The aim this integrative review was to conduct a thorough assessment of the available evidence on PR in the adult critical care unit focusing on: 1) the prevalence of PR; 2) factors that influenced PR; 3) the role the nurses in the implementation of PR; 4) the nurse’s experiences in caring for patients with PR; and 5) the complications associated with applying PR.

This integrative review included quantitative, qualitative, mixed methodologies, and quality improvement projects from accessible literature, referring to studies published between January 2009 and December 2019 and a literature search was conducted in July 2020.

EBSCOhost, Ovid, ProQuest, PubMed, Wiley Online Library, SCOPUS, and ScienceDirect were among the databases searched. Restraint, critical care, intensive care, ICU, mechanical ventilation, intubation, nurses, and experience were utilized as search terms. The “AND” and “OR” Boolean operators were employed.

To map the search process, a flow diagram was used ( Figure 1). Initially, the database identified 3840 articles, and 852 duplicates were removed. From the remaining articles, 1205 were excluded electronically using exclusion criteria. Out of 1783, articles were found relatively relevant to the topic and 1498 articles were eliminated after limiting the search using the Boolean operator “AND”. A further seven were sourced from bibliographies and three of them were considered relevant to the screening. A total of 288 articles were considered for the title and abstract screening. 238 non-relevant articles were excluded and 50 full-text articles were retrieved. The screening process was conducted by two people independently. Twenty-nine of them were excluded after screening the full text and finally, 21 articles were considered for synthesis. EndNote X9.3.3 for Windows was used to manage the articles.

Inclusion criteria

Only peer-reviewed research involving patients or nurses in all types of adult ICUs was evaluated for this review, which was solely focused on the use of PR in adult critical care units. All full-text English papers with various techniques (quantitative research, qualitative studies, and quality improvement programs) are included.

Exclusion criteria

Exclusion criteria for this review included review papers, editorials, conference abstracts, and commentaries. Additionally, studies focusing on chemical restraints, studies conducted in settings outside of ICUs only, and previous literature reviews were excluded. EndNote X9.3.3 electronically removed words in the title that contained animals ( e.g. rat), psychiatric, neonatal, and pediatric. Restraint and intensive care unit/critical care unit/ventilated/intubated/nurses/experience were used with the Boolean operator “AND” to narrow the search. Non-relevant papers were eliminated by manually scanning the title and abstract for exclusion criteria such as animal research, psychiatric/pediatric/neonatal units, and so on. The decision to include an article in the final stage was based on the study’s eligibility ³⁶ and the content’s relevance.

Eligibility of data

The systematic review of literature for a particular intervention, condition, or issue is at the heart of evidence synthesis. The systematic review entails a number of sophisticated procedures that include an analysis of the existing literature (that is, evidence) and a judgment of the effectiveness or otherwise of practice. ¹¹ The initial assessment of articles was done using the Critical Appraisal Sklls Programme (CASP) ¹² checklist tool. This paradigm was chosen because it allowed for a systematic and rigorous approach to analyzing and evaluating article quality that could be used in both quantitative and qualitative methodologies. ¹³ The initial papers were screened using the inclusion/exclusion criteria. The literature was evaluated using the CASP tool’s three major questions: i) Is the research valid? (ii) What were the outcomes? (iii) Will the findings be usefully locally?. ¹² The articles answered sufficiently for these three questions were further considered. According to Whittemore & Knafl, ¹⁴ two quality criterion instruments should be utilized during the review process to assure the authenticity, methodological quality, and informative value of the included research. Because of the wide sample frame, evaluating the quality of the research paper in integrative reviews might be difficult. The articles were re-evaluated using the more specific questions from the CASP ¹² checklist tool and JBI (Joanna Briggs Institute) Critical Appraisal Tool. ^{11 , 36}

The CASP tool and JBI Critical Appraisal Tool includes a set of detailed questions to enhance visualization of the data and to simplify this complex comparison of CASP tool and JBI Critical Appraisal Tool, a common checklist was adapted. Each article was assessed using the adapted checklist and used effectively to analyze the articles clearly and succinctly. While these checklists were designed to assess qualitative research papers, they were adjusted to include the essential criteria stated in both the CASP tool and the JBI Critical Appraisal Tool for a qualitative and quantitative approach. The modified checklist includes a clear aim, methodology, design, recruitment, data collection, researcher and participant relations, participants and their voices, ethical considerations, data analysis and rigor, findings, and research value.

