Tibial nerve stimulation compared with sham to reduce incontinence in care home residents: ELECTRIC RCT

Urinary incontinence in care homes

Urinary incontinence (UI) is defined by the International Continence Society as ‘any involuntary loss of urine’. 2 Prevalence of UI increases with age,3 with the highest prevalence found in older adults living in residential or nursing care homes. 3 Adults living with dementia are three times more likely to have UI than individuals of equivalent age and characteristics who do not have dementia,4 with the development of UI often being a catalyst for individuals requiring full-time care. 5,6 The most common type of UI in the care home population is mixed UI,3 in which symptoms of urgency UI and stress UI occur together, and are often accompanied by additional functional leakage due to physical and cognitive frailty. Mixed UI is considered the most challenging form of incontinence to manage as it includes symptoms of both of the common types of UI (urgency and stress) and is often resistant to intervention. 7

Burden of urinary incontinence

The burden of UI is significant, and is increasing for older adults and care services. 8 It is a distressing condition for older adults, which has a negative impact on their dignity and quality of life. 9 UI is associated with cognitive and physical impairment,8,10 increased risk of falls and fractures,11 sleep disturbance,9 hygiene problems and tissue viability problems. 12 It also has a significant impact on social participation, which can lead to isolation and clinical depression. 13,14 Although the personal costs associated with reduced quality of life and social withdrawal are not quantifiable, the cost of treatment and management of UI remains high. The NHS currently spends upwards of £80M per year on absorbent pads alone for the purpose of containing UI,15 which excludes the costs of associated care needed to manage use of absorbent pads in older, frailer individuals or those living with dementia. With the number of such individuals predicted to increase rapidly as the population ages, the cost of managing UI is also set to increase, with significant implications for future care provision. 16

Management of continence in care homes

Urinary incontinence is viewed as a generic condition in care homes: assessment and identification of specific types of UI or bladder problems is not routine, and interventions targeted towards treatment of specific UI types, such as mixed or urgency UI, have not been extensively demonstrated in this setting. The most common management option for bladder and bowel incontinence in care homes is the use of absorbent products to contain the leakage, combined with scheduled toileting. 17,18 Other conservative, non-pharmacological options to aid management of UI, such as individualised voiding programmes, bladder training or pelvic floor muscle exercise programmes, are often considered unsuitable for care home environments because they are labour intensive and require a degree of co-operation and engagement, which may not be suitable for those with cognitive impairment or dementia. Antimuscarinic drugs may be prescribed for the management of urgency UI or overactive bladder problems; however, their use is associated with significant adverse effects in frail older people and may counteract the functional benefits of anticholinesterase inhibitors used in the treatment of dementia. 19,20 Active treatment of UI or the promotion of continence in care homes is rare and a containment approach predominates. 18

Use of transcutaneous posterior tibial nerve stimulation to treat urinary incontinence and possible mechanism of effect

Transcutaneous posterior tibial nerve stimulation (TPTNS) is a form of non-invasive neuromodulation used to treat the symptoms of an overactive bladder (OAB), including urgency UI and mixed UI. TPTNS uses surface electrodes applied behind the medial malleolus to electrically stimulate the tibial nerve. A full treatment programme is delivered in 30-minute bouts, twice per week over a 6-week period. Although the exact mechanism of action is yet to be fully understood, TPTNS is believed to restore the balance between excitatory and inhibitory bladder functioning by modulating the signal traffic to and from the bladder through the sacral plexus. 21 Current understanding suggests that stimulating afferent sacral nerves in the lower leg increases inhibition of the efferent pelvic nerve activity, which reduces detrusor contractility and increases bladder capacity. 21–23 By these means, TPTNS reduces the sensation of urgency and the frequency of voiding demanded, thus enabling improved bladder control. These mechanisms may also reduce the volume of urine retained in the bladder after voiding. 23 For the care home population, TPTNS may reduce the sudden urge to urinate, allowing residents more time to reach a toilet, and because the bladder capacity may increase the frequency of daily voiding may also reduce. In turn these changes may enable more appropriate use of the toilet to void, rather than relying on containment of leakage by absorbent products.

