Effectiveness of a multi‐layer foam dressing in preventing sacral pressure ulcers for the early acute care of patients with a traumatic spinal cord injury: comparison with the use of a gel mattress

Abstract

Individuals with spinal cord injury are at risk of sacral pressure ulcers due to, among other reasons, prolonged immobilisation. The effectiveness of a multi‐layer foam dressing installed pre‐operatively in reducing sacral pressure ulcer occurrence in spinal cord injured patients was compared to that of using a gel mattress, and stratified analyses were performed on patients with complete tetraplegia and paraplegia. Socio‐demographic and clinical data were collected from 315 patients admitted in a level‐I trauma centre following a spinal cord injury between April 2010 and March 2016. Upon arrival to the emergency room and until surgery, patients were transferred on a foam stretcher pad with a viscoelastic polymer gel mattress (before 1 October 2014) or received a multi‐layer foam dressing applied to their sacral‐coccygeal area (after 1 October 2014). The occurrence of sacral pressure ulcer during acute hospitalisation was similar irrespective of whether patients received the dressing or the gel mattress. It was found that 82% of patients with complete tetraplegia receiving the preventive dressing developed sacral ulcers as compared to only 36% of patients using the gel mattress. Although multi‐layer dressings were suggested to improve skin protection and decrease pressure ulcer occurrence in critically ill patients, such preventive dressings are not superior to gel mattresses in spinal cord injured patients and should be used with precaution, especially in complete tetraplegia.

Introduction

Pressure ulcers (PUs) are an important and frequent secondary complication related to spinal cord injury (SCI), occurring in 18–37% of traumatic SCI patients during acute hospitalisation 1, 2. The occurrence of PU has devastating effects on patients with SCI as it may lead to severe medical complications, such as local and systemic infections, and death in 8% of cases 3. It may also interfere with the rehabilitation process and community reintegration 4.

Individuals with traumatic SCI are particularly at risk of PU due to prolonged immobilisation, moisture exposure related to incontinence problems, friction and shear forces associated with inappropriate transfers and mobility, as well as abnormal microvascular blood flow secondary to a disrupted autonomic function below the level of injury. These result in ischaemia of underlying soft tissues and development of PU 2, 4, 5, 6, 7. The acute care hospitalisation is a critical period for the development of PU following a traumatic SCI, since the neurological impairment is at its peak, individuals may have other traumatic injuries and could also develop nutritional deficits 8, 9, 10. Complete tetraplegia (as opposed to incomplete SCI and paraplegia), and factors prolonging the pre‐operative period where spinal stability is still compromised are also known to increase the risk of developing PU in patients with a SCI 1, 4, 11, 12.

While the occurrence of PU is known to increase the cost of treatment in SCI 13, a cost‐effectiveness study showed that hospital‐acquired PU prevention programmes are cost saving and provide better outcomes to patients 14. Prevention strategies are generally divided into five major areas: risk assessment, management plan, support surface, repositioning and education 15, 16. Latimer et al. 16 showed important gaps between planning and implementation of strategies pertaining to risk assessment, regular repositioning and patients' education. This is also the case at our facility, where PU risk assessment and information pertaining to the patients' repositioning are not thoroughly documented in the medical charts. While education likely remains the best and most powerful ally in PU prevention 17, patients' collaboration is also a factor that should be considered as they are not always compliant and often refuse to be repositioned and have their skin integrity assessed.

Recent studies have focused on the use of prophylactic dressing application as a strategy to prevent hospital‐acquired PU 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28. Those studies were conducted in critically ill patients hospitalised in an intensive care unit (ICU) for respiratory, coronary or gastrointestinal diseases, falls, infections or cancer and at high‐risk of developing PU, but excluded individuals with SCI. They suggested that the application of a preventive dressing on the sacral region reduced the occurrence of PUs 18, 20, 23. When SCI population was included, they represented such a small proportion of the study population that it was not possible to draw any valid conclusion of the effectiveness of the dressing in those specific cases 19, 20, 21, 22, 23, 25. From the very few studies including patients with a SCI, Walsh et al. 26 showed that only 4 of the 69 patients admitted to the ICU who received a silicone border foam dressing to the sacral area had a SCI, but no details were provided on the severity or level of SCI nor on whether they developed a PU or not. There is thus a lack of information on the benefits that could be associated with the use of preventive sacral dressing in SCI patients.