Data analysis

By permitting the use of a combination of studies with diverse methodologies, the integrative review method is the only strategy that has the potential to play a larger role in nursing evidence-based practice. ¹⁴ Data analysis in research reviews necessitates the ordering, coding, categorization, and summarization of primary source data into a cohesive and integrated conclusion regarding the research problem. ¹⁵ The data collected from multiple resources were compiled into a table to provide a more complete picture of the subject at hand. Combining many data sources, on the other hand, is difficult and time-consuming. ¹⁴ These challenges were approached by organizing data in a manageable framework. According to Whittemore & Knafl, ¹⁴ data reduction, data presentation, and data comparison are all part of the data analysis phase of integrative reviews. The phases of data reduction and presentation involved extracting and coding data to be organized into a manageable framework, which then increased the visualization of their linkages and patterns. This approach suited the review framework due to the varied data collected by the researchers through various approaches. Whittemore & Knafl, ¹⁴ proposed a consistent comparative approach for comparing integrative reviews, which was used in this review and presented in Table 2.

The fact that PRs are were commonly employed in the adult critical care unit was obvious. Descriptive research in China found that ICUs used PRs more frequently than medical-surgical wards. ¹⁶ In this review, we found that 23% to 75% of the patients were restrained during their stay in the critical care unit. ^{4 , 17 – 19} Critical care nurses (CCNs) (68%) reported that PRs were most commonly used in ICUs. ²⁰ According to 94.5% of CCNs physical restraint must be applied in ICU. ²¹ Most nurses working in critical care units experienced (68% to 100%) the use of PRs. ^{1 , 21} According to Langley et al., ²² all of the participants in their study agreed that PRs have a place in the ICU. And they stated that, before using PR the patient’s agitation must be assessed.

According to a study conducted in Canada by Luk et al., ²³ 83% of restraints were employed on the night shift and Jiang et al. ¹⁶ support that PR is more commonly used on the night shift. Turgay et al. ²⁴ on the other hand, found no variations in the frequency of PR use between night and day shifts. PRs were utilized by 57% of nurses prior to the patient’s interference with medical equipment. ¹⁷

Five studies looked into nurse-to-patient ratios. According to Kandeel & Attia, ²⁰ nurse workloads play a significant effect on the use of PR. Benbenbishty et al. ²⁵ found that PRs were frequently utilized for sedated and mechanically ventilated patients in units with a lower daytime nurse-to-patient ratio. And according to Hamilton et al. ¹⁷ restraint use was more likely with low nurse-to-patient ratios. The nurse-to-patient ratio was mentioned in the other two studies, but there was no discussion of the relationship between the ratio and PR applications. ^{18 , 22} Regardless of personnel resources or the number of beds in private rooms, PR usage varied among ICUs in a cross-sectional online survey conducted in Japan. ²⁶

According to Ertuğrul & Özden, ²⁷ out of total restrained patients, 85.6% were initiated in the first 24 hours of their hospitalization. The duration of PRs varied from hours to days. An article from Canada explored the median duration of restraint as 21 hours ²⁸ while a study from China found that patients stayed restrained for up to 20 shifts (8 hrs shift). ⁴ According to Langley et al. ²² and Luk et al. ¹⁹ the patients were restrained for an average of 9 and 4.1 days respectively. In a study, Hamilton et al. ¹⁷ found a prevalence rate of 358 restraint days per 1,000 ICU days. From two observational studies, out of total restrained patients, 42.9% to 48% were restrained for more than 24 hours. ^{4 , 28} Repeated application of PRs highlighted in some studies however, in most of the cases (75.9% - 83%) the PR was applied only once during ICU stay. ^{4 , 19}

Documentation is vital when a nurse takes care of a patient. It protects them from legal consequences arising from the outcome or lack of restraint. In a study, patients’ observational data were recorded in 51.3% of nursing notes and no data for removal time, patient responses, or complications were accurately recorded. ⁴ The indication for PR was reported for 91 patients and two individuals were not documented, according to Guenette et al. ²⁸ More than 59% had no documentation of PR as per Turgay et al. ²⁴ According to Kandeel & Attia, ²⁰ patient records did not contain information about the assessment of restraint sites. According to them, experienced nurses maintained better documentation regarding PR application and assessment than novice nurses.

