Received: 2 May 2020 | Revised: 29 June 2021 | Accepted: 30 June 2021

DOI: 10.1111/jocn.15974

ORIGINAL ARTICLE

Risk factors of medical device-­related pressure injury in

intensive care units

Wen Dang MD1,2 | Yuan Liu MD3 | Qing Zhou MD3 | Yuyu Duan MD4 |
Huaxiu Gan Bsc5 | Lin Wang MD6 | Qiongli Zhu Bsc7 | Chunyan Xie Bsc8 | Ailing Hu MD3

1
School of Nursing, Sun Yat-­sen
University, Guangzhou, China Abstract
Objectives: Intensive care unit (ICU) patients are at high risk of medical device related
2
Mental Health Centre, West China
Hospital, Sichuan University, Chengdu,
China
pressure injury (MDRPI). This study aims to ascertain the MDRPI prevalence in ICU
3
The Third Affilated Hospital of Sun patients and analyse the risk factors of MDRPI.
Yat-­sen University, Lingnan Hospital, Background: The occurrence of MDRPI not only increases hospitalisation time with
Guangzhou, China
4 pain and economic burden, but also causes medical disputes. A better understanding
State Key Laboratory of Oncology in
South China, Sun Yat-­sen University of this condition will increase knowledge and facilitate the ability to recognise and
Cancer Center, Collaborative Innovation
prevent MDRPI for clinical nursing staff. However, there are few multicentre studies
Center for Cancer Medicine, Guangzhou,
China of MDRPI prevalence in ICU patients in China.
5
Zhujiang Hospital of Southern Medical Design: A cross-­sectional study design was employed.
University, Guangzhou, China
6 Methods: Data from 694 patients in 66 adult ICU at 30 hospitals in China were in-
Zhongshan People’s Hospital, Zhongshan,
China cluded between October 2018 and March 2019. The stage of each MDRPI was de-
7
Meizhou People's Hospital, Meizhou, termined according to the definitions of National Pressure Ulcer Advisory Panel. The
China
8
study methods were followed by the STORBE guidelines.
People’s Hospital of Yangjiang, Yangjiang,
China Results: The overall prevalence rate of MDRPI was 13.1% (91/694), with 98 anatomic
locations in total. The most common stages of MDRPI were stage 1 (54.1%, 53/98),
Correspondence
Ailing Hu, Third Affiliated Hospital of Sun stage 2 (15.3%, 15/98) and mucosal membrane pressure injury (15.3%, 15/98). MDRPI
Yat-­Sen University, 600 Tianhe Road, mainly occurred in the finger (32.7%, 32/98), followed by nose (18.4%, 18/98). The
Guangzhou, 510630, China.
Email: [email protected] prevalence rate of MDRPI caused by CPAP or BiPAP masks (25%) was highest. Lower
Braden scores and having skin oedema were risk factors for MDRPI in adult ICU
Funding information
This study was supported financially by patients.
the grants from the Guangdong Nursing Conclusion: The prevalence of MDRPI in this study was still high. Nurses should take
Association (gdhlxueh2019zx 316, China).
these related factors into consideration when taking care of ICU patients, and appro-
priate prevention measures should be adopted to decrease the prevalence of MDRPI.
Relevance to clinical practice: The study can help to improve the PI prevention efforts
in ICU patients specific to medical device related PI.

KEYWORDS
intensive care, medical device related pressure injury, prevalence, risk factors

J Clin Nurs. 2021;00:1–10. wileyonlinelibrary.com/journal/jocn© 2021 John Wiley & Sons Ltd | 1
2 | DANG et al.

1 | I NTRO D U C TI O N in Saudi Arabia and found that 115 patients (26.7%) had at least 1
case of MDRPI. Barakat-­Johnson et al. (2017) conducted a 13-­month
Pressure injury (PI) is localised damage to the skin and underlying survey of all inpatients in a tertiary hospital in Australia, and the re-
soft tissue that usually occurs over a bony prominence or in relation sults showed that the overall incidence of MDRPI was 27.9%, with
to the use of a medical or other device (Edsberg et al., 2016). For the highest incidence in the ICU.
a long time, researchers have focused on preventive interventions The occurrence of pressure injury not only aggravates the suf-
and monitoring of patients at high risk of pressure injury at a bony fering of patients but also prolongs their hospitalisation time, in-
prominence. Increasing authors have realised that PI not only occurs creases the economic burden, and even causes medical disputes
over bony prominences but also may occur due to medical device (Dealey et al., 2012; Demarre et al., 2015). Thus, to decrease the
use. The term medical device related pressure ulcer (MDRPU) was pain of ICU patients and improve their health, it is necessary for
first mentioned by Black et al. (2010) which attracted a lot of atten- us to focus on the problem to avoid the occurrence of MDRPI. An
tion. National Pressure Ulcer Advisory Panel (NPUAP) defined the increasing number of researchers have begun to pay attention to
MDRPU as pressure injuries associated with the use of devices ap- the prevalence of ICU medical device-­ related pressure injury and
plied for diagnostic or therapeutic purposes wherein the PU that de- explore its influencing factors. However, due to the differences in
velops has the same configuration as the device (Black et al., 2013). population ethnicity, research methods and assessment tools, the
Until 2016, NPUAP officially included medical device-­related pres- results are inconsistent. At present, domestic large sample inves-
sure injury (MDRPI) into its own category of pressure injury and de- tigations mostly focus on pressure injury at the bony prominence,
fined it as a PI caused by the use of medical devices for diagnosis or even including those PIs caused by related medical devices, but do
treatment, with the shape of the injured part being consistent with not analyse the related risk factors. The research on MDRPI is mainly
that of medical devices (Edsberg et al., 2016). focused on the investigation of small samples and single centre, so
With the development of science and technology, there are the sample lacks representation. In addition, as a sensitive indicator
many kinds of medical devices, but for patients, the use of medical of the quality of care, clinical nurses aim to achieve a ‘zero defect’
devices can not only promote their recovery, but also may introduce level of PI (Zhou et al., 2017), so the occurrence of MDRPI is eas-
the possibility of secondary injury. ily overlooked. Additionally, the knowledge of nurses about MDRPI
is limited (Karadag et al., 2017). Therefore, research results to data
have not well reflected the current situation of MDRPI in China.
2 | BAC KG RO U N D

