Optimizing Seating in the Intensive Care Unit for Patients With Impaired Mobility

Approval for the project (phases 1 and 2) was granted by the institutional ethics committee. The need to obtain written consent from patients was waived in accordance with the National Health and Medical Research guidelines, as the project represented no substantial change to patient care procedures or increased risk.¹²  Each participant was given a verbal explanation of procedures as part of the process of sitting out of bed. The study was designed as a nonrandomized crossover trial in 2 phases for comparison of seating surfaces. Each patient acted as his or her own control, and seating surfaces were compared within patients, not between patients.

Patients who had impaired mobility and were scheduled to be sitting out of bed in the regular ICU chair were recruited. Patients exluded from the study were those who were unsuitable for sitting out of bed, had gross diarrhea, or were able to bear weight and therefore could sit out of bed in the ICU in a recliner rocker or high-back chair. The study was conducted in a 22-bed general ICU, a closed unit in Royal Perth Hospital, a tertiary-referral hospital in Perth, Australia. Organized turning rounds are used to position patients. Alternating airflow mattresses (Nimbus III, Huntleigh Healthcare, Perth, Australia) are used for patients in bed, except for patients with spinal injuries. Patients are sat out of bed by the nurses and/or physiotherapists with the assistance of a “turning team” as soon as their condition permits.

Patients were recruited between October 2007 and May 2008 for phase 1 and between August 2008 and January 2009 for phase 2. In phase 1, 3 seating surfaces were tested: the regular chair (TotaLift-II trolley chair, Wy’East, Clackamas, Oregon), the regular chair with a gel overlay (Gel), and an alternative chair (Hausted APC, SterisCorp, Mentor, Ohio). The regular chair (Figure 1) is a trolley chair designed for surface to surface transfer of patients (a feature not shared by the alternative chair) and is suited for patients unable to bear weight, facilitating safe manual handling practices for patients and staff. Patients are transferred from the bed to the chair while supine and then the chair is moved into an upright position. Patients can be placed supine for care during the seating episode. The seating surface of the regular chair is composed of a single cushion whereas the alternative chair has 4 separate cushions. A gel overlay (45x45 cm) placed beneath the buttocks on the regular ICU chair was the third surface tested in this phase.

In phase 2, a new surface was designed (by R.B., rehabilitation engineer), informed by the results of phase 1, and was compared with the regular chair. The new surface (Figure 2) consisted of 3 cushions (back rest, cushion under the buttocks, and cushion under the legs) made from a combination of high-and low-density foam and covered with a material to limit pressure and facilitate patients’ comfort. The cushion beneath the buttocks was shaped to support patients in a seated position. Preliminary testing of the new surface was conducted among 7 volunteer staff to ensure that it was safe to use before it was tried out with patients.

A seating protocol was used for both phases (Table 1). In phase 1, the decision on which seating surface was used for each seating session was determined by computer-generated random number. The randomization could not always be complied with because of the unavailability or lack of suitability of the chair, for example, some patients were too tall to be seated in the alternative chair. In both phases, the regular ICU chair was used for at least 1 seating episode for all patients. Unlike phase 1, in phase 2, seating interface pressure was measured first on the regular seating surface and second on the new surface at a subsequent session with the patient sitting out of bed. Patients who were able to do so were asked which surface was more comfortable in both phases of the study. This rating was subjective.

A Force Sensing Array (FSA version 4.0) pressure-mapping system (Vista Medical Ltd, Winnipeg, Canada) with a single standard 45x45-cm pressure map was used to measure contact pressure at the buttock-seat interface. The pressure-sensing mat contains 256 (16x16) sensors connected to the computer software (v4.1) via a type 4 interface module. Data were recorded as color-coded maps of pressure distribution, 3-dimensional grids, and numeric output parameters. The system was calibrated with an autocalibrator specific to the system and according to the instructions in the manufacturer’s manual.

