Apressure wound is an injury to the skin and deeper tissues, generally due to prolonged compression of the body which causes underlying tissue ischaemia, usually over a bony structure. These injuries may especially be seen in people who have deficiencies in mobility or sensation, such as patients with neurological disorders, sedation, peri- or postoperative immobilisation status, and hospitalisation.1 Pressure injury (PI), when occurring during a hospital stay, is regarded as a hospital-acquired PI (HAPI), which might cause medical, legal, and economic problems and could be a burden on health systems. The prevalence of HAPIs among hospitalised patients is between 3–14%, and as high as 70% in high-risk patients.2 Health professionals in hospitals should primarily focus on high-risk patients to avoid PIs, but if such PIs have already occurred, treatment should be started as soon as possible.
During the COVID-19 pandemic, increased use of positive pressure ventilating masks, or making the patient lie in a prone position, led to an increase in medical device-associated PIs in patients.3,4,5
Aim
The purpose of this single-centred retrospective study was to investigate the risk factors that contribute to PIs in patients during hospitalisation and to analyse them to determine whether any agent can help hospital healthcare providers prevent HAPI from occurring, such as application of zinc-containing and/or barrier creams (ZnBC) or airbed use. Frequent and regular repositioning is also a recommended practice for HAPI prevention.5
Method
Between June 2014 and June 2021, hospitalised patients' files, from Bayindir Sögütözü Hospital in Ankara, Turkey, were screened retrospectively for Stage 2 and deeper PI development. Patients who had PIs on the day of hospitalisation, or who developed Stage 1 persistent erythema, were excluded from the study.
Patient demographics
The patients' age, sex, Waterlow scale score, mobilisation status, application of ZnBC, airbed use, hospitalisation period, and the day on which the PI opened were recorded.
The Waterlow scale6 has a good predictive validity for PIs in hospitalised patients. It is a scale to assess the risk status of the patient, based on age, sex, weight and size relationship, continence status, and other comorbidities. The risk scale of the patients was grouped into three: ≥20 points; 15–19 points; and ≤14 points.
The mobilisation status of the patient was divided into four subgroups, as follows:
- Immobile: without any motion
- In-bed mobile: only mobilisation in the lying position with the likelihood of turning from one side to another in bed
- Limited mobilisation: continuing normal daily physical activity but with the help of another person if necessary
- Mobile: capable of normal physical activities of daily life.
For patients other than those designated as ‘mobile’, nurses changed the lying position of these patients every two hours when possible.
The application of ZnBC was recorded as ‘yes’/‘no’. Barrier creams such as Rapider cream (Vitamax, Turkey), Oksizink cream (Dermotek, Turkey), Zinko cream (Mega Farma, Turkey) or Coloplast barrier cream (Coloplast, Denmark were applied as ZnBCs. When used, the application was twice daily on PI risk areas, such as bony structures.
An airbed is an electrical device/mattress that contains separate air-filled sacs, inflating and deflating one after the other, causing a wave effect under the patient, and relieving pressure-bearing areas. In our study, we considered this parameter as ‘yes’/‘no’ for the use or not of airbeds.
The hospitalisation period was another parameter recorded for the patients, being the total period of time during which the patient stayed at the hospital, and was grouped into five time periods: 1–6 days; 7–14 days; 15–29 days; 30–89 days; and ≥90 days.
The wound ‘open day’ was considered as the day on which the PI opened after hospitalisation. Stage 1, the persistent erythema phase of the PU, was not included. The ‘open day’ was divided into three periods: 1–5 days; 6–15 days; and ≥16 days after the day the patient was first admitted to hospital.
Study endpoints
In our study, our aim was to present patients' demographic data, and determine whether there was a difference between patients who developed HAPIs and those who did not. Secondly, we wanted to determine whether use of ZnBC and airbeds helped with the prevention of HAPIs. We also wanted to detect the risk ratios between the subgroups of every single parameter that may help further with caregiving to the hospitalised patients.
