Practices for promoting sleep in intensive care units in Brazil: a national survey

Objective To conduct a national survey of intensive care professionals to identify the practices for promoting sleep in adult intensive care units in Brazil and describe the professionals’ perceptions of the importance of sleep for patients. Methods An electronic questionnaire was distributed by the clinical research cooperation network of the Associação de Medicina Intensiva Brasileira and by the Brazilian Research in Intensive Care Network to physicians and nurses registered with the association. The questionnaire evaluated the profile of the respondents, the profile of their intensive care units, whether protocols for promoting sleep were present, the pharmacological and nonpharmacological measures typically employed in the unit, and the professionals’ perceptions regarding sleep in critically ill patients. Results A total of 118 questionnaires were evaluated. The Southeast region of the country was the most represented (50 questionnaires, 42.4%). The majority of units had a clinical-surgical profile (93 questionnaires; 78.8%), and 26 had a continuous visitation policy (22.0%). Only 18 intensive care units (15.3%) reported having protocols for promoting sleep. The most cited measure for sleep promotion was reducing light during the night (95 questionnaires; 80.5%), which was more often performed in private intensive care units. Almost all of the responders (99%) believed that poor-quality sleep has a negative impact on patient recovery. Conclusion The responses to this Brazilian survey revealed that few intensive care units had a program for promoting sleep, although almost all participants recognized the importance of sleep in patient recovery.

Although a significant portion of health care professionals recognize the importance of sleep for critically ill patients, fewer than 30% of ICUs report having a program for promoting sleep. (6) A recent American College of Critical Care Medicine (ACCCM) guideline on pain, agitation/sedation, delirium, immobility, and sleep disruption (PADIS) recommends that sleep be promoted by optimizing the environment by reducing noise, light, and nocturnal stimuli to protect the sleep-wake cycle. (7) However, in Brazil, there is little information on practices for promoting sleep in the intensive care setting and how these recommendations are applied.
Thus, due to the scarcity of Brazilian data, the present study aimed to evaluate the main sleep promotion practices in adult ICUs in Brazil.

METHODS
A national descriptive survey was conducted. Intensive care physicians and nurses with an electronic address listed in the research groups of the clinical research cooperation network of the Associação de Medicina Intensiva Brasileira (AMIBNet) and Brazilian Research in Intensive Care Network (BRICNet) were invited to participate in the study. A total of 1,622 electronic invitations were sent to medical professionals and 523 to nursing professionals from all Brazilian regions listed in AMIBNet. Additionally, an electronic invitation was sent to physicians and nurses with e-mails listed in a Yahoo ® group.
Study participants received an invitation to participate in the study and a link to the SurveyMonkey ® platform. A reminder to participate in the study was forwarded by AMIBNet to the same mailing lists after 1 month. The link remained valid from April 3, 2019 to June 10, 2019.
After agreeing with the participation conditions, the study participants were directed to fill out the electronic questionnaire (Appendix 1), which contained 44 questions on ICU identification, participant profile (profession and position), ICU structure (type, funding source, number of beds, type of division between beds), visitation policy (number of periods and length of each visit), nonpharmacological measures for promoting sleep (nocturnal noise reduction, light reduction, use of ear plugs or eye mask, sleep interruptions for tests, and/or radiography), and pharmacological measures for inducing sleep (melatonin, neuroleptics, benzodiazepines, and other drugs). The questionnaire exclusion criteria were as follows: not filling out the hospital name or city of the ICU, answering fewer than 50% of the questions, nonadult ICU, and duplicate ICU. In cases of duplicate responses from the same unit, only one questionnaire was considered valid, with the first choice being that from the physician coordinator, followed by the staff physician, nurse coordinator, on-duty physician, and on-duty nurse. Thus, each questionnaire analyzed represented a single ICU. The study was approved by the Research Ethics Committee of Hospital Beneficência Portuguesa de São Paulo under opinion number 3,238,465.