Data were monitored continuously, and representative frames of interface pressure maps were recorded every second, but data at 1-minute intervals were analyzed and reported for the period of 5 to 29 minutes of sitting out of bed. Data from the first 5 minutes of pressure mapping were not used, to allow for a settling-in period, and rehabilitation physiotherapy often commenced after 30 minutes. Mean interface pressures are not reported because many recordings exceeded the calibration capability of the measuring system (ie, 250 mm Hg or higher). Pressures of 200 mm Hg or higher were considered excessive on the basis of the work by Reswick and Rogers,¹⁴  who suggested that sustained pressure should be kept below 30 mm Hg on ischii and below 100 mm Hg on thighs and that the maximum time for pressures of 200 mm Hg over bony prominences should be less than 2 hours for young, healthy volunteers. We also wanted to capture the area of highest pressure (ie, the area over the ischial tuberosities) and found that defining excessive pressures as pressures exceeding 200 mm Hg was useful for this purpose. Each patient required at least 2 sets of data for analysis. Severity of illness was estimated with the Acute Physiology and Chronic Health Evaluation (APACHE) II score (worst in first 24 hours).¹⁵  The Braden score was used to estimate the risk for pressure ulcers.¹⁶  Data were abstracted by the research nurse (T.W.): age, sex, APACHE II score, diagnosis, and length of stay were obtained from the ICU clinical database, and body mass index (BMI) and Braden score were obtained from the patient’s nursing charts.

Data analysis was conducted by using SPSS version 17.0 (SPSS, Chicago, Illinois). Continuous data are reported as medians and interquartile range (IQR). The interface pressures of the different surfaces were compared by using the Wilcoxon non-parametric test for paired data. In phase 1, 2 comparisons were made for the 3 interface pressure groups: (1) the gel surface compared with the regular chair and (2) the alternative surface compared with the regular chair. In phase 2, the regular and new surfaces were compared. Univariate logistic regression was used to examine the association of patients’ characteristics with excessive pressure for patients sitting out of bed in the regular chair. Odds ratios and 95% confidence intervals were reported. Results were considered significant if the P value was less than .05.

Eighteen patients had at least 2 sets of analyzable data (11 patients had data for 3 surfaces, 7 patients had data for regular chair and gel surfaces). The median age was 66 years and most patients (72%) were male (Table 2). Among the 18 patients, 8 were admitted for sepsis/pneumonia and 3 were admitted with trauma. The median length of stay was 14 days. The median APACHE II score was 17, and the mean Braden score was 12. The mean BMI was 27.

For the 18 patients, 53 seating surface measurement periods (1532 pressure maps) were compared. Pressure mapping was conducted for up to 120 minutes. One patient had to be put back to bed within minutes of starting to sit out of bed because of hypotension, and this seating episode was not mapped. No other events were encountered. An example of the pressure maps used to compare the regular chair and alternative chair at 5, 15, and 25 minutes of sitting out of bed is shown in Figure 3.

For patients who had pressure maps that showed excessive pressures (≥200 mm Hg), 46% of pressures recorded for the regular chair were higher than pressures for the gel chair, and on 11% of maps, the pressures were similar for the regular and gel seating surfaces (z = 2.0, P = .04). In comparison, the alternative chair had significantly fewer excessive seating interface pressures compared with the regular chair (67% of regular surfaces had higher pressures than the alternative surface, z = 4.0, P < .001), whereas 7% had similar interface pressures for both surfaces. Patients who were able to do so indicated that the alternative chair was more comfortable than the regular chair, and 2 patients refused to sit on the regular surface after sitting on the alternative surface. The alternative chair also had fewer excessive pressures (69% of surfaces) when compared with the gel overlay (P < .001). However, the alternative chair lacked the practical utility of the regular chair, as it was difficult to manage transfer of patients from bed to chair and the alternative chair had limited adjustment options for supporting the patient once out of bed. The gel overlay did not reduce seating interface pressures. It was concluded that no surface addressed all requirements for patients sitting out of bed, and therefore a new surface was designed to be adapted to our original chair.

