Sleep hygiene promotion by nurses in adult psychiatric inpatient care: anonymized dataset
Description
This dataset contains the anonymized responses of a cross-sectional questionnaire survey describing how nurses promote sleep hygiene in adult psychiatric inpatient care, and comparing the practices of day-shift and night-shift nurses. The study tested two hypotheses: (1) that sleep hygiene promotion is only partially implemented by nurses, and (2) that practices differ between day- and night-shift teams. Data were collected between 20 January and 13 April 2025 in three French public hospitals located in two regions, using a 14-item self-administered questionnaire (pre-tested with five nurses). Registered nurses working either day or night shifts in full-time adult psychiatric inpatient units were eligible; nurses working interchangeably on both shifts were excluded. Participants answered on paper or online (Google Forms). The dataset is provided as a single table (128 rows, one per participant; 40 variables) in CSV and XLSX formats, accompanied by a codebook describing every variable. Categorical variables (sex, years of experience, work shift, unit type, centre, training, self-rated knowledge, frequency of advice given, interest in a support tool) are coded with readable labels. Multiple-answer questions (methods used to identify needs, tools/approaches, recommendations given, and perceived barriers) are coded as binary variables (1 = selected, 0 = not selected). To protect participant confidentiality, submission timestamps were removed, hospital names were recoded as Centre A/B/C, the free-text unit field was grouped into categories, and all free-text answers (including "other" specifications) were removed. What the data show: the transmission of sleep hygiene recommendations was generally low (most nurses had advised only one to five patients over the previous four weeks, and 13% none), and lack of theoretical knowledge was the most frequently reported barrier. Several clinically relevant recommendations (smoking, alcohol, heavy meals) were seldom given. Night-shift nurses reported better knowledge, transmitted recommendations more often, gathered patients' medical history more frequently, and more often advised stress management and CBT-I, whereas day-shift nurses more often reported a lack of time. How to interpret and use the data: the dataset reproduces the descriptive statistics and the day/night comparisons reported in the associated article. Frequencies and percentages can be computed on the categorical and binary variables; between-group (day vs night) comparisons can be reproduced using the chi-square test (expected counts ≥ 5) or Fisher's exact test otherwise, with p ≤ 0.05 considered significant. Note that the published Table 1 further split the "Other" unit category into Rehabilitation (n = 7) and Other (n = 8) from manual review of free-text answers, which is not reproducible from the coded data alone. Refer to the codebook for the full definition and coding of each variable.
Files
Steps to reproduce
Steps to reproduce Data collection A 14-item self-administered questionnaire on sleep hygiene promotion was designed and pre-tested with five nurses (no changes required). Administrative and nurse-management authorisations were obtained in three French public hospitals. The questionnaire was distributed to registered nurses working either day or night shifts in full-time adult psychiatric inpatient units, on paper and online (Google Forms), from 20 January to 13 April 2025. Nurses working interchangeably on day and night shifts were not included. Responses were compiled into a single spreadsheet (one row per respondent). Of 133 questionnaires collected, five were excluded (two not meeting the inclusion criteria, three meeting the exclusion criterion), leaving 128 participants. Data preparation (anonymisation and coding) 5. Identifying information was removed: submission timestamps deleted, hospital names recoded as Centre A/B/C, the free-text unit field grouped into categories (Admission, Long-stay, Addiction, Other), and all free-text answers (including "other" specifications) removed. 6. Single-answer questions were coded with readable labels; multiple-answer questions were coded as binary variables (1 = selected, 0 = not selected). Variable definitions are given in the accompanying codebook. Reproducing the results 7. Open HYSPSY_dataset_anonymised.csv (or .xlsx). Each row is one nurse; each column is one variable (see codebook). 8. Compute descriptive statistics: frequencies and percentages (denominator N = 128) for every variable. For binary items, the count of 1s divided by 128 gives the percentage reported in the article. 9. Compare day- and night-shift nurses using the "shift" variable: build a 2×k contingency table for each item and apply the chi-square test when all expected counts are ≥ 5, otherwise Fisher's exact test. A p-value ≤ 0.05 indicates a statistically significant difference. (In the original study, calculations used Microsoft Excel and the online tools at biostatgv.sentiweb.com; any statistical software, e.g. R or Python, will reproduce the same results.) Notes Percentages may not total 100 because of rounding or because several items allowed multiple answers. The published Table 1 further split the "Other" unit category into Rehabilitation (n = 7) and Other (n = 8) based on manual review of the original free-text answers; this finer split cannot be reproduced from the coded data alone.