Statistical analysis
The data collected were analyzed using the Statistical Package for Social Sciences (SPSS), version 20 (IBM, Illinois, USA). Categorical variables are described as absolute and relative frequencies. Quantitative variables are expressed as measures of central tendency (mean and median) and dispersion. To compare data on sleep practices between different kinds of ICU, the chi-squared test or Fisher's exact tests was used for categorical data. Values of p < 0.05 (two-tailed) were considered significant.

RESULTS
A total of 190 responses to the electronic survey were received. After assessing these questionnaires, 72 were excluded. The main causes of exclusion were not identifying the hospital or city of origin of the ICU and duplicate questionnaires from the same ICU. The other 118 questionnaires were considered valid and were included in the statistical analysis. Table 1 shows the profile of the units according to location. Among the study participants, 80.5% were physicians, and 29.7% of the sample identified themselves as physician coordinators of the unit. The results regarding the structural characteristics of the evaluated units are presented in table 2. Of the 118 evaluated ICUs, 96 (81.4%) reported having a sedation and analgesia protocol. The Richmond Agitation and Sedation Scale (RASS) was the most frequently used in the ICUs that had sedation and analgesia protocols (94 units).
The main sedation strategy used in these ICUs was daily awakening (55 ICUs; 57.3%), and the intensive care physician was responsible for implementing the sedation protocol in 85 (88.5%) units. In the evaluated sample, only 18 units (15.3%) reported having a sleep promotion protocol.
The main medications used among participants who reported having pharmacological sleep promotion protocols in the ICU were dexmedetomidine (11 units) and neuroleptics (10 units). The most widely used measure to promote sleep was reducing light at night, and the least performed measure was offering relaxing music to help induce sleep (Table 3). Figure 1 shows the perceptions of the study participants about the main sleep promotion measures in the ICU. Approximately 99% of the participants believed that poor-quality sleep in the ICU could negatively affect patient recovery. Table 4 shows the relationship between the funding source and sleep promotion measures in the ICU: there was no difference in the presence of sleep promotion protocols according to funding source. Reducing light at night was the only sleep promotion measure with a significant difference in prevalence between public and private ICUs (p = 0.01).

