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

IN (IN ra

W Eli Bri Sh a NH b Cen c Me d Alf e Sch f Sch g Ro h Nu i Uni j Uni k Fac l Gri m Un

International Journal of Nursing Studies 52 (2015) 1659–1668

A R

Artic

Rece

Rece

Acce

Keyw

Care

Pati

Pati

Pres

Pres

*

http

002

Troducing A Care bundle To prevent pressure injury TACT) in at-risk patients: A protocol for a cluster

ndomised trial

endy Chaboyer a,b,c, Tracey Bucknall d,f, Joan Webster a,g, zabeth McInnes e,h, Merrilyn Banks g, Marianne Wallis b,i, gid M. Gillespie a,b,c, Jennifer A. Whitty a,c,j, Lukman Thalib k,l, elley Roberts a,b,c,*, Nicky Cullum m

MRC Centre of Research Excellence in Nursing, Griffith University, Australia

tre for Health Practice Innovation, Griffith University, Australia

nzies Health Institute Queensland, Griffith University, Australia

red Health, Australia

ool of Nursing, Midwifery and Paramedicine, Australian Catholic University, Australia

ool of Nursing and Midwifery, Deakin University, Australia

yal Brisbane and Women’s Hospital, Australia

rsing Research Institute, St Vincent’s Health Australia (Sydney), Australia

versity of the Sunshine Coast, Australia

versity of Queensland, Australia

ulty of Medicine, University of Kuwait, Kuwait

ffith University, Australia

iversity of Manchester, United Kingdom

T I C L E I N F O

le history:

ived 19 September 2014

ived in revised form 16 April 2015

pted 28 April 2015

ords:

bundle

ent centred care

ent participation

sure injury prevention

sure ulcer prevention

A B S T R A C T

Background: Pressure injuries are a significant clinical and economic issue, affecting both

patients and the health care system. Many pressure injuries in hospitals are facility

acquired, and are largely preventable. Despite growing evidence and directives for

pressure injury prevention, implementation of preventative strategies is suboptimal, and

pressure injuries remain a serious problem in hospitals.

Objectives: This study will test the effectiveness and cost-effectiveness of a patient-

centred pressure injury prevention care bundle on the development of hospital acquired

pressure injury in at-risk patients.

Design: This is a multi-site, parallel group cluster randomised trial. The hospital is the unit

of randomisation.

Methods: Adult medical and surgical patients admitted to the study wards of eight

hospitals who are (a) deemed to be at risk of pressure injury (i.e. have reduced mobility),

(b) expected to stay in hospital for �48 h, (c) admitted to hospital in the past 36 h; and (d) able to provide informed consent will be eligible to participate. Consenting patients will

receive either the pressure injury prevention care bundle or standard care. The care bundle

contains three main messages: (1) keep moving; (2) look after your skin; and (3) eat a

healthy diet. Nurses will receive education about the intervention. Patients will exit the

study upon development of a pressure injury, hospital discharge or 28 days, whichever

Corresponding author at: NHMRC Centre of Research Excellence in Nursing, Griffith University, Australia. Tel.: +61 755529557.

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

Contents lists available at ScienceDirect

International Journal of Nursing Studies

journal homepage: www.elsevier.com/ijns

://dx.doi.org/10.1016/j.ijnurstu.2015.04.018

0-7489/� 2015 Elsevier Ltd. All rights reserved.

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–16681660

What is already known about the topic?

� Pressure injury is a significant clinical and economic issue. � Most hospital acquired pressure injuries are preventable. � Strategies to prevent pressure injuries must be complex

and multidimensional.

What this paper adds

� This paper proposes a protocol for a complex interven- tion for pressure injury prevention, i.e. a pressure injury prevention care bundle. � The proposed care bundle includes three main messages:

(1) keep moving; (2) look after your skin; and (3) eat a healthy diet. � The care bundle incorporates patient participation in

care and encourages nurse and patient partnership.

1. Background

A pressure injury (PI), also known as pressure ulcer, is defined as ‘‘localised injury to the skin and/or underlying tissue usually over a bony prominence, as a result of pressure, or pressure in combination with shear’’ (European Pressure Ulcer Advisory Panel et al., 2014). PIs are a major problem in hospitals, affecting approximately 10–30% of patients worldwide, depending on the country, setting, patient population and presence of pressure injury preven- tion (PIP) strategies (Banks et al., 2010; Gunningberg et al., 2013; Igarashi et al., 2013; James et al., 2010). The majority of PIs in the clinical setting are hospital acquired pressure injury (HAPI) (Gallagher et al., 2008; Gunningberg et al., 2011; Lyder et al., 2012). Two European studies of nearly 2000 hospitalised patients reported PI prevalence between 15 and 19%, and 77–78% of these were HAPI (Gallagher et al., 2008; Gunningberg et al., 2011). In Australia, the incidence of HAPI is 7.4–17.4% (Mulligan et al., 2011).

PIs result in significant physical, social and physiologi- cal problems for patients, including pain, wound exudate and odour, decreased mobility and independence, poor body image and emotional issues (Gorecki et al., 2009,

2011; Latimer et al., 2014). They also place a large burden on the health care system, by increasing length of stay (Graves et al., 2005) and hospital costs (Graves and Zheng, 2014). In fact, a recent estimate of the total cost of PI in public and private hospitals in Australia in 2010–11 was US$1.64 billion (�US$1.05 billion) (Graves and Zheng, 2014). PI treatment costs are estimated to represent 1.9% of all public hospital expenditure in Australia (Nguyen et al., 2015). Considering the significant negative outcomes for both patients and hospitals, PIP is of high importance.