Because the alternative chair was considered unsuitable for our ICU (eg, it had no built-in slide board facility, minimal footplate adjustment) and little benefit was seen with the gel surface, the new surface was compared only with the old surface by using the existing chair frame for both. The 2 surfaces were assessed among 7 volunteer ICU staff for practicality and safety before being introduced to the patient cohort. Interface pressures were assessed and revealed significant differences between them (P < .005 for all t-test paired comparisons) in favor of the new surface. The highest peak interface pressures were 90 to 101 mm Hg for the regular and 70 to 78 mm Hg for the new surface. The volunteers all remarked that the new seating cushions were more comfortable than the regular ICU chair.

Twenty phase 2 participants had at least 2 sets of analyzable data, with slightly more women (55%) than men. The median age was 62 years, and length of stay in the ICU was 17 days (Table 2). The median worst APACHE II score in 24 hours was 20. The mean Braden score was 12, and the mean BMI was 27.

More than half (55%) of the patients had seating interface pressures of 200 mm Hg or greater (excessive pressures), and of these 11 patients, 10 (93%) had fewer episodes of excessive pressures on the new surface (z = −4.2, P < .001). Nine patients had pressures of 200 mm Hg or higher at 5 minutes of sitting, and the excessive pressures were present in all members of this subgroup by 7 minutes of sitting out of bed. An example of the pressure maps that compares the regular surface with the new surface at 5, 15, and 25 minutes of sitting is shown in Figure 4. No events occurred that required return to bed or dissatisfaction with the new seating surface in phase 2. Patients able to do so indicated that the new surface was more comfortable.

For the remaining 9 patients, seating interface pressures were lower than 200 mm Hg. Among patients who had pressures of 150 to less than 200 mm Hg, 40% had fewer episodes of higher interface pressure with the new surface than with the regular surface (z = −7.1, P < .001), 28% had similar interface pressures on either surface, and the remaining 32% had higher pressures on the new surface.

The only factor (unadjusted) that was associated with excessive pressures when seated on the regular chair was the patient’s sex. Women (odds ratio, 0.12; 95% confidence interval, 0.03–0.59) were less likely to experience excessive pressures (Table 3). Age, APACHE II score, BMI, and Braden score were not significantly associated with excessive pressures. Women and men (n = 38) were similar for age, APACHE II score, Braden score, and BMI (P values ranged from .80 to .97; mean age, 64 years for women vs 63 years for men; mean APACHE II score, 22 for women vs 21 for men; mean Braden score, 12.3 for both women and men; and BMI, 27 for both women and men).

This study showed for the first time that most ICU patients sitting out of bed experience high seating interface pressures, and the cushions used to support critically ill patients have significantly different seating interface pressures. Furthermore, high pressures are often experienced early in the episode of sitting out of bed and are often sustained for the duration of the seated period. Practical design features, which include other considerations such as occupational risk reduction in moving patients and positioning devices such as adjustable footboards, are also important in any chair for vulnerable ICU patients sitting out of bed.

None of the tested surfaces in phase 1 of our study combined lower interface pressures with the required design features for a chair to sit ICU patients out of bed. Consequently, a new surface was developed that could be used with the existing chair frame.

Although we found high interface pressures with all surfaces assessed, we found significantly fewer episodes of excessive seating interface pressures with the new surface. These findings are important for promoting patients’ comfort and managing patients sitting out of bed in the ICU. Researchers in 2 studies^11,17  have used pressure mapping to examine interface pressure for ICU patients but, to our knowledge, no one has evaluated the seating interface pressures for patients sitting out of bed in the ICU. Sakai and colleagues¹⁷  assessed the feasibility of continuous interface pressures among 30 postoperative patients in ICU while the patients were in bed for up to 48 hours. They report the technique as feasible, although they excluded emergency admissions and the very ill patients who constitute a substantial proportion of ICU patients.