Therefore, the purpose of this study was to determine the effectiveness of a multi‐layer foam dressing installed before surgical spine stabilisation for reducing PU occurrence in the sacral area in patients who sustained a traumatic SCI, as compared to the effectiveness of using a gel mattress pre‐operatively.

Methods

Patients

This is a retrospective study conducted on a prospective cohort of 315 consecutive patients who sustained a traumatic SCI and were admitted in a single level‐I SCI‐specialised trauma centre between 1 April 2010 and 31 March 2016 (248 males and 67 females; mean age ± standard deviation: 48·6 ± 19·3 years). Patients entered the cohort at the time of admission and were followed up until discharge from the acute care centre. They were included if they sustained a spine trauma that involved a SCI above the L1‐L2 intervertebral disc and had surgery performed in our institution. The study was approved by the institutional review board and all patients were enrolled on a voluntary basis.

Intervention

Protocols for prevention of pressure ulcers at our institution aim at maintaining skin integrity, decrease pressure, shear and friction, control moisture level and optimise nutrition, hydration and mobility. Prior to 1 October 2014, standard care to prevent PU in SCI patients involved transfer on a foam stretcher pad with a viscoelastic polymer gel mattress (Blue Cloud™; Batrik Medical Manufacturing, Montreal, Canada) upon arrival at the emergency room until spine stabilisation surgery. Patients also underwent log roll mobilisation once every 2 hours during the pre‐operative period. In the post‐operative period, patients were cared for on a therapeutic, low air loss pressure‐relieving mattress (Versacare A.I.R.® Surface; Rom‐Hill, Mississauga, Canada), with regular repositioning (every 2 hours) and skin care/assessment.

Since 1 October 2014, all patients with suspected or confirmed SCI systematically had preventive multi‐layer foam dressing applied to their sacral‐coccygeal area upon arrival in the emergency room until surgery (pre‐operative period). This self‐adherent multi‐layer foam dressing is advertised as a dressing redistributing shear forces and pressure, reducing friction and balancing the microclimate during wear time, while maintaining an optimal moisture level. Since immobilisation is preconised in SCI patients in the period prior to spine stabilisation surgery 9, the rationale for implementing the multi‐layer foam dressing protocol in the pre‐operative period was to benefit from the dressing's potential protective properties on the sacral area and to limit friction and shear forces by avoiding transfer of the patients on a gel mattress. The sacral area was selected because it is the most frequent area involved in PU development for SCI patients during acute care 13, 29, 30. The attending nurse would also reposition the dressing after skin assessment every 8 hours if needed. The dressing was replaced by a new one if soiled in the pre‐operative period. The dressing was removed during surgery and not put back in place afterwards. Because the gel mattress was not used for patients receiving the multi‐layer foam dressing, local gel pads were placed under the heels and occiput for these patients in order to prevent PU at these areas.

Basic PU prevention protocol, including periodic log roll mobilisation and skin assessment, was similar for all patients. They were also followed up by a specialised multidisciplinary rehabilitation team (including physical therapy and nutrition). Patients enrolled in this study were sub‐grouped on the basis of the local PU prevention method used for sacral‐coccygeal area: Group 1 gel mattress for 226 consecutive patients admitted between 1 April 2010 and 30 September 2014 versus Group 2 multi‐layer foam dressing for 89 consecutive patients admitted from 1 October 2014 to 31 March 2016.

