Urinary incontinence (UI) and lower urinary tract symptoms (LUTS) are commonly experienced by adults admitted to hospitals, also referred to as inpatients, and contribute to the complexity and cost of providing care to these individuals. UI types include functional, neurogenic, stress, overflow, continuous, urgency, and mixed UI [1,2]. LUTS include acute and chronic urinary retention, frequency, urgency, and nocturia [1,2]. Surprisingly, data on the prevalence of UI and LUTS in adult inpatients are limited. However, these conditions have been reported to range from 10% to 45% of patients receiving acute and subacute care in hospitals [3-5]. The often-taboo subject of UI is significantly associated with poorer patient outcomes, including urinary tract or urinary catheter–associated infections [6,7], incontinence-associated pressure injury [8], falls [9], and pain associated with these conditions [6,7]. People with UI are twice as likely to experience depression and are more often socially isolated [6,7,10]. UI is associated with increased carer stress and is a main reason for carers feeling unable to continue in the carer role, leading to residential care admissions [11]. Although UI and LUTS are often complex and not always curable, with appropriate clinical care, symptoms can be managed and complications can be avoided.

International and Australian clinical practice guidelines provide recommendations for optimal care for UI and LUTS, based on the current, albeit limited research evidence [1,12-16]. Stroke is an example where UI and LUTS care has been included in condition-specific guideline recommendations [17-19]. Australian stroke guideline recommendations for optimal care are that all people poststroke are screened for continence issues and that those with symptoms receive an assessment, diagnosis, and a tailored inpatient and postdischarge management plan [17]. In the 2017 National Stroke Audit Acute Services, of the one-third of inpatients with UI, only 18% received a management plan [20]. In the 2018 National Stroke Audit Rehabilitation Services, of the 41% of inpatients who had UI, 52% had a documented management plan [21]. These results indicate an evidence-practice gap in current inpatient UI/LUTS care.

Although peak bodies recommend that health services have systems for optimal UI and LUTS care [1,15,17], there is little reporting of effective inpatient interventions to systematically deliver this care, as demonstrated in stroke care. In a recent Cochrane review, it was identified that there was limited evidence for the effectiveness of UI interventions poststroke [6]. The review included 20 trials (with 1338 participants, reporting 21 comparisons), with the authors reporting that the risk of bias was impossible to judge for many of the included studies because of poor reporting. The authors call for more robust multicenter trials.

As part of our formative quality improvement research [22], in 2009-2010, we translated high-level UI and LUTS guideline recommendations into an intervention that presents clear, concise, and explicit optimal inpatient care in a user-friendly format [23]. We collaborated with health service clinicians and managers from 3 rehabilitation services in the Hunter Region, Australia, to synthesize the best-available evidence into our Structured urinary Continence Assessment and Management Plan (SCAMP) intervention that was specifically designed for inpatients poststroke in metropolitan rehabilitation units [22]. Our SCAMP intervention consists of (1) a 4-page clinical decision support tool guiding comprehensive UI and LUTS assessment, diagnosis, and management; (2) the associated clinical practice guideline; and (3) supporting web-based education modules [22]. Stroke clinicians using the SCAMP intervention identified that it has the potential to be applicable across a range of hospitals in different health districts, for inpatients with a range of diagnoses including stroke, and across the phases of inpatient care.

The aim of this study is to determine if the implementation of our SCAMP intervention is feasible and effective across this range of clinical scenarios.

This will be a pragmatic, real-world, before- and after-implementation study conducted at 12 hospitals. Data will be collected at 3 time points: before implementation (T₀), immediately after the 6-month implementation period (T₁), and again after a 6-month maintenance period (T₂; Figure 1). Data will be collected from medical record audits and clinician questionnaires. An economic evaluation from the perspective of hospitals will be conducted. We will report our primary findings according to the Standards for Reporting Implementation Studies (StaRI) [24,25].

