM: Writing – original draft, Conceptualization, Methodology, Investigation, Project administration, Resources, Visualization; NA: Writing – review & editing, Conceptualization, Formal analysis, Supervision, Methodology; SA: Writing – review & editing, Resources; FA: Writing – review & editing, Resources.

Funding

No financial support received for this work.

Data availability

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

Declarations

Ethics approval

An institutional Review Board (IRB) approval was obtained from the NRC research ethics committee in order to conduct this study (Reference number: #1_2022). The study was conducted in accordance with the local legislation and institutional requirements. Written informed consent for participation in this study was provided by the principal author.

Competing interests

The authors declare no competing interests.