Nursing Research Utilization: Literature Review and Solution in reducing pressure ulcers

Fa pr

Ev a Clin b Cen c De d De e De f Dep g Sch

International Journal of Nursing Studies 51 (2014) 717–725

A R

Artic

Rece

Rece

Acce

Keyw

Asso

Hos

Nur

Pati

Pres

Unit

Wor

*

002

http

ctors contributing to evidence-based pressure ulcer evention. A cross-sectional study

a Sving a,b,c,*, Ewa Idvall d,e, Hans Högberg b,f, Lena Gunningberg c,g

ical Training Centre, County Council of Gävleborg, Sweden

tre for Research & Development, Uppsala University/County Council of Gävleborg, Sweden

partment of Public Health and Caring Sciences, Caring Sciences, Uppsala University, Sweden

partment of Care Science, Faculty of Health and Society, Malmö University, Sweden

partment of Intensive Care and Perioperative Medicine, Skåne University Hospital, Malmö, Sweden

artment of Public Health Medicine, County Council of Gävleborg, Gävle, Sweden

ool of Nursing, University of California, San Francisco, USA

T I C L E I N F O

le history:

ived 3 September 2012

ived in revised form 13 August 2013

pted 18 September 2013

ords:

ciations

pital type

se staffing

ent characteristic

sure ulcer prevention

type

kload

A B S T R A C T

Background: Implementation of evidence-based care for pressure ulcer prevention is

lacking. As the hospital organization is complex, more knowledge is needed to understand

how nursing care in this area can be improved.

Objectives: The present study investigated the associations between variables on different

levels in the healthcare setting (patient, unit, hospital) and the documentation of (1) risk

assessment and (2) skin assessment within 24 h of admission, the use of (3) pressure-

reducing mattresses and (4) planned repositioning in bed.

Design: A cross-sectional study.

Settings: One university hospital and one general hospital.

Participants: Geriatric (n = 8), medical (n = 24) and surgical (n = 19) units. All adult patients

(>17 years), in total 825, were included.

Methods: A one-day prevalence study was conducted using the methodology specified by

the European Pressure Ulcer Advisory Panel, together with the established methods used

by the Collaborative Alliance for Nursing Outcomes. Independent variables were patient

characteristics, hospital type, unit type, nurse staffing and workload. Dependent variables

were documented risk and skin assessment within 24 h of admission, pressure-reducing

mattresses and planned repositioning in bed. The data were analysed with Logistic

regression using the Generalized Estimating Equation (GEE) approach.

Results: Patients at risk of developing pressure ulcers (Braden < 17) had higher odds of

having risk assessment documented, and of receiving pressure-reducing mattresses and

planned repositioning. Patients at the general hospital were less likely to have risk and skin

assessment documented and to receive pressure-reducing mattresses. On the other hand,

planned repositioning was more likely to be used at the general hospital. When total hours

of nursing care was lower, patients had higher odds of having pressure-reducing

mattresses but were less likely to have planned repositioning.

Conclusion: Patient characteristics (high age and risk score) and hospital type were

associated with pressure ulcer prevention. Surprisingly, nurse staffing played only a minor

role. Leaders in healthcare organizations should establish routines on different levels that

support evidence-based pressure ulcer prevention, and registered nurses need to assume

responsibility for bedside care.

� 2013 Elsevier Ltd. All rights reserved.

Corresponding author at: Centre for Research and Development, Uppsala university/County Council of Gävleborg, Sweden.

E-mail address: [email protected] (E. Sving).

Contents lists available at ScienceDirect

International Journal of Nursing Studies

journal homepage: www.elsevier.com/ijns

0-7489/$ – see front matter � 2013 Elsevier Ltd. All rights reserved. ://dx.doi.org/10.1016/j.ijnurstu.2013.09.007

E. Sving et al. / International Journal of Nursing Studies 51 (2014) 717–725718

What is already known about the topic?

� Pressure ulcer prevention remains a problem in health care. � Pressure ulcer prevention is lacking for patients with a

high risk for developing pressure ulcers.

What this paper adds

� Factors on different levels in the healthcare organization were examined simultaneously to determine which of them contribute to evidence-based pressure ulcer prevention in hospital settings. � When combining patient characteristic, hospital and unit

type, nurse staffing and workload in regression models, patients’ age, risk score, type of hospital and type of unit were associated with pressure ulcer prevention. � In the regression models, nurse staffing played a minor

role in the pressure ulcer prevention care.

1. Background

Pressure ulcers are related to reduced quality of life for patients (Gorecki et al., 2009) and high costs for the healthcare system (Bennett et al., 2004). Evidence-based international guidelines built on current research are available to health care professionals. However, studies show that pressure ulcer prevention is lacking (Moore and Cowman, 2012; Vanderwee et al., 2011). Although pressure ulcers are considered adverse events in health- care, the problem remains. In European studies, the reported prevalence of pressure ulcers varies between 9 and 18% (Barrois et al., 2008; Beeckman et al., 2011; Tannen et al., 2008). Pressure ulcers are classified according to four categories, where Category 1 is intact skin with non-blanchable redness and Category 4 is full thickness tissue loss with exposed bone, tendon or muscle (NPUAP & EPUAP, 2009). A hospital-acquired pressure ulcer is defined as a case with no documentation of an existing pressure ulcer in the patient’s records within 24 h of admission (Brown et al., 2010a).