With the ageing of the population and the development of medi- 2.1 | Objectives
cal technology, an increasing number of medical devices used for
life support and auxiliary treatment have been put into clinical use, The aim of this study was to ascertain (i) the MDRPI prevalence of
which increases the risk of MDRPI (Arnold-­Long et al., 2017, Kayser ICU patients, (ii) describe their characteristics, and analyse the pos-
et al., 2018). A study of 86,932 acute care patients in the United sible risk factors for MDRPI in general hospitals in different areas of
States found that among all identified PI, 9.1% of PI was described as Guangdong Province in China.
‘device-­related’ (VanGilder et al., 2009). Black et al., (2010) analysed
the data from a cross-­sectional point prevalence study and found
that an overall rate of nosocomial acquired pressure injury was 5.4%, 3 | M E TH O D S
among which the proportion of MDRPI was 34.5%. A severe PI re-
port from 34 hospitals in Minnesota found that 29% of severe PI was 3.1 | Study design
caused by the use of medical devices (Apold and Rydrych, 2012).
Therefore, when medical devices become an important part of pa- A cross-­sectional study design was employed, according to
tients’ diagnosis or treatment, the chances of patients developing the Strengthening the Reporting of Observational studies in
MDRPI may be greatly increased, especially in the ICU environment. Epidemiology (STORBE) guidelines (File S1). Data were collected be-
ICU patients are at high risk of MDRPI due to their critical con- tween October 2018 and March 2019.
dition, the inability of some unconscious patients to express feel-
ings of pain, and the long use of a large number of medical devices.
Therefore, ICU patients, especially those who depend on medical 3.2 | Settings
devices to maintain their lives, are more likely to develop MDRPI. A
prospective descriptive study of five adult intensive care units at a The study was conducted at 30 hospitals located in 14 cities in
university hospital in Turkey found that the prevalence of nosoco- Guangdong Province, China. All patients admitted to adult intensive
mial acquired pressure injury was 15%, and 40% had MDRPI during care units (surgical, medical, cardiovascular, neurosurgical and other
the course of the study (Hanonu and Karadag, 2016). Amirah et al., intensive care units) at the participating hospitals were surveyed on
(2017) investigated 431 adult ICU patients in a large tertiary hospital a single predetermined day.
DANG et al. | 3

3.3 | Participants 3.5 | Data collection

Participants were selected by mixed sampling method. According Three days before each investigation, the researcher contacted the
to the administrative division and location, stratified sampling was nursing department at the local hospital and trained the enteros-
used to divide Guangdong Province into four parts, including the tomal therapists or nurse specialists, who were trained in assessing
Pearl River Delta, Eastern region, Western region and Northern MDRPI, to explain the aim of this study and the method to collect
region. Then, convenience sampling was carried out in each region information. To ensure the consistency of methodology and the
according to the numbers and cooperation of hospitals. Finally, all accuracy of participating nurses in recognising MDRPI, ten photo-
ICU patients who met the inclusion and exclusion criteria were se- graphs of each MDRPI were identified, and the stage of MDRPI was
lected by cluster sampling. The inclusion criteria for the hospitals determined according to the definitions by the National Pressure
were as follows: (i) tertiary public hospitals in Guangdong Province Ulcer Advisory Panel. According to the accuracy of recognition
and (ii) agreement and support were obtained from the manage- score, scores ≥90 were qualified for participation in the study. This
ment department. Specialised hospitals were excluded. Adult pa- purpose is to ensure that investigators have the ability to accurately
tients (age ≥18 years) in the ICU who were more than 24 hr after identify MDRPI, so as to reduce the Investigators’ bias. The pass-
admission were eligible for inclusion. The exclusion criteria for the ing rate of the test was 100%. During the investigation, the re-
patients were as follows: (i) patients who had skin diseases such as searcher and the trained ET or nurse specialist evaluated the ICU
systemic lupus erythaematosus and burns that may have influenced patient together, and the data collection was recorded by the re-
the judgement of MDRPI; (ii) patients who were not in the ICU for searcher alone. If there was a difference of opinion on judgement
treatment or examination during the investigation; and (iii) patients of the MDRPI, the third ET or nurse specialist was invited to make
with MDRPI before they were admitted to the ICU. a decision.