Statistical analysis
In the statistical part of the study, in the descriptive tables, categorical variables are shown as numbers and percentages. Pearson Chi-squared test was used in the analysis of categorical variables. For numerical variables (only age was numerical) normal distribution was evaluated with the Kolmogorov–Smirnov test. Mann–Whitney U test was used in the analysis of numerical data that did not fit the normal distribution.
Logistic regression is a statistical analysis method used to obtain an odds ratio (OR) in the presence of more than one explanatory variable. The procedure is similar to multiple linear regression analysis but with the exception that the response variable is binomial. The result is the impact of each variable on the OR of the observed event of interest.7 Considering the development of HAPI as a dependent variable, in the logistic regression analysis, those with p<0.2 in pairwise comparisons were included in the model, ‘enter’ was used as the variable selection method. A p-value of <0.05 was considered statistically significant.
Ethical approval
Approval was obtained from the Scientific, Medical Ethics and Deontology Committee of Bayindir Hospital (Approval number: BTEDK-15/21, 5 November 2021). No written patient consent was required by the Ethics Committee as the study was retrospectively designed.
Results and statistical analysis
The demographic and descriptive data of all 2327 hospitalised patients included in the study are listed in Table 1. Of the hospitalised patients, 303 (13%) developed Stage 2 and deeper PIs, and 2024 were hospitalised and discharged without the HAPI opening.
Table 1. Demographic and descriptive data of hospitalised patients (n=2327)
| Variable | |
|---|---|
| Age, years, mean±SD, median (min–max) | 73.5±15.1, 75 (1–107) |
| Sex, n (%) | |
| Male | 1047 (45.0) |
| Female | 1280 (55.0) |
| Waterlow score, n (%) | |
| ≥20 | 761 (32.7) |
| 15–19 | 1457 (62.6) |
| ≤14 | 109 (4.7) |
| Mobilisation, n (%) | |
| Immobile | 1087 (46.7) |
| In-bed mobile | 125 (5.4) |
| Limited | 710 (30.5) |
| Mobile | 405 (17.4) |
| HAPI occurrence, n (%) | |
| HAPI occurred | 303 (13.0) |
| HAPI did not occur | 2024 (87.0) |
| Wound opening occurrence (n=303), days, n (%) | |
| 1–5 | 63 (20.8) |
| 6–15 | 129 (42.6) |
| ≥16 | 111 (36.6) |
| Airbed, n (%) | |
| Used | 700 (30.1) |
| Not used | 1627 (69.9) |
| Zinc-containing and/or barrier cream, n (%) | |
| Used | 1757 (75.5) |
| Not used | 570 (24.5) |
| Hospitalisation period, days, n (%) | |
| 1–6 | 217 (9.3) |
| 7–14 | 342 (14.7) |
| 15–29 | 1439 (61.9) |
| 30–89 | 249 (10.7) |
| ≥90 | 80 (3.4) |
HAPI—hospital-acquired pressure injury; max—maximum; min—minimum; SD—standard deviation
When patient age, sex, Waterlow scale score, mobilisation, ZnBC use, airbed use and hospitalisation period between patients who had a HAPI and those who did not, were compared, all parameters except age were statistically significant (p<0.05) (Table 2). The median age was 74 years in patients with a HAPI compared with 76 years in patients without. The difference was not statistically significant (p=0.692) (Table 2).