Many studies support the lack of documentation of PR applications among patients. ^{4 , 20 , 24 , 28} Hence, the prevalence rate of PR applications can be more than the available data suggests. The high prevalence rate of PRs shows its importance in the adult critical care unit, making it inevitable in the care of patients.

The primary justification for restraints is patient safety. PRs were designed to prevent the unintentional removal of medical or therapeutic devices, hence reducing the threat to life. ^{4 , 16 , 17 , 20 , 21 , 23 , 24 , 29 – 31} The presence of an endotracheal tube raised the probabilities of PR application eight-fold, according to Hamilton et al. ¹⁷ who also emphasized the incidence of self-extubation in patients who were not restrained. Restraints were utilized as a preventative measure to prevent self-extubation. ²⁹ The most prevalent documented indication in a observational study was the prevention of treatment interference. ²⁸ Other indications for PR included prevention of falls or attempts to get out of bed, maintenance of posture & calming down the patient, and managing treatment resistance or care. ^{20 , 21}

In this review, two studies showed heavy workload and ward overflow with staff shortages as common reasons for PR. ^{16 , 31} According to the response of more than half of the samples in qualitative data collected by Langley et al., ²² PRs were used primarily for the benefit of the clinical staff and to leave the patient unattended. Nurses face difficulties to identify an alternative as a barrier to the removal of PRs to ensure patient safety. ^{17 , 30 , 31} However, in a study, for 33% of patients, alternative measures were considered prior to PR, and in 21% of cases chemical restraints were used instead. ²³

Restraints were sometimes employed to prevent injury to personnel or to oneself as a result of patient behavior such as agitation, delirium, restlessness, and varying degrees of impaired mental status. ^{17 , 23 , 25 , 29} Patients with delirium or coma, who were taking psychoactive or sedative drugs, and who were unable to speak vocally, were at a higher risk of receiving PR. ¹⁸ Establishing an alternative measure for reducing the use of PRs while ensuring patient safety remains a predominant challenge for nurses.

Patient’s requirements and desires must be addressed when applying for PR. Due to the variable responses to restraint treatment in this patient population, special consideration to be given to delirium patients. The nurse’s capacity to maintain vigilance in patient care creates a safe environment. ²⁹

The decision to release restraints is based on the patient’s behavior. Improved mental status (68.9%) is a vital indicator of the removal of restraint. ²⁴ Restraints were removed from 57% patients who were generally calm and cooperative for more than two hours. ²³ The nurses described that successful withdrawal of restraints without perceived negative effects is predicted according to patient behaviors. ²⁹ According to the CCNs, the decision to remove the restraint was made when the patient cooperated with the nurse, was awake and aware of their surroundings, and did not attempt to touch the tube, or when the patient was ready to have the endotracheal tube removed. ¹⁷ Before stopping PR, however, it should be determined if there is a risk to the patient’s safety. Falls, tissue injuries, pulling or removing connections, delays in therapies, extended hospital stays, high treatment costs, and even death are all risks of not utilizing PR. ³¹

In critical care units, bilateral upper limb restraints were most commonly used. ^{27 , 28} ICU patients were also restrained using four-point restraints. ^{4 , 20 , 24} Among 141 patients in Canada, most were restrained using wrist restraints (91%). All-four-limb restraints were used on a few patients (4%) and combinations of unilateral wrist-ankle and wrist-mitten restraints were rarely employed (2%). ²³

Benbenbishty et al. ²⁵ explored the common use of commercial wrist restraints. In a study in Turkey, PR was used with tough cuff material, green foam tie, and a roll of gauze at a rate of 17.6%, 10.8%, and 71%, respectively. The roll of gauze was used most commonly on the wrist. ²⁷ Hence, it can be articulated that the type of PR depends on the need of the patient and the material available in the unit.