3.4 | Variables and measurement 3.6 | Statistical analysis

The survey consisted of 3 main sections: patient characteristics in- Quantitative data were expressed as the means and standard
formation, the Braden scale and the MDRPI data. deviations, whereas qualitative data were described as absolute
Patient characteristics information was designed by the research- frequencies and percentages. Data were analysed using independ-
ers based on the literature review and nursing experts, including 6 ent t test, chi-­square test, and Fisher's exact test. Variables that
items continuous variables (age, length of ICU stay, albumin, haemo- were significantly associated with the presence of MDRPI in uni-
globin, lactic acid and the number of medical devices that patients variate analyses were introduced in a logistic regression model to
used) and 11 items categorical variables (location of hospital, type examine the associations between the independent variables and
of hospital, type of ICU, gender, primary diagnosis, chronic disease presence of MDRPI. All analyses were performed with SPSS 20.0.
history, nutrition mode, skin oedema, medicine usage, mechanical Results with a p value less than 0.05 were considered statistically
ventilation usage and surgical treatment). According to the medical significant.
device used in the clinic and the results of other scholars, the poten-
tial medical devices related to pressure injury were separated into six
different types, including but not limited to respiratory devices, mon- 3.7 | Ethical consideration
itoring devices, catheters, orthopaedic devices and other devices.
The Braden scale is designed by Braden and Bergstrom in 1987 The investigation on each hospital was under the permission of in-
which has been widely used with high reliability and validity (Braden struction of Guangdong Nursing Association. The researcher sent
and Bergstrom, 1996; Wei et al., 2020). This scale has 6 items, in- the invitation letters to the selected hospital and obtained agree-
cluding perception, moisture, mobility, activity, friction, shear and ment before the investigation. During the research, researchers
nutrition. The total score ranges 6 to 23. The lower score means a need to check the patients’ skin with their consent, and patients had
higher risk. The Cronbach alpha coefficient for the Braden scale in the right to refuse. Finally, after we finished this survey, the data of
this study was 0.752. every patient was recorded on the prepared paper, and put in an
The MDRPI data collection form was developed by researchers opaque envelope to make sure the personal information was safety.
to record the prevalence and characteristics of the MDRPI. Items
on the instrument included the location and classification of MDRPI
as well as the type of medical device that led to MDRPI. As the out- 4 | R E S U LT S
come of this study, the identification of MDRPI conducted by the
researchers and trained ETs or nursing specialists, and the stage of In this study, there were 21 hospitals from 8 cities in the Pearl River
each MDRPI was determined according to the definitions by the Delta, 3 hospitals in the eastern region, 2 hospitals in the western
National Pressure Ulcer Advisory Panel (Edsberg et al., 2016). region and 4 hospitals in the northern region. A total of 694 ICU
4 | DANG et al.

patients at 30 general hospitals in Guangdong Province were inves- Sedatives (24.6%) and vasopressors (42.2%) were the common med-
tigated in this study, 91 of whom had MDRPI. ications used by these ICU patients. 65.9% patients received tube
feeding as nutritional support. Nearly a half of the patients (45.0%)
received mechanical ventilation, and 25.2% had surgical experience.
4.1 | Prevalence rate of MDRPI There were various types of medical devices used in the ICU, and
the most commonly used medical devices were monitoring devices.
The prevalence rate of MDRPI in each of the 30 hospitals ranged In addition, at least one medical device was used to diagnose or treat
from 0% to 33.3%, indicating an average prevalence of 13.1%. When each patient, with a mean number of 7.22 (Table 2).
stage 1 MDRPI was excluded, the total prevalence rate was 5.9%.
Compared with the other regions in Guangdong Province, the prev-
alence rate of MDRPI in the Pearl River Delta was higher (18.8%) 4.3 | MDRPI characteristics
(χ2 = 38.223, p < 0.001). There was no significant difference in
MDRPI prevalence between university hospitals (12.4%) and general There were 91 patients with MDRPI in total of 98 anatomic locations.
2
hospitals (15.1%) (χ = 0.823, p = 0.364). Patients from the general The observed pressure injuries caused by medical devices occurred
ICU accounted for the largest proportion (62.5%). However, the dif- in both the skin (84.7%) and the mucous membrane (15.3%), and
ference in MDRPI prevalence between the surgical ICU and inter- these patients experienced stage 1, stage 2, deep tissue injury and
nal medicine ICU was statistically significant (χ2 = 7.939, p = 0.044) unstageable PI, while no stage 3 or stage 4 was found (Figure 1). The
(Table 1). most common anatomic locations of MDRPI were fingers (32.7%),
nose (18.4%), mouth and lips (16.3%), cheeks (7.1%) and legs (7.1%)
(Figure 2). There were 16 types of medical devices related to the
4.2 | Patient characteristics MDRPIs. In total of 98 MDRPIs, 32 MDRPIs were caused by the use
of oximetry probes, followed by endotracheal tubes (21/98) and na-
Among the 694 patients, there were more males (67.3%) than fe- sogastric tubes (15/98) (Figure 3). The prevalence rate of PI caused
males (32.7%). More than 50% of the patients were older than 60, by different medical devices varies. The prevalence rate of PI caused
with a mean age of 65.0 years (±17.4). A total of 65.6% of patients by CPAP or BiPAP masks (25%) was highest, followed by the figure
were admitted to the ICU for less than 10 days. for Bone tractors (20%), while the prevalence rate of PI caused by
Nearly one-­third of the patients (30.4%) had been admitted for ECG electrodes was lowest, with 0.14% (Table 3).
respiratory disease reasons. Most ICU patients (60.1%) had a his-
tory of chronic illness. According to the skin evaluation results, only
2.7% of the patients were not at risk of pressure injury and most of 4.4 | Risk factors related to MDRPI in the ICU
them (68%) were at high risk (Braden≤12). There were 217 patients
who had skin oedema. The laboratory results showed that albumin The results of the univariate analysis showed that MDRPI was more
and haemoglobin levels were below normal in most ICU patients. likely to occur in ICU patients with the following characteristics:

