Interface Pressures of New and Worn Standard and Viscoelastic Hospital Mattresses: A Comparative Study

Pressure ulcers are a target area for reduction of adverse events in the Norwegian patient safety program. A pressure injury/ulcer is localized damage to the skin and/or underlying tissue as a result of pressure, or pressure in combination with shear.¹ Pressure ulcers are costly both for the patient and for society.² Patients in whom a pressure ulcer develops during their hospitalization have longer length of stay, greater likelihood of readmission within 30 days, and higher risk of dying during their stay.³ Moreover, prevention is less expensive than treatment.² A pressure ulcer may cost nearly 500 Euros per patient per day depending on the severity of the pressure ulcer and the care setting.² Most pressure ulcers are preventable if preventive measures, such as pressure-redistribution mattresses or cushions, planned repositioning in bed or chair, and offloaded heels in bed, are implemented.¹ There is evidence that a pressure-redistribution mattress is better than a standard hospital mattress for pressure ulcer prevention,^4,5 and clinical practice guidelines recommend that all patients at risk or with a pressure ulcer should receive a pressure-redistribution mattress.¹ However, there is no evidence to conclude that one type of pressure-redistribution mattress is better than others.^4,6 Further, a better mattress does not have the desired effect if the patient does not receive it early enough. A previous Norwegian cross-sectional study (N = 1209) revealed that only 135 of 305 patients (44.3%) at risk for or with a pressure ulcer had a pressure-redistribution mattress; the rest were positioned on a standard hospital mattress.⁷ Furthermore, hospital mattresses in Norway have been found to be used for up to 10 years.^8,9 Few studies have looked into how wear and tear affects the pressure-redistribution qualities of mattresses over time. A prospective comparison study that examined 4 types of foam replacement hospital mattresses (5 mattresses of each type) was published in 1998. The study showed that of the 25 included mattresses, only 10 were available for inspection. Of these, only 1 was still usable after 3.5 years of daily use.¹⁰ It has been suggested that the expected lifespan of properly used and maintained mattresses is 2 to 7 years.¹¹ 

High interface pressure may contribute to pressure ulcer development.¹² High patient turnover requiring frequent mattress cleaning, and high occupancy rates resulting in daily use, can reduce the usable device life of a mattress. 

The purpose of this comparative study was to investigate interface pressures in a sample of new and worn hospital standard mattresses and pressure-redistribution mattresses. Participant comfort and mobility level using new versus worn standard and pressure-redistribution mattresses were also examined. Mean interface pressure as well as the comfort and the ease of turning oneself on the selected mattresses were evaluated.

The study was conducted using a convenience sample of mattresses that were in regular use in 1 hospital in Norway between July and August 2018. The following mattresses, with the year they were put into use in parentheses, were included: 

• standard compact foam hospital mattress (2011) 

• standard compact foam hospital mattress (2014) 

• standard compact foam hospital mattress with egg-crated surface (2018) 

• new viscoelastic overlay (2018) placed on a standard hospital mattress from 2014

• worn viscoelastic full mattress (no date available). 

The standard mattresses had a thickness of 13 cm, a weight limit of 120 to 130 kg, and density of 35 kg/m³. The manufacturer of all standard mattresses was Puls Brann (Norway); since 2018, the standard foam has an egg-crate surface layer.

Both viscoelastic mattresses had 7-cm thick viscoelastic foam with open cells and a high density of 85 to 110 kg/m³. The mattresses had a polyurethane shield and a weight limit of 100 kg. The full viscoelastic mattress had an 8-cm, high-quality foam layer as well. Both viscoelastic mattresses used in this study were manufactured by Tempur MED (Denmark).

Volunteer sample. Using convenience sampling methods, including mail and posters, 20 hospital employees were recruited to participate. All were healthy, mobile volunteers, and the only exclusion criterion was body temperature above 37.5°C. 

All mattresses were tested with volunteers lying in the flat supine position for 10 minutes on all 5 surfaces. The order of mattress selection for each participant was random. The website www.randomization.com was used for the random order, and each order was put in a sealed envelope. Most participants were dressed in their work uniform but some were dressed in pants and a T-shirt.