Data collection

Socio‐demographic and clinical data were collected prospectively and updated on a daily basis during acute care hospitalisation. Collected data included age, gender, smoking status (active smoker or past/non‐smoker), obesity defined as a body mass index ≥30·0, co‐morbidities quantified by the Charlson co‐morbidity index (CCI) and trauma severity measured by the injury severity score (ISS). The neurological level was used to discriminate between tetraplegia (C1‐C8) and paraplegia (T1‐L1). The severity of the SCI was assessed upon arrival at the SCI centre and was reported using the International standards for neurological classification of SCI (ISNCSCI) by grades A–D 27. The presence of a concomitant traumatic brain injury (TBI), defined as a non‐degenerative insult to the brain caused by an external mechanical force leading to an altered or diminished conscious state according to the Glasgow coma scale, was determined upon arrival at the SCI centre. Since the great majority of TBI were of mild severity, we did not consider the severity of TBI in our analyses. The transfer delay (time interval between trauma and arrival at the SCI centre, in hours), as well as the surgical delay (time interval between trauma and time of skin incision for surgery, in hours) were compared. The length of stay in acute care (days) was also considered.

Outcome measures

The main outcome measure was the occurrence of sacral PU developed during acute hospitalisation. This information was collected prospectively in a document pertaining to the evaluation, observation and treatment of all types of wounds (nosocomial, diabetic, traumatic, pressure and surgical) that is systematically included in all medical charts. The severity of the sacral PU was considered as a secondary outcome. PUs located on the sacrum, coccyx and/or gluteal cleft were included in the analyses as sacral PUs. PU development and staging was based on the clinical practice guidelines from the NPUAP (National Pressure Ulcer Advisory Panel) 28. PUs were categorised as stages 1–4, suspected deep tissue injury (SDTI) or unstageable.

Statistical analyses

Continuous data were compared between Groups 1 and 2 using Student's t‐tests, while categorical data were compared using chi‐square tests. Continuous data were reported as mean ± 1 standard deviation (SD). The primary outcome, i.e. occurrence of PU, the secondary outcome, i.e. severity of PU, as well as the other categorical covariables were compared between both groups using chi‐square tests, and reported as proportions and percentages. We determined potential predictors of sacral PU by investigating 12 potential predictors in a multi‐variate logistic regression: (i) age; (ii) gender; (iii) pre‐operative severity of neurological deficit (ISNCSCI grades A–D); (iv) neurological level of injury – NLI (tetraplegia versus paraplegia); (v) ISS; (vi) presence of concomitant TBI; (vii) transfer delay; (viii) surgical delay ; (ix) obesity; (x) CCI; (xi) use of preventive PU dressing and (xii) smoking status. Since the level (tetraplegia versus paraplegia) and severity (complete versus incomplete) of the SCI are recognised to be important risk factors of PU, stratified analyses were then performed for individuals with complete tetraplegia and complete paraplegia. All statistical analyses were performed using the IBM SPSS Statistics 21 software (Chicago, IL, USA), and results were considered statistically significant when P < 0·05.

Results

All patients

Our cohort included 315 patients, where 226 received a gel mattress (Group 1) and 89 received multi‐layer foam dressing (Group 2). Comparison of socio‐demographic and clinical characteristics is shown in Table 1. Groups 1 and 2 were similar in terms of age, gender, burden of co‐morbidity, surgical approach and smoking status. They sustained traumas of similar severity, and approximately half of the patients from both groups had a concomitant TBI. The proportion of tetraplegic and paraplegic patients was the same in Groups 1 and 2. The majority of patients were tetraplegic in both groups. Moreover, the severity of the neurological deficit, as evaluated by the ISNCSCI grade, was also similar between Groups 1 and 2, although patients who received the multi‐layer foam dressing (Group 2) tended to have less severe neurological deficits than patients who only received standard PU prevention care (Table 1). The average delay between trauma and arrival at the SCI centre, as well as the average surgical delay were similar between Groups 1 and 2. The average acute care length of stay was also not different, with approximately 1 month in both groups.

Table 1.

