Effects and Interventions of Pressure Injury Prevention Bundles of Care in Critically Ill Patients: A Systematic Review

Corresponding Author: Wenny Trisnaningtyas Student of Master Program in Nursing, Department of Nursing, Faculty of Medicine, Universitas Diponegoro, Indonesia Email: wennytrisnaningtyas@gmail.com Background: Many studies on pressure injury prevention bundles have been conducted outside the Intensive Care Unit (ICU). The bundles, which include multi interventions, have proven effective in reducing pressure ulcer incidents compared to a single intervention. However, the existing review studies on pressure injury prevention in ICUs still only investigate a single intervention rather than multi interventions. Only few reviews, to our knowledge, involves prevention bundle strategies in the ICU. Purpose: This study aims to review the effects of the pressure injury prevention bundles of care on the incidents of pressure injury in critically ill patients and the intervention measures of the care bundles. Methods: This review searched published articles from several databases, namely EBSCO, ScienceDirect, PubMed, ProQuest, Google Scholar, and Scopus from 2009 up to 2020. PRISMA flowchart was used to select relevant articles using several inclusion and exclusion criteria, resulting in 17 articles from 50 eligible full-text articles for assessment. The included studies were assessed for their quality using Joanna Briggs Institute (JBI) critical appraisal tools. The synthesis was then conducted narratively. Results: As many as 17 studies, which mostly had good quality yet evidence level of II, were included in the analysis. The findings showed that the pressure injury prevention bundles of care decreased pressure injury incidents as many as 4.3%36.2% in developed countries and 4.16%-25.72% in developing countries. Moreover, the bundles of care which significantly reduced the incidents of pressure injury consisted of 7 intervention measures, which were pressure injury risk assessment using Cubbin Jackson scale, skin assessment and care, repositioning, nutrition, education, support surface, and medical device care. Conclusion: The review concluded that the pressure injury prevention bundles of care in critically ill patients significantly reduced the incidents of pressure injury. The study recommends more studies with stronger evidence levels to carry out and utilize 7 intervention measures as a preventive standard of care in critically ill patients.


Introduction
Pressure injuries still become a health problem related to patient safety in various parts of the world, regardless of the current progress in the quality of health care. The incidence of pressure injuries in the hospital is one of the unexpected events during hospitalization in the context of patient safety (Slawomirski et al., 2020). The incidence of pressure injuries in the Intensive Care Unit (ICU) in some countries is reported to be around 8. 1%-63% (Clough et al., 2016;Dale et al., 2018). At the same time, the incidence of pressure injuries in some developing countries is about 4.16%-39.3% (Akhkand et al., 2020;Anderson et al., 2015;Girard et al., 2014;Lupe et al., 2015;Mallah et al., 2015;

Inclusion and exclusion criteria
The articles obtained from the searching process were then screened for their eligibility using some inclusion and exclusion criteria. The inclusion criteria were randomized controlled trial and experimental research studies, conducted in the critical care unit, examining pressure injury incidents and interventions, involving adult participants of 18-69 years old, participants without underlying diseases and pressure injuries before ICU's commencements, treated in the critical units more than 24 hours, and utilizing more than one intervention of pressure injury preventions. The review excluded papers that were written in non-English and non-Indonesian languages and of their results which were difficult to be separated from other populations or interventions.

Screening of articles
The Mendeley citation manager was used to pool and screen the search results. Two authors (WT and NR) independently screened the articles according to titles and abstracts, and also languages after duplicate removal. Full-text articles were then screened for their eligibility based on the inclusion and exclusion criteria. Any discrepancies were discussed between the two authors and consulted to the third author (RR).

Data extraction
Quality assessment from the included studies and extraction of the data were separately done by two authors (WT, NR). Some disagreements were discussed between the two authors and resolved by the third author (RR). Furthermore, data extraction was done using predetermined extraction table consisting of author/year, city/country, aim, design, sample and setting, intervention measures, main result, and outcomes (Table 1) (see Appendix 1).

Quality appraisal
The quality of studies was critically assessed using the Joanna Briggs Institute (2017) (JBI) critical appraisal tools and their level of evidence (Table 2). Checklists for quasi-experimental and randomized controlled trials were used to appraise the included studies (Munn et al., 2014). There were 9 questions of the quasi-experimental checklist and 13 questions of randomized controlled trials checklist with 4 options for the answer: yes, no, unclear, and not applicable. The answer of yes scored 1, while the answer of no, unclear, and not applicable scored 0. The studies were classified as good if the total score was >80%, fair if the total score was 50-80%, and poor if the total score was <50% (Reilly et al., 2016). Five level classical pyramids of evidence was also used to classify the included studies' level of evidence. Level I and II included randomized controlled trials and cohort studies, respectively, while case-control studies were classified as level III (Greenhalgh et al., 2011;Higgins et al., 2021). There were no level IV and V studies.