TA B L E 1 Prevalence of MDRPI in Guangdong Province

No. of hospitals No. of patients No. of MDRPI

n (%) N (%) patients N (%) Prevalence(%) 𝝌2 p
a
Location of the hospital 38.223 <0.001
Pearl River Delta (8) 21 (70.0) 430 (62.0) 81 (89.0) 18.8
East region (2) 3 (10.0) 123 (17.7) 0 (0.0) 0.0
West region (2) 2 (6.7) 61 (8.7) 1 (1.1) 1.6
North region (2) 4 (13.3) 80 (11.5) 9 (9.9) 11.3
Type of hospital 0.823 0.364
University hospital 22 (71.0) 515 (74.2) 64 (70.3) 12.4
General hospital 9 (29.0) 179 (25.8) 27 (29.7) 15.1
Type of ICU 7.939 0.044
General ICU 434 (62.5) 59 (64.8) 13.6
Internal medicine ICU 143 (20.6) 11 (12.1) 7.7
Surgical ICU 82 (11.8) 17 (18.7) 20.7
Emergency ICU 35 (5.0) 4 (4.4) 11.4
Total 30 (100.0) 694 (100.0) 91 (100.0) 13.1
DANG et al. | 5

TA B L E 2 Patient characteristics TA B L E 2 (Continued)

Percentage Percentage
Characteristics No. (%) Mean (SD) Characteristics No. (%) Mean (SD)

Sex Below normal 487 70.2

Male 467 67.3 Normal 187 26.9
Female 227 32.7 Above normal 20 2.9
Age (year) 64.97 ± 17.43 Lactic acid (mmol/L) 2.09 ± 2.97
18~ 99 14.3 Below normal 163 23.5
45~ 144 20.7 Normal 315 45.5
60~ 212 30.5 Above normal 216 31.1
75~ 207 29.8 Medicine usage
90~99 32 4.6 Vasopressors 293 42.2
Length of ICU stay (day) 19.74 ± 61.93 Sedatives 171 24.6
≤10 455 65.6 Anticoagulants 76 11.0
>10 239 34.4 Steroids 18 2.6
Primary diagnosis Cytotoxic agents 3 0.4
Neurological 173 24.9 Mechanical ventilation
Cardiovascular 160 23.1 Yes 312 45.0
Respiratory system 211 30.4 No 382 55.0
Gastrointestinal 60 8.6 Surgical experience
Urinary 27 3.9 Yes 175 25.2
Endocrine system 11 1.6 No 519 74.8
Other 52 7.5 Number of medical 7.22 ± 1.53
devices
History of chronic illness
No 271 39.0
Yes 423 60.1 surgical ICU stay, longer length of stay, use of a large number of
Hypertension 290 41.8 medical devices, lower Braden score, skin oedema, parenteral nutri-
Diabetes 151 21.8 tion and use of vasopressors (p < 0.05) (Table 4).
Stoke 96 13.8 Factors that were significantly associated with MDRPI in uni-
Other 62 8.9 variate analyses were included in a logistic regression model. The
Nutrition result showed that the higher Braden score, the lower risk of MDRPI
Oral 168 24.2 (OR = 0.83, 95% CI: 0.76~0.91); having skin oedema (OR = 3.43, 95%

Enteral 457 65.9

CI: 2.16~5.45) is the risk factor for MDRPI in ICU patients (Table 5).

Parenteral 69 9.9
Oedema
5 | DISCUSSION
Yes 217 31.3
No 477 68.7
5.1 | The prevalence rate of MDRPI in ICU patients
Braden score (0~23) 11.87 ± 2.66
is still at a high level
No risk (≥19) 19 2.7
Low risk (15~18) 90 13.0 ICU patients are a high-­risk group for MDRPI, and the incidence of
Moderate risk (13~14) 113 16.3 MDRPI in adult ICU patients has been reported in literature to be
High risk (10~12) 341 49.1 3.1% to 40% (Amirah et al., 2017; Coyer et al., 2014; Hanonu and
Extreme risk (≤9) 131 18.9 Karadag, 2016; Mehta et al., 2019; Koo et al., 2019). Methodological
Albumin (g/L) 34.01 ± 9.22 differences in research may also account for these differences. In

Below normal 596 85.9 this study, the prevalence of MDRPI was 13.1%, which was lower
than the cross-­sectional point prevalence reported by Mehta et al.
Normal 73 10.5
(2019). The difference may be related to the different sample sizes
Above normal 25 3.6
and ICU types. 146 patients who admitted in three ICU type (medi-
Haemoglobin (g/L) 103.65 ± 62.90
cal, cardiovascular and neurosurgical) in one tertiary care hospital
(Continues) in India were included in Mehta’s study, and the ICU type was quite
6 | DANG et al.