Socio‐demographic and clinical characteristics for the total cohort of patients with SCI based on the pressure ulcer prevention care received

		Pressure ulcer prevention care
		Gel mattress (Group 1)	Multi‐layer foam dressing (Group 2)	P
N		226	89	—
Age	Mean	47·8	50·7	0·23
	SD	19·4	18·3
Gender	Male	81·0	73·0	0·12
	Female	19·0	27·0
ISS	Mean	23·8	22·4	0·24
	SD	10·6	5·1
NLI	Tetraplegia	56·2	60·7	0·47
	Paraplegia	43·8	39·3
ISNCSCI grade (%)	A	38·1	24·7	0·11
	B	10·8	11·2
	C	15·7	16·9
	D	35·0	47·2
TBI	% TBI	46·4	50·0	0·63
Obesity	% Obese	15·0	18·2	0·50
Transfer delay (hours)	Mean	71·9	61·8	0·76
	SD	301·8	122·3
Surgical delay (hours)	Mean	98·4	81·8	0·63
	SD	309·7	135·5
CCI	Mean	0·37	0·43	0·60
	SD	0·79	0·95
Smoking status	% Active smoker	25·7	28·1	0·62
Length of stay (days)	Mean	30·8	26·2	0·11
	SD	25·0	16·9

CCI, Charlson comorbidity index; ISNCSCI, International Standards for Neurological Classification of Spinal Cord Injury; ISS, injury severity score; NLI, neurologic level of injury; SCI, spinal cord injury; SD, standard deviation; TBI: traumatic brain injury.

Table 2 shows the occurrence and severity of PU for the total cohort of patients. There were no significant differences in the proportion of patients in Groups 1 and 2 developing a PU during acute care hospitalisation (P = 0·77), nor were there any differences in the severity of PUs (P = 0·71). PUs were observed in 40 patients from Group 1 (17·7%) and 17 patients from Group 2 (19·1%). More than half of the PUs were stage 2 severity for both groups. There was only one patient with a stage 3 PU and two patients with a stage 4 PU in Group 1, while no patient from Group 2 developed a PU staged 3 or 4 (Table 2).

Table 2.

Occurrence and severity of sacral pressure ulcer (PU) for the total cohort of patients with spinal cord injury based on the preventive care receive

		Pressure ulcer prevention care
		Gel mattress (Group 1)	Multi‐layer foam dressing (Group 2)	P
Occurrence of sacral PU (%)		17·7	19·1	0·77
Severity of PU (%)	Grade 1	30·0	29·4	0·71
	Grade 2	62·5	70·6
	Grade 3	2·5	0·0
	Grade 4	5·0	0·0

Using a multivariate logistic regression model to identify potential predictors of PU of any stage during acute hospitalisation, the results showed that ISNCSCI‐A injuries, tetraplegia, older age and higher trauma severity were significantly associated with sacral PU occurrence (Table 3). Patients with a complete ISNCSCI‐A injury were seven times more likely to develop a PU than individuals sustaining an ISNCSCI‐D SCI. Tetraplegia was also an important predictor, since patients with a cervical SCI were nearly three times more likely to have a sacral PU as compared to paraplegic patients. This model was statistically significant (χ ² = 50·359; P < 10⁻³) and explained 27% (Nagelkerke's R) of the variance in sacral PU.

Table 3.

Factors associated with sacral pressure ulcer development for the total cohort of spinal cord injury patients: multivariate logistic regression results

Factors	Beta	OR (95% CI)	P
Age	0·038	1·039 (1·018–1·061)	<10−3
ISS	0·044	1·045 (1·003–1·088)	0·033
Tetraplegia	0·972	2·644 (1·194–5·853)	0·017
Paraplegia	Reference category
ISNCSCI A	1·959	7·089 (2·370–21·205)	<10−3
ISNCSCI B	0·991	2·693 (0·757–9·574)	0·126
ISNCSCI C	0·484	1·623 (0·475–5·548)	0·440
ISNCSCI D	Reference category

CI, Confidence interval; ISNCSCI grade, International Standards for Nclassification of Spinal Cord Injury; ISS, injury severity score; OR, odd ratio. Bold numbers represent significant results.