Pressure ulcers are debilitating and costly.¹⁸  Risk assessment and implementation of minimization strategies for pressure ulcers include pressure-relieving surfaces,¹⁹  reduction of mechanical load, appropriate skin care and control of moisture,²⁰  and adequate nutrition. The majority of these recommendations are based on expert opinion and consensus rather than controlled trials. Appropriate seating surfaces are crucial to ameliorate the risk for pressure ulcers and should be selected on the basis of the needs and characteristics of the patient.^21,22  None of the patients in this study had pressure ulcers develop once they started sitting out of bed, but we could not determine if pressure ulcers were caused by sitting out of bed after the ICU, and most patients in the study were discharged to the general care area soon after they were able to sit out of bed. Comfort and minimizing the risk for pressure ulcer development are particularly important for these patients because they are likely to have a long recovery phase that includes regularly sitting out of bed.

The volunteer staff participants had comparatively lower seating interface pressures with the new surface and reported it to be subjectively more comfortable. Interestingly, the seating interface pressures on both surfaces for the volunteers were considerably lower than those for most patients sitting out of bed in the ICU. These results provide further evidence of the lack of generalizability between the “healthy volunteer” and the critically ill patient in this context and might partly be explained by factors such as sedation of the patient, debilitated condition, and nutritional status.

We were able to translate our research findings directly into practice. Although the alternative chair had lower seating interface pressures, it lacked clinical utility. There was no built-in slide-board, which facilitates transfer of patients and reduces occupational injury of staff involved in manual handling of patients. As a result of the study, a decision was made not to purchase the alternative chair because it did not meet all the seating needs of the patients in our ICU. The useful features of the regular ICU chair were then incorporated with the new seating surface that could be used on the existing chair frame.

This study had several limitations. We anticipated that we would be able to recruit 20 patients within 6 months in phase 1. However, recruitment was restricted to business hours because of the availability of equipment and technical staff and did not include patients once they were transferred to the general care area or patients who were well enough to sit in a recliner rocker or high-back chair. As a consequence, recruitment was slow.

The potential for error in measurement of interface pressure is an important consideration. Pressure mapping is a comparative tool. Patients were used as their own controls to compare interface pressures between surfaces so as to limit variation in physical characteristics that would have been a factor if between-patient analyses were used. The contributions of factors such as shear that cannot be measured by pressure mapping are important.

We made every effort to reduce errors. Nevertheless, pressure mapping is only one tool in the assessment of risk for pressure ulcers. Peak mean pressures or the standard deviation are commonly used in the reporting of interface pressure. The calibration of the FSA equipment is limited to 250 mm Hg, and many of our patients experienced seating interface pressures of at least 250 mm Hg. We therefore compared the area of highest pressure by comparing the number of cells with excessive pressure (≥200 mm Hg) between surfaces for each patient. This strategy also minimized the bias from the differences in the contact area that occurred from the use of different seating surfaces. As only 9 patients had excessive pressure in phase 2, we then compared the number of cells registering pressures between 150 and 200 mm Hg. Measuring these pressures did not adequately capture the areas of peak pressure. Sixty-eight percent of the patients had lower or similar interface pressures with the new surface whereas nearly a third of patients (32%) had lower pressures with the old surface in this range. Pressure mapping provides information about individual patients based on their unique physical attributes. Patients may assume different seating postures between seating episodes. In addition, seating patterns vary on different surfaces. We used a seating protocol for sitting patients out of bed and maintained a similar inclination of the chair to minimize these effects.

Consideration of strategies for minimizing pressure ulcers, including appropriate seating surfaces for patients sitting out of bed in the ICU, is important. A combination of low- and high-density foam covered with pressure minimization material and the use of multiple cushions were key elements of our new seating surface. Evaluation of seating surfaces in the ICU by using interface pressure measurement is an important component of a quality initiative to optimize seating for critically ill patients and has utility within the wider hospital community, but further research is needed.

We thank the physiotherapists and nursing staff in the ICU at Royal Perth Hospital and Huntleigh Health Care for extending their trial period of the Steris HAUSTED APC chair.