DISCUSSION
Our study, one of the first to evaluate practices for promoting sleep in Brazilian ICUs, has the following main findings: a small portion of the evaluated ICUs had a sleep promotion program; the most frequent measure for promoting sleep was reducing light at night; approximately 30% of the participants reported that they did not routinely assess the sleep of their patients, but 65% reported that more than half of their patients received sleep induction medication (dexmedetomidine being the main medication used); almost all the participants believed that poor sleep quality can negatively affect patient recovery and believed that adequate sleep reduces the incidence of delirium.
The concept of ICU humanization is broad and should involve much more than extended or 24-hour visitation. Measures aimed at patient well-being are part of this humanized care. (8) Noise and poor sleep quality are among the main complaints of ICU patients and their relatives. (1,2)  Excessive noise peaks may occur in the ICU setting and interfere with patient sleep. (9)(10)(11) Our findings suggest that a minority of the evaluated units measure the noise level (14.4%) or perform interventions to minimize it (such as reducing the volume of alarms at night), revealing an opportunity for improvement in the quality of patient care. Ensuring an environment with controlled noise is a simple, low-cost intervention and falls within the concept of humanized and patient-centered care.
The effect of natural light on critically ill patients has been evaluated for at least a decade. The deprivation of exposure to natural light and exposure to artificial light have been related to changes in the circadian rhythm, sleep disorders, and delirium. (3,12) However, Khon et al. found no effect of the presence of windows and exposure to sunlight on outcomes in critically ill patients. (13) Measures such as exposure to natural light, maintenance of the circadian rhythm, and minimization of artificial light exposure at night are part of a bundle for promoting sleep, with satisfactory results. (14) We observed that the majority of the units reduced light at night (80.5%), which is also a feasible and inexpensive intervention for promoting sleep in patients.
The Clinical Practice Guidelines for the Prevention and Management of Pain, Agitation/Sedation, Delirium, Immobility, and Sleep Disruption in Adult Patients in the ICU and the Guidelines for Family-Centered Care in the Neonatal, Pediatric, and Adult ICU, both from ACCCM, make recommendations for sleep promotion in the ICU. (7,15) The American PADIS guidelines recognize that some measures, especially nonpharmacological measures, are safe and should be implemented as a sleep promotion bundle, including adjustment of the ventilatory mode at night and reduction of noise and light. These same guidelines do not make recommendations on the use of medications to promote sleep, exception that they recommend against propofol to improve sleep in critically ill patients. (7) The guidelines for family-centered care in the ICU recognize that sleep deprivation in the ICU has a negative impact not only on the patient but also on the family and recommend that adequate sleep conditions be offered to the relatives of patients in the ICU. (15) Nevertheless, in the evaluated sample, only a small portion of units (15.3%) reported having a sleep promotion protocol. There was no difference between private and public units regarding the institution of sleep promotion measures, although the pharmacological protocol was more frequent in private units. Light reduction at night was also more common in private units (p = 0.01), which is a simple measure that does not require financial investment and which should be more encouraged in Brazilian ICUs. Among the study participants, 99% answered that they believed that poor sleep can negatively affect patient recovery and that adequate sleep reduces the incidence of delirium. An international survey study with 1,223 participants, in which 84% of the ICUs were located in the US and 1% in Brazil, showed that a small portion of the units had a sleep promotion protocol (32%). As in our study, most participants believed that poor sleep can affect the patient recovery process. (6) This study has some limitations. First, the analyzed sample was small. E-mails were sent to the AMIBnet list and to the e-group with a link to the survey. According to the AMIB census conducted in 2016, Brazil has 1,951 establishments with adult ICUs, 77 coronary ICUs, and 50 burn ICUs. (16) A total of 190 questionnaires were filled out, but only 118 were included in the analysis. This factor may have caused sampling bias and limited the external validity of our results because the response rate was approximately 5%. Nevertheless, this is the first description of sleep promotion practices in Brazilian ICUs, and the distribution of responses was similar to the distribution of ICUs among Brazilian regions, with representatives from all regions. Second, the survey was based on previous studies, and it was validated only among our researchers using a method similar to Delphi and not previously externally validated. Third, although the questionnaire evaluates practical aspects of sleep promotion and sedation protocols, we did not specifically evaluate the details of the protocols of units that reported having sleep promotion measures. In addition, we did not ask the units about protocols for the prevention and management of delirium. We know that analgesiasedation and delirium protocols are contiguous with each other. Lastly, we did not assess the ventilatory mode setting at night, another measure that can impact sleep improvement. (17,18) The implementation of measures to promote sleep in ICUs is challenging and involves the participation of a multidisciplinary team. Sleep promotion bundles include nonpharmacological and pharmacological measures.
Among the nonpharmacological measures, a noise reduction program, light reduction, and patient care adjustment are environmental modification actions that are simple to implement, cost little, have shown several benefits to patients, and may be the first measure to be adopted in the development of a sleep promotion program, but we emphasize that such actions require the engagement of the team to become effective.

CONCLUSION
In Brazil, there is a gap between the perception of health professionals about the impact of sleep in the intensive care unit and the implementation of measures for promoting sleep.
The simplest and least expensive measures for promoting sleep could be implemented in a broad multidisciplinary program. Studies with a more robust methodology should be performed to evaluate the actual impact of a sleep promotion protocol on critically ill patients in various outcomes in this population.

ACKNOWLEDGMENTS
The authors thank the clinical research cooperation network of Associação de Medicina Intensiva Brasileira (AMIBNet) and Brazilian Research in Intensive Care Network (BRICNet) for the support and dissemination of the study.
Objetivo: Realizar um inquérito nacional com profissionais de terapia intensiva para determinar as práticas de promoção do sono em unidades de terapia intensiva para adultos no Brasil, e descrever suas percepções sobre a importância do sono para os pacientes.