Stratified analysis: patients with a complete ISNCSCI‐A tetraplegia and paraplegia

The cohort included 44 patients with complete ISNCSCI‐A tetraplegia (33 received a gel mattress; 11 had a multi‐layer foam dressing) and 64 with complete ISNCSCI‐A paraplegia (53 received a gel mattress; 11 had a multi‐layer foam dressing). Complete tetraplegic patients who received a gel mattress were 11 years older than those who had a multi‐layer foam dressing, although this difference did not reach statistical significance (P = 0·06). Moreover, all patients with complete tetraplegia who received a multi‐layer foam dressing had a mild concomitant TBI, while this proportion was only 58% for patients who received a gel mattress (P = 0·03). All other socio‐demographic and clinical characteristics were similar in both sub‐groups of complete tetraplegic patients. Patients with complete paraplegia who received a gel mattress were similar to complete paraplegic patients who had a multi‐layer foam dressing with respect to their socio‐demographic and clinical characteristics, except for the burden of co‐morbidities which was higher for patients who received a gel mattress. Moreover, paraplegic patients who received a gel mattress were transferred to the SCI centre and had surgery earlier than paraplegic patients who received a multi‐layer foam dressing (Tables 4 and 5).

Table 4.

Socio‐demographic and clinical characteristics of patients with a complete tetraplegia or complete paraplegia based on the pressure ulcer prevention care received

		Complete tetraplegia			Complete paraplegia
		Gel mattress	Multi‐layer foam dressing	P	Gel mattress	Multi‐layer foam dressing	P
N		33	11	—	53	11	—
Age	Mean	42·3	53·6	0·06	40·5	47·7	0·21
	SD	16·6	14·5		17·2	18·1
Gender	Male	69·7	72·7	0·85	88·7	90·9	0·83
	Female	30·3	27·3		11·3	9·1
ISS	Mean	34·0	25·1	0·11	30·7	29·2	0·48
	SD	17·8	3·7		6·5	6·8
TBI	% TBI	57·6	100·0	0·03	50·9	40·0	0·53
Obesity	% Obese	21·2	9·1	0·37	20·2	16·7	0·63
Transfer delay	Mean	17·7	50·4	0·14	17·8	112·0	0·001
	SD	39·6	108·3		25·4	201·6
Surgical delay	Mean	34·9	54·4	0·49	27·2	157·6	<10−3
	SD	69·9	107·4		31·4	222·6
CCI	Mean	0·3	0·6	0·20	0·2	0·7	0·02
	SD	0·8	0·9		0·5	1·0
Length of stay (days)	Mean	46·6	27·3	0·07	34·1	35·5	0·89
	SD	32·2	21·0		32·2	16·5
Smoking status	% Active smoker	21·2	9·1	0·24	35·8	18·2	0·43

CCI, Charlson comorbidity index; ISS, injury severity score; SD, standard deviation; TBI, traumatic brain injury.

Table 5.

Occurrence and severity of sacral pressure ulcers (PU) of patients with a complete tetraplegia or complete paraplegia based on the PU prevention care received

		Complete tetraplegia			Complete paraplegia
		Gel mattress	Multi‐layer foam dressing	P	Gel mattress	Multi‐layer foam dressing	P
Occurrence of sacral PU (%)	PU	36·4	81·8	0·009	20·8	27·3	0·63
Severity of PU	Grade 1	16·7	44·4	0·30	54·5	0·0	0·15
	Grade 2	75·0	55·6		36·4	100·0
	Grade 3	8·3	0·0		0·0	0·0
	Grade 4	0·0	0·0		9·1	0·0

Sacral PU developed in 9 (82%) complete tetraplegic patients with a multi‐layer foam dressing, while 21 (64%) complete tetraplegic patients who received a gel mattress (N = 21; P = 0·009) developed a PU during the course of acute hospitalisation. The severity of PU was not significantly different between patients with complete tetraplegia whether they received a gel mattress or a multi‐layer foam dressing, although a grade 3 PU developed in one patient receiving gel mattress while there were none developed in patients with a multi‐layer foam dressing. Whether they received a gel mattress or a multi‐layer foam dressing, patients with complete paraplegia developed sacral PUs in similar proportions (20·8% versus 27·3% for gel mattress and multi‐layer foam dressing, respectively; P = 0·63).