Data analysis
Heterogeneity in the methods and statistical values of the outcomes caused this review could not perform a meta-analysis of the pressure injury incidents. Therefore, a narrative synthesis was conducted. The studies' characteristics, the incidents of pressure injury, and the intervention measures of pressure injury's prevention bundle were separately summarized to enhance understanding and thorough analysis of the outcomes. The steps of narrative analysis were carried out according to nine syntheses without meta-analysis (SWiM) reporting items of systematic reviews (Campbell et al., 2020).  Lewis et al., (2015) II Fair *Notes. Critical appraisal Score: Good (>80% quality score), moderate (50%-80% quality score, and poor (<50%) (Alshahrani et al., 2021), *I: Systematic review or meta-analysis, prospective cohort, II: RCT, cohort studies, III: Case control, IV: Case series, V: Expert opinion, report or clinical example (Greenhalgh et al., 2011)

Characteristic of the included studies
A PRISMA flowchart was used to select the included studies ( Figure 1). In this review, 17 papers were included, consisting of quasi-experimental (n=14), RCT (n=2) and cohort (n=1) studies. Of this number, 9 studies were in developing countries (Anderson et al., 2015;Chaboyer et al., 2016;Loudet et al., 2017;Lupe et al., 2015;Rogenski et al., 2015;Siracusa & Schrier, 2011;Swafford et al., 2016;Vasconcelos & Caliri, 2017), while 7 studies were carried out in developed countries (Amr et al., 2017;Avşar & Karadağ, 2018;He et al., 2016;Lewis et al., 2015;Mallah et al., 2015;Tayyib et al., 2015;Uzun et al., 2015). These studies were conducted in Australia, Saudi Arabia, Turkey, Lebanon, United Sates, Egypt, Brazil, United Kingdom, China, and Argentina; and included a total of 7.439 patients. Based on the review, the bundles of pressure injury were performed in various types of ICU such as medical, surgical, trauma, neurology, cardiovascular, and oncology. Twelve studies used a large sample of tertiary referral hospitals, 3 studies used small samples in an ICU, and 2 studies did not include sample sizes. The details for each study are presented in Table 1.
The results of the initial search identified 302,569 titles from online databases; their duplicates were removed. As many as 301,733 titles were excluded after a further screening on the titles and abstracts, and language, leaving 836 full-text. The results of the critical appraisal showed that most of the articles have good quality but with level II evidence. Only 2 articles show a high-quality and high level of evidence ( Figure 1).

Primary outcome measurement: Incidence of pressure injury
The outcome of all these studies was the incidence of pressure injuries. In this review, 14 of 17 studies reported a decreased incidence of pressure injuries (Amr et al., 2017;Anderson et al., 2015;Avşar & Karadağ, 2018;Chaboyer et al., 2016;Coyer et al., 2017;He et al., 2016;Loudet et al., 2017;Lupe et al., 2015;Mallah et al., 2015;Rogenski et al., 2015;Siracusa & Schrier, 2011;Swafford et al., 2016;Tayyib et al., 2015;Uzun et al., 2015). This review showed that the incidence of pressure injuries in patients after the implementation of pressure injury prevention bundle of care in ICU ranged from 4.3% to 36.2% in developed countries and from 4.16% to 25.72% in developing countries (Table 3). The incidence of pressure injuries in patients after the implementation of pressure injury prevention bundle of care is presented in Table 3.

Secondary outcomes: Pressure injury prevention strategies
The interventions used to prevent pressure injuries were analyzed and reported by the type of intervention. This review found that multi-interventions (bundles) that significantly reduced the incidence of pressure injury consist of seven (7) interventions: risk assessment, skin care, reposition, nutrition, support surface, education, and medical device maintenance. It is shown that multi interventions with significant effects on reducing the incidence of pressure injury were risk assessment, skincare, support surface, and repositioning (Table 1).