The participants and the researcher were not blinded due to the nature of the study.

Variables/measures. The researcher used a paper form to document the room temperature in Celsius, participant characteristics (sex, age in years, height in centimeters, weight in kilograms, body temperature in Celsius, oxygen saturation [%] measured using the Mindray VS 900, and blood pressure measured by Mindray VS 900 [China] in mm Hg). These measurements were obtained before the participants laid down on the mattresses. Software for the pressure mapping mat was installed on the researcher’s computer, and the interface pressure (mm Hg) and sensing area (cm²) in the supine position was measured at baseline (0 minutes) and after 10 minutes. After 10 minutes, participants completed 1 question for the comfort and 1 question for the mobility levels for each mattress. The comfort and mobility level questions were answered using a 5-point Likert scale (1 = very poor and 5 = very good). 

Pressure mapping system. The Force Sensing Array pressure mapping bed system from BodiTrak (Vista Medical, Canada) was used to measure pressure distribution of the participant’s body. The mat was of bed size (193 × 76 cm). The mapping system measured pressure from 0 to 100 mm Hg. The mat had 1809 sensors (67 × 27). Pressure above 90 mm Hg in an area indicated risk of pressure ulcer.¹³ Pressure mapping is an objective, noninvasive method to measure interface pressure and pressure redistribution. 

Procedures. The pressure mat was placed on top of the mattresses covered with a thin loose sheet. All participants were allowed to have a pillow under their head for comfort. Five (5) minutes is required before a measure should be done to allow for the person’s acclimation.¹⁴ After the 10-minute pressure mapping session, participants were asked to move from side to side to assess their mobility. The researcher was present at all times and used a stopwatch app on a mobile phone as a timer. 

Ethics. The study was approved by the hospital’s privacy protection officer. All participants received oral and written information about the purpose of the study and provided written informed consent. Participant anonymity and confidentiality was maintained by aggregating the data. The list of participants’ names was shredded when the data collection was done. The study involved minimal risk of harm as all participants were healthy volunteers and laid on the mattresses for a short period.

Statistics. All data were entered into Excel 2013 (Microsoft, USA) and SPSS version 25 (IBM, USA). Descriptive statistics of participants’ characteristics and sensing area (cm²) (which refers to the area of the mat whose readings are at or above 0) and mean interface pressure (mm Hg) were entered, and the sensor’s data from scanning were exported from the software into Excel and SPSS. The mean numbers of cells reaching 60 mm Hg, 80 mm Hg, 90 mm Hg, and 100 mm Hg after 10 minutes were calculated. The authors eliminated pressure values of 0 to avoid underestimation of pressure. Nonparametric statistics method was used due to the small sample. Kruskal-Wallis test was used to compare type of mattress with mean interface pressure and mean number of cells (60–100 mm Hg). Spearman correlation was used to examine associations between comfort level and ease of mobility. The correlation value ranged from ± 1.00, and large strength of relationship was considered for r = .50 to 1.0.¹⁵ The Mann-Whitney U test was used to compare mean interface pressure between persons weighing more than and less than 100 kg.

Most participants were female (75%), and the age range was from 25 to 64 years (average 41.3 ± 12.25). The sample had a range of body types; weight varied from 55.1 to 113.5 kg, height from 161 to 187 cm, and body mass index (BMI) from 19.1 to 36.2 kg/m² (Table 1). Body temperature ranged from 35.3 to 36.9°C.

The sensing area (cm²) differed for the different mattresses. The largest area was found for the old viscoelastic mattress (average 6315.20 cm²  ± 1105.22), and the least for the standard mattress from 2014 (average 5132.96 cm² ± 1032.02). The mean interface pressure at baseline was higher on one or more of the mattresses for 14 of 20 participants compared to the 10-minute score. The Kruskal-Wallis test showed a significant difference between mean interface pressures among the mattresses after 10 minutes. The mean interface pressure ranged from 30.28 to 38.37. The standard mattress from 2014 had the highest mean interface pressure (Table 2). 