Table 2. The comparative data of the patients with or without hospital-acquired pressure injury (HAPI) (n=2327)
| Variable | HAPI occurred | HAPI did not occur | p-value |
|---|---|---|---|
| Age, years, median (min–max) | 74 (1–98) | 76 (1–107) | 0.692 |
| Sex, n (%) | |||
| Male | 169 (16.1) | 878 (83.9) | <0.001 |
| Female | 134 (10.5) | 1146 (89.5) | |
| Waterlow score, n (%) | |||
| ≥20 | 137 (18.0) | 624 (82.0) | <0.001 |
| 15–19 | 139 (9.5) | 1318 (90.5) | |
| ≤14 | 27 (24.8) | 82 (75.2) | |
| Mobilisation, n (%) | |||
| Immobile | 212 (19.5) | 875 (80.5) | <0.001 |
| In-bed mobile | 5 (4.0) | 120 (96.0) | |
| Limited | 58 (8.2) | 652 (91.8) | |
| Mobile | 28 (6.9) | 377 (93.1) | |
| Airbed, n (%) | |||
| Used | 126 (18.0) | 574 (82.0) | <0.001 |
| Not used | 177 (10.9) | 1450 (89.1) | |
| Zinc-containing and/or barrier cream, n (%) | |||
| Used | 277 (15.8) | 1480 (84.2) | <0.001 |
| Not used | 26 (4.6) | 544 (95.4) | |
| Hospitalisation period, days, n (%) | |||
| 1–6 | 2 (0.9) | 215 (99.1) | <0.001 |
| 7–14 | 28 (8.2) | 314 (91.8) | |
| 15–29 | 132 (9.2) | 1307 (90.8) | |
| 30–89 | 101 (40.6) | 148 (59.4) | |
| ≥90 | 40 (50.0) | 40 (50.0) | |
max—maximum; min—minimum
Among the 2327 hospitalised patients, we also evaluated the differences in ZnBC use. We compared the patients' age, sex, Waterlow scale score, mobilisation status, whether HAPI occurred or not and, if it did, when the day the HAPI opened, airbed use and hospitalisation period. All parameters were statistically significant except for sex, although this was close to significance (p=0.059) (Table 3).
Table 3. The comparative data of the patients according to the usage of ZnBC (n=2327)
| Variable | ZnBC used | ZnBC not used | p-value |
|---|---|---|---|
| Age, years, median (min–max) | 77 (1–107) | 72.5 (1–100) | <0.001 |
| Sex, n (%) | |||
| Male | 810 (46.1) | 237 (41.6) | 0.059 |
| Female | 947 (53.9) | 333 (58.4) | |
| Waterlow score, n (%) | |||
| ≥20 | 633 (36.0) | 128 (22.5) | <0.001 |
| 15–19 | 1046 (59.5) | 411 (72.1) | |
| ≤14 | 78 (4.5) | 31 (5.4) | |
| Mobilisation, n (%) | |||
| Immobile | 953 (54.2) | 134 (23.5) | <0.001 |
| In-bed mobile | 72 (4.1) | 53 (9.3) | |
| Limited | 512 (29.2) | 198 (34.7) | |
| Mobile | 220 (12.5) | 185 (32.5) | |
| HAPI occurrence, n (%) | |||
| HAPI occurred | 277 (15.8) | 26 (4.6) | <0.001 |
| HAPI did not occur | 1480 (84.2) | 544 (95.4) | |
| Wound opening occurrence (n=303), n (%) | |||
| 1–5 | 56 (20.2) | 7 (26.9) | 0.020 |
| 6–15 | 113 (40.8) | 16 (61.6) | |
| ≥16 | 108 (39.0) | 3 (11.5) | |
| Airbed, n (%) | |||
| Used | 627 (35.7) | 73 (12.8) | <0.001 |
| Not used | 1130 (64.3) | 497 (87.2) | |
| Hospitalisation period, days, n (%) | |||
| 1–6 | 141 (8.0) | 76 (13.3) | <0.001 |
| 7–14 | 254 (14.5) | 88 (15.4) | |
| 15–29 | 1050 (59.8) | 389 (68.3) | |
| 30–89 | 236 (13.4) | 13 (2.3) | |
| ≥90 | 76 (4.3) | 4 (0.7) | |
HAPI—hospital-acquired pressure injury; max—maximum; min—minimum; ZnBC—zinc-containing and/or barrier cream
For ZnBC use, the median age in the group of patients in which cream was used was 77 years compared with 72.5 years in the group of patients in which cream was not used. The difference between age groups was statistically significant (p<0.05) (Table 3).
The OR values of the subgroups for each variable according to HAPI occurrence are listed in Table 4. An OR value of 1 is accepted as the reference value. Males had a 1.49-times greater risk of HAPI occurrence than females, which was statistically significant (p<0.05).