It is recognised that the majority of PI, particularly HAPI, are preventable (Australian Commission on Safety and Quality in Health Care, 2011; Ayello and Lyder, 2008; VanGilder et al., 2009). In 2008, the US Centres for Medicare and Medicaid Services ceased reimbursing facilities for HAPIs, considering them ‘‘never events’’, that is, preventable events that should never occur (Centers for Medicare and Medicaid Services, 2007). Similarly, in the UK, facilities now incur financial penalties for failing to meet PIP goals, or receive financial incentives if goals are achieved (Department of Health, 2012). In Australia, Queensland public hospitals are fined AU$30,000 and AU$50,000 for stage III and IV HAPI, respectively (Queensland Government and Department of Health, 2014). Preventing PI is one of the Australian National Safety and Quality Health Service Standards, which are used to assess performance and accreditation of Australian health care facilities (Australian Commission on Safety and Quality in Health Care, 2011). The importance of PIP is also reflected in international PI guidelines (European Pressure Ulcer Advisory Panel et al., 2014). Due to the number of complex and inter-related risk factors for PI, preventative interventions must be complex and multifaceted (Coleman et al., 2013). Core PIP strategies include: encouraging mobility and appropriate support surfaces; good skin care; and nutritional assessment and intervention if required (European Pressure Ulcer Advisory Panel et al., 2014).

The evidence for improvements in HAPI rates with multifaceted PIP interventions is limited to quality improvement projects with a lack of large-scale RCTs. However, systematic reviews and theory suggest that a combination of quality improvement intervention strategies

comes first; transfer to another hospital or transfer to critical care and mechanically

ventilated. The primary outcome is incidence of hospital acquired pressure injury.

Secondary outcomes are pressure injury stage, patient participation in care and health care

costs. A health economic sub-study and a process evaluation will be undertaken alongside

the trial. Data will be analysed at the cluster (hospital) and patient level. Estimates of

hospital acquired pressure injury incidence in each group, group differences and 95%

confidence interval and p values will be reported.

Discussion: To our knowledge, this is the first trial of an intervention to incorporate a

number of pressure injury prevention strategies into a care bundle focusing on patient

participation and nurse–patient partnership. The results of this study will provide

important information on the effectiveness and cost-effectiveness of this intervention in

preventing pressure injuries in at-risk patients. If the results confirm the utility of the

developed care bundle, it could have a significant impact on clinical practice worldwide.

Trial registration: This trial is registered with the Australian New Zealand Clinical Trials

Registry, ACTRN12613001343796.

� 2015 Elsevier Ltd. All rights reserved.

ma et a Sev staf ma pro Sob the rev imp PIP of c pro

bas pro and (In et car nut alth tow ana A c a p adv and Gaa imp inv sim gui (va foc evi fam dri et a

sho safe out Nat Com US Org (20 Imp 201 dar the (Au Car pro tha hea (Da exa pat (Co

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–1668 1661

y reduce the incidence of PI in the hospital setting (Padula l., 2014; Soban et al., 2011; Sullivan and Schoelles, 2013). eral core elements are consistent across studies, including f education and information sharing, leadership, perfor- nce monitoring, and PI-specific interventions such as tocols, guidelines and risk assessment (Padula et al., 2014; an et al., 2011; Sullivan and Schoelles, 2013), emphasising

value of multifaceted interventions. One systematic iew highlighted a need for further understanding around lementation, causal pathways and the role of context in

interventions (Soban et al., 2011), hence the importance onducting definitive trials with adequate piloting and cess evaluation. A care bundle is a structured group of interventions ed on clinical practice guidelines that improve cesses of care, encourage compliance to guidelines,

have been shown to improve patient outcomes stitute for Healthcare Improvement, 2014; Rello al., 2011). One US group developed an 8-item PIP e bundle that included skin care, turning and ritional assessment, directed at nursing staff, and ough their annual PI prevalence data showed trends ards improvements in PI prevalence, no formal lysis was undertaken (Baldelli and Paciella, 2008).

luster randomised trial in the Netherlands found that atient safety care bundle was effective in reducing erse events, specifically PIs, urinary tract infections

falls among hospital and nursing home patients (van l et al., 2011). The study employed a multifaceted lementation strategy using staff education, patient

olvement and feedback on processes and outcomes to ultaneously implement multiple best practice delines (i.e. for each adverse event) in study wards n Gaal et al., 2011). To date, care bundles have mostly used on guiding clinicians in their practice, yet dence suggests involvement of patients and their ilies working alongside clinicians could be a major

ver in the use of care bundles (Coulter, 2006; van Gaal l., 2011).

Patient participation in their health care has been wn to reduce adverse events and improve patient ty (Weingart et al., 2011) and result in improved health

comes (Arnetz et al., 2010; Dwamena et al., 2012). ional and international groups such as the Australian mission on Safety and Quality in Health Care (2011), Joint Commission on Accreditation in Healthcare

anizations (2006) and the World Health Organisation 07) advocate for consumer participation in health care. ortantly, ‘‘Partnering with consumers’’ is one of the 1 Australian Safety and Quality Health Service Stan- ds, and is expected to be applied in conjunction with

standard ‘‘Preventing and Managing Pressure Injuries’’ stralian Commission on Safety and Quality in Health e, 2011). Recent research on patient and health fessional views about patient participation suggests t many interventions work optimally when both the lth professional and patient have a role in the initiative vis et al., 2012). However, rigorous studies that have mined the impact of patient–clinician partnerships on ient outcomes in the acute care setting are lacking ulter, 2006).

In summary, PIs contribute to significant burden for patients and the health care system. As many HAPI are considered preventable, strategies for PIP warrant special consideration, yet research suggests they are poorly implemented. A patient-centred PIP care bundle may be an effective approach to PIP, and is consistent with evidence on the benefits of care bundles and mandates for patient participation in care. Patients, who have a vested interest in PIP, may be an untapped resource to prompt timely use of PIP strategies.

Consequently, an innovative PIP strategy was devel- oped (Chaboyer and Gillespie, 2014; Gillespie et al., 2014) that is intended to optimise efficacy and sustainability, namely a care bundle that incorporates patient participa- tion in care, easy access to PIP information and nursing staff engagement. This paper reports the protocol currently being used to test this care bundle, namely the INTACT trial (INTroducing A Care bundle To prevent pressure injury in at-risk patients).

