Cross-cultural Adaptation and Psychometric Assessment of the Freedman Sleep Questionnaire for Intensive Care Unit Patients in Morocco

3.1. Study Design

This psychometric study was conducted in the ICU of three hospitals in the Souss-Massa region, Morocco. It took place in the multi-purpose ICUs (26 beds) of the three hospitals. The study aimed to assess the reliability, stability, and conceptual and content validity of the tool.

3.2. Participants

The study included patients admitted to the ICU between May 15, 2024, and December 10, 2024. A convenience sample was used, following the recommendations of the Streiner guide, which were applied to estimate the sample size (18). The minimum number of subjects required is at least 140, considering an intra-class correlation coefficient of approximately 0.70 and a precision level of 0.10. The study included patients who met the following criteria: Critically ill patients admitted to the ICU who were over 18 years old, Arabic-speaking, and provided informed consent. Patients with hearing or speech impairments, pre-existing dementia diagnoses, documented substance use disorders, or Glasgow Coma Scale (GCS) scores below 15 were excluded from the study.

3.3. Instrument

The subjective sleep quality of critically ill patients is evaluated using the FSQ, which considers environmental factors. The original questionnaire was developed by Freedman et al. (19) using a 10-point Likert scale to assess sleep quality. On this scale, 1 represents 'poor sleep quality' and 10 corresponds to 'excellent sleep quality'. A value of 1 denotes "inability to stay awake", while 10 represents "completely alert and awake". A value of 10 indicates a "significant interruption", while a value of 1 means "no interruption" in terms of environmental conditions.

3.4. Procedure

This study received ethical approval from the Biomedical Research Ethics Committee (CERB) of Rabat’s Faculty of Medicine and Pharmacy (approval No. 85/24). All procedures were conducted in accordance with institutional ethical standards and the principles outlined in the Declaration of Helsinki (20). Authorization for data collection was obtained from the regional Department of Health and Social Protection in the Souss-Massa region. The original questionnaire developers and copyright holders authorized the cultural adaptation and linguistic validation of the FSQ in Arabic for Morocco. The tool used in this study was the FSQ, which subjectively assesses sleep quality in critically ill patients, considering environmental factors.

3.5. Translation and Cultural Adaptation of the Arabic for Morocco Version of the Freedman Sleep Questionnaire

The translation of the FSQ aimed to provide an Arabic for Morocco version of the questionnaire that maintained conceptual equivalence with the original English version. Dr. Nil Freedman, the original developer of the FSQ, was consulted, and permission was granted to translate the questionnaire. The AM-FSQ was developed according to established guidelines (20). Initially, two healthcare professionals specializing in intensive care, both fluent in English, independently translated the FSQ from English to Arabic for Morocco. These two versions were then compared and consolidated by an ICU expert with over 15 years of experience in critical care. A linguist and a sleep specialist, fluent in Arabic and English, subsequently performed a back-translation into English. A panel of eight bilingual healthcare professionals, all actively involved in the study, reviewed the six versions of the AM-FSQ. The translation process continued until all discrepancies were resolved, ensuring precision and accuracy. Finally, cognitive debriefing was conducted with four participants (four ICU physicians and one sleep specialist) to verify that each question of the AM-FSQ was clearly understood by all participants (Figure 1). The degree of convergence of the experts’ answers during the final validation was measured by the Aiken coefficient, which showed a very high level of agreement (V = 0.78). The validated version of the AM-FSQ was tested on 14 intensive care patients. Participants reported no ambiguity in their understanding of the questions. The final version of the AM-FSQ was not re-adjusted after testing. Data were collected at three time points: Upon ICU admission (baseline), at the midpoint of the ICU stay, and on the day of ICU discharge.

Figure 1.

The process of translating and adapting the Arabic for Morocco version of the Freedman Sleep Questionnaire (AM-FSQ)

3.6. Statistical Analysis

A descriptive analysis was performed for all study variables. Qualitative variables were expressed as frequencies and percentages, while quantitative variables were presented as means ± standard deviations. Content validity was assessed through expert consensus, with the Content Validity Index (CVI) calculated using Davis’ method (21). Questionnaire reliability was evaluated through internal consistency (Cronbach’s alpha coefficient) and temporal stability [intraclass correlation coefficient (ICC) in a test-retest approach] (22, 23). Construct validity was examined through an exploratory factor analysis with an Oblimin rotation (loading threshold ≥ 0.4), which identified four dimensions: Sleep quality, daytime sleepiness, environmentally induced sleep disturbances, and human factor-induced sleep disturbances. This analysis was preceded by adequacy tests: Kaiser-Meyer-Olkin (KMO) measure (acceptable ≥ 0.5; good ≥ 0.75) and Bartlett’s sphericity test (P < 0.05 required) (24). All analyses were conducted using JAMOVI software (version 2.3.28), with statistical significance set at P ≤ 0.05.