Eleven hospitals in New South Wales (NSW) and 1 in Queensland, Australia, participated in our study. The hospitals are located in 4 health service districts. The hospitals were a convenience sample. Ten of the hospitals are located in 2 Local Health Districts that form part of the NSW Regional Health Partners, a Centre for Innovation in Regional Health (accredited by the National Health and Medical Research Council). Lead clinicians from the other 2 hospitals heard about our SCAMP intervention at conference presentations and approached the lead author about adopting the intervention.

Fifteen wards where patients after stroke are admitted are participating (Table 1). In Australia, people after stroke are cared for on wards that admit people with a range of conditions. This care may be provided in a stroke unit embedded in the ward or as part of the general ward population. This study was instigated by stroke clinicians who identified UI/LUTS inpatient care needed to be improved on their ward for people after stroke and potentially for other inpatient populations. To be eligible to participate, key ward clinicians and managers had to identify that UI/LUTS care was an issue for their ward and that they were willing to commit resources toward improving optimal UI and LUTS care by implementing the SCAMP intervention. The characteristics of each ward are outlined in Table 1. A total of 47% (7/15) of the wards are in 4 hospitals in 2 major cities, and the other 53% (8/15) wards are in 8 hospitals in inner regional locations [26]. The wards included 43% (7/15) rehabilitation wards, 27% (4/15) acute medical wards with an embedded stroke unit, 13% (2/15) wards with both acute and rehabilitation inpatients, 7% (1/15) rehabilitation ward with an embedded stroke unit, and 7% (1/15) medical ward. Before commencing the study, the investigators from each ward nominated the target adult inpatient populations from their ward to be included in the study of acute stroke, acute medicine, and/or rehabilitation for any condition, including stroke (Table 1). Clinician representatives from each site have been project team members from the outset (including authors KB, JAD, JS, FM, AS, JB, SO, AB, and SL).

The population targeted by our practice-change package is clinicians (full time, part time, and casual) employed in each participating ward (including nurses, Nurse Unit managers, physiotherapists, occupational therapists, speech pathologists, social workers, and doctors). Participating clinicians are general medical, rehabilitation, or neuroscience clinicians who are not identified as continence or urology specialists. There are no exclusion criteria, as the study is a service improvement initiative, clinicians will not be consented to receive our practice-change package. The unit of analysis is hospital performance, based on patient-level data.

Our practice-change package is designed to support clinicians and health services to deliver guideline-recommended UI and LUTS care. It consists of our SCAMP intervention that we will implement using evidence-based implementation strategies.

Our primary outcome is the change (T₁-T₀) in the proportion of inpatients who have an individually tailored UI/LUTS management plan. This will be determined via a medical record audit.

Our secondary research outcomes are:

Data will be collected at 3 time points (Figure 1): before implementation (T₀), after a 6-month implementation period (T₁), and after a 6-month maintenance period (T₂). Before-implementation data will be used to tailor the intervention to each ward.

For the primary outcome, 15 consecutive medical record audits per site per month (ie, a pooled sample of 675 audits anticipated per data collection period) will provide >90% power to detect a 10% absolute increase (from before intervention) in the proportion of incontinent patients with a continence management plan (type 1 error rate of 5%). This calculation conservatively assumes that 20% of patients in acute and 50% of rehabilitation sites have a plan before intervention (based on the Australian Stroke Foundation National Audit results for included sites [20,39]).

The before-intervention group, after-intervention group, and maintenance period group results will be presented with descriptive statistics, including site, clinician, and medical record data for characteristics and demographics. No individual will be identifiable. All results will be presented as aggregated summary measures, with their variance depending on the distribution of the data (eg, mean and standard deviation, medians, and interquartile range). Groups will be compared with respect to change, from baseline (T₀) to immediately postintervention period (T₁) and from baseline to maintenance period (T₂) using mixed effects logistic regression models, with a random intercept for site, and fixed effect for period. Results are presented as odds ratios with 95% CI and P values.

There are no criteria for study discontinuation as it is not anticipated that there are any events that would warrant discontinuation of this study. Any unforeseen adverse events will be reported to the Hunter New England Human Research Ethics Committee (the primary approval committee) and advice sought out regarding the required action. Any deviations from this original protocol will be reported in our study outcomes papers.