The multivariate logistic regression model revealed that patients with complete tetraplegia were nearly 18 times more likely to develop a PU if they received a multi‐layer foam dressing (odds ratio = 17·718). The severity of trauma (ISS) was also associated to greater odds of developing a sacral PU (odds ratio = 1·063). This prediction model was significant (χ ² = 19·995; P < 10⁻³) and predicted 37% (Nagelkerke's R = 0·374) of variance in the development of sacral PU. None of the covariables entered in the multi‐variate logistic regression model were found to be significant predictors of sacral PU development in complete paraplegia (Table 6).

Table 6.

Factors associated with sacral pressure ulcer (PU) development in patients with a complete tetraplegia and complete paraplegia: multivariate logistic regression results

	Predictors of sacral PU	Beta	OR (95% CI)	P
Complete tetraplegia	Multi‐layer foam dressing	2·875	17·718 (1·766–177·798)	0·015
	Gel mattress	Reference category
	ISS	0·061	1·063 (1·008–1·122)	0·026
Complete paraplegia	No covariable was significantly associated with the occurrence of sacral PU in patients with complete paraplegia

ISS, Injury severity score; OR, odds ratio.

Discussion

This study aimed at examining the effectiveness of a multi‐layer foam dressing applied to the sacral region as compared to transfer on a gel mattress in preventing pressure ulcers in patients with a traumatic SCI upon arrival at a level I SCI specialised trauma centre for the period prior to spine surgery. Nearly 19% of patients developed a PU during the acute care hospitalisation, which is within the range reported in the SCI literature 4, 31. Results of this study suggest that using a multi‐layer foam dressing or a gel mattress pre‐operatively provide a similar effectiveness in preventing PUs in those with SCI, suggesting that the preventive dressing failed to be superior to the gel mattress in preventing sacral PUs. However, although the severity of SCI was not different between both sub‐groups, patients receiving the multi‐layer foam dressing tended to have less severe SCI. Although less severe SCI is known to reduce the occurrence of PUs 1, the multi‐layer was not superior in preventing PUs in our study.

Results of the logistic regression model for the total cohort of patients allowed identification of SCI patients at higher risk of developing PU during acute care, namely patients with a complete SCI and tetraplegia, which is in accordance with previous studies 9. Older age and increased burden of traumatic associated injuries were also predictors of PU, however to a lesser impact, as shown by the low odd ratios. It is then recommended that these patients, particularly individuals sustaining a complete tetraplegia should be monitored more closely towards the prevention of PUs.

Although it has been suggested that preventive dressings might redistribute shear and friction forces, provide an optimal skin microclimate and therefore represent an efficient tool in preventing PUs 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 32, our results showed that the application of a preventive dressing was rather associated with increased occurrence of PUs in individuals with complete tetraplegia. In addition to being immobilised for prolonged periods while experiencing the most severe neurologic and functional impairments, individuals with complete SCI also generally present severe incontinence problems during acute care hospitalisation, resulting in an increased risk of developing PUs 7. The negative consequences of prolonged immobilisation and incontinence problems on skin integrity may be prevented by regular assessment of the skin, repositioning and weight shifts, as well as meticulous sphincter management 9, 13, 17, 33, which are recognised as being the most important measures to prevent PUs 9, 34. However, most individuals with severe SCI are unable to complete independently these measures during acute care and will develop this capacity only during intensive functional rehabilitation. They therefore need to rely on the nursing staff, who must adopt a very proactive behaviour to ensure appropriate and efficient prevention of PUs, upon admission 9, 35. The use of a multi‐layer dressing as a preventive measure of PUs in complete tetraplegic patients could also have led to a sense of false security, resulting in decreased adherence to the essential preventive measures of PUs that are usually provided, such as frequent repositioning and assessment of skin integrity. This could partly explain the increased occurrence of sacral PUs in complete tetraplegic patients with the use of the multi‐layer dressings. Results of this study thus reinforce the importance of meticulous preventive care, which is considered as the key factor for PU prevention, and cannot be replaced by any kind of dressing 15, 17.