1.1. Aims and hypotheses

This study aims to provide rigorous evidence regarding the effect of a PIP care bundle on the development of HAPI in patients at risk of PI.

Primary hypothesis: The incidence of HAPI in ‘‘at risk’’ hospitalised patients who receive a PIP care bundle will be lower than that in at risk hospitalised patients receiving standard care.

Secondary hypotheses: The intervention group will have better outcomes than the standard care group in terms of: (a) PI stage, i.e. depth of tissue damage; (b) patient participation in care; and (c) healthcare costs.

All hypotheses apply to both the cluster level and the patient level (see data analysis).

2. Methods/design

2.1. Study design

This study is a multi-site, parallel group pragmatic cluster randomised trial (c-RT). The unit of randomisation will be hospitals to prevent contamination between groups. Wards cannot be randomised as there is substan- tial patient movement between wards. Patients will receive either the intervention or standard care. The primary outcome is incidence of HAPI, and secondary outcomes are PI stage, patient participation in care and health care costs. Patients will exit the study upon: reaching study day 28; development of a PI; hospital discharge; transfer to another hospital; or transfer to critical care and requiring mechanical ventilation. An economic sub-study will evaluate the cost effectiveness of the intervention compared to routine care. A project manager (based at a university) will coordinate overall management of the project. The project manager’s role will entail assisting with the development of the database, coordinating ethics applications, overseeing and/or deliv- ering research assistant training, site monitoring both during start up and during the trial, assessment of data quality. At each site, a study investigator who is a senior

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–16681662

researcher and four separate groups of research assistants will be involved: (1) to recruit patients; (2) to deliver the intervention; (3) to collect data for the economic evalua- tion; and (4) to collect daily data to assess the outcomes (incidence of HAPI). The trial has been approved by the Queensland Health Human Research Ethics Committee (reference number HREC/13/QGC/192), and by five hospi- tals and one university. The trial is registered with the Australian New Zealand Clinical Trials Registry (ACTRN12613001343796) and is funded by the National Health and Medical Research Council (APP1058963).

2.2. Setting

To be eligible for the study, hospitals must (a) be metropolitan referral hospitals that cater for diverse patient adult populations and case-mix groups; (b) offer acute medical and surgical and rehabilitative services; and (c) have 200 or more beds. All patients from wards except day-surgery, critical care, emergency, maternity, paediat- rics, mental health and dialysis units will be eligible to participate. There will be 5–15 wards able to participate at each site, which ensures diverse geographical, unit size and sub specialty representation. Methodological and prag- matic issues were considered in determining the number of sites. A total of eight acute care hospitals in Queensland, New South Wales and Victoria, Australia, have agreed to participate.

2.2.1. Randomisation

Generation of the allocation sequence and random 1:1 block allocation of hospitals to intervention or control group will be undertaken using a central randomisation service not involved in the study in any way. Hospitals will be stratified by the current HAPI rates as reported in each hospital. Allocation based on clusters will be concealed until the intervention is assigned. A statistician, not involved in recruitment and blinded to groups (treat- ment/control) will generate a series of random number lists for 4, 6, 8, 10 and 12 wards. The lists will then be sent to recruiters, who will determine the number of wards they will recruit from and the order in which wards will be approached. Hospitals will be informed of which arm of the study they are randomised to on completion of data collection. The chief investigators and interventionists will not be blinded to group allocation but the recruiters, outcome assessors and analysts will be blinded. These latter research assistants will be informed about group allocation at the end of the data collection.

2.3. Participants

A consecutive sample of patients in the study wards who meet all inclusion and no exclusion criteria will be recruited into the sample.

Inclusion criteria: (i) adults admitted to a study ward; (ii) expected hospital length of stay of �48 h; (iii) at risk of PI as measured by limited mobility (i.e. requiring physical or mechanical assistance to reposition or ambulate); (iv) able to read English; and (v) able to provide informed consent.

Exclusion criteria: (i) previous participation in this trial; (ii) admitted to the hospital for >36 h prior to recruitment; and (iii) palliative or dying patients.

2.3.1. Recruitment

We plan to commence recruiting at one site, to test procedures and other processes before recruitment com- mences at other sites. For each site, a computer-generated ward randomisation schedule will be developed and used to determine the order of approaching wards for recruit- ment of patients. This is to ensure that all wards in a particular institution are fairly represented. The number of wards approached each week will depend on recruitment rates at that site. The recruiters, from Monday to Friday, will screen eligibility of all patients admitted to study wards using a screening tool reflecting the inclusion and exclusion criteria. Patients will only be screened once. The nurse unit manager will ask potential participants if they are willing to be approached about this study, and if patients agree, they will be consecutively approached by recruiters according to the ward randomisation schedule. Verbal and written information about the study will be provided before seeking consent from the patient. Recruitment will continue until 200 patients have been recruited at each site. The trial period (i.e. time from first recruitment to final completion) is expected to be nine months. Fig. 1 shows anticipated participant flow through the study.

2.3.2. Sample size

The incidence of HAPI in Australian hospitals ranges from 7.4% to 17.4% (Mulligan et al., 2011). A Cochrane review on support surfaces reporting a meta-analysis of five studies indicated a reduction in PI in at risk patients (RR 0.40, 95% CI 0.21–0.74) (McInnes et al., 2012). In our sample size estimate we have taken a conservative approach of incidence of 10% HAPI (control) with an expected reduction of 50% or an absolute reduction of 5% (from 10% to 5%) in the intervention group and an intra- class correlation of 0.001, the actual intra-class correlation in another c-RT of a PIP strategy (Moore et al., 2011). To obtain 90% power with a two-sided a level of 0.05, 8 hospitals with 169 patients per hospital are required (PASS–Power Analysis and Sample Size system, NCSS, UT). Thus, the total sample required will be 1352. Data from our pilot study indicated an attrition rate of just over 10%, hence additional patients will be recruited for a total sample of 1600 (n = 200 per site).