4.1. Socio-demographic and Clinical Characteristics of Study Participants

Based on the inclusion criteria, 260 records of ICU patients were valid and included in the analysis. Their average age was 49.6 ± 18.42 years (range, 25 - 85); males constituted 46.6% of the sample. The mean length of ICU stay was 5.20 ± 2.47 days, with an observed range of 2 to 18 days. Regarding the cause of admission to the ICU, 72.2% were for medical conditions, 27.3% for postoperative care, 44.2% for acute heart disease, and 26.9% were related to respiratory diseases. Twenty percent of patients were smokers. Concerning clinical scores, the mean pain score was 3.79 ± 3.18, the Charlson comorbidities score was 1.02 ± 1.34, and the APACHE-II score was 9.60 ± 5.87 (Table 1).

4.2. Validity

Experts were consulted to determine the definition and relevance of each variable, thereby assessing the content validity of the questionnaire. The fact that most respondents answered 'more than 3' to the question indicates that this definition is accurate. Expert consensus was used to evaluate the content validity. Since "bed telephones" and "televisions" are not available in the ICU in Morocco, the assessments of these items were substituted with "personal telephone" and "conversation", respectively, by expert recommendations. Since these items were not present in intensive care units (ICUs), they were considered irrelevant. Following these modifications, the CVI of the Davis technique, which was previously determined to be 0.88, demonstrated a high level of content validity (25). The content validity analysis resulted in a final revised questionnaire, to which the responses were subsequently subjected during data collection. The KMO value was 0.845, and Bartlett’s test of sphericity indicated adequate factorability of the data [χ² (351) = 8515, P < 0.001]. Our sample satisfies assumptions for factor analysis by varimax rotation (Table 2). The Oblimin rotation revealed a four-factor structure accounting for 70.8% of the total variance in the measured constructs (Table 2). Individual factor loadings ranged from 0.40 to 0.99 across all questionnaire items (Table 3).

4.3. Reliability

5.1. Conclusions

The Moroccan version of the Freedman questionnaire shows excellent reliability and validity for assessing ICU sleep quality in Morocco. Its simple administration makes it practical for routine clinical use by ICU teams to identify sleep disturbances early and determine the factors affecting patients’ sleep. We suggest that the AM-FSQ be applied in multicenter and longitudinal studies to evaluate its responsiveness to interventions, thereby providing further insights into its utility in assessing sleep quality improvements across different clinical contexts. Implementing targeted interventions to promote healthy sleep and improve overall sleep quality in ICU patients may be beneficial. Additionally, it can be utilized in clinical research to assess sleep quality in correlation with other variables, such as depression, anxiety, and quality of life, in ICU settings.

5.2. Strengths and Limitations

The researchers implemented several strategies to minimize bias throughout this study. The translation process rigorously followed established protocols and underwent multiple rounds of expert review, significantly strengthening the validity of the translated version. A single investigator carried out data collection, which helped mitigate the risk of performance bias. Additionally, the diverse range of diagnoses within the sample improved the generalizability of the results and reduced the potential for selection bias. Despite the study’s findings, it is essential to recognize that questionnaire-based assessments are inherently subjective and can only serve as a substitute for objective evaluation methods when the latter are not feasible.

The study adopted a convenience sampling approach, wherein patients were selected based on their availability and willingness to participate. While this method is efficient, it may introduce potential selection bias. To minimize this bias, the sample was diverse in terms of age, pathologies, and severity of illness. A comparison between the characteristics of participants and non-participants revealed no significant differences. Moreover, the participation rate exceeded 94% of the target population, which further mitigates the risk of selection bias.

The exclusion of patients with a GCS score of less than 15 was a deliberate decision made to enhance the validity of the collected data. By excluding these patients, we aimed to ensure that the participants included in the study could provide clear and valid responses, thereby ensuring the reliability of the psychometric evaluation of the Arabic version of the AM-FSQ. While this exclusion limits the generalizability of the findings to more critically ill patients, it was necessary to maintain the internal validity of the study. Future research should focus on recruiting a larger and more diverse sample from multiple centers to explore further the correlations between patient demographics, clinical characteristics, perceptions of sleep quality, and the underlying factors contributing to poor sleep in critically ill patients.