The main patient-related risk factors are activity and mobility problems due to the risk they pose for pressure and/or shear on the skin. Early risk assessment including use of a combination of a validated risk assessment instrument, skin assessment and clinical judgement is crucial (NPUAP & EPUAP, 2009). When a patient is at risk for developing pressure ulcers, the multidisciplinary team should establish goals and a prevention plan together with the patient, and evaluate these goals regularly. Higher specification foam mattresses are recommended for all risk patients, according to a Cochrane review (McInnes et al., 2011). Repositioning should be undertaken to reduce the duration and magnitude of the pressure over vulnerable areas of the body (NPUAP & EPUAP, 2009). Defloor et al. (2005) reported that patients on pressure-reducing mattresses need less repositioning compared to patients on standard mattresses. Studies have also revealed that when preventive guidelines are implemented, the preva- lence of pressure ulcers decreases (Cuddigan et al., 2001; Lahmann et al., 2010).

Various factors on different levels in the healthcare organization have been examined in an attempt to understand how they affect patient outcomes. Examples include hospital organization, nurse staffing, workload and factors in direct patient care. The results are not clearcut. Estabrooks et al. (2007) reported that the only significant factor for higher levels of application of research findings by registered nurses in clinical practice was hospital size. On the other hand, Brown et al. (2010a) concluded that hospital size did not play a role when nurse-sensitive care outcomes were studied. Nurses provide a large proportion of inpatient care, and research has focused on increasing our understanding of how nurse staffing affects care outcomes. A systematic review (Lake and Cheung, 2006) found mixed results, with some studies showing a significant link between nurse staffing and pressure ulcers and others not showing such an association. A study performed in Belgian hospitals found no correlation between pressure ulcers and nurse staffing (Van den Heede et al., 2009). Furthermore, other studies have shown that having a greater number of patients assigned to a nurse was correlated with more perceived pressure ulcer events (Al-Kandari and Thomas, 2009) and that excessive workload and insufficient time contributed to a higher degree of pressure ulcers in long-term care (Pekkarinen et al., 2008).

A nursing-sensitive benchmarking registry, the Col- laborative Alliance for Nursing Outcomes (CALNOC), combines nurse staffing and workload with patient outcomes, e.g. pressure ulcers (Aydin et al., 2004). To learn more about the organization and pressure ulcer prevention, a benchmarking study comparing CALNOC hospitals (n = 207) and two Swedish hospitals (a univer- sity and a general hospital) was conducted (Gunningberg et al., 2011). The findings revealed that CALNOC hospitals had lower prevalence of hospital-acquired pressure ulcers (2.0%), higher staffing and higher patient turnover (workload) than the Swedish hospitals did. Some differences between the two Swedish hospitals were also identified.

The healthcare system is a complex organization, and professionals within the organization need to cooperate on different levels to prevent adverse events, such as pressure ulcers. Thus far, we do not know which factors in the organization contribute most to evidence-based pressure ulcer prevention. Previous studies have primarily investi- gated factors such as hospital size, nurse staffing and workload separately (Brown et al., 2010a; Lake and Cheung, 2006; Al-Kandari and Thomas, 2009). Simulta- neous testing of the associations between these variables could help to explain their relationships to each other. Therefore, we wished to gain a deeper understanding and set out to analyze the combined effect of different factors using data from the two Swedish hospitals (Gunningberg et al., 2011). The aim of the present study was to investigate the associations between variables on different levels in the health care setting (patient, unit, hospital) and documentation of (1) risk assessment and (2) skin assessment within 24 h of admission, the use of (3) pressure-reducing mattresses and (4) planned reposition- ing in bed.

2. M

2.1.

(pr we dat mo

2.2.

(11 cou had rela edu ass had had pre inte sur fast Sto lish gen hos pre pre hos dire hos hos exp 201

2.3.

two sur 825 stu (n = nee

2.4.

ass adm and

hos sur typ hou reg per

E. Sving et al. / International Journal of Nursing Studies 51 (2014) 717–725 719

ethod

Design

A cross-sectional study design was used. Clinical data essure ulcer prevention and pressure ulcer prevalence) re collected during one day, September 30, 2009, and a on nurse staffing and workload were collected for one nth, September 2009 (Gunningberg et al., 2011).

Settings

The study was conducted in one university hospital 00 beds) and one general hospital (350 beds) in two nty council districts in Sweden. The university hospital

developed a hospital-wide quality improvement plan ted to pressure ulcer prevention. For example, an cation programme, seminars for registered nurses and

istant nurses, and a network for pressure ulcer nurses been established. Multidisciplinary clinical guidelines been developed. Templates for risk assessment, ssure ulcer classification and standard care plans were grated into the electronic health record. Prevalence

veys had been conducted regularly and repeated with feedback of results to the units (Gunningberg and tts, 2008). Pressure ulcer prevalence was also estab- ed as a quality indicator at the hospital level. The eral hospital had not worked systematically at the pital level with quality improvement related to ssure ulcers and conducted its first pressure ulcer valence survey in 2009. On the other hand, the general pital had higher percentage of registered nurses in ct patient care (62.8%) compared to the university pital (52.8%) (Gunningberg et al., 2011). The general pital also had registered nurses with longer work erience (>5 years), 14.8% vs. 10.9% (Gunningberg et al., 0).