2.4. Intervention

The intervention is a PIP care bundle aimed at both the individual (patient) and the cluster (hospital). It includes three main messages to promote patient participation in PIP care: (1) keep moving; (2) look after your skin; and (3) eat a healthy diet, which will be delivered to patients through a brochure, poster and DVD. Nurses will also receive training on encouraging patients to engage in PIP self-care activities. Table 1 contains details of the intervention components and materials.

The PIP care bundle is based on the Institute of Healthcare Improvement recommendations that care

bun enc pro ded nur pat me pat info trai pat wil rec nur tion

Tab

Con

M

1.

2.

3.

Note

sum

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–1668 1663

dles should be evidence based and straightforward, ompassing 3–5 items (Institute for Healthcare Im- vement, 2014). The intervention will be delivered by icated health professionals (predominantly registered ses) with a specific interest in wound care to both ients and nurses (patient and cluster level). Imple- nting the intervention will involve: (a) one-to-one ient training including watching a 5 min DVD; (b) an rmation brochure on PI; (c) poster reminders; and (d) ning of nursing staff to promote partnering with ients in PIP care and the care bundle. Interventionists l deliver the PIP care bundle to patients shortly after ruitment and will provide both patient education and se training throughout the study. Thus the interven-

is patient-centred and reflects partnership between

patients and nurses in the delivery of care. The interven- tion was previously developed and piloted in an acute tertiary hospital setting (citations masked for blinded peer review).

Awareness of the INTACT trial will be raised through a range of strategies such as hospital newsletters, e-mails and forums. A trial-specific education programme target- ing all nurses in study wards will be implemented to ensure clinical staff understand and support the study. A run-in period of two months will include formal group inservices and informal discussions with staff to ensure those at intervention sites understand the intervention and engage in partnering with patients in PIP strategies, and those at control sites are aware of the study. This training will commence prior to data collection and will continue as

Assess si te for eligibility / contact

Cluster randomisation

Exclude • Exclusion criteria • Refuse to partici pate

Contro l g roup n=4 sites, n=800 patie nts

Intervention grou p n=4 sites, n=800 patie nts

Lost to follow up n = 80 (10%)

Lost to follow up n = 80 (10%)

E nr

ol m

en t

A ll

oc at

io n

A na

ly si

s F

ol lo

w -u

p

Total analysed n = 720 (90 %)

Total analysed n = 720 (90 %)

Fig. 1. Anticipated participant flow through study.

le 1

tent of patient education.

essage Summary of content

Keep moving � Change position whilst in bed or in a chair � Use pillows for support or ask staff for help when changing position � Keep active by going for walks if possible

Look after your skin � Advise staff of pain, tenderness, redness or blistering over bony areas � Keep skin, clothes and bedding clean and dry � Use moisturising lotion and mild cleanser or moisturising soap to prevent drying out of skin � Special equipment (i.e. air mattresses, pressure relieving cushions, booties) may be used to reduce pressure

Eat a healthy diet � Good nutrition is important for skin protection and wound healing � Ensure good protein sources (examples given) for skin maintenance � Drink plenty of fluids for hydration � Consult with a dietitian or nutritionist � Take nutrition supplements as prescribed

: The intervention materials include a detailed brochure, which will be explained to patients by interventionists; a 5 min DVD; and a poster,

marising the three key messages.

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–16681664

deemed necessary throughout recruitment depending on ward staff turnover. Ideally, all nurses on study wards will receive the training at least once. The study investigator at each site will be responsible for organising training at that site.

2.5. Standard care

For the duration of the trial, patients in the control group hospitals will receive the standard care provided for PIP on the particular ward. It is likely that standard care may vary from hospital to hospital and even between wards; consequently, we will document ‘standard care’ at each recruiting site. Patients in intervention group hospitals will also receive standard PIP care along with the intervention.

2.6. Blinding

Selection bias is a potential issue when undertaking cluster trials, especially when individual participants are recruited (Giraudeau and Ravaud, 2009). In order to minimise this potential bias, three strategies will be employed. Firstly, as described previously, recruiters will use a randomly generated ward list to determine the order by which they will approach wards during recruitment. Secondly, recruiters will not be the same individuals as those delivering the intervention. Finally, we will blind the recruiters to the actual intervention. The patient informa- tion and consent forms only state we will be examining various PIP strategies and in training recruiters will also be told this. Recruiters at each site will also be trained at their own hospital only.

Full blinding in this study is difficult due to the study design (i.e. patients participating in their care) and the need for informed consent. However recruiters, nursing staff and patients will only be told broad details of their arm of the study, for example that the study involves documenting PI and various strategies for PIP, and will be unaware that the study is a c-RT. All research assistant groups (recruiters, interventionists and outcome asses- sors) will be blinded to study design and hypotheses, and will be trained separately to avoid the intervention being known to all groups. Outcome assessors will be blinded to group allocation and will be assigned to one hospital only. Success of outcome assessor blinding will be assessed at the end of the study using the James Blinding Index (James et al., 1996). Finally, data analysts will be blinded to group allocation.

2.7. Research assistant training and treatment fidelity

All research assistants will undergo separate and group specific onsite training tailored to their role to ensure consistency across all sites. All research assistants will use standardised procedure manuals providing specific detail on their roles and data collection, plans for dealing with intervention fidelity issues and monitoring the delivery of the intervention. Participants in all groups will receive identical information and instructions regarding the study, except for the actual intervention.

2.8. Data collection

One research assistant will recruit the patients, obtain consent and collect demographic and clinical data. A second research assistant will deliver the intervention. Trained outcome assessors will collect daily data and assess the outcomes (blind to allocation). An electronic case record form will be developed to capture cluster level and patient level data and will be completed by the recruiter and/or the project manager. At the cluster level, the number of hospital beds, most recent HAPI rate and hospital PIP resources will be collected at baseline. At the individual patient level, demographic and clinical data such as gender, diagnosis, body mass index and risk factors for PI will be collected at baseline. This patient data will be collected from the medical records and the patient. The use of any other PIP interventions (i.e. as part of standard care) for each patient such as pressure relieving devices, special diets etc. will be recorded from the daily care plan, medical records or by direct observation by the outcome assessor. The outcome assessors will monitor and collect the same information on actual PIP interventions delivered to patients in the control wards as for patients in the intervention wards.