The cut-off for high pressure was set at ≥ 90 mm Hg. The standard mattress from 2014 had the highest mean number of cells ≥ 90 mm Hg (144.25 ± 57.39), while the lowest number was found for the new viscoelastic mattress (96.05 ± 69.66). There were significant differences among the mattresses for number of cells reaching 60 (range 129.45–175.80; P = .025) and 80 mm Hg (range 100.10–148.10; P = .046) (Table 2).

Participant-assessed outcome measures showed the highest comfort on both the viscoelastic mattresses (75% good or very good) and lowest for the standard mattress from 2011 (30% good or very good). Ability to move with ease was greater on all standard mattresses compared to the viscoelastic mattresses (Table 3). All mattress types had a strong, positive correlation between comfort and mobility, with the lowest between the comfort and mobility for the new viscoelastic mattress (r = .572) and the highest for the standard mattress from 2018 (r = .874). 

BMI used as a continuous variable was not a significant factor for comfort and mobility level. However, BMI was a significant factor for mean interface pressure on each mattress as well as for the number of cells above 60 and 80 mm Hg. 

The mean interface pressures were 46.50 ± 4.83 for the participants weighing more than 100 kg and 33.86 ± 5.83 for participants weighing less than 100 kg on the standard mattress from 2011 (P = .012). Similarly, the mean interface pressures were 41.25 ± 7.70 for the participants weighing over 100 kg and 29.78 ± 5.99 for participants weighing less than 100 kg on the new viscoelastic mattress (P = .040), and 42.87 ± 4.09 for the participants weighing more than 100 kg and 28.05 ± 6.16 for participants weighing less than 100 kg on the old viscoelastic mattress (P = .012).

We know little about the pressure-redistributing properties of mattresses used in Norwegian hospitals or how their potential to redistribute pressure changes over time. A recent Korean study with 110 patients randomized in 2 groups found a significant difference in pressure ulcer development by mattress type. The experimental group lying on viscoelastic foam had fewer pressure ulcers compared to the group lying on standard hospital mattresses (3.6% vs. 27.3%; P = .001). However, whether it was the viscoelastic overlay or the double thickness of the mattress (6 cm + 7 cm) compared to the standard mattress of 6 cm in thickness may be questioned.¹⁶ Our study showed that volunteers achieved better pressure-redistribution lying on viscoelastic foam mattresses than on the hospital standard mattresses. Furthermore, the viscoelastic mattresses had the lowest number of cells reaching high pressure, although the differences were not statistically significant. This finding is similar to Defloor’s results. In that study, 62 healthy volunteers laid on 2 mattresses in 10 different positions for 1 minute. In the supine position, comparable to this study, that author found a maximum pressure (mm Hg) of 39.5 ± 7.0 for standard mattress and 27.7 ± 4.1 for a 14-cm viscoelastic mattress (Tempur). The author also found a 20% to 30% lower maximum interface pressure on a viscoelastic compared to a standard mattress in the supine position.¹⁷  

The baseline and 10-minute scan differed for each mattress. Foams require time to reach full function, as other studies have shown.^14,18 One expects older foam mattresses to be harder and more compact than newer foam. However, the average interface pressure was highest for the 2014 standard hospital mattress. This mattress had similar foam to the 2011 mattress that had a lower interface pressure. This difference could be related to the density of the foam. Whereas the mattresses from 2011 and 2014 have a flat foam structure, the 2018 mattress has an egg-crate shape that may distribute pressure differently. The range of interface pressure among the participants showed that hospitals need different types of mattresses to achieve the best redistribution properties for each patient. A cross-sectional study by Bergstrand et al,¹⁹ which included 35 inpatients 65 years or older and 80 healthy individuals 18 years or older, showed that although the viscoelastic foam/air mattress had significantly lower average sacral pressure values (23.5 ± 2.5) compared with other mattresses (25.6 ± 2.3 and 27.4 ± 2.8), it also had the highest proportion of subjects with decreased blood flow. A randomized controlled study (N = 15) found that on foam mattresses, participants had a 3 times higher risk of erythema after 2 hours than on gel mattresses.²⁰ This indicates that all surfaces in a bed probably should be pressure-mapped to meet patients’ individual needs and inform the selection of mattresses with the most appropriate pressure-redistribution properties.¹⁴  