Participants

The study included participants from 44 units in the hospitals: geriatric (n = 8), medical (n = 24) and

gical (n = 19). All adult patients (>17 years), in total , admitted to those units before midnight the day of the

dy gave their consent to participate. Intensive care units 6) were excluded due to the different nursing care ds and nurse staffing levels.

Data collection

Dependent variables were documentation of (1) risk essment and (2) skin assessment within 24 h of

ission, the use of (3) pressure-reducing mattresses, (4) planned repositioning in bed.

Independent variables were age, gender, days of pitalization, risk score (Braden <17) on the day of vey, hospital type (university and general hospital), unit e (geriatric, medical and surgical), nurse staffing (total rs of care per patient day, percent of care provided by

istered nurses) and workload (percent patient turnover patient day). Patients’ pressure ulcer risk on the day of

survey was assessed using the Braden scale, which consists of six subscales (sensory perception, activity, nutrition, moisture, shear and friction) and has demonstrated good validity and reliability (Braden and Maklebust, 2005). Data collection sources, variables, and definitions are presented in Table 1.

2.5. Procedure

The procedure followed the methodology outlined by the European Pressure Ulcer Advisory Panel (Vanderwee et al., 2007) and the Collaborative Alliance for Nursing Outcomes (Brown et al., 2010b).

Prior to the study, the data collectors (registered nurses) participated in a half-day seminar; they were presented information on study procedures and data collection with a specific focus on clinical assessment of patients. The registered nurses also completed a 1-h e- learning session that included training on classification of pressure ulcers and differentiation of these from moisture lesions (Gunningberg et al., 2011). The feasibility of identifying variables from the electronic health record was pilot tested, and it was confirmed that accurate data could be obtained.

On the day of data collection, a team consisting of two registered nurses on the participating units (one staff nurse and one nurse from a different unit) visited each patient at the unit and gathered the clinical data at bedside: risk assessment (Braden score), skin assessment, type of mattresses and the use of planned repositioning (defined as having a turning schedule in place at the bedside). Afterwards, a retrospective audit of the electronic health records was conducted looking at documented risk and skin assessment within 24 h of admission.

Nurse staffing and workload variables were collected at the unit level and were extracted from the computerized patient and staff administration system for September 2009 (Table 1).

2.6. Data analysis

For descriptive purposes, frequency, mean, standard deviation (SD), median, interquartile range (IQR) and percentage data were used. Logistic regressions were performed using the Generalized Estimating Equation (GEE) approach to adjust for the clustering of patients within units (Agresti, 2002). To investigate the simulta- neous effect of the included independent variables, all independent variables were included simultaneously for each of the four dependent variables. The results of the logistic regressions are presented as odds ratios (OR), 95% confidence intervals (CI) and p-values of the Wald tests of the null hypothesis of odds ratios equal to one. Odds ratios should be interpreted as the change in the ratio of receiving preventive care to not receiving preventive care, compar- ing two patients with a one-unit change in an independent variable, keeping the other independent variables con- stant. Results were considered statistically significant when the p-value was less than 0.05. All tests were two- tailed. Data were analyzed using the statistical software package SPSS 19.0.

E. Sving et al. / International Journal of Nursing Studies 51 (2014) 717–725720

3. Ethical description

The study was approved by the Research Ethics Committee of the Faculty of Medicine at Uppsala University (No. 01-502). It followed the principles of the Declaration of Helsinki and the national and local ethical guidelines for research (CODEX). The patients received verbal and written information about the study and gave their verbal consent; if necessary relatives were consulted. The patients were informed that they were free to withdraw at any time. All data were treated confiden- tially.

4. Results

The total sample consisted of 825 patients with a mean age of 68.4 years and a median length of stay prior to the survey day of 6.0 days (IQR = 2.0–16.0). The gender distribution was almost equal. Eighteen percent of patients were assessed as being at risk for developing pressure ulcers (Braden score <17). The prevalence of hospital- acquired pressure ulcers Category 1–4 was 12.6% and 4.7% when Category 1 was excluded. The median value of total hours of care at the two hospitals was 8.0. The median value of percentage of total hours of care provided by registered nurses was 56.3. Percentage patient turnover per patient day had a median value of 40.9. Almost 75% of patients were admitted to the university hospital. Surgical and medical units had 40% of the patients each (Table 2).

Pressure ulcer prevention was performed for 44.1–

(Braden score <17). Planned repositioning was performed least often. Data on risk and skin assessment and pressure ulcer prevention are presented in Table 3.