2.9. Outcome assessment

The primary outcome is incidence of HAPI, and secondary outcomes are pressure injury stage, patient participation in care and health care costs. Patients will reach the trial endpoints of: development of a pressure injury, hospital discharge or 28 days, whichever comes first; transfer to another hospital or transfer to critical care requiring mechanical ventilation. The outcome assessors, who will be trained in skin assessment, will visually inspect the skin of all participants daily to determine if a PI is present, and if so, its stage. This skin assessment will follow international consensus guidelines for assessing PI (European Pressure Ulcer Advisory Panel et al., 2014). Outcome assessor training will be led by an experienced Wound Care Nurse Practitioner and will focus on the outcome data to be collected including skin assessment, as well as other daily data to be collected. Each site will be funded for one half-time equivalent outcome assessor position plus weekend half-time, however due to staffing logistics (i.e. holidays, sick leave), 3–4 outcome assessors will be trained per site and are anticipated to be used using a roster system. All outcome assessors will receive a full day training in outcome assessment and data collection. This training will include theory and practice related to full skin integrity assessment, PI identification and staging. All assessors will undertake a test under exam conditions in which they will be asked to observe high definition photographs of PIs and determine whether the photograph shows a PI and if so, what stage. Inter-rater reliability of outcome assessors will be assessed using Fleiss Kappa for multiple raters on the primary outcome (presence of a PI) and secondary outcome (stage of PI). Research suggests that with training, excellent inter-rater reliability (Kappa 1.0, i.e. 100% agreement between trained clinical trials nurses and their team leader) can be achieved (Nixon et al.,

200 pro ent ass inc sca end inp kno pro par (W

2.10

tria obs inte fram tion fram pro del out eva be d stu rec ind ins of inte exp and the be e at a rev and ass rou

Tab

Sum

Re

PI

PI

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–1668 1665

5). During the trial, monitoring will occur, with the ject manager using source data to verify the date ered into the electronic case report form. The outcome essor will also collect data for secondary outcomes luding patients’ responses to the patient participation le (Weingart et al., 2011) when patients reach the trial point. This scale measures patient participation in atient care by asking seven questions around patients’ wledge and understanding of medical conditions and cedures, ability to access health professionals and ticipate in decision making, and hospital experiences eingart et al., 2011).

. Process evaluation

A process evaluation will be undertaken alongside the l to explain discrepancies between expected and erved outcomes and provide insight into why the rvention is successful or not (Craig et al., 2008). A ework for designing and reporting process evalua-

s for c-RTs will be employed (Grant et al., 2013). This ework consists of a number of elements for evaluating

cesses, including recruitment and reach, intervention ivery, response to intervention, maintenance, context, comes and theory (Grant et al., 2013). In the process luation of the INTACT trial, recruitment and reach will escriptively analysed using screening log data from all

dy sites (i.e. total number of patients screened vs. ruited). Intervention delivery to clusters (nurses) and ividuals (patients) will be measured by keeping ervice logs of staff training and documenting delivery each intervention component to individuals at all rventions sites. Response to the intervention will be lored at the cluster and individual level through nurse

patient interviews. The possibility of maintenance of intervention (i.e. translation into clinical practice) will xplored in nurse interviews. Context will be considered local, state and national level regarding PIP care by

iewing hospital policies and procedures, state penalties national standards. Unintended consequences will be

essed through any changes in policy, processes or tine care at each site.

2.11. Cost and resource use data

A trial-based economic evaluation will be undertaken from the health system perspective to compare the direct healthcare costs and effects of a care bundle for PIP, relative to standard care. The resource utilisation data to be recorded and costed are summarised in Table 2. Costs related to provision of the care bundle such as time to educate patients and train nursing staff and costs of developing resources such as brochures, posters and DVDs will be collected for all trial participants, as will length of hospital stay. During a 4-week observational sub-study, micro-costing data will be collected by a separate research assistant, which will allow the calculation of direct costs to the hospital for pressure-related assessment and preven- tion for each participant across all sites. This sub-study sample is expected to represent 320 participants (20% of the trial cohort) and will be sufficient to indicate the mean and distribution of costs related to implementation and other PIP strategies used for each patient. The sub-study will determine the number of repositioning episodes per participant, number of clinical staff required for reposi- tioning, and the nurse time required per turn. Other resource use related to PIP such as the use of special mattresses, skincare products and incontinence care will also be recorded for this sub-study. These resource use data will be collected by directly observing the patients in the sub-study and auditing medical records. Direct costs (AU$ 2014) will then be allocated to each resource unit using standard costing sources. These data from the sub- study will be used alongside intervention and length of stay data to estimate the total resource use and direct healthcare cost per participant for the two groups over the entire study period.

2.12. Data analysis

Blinded analyses of primary and secondary outcome measures will be undertaken at cluster level and patient level. Our primary hypothesis will be tested at the individual patient level but adjusted for the cluster structure as per the recommendation of Cochrane methods

le 2

mary of resource data collection and costing methods.

source to be measured Data collection and costing method Source for unit cost

PCB intervention

Materials including poster, brochure, DVD Fixed cost, averaged across all trial participants Commercial provider

Staff time for patient education and

nurse training

Recorded by RA for all patients in trial and

estimated from nurse training sessions

Staff salary hourly rates

(including on-costs)

P strategies

Clinical staff time for turning Observational sub-study: number of nurses per

turn, turns per patient and time per turn observed,

confirmed with patient, and recorded by RA

Staff salary hourly rates

(including on-costs)

PIP products (including mattresses, wedges,

rings and cushions; skincare products;

incontinence care; dressings for PIP)

Observational sub-study: observed, confirmed

with patient, and recorded by RA

Hospital cost centre

Dietitian consult for the purpose of PIP Observational sub-study: patient notes, confirmed

with patient, and recorded by RA

Staff salary hourly rates

(including on-costs)

Hospital length of stay Length of stay recorded by RA for all Independent Hospital Pricing

patients in trial Authority (2014)

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–16681666

on the analyses of c-RT. Generalised Estimation Equations (GEE) models, hierarchical or generalised mixed models and multi-level models will be used to adjust for clustering of patient-level data. Within each approach, simple analyses such as t-tests, Chi-square tests or more complex approaches such as multivariate Cox regression models will be considered. Both allow for the effect of the intervention on the incidence of HAPI and other secondary outcomes to be tested; however, only complex analyses allow adjustment for potential covariates. Adjustment for individual patient covariates will occur in the appropriate level of analyses using GEE or multi-level models.