Unstageable F I G U R E 1 The classification of the

Mucosal membrane 7, 8.4% MDRPIs
pressure injury
15, 15.3%
Deep tissue injury
8, 9.6%

Stage 2
15, 18.1% Stage 1
Skin pressure injury 53, 63.9%
83, 84.7%

Fingers 32, 32.7% F I G U R E 2 The anatomic locations of

the MDRPIs
Nose 18, 18.4%
Mouth and lips 16, 16.3%
Cheeks 7, 7.1%
Legs 7, 7.1%
Arms 5,5.1%
Ears 4, 4.1%
Neck 4, 4.1%
Abdomen 2, 2.0%
Mandible 1, 1.0%
Chest 1, 1.0%
Foot 1, 1.0%

TA B L E 3 The prevalence rate of PI caused by different medical devices

Total Number of patients who used per Total Number of Pressure injury per Percentage of MDRPU
Type of Medical devices medical device medical device per medical device

CPAP or BiPAP mask 32 8 25%

Bone tractor 5 1 20%
Anti-­embolic stocking 6 1 16.7%
Endotracheal tube 273 21 7.69%
Elastic bandage 15 1 6.67%
Oximetry probe 690 32 4.64%
Tracheostomy tube 132 4 3.03%
Nasogastric tube 495 15 3.03%
Drainage tube 124 2 1.61%
Urinary catheter 544 6 1.10%
Arterial catheter 170 1 0.59%
Oxygen tube 219 1 0.46%
Restraining band 253 1 0.40%
Blood pressure cuff 678 2 0.29%
Central catheter 375 1 0.27%
ECG electrodes 690 1 0.14%

different from the ICU in this study. Additionally, patients admitted In Guangdong Province, Pearl River Delta was the region where
in ICU with different disease; therefore, they had to use different the prevalence rate of MDRPI was highest (18.8%), followed by the
medical devices which may lead to the difference. figure for North region (11.3%). The distribution trend of MDRPI is
DANG et al. | 7

TA B L E 4 Univariate analysis of MDRPI

Non-­MDRPI MDRPI
2
Factors group group t/𝝌 p

Number of medical devices 7.16±1.573 7.62±1.227 −2.514 0.012

Braden scores 12.07±2.67 10.54±2.10 −6.249 <0.001
Oedema 38.405 <0.001
Yes 163 (27.0) 54 (59.3)
No 440 (73.0) 37 (40.7)
Length of ICU stay 6.367 0.012
≤10 406 (67.3) 49 (53.8)
>10 197 (32.7) 42 (46.2)
Type of ICU 7.939 0.044
General ICU 375 (62.2) 59 (64.8)
Internal medicine ICU 132 (21.9) 11 (12.1)
Surgical ICU 65 (10.8) 17 (18.7)
Emergency ICU 31 (5.1) 4 (4.4)
Nutrition 10.184 0.006
Oral 157 (26.0) 11 (12.1)
Enteral 391 (64.8) 66 (72.5)
Parenteral 55 (9.2) 14 (15.4)
Use of Vasopressors 6.953 0.008
Yes 243 (40.3) 50 (54.9)
No 360 (59.7) 41 (45.1)

TA B L E 5 Risk factors associate with the presence of MDRPI The most common stage of MDRPI was stage 1, and this result
was supported by other researchers (Black et al., 2010; Hanonu
95%
Odds Confidence and Karadag, 2016; Koo et al., 2019). However, some researchers
Factors ratio interval p reported that MDRPI was found to be stage 3 or stage 4, which
demonstrates the lack of best practices in assessing and preventing
Braden scores, mean 0.83 0.76–­0.91 <0.001
MDRPI. Apold and Rydrych (2012) considered unstageable PI to be
Oedema
the most common stage of PI caused by devices. Therefore, if ICU
No* (0)
nurses can be trained to more accurately identify high-­risk popula-
Yes* (1) 3.43 2.16–­5.45 <0.001
tions and provide timely preventive measures to prevent the pro-
*Coding in brackets. Reference category = 0. gression of MDRPI to a more serious degree of injury, the prevalence
rate of MDRPI can be greatly reduced. In addition, the appearance of
similar to the distribution trend traditional PI in Guangdong province mucosal membrane injury is one of the differences between MDRPI
(Zhou et al., 2017). The cause of this result may be due to the seri- and traditional PI, which also suggests that nurses’ care of medical
ous condition of patients in the Pearl River Delta region. As one of device users is not limited to the skin area but that observation and
the most economically developed areas in Guangdong Province and care of the mucosal areas are equally important.
even in China, the Pearl River Delta region has gathered a complete
set of medical resources, so as to attract critical patients from the
province and all over the country. In addition, due to the relatively 5.2 | The most common anatomic
developed economy and complete medical facilities in the Pearl location of MDRPI
River Delta region, a large number of medical devices are used for
treatment or diagnosis, while some medical devices and facilities in The result of this study showed that the most common anatomic
ICU may be insufficient in other regions, which increases the proba- locations of MDRPI were fingers (32.7%), nose (18.4%), and mouth
bility of using medical devices in patients, thus increasing the risk of and lips (16.3%). Hanonu et al. (2016), observed the location of most
MDRPI. By contrast, the prevalence of MDRPI in West region was MDRPIs was lips (44.0%), followed by nose (15.6%) and fingers
only 1.6%, and there was no MDRPI found in East region. The reason (7.5%). Black et al. (2010) and VanGilder et al. (2009) observed the
of this result may because that the average number of patients using most common locations of MDRPI were ears, 20% and 35% respec-
medical devices in these two areas was less than that in the other tively. These studies support that the anatomic location of MDRPI
two areas. varies which depends on the site of medical device use.
8 | DANG et al.