4.1. Results from logistic regression

Table 4 presents the results of the associations between patients’ characteristics, hospital type, unit type, nurse staffing and workload and risk and skin assessment documented within 24 h of admission, the use of pres- sure-reducing mattresses and planned repositioning in bed.

Patients at risk for developing pressure ulcers (Braden <17) had higher odds of having skin assessment docu- mented, receiving pressure-reducing mattresses and planned repositioning. Patients with higher age were more likely to have risk and skin assessment documented. Hospital and unit type revealed significant results on all four dependent variables. Patients at the university hospital had higher odds of having risk and skin assess- ment documented and of having a pressure-reducing mattresses. On the other hand, planned repositioning was less likely to be used at the university hospital. Medical units revealed significant results but in different direc- tions. Documented risk and skin assessment and the use of planned repositioning were more likely to be found on medical units, but not the use of pressure-reducing mattresses. The only significant result for nurse staffing was total hours of care. When total hours of care was lower, patients had higher odds of having pressure-reducing mattresses but were less likely to have planned reposi-

Table 1

Data collection sources, variables, definitions of dependent and independent variables according to the European Pressure Ulcer Advisory Panel (Vanderwee

et al., 2007) and the Collaborative Alliance for Nursing Outcomes (CALNOC 2010).

Data collection source Variable Definition

Dependent variables

Electronic Health record Risk assessment Documented risk assessment performed by a validated

risk assessment scale or clinical judgement within 24 h of

admission to the hospital

Electronic Health record Skin assessment Documented skin assessment within 24 h of admission to the hospital

Clinical observations Pressure-reducing mattress Non-powered devices, e.g., visco-elastic foam or powered devices,

e.g., alternating pressure mattress

Clinical observations Planned repositioning in bed As turning schedule in place at the bedside with planned manual

repositioning every second, third and fourth hour

Independent variables

Patients records Patient characteristics Age Gender

Electronic Health record Patient characteristics Days of

hospitalization

Days of hospitalization prior to the study day

Clinical observations Risk assessment The Braden scale. Risk score <17 was defined as at risk for developing

pressure ulcers

Hospital type University hospital

General hospital Unite type Geriatric

units Medical units Surgical units

The type of hospital and unit the patient was admitted to

Staff administration system Nurse staffing Total hours of care

per patient days

The sum of productive hours for one month divided by the total number

of patient days on the unit. Included were all RNs’ and ANs’ hours

spent on direct patient care responsibilities

Nurse staffing RNs’ percent of the

total hours of care for one month

The registered nurses’ total hours of care divided by the total hours of

care multiplied by one hundred

Patient administration system Workload Patient turnover per

patient day

The total number of patients admitted, discharged and transferred

divided by patient days and multiplied by one hundred. Thus,

one hundred percent turnover indicates that for each patient day

there was one admission, discharge or transfer

tioning.

58.7% of patients at risk for developing a pressure ulcer

Table 3

Preventive care for all patients and for patients at risk of developing pressure ulcers.

Total

n = 825 n (%)

University hospital n = 610 n (%) General hospital n = 215 n (%) Geriatric

units n = 133 n (%)

Medical units n = 343 n (%) Surgical units n = 349 n (%)

Risk assessment n = 825 380 (46.1) 366 (60.0) 14 (6.5) 70 (52.6) 167 (48.7) 143 (41.0)

Skin assessment n = 825 388 (47.0) 365 (59.8) 23 (10.7) 62 (46.6) 188 (54.8) 138 (39.5)

Pressure-reducing mattresses n = 822 263 (32.0) 210 (34.6) 53 (24.7) 76 (57.1) 95 (27.9) 92 (26.4)

Planned repositioning n = 825 83 (12.8) 59 (13.6) 24 (11.3) 22 (16.5) 53 (15.5) 8 (2.3)

Risk patients (Braden score < 17) n = 143 n = 112 n = 31 n = 42 n = 56 n = 45 Risk assessment n = 143 77 (53.8) 70 (62.5) 7 (22.6) 29 (68.0) 34 (60.7) 14 (31.1)

Skin assessment n = 143 84 (58.7) 76 (67.9) 8 (25.8) 25 (59.5) 41 (73.2) 18 (40.0)

Pressure-reducing mattresses n = 142 67 (47.2) 48 (43.2) 19 (61.3) 26 (61.9) 23 (41.8) 18 (40.0)

Planned repositioning n = 143 63 (44.1) 47 (42.0) 16 (51.6) 15 (36.7) 41 (73.2) 7 (15.6)

Table 2

Patient characteristics, hospital-acquired pressure ulcer, nurse staffing and workload.