Table 4. The odds ratio (OR) values of the subgroups for each variable according to hospital-acquired pressure injury (HAPI) occurrence (1 is accepted as reference value) (n=2327)
| Variable | β | SE | OR | 95% CI | p-value |
|---|---|---|---|---|---|
| Sex, n (%) | |||||
| Male | 0.400 | 0.142 | 1.492 | 1.130–1.970 | 0.005 |
| Female | 1 (ref) | ||||
| Waterlow score, n (%) | |||||
| ≥20 | 1 (ref) | ||||
| 15–19 | -0.173 | 0.158 | 0.841 | 0.617–1.146 | 0.272 |
| ≤14 | 1.183 | 0.294 | 3.265 | 1.834–5.813 | <0.001 |
| Mobilisation, n (%) | |||||
| Immobile | 0.161 | 0.256 | 1.175 | 0.711–1.941 | 0.529 |
| In-bed mobile | –1.230 | 0.545 | 0.292 | 0.100–0.851 | 0.024 |
| Limited | –0.245 | 0.260 | 0.782 | 0.470–1.301 | 0.345 |
| Mobile | 1 (ref) | ||||
| Airbed, n (%) | |||||
| Used | 1 (ref) | ||||
| Not used | 0.476 | 0.164 | 1.609 | 1.167–2.218 | 0.004 |
| Zinc-containing and/or barrier cream, n (%) | |||||
| Used | 0.570 | 0.235 | 1.768 | 1.116–2.800 | 0.015 |
| Not used | 1 (ref) | ||||
| Hospitalisation period, days, n (%) | |||||
| 1–6 | 1 (ref) | ||||
| 7–14 | 2.071 | 0.744 | 7.932 | 1.844–34.115 | 0.005 |
| 15–29 | 2.314 | 0.721 | 10.115 | 2.460–41.590 | 0.001 |
| 30–89 | 3.684 | 0.731 | 39.811 | 9.501–166.815 | <0.001 |
| ≥90 | 4.026 | 0.758 | 56.046 | 12.683–247.673 | <0.001 |
β—regression coefficient; CI—confidence interval; SE—standard error
With regard to wound opening occurrence in patients with a HAPI and comparing their timing with airbed use, 49.2% of the patients who used airbeds had a HAPI after 16 days. Meanwhile, only 27.7% of patients without airbed use had a HAPI after 16 days and that difference was statistically significant (p<0.05) (Table 5). These results can be summarised as: airbed use protects against HAPI formation and in patients with a HAPI, airbed use postpones wound opening when used.
Table 5. Hospital-acquired pressure injury ‘open day’ in patients using an airbed
| 1–5 days | 6–15 days | ≥16 days | Total | p-value | |
|---|---|---|---|---|---|
| Airbed used, n (%) | 15 (11.9) | 49 (38.9) | 62 (49.2) | 126 (100) | <0.001 |
| Airbed not used, n (%) | 48 (27.1) | 80 (45.2) | 49 (27.7) | 177 (100) | |
| Total, n (%) | 63 (20.8) | 129 (42.6) | 111 (36.6) | 303 (100) |
Using ZnBC did not reduce the risk of developing a HAPI and was statistically significant (p<0.05). On the contrary, when we take patients not using ZnBC as reference, the relative risk increased 1.8 times with the use of ZnBC (Table 4). However, when we consider wound opening occurrence in patients with a HAPI in patients who used ZnBC and in patients who did not use ZnBC, in 39% of the patients who used ZnBC the HAPI opened after 16 days. Meanwhile, only 11.5% of the patients who did not use ZnBC had a HAPI open after 16 days, and that difference was statistically significant (p<0.05). This result showed that ZnBC did not prevent but may have postponed the time of HAPI occurrence (Table 6).