Cluster level analysis: The incidence rates per 1000 hos- pitalised days between intervention and standard care groups will be compared at cluster level. Estimates of the HAPI incidence in each group, group differences as well as the 95% CI and p values will be reported. Other outcomes will also be compared at the cluster levels between the intervention and control group. Baseline variables and other covariates will be compared between the two groups to ensure the intervention and control groups do not differ in their baseline characters.

Patient level analysis: Patient level analyses will primarily account for the intra-cluster correlation, thus increasing the statistical power of the analysis. Cluster adjusted Z-tests (to compare the proportion between the intervention and control groups) or t-tests (to compare means) will be undertaken to avoid spuriously low p-value and overly narrow confidence levels, over-emphasizing the impact of the intervention. However, the more compre- hensive inferences in our study will be based on the use of modelling techniques to incorporate patient level data adjusted for nested structure. Cox regression with robust standard error will be used to estimate the rate ratios given the incidence rate per person time is the main outcome in this study. Robust standard error accounts for the correlation outcome within each hospital. Cox regression will also adjust for a number of pre-specified covariates in comparing the incidence of HAPI between the intervention and standard care group. These modelling techniques allow the direct correlation within clusters to be modelled explicitly, and many potential confounding factors will be included in the model when comparing the effect of the intervention. Such models will be developed for primary and secondary outcomes. We will use STATA for the analyses of our data. An intention to treat analysis will be used but we will also undertake a per protocol analysis to elucidate differences in outcomes depending on the intended vs. actual intervention received.

Cost-effectiveness analysis: Detailed within-trial re- source utilisation and costs will be used to undertake a stepped economic evaluation and estimate (i) the comparative per patient HAPI related healthcare preven- tion costs for each group, and (ii) the comparative incremental cost of preventing an additional case of HAPI. As the duration of the trial is less than one year, discounting will not be applied to costs or benefits. Hierarchical modelling approaches and cluster-adjusted non-parametric bootstrapping techniques will be employed to compare mean difference in the total costs between groups, and to estimate a confidence interval

around the mean (Bachmann et al., 2007). In addition, a cost-effectiveness analysis will be undertaken based on the primary outcome measure (incidence of HAPI), to estimate the incremental cost per additional person remaining free from PI.

3. Discussion

This study will provide important information on the effectiveness of a patient-centred PIP care bundle to reduce the incidence and severity of hospital-acquired PI in at-risk patients, improve patient participation in their care and reduce associated health care costs. To our knowledge, this is the first randomised study to incorporate a number of recognised PIP strategies into a care bundle with a focus on patient engagement and nurse–patient partnership. This is a novel approach to a significant clinical problem. It is multifaceted, patient-focused and expected to maximise effectiveness and sustainability.

PIP is of high importance considering the significant patient consequences and economic burden to hospitals. As noted by an international Expert Wound Care Advisory Panel, PI research lags behind that in other areas of medicine and very few RCTs have been conducted (Armstrong et al., 2008). Whilst recent systematic reviews show promising evidence around care bundles (i.e. a set of evidence based practices) for PIP (Soban et al., 2011; Sullivan and Schoelles, 2013), it has been suggested that further research is needed to better understand imple- mentation, mechanisms (or causal pathways) and the role of context in intervention success or failure (Soban et al., 2011). This proposed c-RT will be cutting edge. If it confirms the utility of a patient centred PIP care bundle, it could change practice worldwide.

This study has a number of strengths, including its robust design. Cluster randomisation (i.e. by hospital) will ensure there is no contamination between study groups. Thorough training of interventionists and data collectors will optimise fidelity and reliability. Whilst blinding of patients is not possible due to the need for informed consent, blinding of data collectors and nurses to study design and hypotheses, and blinding of outcome assessors and data analysts to group allocation will minimise bias. The study also has a strong theoretical base. Care bundles have been shown to improve patient outcomes and are based on core practice guidelines (Institute for Healthcare Improvement, 2014; Rello et al., 2011). Patient participa- tion in their health care results in improved outcomes and responses to interventions, and is promoted by leading healthcare organisations worldwide (Australian Commis- sion on Safety and Quality in Health Care, 2011; Joint Commission on Accreditation in Healthcare Organizations, 2006; World Health Organisation, 2007). The inclusion of an economic sub-study will strengthen the impact of the findings by providing meaningful cost-analysis data, and may inform recommendations on the adoption of the intervention for PIP in clinical practice. Finally, the inclusion of a process evaluation will provide important information on implementation and give insight into why the intervention is successful or not and how it may be optimised.

rec of w pat out disc ens out wil avo usi con

hos rele ma nur hea wil clin test ind Aus

3.1.

has

Con

Fun

Res (AP

Eth

num

Aut

to s

acq

inte

and

Ref

Arm

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–1668 1667

Challenges in this trial may include: (1) reaching ruitment targets, but we plan to increase the number

ards participating at each site if required; (2) when ients are unexpectedly discharged, resulting in missing come data, although we have considered hospital harge practice in timing outcome assessments; (3) uring the blinding of research assistants (including come assessors) to study site group allocation, but we l assess their perception of group allocation; and (4) iding selection bias, however this will be addressed by

ng a ward randomisation schedule and recruiting secutive patients according to the schedule. The results of this study will be presented at local pital, patient safety and university fora, and a press ase will be prepared. Abstracts will be submitted to

jor international meetings such as patient safety, sing and medicine and published in high-ranking lth services and medical/nursing journals. The results l have international application and implications for ical practice and nursing education. If successful, ing of the intervention internationally would be icated, and it should be rapidly adopted and cited in tralian core practice guidelines.