Total University hospital General hospital Geriatric units Medical units Surgical units

Patients – n (%) 825 610 (73.9) 215 (26.1) 133 (16.1) 343 (41.6) 349 (42.3)

Age – Mean (SD) 68.4 (16.8) 67.8 (17.3) 70.3 (15.2) 80.4 (8.2) 67.12 (17.306) 65.1 (16.8)

Gender male – n (%) 385 (46.7) 281 (46.1) 104 (48.4) 49 (38.6) 178 (52) 158 (45.3)

Days of hospitalization – Median (IQR)b 6.0 (2.0, 16.0) 7.0 (2.0, 19.0) 5.00 (2.0, 10.0) 22 (11.5, 22.0) 5 (2.0, 14.5) 4.0 (2.0, 8.0)

Braden score <17 – n (%) 143 (17.6) 112 (18.8) 31 (14.4) 42 (32.1) 56 (15.5) 45 (13.2)

HAPU category 1–4 – n (%) 104 (12.6) 87 (14.3) 17 (7.9) 28 (21.1) 36 (10.5) 40 (11.5)

HAPU category 2–4 – n (%) 39 (4.7) 31 (5.1) 8 (3.7) 19 (13.3) 14 (3.7) 8 (2.2)

Total hours of care per patient day – Median (IQR)b 8.0 (7.2, 9.0) 8.2 (7.3, 9.0) 7.9 (7.2, 8.4) 8.3 (7.9, 9.0) 8.5a (8.0, 9.9) 7.4 (6.6, 7.8)

RNs’ percent of total hours of care – Median (IQR)b 56.3 (47.3, 60.8) 55.2a (41.7, 57.6) 64.8 (57.4, 65.8) 40.6 (37.7, 41.0) 57.15 (54.8, 65.8) 56.7 (52.0, 60.8)

Percent patient turnover per patient day – Median (IQR)b 40.9 (31.7, 54.7) 40.25 (28.4, 54.7) 42.2 (35.5, 47.8) 13.3 (12.0, 17.3) 49.42 (31.9, 62.8) 42.2 (39.6, 54.7)

a One unit with 100% RNs’. bInter-quartile range (IQR) 25th–75th percentiles.

E .

S v

in g

e t

a l.

/ In

te rn

a tio

n a

l Jo

u rn

a l

o f

N u

rsin g

S tu

d ie

s 5

1 (2

0 1

4 )

7 1

7 –

7 2

5

7 2

1

Table 4

Prevention activities. Estimated odds ratio of pressure ulcer prevention. Logistic regression by Generalized Estimating Equations (GEE). OR = odds ratio, 95% CI = 95% confidence intervals.

Multivariate estimates

Risk assessment documented

n = 806b Skin assessment documented

n = 806b Pressure-reducing mattresses

n = 803b Planned repositioning in bed n = 806b

OR 95%CI p-Value OR 95%CI p-Value OR 95%CI p-Value OR 95%CI p-Value

Patient characteristics Patient’s age 1.015 1.004–1.026 0.007 1.020 1.009–1.031 <0.001 1.006 0.995–1.016 0.306 1.016 0.988–1.046 0.286

Gender Mana0) 1.000 1.000 1.000 1.000

Womana(1) 0.877 0.631–1.218 0.432 0.893 0.646–1.235 0.495 1.062 0.767–1.470 0.718 1.484 0.810–2.689 0.193

Day of hospitalization 0.994 0.987–1.001 0.093 0.995 0.989–1.002 0.173 1.022 1.011–1.033 <0.001 1.009 0.999–1.019 0.088

Braden �17 Not at riska(0) 1.000 1.000 1.000 1.000 Braden < 17 At riska(1) 1.461 0.906–2.255 0.120 1.916 1.216–3.019 0.005 1.751 1.153–2.658 0.009 26.824 13.540–51.140 <0.001

Type of hospital and unit University hospitala(0) 1.000 1.000 1.000 1.000

General hospitala(1) 0.038 0.021–0.070 <0.001 0.061 0.038–0.100 <0.001 0.673 0.453–0.999 0.050 3.226 1.564–6.647 0.002

Geriatrica(0) 1.000 0.006 1.000 < 0.001 1.000 0.014 1.000 <0.001

Medicala(1) 2.601 1.270–5.237 0.009 4.320 2.134–8.745 <0.001 0.432 0.231–0.809 0.009 4.112 1.629–10.380 0.003

Surgicala(2) 1.630 0.801–3.318 0.178 1.847 0.938–3.637 0.076 0.395 0.216–0.722 0.003 0.412 0.142–1.190 0.101

Nurse staffing Total hours of care per patient day 0.989 0.989–1.090 0.829 1.025 0.937–1.122 0.584 0.856 0.785–0.933 <0.001 1.195 1.046–1.365 0.009

RNs’ percent of care 1.006 0.986–1.026 0.570 0.997 0.978–1.017 0.782 1.007 0.989–1.024 0.463 0.961 0.923–1.000 0.839

Workload Percent patient turnover per patient day 0.981 0.971–0.992 0.001 0.993 0.984–1.002 0.135 0.998 0.991–1.006 0.634 0.999 0.980–1.009 0.839

a Coding in brackets. Reference category = 0. b Number of valid observations.

E .

S v

in g

e t

a l.