Complete tetraplegic patients receiving the multi‐layer foam dressing were older, had less associated injuries and a shorter duration of acute care hospitalisation. Although the differences did not reach statistical significance, these factors still represent potential confounding variables. Indeed, older age is recognised as a risk factor of PU. Loss of dermal vessels, thinning of the epidermis and loss of elastic fibres associated with ageing increase susceptibility to skin breakdown 35. Individuals with preventive dressing were approximately 11 years older than individuals treated with a gel mattress. However, age was not a predictive factor of PUs in our regression model for complete tetraplegia.

On the other hand, the number and severity of associated injuries was higher in the complete tetraplegic patients receiving the mattress gel (Group 1), which might have prolonged the immobilisation period and further increased the risk of PUs during acute care 36. Accordingly, the ISS score was also shown to be a significant predictor of sacral PUs (odds ratio: 1·06) in our study. However, applying the multi‐layer foam dressing did not result in a decrease rate of PUs in complete tetraplegic patients despite a decreased ISS score, further emphasising the lack of benefit from applying this preventive dressing in complete tetraplegic patients. Additionally, the rate of TBI was significantly higher in the preventive dressing group but is unlikely to have influenced the rate of PU development. TBI reported in the preventive dressing group were all of mild severity, so less likely to have caused serious additional neurologic impairments that may have increased vulnerability to PUs. Headaches, dizziness and/or fatigue associated with acute mild TBI could potentially influence the rehabilitation process 37, but TBI was not revealed as a significant predictor of PU in the regression analysis.

Finally, results showed a tendency towards a longer acute care length of stay in patients who received the preventive dressing. The interpretation of this result is twofold. Indeed, the occurrence of PU may lead to prolonged hospitalisation stay, while extended hospitalisation stay is also recognised as a risk factor of developing PU 35. Patients with complete tetraplegia receiving the preventive dressing did not develop more severe PU, although they presented a higher rate of sacral PUs than patients with the gel mattress. With regard to the resources utilisation as related to the length of stay, the use of the multi‐layer foam dressing as a preventive measure did not provide a significant benefit, especially in patients with complete tetraplegia who required the longest length of hospitalisation.

Limitations

The main limitation of this study is its retrospective nature. Indeed, a prospective study would have allowed a better identification of factors leading to our results. The delay for PU appearance was not available, which would have helped in evaluating the efficacy of the preventive dressing during the pre‐operative period. Further studies should investigate and document in a more robust and rigorous manner the exact moment the PU is noticed as well as when the dressing is removed.

Moreover, nutritional deficits, bladder and bowel dysfunction resulting in skin moisture, hypoalbunemia and anaemia 1, 4, 38, 39 have also been linked to PU after a SCI. These have not been considered in the present study and could possibly be different between patients who received the gel mattress and those who received the multi‐layer foam dressing. However, our facility comprises a specialised multidisciplinary rehabilitation team including a dietician, which follows the same protocol and procedures pertaining to these factors. Our results would unlikely be affected or biased by any difference in one of these elements. Different adherence to other preventive measures such as repositioning and skin assessments could also have influenced the results as mentioned previously. However, this study was performed under real clinical conditions in a specialised SCI centre in which the health care providers are practising under the same standards and did not bring any modifications to their usual practice

As mentioned above, the Braden score was not considered since it was not available for all patients. However, the great majority of the Braden scale items mainly depend on the level and severity of the SCI, which was accounted for in our analyses.

Conclusions

Although pre‐operative use of multi‐layer dressings was suggested to improve the protection of the skin and to decrease the occurrence of PU in critically ill patients, this study suggests that such preventive measure for patients sustaining acute SCI is not superior to the conventional use of a gel mattress and that multi‐layer dressings should be used with precaution in SCI patients, particularly in complete tetraplegia. Early and rigorous assessment of PU risk factors, the use of standardised checklists to systematically apply all preventive measures and optimisation of education is then highly recommended when using preventive dressing in SCI patients.