Table 6. Hospital-acquired pressure injury ‘open day’ in patients using ZnBC
| 1–5 days | 6–15 days | ≥16 days | Total | p-value | |
|---|---|---|---|---|---|
| ZnBC used, n (%) | 56 (20.2) | 113 (40.8) | 108 (39.0) | 277 (100) | 0.020 |
| ZnBC not used, n (%) | 7 (26.9) | 16 (61.5) | 3 (11.5) | 26 (100) | |
| Total, n (%) | 63 (20.8) | 129 (42.6) | 111 (36.6) | 303 (100) |
ZnBC—zinc-containing and/or barrier cream
There was a significant increase in HAPI occurrence related to the hospitalisation period. The risk increased by: 7.9-times in patients hospitalised for 7–14 days; 10.11-times in patients hospitalised for 15–29 days; 39.81-times in patients hospitalised for 30–89 days; and 56.04-times in patients hospitalised for ≥90 days, compared with patients hospitalised for 1–6 days (Table 4).
Discussion
The incidence of HAPI is accepted as a marker for patient care quality. Dos Santos et al.8 focused on high rates of prevalence in patients admitted to intensive care units (ICU), ranging from 8–23% of cases, depending on the patients' severity of illness and the risk factors detected. In Europe, the cost of wound care is known to be 2–4% of healthcare budgets.9,10 In the future, skin lesions are expected to increase in parallel with the increase in life expectancy and accompanying medical conditions, such as diabetes, heart failure, obesity, pulmonary and vascular diseases, which will make the wound care market continue to expand.9,10 In this study, we present our patients' demographic data from Bayindir Hospital, a private hospital, located in the capital of Turkey, Ankara, with intensive care unit, internal medicine and surgical wards.
Among 2327 patients, the majority were older (mean: 73.5 years), with an approximately equal male:female distribution; most of them scored 15–19 points on the Waterlow scale (high-risk patients for PU development) and were immobile. We expected to find an increase in the prevalence of HAPI with age, particularly because of the flattening of the dermo-epidermal junction and increase in skin stiffness with ageing, leading to an increased risk of shear-type injuries, such as skin tears or deeper wounds.11 However, in our study, there was no statistically significant age difference between patients where HAPI occurred and where it did not occur, and this was probably a consequence of having similar older ages in both groups (median: 74 years versus 76 years).
Despite the use of known protective factors, such as airbeds and ZnBC, HAPIs may still occur. Therefore, we compared parameters including age, sex, Waterlow scale score, mobilisation, ZnBC use, airbed use and hospitalisation period among patients who developed HAPIs, and those who did not. In our study, male sex had a poor prognostic effect on HAPI development. In the literature, multiple risk factors are associated with the development of PIs, and these factors can be grouped as ‘intrinsic’, which are related to a patient's pre-existing medical status, and ‘extrinsic’, which are related to a patient's environment.
Intrinsic risk factors include: neurological disease; motor impairment; cognitive impairment; sensory deficits; malnutrition; and hypoperfusion due to peripheral vascular disease or congestive heart failure. Extrinsic risk factors include: inadequate mobilisation by caregivers; trauma and sedation; application of physical restraints; incorrect positioning; moisture; and shearing forces. Of these risk factors, failure to frequently change position is thought to be the biggest contributor to PU formation. A combination of incorrect positioning and moisture on the skin surface is a frequent cause of PI formation in critically ill patients.12 These PU parameters do not display sex predominance; however, in our study, we found that the male sex is more prone to HAPI formation (OR: 1.49).
The Waterlow scale consists of seven parameters: build or weight; height; visual assessment of the skin; sex and age; continence; mobility; and appetite, in addition to some particular risk factors, such as tissue malnutrition; neurological deficit; major surgery or trauma; and patient medication. This scale finally identifies three ‘risk’ categories: a score of 10–14 categorises the patient as ‘at risk’; a score of 15–19 categorises the patient as ‘high risk; and a score of ≥20 categorises the patient as ‘very high risk’.13 The Waterlow scale, which our nurses have been familiar with for many years, has been used in our hospital. In our study, for Waterlow scale we would have expected to find that the higher the score, the higher the risk for HAPI formation, similar to Olivo et al.14 However, we found the highest proportion at the lower end of the Waterlow scale (≤14 points). This may be attributed to the lower intensity of care and regular repositioning given to patients in the low-risk group compared to the high-risk patients.