/ In

te rn

a tio

n a

l Jo

u rn

a l

o f

N u

rsin g

S tu

d ie

s 5

1 (2

0 1

4 )

7 1

7 –

7 2

5 7

2 2

5. D

hos reg me use tion res

sco tion 44– rec siti bee 201

ind stru me (NP pre diff per to p wit pat be pat com hos adm (Gu

Pat pre pla per in a One esta bet adm risk red EPU

dem at t stru of c phy mu nec kno sho

risk pro fou

E. Sving et al. / International Journal of Nursing Studies 51 (2014) 717–725 723

iscussion

The present study combined patient characteristics, pital type, unit type, nurse staffing and workload in ression models to identify associations between docu- nted risk and skin assessment within 24 h of admission,

of pressure-reducing mattresses and planned reposi- ing. The findings show that statistically significant

ults varied between the four dependent variables. Prevention care was related to patients’ high age and risk res (Braden <17), which is in accordance with interna- al guidelines (NPUAP & EPUAP, 2009). However, only 47% of patients at risk for developing pressure ulcers

eived pressure-reducing mattresses and planned repo- oning. Insufficient pressure ulcer prevention has also n shown in other studies in Europe (Lahmann et al., 0; Tannen et al., 2008; Vanderwee et al., 2011).

Prevention care should be based on each patient’s ividual needs and grounded on a risk assessment. A ctured risk assessment includes use of a risk assess-

nt scale, skin assessment and clinical judgement UAP & EPUAP, 2009). In the present study, we analyzed ssure ulcer prevention in two hospitals operating under erent conditions. The university hospital had been forming continuous quality improvement work related ressure ulcer prevention. The general hospital had staff h more work experience. Our findings show that ients at the university hospital were more likely to risk and skin assessed, although only 60% of these ients had documented risk and skin assessment. In parison, almost 100% of patients in the CALNOC

pitals were risk and skin assessed within 24 h of ission, and care plans were in place for patients at risk

nningberg et al., 2011). The present results on unit type were interesting. ients admitted to medical units were less likely to have ssure-reducing mattresses but more likely to have nned repositioning. Differences between units in the formance of pressure ulcer prevention were also found

study performed in Belgium (Vanderwee et al., 2011). explanation could be differences in culture and blished routines for prevention of pressure ulcers

ween unit types. Patients at risk for pressure ulcers are itted to all kinds of hospitals and units, and patients at

for developing pressure ulcers need both pressure- ucing mattresses and planned repositioning (NPUAP & AP, 2009).

The present results on hospital type and unit type onstrate the importance of carrying out quality work

he hospital level. Having a hospital-wide routine for ctured risk and skin assessment and for establishment are plans is important. The staff (physicians, nurses, sical therapists and occupational therapists) in all units st have the knowledge and skills, as well as routines, essary for preventing pressure ulcers. They need to w what, when and how pressure ulcer prevention uld be performed. The variable total hours of care was not significant for

and skin assessment. Furthermore, percentage of care vided by registered nurses was not related to any of the r dependent variables. These findings were surprising,

as Swedish registered nurses often argue that low staffing levels can explain insufficiencies in nursing care. There is consensus across the globe that registered nurses should be responsible for pressure ulcer prevention. However, Swedish studies show that registered nurses trusted and delegated nursing tasks to assistant nurses (Sving et al., 2012) and did not prioritize pressure ulcer prevention (Athlin et al., 2010, Sving et al., 2012). This could be an expression of negative attitudes towards pressure ulcers (Beeckman et al., 2011) or insufficient awareness of the importance of evidence-based prevention in ensuring patient safety.

In the present study, we have investigated factors at three different levels (patient, unit and hospital) of the healthcare organization simultaneously, as it is important to understand how the various parts of this complex system are linked together. These levels must be organized in ways that support patients’ journey through the system (Nelson et al., 2008). If such a salutary organization can be achieved, evidence-based pressure ulcer prevention will be possible.

6. Strengths and limitations

The present study employed a cross-sectional design, which is appropriate for describing the status of phenom- ena and for describing relationships among phenomena (Polit and Beck, 2008). A cross-sectional design provides only a ‘‘snapshot’’ of the prevalence of a phenomenon at a particular point in time. In our study, the month of September was selected purposely, because staffing and hospital activities are normal and not affected by, for example, vacations.

The pressure ulcer prevalence methodology employed in the present study is used internationally (Vanderwee et al., 2007). The clinical observation data were based on examination of patients by two registered nurses, who were trained specifically for this task, which increased the reliability of the observations. The data on planned repositioning were based on having a turning schedule at the bedside. This could entail an overestimation of how often repositioning of patients was actually executed.

Nurse staffing and workload were defined according to the Collaborative Alliance for Nursing Outcomes (CALNOC, 2007). The criteria were developed in another health care organization, and therefore the Swedish researchers worked in collaboration with CALNOC leaders to adhere to the methodology as closely as possible. The computer- ized administrative and patient system is considered a reliable and valid source for data on nurse staffing and patient turnover.

