There is evidence supporting the effectiveness of some pressure ulcer preventive measures (Cullum, McInnes, Bell-Syer, & Legood, 2004; Reddy, Gill, & Rochon, 2006), and national clinical guidelines for prevention of pressure ulcers advise that all patients at risk of or with pressure ulcers should receive preventive measures such as risk assessment (Bergstrom et al., 1992; Frantz, 2004; National Collaborating Centre for Nursing and Supportive Care, 2003; Registered Nurses Association of Ontario [RNAO], 2005; Royal College of Nursing [RCN], 2001; Wound, Ostomy, and Continence Nurses Society [WOCN], 2003) using an instrument such as the Braden (Bergstrom, Braden, Laguzza, & Holman, 1987) or Norton (Norton, McLaren, & Exton-Smith, 1975) scales, pressure-reducing devices such as overlays or mattresses (Bergstrom et al., 1992; Frantz, 2004; National Collaborating Centre for Nursing and Supportive Care, 2003; RNAO, 2005; RCN, 2001; WOCN, 2003), avoidance of exposure of skin to moisture from urinary or fecal incontinence (Bergstrom et al., 1992; RNAO, 2005; RCN, 2001; WOCN, 2003), reduction of shear forces by limiting the head of the bed to an angle below 30° (Frantz, 2004; RNAO, 2005; WOCN, 2003), and regular re positioning of immobile patients (Bergstrom et al., 1992; Frantz, 2004; National Collaborating Centre for Nursing and Supportive Care, 2003; RNAO, 2005; RCN, 2001; WOCN, 2003). Also suggested is careful documentation of pressure ulcers and of preventive interventions (Bergstrom et al., 1992; Frantz, 2004; National Collaborating Centre for Nursing and Supportive Care, 2003; RNAO, 2007; WOCN, 2003). Most of these recommendations are based on limited evidence (primarily expert opinion), except for the use of pressure-reducing devices (Bergstrom et al., 1992; RNAO, 2005; WOCN, 2003) for which the evidence includes systematic reviews of randomized controlled trials (Cullum et al., 2004; Reddy et al., 2006; Vanderwee, Grypdonck, & Defloor, 2008). Results of these trials suggest that specialized foam mattresses can reduce pressure ulcer incidence compared with standard hospital mattresses and that alternating pressure mattresses are also likely to reduce pressure ulcer incidence (although this evidence is less clear because of methodological issues).

It has been shown that preventive measures are under-used in at-risk populations of hospital patients (Gunningberg, 2005; Gunningberg, Lindholm, Carlsson, & Sjoden, 2001; Perneger et al., 1998; Pieper, Sugrue, Weiland, Sprague, & Heimann, 1997; Pieper, Sugrue, Weiland, Sprague, & Heiman, 1998). In two studies, fewer than 25% of patients had any preventive device in use (Gunningberg, 2005; Perneger et al., 1998). Use of preventive measures, and devices in particular, generally has been found to be more frequent for patients at high risk, as measured by the Norton (Allman et al., 1995) or Braden (Gunningberg, 2005; Pieper et al., 1997) scale, and for patients with a pressure ulcer (Perneger et al., 1998). Patients with pressure ulcers may benefit from these devices in terms of improved healing of existing pressure ulcers and prevention of additional pressure ulcers (WOCN, 2003).

Evidence from studies in animals (Kosiak, 1961) and humans (Defloor, De Bacquer, & Grypdonck, 2005; Norton et al., 1975) suggests that pressure ulcers can arise after short periods of immobility. This evidence suggests that preventive measures should be implemented early in the hospital stay. In prior studies of use of preventive measures, cross-sections of the hospital population have been used (Gunningberg, 2005; Gunningberg et al., 2001; Perneger et al., 1998) or patients have been followed prospectively across their stay (Pieper et al., 1997, 1998), but none have examined the use of preventive measures early in the hospital stay. It is important to determine whether guidelines for prevention of pressure ulcers are being implemented early in the hospital stay to maximize effectiveness of preventive measures.

The aims of this study were to determine the frequency and correlates of use of pressure ulcer preventive devices early in the hospital stay of elderly patients and to determine the frequency and correlates of recording of pressure ulcers in the patient record. The hypotheses were that, if guidelines are being applied appropriately, device use would be more frequent among patients with pressure ulcers or who are at higher risk of pressure ulcers and that all pressure ulcers observed in the research assessment would be documented in the patient record.

