Intervention to Improve Quality of Sleep of Palliative Patient Carers in the Community: Clinical Trial

Introduction Sleeping problems experienced by cancer patients carers are frequent with 72% of carers reporting moderate to severe problems. Another consequence for this carers is the increase of suffering from depression and it is usually associated to carer overload. Objectives Main objective: \- To evaluate the effects of listening to music in sleep quality for oncology patients non-professional carers at home, and to evaluate the influence of specific factors that may modify that effect. Secondary objectives: * To assess the relationship between the waking state and the consequences during the day according to sleep characteristics. These include: 1\. Psychological results: 1. Quality of life. 2. Carer overload. 2\. Physical results: 1. Sleepiness during the day. 2. Physical activity, intensity and length evaluation. * To assess carers satisfaction with the intervention. Method Randomized clinical trial, single blinding and performed in oncological patients carers in several practices. 2 samples of 40 carers. Intervention group will be taken through a seven session intervention with music; control group will undergo seven sessions of therapeutic education (as a reminder). Results will be evaluated using Pittsburgh Quality of Sleep Index, triaxial accelerometer, EuroQol-5D-5L, Caregiver Strain Index, Epworth Sleepiness Scale and Client Satisfaction Questionnaire. Statistical analysis Basal values will be compared for both groups. Then the values before and after the intervention will be compared using Student's t distribution for normal continuous variables, and Wilcoxon's T test for paired data in continuous not normal variables. A multiple linear regression will be carried out. The system developed for the PREDIMEDPLUS study will be used to process accelerometer data results. This semi-automated system manages data capture in a continuous stream in a central server of Malaga University. Processing and analysis of this data is also centralised, with an open source package, GGIR v. 1.5-1248

BACKGROUND AND CURRENT STATUS Sleep is a complex vital process in which humans spend a third of our lives. Insomnia is related to discontent with quality, length or continuity in that sleep, usually seen as problems to fall sleep, waking up frequently or a sleep that doesn't leave us rested. Progressive shortening on sleep length is a risk factor on developing obesity, diabetes, hypertension, cardiac related conditions and cerebrovascular accident, and can lead to premature death. Insomnia affects work performance: in severe cases it may lead to a reduction in care or reflexes, increased risk of being involved in accidents, poor morale or social relationships, absence from work and quality of life and immunity reduction. Regarding sleep decrease in informal carers it is observed that during the condition progress there are several changes in the patients. These changes are related to an increase on sleep alterations in carers, with a mayor prevalence (42 to 95%) in carers of patients with late stage cancer, compared to early stage patients carer (36 to 80%). Another aspect that may be overlooked frequently is physical activity conducted by carers as a health preventive measure, due to the benefits that physical exercise reports on the health and well-being, including a reduction on risk factors for cardiovascular diseases, cancer, stress and depression reduction, carer overload, mental and cognitive health, and general well-being. Sleep quality improves when physical activity increases in carers. Insomnia treatment is currently based in pharmaceutical treatment, including sedatives and hypnotic drugs which can provoke dependency and tolerance, reducing their efficacy. These drugs can induce cognitive and behavioural changes, and can lead to severe if not uncommon consequences. Other non-drug related treatments have been developed to deal with insomnia, including music therapy as this induces emotional responses in people. These responses are classified according to good or bad, and also in their intensity, being high or low. Several theories have tried to explain how our brain processes emotions. One of them is the classic subcortical route in which our limbic system plays a fundamental role. The type of melody influences recognition as happy or sad, and this recognition happens thanks to the inferior frontal gyrus, midline nuclear group and the anterior cingulate cortex. Music capacity to produce such an intense pleasure and the stimulation it provides to the endogenous rewarding systems suggest that, if not essential, music it is still a great benefit for our mental and physical well-being. A Cochrane review of 2015 evaluated the effects on listening to music for adults who suffered insomnia. 5 studies were included in the analysis (N = 264) to provide data about sleep quality on Pittsburgh Index, with results proving a benefit on listening to music. The effect was an increase in quality of sleep approximate to a standard deviation compared to no treatment or standard treatments. Only one study (N = 50, low quality evidence) showed data regarding latency at the beginning of the sleep cycle, total sleep time, sleep interruption and sleep efficacy. However it didn't show any evidence of improvement thanks to the music intervention. Authors concluded music can be effective to improve subjective quality of sleep in adults with insomnia. The intervention is safe and easy to administered, but more research is required to stablish the effects on other variables, as well as daytime consequences of insomnia. A systematic review in 2018 regarding the possibility that music can improve quality of sleep in primary insomnia adults concluded that the improvement is only present with the relaxation that is associated to music, compared to the normal state of the patients. This research is justified with the goal of evaluation in an objective way the quality of sleep in carers using the accelerometer, as well as subjective measures related to how listening to music can affect other sleeps variables (sleep interruptions, quality, total time, daytime consequences). The Cochrane review showed poor ratings of evidence in these variables except for quality of sleep being medium, as they were evaluated only in one study. High rating research is required to determine and establish the effect on daytime activity, which is why this clinical trial will take into account the authors recommendations to avoid bias. The use of the accelerometer will also provide data on physical activity in carers setting the ground for an intervention in the field. Objectives Main objective: \- To evaluate the effects of listening to music in sleep quality for oncology patients non-professional carers at home, and to evaluate the influence of specific factors that may modify that effect. Secondary objectives: * To assess the relationship between the waking state and the consequences during the day according to sleep characteristics. These include: 3\. Psychological results: 1\. Quality of life. 2. Carer overload. 