Facial ulcers in patients with COVID-19 admitted to ICU: review of the evidence

In December 2019, the first cases of pneumonia from a novel coronavirus were identified in Wuhan, China, and the pathogen was later named as severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). According to official figures (https://coronavirus.data.gov.uk) up to 1 March 2021, 4 182 009 people had tested positive in the UK, 439 352 of them requiring hospital admission. At the peak of COVID-19 admissions by the end of January 2021, more than 4000 occupied an intensive care unit (ICU) level 3 bed, those reserved for patients requiring either advanced respiratory support or two or more organ system support. Patients with COVID-19 often present with acute respiratory distress syndrome (ARDS), requiring management in the prone position for up to 16 hours, several times, which presents a challenge for health professionals to maintain skin integrity (Worsley et al, 2020).

Prone positioning is a postural therapy where the patient lies in a horizontal position facing down, which aims to enhance the oxygenation and lung compliance of patients who develop moderate or severe ARDS, a condition experienced by 67-85% of critically ill COVID-19 patients admitted to ICU (Moore et al, 2020; Stephen-Haynes and Maries, 2020). The APRONET analysis, a prospective international study performed between April 2016 and January 2017 in 141 ICUs from 20 countries, revealed that 32.9% of patients with serious ARDS were ventilated in the prone position for prolonged periods of time, increasing their risk of facial pressure ulcers (PUs) (Moore et al, 2020).

The total cost of PUs to the NHS in England has been estimated to be more than £530 million, based on an audit between May 2012 and April 2013 (Guest et al, 2018). However, this does not consider the psychological and emotional costs to patients, which contribute to the direct and indirect costs of their care. Facial PUs, such as those related to endotracheal tubes (ETTs) or nasogastric tubes, can have a long-term social and psychological impact for a patient, with the possibility of disfiguration and scarring of ulcers on the face and neck (Gefen et al, 2020). Demarré et al (2015) estimated that the daily cost per patient of PU prevention in their region of Belgium ranged from €15.70 to €87.57, whereas the cost of treatment per day ranged from €1.71 to €470.49. Preventing PUs and their associated costs socially, psychologically and financially is therefore of great significance (Sleiwah et al, 2020) and identifying evidence-based recommendations to prevent or reduce their incidence is the focus of this study.

Although the National Institute for Health and Care Excellence (2014) guideline for prevention and management of PUs advises repositioning every 4 hours for critically ill patients in ICU, this is limited to only arms, legs and head in patients in prone position (Capasso et al, 2020; Worsley et al, 2020). Moreover, even if regular repositioning of the head is implemented, device-related PUs still occur, such as those caused by pressure from the ETT on the corner of the mouth (Worsley et al, 2020).

Devices used in intensive care are particularly associated with PUs. Gefen and Ousey (2020a) estimated the incidence of device-related PUs in ICU at around 20%, with a pessimistic estimate of 40%, and further analysis identified that 68% of device-related PUs were associated with respiratory devices, of which 20% were non-invasive ventilation (NIV) devices. This is similar to the findings of Moore et al (2020) where prevalence of facial PUs ranged from 0.38% to 53.2% and their incidence ranged from 1.9% to 71.6%.

Prone positioning is not a new phenomenon and has been used in patients with ARDS over 20 years, with PU the main complication highlighted widely in the literature, including a Cochrane review (Bloomfield et al, 2015). Evidence suggests that the prone position is associated with reduced mortality rates in patients with COVID-19, hence it is widely used (Shelhamer et al, 2021). However, prone position has been found to increase the number of facial PUs, with data from 26 ICUs in France and one in Spain showing that the prevalence of facial PUs among patients with severe ARDS who were in prone position was almost seven times greater than the prevalence among patients who were managed in a supine position (Girard et al, 2014).

Although prone positioning is not new, the scale of the problem presented by the COVID-19 pandemic is new. In Brigham and Women's Hospital in Boston, USA, the average daily number of patients in ICU increased by 27.5% during April and May 2020, with an increase of 275% in hospital-acquired PUs (Martel and Orgill, 2020). Almost 50% were attributed to the prone position and of those PUs that were device-related, 73% were caused by the ETT and its securement.