Pressure injury risk assessment (1st strategy)
Several risk instruments were used in the studies to identify participants who were prone to stress injuries. The study showed that the incidence of pressure injury patients admitted to the ICU were assessed by Waterlow, Cubbin Jackson, Braden. The Braden scale was the most frequently used pressure injury risk assessment instrument in this study (Amr et al., 2017;He et al., 2016;Loudet et al., 2017;Lupe et al., 2015;Mallah et al., 2015;Rogenski et al., 2015;Siracusa & Schrier, 2011;Swafford et al., 2016;Tayyib et al., 2015;Uzun et al., 2015;Vasconcelos & Caliri, 2017). However, Cubbin Jackson scale had more comprehensive risk assessment aspects for critically ill patients.

Reposition (3rd strategy)
Repositioning is an intervention to reduce the duration of pressure on tubal areas that are prone to pressure injuries, such as areas of bony prominence, in addition to providing a sense of security and maintaining the patient's functional ability. Studies related to reposition are presented in Table  1. As many as 15 studies (Amr et al., 2017;Avşar & Karadağ, 2018;Chaboyer et al., 2016;Coyer et al., 2017;He et al., 2016;Lewis et al., 2015;Lupe et al., 2015;Mallah et al., 2015;Manzano et al., 2014;Siracusa & Schrier, 2011;Swafford et al., 2016;Tayyib et al., 2015;Uzun et al., 2015;Vasconcelos & Caliri, 2017) reported the benefits of patients' repositioning. The repositioning used a three-or twoor six-hour play schedule using a "play clock." The legs of the bed were elevated by 20 degrees if clinically permitted. The patient's heel was elevated and supported. Also, draw sheets were used to transfer and lift patients (Gillespie et al., 2012;Zhang, 2021;Zuo & Meng, 2015). Behrendt et al. (2014) reported that a bedside pressure mapping system was able to support clinical staff to optimize repositioning and duration of repositioning, allowing interventions to reduce initial pressures. When two methods were compared for patient repositioning, a clockwise rotation system was effective in reducing the incidence of pressure injury compared to standard pillow care (Whitty et al., 2017).

Nutritional intervention (4th strategy)
Nutrition plays an important role in the prevention and treatment of pressure injuries (He et al., 2016). Macro and micronutrients were needed by each organ system in a certain amount to promote the growth, development, maintenance, and repair of body tissues. Nutrition-related studies were presented in Table 1. Four studies (Amr et al., 2017;Chaboyer et al., 2016;Gage, 2015;Mallah et al., 2015;Siracusa & Schrier, 2011) related to nutrition for the prevention of pressure injuries. In addition, albumin levels as an indicator of malnutrition should be assessed routinely (weekly or biweekly) to determine trends in nutritional therapy adequacy. A recent study reported by Amr et al., Nurse Media Journal of Nursing, 11(2), 2021, 160 Copyright © 2021, NMJN, e-ISSN 2406-8799, p-ISSN 2087-7811 (2017) pointed out that enteral nutritional formula fortified with fish oil containing polyunsaturated fatty acids, significantly reduced the incidence of stress injuries in critical units.

Support surface (5th strategy)
The support surface is the surface on which a patient is placed to manage pressure, shear, and microclimate loads. This surface includes mattresses, trolleys, operating table mattresses, integrated bed systems, and pillow chairs. The support surface is designed to reduce interface pressure ulcers by increasing the surface area of the body or alternating the area of the body in contact with the support surface. Table 1 shows the results of the support surface of 10 studies (Anderson et al., 2015;Gage, 2015;He et al., 2016;Loudet et al., 2017;Lupe et al., 2015;Mallah et al., 2015;Ozyurek & Yavuz, 2015;Tayyib et al., 2015;Uzun et al., 2015). These studies only evaluated mattresses; no studies related to pillow performance were found. Ozyurek & Yavuz (2015) assessed the effect of two types of viscoelastic mattresses (viscoelastic foam 1 consisting of two layers and viscoelastic foam 2 consisting of three layers) in patient admitted to critical units. They found no statistically significant difference between patients using plain foam and patients using viscoelastic foam to decrease the incidence of pressure injury grade ≥ II when compared with the overlay of APAM (alternating pressure air mattress) in patients admitted to critical units.

Educational (6th strategy)
This review showed that staff education was an important component of pressure injury prevention. The educational programs should include a variety of factors that reflect the multifactorial nature of pressure injury. There were 6 studies (Chaboyer et al., 2016;Gage, 2015;Lupe et al., 2015;Swafford et al., 2016;Tayyib et al., 2015;Vasconcelos & Caliri, 2017) in Table 1 which showed the results of education and education of health workers. The education with a focus on preventive care can be effective in reducing the incidence of pressure injury in critical care settings.