35 32 F I G U R E 3 The frequencies of
MDRPI caused by different medical
30
devices
25
21
20
15
15

10 8
6
4
5 2 2
1 1 1 1 1 1 1 1
0

5.3 | The most common device that causes MDRPI disorder that causes local limb swelling. The use of certain medical
devices, such as anti-­embolism stocking, may result in a ‘tourniquet’
Although the number of MDRPI caused by oximetry probe was high- effect (Bowling et al., 2015). Therefore, it is particularly important to
est, the prevalence rate of MDRPI by oximetry probe was lower due select medical devices of appropriate size for each patient, to ensure
to the large number of patients using. By contrast, the number of pa- that the devices are worn correctly and to adjust the devices at any
tients using CPAP or BiPAP mask was smaller, a higher rate of MDRPI time according to the patients’ conditions. It is recommended that
(25%) development was observed in these patients. Mehta et al. the frequency of examination below the medical device should be in-
(2019) in a point prevalence study found that MDRPIs were most creased for patients with oedema (National Pressure Ulcer Advisory
commonly associated with the use of non-­invasive ventilation mask Panel, European Pressure Ulcer Advisory Panel and Pan Pacific
(20%). CPAP and BIPAP, as common respiratory support methods Pressure Injury Alliance, 2014). As a high-­risk group for MDRPI, the
for respiratory distress syndrome, are widely used in ICU patients. skin care of patients with oedema should receive increased attention
However, in order to ensure the effectiveness of treatment, the mask from nursing staff, and active measures should be taken to improve
should be kept close to the skin during treatment to avoid air leakage. the primary disease of patients and reduce the oedema caused by
Therefore, when the patient uses the device, the nose, face and ear long-­term bed rest or malnutrition.
become the most pressurised parts, finally leading to MDRPI devel- Besides, the univariate analysis represented use numerous med-
opment. Nursing staff should pay more attention to these patients. ical devices, longer length of stay, surgical ICU stay, parenteral nutri-
tion and use of vasopressors also related to the presence of MDRPI.
Research from Black et al. (2010) showed that patient with a
5.4 | Risk Factors related to MDRPI in the ICU medical device in place was 2.4 times more likely to develop pres-
sure injury of any kind. Our research also emphasis that MDRPI is
The results of this study indicated that the higher the Braden score, more likely to occur in patients who use more medical devices. This
the lower the risk of MDRPI; Patients with skin oedema had a 3.43 may be attributed to the fact that patients who used numerous med-
times higher risk of developing MDRPI than those without skin ical devices are usually at higher risk with respect to life functions,
oedema. thus they may exist in lower nutrition, limited sensory etc. In addi-
MDRPI develops in ICU patients whose Braden Scale is low. The tion, the more medical devices patients used, the more time nurses
findings of the current study are supported by the literature (Apold paid. Once in neglect, it may lead to the presence of MDRPI.
and Rydrych, 2012; Black et al., 2010; Coyer et al., 2014; VanGilder This research indicated that patients with a longer ICU stays
et al., 2009). The conclusion of this study demonstrates that MDRPI (particularly ≥10 days) were more likely to have MDRPI. Mehta et al.
has similar risk factors as traditional PI. Patients with low Braden (2019) found that a higher length of ICU stays was associated with
scores may experience sensory limitations, reduced mobility and nu- presence of MDRPI (OR = 9.8, p < 0.0001). Hanonu and Karadag
tritional deficiencies. Therefore, MDRPI may be caused by decreased (2016) also found the incidence of MDRPI increased with the length
sensation or inability to express pain and feeling in the injured parts, of hospital stay, and MDRPIs occurred as early as 24 hr after admis-
which may lead to the occurrence or delayed detection of MDRPI. sion to the ICU and continued to occur through day 11. This result
Oedema increases pressure and deterioration of nutrient ex- highlights the importance of early identification of high-­risk patients
change in tissues. Long-­term use of medical devices without pro- and timely provision of preventive measures to prevent MDRPI.
tection may increase the vulnerability of the skin and soft tissues in Differences were found in the prevalence rates of MDRPI in dif-
oedema patients. When a patient has oedema, there may be a reflux ferent types of ICUs. In this study, the highest prevalence rate of
DANG et al. | 9