A regional medical centre in California, USA, introduced a pressure injury prevention bundle in January 2019, decreasing the incidence of PUs by 88% (Singh et al, 2020). However, at the start of the COVID-19 pandemic, the authors noted that patients developed extensive skin injury despite the implementation of the entire bundle, highlighting that further research was needed to understand the pathophysiological processes in patients diagnosed with COVID-19 (Singh et al, 2020).

Methods

A multi-case research study approach, using secondary data, was used to critically analyse the development of pressure ulcers in COVID-19 patients admitted to ICU and nursed in prone position. A case study approach was selected in order to enable a detailed, intensive study of a particular contextual, and bounded, phenomenon (Luck et al, 2006). For the purpose of this study, the phenomenon was COVID-19 patients who underwent prone positioning. The context was COVID ICUs and the high incidence of PUs and preventive measures.

In this study, retrospective research was undertaken covering the period from the beginning of COVID-19 up to April 2021. A multiple case research study was used to enable in-depth understanding of the cases as a unit, through comparison with evidence from multiple case studies (Heale and Twycross, 2018). The inclusion criteria for the sources of information were articles or other documentation in English of COVID-19 patients cared for in ICU who underwent at least one period of prone positioning and suffered facial pressure ulcers as a result. The electronic databases searched were CINAHL and MEDLINE. The search terms were:

• COVID OR coronavirus OR 2019-ncov OR SARS-CoV-2 OR cov-19
• Pressure ulcers OR bed sores OR pressure sores OR pressure injury
• Prone position OR prone positioning OR prone OR proning position

A total of 131 studies were retrieved from the database search. Other sources of literature were sought, for example, national guidelines to deepen the understanding of the case, which led to two additional sources of information. These sources were identified from the reference lists of the retrieved studies and an initial scoping search within a university online library. A total of 16 sources met the inclusion criteria (Figure 1).

Critical appraisal was undertaken using the Joanna Briggs Institute (JBI) tools and the JBI's hierarchy of evidence classification (JBI Levels of Evidence and Grades of Recommendation Working Party, 2014). Overall, the quality of evidence is low, as highlighted by the JBI level of evidence scores in Table 1.

Findings

The information sources analysed to illuminate the case for this study provided in-depth insights into causative factors and preventive measures for facial PUs.

Specialists at Kettering General Hospital NHS Trust, UK, identified five potential causes for the large number of PUs in patients with COVID-19: the physiological changes related to COVID-19, the increase in patients needing medical devices to support their treatment, poor nutrition due to their critical condition, immobility and understaffing (Percy, 2020). Percy outlined that with COVID-19, the poor oxygen supply to the tissues leads to ischaemia and increases the risk of developing PUs, risk is also increased due to the hyper-inflammation and coagulation problems commonly encountered in critically ill COVID-19 patients. Gefen and Ousey (2020a) added that the increased incidence of PUs in patients with COVID-19 is caused by the cytokine storm seen in severe disease, where endothelial dysfunction, hypoxia and hypercoagulability, create microthrombi, which interact with two of the three primary aetiological factors in device-related PUs, inflammation and ischaemia, causing an increase in PUs in patients with COVID-19.

Hocková et al (2021) reviewed 14 studies and concluded that the main causative factors of oral complications, mainly perioral PU, were prolonged prone positioning, less-experienced staff undertaking positioning and the use of respiratory support devices. COVID-19 patients who are in prone position cannot be fed by nasogastric tubes, contributing to weight loss and poor nutrition, risk factors for developing PU and impairing wound healing once the PU has developed (Gefen et al, 2020). Gefen et al added as a causative factor the patient's inability to sense the device and associated pressure, friction and shear on the skin due to sedation, and health professionals excessively tightening devices such as NIV and ETT holders to ensure correct functioning and securement, leading to oedema and tissue damage. Furthermore, patients with COVID-19 cared for in ICU were identified as needing full support for their activities of daily living, resulting in immobility for long periods, and with understaffing and increased patient:nurse ratios regular repositioning was hindered (Gefen et al, 2020; Percy, 2020). Patients can be in prone position for up to 16 hours with only minimal head and arm repositioning, increasing their risk of developing PUs. Moreover, understaffing can prolong those 16 hours further, until experienced staff such as anaesthetics and ICU nurses are available to reposition the patient back to supine position, which is challenging and time-demanding, particularly if the patient is ventilated, on inotropes and on continuous veno-venous hemofiltration. Understaffing during the COVID-19 pandemic was due to several factors, such as staff shielding due to their own or family member's health issues, staff being infected with SARS-CoV-2 and the increased staff demand being covered with redeployed staff who were unfamiliar with working in an ICU (Percy, 2020).