Medical device maintenance (7th strategy)
Treatment using medical devices is proven to reduce pressure injuries by using several dressings on the skin area where the medical device is attached. Based on Table 1, there were two studies Zakaria et al., 2018) reporting that dressing on medical devices was effective in preventing pressure injuries. There was one type of dressing that showed a statistically significant reduction in the incidence of pressure injuries, such as polyurethane film dressings rather than hydrocolloid dressings (Boyko et al., 2018). Lewis et al. (2015) and Tayyib et al. (2021) found that coated thin hydrocolloid and silicone single-layer dressings were effective in reducing pressure injuries on the medical device fixated areas (non-invasive ventilator, endotracheal tubes/ tracheal tubes, nasogastric tubes, urinary catheter, non-rebreathing/oxygenation mask) in critically ill patients when previously applied within 24 hours of admission to the critical care room.

Discussion
This review aimed to identify the impact of the pressure injury prevention bundles of care on the pressure injury incidents in critically ill patients and to identify the intervention measures of the care bundles. The result of this review showed that the bundles of care reduced the incidents of pressure injury in critically ill patients both in developed and developing countries although the reduction was higher in the former countries. Furthermore, risk assessment, skin care, reposition, nutrition, support surface, education, and medical device maintenance were identified as effective strategies for preventing pressure injuries in critically ill patients.

1 The incidence of pressure injury
The use of multiple interventions reduces the incidence of pressure injury in critically ill patients by 4.3% to 36.2% in 9 developed countries, and 4.16%-25.72% in 5 developing countries (The World Bank, 2021). The success of the bundles in reducing the incidence of pressure injury is affected by nurses and is dependent on standard operating procedures, nurses' knowledge, compliance with nursing actions supported with clear documentation, and complete data interpretation (Lavallée et Nurse Media Journal of Nursing, 11(2), 2021, 161 Copyright © 2021, NMJN, e-ISSN 2406-8799, p-ISSN 2087-7811 al., 2017; Zuo & Meng, 2015). This review cannot perform meta-analysis due to a heterogeneity of the included studies, and lacking of RCTs. The decrease in the incidence of pressure injury in developed countries is higher due to nurse compliance, the number and ratio of nurses caring for the number of patients (1:1.96 or 1: 3.24) (Zhang, 2021). Low nurse workload (Lee et al., 2019), support from ICU nursing organizational institution (hospital management conducts continuous audits regarding the performance of nurses) (Amaravadi et al., 2000;Driscoll et al., 2018), The success of pressure injury prevention bundles was influenced by education and training of nurses in prevention bundles, nurse awareness related to pressure injury prevention (Rivera et al., 2020), collaboration with unit-based wound care expertise, audit feedback to doctors, and consistency of nurses identifying pressure ulcer events (Floyd et al., 2021;Krupp & Monfre, 2015). Moreover, The bundle approach is more effective than a single method in the Chinese ICU (Zhang, 2021).