Trial status

At the time of manuscript submission, data collection commenced at all study sites.

flict of interest

None declared.

ding

This study is funded by the National Health and Medical earch Council (NHMRC) through a Project Grant P1058963).

ical approval

Ethical approval given by Queensland Health (reference ber HREC/13/QGC/192).

hor contributions

WC, TB, JWe, EM, MW, BG, JWh, LT and NC contributed tudy conception.

All authors except SR contributed to study design. WC, TB, JWe, EM, MB, MW and BG contributed to data uisition. All authors will contribute to data analysis and rpretation.

All authors contributed to article drafting and revision, have approved the final version of the manuscript.

erences

strong, D.G., Ayello, E.A., Capitulo, K.L., Fowler, E., Krasner, D.L., Levine, J.M., Sibbald, R.G., Smith, A.P., 2008. New opportunities to

hospital-acquired conditions policy: a consensus paper from the International Expert Wound Care Advisory Panel. Adv. Skin Wound Care 21 (10), 469–478.

Arnetz, J.E., Winblad, U., Höglund, A.T., Lindahl, B., Spångberg, K., Wal- lentin, L., Wang, Y., Ager, J., Arnetz, B.B., 2010. Is patient involvement during hospitalization for acute myocardial infarction associated with post discharge treatment outcome? An exploratory study. Health Expect. 13 (3), 298–311.

Australian Commission on Safety and Quality in Health Care, 2011. National Safety and Quality Health Service Standards. ACSQHC, Sydney.

Ayello, E.A., Lyder, C.H., 2008. A new era of pressure ulcer accountability in acute care. Adv. Skin Wound Care 21 (3), 134–140.

Bachmann, M.O., Fairall, L., Clark, A., Mugford, M., 2007. Methods for analyzing cost effectiveness data from cluster randomized trials. Cost. Eff. Resour. Alloc. 5 (1), 12.

Baldelli, P., Paciella, M., 2008. Creation and implementation of a pressure ulcer prevention bundle improves patient outcomes. Am. J. Med. Qual. 23 (2), 136–142.

Banks, M.D., Graves, N., Bauer, J.D., Ash, S., 2010. The costs arising from pressure ulcers attributable to malnutrition. Clin. Nutr. 29 (2), 180–186.

Centers for Medicare, Medicaid Services, 2007. Medicare program: changes to the hospital inpatient prospective payment systems and fiscal year 2008 rates. Fed. Reg. 72 (162), 47129.

Chaboyer, W., Gillespie, B.M., 2014. Understanding nurses’ views on a pressure ulcer prevention care bundle: a first step towards successful implementation. J. Clin. Nurs. 23 (23–24), 3415–3423.

Coleman, S., Gorecki, C., Nelson, E.A., Closs, S.J., Defloor, T., Halfens, R., Farrin, A., Brown, J., Schoonhoven, L., Nixon, J., 2013. Patient risk factors for pressure ulcer development: systematic review. Int. J. Nurs. Stud. 50 (7), 974–1003.

Coulter, A., 2006. Engaging Patients in their Healthcare: How is the UK doing Relative to Other Countries? Picker Institute Europe, Oxford, UK.

Craig, P., Dieppe, P., Macintyre, S., Michie, S., Nazareth, I., Petticrew, M., 2008. Developing and Evaluating Complex Interventions: The New Medical Research Council Guidance.

Davis, R.E., Pinto, A., Sevdalis, N., Vincent, C., Massey, R., Darzi, A., 2012. Patients’ and health care professionals’ attitudes towards the PINK patient safety video. J. Eval. Clin. Pract. 18 (4), 848–853.

Department of Health, 2012. Delivering the NHS Safety Thermometer Commissioning for Quality and Innovation 2012/13: A Preliminary Guide to Delivering ‘Harm Free’ Care. Department of Health, London.

Dwamena, F., Holmes-Rovner, M., Gaulden, C.M., Jorgenson, S., Sadigh, G., Sikorskii, A., Lewin, S., Smith, R.C., Coffey, J., Olomu, A., 2012. Inter- ventions for providers to promote a patient-centred approach in clinical consultations. Cochrane Database Syst. Rev. 12 .

European Pressure Ulcer Advisory Panel, National Pressure Ulcer Advisory Panel, Pan Pacific Pressure Injury Alliance, 2014. Prevention and Treatment of Pressure Ulcers: Quick Reference Guide. Media, C, Perth, Australia.

Gallagher, P., Barry, P., Hartigan, I., McCluskey, P., O’Connor, K., O’Connor, M., 2008. Prevalence of pressure ulcers in three university teaching hospitals in Ireland. J. Tissue Viability 17 (4), 103–109.

Gillespie, B.M., Chaboyer, W., Sykes, M., O’Brien, J., Brandis, S., 2014. Development and pilot testing of a patient-participatory pressure ulcer prevention care bundle. J. Nurs. Care Qual. 29 (1), 74–82.

Giraudeau, B., Ravaud, P., 2009. Preventing bias in cluster randomised trials. PLoS Med. 6 (5), e1000065.

Gorecki, C., Brown, J.M., Nelson, E.A., Briggs, M., Schoonhoven, L., Dealey, C., Defloor, T., Nixon, J., European Quality of Life Pressure Ulcer Project group, 2009. Impact of pressure ulcers on quality of life in older patients: a systematic review. J. Am. Geriatr. Soc. 57 (7), 1175–1183.

Gorecki, C., Closs, S.J., Nixon, J., Briggs, M., 2011. Patient-reported pressure ulcer pain: a mixed-methods systematic review. J. Pain Symptom Manage. 42 (3), 443–459.