Results of a meta-analysis including 2 publications from medical intensive care units (N = 1049) showed that the use of pressure mapping as a tool for systematic registration of redistribution in clinical practice decreased the risk of pressure ulcer development nearly 90%.²¹ In the current study, the authors investigated only 1 mattress from each year. However, the different pressure-redistributing effects according to BMI and weight showed that the function of the mattress differed among patients. Therefore, hospitals need a selection of mattresses that are in good condition. Although a significant correlation between comfort and ability to move was found for each mattress type, the present study sample was too small to draw definite conclusions about mattress type comfort and ability to move. Within-person differences were also observed as some prefered harder while others prefered softer mattresses. However, ease of mobility was much better on the standard mattresses compared with the viscoelastic mattresses, showing that patients may need help to shift position on a viscoelastic mattress. Body mechanics and morphology are important characteristics of individuals to consider when selecting a support surface.

Randomized order in which participants were exposed to pressure on each mattress as well as the comparison of new to worn mattresses. Being overweight is a growing problem throughout the world, and hospitalized patients often weigh more than 100 kg. We do know that patients weighing more than 100 kg are probably positioned on a viscoelastic overlay mattress at our hospital even though the weight limit is 100 kg. Therefore, it is a strength of this study that we included participants who weighed more than 100 kg, to be reflective of the patients admitted to our hospital.

There are several limitations to this study. First, the authors limited measurements to the supine position as this is the position with increased risk of pressure over bony prominences. Moreover, the supine position allows for equal positioning on all mattresses for all participants. The interface pressure measurement could well have been performed with the head of the bed elevated and in the lateral position for a more complete investigation.²² However, due to time limitations, only data collection in the supine position on each mattress was performed. Second, the room temperature was high because the study was conducted during a heat wave in the summer. The temperature was up to 27°C at 1 of the measurement points. This may have affected the density of the foam. Third, the pressure mat may have camouflaged the true effects of the mattresses. Fourth, unfortunately, no measures of bottoming out, hand compression, envelopment, immersion, or microclimate were performed, although several of these measurements are recommended for a mattress study.²³ Fifth, more body morphology measurements could have been included such as waist, hip, leg length, and shoe size.²⁴

Other limitations of this study are the small sample size, the use of healthy volunteers, and short time of measurement. Ten (10) minutes were chosen because foam uses 7 to 8 minutes to fully obtain its function. Crawford et al found a change in interface pressure up to 8 minutes in a sitting sample.¹⁸ Although other studies have found no change in interface readings between 3 and 30 minutes of acclimation,²⁵ this may have been too short a test period. On the other hand, other studies have measured the interface pressure after 1 to 5 minutes.^17,25,26 

The egg crate–shaped mattress distributed the pressure better than compact foam.  However, this could have been related to the age of the 2 mattresses evaluated.

In this comparative study of 20 volunteers laying on 5 different mattresses, older compact foam standard mattresses were found to be less comfortable and have higher interface pressure compared to the new type of standard foam and viscoelastic foam mattresses. A person’s body mechanics and morphology’s impact on the pressure redistribution, comfort, and ease of mobility need to be investigated in further study. The results of this study may inform hospitals’ procedures for mattress replacement.

Dr. Bredesen is a nurse researcher, Division of Orthopedic Surgery, Oslo University Hospital, Oslo, Norway, and an associate professor, Department of Nursing and Health Sciences, University of South-Eastern Norway, Campus Drammen, Norway. Address all correspondence to: Dr. Ida Marie Bredesen, Research and Development, Division of Orthopedic Surgery, Oslo University Hospital, Postboks 4950 Nydalen, 0424 Oslo, Norway; email: uxidbr@ous-hf.no.

This study received a small grant from the Orthopedic Nurses subgroup of the Norwegian Nurses Organisation.