Scale of pressure injury (1st strategy)
The developed scales for ICU patients are 7, such as Braden, Norton, PURAS, Waterlow, Cubbin Jackson, EVARUCI, and Suriadi Sanda. A meta-analysis by Chen et al. (2016) showed that the Braden scale had a moderate prediction value (AUC of 0.7686, moderate sensitivity (0.80) and low specificity (0.42)) to predict pressure injury risks. While meta-analysis Wei et al. (2020) stated the opposite; to predict pressure injury risk in ICU patients, the Braden scale had AUC of 0.7812, sensitivity of 0.89, and specificity of 0.28. Deng et al. (2017) stated that the Braden scale was more suitable to assess pressure injury in general ward because it evaluated skin damages in seven domains: sensory perception, moistness, activity, mobility, nutrient, and friction, and shear (Gomes et al., 2011;Lewicki et al., 2000).
Components of the Braden scale were absolutely less proper to be applied to ICU patients, the risk factors of pressure injury incident were generally caused by age, some comorbidities, unstable hemodynamic status, sedation, peripheral perfusion change, hypotension, vasoactive and vasopressor medication, frequent incontinence and edema which needed sustainable assessment due to rapidly changing condition of critically ill patients (Hyun et al., 2013;Kim et al., 2013;Manzano et al., 2014;Seongsook et al., 2004). Some studies (Cooper, 2013;Seongsook et al., 2004;Manzano et al., 2014;Shi et al., 2018) stated that the Cubbin & Jackson scale had better validity compared to the Braden scale because the domains in the Cubbin & Jackson scale was specifically used to assess pressure injury in ICU patients. Cubbin and Jackson (1991), Jackson (1999), and Shi et al. (2018) reported that the Cubbin & Jackson scale had a value of sensitivity by 89 % and 61% lower than Braden scale for its specificity. Cubbin & Jackson scale reflected the complex condition of critically ill patients such as patients with comorbidities, unstable hemodynamic status, ventilated, vasoactive medication who needed sustainable assessment due to rapidly changing condition of critically ill patients.
The Suriadi & Sanda scale had three domains, namely interface pressure, temperature, and cigarette smoking history, which effectively detected pressure injury in ICU patients. However, the weakness of this scale was that it was conducted only in one city in a country despite the sensitivity value of 81%, and specificity of 83% (Suriadi et al., 2007). EVARUCI scale had five domains containing consciousness, hemodynamic status, respiratory status, mobility and others (Souza et al., 2018). This scale was effective to screen pressure injuries in ICU patients due to its relatively high predictive value; the ROC value was around 0.938, which was a very good value. However, it had a limitation for its small number of samples. The writers compared scales in this review and concluded that the use of pressure injury risk scale in ICU was better than using prediction capacity. The reduction of pressure injury incidents was pertained to risk factor identification and the use of care intervention to prevent pressure injuries. Therefore, scale identification with a good predictive capacity would contribute to improving accuracy in care decisions which could provide support in critical care.

Skin care (2nd strategy)
The skin care component of the prevention bundle in this review was taken from some studies. Skin care in the prevention of pressure injury able to reduce the incidence of pressure injury included risk assessments using physical assessments which were done comprehensively from the patient's skin and documented within 24 hours after entering ICUs, and evaluated every shift; bathing the patient using antiseptic soap with 2% of hydrogen perioxide / prepackaged washclotes (pH balanced once per day; giving basic skin moisturizer (petroleum jelly, VCO, sorbolence) every bath (Zuo & Meng, 2015). Foam dressing had the ability to move water vapor out of the dressing to minimalize the accumulation of heat and excess moisture so it could prevent skin maceration in patients with lateral positions. Multilayer materials in foam dressing, i.e., silicon foam composite, could minimalize friction when the dressing touches skin surface and protect skin from mechanical wound (Huang et al., 2015). Ohura et al., (2005) stated that foam could be compressed, but it was easily damaged; it also effectively reduced pressure and friction. Film dressing, based on the review, effectively reduced pressure injury because it was semi-occlusive with multilevel permeability, but it could not be used in patients with high fever and excessive sweat (Sood et al., 2014;Weller et al., 2020). Tayyib et al. (2015) stated that most frequent pressure injuries were located on the sacrum area, tailbone, and heel, so the high-risk of friction was likely to occur. It became a challenge for nurses to protect those areas because dressings were easily saturated, or even dirty. However, some studies had limitations related to this topic. Studies by Avsar and Karadag (2018), Tayyib et al. (2015); and Zakaria et al. (2018) were conducted in a relatively short period, while a study by Loudet et al. (2017) in this review did not document their risk assessment.

Reposition (3 rd strategy)
Repositioning was to move a person to a different position to distribute pressure from a certain body parts. Amr et al. (2017) suggested the time distributed for one position was not more than 2 hours. However, the repositioning frequency had to consider the general medical condition, skin condition, and comfort. Repositioning could be a hard duty, so it had to be performed by trained personnel by practicing correct techniques to prevent further pressure injury. A study by Tayyib et al. (2015) in medical ICU found that employment of mobility team consisting of pressure injury prevention nurses, skin care mobility assistants, and patient mobility assistants was pertained to significantly reduce pressure injury in ICUs (6.1% vs. 9.2%, p=0.04). Chaboyer et al. (2018) and Wayunah (2018) found that repositioning in critical care patients (by an hourly-repositioning) only occurred around 50% from time to time. Lateral position often became a choice when intervention and care were frequently performed in ICU, which might contribute to pressure injury incidents in that area. The problem was in the practice of lifting the patient's head until 45° to prevent pneumonia, considering that the position of mechanically ventilated patients received bigger pressure in the sacrum and heel (Gillespie et al., 2012;Lewis et al., 2015). Patients in this position tended to slide down the bed and increased the risk of skin exposure with friction and shear . Even though there was no definite evidence of repositioning frequency and determination in critically ill patients, it was clearly shown that regular re-repositioning was important prevention. A study by Al-Dorzi (2019) showed a-2 hour-repositioning reduced pressure injury about 1.35±0.520 compared to control group (1.73±0.790, p=0.000).