For mobilisation status, the highest risk for HAPI formation was in patients who were immobile, as was expected. This was followed, by patients who were mobile patients and patients who had limited mobility, but these data were not statistically significant (p>0.05). Surprisingly, it was noticed that in-bed mobile patients had the lowest percentage of HAPI formation (p<0.05). This might have been a consequence of the limited number of patients in this group (Table 2).
When we analysed the effect of airbed use on HAPI occurrence, we observed its protective effect (p<0.05). Previous reports have pointed out the preventive effect of air surfaces in PIs. Alternating air pressure surfaces may reduce the risk of developing a new PI.15,16 Some reports show conflicting results.17,18 These inconsistencies may stem from their retrospective nature and the absence of comparable settings for evaluation. In our study, the risk of HAPI increased to 1.6-times without the use of an airbed. Furthermore, even in cases where HAPI occurred, the onset day was notably delayed. Therefore, we recommend using active air surfaces, i.e., airbeds, in high-risk patients as soon as possible after they are hospitalised.
There are some avoidable risk factors in ideal treatment, such as decreasing pressure and controlling moisture. The National Institute for Health and Care Excellence (NICE) guidelines recommend using barrier preparations to prevent skin damage and control moisture.2,19 The application of barrier protectants based on petrolatum, silicon, zinc and acrylates, which create a physical barrier, has been described.20 Although these skin protectants are used to minimise friction, remove fluid and protect the skin from chemical irritants, results are still inconsistent.20 Groom et al.21 reported a statistically significant reduction of skin tears after applying a phospholipid-based cleanser combined with a dimethicone-containing moisturiser compared with a surfactant-based cleanser and two dimethicone- and/or zinc-containing skin protectants.11,21 In our study, we could not detect the protective effect of ZnBC application on HAPI formation. This may have been a consequence of the characteristics of the patient group in which ZnBC was used. In our study, ZnBC was more often applied in older patients, who had high Waterlow scale scores and worse mobilisation status. These high risk parameters may be the reason for the loss of the known protective feature of the ZnBC in our study. Further studies are needed to overcome this bias. Additionally, in this present study, we found that the day of HAPI opening was later in patients who used protective creams than in patients who did not. This finding may be accepted as a beneficial effect of the use of ZnBC.
The hospitalisation period is an important parameter in HAPI formation. Similar to the findings of Olivo et al.,14 we found the longer the patient stayed at the hospital, the greater the risk of HAPI occurrence. In our study, the risk increased 7.9 times in 7–14 days, 10.11 times in 15–29 days, 39.81 times in 30–89 days, and 56.04 times in ≥90 days staying at the hospital when compared with the 1–6 days group. This may be interpreted in two ways. First, the patients who need to stay longer at the hospital have more severe medical conditions which lead to an increased risk of PI formation. Second, the sooner we ameliorate the cause of patients' hospitalisation and discharge them, the lower the risk of PI occurrence. This condition is similar in healthcare-associated infections (HAI)—the longer the patient stays in a healthcare facility the greater the risk for HAI.22
Limitations
The main limitations of this study are its retrospective design without randomisation, as well as lack of regular positioning documentation. The patient selection for ZnBC and airbed use was not randomised. Nevertheless, we believe that a prospective and randomised study should also be performed to validate these findings. We also tried to regularly reposition all of our patients; however, this could not be achieved, especially with regards to those patients who could not tolerate repositioning.
Conclusion
PIs, especially when they occur during a hospital stay, may have medical, legal and economic implications. In our study, we found an increased risk of HAPIs in male patients, lack of efficacy of ZnBC on HAPI formation, delay in the day of the HAPI opening with ZnBC use, the protective effect of airbed use, and an increasing risk of HAPI formation with a longer hospitalisation period.
Reflective questions
- What is a pressure injury (PI) and how does it occur?
- What effect did the COVID-19 pandemic have on PI incidence?
- Which factors predispose to PIs?
- In our daily practice, how can we safeguard against the risk of developing hospital-acquired PI?