The majority of the research in this review supports the nurse’s role as the principal decision-maker in the application and removal of PRs. In Egypt, nurses were responsible for 58.2% of the choice to use PR, while 41.1% was decided by both physicians and nurses. ²⁰ According to Turgay et al., ²⁴ 84.7% of PR were applied without a physician’s order in Turkey and according to Balcı & Arslan, ³² nurses deciding on the use of PR have higher scores than physicians. However, Yönt et al. ²¹ stated that 70.9% were decided jointly by the nurse and the physician and 25.5% by the physician alone. Before restricting a patient in Canada, nurses did not confer with the physician. ¹⁷ Nurses were the primary decision-makers in PR, according to Langley et al. ²² from South Africa. Nurses in Iran were not allowed to use PR without a medical order; however, restraints are used without a physician’s order since nurses believe physicians are unsupportive if further complications arise from not using PR ( e.g. if a patient falls). ³¹ In California, 77% of nurses felt confident releasing restraints without a doctor’s order. Nurses wanted to rely on their own clinical judgment when determining whether or not to restrain a patient. ³⁰

The literature shows that nurses play a significant involvement in the decision to use restraints on patients. In this review, no studies mention the role of patients or their families. Inconsistencies in the role should be addressed with a standard guideline that specifies each individual’s role (physician, nurse, patient, or family), and it should be very clear about a physician’s written order for PRs.

Moral distress may increase among the nurses who care for patients under PR. ³³ Many nurses voiced dissatisfaction with the use of PRs. Most of them empathized with restrained patients and established innovative ways to diminish PR use or lessen the impact on the patient’s freedom. ¹⁷ The ethical issues for CCNs are caused by the outcomes of utilizing PR. Because of the overwhelming workload and nursing unit overflow, they felt obligated to disregard patient rights and autonomy in favor of PR. Due to the breach of patient rights and experience, nurses were conflicted. Even after leaving their shift, some of them felt uncertain. ³¹ Nurses encounter ethical issues due to harm and benefit principles, according to Yönt et al. ²¹ In their study, according to 65.5 % of nurses, no family consent was obtained for PR, and 63.6 % of nurses have no reservations about PR.

In a Chinese study, only about a third of restrained patients gave informed consent for PR. ⁴ Communication between physicians, patients, and families is paramount. The proper application of PR should be a balancing act between its benefits and drawbacks. ²² These studies show the moral and ethical dilemmas facing nurses in the application of PR. Appropriate organizational policies will help nurses to overcome moral and ethical dilemmas. In addition, adequate training and education may be needed to prevent the psychological impact on the nurses.

A quality improvement project conducted in the US ³³ showed a reduction in restraint rate after restraint collaboration and the restraint culture shifted from heavy to minimal use. The use of restraints was included in the daily-goals checklist, and the need for restraint was assessed daily during the multidisciplinary rounds. Furthermore the researcher added that, nurses who engage in evidence-based practice use the most up-to-date evidence and gladly share it with their colleagues. The impact of a restraint management bundle in an ICU on restraint utilization was studied by Hall et al. ³⁴ When compared to the pre-cohort group, there was a reduction in the total number of restraint patients in the ICU in the post-cohort group (24.3% vs 20.9%). The average number of restrained patients per patient day decreased among the group (0.075 vs 0.059), and the average of restraint events per patient day also decreased (0.191 vs 0.133). According to the data, restraint use, and duration were reduced dramatically. The length of stay in the ICU, on the other hand, remained consistent over time (3.64 vs 3.6 days).

Targeted nurse-led interventions can help reduce the usage of restraints, especially in delirious patients. ²⁹ Hevener et al. ³⁰ conducted a quasi-experimental study in an ICU with the implementation of a restraint decision wheel (RDW) to reduce the usage of restraints, and the results demonstrate that restraints are gradually reduced in critical care units. In the first month, 64 devices were dislodged, and 38% of the patients were restrained; in the second to fourth months, only 51 devices were dislodged and 13% were restrained. During the research, the restraint used decreased to 32%. This study suggested 81% of nurses strongly agreed that the RDW was easy to use, and 62% thought that it was helpful in deciding whether or not to employ restraints, and 84% of nurses supported using the RDW on a regular basis. Many of them, however, stated that having another resource would be good.