Nutritional intervention (4 th strategy)
Nutrition components in this review were found only in two studies (Gage, 2015; Siracusa & Schrier, 2011) which included nutrition in pressure injury prevention bundles. These studies reported that nutrition intervention significantly reduced pressure injury (p=0.05). This was correlated to patients with malnutrition status that increased pressure injury risks. Generally, nutrition support had to target prevention or nutrition deficit correction. Based on the circumstantial evidence, guidelines recommended the supply of 30 to 35 kcal/kg weight per day for malnutrition patients or in malnutrition risk on pressure injuries (Al-Dorzi, 2019; EPUAP et al., 2019). Recommended protein and meta-analysis from eight studies (6,062 patients) which compared mixed nutritional supplement to hospital standard diet did not find an obvious effect from supplemented nutrition in pressure injury development (risk ratio 0.86; confidence interval 95%, 0.73-1.00; p=0.05) (Al-Dorzi, 2019).

Support surface (5 th strategy)
Support surface component was based on 10 studies reviewed about pressure injury prevention bundles (Anderson et al., 2015;Gage, 2015;He et al., 2016;Loudet et al., 2017;Lupe et al., 2015;Mallah et al., 2015;Ozyurek & Yavuz, 2015;Tayyib et al., 2015;Uzun et al., 2015). Ozyurek and Yavuz (2015) conducted a study by comparing the efficacy of two viscoelastic beds; one was with two layers, while the other was with three layers. The result showed that there was a significant difference in pressure injury prevention (p=0.44). This was in line with a study by de Camargo et al. (2018) which found that pressure injury incidence with support surface was about 32.2% compared to the group with no support surface 80.6% (p≤0.001); viscoelastic could redistribute pressure and affect micro climate that could reduce skin moisture to prevent pressure injuries. Studies showed that the use of pressure air mattress and pillow to position patients successfully reduced pressure injuries, statistically significant about 4.1-17% (Lupe et al., 2015;Uzun et al., 2015). In a study by Tayyib et al., (2015), the use of support surfaces to manage the weight and pressure of patients along with a-two hour-repositioning reduced the pressure injury incidence for 39%. Also, Anderson et al. (2015) and Tayyib et al. (2015) reported that the use of two air mattresses reduced pressure injury incidence by 2.1%-7.14%. The use of support surface aimed to reduce the duration of pressure between the patient body and support surfaces in pressure injury prevention (McInnes et al., 2015).

Education (6 th strategy)
There were not many studies about pressure injury prevention bundles that discussed education component, but the study by Anderson et al. (2015), Tayyib et al. (2015), and Uzun et al. (2015) showed that an education of element bundles to the ICU nurses was an important action that can reduce pressure injury by 2.1% to 39.3%. Giving education was considered important to improve staff awareness in pressure injury prevention. Staff education and training also contribute to the successful result in care improvement and patient outcomes. Education pertained to effective bundles improves care quality because it combines some core interventions to "care package", like VAP bundle approach to reduce pneumonia pertained to ventilator usage and respiratory tract infections in critically ill patients (Anderson et al., 2015;Tweed & Tweed, 2008).