Another causative factor is heat trapping, with some medical device designs trapping heat between the device and the skin, increasing moisture and skin fragility and elevating the metabolic demands of the tissue at a time when there is shortage of metabolic supplies (Gefen et al, 2020). In addition to this, fever develops in the majority of patients with COVID-19 cared for in ICU, with the highest facial temperature being on the forehead.

Many patients with COVID-19 are cared for in ICU with either mechanical ventilation or NIV (in either supine or prone position) and although normally these treatments are used for short periods of time, patients with COVID-19 require prolonged periods of respiratory support (Percy, 2020). NIV masks, for example, often ulcerate the bridge of the nose or the nasolabial fold (Gefen et al, 2020; Percy 2020). In addition, nasogastric tubes are also used for long periods, resulting in an increase of device-related PUs around the nostrils (Percy 2020). Devices associated with PUs are oxygen face masks, NIV masks, ETTs and their securement devices, nasal prongs and tubing for high-flow nasal oxygen, donut head support, nasogastric tubes, orogastric tubes, and items dropped on beds such as pens or clamps (Gefen et al, 2020).

Perrillat et al (2020) published two case reports where patients who underwent prone positioning using a semi-lunar head cushion developed multiple facial PUs under the tape used to secure the ETT, on the labial commissure. Both patients also developed PUs on the cheek due to the semi-lunar cushion, which was normally used for supine patients and not designed for prone position, being too hard and the distribution of pressure points too restricted. Moreover, one of the patients sustained a corneal ulcer due to incomplete eyelid closure, which was only taped shut without using a dressing (Perrillat et al, 2020).

The physiological, psychological and social impact for patients and their caregivers is considerable, with PUs increasing the risk of life-threatening infections such as sepsis, causing pain, and leaving scars that may be highly visible or cause hair loss, leading to altered body image and a reduction in quality of life, with a devastating impact for the patient's wellbeing (Gefen et al, 2020). Some facial PUs are deep tissue injuries, which are particularly difficult to heal without significant scarring. It is important to note that the face has an important aesthetic function and thus scarring results in distress and a profound psychological impact (Peko et al, 2020; Sleiwah et al, 2020). Moreover, the increased length of hospital stay resulting from PUs and their complications consequently increases the financial and human resources used (Gefen et al, 2020).

Prevention

Gefen et al (2020) identified as main preventive measures the redistribution and relief of pressure points with periodic repositioning, pressure relief through application of an alternative device such as prophylactic dressings, improved device designs, incontinence and moisture management, nutrition and hydration and education, as well as advocating the SECURE mnemonic: Skin, Education, Champion, Understanding, Report and Evaluate.

Skin condition

The evidence highlights that the skin must be kept clean and moisturised, to remove dirt, prevent maceration and minimise friction, with many patients with ARDS being kept in negative fluid balance, impacting the skin and mucosal surface (Moore et al, 2020). Over-cleansing can dehydrate the skin further due to damage of the skin's natural moisturising and although traditionally soap and water have been used for cleansing, evidence exists that this impacts negatively on the pH balance of the skin (Moore et al, 2020). Gefen et al (2020) suggested using cleansers ranging from pH 4.0 to 7.0 to reduce dryness, erythema and irritation. Moore et al (2020) analysed four RCTs comparing different strategies for the prevention of facial PUs related to the use of NIV, finding conflicting results regarding different moisturisers, with one RCT not finding a statistically significant difference when hyperoxigenated fatty acid moisturiser was compared with a placebo (Moore et al, 2020). Suctioning oral secretions, using liquid skin protectants on the face, changing dressings as often as required and using hydrofibre/calcium alginate dressings are recommended by Capasso et al (2020) to manage moisture. A hand-held device, the SEM Scanner (Bruin Biometrics), assesses sub-epidermal moisture, which in principle predicts tissue break down several days before damage becomes visible; however, its size makes it unsuitable for assessing relatively small regions such as the nose or lips (Gefen et al, 2020).