In a descriptive and correlational study regarding PR, physician’s practice scores were found to be higher than those of nurses. Nurses have a sufficient level of information regarding restraint but have negative attitudes towards it. However, nurses’ attitude scores were found to be higher than those of physicians. And in practice, nurses’ skillset is lacking/insufficient. ³² This opened a floor for the training need of nurses about PRs. Ahmadi et al. ¹ conducted Interventional Educational Program for ICU nurses to modify negative attitudes regarding the use of PRs. It leads to a positive attitude and improved the level of knowledge, perception, and practice among them.

In the descriptive and cross-sectional study, Yönt et al. ²¹ concluded that 36.4% of nurses felt difficulty in deciding on PR use and ethical dilemmas. The majority of them (78.2%) did not participate in training regarding ethical dilemmas. They recommended providing training for the nurses regarding ethical principles related to PR and PR application. In addition, they suggested having established policies for the use of PR in the ICU.

In a prospective multicenter study at Dutch ICUs, only 31% of nurses reported the use of a protocol for PR. ¹⁸ In this review, two studies highlighted the need for a set policy or guideline for the restraint practice ^{4 , 16} and Unoki et al. ²⁶ suggested in-service education and the establishment of a systematic approach to reducing PR use in mechanically ventilated patients. However, the effectiveness and possibility of de-escalation in critically ill patients were not tested in any of the studies in this review.

The overall analysis of the data in this theme shows the high need for training for nurses regarding PR application. The establishment of a protocol or guidance will be more beneficial in the management of a patient with restraint. Implementation of Restraint Management Bundle, restraint decision wheel, and Targeted nurse-driven interventions can improve restraint application in ICU.

Studies in this review show the negative outcomes of PRs like neuromuscular complications, increasing the need for sedation, analgesics, and antipsychotics. Turgay et al. ²⁴ reported 36.8% complications and the most common one was skin breakdown. In Iran, 66.7% of the nurse had experienced complications caused by PRs. ¹ According to prospective observational cohort research, most issues increased after the first 24 hours. An increase in redness was generated by the roll of gauze and tough cuff material. The length of PR raises the risk of neurovascular problems. Nurses failed to pay attention to peripheral circulation and failed to check the restrained wrist on a frequent basis. ²⁷ Observational data from South Africa shows that 21.46% of patients with analgesics and/or sedatives. There were patients managed with restraint (21.46%) and no sedation. ²²

More patients obtained psychiatric medicines after PRs were applied, according to a single-center, prospective, observational study done in Canada. According to that study, following the adoption of PRs, opioid administration became more widespread, accounting for more pharmacological interventions. Prior to PR, fewer people were treated with psychiatric drugs than subsequently (56% vs 86%). The administration of opioids was more common after the use of PRs (20 % vs 54%) and accounted for more pharmacological interventions (29% vs 45%). Before and after PR, 16% of 50 patients remained oversedated and disturbed. However, 20% of individuals got overly sedated after receiving PR, whereas 6% became less sedated. Furthermore, 10% of patients become more agitated, compared to eight percent who become less agitated. ²⁸ According to Hamilton et al., ¹⁷ all opioid or midazolam administration (95 %) increased the likelihood of restraint use. Analgesic, sedative, and antipsychotic medication use, excessive sedation, agitation, and the incidence of an adverse event all predicted PR use or days utilized. ¹⁹ However, according to Gu et al., ⁴ analgesic use can be an independent protective factor for PR use.

PRs have a psychological effect on patients, causing issues including fear of the ICU, depression, anger, aggression, restlessness, agitation, and anxiety. ¹⁹ The impact of PR on the outcome of the physical and psychological elements was disclosed in this overview of studies on this topic. Restraint complications may have an impact on the patient’s long-term and short-term outcomes. Those issues can be avoided by assessing the restraints on a frequent basis.