Medical device (7 th strategy)
The incidence of pressure injury due to the use of medical equipment was estimated at about 10-12% (Brooke et al., 2019). According to Padula et al. (2015) the incidence of pressure injury related to the use of medical equipment was about 30-70% and mostly it was caused by external factors. Medical equipment which were usually used in critically ill patients, such as intravenous catheters, splints, niv mask, servical collar, nasogastric tube, and endotracheal tube. A recent study by Hanonu and Karadag, (2016) showed that the incidence of pressure injury related to the use of breathing assist devices was the highest percentage (about 30-70%) started about 3.3 days after administration and 20% of nurses were unaware of the fact that medical devices could also lead to ulcer formation. While a study in the Netherlands Ham et al. (2017) showed that support surface intervention, reposition every 2-4 hours, and nutrition factors gave contribution to the incidence of pressure injury related to immobilized patients, and also the use of manual tightening rope in immobilized patient contributed 20.1% of pressure injury incident. A study by Zakaria et al. (2018) showed that manipulating the position of breathing assist tube will reduce pressure injury from 77.8% to 13.1%. Tayyib et al. (2021) describe the maintenance of medical equipment in the prevention of pressure injury. The procedures begin with securing medical equipment, protecting the skin with silicon hydrocolloids in areas at high risk of pressure ulcers, and avoiding to directly install medical devices with the patient's skin unless the patient's condition is not possible. Next, checking the skin under the medical device more than 2 times a day, monitoring nutrient intake, and choosing the right Nurse Media Journal of Nursing, 11(2), 2021, 164 Copyright © 2021, NMJN, e-ISSN 2406-8799, p-ISSN 2087-7811 size and type of medical device to suit the patient's condition were conducted. Last, monitoring the incidence and prevalence of pressure ulcers related to the installation of medical devices is carried out in a regular basis (Mehta et al., 2019;Tayyib et al., 2021). Risk identification instrument of pressure injury (Gómez et al., 2017;Jackson, 1999;Lyder et al., 1999;Suriadi et al., 2007) used by most hospitals today was not sufficient to identify the risks of pressure injury related to the attached medical device. There were lots of medical devices which could initiate pressure injuries, and those instruments alone were ineffective to examine and identify pressure injuries in mucus membrane. According to this finding, we need future studies to develop appropriate instruments that can be used specifically to examine and identify patients with a high risk of developing pressure injuries related to medical devices.

Implication and limitation
This review may implicate the critical health care providers especially critical nurses and the ICUs for providing comprehensive overview through a thorough and systematic study to prevent pressure injuries in the critically ill patients. Intervention measures, such as risk assessment, skin care, reposition, nutrition, support surface, education, and medical device maintenance could be applied as a bundle of care for the better patient nursing care, compared to the single intervention.
The strength of this review is the importance of raising the topic of the pressure injury prevention bundle guided by the librarian's comprehensive search strategy and the systematic review of this review represents a novelty in synthesizing related literature containing multi-component prevention of pressure injury and incidents of pressure injury in critically ill patients. With these facts, the bundles, by such findings in this review, might be implemented as a multi-intervention alternative to prevent pressure injury in critical care units. This review might be chosen as a reference of Indonesian nursing as a guideline in daily practice.
One considerable limitation from this review was unable to perform a meta-analysis as the included studies varied in the methods and statistical values of the outcomes. From 13 studies, there were only 2 studies which used true high quality RCT method, while other studies used low quality method. In addition, there were so many components of the bundle for pressure ulcer prevention, and each hospital implemented this bundle in various techniques that differ from one hospital to the others. Thus, clinical staffs who were involved in this bundle should have more attention to detail when implementing this bundle to patients with complex medical situations and diseases. There was no single component of this bundle that was proven to be the most effective to prevent the incidence of pressure injuries. The bundle prevention of pressure injury in this review could not show metaanalysis purpose. Besides, this review did not evaluate the working quality of medical and nursing staff while implementing this bundle, but only showed that this bundle prove to be more effective in lowering the incidence of pressure injury in critically ill patients.

Conclusion
Pressure injury prevention bundles can reduce the incidence of pressure by 4.3% to 36.2% in developed countries and 4.16% to 25.72% in developing countries. The intervention strategies in the prevention bundles included the risk assessment of pressure injury using the Cubbin Jackson scale, skincare, repositioning, nutrition, education, support surfaces, and medical device care. This bundle was a multi-intervention method of prevention, which consists of 7 components, and still need supervision from higher-rank of nurses in critical care units in order to early detect the incidence of pressure injuries. Although the prevention bundle can reduce the incidence of pressure injuries, this review found that scientific evidence of its effect is still lacking. Only 2 studies used high quality RCT. Therefore, it is recommended to conduct randomized clinical trials for next studies to investigate the effect of prevention bundles to reduce the incidence of pressure injury. The incidence of pressure injuries in the intervention group was 7.14% and the control group was 32.86% from 784 days of observations. Poisson regression reveal the incidence of pressure injuries in the intervention group was 70% lower. The intervention group experienced a significant decrease in stage 1 (p=0.002) and stage 2 (p=0.026).

Incident
3 Anderson et al.,  (4) The incidence of pressure injuries in the control and intervention groups decreased by around 4.6% to 0.3%. There was a significant decrease of approximately (p<0.0001) in the incidence of pressure injuries in the sacral at 2 months of treatment (n=1, 0.3%) compared to the control group around 4.6% (n 16, 4.6%). There was a significant reduction in the incidence (p<0.001). The incidence of pressure injuries in the sacral area of the intervention group was 4.75%.