Patients were observed within units. For a few variables the observed data for patients within a unit were the same and, for most of the variables, the observations were related. The GEE approach models this correlation in a relatively simple way. Explicit power calculations were not made. The study design was not confirmatory, and no explicit hypotheses with pre-specified effect sizes were stated. The p-values are in our study were supplemented by confidence intervals, which indicate the precision of the estimates. Adequate sample size was still considered important. Some guidelines suggest at least 10 outcomes

E. Sving et al. / International Journal of Nursing Studies 51 (2014) 717–725724

for each of the binary responses for every independent variable (Agresti, 2002). Small sample size, unbalanced data, and many categorical variables may cause unstable estimates and numerical problems in estimation. Large standard errors constitute one important sign of such problems. We found no clear signs of this in our study. Furthermore, we used variables previously described as having an impact on patient outcome and pressure ulcer prevention. The study design did not permit us to infer causality, but only to see associations between variables. Because pressure ulcer prevention in hospital is performed in a complex organization, it is possible that additional factors facilitated or hindered pressure ulcer prevention. For example, clinical patient data in addition to the Braden score, gender, age and days of hospitalization could be considered in future studies.

One of the limitations is that data collection was performed at only two hospitals, and therefore the results cannot be generalized. However the hospitals are from different parts of Sweden, which means that they have separate hospital organizations. Inclusion of a greater number of patients may have resulted in a greater number of significant variables.

7. Conclusion

Patients at risk of developing pressure ulcers (Braden <17) had higher odds of having risk assessment docu- mented within 24 h of admission, and of receiving pressure-reducing mattresses and planned repositioning. Hospital type was also associated with the four dependent variables. Surprisingly, nurse staffing played only a minor role. Leaders in healthcare organizations should establish routines on different levels that support evidence-based pressure ulcer prevention, and registered nurses need to assume responsibility for bedside care.

Acknowledgements

We wish to thank Nancy Donaldson, RN, DNSC., FAAN CALNOC, senior scientist, for her inspiring collaboration in the research project. Lena Hagman was responsible for the data collection at the general hospital. Grants from Uppsala-Örebro Regional Research Council are gratefully acknowledged, as is the Centre for Research & Develop- ment, Uppsala University/County Council of Gävleborg.- Conflict of interest: None.Funding: Uppsala-Örebro Regional Research Council are gratefully acknowledged and Center for Research & Development, Uppsala university/County Council of Gävleborg. The founders did not participate in the conduct of the research.Ethical approval: The Research Ethics Committee of the Faculty of Medicine at Uppsala University (No. 01-502).

References

Agresti, A., 2002. Categorical Data Analysis, 2nd ed. Wiley, New York. Al-Kandari, F., Thomas, D., 2009. Perceived adverse patient outcomes

correlated to nurses’ workload in medical and surgical wards of

Athlin, E., Idvall, E., Jernfalt, M., Johansson, I., 2010. Factors of importance to the development of pressure ulcers in the care trajectory: percep- tions of hospital and community care nurses. Journal of Clinical Nursing 19 (15–16) 2252–2258.

Aydin, C.E., Bolton, L.B., Donaldson, N., Brown, D.S., Buffum, M., Elashoff, J.D., Sandhu, M., 2004. Creating and analyzing a statewide nursing quality measurement database. Journal of Nursing Scholarship 36 (4) 371–378.

Barrois, B., Labalette, C., Rousseau, P., Corbin, A., Colin, D., Allaert, F., Saumet, J.L., 2008. A national prevalence study of pressure ulcers in French hospital inpatients. Journal of Wound Care 17 (9) 373–376.

Beeckman, D., Defloor, T., Schoonhoven, L., Vanderwee, K., 2011. Knowl- edge and attitudes of nurses on pressure ulcer prevention: a cross- sectional multicenter study in Belgian hospitals. Worldviews on Evidence-Based Nursing 8 (3) 166–176.

Bennett, G., Dealey, C., Posnett, J., 2004. The cost of pressure ulcers in the UK. Age Ageing 33 (3) 230–235.

Braden, B.J., Maklebust, J., 2005. Preventing pressure ulcers with the Braden scale: an update on this easy-to-use tool that assesses a patient’s risk. American Journal of Nursing 105 (6) 70–72.

Brown, D.S., Aydin, C.E., Donaldson, N., Fridman, M., Sandhu, M., 2010a. Benchmarking for small hospitals: size didn’t matter! Journal for Healthcare Quality 32 (4) 50–60.

Brown, D.S., Donaldson, N., Burnes Bolton, L., Aydin, C.E., 2010. Nursing- sensitive benchmarks for hospitals to gauge high-reliability perfor- mance. Journal for Healthcare Quality 32 (6) 9–17.

CALNOC, 2007. CALNOC Code-Book Part I & II, Coordinating and Using CALNOC Data in the Hospital Setting. UCSF School of Nursing, Center for Nursing Research & Innovation, Collaborative Alliance for Nursing Outcomes, San Francisco, CA.

CODEX. Rules and guidelines for research. http://www.codex.vr.se/ (accessed 19.04.2013).

Cuddigan, J., Berlowitz, D., Ayello, E., 2001. Pressure ulcers in America: prevalence, incidence, and implications for the future. An executive summary of the National Pressure Ulcer Advisory Panel monograph. Advances in Skin and Wound Care 14 (4) 208–215.

Defloor, T., De Bacquer, D., Grypdonck, M.H., 2005. The effect of various combinations of turning and pressure reducing devices on the inci- dence of pressure ulcers. International Journal of Nursing Studies 42 (1) 37–46.