MDRPI was in the surgical ICU (20.7%), followed by the general ICU can improve the attention of the clinical nurses on MDRPI of ICU
(13.6%). The reason can be attributed to the greater use of medical patients. Additionally, the knowledge on the risk factors of MDRPI
instruments in surgical ICUs. Hanonu and Karadag (2016) showed can help nurses to identify the MDRPI earlier and take measures
that prevalence in the internal medicine ICU ranked first, while preventing patients from MDRPI.
VanGilder et al. (2009) represented that the most of the MDRPIs
occurred in general ICU. However, in these studies, the classification
of ICU types is different, which limits the possibility of comparing 6.2 | What does this paper contribute to the wider
studies with respect to MDRPI developing patients. global clinical community?
Different nutritional patterns are related to the occurrence of
MDRPI. Although the number of patients receiving parenteral nu- 1. Ascertain the clinical data of MDRPI in ICU patients in China
trition was smaller, a higher rate of MDRPI (20.3%) development area which can communicate with other countries and areas.
was observed in these patients than those patients receiving enteral 2. The risk factors of MDRPI in ICU patients is suitable for all
feeding (14.4%) and oral feeding (6.5%). This result may attribute to countries.
the fact that compared to parenteral nutrition, the use of enteral 3. It is suggested that high-­risk devices may cause MDRPI in clinic.
nutrition was associated with a significant reduction in infectious 4. Findings indicate a need for nursing staff to have access the
complications and ICU length of stay (Elke et al., 2016), so the risk knowledge of prevention MDRPI.
of developing MDRPI also was reduced. Yet in the prospective study
by Hanonu and Karadag (2016), MDRPI risk increased 2.12 times in
enteral feeding patients. Because of the different study design, this 6.3 | Limitation
result need to be demonstrate in the future researches.
The use of vasopressors can increase mean arterial pressure and Although the survey covered 30 tertiary general hospitals in differ-
can lead to tissue ischaemia, hypoxia damage, and even necrosis of ent regions of Guangdong Province, the sample in this study was
the extremities, mesenteric organs and kidneys (Cox and Roche, limited to adult ICU patients. It did not include all inpatients with
2015). Therefore, patients who use vasopressors have a higher prev- MDRPI, and the occurrence of MDRPI in general departments is
alence of PI as well as MDRPI. Koo et al. (2019) supported this re- not described. The research in the future can increase the sample
sult. However, the use of steroids and sedatives can increase the risk size and expand the study group to patients in the whole hospital to
of MDRPI in other literature (Hanonu and Karadag, 2016). Steroids compensate for the deficiency of domestic epidemiological research
prevent the formation of collagen fibres, and the use of sedatives results regarding MDRPI. Moreover, because of the wide variety of
can affect patients’ sensory ability, which can prevent patients from devices that lead to MDRPI, multidisciplinary research is also impor-
adequately expressing any discomfort caused by the use of medical tant to clinical practice.
devices. Besides, the ability of nursing staffs to mitigate MDRPI risk
for ICU patients receiving these medications may be limited because C O N FL I C T O F I N T E R E S T
these agents are life-­saving modalities that cannot be terminated to None.
prevent MDRPI development. So patients who use this kind of medi-
cines should be paid more attention to ensure their safety. AU T H O R C O N T R I B U T I O N
Wen Dang, Yuan Liu and Ailing Hu involved in study design. Wen
Dang, Huaxiu Gan, Lin Wang, Qiongli Zhu and Chunyan Xie involved
6 | CO N C LU S I O N in data collection. Wen Dang, Yuyu Duan and Qing Zhou involved in
data analysis. Wen Dang, Yuan Liu, Ailing Hu, Yuyu Duan and Qing
ICU patients are at high risk for MDRPI. The prevalence of MDRPI Zhou involved in manuscript preparation.
in this study was still high, especially in hospitals located in the Pearl
River Delta in Guangdong Province. Skin injury was more common
than mucosal membrane injury. There were a wide variety of de- ORCID
vices that caused MDRPI. Nurses should identify the risk factors Wen Dang https://orcid.org/0000-0001-5554-9103
for MDRPI early and intervene early to prevent the occurrence of
MDRPI in ICU patients. REFERENCES
Amirah, M. F., Rasheed, A. M., Parameaswari, P. J., Nu’man, O. S., &
Al Muteb, M. (2017). A cross-­sectional study on medical device-­
related pressure injuries among critically ill patients in Riyadh,
6.1 | Relevance to clinical practice
Kingdom of Saudi Arabia. WCET Journal, 37(1), 8–­11.
Apold, J., & Rydrych, D. (2012). Preventing device-­related pressure ul-
As the high-­risk group, ICU patients have more potential to have cers: Using data to guide statewide change. Journal of Nursing Care
MDRPI during the time in intensive care unit. This study showed Quality, 27(1), 28–­3 4. https://doi.org/10.1097/NCQ.0b013​e3182​
2b1fd9.
the present occurrence of MDRPI in ICU patients in China, which
10 | DANG et al.