Pressure redistribution

Pressure redistribution or positioning devices should be used in patients undergoing prone positioning, to offload pressure points on the face (Moore et al, 2020). It is unclear which device is best with evidence not showing any statistically significant difference when prone head support was compared with a control group without the prone head support system (Moore et al, 2020). When three different support pillows were compared on spinal surgery patients, a polyurethane foam facial pillow did not prevent the development of PU in 45% of patients, whereas those using a protective helmet system of polyurethane foam or a neoprene air-filled device did not sustain any PU (Moore et al, 2020). Jiang et al (2021) experienced in their ICU that several COVID patients who underwent prone positioning developed facial PUs at the site of maximal contact against a foam donut-shaped pillow used to offload pressure, which were discovered when patients were returned to a supine position. Capasso et al (2020) and Peko et al (2020) recommended shifting the patient's head every 2 hours and re-positioning the head every 4 hours, as well as offloading the head with offloading devices. In addition Bamford et al (2019) suggested that the bed should be placed 30 degrees head up to minimise the development of facial oedema, and when using an alternating-air mattress it must be flat and in static mode for prone positioning.

Eyes covering

The eyes should be taped shut, with perioperative nursing guidelines highlighting the benefits in using dressings prophylactically to pad the eyes and facial tissues when patients underwent prone positioning during spinal surgeries (Peko et al, 2020; Stephen-Haynes and Maries, 2020). Capasso et al (2020) recommended applying ophthalmic lubricant before taping the eyelids shut in order to prevent corneal abrasions and Bamford et al (2019) suggested that the dressing used to cover the eyes should be a gel pad or similar, secured with micropore tape horizontally, ensuring eyelashes and eyelids are free from ointment, and avoiding applying any pressure on the eyes.

Devices and prophylactic dressings

The medical device must be designed to manage tissue deformation and stresses, as well as managing the transfer of thermal energy to tissues, heat trapping and moisture at the skin-device interface (Gefen et al, 2020). In order to achieve this, materials with mechanical properties that reduce pressure and shear, with low-friction surfaces and coatings and contours that do not include sharp surfaces or highly curved regions, are important (Gefen et al, 2020). A balance needs to be found between the purpose, such as NIV masks that must maintain a seal to function, and minimising pressure and frictional forces (Gefen et al, 2020). Some studies suggest that silicone-foam multi-layered prophylactic dressings are well-tolerated and minimise shearing of the skin when the patient is placed in prone position and also allow moisture to evaporate (Peko et al, 2020). When a twill endotracheal securement was used with a silicone strip underneath to reduce the pressure, it decreased device-related PUs from 20.7% before introduction to 5.2% with the silicone strip (Moore et al, 2020). Silicone gel padding redistributes pressure well and is available in a wide range of shapes and sizes, they are cost effective as they can be cleaned and repositioned, always on the same patient; however, they do not offer any absorbency (Stephen-Haynes and Maries, 2020). Sleiwah et al (2020) identified 16 patients with COVID-19 who had PUs around the lips, with 15 of them having had their ETT secured with devices. As a result of their research, Sleiwah et al recommended replacing devices with ties before proning a patient, with the option of adding silicone padding to decrease the shearing force between the skin and the device. Capasso et al (2020) also recommended that commercially available ETT securement devices could contribute to skin breakdown and suggested health professionals should consider using tape instead, while also recommending thin foam dressings under medical devices to reduce device-related PUs on the face. This recommendation was previously made by Bamford et al (2019) in the guidance for prone positioning in adult critical care, which advised health professionals to securely tape or tie the ETT, removing any anchor fast device, and to ensure the use of padding between any ties used and skin. Shearer et al (2021) noted that of the 143 patients at two hospitals in Washington DC, USA, who were COVID-19 positive and underwent prone positioning in ICU, 47.6% developed facial PUs, mostly on the cheeks, attributed to the commercial ETT fasteners. The authors began placing foam dressing under the commercial tube fastener, however, they were unable to analyse their impact as there was no protocol in place (Shearer et al, 2021).