Estabrooks, C.A., Midodzi, W.K., Cummings, G.G., Wallin, L., 2007. Pre- dicting research use in nursing organizations: a multilevel analysis. Nursing Research 56 (Suppl. 4) 7–23.

Gorecki, C., Brown, J.M., Nelson, E.A., Briggs, M., Schoonhoven, L., Dealey, C., Defloor, T., Nixon, J., 2009. Impact of pressure ulcers on quality of life in older patients: a systematic review. Journal of the American Geriatrics Society 57 (7) 1175–1183.

Gunningberg, L., Brudin, L., Idvall, E., 2010. Nurse Managers’ prerequisite for nursing development: a survey on pressure ulcers and contextual factors in hospital organizations. Journal of Nursing Management 18 (6) 757–766.

Gunningberg, L., Donaldson, N., Aydin, C., Idvall, E., 2011. Exploring variation in pressure ulcer prevalence in Sweden and the USA: benchmarking in action. Journal of Evaluation in Clinical Practice 18 (4) 904–910.

Gunningberg, L., Stotts, N.A., 2008. Tracking quality over time: what do pressure ulcer data show? International Journal for Quality in Health Care 20 (4) 246–253.

Gunningberg, L., Stotts, N.A., Idvall, E., 2011. Hospital-acquired pressure ulcers in two Swedish County Councils: cross-sectional data as the foundation for future quality improvement. International Wound Journal 8 (5) 465–473.

Lahmann, N.A., Halfens, R.J., Dassen, T., 2010. Impact of prevention structures and processes on pressure ulcer prevalence in nursing homes and acute-care hospitals. Journal of Evaluation in Clinical Practice 16 (1) 50–56.

Lake, E.T., Cheung, R.B., 2006. Are patient falls and pressure ulcers sensi- tive to nurse staffing? Western Journal of Nursing Research 28 (6) 654–677.

McInnes, E., Dumville, J.C., Jammali-Blasi, A., Bell-Syer, S.E., 2011. Support surfaces for treating pressure ulcers. Cochrane Database of Systematic Reviews 12, CD009490.

Moore, Z., Cowman, S., 2012. Pressure ulcer prevalence and prevention practices in care of the older person in the Republic of Ireland. Journal of Clinical Nursing 21 (3–4) 362–371.

Nelson, E.C.G., Batalden, M.M., Berry, P.B., Bothe, S.A., McKinley Jr., A.E., Melin, K.E., Muething, C.N., Moore, S.E., Wasson, L.G., Nolan, J.H.T.W., 2008. Clinical microsystems. Part 1. The building blocks of health

systems. Joint Commission. Journal of Quality and Patient Safety 34 (7) 367–378.

selected hospitals in Kuwait. Journal of Clinical Nursing 18 (4) 581–590.

NPU

Pekk

Poli

Svin

E. Sving et al. / International Journal of Nursing Studies 51 (2014) 717–725 725

AP & EPUAP, 2009. International Guideline. Pressure Ulcer Prevention & Treatment. Clinical Practice Guideline. National Pressure Ulcer Advisory Panel & European Pressure Ulcer Advisory Panel, Washing- ton, DC. Available through the National Pressure Ulcer Advisory Panel (www.npuap.org). arinen, L., Sinervo, T., Elovainio, M., Noro, A., Finne-Soveri, H., 2008.

Drug use and pressure ulcers in long-term care units: do nurse time pressure and unfair management increase the prevalence? Journal of Clinical Nursing 17 (22) 3067–3073. t, D., Beck, C.T., 2008. Nursing Research: Generating and Assessing Evidence for Nursing Practice. Lippincott Williams & Wilkins, Phila- delphia. g, E., Gunningberg, L., Hogman, M., Mamhidir, A.G., 2012. Registered nurses’ attention to and perceptions of pressure ulcer prevention in hospital settings. Journal of Clinical Nursing 21 (9–10) 1293–1303.

Tannen, A., Dassen, T., Halfens, R., 2008. Differences in prevalence of pressure ulcers between the Netherlands and Germany – associations between risk, prevention and occurrence of pressure ulcers in hospi- tals and nursing homes. Journal of Clinical Nursing 17 (9) 1237–1244.

Van den Heede, K., Sermeus, W., Diya, L., Clarke, S.P., Lesaffre, E., Vleugels, A., Aiken, L.H., 2009. Nurse staffing and patient outcomes in Belgian acute hospitals: cross-sectional analysis of administrative data. In- ternational Journal of Nursing Studies 46 (7) 928–939.

Vanderwee, K., Clark, M., Dealey, C., Gunningberg, L., Defloor, T., 2007. Pressure ulcer prevalence in Europe: a pilot study. Journal of Evalua- tion in Clinical Practice 13 (2) 227–235.

Vanderwee, K., Defloor, T., Beeckman, D., Demarre, L., Verhaeghe, S., Van Durme, T., Gobert, M., 2011. Assessing the adequacy of pressure ulcer prevention in hospitals: a nationwide prevalence survey. BMJ Quality and Safety 20 (3) 260–267.