Arnold-­Long, M., Ayer, M., & Borchert, K. (2017). Medical device-­related Karadag, A., Hanönü, S. C., & Eyikara, E. (2017). A Prospective, descrip-
pressure injuries in long-­term acute care hospital setting. Journal of tive study to assess nursing staff perceptions of and interventions
Wound Ostomy & Continence Nursing, 44, 325–­330. to prevent medical device-­related pressure injury. Ostomy Wound
Barakat-­Johnson, M., Barnett, C., Wand, T., & White, K. (2017). Medical Manage, 63(10), 34–­41.
device-­related pressure injuries: An exploratory descriptive study Kayser, S. A., VanGilder, C. A., Ayello, E. A., & Lachenbruch, C. (2018).
in an acute tertiary hospital in Australia. Journal of Tissue Viability, Prevalence and analysis of medical device-­related pressure in-
26(4), 246–­253. https://doi.org/10.1016/j.jtv.2017.09.008. juries: Results from the International Pressure Ulcer Prevalence
Black, J., Alves, P., Brindle, C. T., Dealey, C., Santamaria, N., Call, E., & Survey. Advances in Skin & Wound Care, 31(6), 276–­285. https://doi.
Clark, M. (2013). Use of wound dressings to enhance prevention org/10.1097/01.ASW.00005​32475.11971.aa.
of pressure ulcers caused by medical devices. International Wound Koo, M., Sim, Y., & Kang, I. (2019). Risk factors of medical device-­related
Journal, 12(3), 322–­327. https://doi.org/10.1111/iwj.12111. pressure ulcer in intensive care units. Journal of Korean Academy
Black, J. M., Cuddigan, J. E., Walko, M. A., Didier, L. A., Lander, M. J., & of Nursing, 49, 36–­45. https://doi.org/10.4040/jkan.2019.49.1.36.
Kelpe, M. R. (2010). Medical device related pressure ulcers in hospi- Mehta, C., Ali, M., Mehta, Y., George, J. V., & Singh, M. K. (2019).
talized patients. International Wound Journal, 7(5), 358–­365. https:// MDRPU-­an uncommonly recognized common problem in ICU:
doi.org/10.1111/j.1742-­481X.2010.00699.x. A point prevalence study. Journal of Tissue Viability, 28(1), 35–­39.
Bowling, K., Ratcliffe, C., Townsend, J., & Kirkpatrick, U. (2015). Clinical https://doi.org/10.1016/j.jtv.2018.12.002.
thromboembolic detterrent stockings application: Are thrombo- Park, 2014National Pressure Ulcer Advisory Panel, European Pressure
embolic detterrent stockings in practice matching manufacturers Ulcer Advisory Panel and Pan Pacific Pressure Injury Alliance.
application guidelines. Phlebology: The Journal of Venous Disease, Prevention and Treatment of Pressure Ulcers: Quick Reference
30(3), 200–­203. https://doi.org/10.1177/02683​55514​542843. Guide.[M]. Cambridge Media: Osborne Park, Western Australia,
Braden, B. J., & Bergstrom, N. (1996). Risk assessment and risk-­based 2014.
programs of prevention in various settings. Ostomy/Wound VanGilder, C., Amlung, S., Harrison, P., & Meyer, S. (2009). Results of the
Management, 42(10A Suppl), 6S–­12S. 2008–­2009 international pressure ulcer prevalenceTM survey and
Cox, J., & Roche, S. (2015). Vasopressors and development of pressure a 3-­year, acute care, unit-­specific analysis. Ostomy Wound Manage,
ulcers in adult critical care patients. American Journal of Critical 55(11), 39–­45.
Care, 24(6), 501–­510. https://doi.org/10.4037/ajcc2​015123. Wei, M., Wu, L., Chen, Y., Fu, Q., Chen, W., & Yang, D. (2020). Predictive
Coyer, F. M., Stotts, N. A., & Blackman, V. S. (2014). A prospective validity of the braden scale for pressure ulcer risk in critical care: A
window into medical device-­related pressure ulcers in intensive meta-­analysis. Nursing in Critical Care, 25(3), 165–­170. https://doi.
care. International Wound Journal, 11(6), 656–­664. https://doi. org/10.1111/nicc.12500.
org/10.1111/iwj.12026. Zhou, Q., Yu, T., Liu, Y., Shi, R., Tian, S., Yang, C., & Hu, A. (2017). The
Dealey, C., Posnett, J., & Walker, A. (2012). The cost of pressure ulcers in prevalence and specific characteristics of hospitalised pressure
the United Kingdom. Journal of Wound Care, 21(6), 261–­266. ulcer patients: A multicentre cross-­sectional study. Journal of
Demarré, L., Van Lancker, A., Van Hecke, A., Verhaeghe, S., Grypdonck, Clinical Nursing, 27(3–­4), 694–­704.
M., Lemey, J., Annemans, L., & Beeckman, D. (2015). The cost of
prevention and treatment of pressure ulcers: a systematic review.
International Journal of Nursing Studies, 52(11), 1754–­1774. https:// S U P P O R T I N G I N FO R M AT I O N
doi.org/10.1016/j.ijnur​stu.2015.06.006.
Additional supporting information may be found online in the
Edsberg, L. E., Black, J. M., Goldberg, M., McNichol, L., Moore, L., &
Sieggreen, M. (2016). Revised national pressure ulcer advisory Supporting Information section.
panel pressure injury staging system. Journal of Wound, Ostomy
and Continence Nursing, 43(6), 585–­597. https://doi.org/10.1097/
WON.00000​0 0000​0 00281. How to cite this article: Dang, W., Liu, Y., Zhou, Q., Duan, Y.,
Elke, G., van Zanten, A. R., Lemieux, M., McCall, M., Jeejeebhoy, K. N.,
Gan, H., Wang, L., Zhu, Q., Xie, C., & Hu, A. (2021). Risk
Kott, M., Jiang, X., Day, A. G., & Heyland, D. K. (2016). Enteral versus
parenteral nutrition in critically ill patients: an updated systematic factors of medical device-­related pressure injury in intensive
review and meta-­analysis of randomized controlled trials. Critical care units. Journal of Clinical Nursing, 00, 1–­10. https://doi.
Care, 20(1), 117. https://doi.org/10.1186/s1305​4-­016-­1298-­1. org/10.1111/jocn.15974
Hanonu, S., & Karadag, A. (2016). A Prospective, descriptive study to
determine the rate and characteristics of and risk factors for the
development of medical device-­related pressure ulcers in intensive
care units. Ostomy Wound Manage, 62(2), 12–­22.