4. Physical results: 1. Sleepiness during the day. 2. Physical activity, intensity and length evaluation. * To assess carers satisfaction with the intervention. Method: Randomised clinical trail, single blinding carried out in several practices in the field of Primary Care of 6 units of Clinical Management belonging to the Malaga-Guadalhorce Health District, with informal carers of domiciliary oncological palliative patients. Subject recruitment will be randomised and performed using a list of random numbers obtained using the program Epidat 3.1. This sample will take the oncological patients included in the Palliative Care Assistance Procedure showed on the Digital Clinical Records (DIRAYA) and having an informal care; every carer will be contacted and evaluated to the inclusion criteria, offering the possibility on taking part of the study if agreed and providing an informative leaflet. If accepted, a consent must be signed. Randomising method will be as followed: several cards will be prepared labelled with "you have been included in the intervention group" or "you have been included in the control group", making equal numbers for the groups. They will be sealed on non-transparent envelopes, will be mixed and numbered after shuffling. The participants will be given a number (in order of acceptance to take part) and that number will be the assigned envelope and the group it contains. This procedure makes bias harder to occur. Carers are offered the different groups and classified in intervention or control group, and this procedure it's performed by a sole person who will be responsible for randomising. Once the researcher in charge of randomising knows the group the carer belongs to, he will proceed to inform them of the procedure the participants will be subject to without mentioning if they are intervention or control group so the patient will be oblivious. This researcher will also inform the patient on how to use the accelerometer, the music program Spotify (Premium version) or the audio file through Google Drive (intervention or control group respectively). Sample size: Based on data published by Jespersen et al. 201543 about quality of sleep the researchers are expecting to find an equivalent effect of medium standard differences of 0,776, with an alpha error of 0,05 and a beta error of 0,10. The sample size will be two groups of 35 patients (Epidat 3.1) increased to 40 to compensate for possible droppings/leaving. Procedure for each group: * Control group: The treatment will be performed according to what it is established in the Andalusian Palliative Care plan from the Health Council regarding carers, which include a complete initial evaluation regarding the 14 needs of Virginia Henderson, and the subsequent care plan and follow-up for problems or symptoms check; a basic health education for self-care and patient care will be administered, regarding feeding and hydration, physical exercise and leisure, medication, effective communication, skin care, prevention and treatment of constipation, and sleep hygiene. Carers from the control group will also receive the conventional health education that the carers will need to listen through headphones/earphones in the morning, using the mobile phone and the audio file shared in Google Drive, in daily sessions of 30 minutes during 7 days, while also wearing the accelerometer a week before and the week of the intervention. This way the caseload management nursing staff will be blinded on patient evaluation (see Masking and Randomising). * Intervention group: carers will receive conventional health assistance complemented with music therapy with preselected music according to carers personal choice. The participants will have to listen to it half an hour before going to bed. The chosen music must be music that produces them pleasure or a nice feeling. This will be listened to via headphones/earphones and the mobile app Spotify (Premium version with free month trial). Daily session of 30 minutes during 7 days and the participants will wear the accelerometer the week before and the week of the intervention. Statistical analysis: A descriptive study will be performed with the collected variable using the medium and standard deviation in continuous normal variables, the confidence interval in punctual estimations; median and interquartile range in non-normal continuous variables, and frequency and percentages in categorical variables. The adjustment to normality will be determined with the Shapiro-Wilk test. The basal values of both groups will be compared. Pre and post-intervention values will be analysed in both groups using the Student's t distribution for normal continuous variables and Wilcoxon's T for paired data in non-normal continuous variables. A multiple lineal regression will be also carried out where the dependant variable will be: self-perceived quality of sleep, total sleep time, total resting time, sleep interruption, sleep efficiency, quality of life, carer overload, daytime sleepiness, physical activity, intensity and length. The independent variables will be the intervention ones and the socio-demographic such as sex, age, education level, civil status, time dedicated to care, help received in care, time that the participants have been taking care of the patient, and relationship to person care for. The analysis will be similar for the use of painkillers, anxiolytic and sleeping pill medication. The system developed for the PREDIMEDPLUS study will be used to process accelerometer data results. Processing and analysis of this data is also centralised, with an open source package, GGIR v. 1.5-1248 If statistically significant data are obtained in normal variables, the confidence intervals will be calculated at 95% to estimate values for the differences between the variables. The software to be used will be SPSS 23.0 and Epidat 3.01. Confidence interval will be 95% meaning only p values below 0,05 will be considered as statistically significant. The present studio adjusts to the ethical recommendations: 1975 Helsinki Accords revised in 2013 (Ethical principles for research in humans (Helsinki, Adapted by the 64th General Assembly, Fortaleza, Brasil, October 2013). All the personal information obtained in this study are confidential and will be dealt with in agreement to the EU 2016/679 regulation and the 27th of April 2016 GDPR (General Data Protection Regulation). The 14th November 41/2002 law regulates the patient's autonomy and all the rights and duties regarding information and clinical records. All the personal information collected in this study are confidential and will be treated in agreement with the 5th December 3/2018 Organic Law of Personal Data Protection and digital rights guarantee. Participation is fully voluntary and an Informed consent will be signed to be admitted into the study.