Grant, A., Treweek, S., Dreischulte, T., Foy, R., Guthrie, B., 2013. Process evaluations for cluster-randomised trials of complex interventions: a proposed framework for design and reporting. Trials 14 (1), 15.

Graves, N., Birrell, F., Whitby, M., 2005. Effect of pressure ulcers on length of hospital stay. Infect. Control Hosp. Epidemiol. 26 (3), 293–297.

Graves, N., Zheng, H., 2014. Modelling the direct health care costs of chronic wounds in Australia. Wound Pract. Res. 22 (1), 20–33.

Gunningberg, L., Hommel, A., Bååth, C., Idvall, E., 2013. The first national pressure ulcer prevalence survey in county council and municipality settings in Sweden. J. Eval. Clin. Pract. 19 (5), 862–867.

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. Int. Wound J. 8 (5), 465–473.

improve pressure ulcer prevention and treatment: implications of the CMS inpatient hospital care present on admission indicators/

W. Chaboyer et al. / International Journal of Nursing Studies 52 (2015) 1659–16681668

Igarashi, A., Yamamoto-Mitani, N., Gushiken, Y., Takai, Y., Tanaka, M., Okamoto, Y., 2013. Prevalence and incidence of pressure ulcers in Japanese long-term-care hospitals. Arch. Gerontol. Geriatr. 56 (1), 220–226.

Institute for Healthcare Improvement, 2014. What is a Bundle? . James, J., Evans, J.A., Young, T., Clark, M., 2010. Pressure ulcer prevalence

across Welsh orthopaedic units and community hospitals: surveys based on the European Pressure Ulcer Advisory Panel minimum data set. Int. Wound J. 7 (3), 147–152.

James, K.E., Bloch, D.A., Lee, K.K., Kraemer, H.C., Fuller, R.K., 1996. An index for assessing blindness in a multi centre clinical trial: disulfiram for alcohol cessation: a VA cooperative study. Stat. Med. 15 (13), 1421– 1434.

Joint Commission on Accreditation in Healthcare Organizations, 2006. Patient-Inclusive Care: Encouraging Patients to be Active Participants in their Care.

Latimer, S., Chaboyer, W., Gillespie, B., 2014. Patient participation in pressure injury prevention: giving patient’s a voice. Scand. J. Caring Sci. 28 (4), 648–656.

Lyder, C.H., Wang, Y., Metersky, M., Curry, M., Kliman, R., Verzier, N.R., Hunt, D.R., 2012. Hospital-acquired pressure ulcers: results from the National Medicare Patient Safety Monitoring System Study. J. Am. Geriatr. Soc. 60 (9), 1603–1608.

McInnes, E., Jammali-Blasi, A., Bell-Syer, S., Dumville, J., Cullum, N., 2012. Preventing pressure ulcers—are pressure-redistributing support sur- faces effective? A Cochrane systematic review and meta-analysis. Int. J. Nurs. Stud. 49 (3), 345–359.

Moore, Z., Cowman, S., Conroy, R.M., 2011. A randomised controlled clinical trial of repositioning, using the 30 tilt, for the prevention of pressure ulcers. J. Clin. Nurs. 20 (17–18), 2633–2644.

Mulligan, S., Prentice, J., Scott, L., 2011. WoundsWest Wound Prevalence Survey State-Wide Overview Report. Perth, WA.

Nguyen, K.-H., Chaboyer, W., Whitty, J., 2015. Pressure injury in Austra- lian public hospitals: a cost of illness study. Aust. Health Rev.. http:// www.publish.csiro.au/?act=view_file&file_id=AH14088.pdf.

Nixon, J., Thorpe, H., Barrow, H., Phillips, A., Andrea Nelson, E., Mason, S.A., Cullum, N., 2005. Reliability of pressure ulcer classification and diagnosis. J. Adv. Nurs. 50 (6), 613–623.

Padula, W.V., Mishra, M.K., Makic, M.B.F., Valuck, R.J., 2014. A framework of quality improvement interventions to implement evidence-based practices for pressure ulcer prevention. Adv. Skin Wound Care 27 (6), 280–284.

Queensland Government, Department of Health, 2014. Health Funding Principles and Guidelines 2013–14.

Rello, J., Chastre, J., Cornaglia, G., Masterton, R., 2011. A European care bundle for management of ventilator-associated pneumonia. J. Crit. Care 26 (1), 3–10.

Soban, L.M., Hempel, S., Munjas, B.A., Miles, J., Rubenstein, L.V., 2011. Preventing pressure ulcers in hospitals: a systematic review of nurse- focused quality improvement interventions. Joint Comm. J. Qual. Patient Saf. 37 (6), 245.

Sullivan, N., Schoelles, K.M., 2013. Preventing in-facility pressure ulcers as a patient safety strategy: a systematic review. Ann. Intern. Med. 158 (5 (Part 2)), 410–416.

van Gaal, B.G.I., Schoonhoven, L., Mintjes, J.A.J., Borm, G.F., Hulscher, M.E.J.L., Defloor, T., Habets, H., Voss, A., Vloet, L.C.M., Koopmans, R.T.C.M., van Achterberg, T., 2011. Fewer adverse events as a result of the SAFE or SORRY? programme in hospitals and nursing homes. Part I: Primary outcome of a cluster randomised trial. Int. J. Nurs. Stud. 48 (9), 1040–1048.

VanGilder, C., Amlung, S., Harrison, P., Meyer, S., 2009. Results of the 2008–2009 International Pressure Ulcer Prevalence Survey and a 3- year, acute care, unit-specific analysis. Ostomy Wound Manage. 55 (11), 39.

Weingart, S.N., Zhu, J., Chiappetta, L., Stuver, S.O., Schneider, E.C., Epstein, A.M., David-Kasdan, J.A., Annas, C.L., Fowler, F.J., Weissman, J.S., 2011. Hospitalized patients’ participation and its impact on quality of care and patient safety. Int. J. Qual. Health Care 23 (3), 269–277, mzr002.

World Health Organisation, 2007. Patients for Patient Safety. World Health Organisation, Geneva.