There is limited evidence on the effectiveness of many preventive measures, without any commercial dressing being designed specifically to prevent device-related PUs; however, hydrocolloid and film dressings appeared to significantly reduce facial PUs when compared with a control group (Gefen et al, 2020; Moore et al, 2020). Hydrocolloid dressings are good in areas of higher friction but provide no pressure relief, whereas multilayer silicone foam products reduce pressure, shear and friction and have good absorbency, but they may not fit into very narrow or small areas (Stephen-Haynes and Maries 2020). Peko et al (2020) analysed the biomechanical performances of the Mepilex Border Flex (Mölnlycke Health Care) dressing design in protecting facial tissues while the head is in prone position, a multi-layered silicone-foam dressing, which absorbs a portion of the mechanical energy associated with bodyweight-related forces, so the energy does not reach the tissues (Peko et al, 2020). The authors used a computational three-dimensional head model where they applied the Mepilex dressing and used another model without the dressing as a control group, using a donut-shaped headrest as the head support, alleviating the facial loads by more than 50% when the Mepilex dressing was used. Gefen and Ousey's (2020b) research further identified that polymeric membrane dressings allowed better clearance of accumulated body-heat when compared with conventional dressings, a property relevant for patients with COVID-19 who often develop fever.

Discussion

The high risk of facial PUs in patients who are in prone position highlights that skin assessment is critical in preventing PUs and must be part of the daily routine, ensuring that pressure points in patients undergoing prone positioning—the forehead, cheeks, nose and chin—are checked as soon as the patient is admitted to ICU, periodically as indicated by the risk assessment tool used, and before and after each time the patient undergoes prone positioning, as highlighted by Capasso et al (2020), Moore et al (2020) and Gefen et al (2020).

The most common risk assessment tools are the Braden Scale, the Norton Scale and the Waterlow. Scale. The Waterlow, being the most sensitive at 82%, nonetheless has low specificity, at only 28% (Gefen et al, 2020). The evidence does not recommend any specific risk assessment tool, acknowledging that they all have generally low sensitivity and specificity, which can also vary depending on the population where they are applied and the experience of the health professional using them and therefore local policies should be followed (Mahalingam et al, 2014). Moreover, assessment is more challenging with darkly pigmented skin, thus some risk assessment tools may have lower sensitivities and specificities in some populations (Gefen et al, 2020). Gefen et al (2020) and Moore et al (2020) identified the key things that should be checked when assessing the skin: colour, humidity, moisture, oedema, turgor, ‘bogginess’, temperature, signs of skin irritation and PU, bruising and dryness. Gefen et al suggest that in order to do this accurately, medical devices should be safely removed to visualise the skin beneath the device, with Stephen-Haynes and Maries (2020) further adding that dressings should also be removed regularly to check and record the skin condition.

Moore et al (2020) and Gefen et al (2020) highlighted the importance of skin care. The skin should be cleaned and moisturised following local policies regarding products, which should have their efficacy audited, as there are currently no specific methods or products demonstrated to have greater effectiveness. Any dressing used for PU prevention needs to allow effective release of heat, otherwise they may promote perspiration, leading to an increase of PU risk or an exacerbation of existing tissue damage (Gefen and Ousey 2020b).

ETTs, nasogastric tubes, oxygen and NIV masks and their tubing have changed little in decades, despite being in contact with vulnerable skin and soft tissue and frequently associated with PUs (Gefen et al, 2020). Gefen et al identified several causes related to the device design that increased the risk of device-related PUs, such as generic designs with stiff material not conforming to tissue shape and inadequate guidance provided on device application. Challenges to overcome include the fact that medical devices have a diagnostic or therapeutic purpose, therefore it may not be possible to remove or reposition them (Gefen et al, 2020). However, the evidence suggests that there are some short-term solutions to reduce device-related PUs. The UK ICU guidelines for putting patients in prone position developed by Bamford et al (2019) suggests ensuring, before and after turning a patient, that they are not lying directly on a medical device, particularly that the ETT is not pressing on the corner of the mouth/lips and the nasogastric tube is not pressing the nostril, while also suggesting that fluidised positioners could aid this by creating a channel for the ETT. However, the use of a fluidised positioner is suggested without citing any evidence and this requires close monitoring if used with the patient in prone position owing to the risk of suffocation (Mölnlycke Healthcare, 2019). Capasso et al (2020) and Bamford et al (2019) also suggested replacing commercial tube holders with ties as a way to minimise device-related PUs.