The study sample included 29 patients with both preexisting and hospital-acquired pressure ulcers, 36 patients with one or more preexisting pressure ulcers but no hospital-acquired pressure ulcers (14 of whom had at least one Stage 2+ ulcer), 166 patients with one or more hospital-acquired pressure ulcers but no preexisting pressure ulcers (71 of whom had at least one Stage 2+ ulcer), and 561 patients with no pressure ulcers. Thus, most patients in the study population had no pressure ulcers (88.3%). Among those with pressure ulcers, it was most common to have just one pressure ulcer (7.4% of the study population), although 1.9% had two pressure ulcers and 2.4% had at least three pressure ulcers at examination.

Approximately 17% of patients in the study cohort were classified as being at risk of pressure ulcers (Norton score ≤14). There were also substantial proportions of patients with the individual pressure ulcer risk factors examined. For example, 24.5% of patients in the study cohort had an intermediate or high risk of nutrition-related complications. Almost 21% of patients in the study cohort were not oriented fully to person, place, and time, whereas 18.9% of patients in the study cohort had urinary incontinence, fecal incontinence, or both (not tabulated).

Fifteen percent of patients had one or more preventive devices for pressure ulcers in use (Table 2). Frequencies for individual devices were 0.4% for replacement mattress, 3.6% for mattress overlay, 2.9% for heel protectors, 1.1% for chair cushion, 11.4% for positioning pillows, and 0.6% for another preventive device.

All examined risk factors were associated significantly with device use (Table 2). In particular, the presence of pressure ulcers was associated with the use of pressure ulcer preventive devices (p < .001). Patients with both hospital-acquired and preexisting pressure ulcers were most likely to have a preventive device (75.9%). Frequency of device use was lower for those with hospital-acquired pressure ulcers than for those with preexisting pressure ulcers (43.2% vs. 50.0% for those with Stage 2+ pressure ulcers, respectively, and 32.4% vs. 64.3% for those with Stage 1 pressure ulcers only). Patients with Stage 1 hospital-acquired pressure ulcers were less likely to have preventive devices in use than were those with Stage 2+ hospital-acquired pressure ulcers. The opposite association with stage was observed for preexisting pressure ulcers.

There was a gradient of decreasing frequency of preventive device use with decreasing pressure ulcer risk (indicated by lower Norton scores) and with decreasing risk of nutrition-related complications. Fifty-one percent of patients classified as at risk by the Norton score (≤14) had a preventive device in use (p < .001). Even in the highest pressure ulcer risk group (defined as Norton score 5–10), only 62.4% of patients had a preventive device in use. Similarly, among patients at high risk of nutrition-related complications, only 36.3% had a preventive device. Greater comorbidity, presence of incontinence, disorientation, and more ADL dependencies were associated with preventive device use (p < .001 for each). Less than half of the patient population were using preventive devices in the high-risk category of each of these risk factors. Higher age was associated also with use of preventive devices (p = .011), but gender (p = .13) and Black race were not (p = .37).

In the unadjusted model, there was a strong association between pressure ulcer type and stage and preventive device use, but this association was no longer significant once the model was adjusted for Norton score and the other covariates. For example, the OR for device use comparing those with both types of pressure ulcers with those with no pressure ulcers was 26.7 (95% confidence interval [CI] = 10.9–65.4) before adjustment but only 1.2 (95% CI = 0.5–2.9) after adjustment. In contrast, the association between Norton score category and preventive device use was strong regardless of adjustment for covariates including pressure ulcer type and stage. Those with Norton scores of 5–10 had a much higher odds of preventive device use (adjusted OR = 41.8, 95% CI = 14.0–124.6) compared with the odds of those with scores of 19–20. The adjusted OR for the other at-risk category (Norton scores 11–14) was also large (OR = 21.1, 95% CI = 8.4–52.9). Comorbidity (OR = 1.1, 95% CI = 1.0–1.3), urinary incontinence (OR = 2.5, 95% CI = 1.3–5.1), and having six ADL dependencies (OR = 2.2, 95% CI = 1.1–4.4) were the only other variables that remained significantly associated with device use in the multivariable models.

The association between Norton at-risk status and use of preventive devices was strongest among those with no pressure ulcers (OR = 9.3, 95% CI = 4.8–18.2; Figure 2) or with hospital-acquired pressure ulcers only (OR = 9.4, 95% CI = 1.4–64.2). The association was also strong among those with preexisting pressure ulcers only (OR = 6.0, 95% CI = 2.5–14.2). Among those with both hospital-acquired and preexisting pressure ulcers, there was no significant association between Norton at-risk status and preventive device use. The interaction between Norton at-risk status (≤14 vs. >14) and pressure ulcer type was not significant (χ²[3, n = 739] = 3.36, p = .34, not tabulated).

