Nursing Intervention in the Patient Being Discharged From the Intensive Care Unit

Method An experimental pre-test-post-test design will be used. Setting and participants The study will be carried out in the ICU of three university hospitals which have 151 level 3 ICU beds and where the nurse/patient ratio is 1:2. The three ICUs admit approximately 2000 patients per year. Participants who meet the following inclusion criteria will be included: age > 18 years; ICU stay > 48h; speaker of one of the two official languages of the region and carrying out a personal interview, Glasgow Coma Score of 15, Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) negative, and provision of signed informed consent. Patients with a pre-admission diagnosis of mental illness will be excluded.

Patients will be selected by simple random probabilistic sampling. Patients will be recruited through the ICU admission list and randomly assigned to the experimental group (EG) and the control group (CG). To calculate the sample number needed, on the basis of a statistical power of 80% and a level of confidence of 95% in a unilateral analysis, and estimating the percentage of patients who have high anxiety at ICU discharge at 50% (between 30% and 75% according to reports), a sample of 86 patients per group will be needed (assuming 15% losses per group) to detect a difference of >20% as statistically significant (expected to be 0.5 for the CG and 0.28 for the EG).

Nursing empowerment intervention The intervention will be carried out during the ICU discharge planning. A "Patient Information Guide for Intensive Care Unit patients" will be developed.

The guide will be based on the findings obtained in a previous qualitative study and on previous studies on patient empowerment and education. The guide structure will take into account the dimensions of empowerment: bio-physiological, functional, social, experience, ethics and the financial aspect. The guide will be reviewed by experts in the development of information guides for patients and family members and will be evaluated by the participating ICUs before use.

EG patients will receive the intervention verbally and in booklet format from the study nurses at each ICU. CG patients will receive the usual nursing treatment provided during ICU discharge, which usually consists of verbal information about the time and place of destination.

Variables:

1. Independent variable: Nursing empowerment intervention administered to the patient during ICU discharge.
2. Dependent variable: Level of anxiety and depression (HADS scale) pre-post ICU discharge.
3. Sociodemographic variables: sex (male/female), age (in years), educational level (none, primary, secondary, higher), profession, family situation (lives alone, with family or friends) and marital status (single, married or cohabiting, widower, divorced).
4. Clinical variables: severity level (APACHE II), reason for ICU admission and days of stay in the ICU.

Data will be collected using a first contact with the patient in the ICU, and the study objectives and informed consent will be explained.

The researcher will administer the HADS scale to the EG and CG groups before ICU discharge and after ICU discharge (in the destination ward).

Instruments:

1. Glasgow Coma Score: The Glasgow scale is designed to assess the level of consciousness, and includes the assessment of the ocular, verbal and motor response. It is simple, objective and easily reproducible.
2. The Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) adapted and translated to Spanish by Tobar et al. (2010), which is recommended and validated in the detection of delirium. The four criteria that characterize delirium are evaluated using four questions for each area (change in mental status, inattention, altered consciousness and disorganized thinking).
3. Hospital Anxiety and Depression Scale (HADS): Adapted and validated in Spanish by Tejero, Guimera, Farré and Peri (1986) it comprises two subscales of 7 items each, one for anxiety and one for depression. The intensity/frequency of symptoms is evaluated on a four-point Likert scale. For each subscale, the score ranges from 0 to 21, with the following cut-off points: 0-7 Normal; 8-10 Doubtful/possible clinical problem; >10 Likely clinical problem. The HADs measures the current status during the last week. Designed to be administered to subjects with organic diseases, the physical aspects that may accompany anxiety or depression have been eliminated, leaving the emotional aspects ones. In addition to the quantitative evaluation, it may be stratified into non-anxious or non-depressed if the HADS scores are < 8; possible or doubtful between 8 and 10, and likely or affirmative if they are ≥ 11. The authors consider that, for research purposes, if a low proportion of false positives are to be obtained, the upper limit with values of 10 or 11 should be used, while values of 8 or 9 would prevent false negatives. For this study, a value of 10 was decided on as a cut-off point between the presence or absence of anxiety and depression.
4. Acute Physiology and Chronic Health Evaluation (APACHE) II: To measure and evaluate the severity of the critically ill, Knaus produced in 1981 the first version, comprising 34 items, which were reduced reducing to 14 in the second version in 1985. APACHE II classifies disease severity using 12 usual physiological measures, the previous health status and age, with a score ranging from 0 to 71 points.

Data analysis The data will be analysed using the SPSS v23 statistical package. A descriptive analysis of the study variables will be made (means and standard deviation and medians and 25-75 percentiles (IQR) for quantitative variables according to the normality of distribution, and frequencies and percentages for qualitative variables). The possible relationship between sociodemographic and clinical variables and the level of anxiety and depression will be analysed using Pearson's correlation coefficient, the Student's t test and ANOVA or their nonparametric equivalents based on the type of variable and its distribution and the number of subgroups compared. The chi square test will be used to compare the variation in categorical variables between the EG and CG. The Mann-Whitney test or the Wilcoxon test will be used to compare the mean anxiety scale score between EG and CG. Statistical significance will be established as p ≤ 0.05.