Regarding dressings, Peko et al (2020) compared a 3D computational head model with Mepilex Border Flex, a multi-layered silicone-foam dressing and a control head model without any dressing, identifying a 50% reduction in pressure forces. However, the study was funded by Mölnlycke Health Care, the manufacturer of the dressing, and the findings therefore need to be viewed with caution. Gefen and Ousey (2020b) suggested that polymeric membrane dressings allowed better clearance of heat when compared with conventional dressings, however, this was based on computational studies and no further research has been undertaken by other authors.

Studies undertaken on patients showed that hydrocolloid and film dressings significantly reduced facial PUs when compared with a control group and are widely recommended (Gefen et al, 2020; Moore et al, 2020; Stephen-Haynes and Maries, 2020). Moore et al (2020), in their clinical review, identified that silicone pressure-reducing strips placed under the tie that secures the ETT reduced device-related PUs: 20.7% of patients with an ETT in situ without silicone strips under the tie developed a PU, versus 5.2% of patients with silicone strips under the ETT tie. Stephen-Haynes and Maries (2020) supported the use of these strips, highlighting that they work well, but they do not offer any absorbency.

Only Gefen et al (2020) discussed education and communication in the articles reviewed and, based on international consensus, it is suggested that ICUs should develop local strategies around these areas. It is important that standard procedures are in place covering device selection, application (including tapes and fixation methods), and that these standard procedures are documented and available to all health professionals. Within each facility there should be a nominated champion to develop and distribute standardised procedures, ensure compliance, and the correct documenting of all device-related PUs. These champions should also suggest design changes to manufacturers in order to reduce the risk of device-related PUs (Gefen et al, 2020). Risk awareness for patients, caregivers and health professionals is also important, for example, by highlighting the risks posed by personal possessions, encouraging patients to inform health professionals of any discomfort or pain at the device site, informing health professionals of any objects left between the patient and support surface, moving or adjusting the device if there are signs that the patient is in discomfort or pain, taking proactive actions to minimise the risk of device-related PUs and undertaking regular skin assessments (Gefen et al, 2020).

Overall, this study was limited by the low quality of evidence found; no RCTs undertaken in patients with COVID-19 relating to facial PUs had been undertaken at the time of this study. Moreover, some studies were based on computational modelling and sponsored by dressing companies and their findings must therefore be interpreted with caution until more robust research is undertaken. However, the evidence base has provided insight into recommends for practice, which are outlined in Table 2.

This study aimed to collect all available evidence related to COVID-19 patients who experience facial PUs within an ICU setting and discovered the high incidence of such PUs due to the wide range of factors. The adverse consequences of these PUs requires application of the prevention measures identified within the literature in order to minimise facial pressure ulcers in patients undergoing prone positioning.

KEY POINTS

• This article analysed evidence around facial pressure ulcers developed on patients with COVID-19 when undergoing prone positioning in the intensive care unit (ICU)
• Facial pressure ulcers have a serious physiological, psychological and social impact on patients and place a financial burden on the healthcare system
• Recommendations based on the evidence include cleaning and moisturising the skin, covering the eyes with gel pads or similar, replacing endotracheal tube holders with ties and using hydrocolloid or film dressings on facial pressure points
• It is recommended that ICUs should develop policies and strategies around skin assessment, skin condition, pressure redistribution surfaces, eyes, education, devices used and prophylactic dressings

CPD reflective questions

• Analyse the current preventive measures in your clinical area and how they align with the recommendations highlighted in this case study. Reflect on the differences and possible reasons for any differences
• How could you reduce the number of facial pressure ulcers in your clinical area?
• Consider what strategies you could use to communicate the recommendations from the evidence in this study to your clinical colleagues