All of the pressure ulcer risk factors of interest were associated with the use of preventive devices, both risk factors evaluated by research nurses directly and those based on the admission nursing assessment in the patient record. Use of preventive devices was more frequent for patients with a pressure ulcer, particularly for those with a preexisting pressure ulcer and those with both preexisting and hospital-acquired pressure ulcers. The association between device use and the presence and type of pressure ulcers was fairly consistent even among patients with only Stage 1 pressure ulcers, despite the difficulties in diagnosing these pressure ulcers. The ORs for device use comparing the high risk to the lowest risk Norton score categories were very large. With only 2.8% of patients in the lowest risk category using preventive devices, the ORs for device use were expected to be very large. The more frequent use of preventive devices among those with pressure ulcers and among at-risk patients is encouraging because it may indicate that nursing staff correctly allocate these devices to individuals most likely to benefit, as recommended by clinical guidelines (Bergstrom et al., 1992). Use of preventive devices is indicated for patients with pressure ulcers to prevent development of additional pressure ulcers and advancement of existing pressure ulcers (WOCN, 2003). A higher Norton score was associated with device use, even after adjustment for pressure ulcer presence and type, indicating that at-risk patients were more likely to be allocated a device even if they did not have a pressure ulcer. This study could not distinguish between patients who were not allocated devices and those who were allocated devices but were not using them. Thus, the association between preventive device use and pressure ulcer risk factors may also represent a higher level of adherence among at-risk patients. In either case, it is encouraging that the patients who are most likely to benefit are those most likely to have these devices in use.

However, the prevalence of use of preventive devices was suboptimal even in the highest risk groups. Although nearly all of these patients would be expected to have a device in use, only approximately three quarters of those with both preexisting and hospital-acquired pressure ulcers had a preventive device in use at the time of the skin examination. Similarly, although almost all patients considered at risk using the conventional cutoff of the Norton score (≤14) would be expected to have a device in use, only approximately one half of those at risk by this definition had a preventive device. Even lower frequencies of device use were found for patients with the other risk factors. These results are similar to those of previous studies (Gunningberg, 2005; Gunningberg et al., 2001; Perneger et al., 1998; Pieper et al., 1997, 1998). This relatively low level of use by high-risk patients is of concern because it indicates a likely failure to follow guidelines (Bergstrom et al., 1992) and implement strategies that have potential to prevent pressure ulcers. As allocation appears to properly target patients at greatest risk, it is possible that the suboptimal levels of device use overall were due to insufficient resources or an overall lack of patient adherence rather than an issue of quality of nursing care. It is important that hospital administrators understand the substantial personal and institutional costs of pressure ulcers and allocate resources to devices for their prevention.

Frequency of documentation of pressure ulcers was also low, as was found previously (Gunningberg & Ehrenberg, 2004; Knox, 1998). In this study, the documentation was not required to correspond to the pressure ulcer noted on examination, and the quality of documentation (e.g., recording of stage, site, and clinical characteristics) was not examined, so significant gaps in recording may have existed even among patients with documentation of a pressure ulcer. Documentation is important to ensure proper communication within the healthcare team, as well as ensuring the accuracy of audits (Gunningberg & Ehrenberg, 2004; Knox, 1998). The low level of documentation found in this study suggests that many healthcare teams may not be maintaining proper communication to ensure effective management of this important health concern. Documentation may have been infrequent because of a lack of knowledge of skin assessment by the nurses, a misunderstanding of the importance of pressure ulcers in patient care, or too little time spent on documentation. Hospitals should ensure that nurses are properly trained in chronic wound prevention and care, including an understanding of the importance of pressure ulcers in patient care. They should also consider implementing standardized wound assessment protocols to ensure that skin integrity is assessed regularly and all wounds are documented and treated properly.

Caution is required when considering the results of this study because the data were gathered between 1998 and 2001. The frequency of device use and of pressure ulcer documentation may have changed in the intervening years, although the prevalence of pressure ulcers has remained fairly stable (Whittington & Briones, 2004). Also, interpretation of these results may be limited because most data were collected at a single brief examination. Therefore, there is the possibility of underdetection of device use and difficulty in assessing the risk of pressure ulcers and nutritional status. There is also the possibility of overdiagnosis of pressure ulcers by the research nurses, which could result in spuriously low levels of pressure ulcer documentation. Although this study was intended to estimate the frequency of device use early in the hospital stay, data collection was limited to an examination on Hospital Day 3, and information about earlier device use cannot be provided. Data sparseness precluded determination of characteristics associated with the use of specific types of preventive devices as the frequency of use of each type was less than 12%. Pressure ulcer status, Norton score, and preventive device use were assessed at the same examination, so temporality of proposed risk factors relative to device use could not be determined. However, pressure ulcer status and Norton subscales are relatively stable over several days, whereas the use of a preventive device may change at any time, so the temporal order is plausible. As the study population is not representative of the general population, caution should be used in generalizing this study’s results.