Ankara, Turkey
To determine the costeffective and efficacy of evidence based nursing intervention on increased tissue tolerance by maintaining tissue integrity.  (3) The incidence of pressure injuries in the control group was 54.5% and the intervention group was 18.3% (p<0.05). The incidence of pressures injuries in the intervention group with skin integrity was 13%, the incidence of MDRPU pressure injuries was 3.9%. While in the control group, the incidence of pressure injuries in skin integrity was 33. The incidence of pressure injuries decreased 69% (from 45 patients around 10% to 17 patients around 3%). The incidence of pressure injures related to oral ETT in the intervention group was around 9/28, whereas in the control group 27/30 p = 0.031, while the incidence of pressure injuries related to NGT in the control group was 35/45 with an incidence of 77.8%, in the intervention group 5/38 with p = 0.012.  (2), & support surface (5) The incidence of pressure injuries in the control group was 37% and the interventions group was 17% with significant differences (χ2 = 8.86, t = 0.593, p = 0.554), pressure injuries generally occurred in the sacrum 46%, trochanters 11% and in the heel 7%, interventions that decreased the occurrence of pressure injuries in this article is detection of pressure injuries, and repositioning.

Incident & intervention 10
Gage,  (2), nutrition (4), support surface (5), & education (6) The incidence of pressure injuries decreased significantly from 18 cases to 2 cases from 2011 to 2013.  (2), & education (6) The incidence of pressure injuries was not explained; the risk assessment of pressure injuries was recorded at around 57.9% in the pre phase and 77.3 in the post phase.

Incident & intervention 12
Coyer et al.,  (1), & reposition (3) The incidence of pressure injuries before being given an InSPIRE intervention was around 32%, whereas after being given the intervention, it was around 15% (p = 0.016), (χ2 = 5.847, df = 1, p = 0.016; there was a significant difference of 17%. There was a significant difference in the interventions for giving acrylate polymer based prophylactic in reducing the incidence of pressure injuries by about 94% in episodes of bathing patients in the intervention group. Intervention

Pennsylva nia
To design an evidencebased PI prevention bundle based and determine its effectiveness on reducing Pi Quasiexperimental 1199 respondents Pressure injury assessment (1), reposition (3), skin assessment and skin care (2), & nutrition (4) The incidence of pressure injuries in the quarter 1 control group was around 5.7%, quarter 2 was around 0.0%, quarter 3 was around 5.2%, quarter 4 was around 0.0% (p = 0.11). The incidence of pressure injuries in the intervention group (PUB): Quarter 1 around 0.0%, quarter 2 around 0.9%, quarter 3 around 0.0% and quarter 4 around 0.0%, reducing the incidence of pressure injuries around 1%.

Incident 14
He et al.,

Hangzhou, China
To determine whether skin barrier factors were associated with the common complication of PrUs in ICU Single center (pre and post) 102 patients in postoperativ e ICU (54 men and 48 women) Pressure injury assessment (1), reposition (3), skin care (2), & support surface (5) The incidence of pressure injury incidence of 62.5% decreased by about 31.4% (p = 0.031); this incident in male sex (33%) was more at risk of developing pressure injuries than women (29.2%) (p = 0.031). The incidence of pressure injuries in the scapula was around 0.4% (p=0.058), sacral around 0.5% (p<0.001), hip 0.7% (p<0.001), and heel 0.3% (p=0.062). Braden scale has a significant difference between patients who have pressure injuries and patients who do not have pressure injuries (p<0.001).

Incident 15
Lupe et al., Miami To identify the prevalence of HAPU at the institution and to implement interventions to reduce the incidence of HAPU The incidence of pressure injuries in April 2009 was around 11.7%, and decreased around 2.8% in September 2012.  (3), & support surface (5) The incidence of pressure injuries from the pre 75% group decreased to 54% (p=0.016); In the post group using pressure-reducing mattresses increased from 48% to 85%; family participation increased from 9% to 39%.

Incident & intervention 17
Lewis et al.,  (2), support surface (5), & reposition (6) The cumulative incidence of pressure injuries was 39.3% (33/84). The incidence of pressure injuries in patients with medical devices was 8.39%; the incidence of pressure injuries in the sacrum area was 24.3% and heels 29.2%, and in the ear area was 37.5%.