A treatment approach, as opposed to a containment approach, engenders respect for the person’s right to use a toilet. Treatment using TPTNS is distinctive because, unlike commonly used behavioural interventions, there is no requirement for the recipient to engage with the treatment. Therefore, it is highly suited to individuals who are frail and/or living with dementia, and the stimulation site at the tibial nerve in the ankle ensures that the person’s dignity and comfort are maintained.

Evidence for effectiveness of transcutaneous posterior tibial nerve stimulation

There is evidence that TPTNS can be effective at reducing symptoms of urgency or mixed UI in women, and in adults with neurogenic bladder dysfunction. 24–27 A systematic review of TPTNS for the treatment of UI using Cochrane methodology identified 10 randomised controlled trials (RCTs), with a total of 472 participants. 24 Improvements were reported in all studies in terms of bladder symptoms and/or UI-related quality of life. Although the studies had small sample sizes and contained methodological weaknesses [Grading of Recommendations Assessment, Development and Evaluation (GRADE) score gave ‘low quality’ ratings], the results indicated that TPTNS showed promise as a safe, low-cost intervention for UI. 28 One small-scale feasibility study has been undertaken in care homes29 and indicated that TPTNS had the potential to be a safe, acceptable and effective treatment for UI in this population. However, definitive evidence on the effectiveness of TPTNS is required before it can be recommended for routine practice in a care home context.

Rationale for research

Current evidence24 to date has been generated from studies based on women only, or on both men and women adult populations. The participants in these studies had both physical and cognitive capacity, and the studies were conducted in outpatient health-care contexts. One small pilot study has shown the safety and acceptability of using TPTNS in care homes,29 but no other research in this care setting has been undertaken. Given the complexity of need, and the physical and mental dependence of care home residents, and the particular challenges of UI and its management in care homes there is a need for definitive evidence on the effectiveness of TPTNS before it can be recommended for routine practice in the care home context.

Aim

The aim of the ELECTRIC (ELECtric Tibial nerve stimulation to Reduce Incontinence in Care homes) trial was to determine the clinical effectiveness of a programme of TPTNS to treat UI in care home residents, and the associated costs and consequences.

Objectives

1. To establish whether or not TPTNS is more effective than sham stimulation for reducing the volume of UI at 6, 12 and 18 weeks in care home residents.

2. To investigate mediating factors that have an impact on the effectiveness of TPTNS in a mixed-methods process evaluation involving fidelity, implementation support and qualitative components.

3. To undertake an economic evaluation of TPTNS in care homes assessing the costs of providing the programme and presenting these costs alongside the key primary and secondary outcomes in a cost–consequences analysis.

4. To explore in an interview study the experiences of TPTNS from the perspectives of care home residents, family carers, care home staff (nurses, carers and senior carers), and care home managers.

Eligibility criteria

The care home residents were eligible for inclusion if they:

• had self- or staff-reported UI more than once per week

• used the toilet or a toilet aid for bladder emptying, with or without assistance

• wore absorbent pads to contain urine.

Exclusion criteria included care home residents:

• with an indwelling urinary catheter

• with a symptomatic urinary tract infection (UTI)

• with a post void residual urine volume (PVRU) > 300 ml

• with a cardiac pacemaker

• with epilepsy that was being treated

• with bilateral leg ulcers

• with pelvic cancer (current)

• with palliative care status

• who were non-English speakers.

Identification and informed consent

One person in each care home (usually the care home manager or a senior nurse) was assigned as the local principal investigator (PI) for that site. The PI was responsible for identifying potentially eligible participants in the care home based on the aforementioned criteria (see Report Supplementary Material 9). However, the process for recruitment differed based on the location of the care home, in accordance with statutory requirements on capacity to provide informed consent to participate in research in Scotland and England (see below). All consent procedures were approved by RECs with responsibility for reviewing applications for research involving adults who lack capacity.

Recruitment in Scotland

The PI compiled a log of potentially eligible residents in the care homes based on the trial eligibility criteria, then provided trial information to the resident if appropriate (depending on capacity), and sought permission for the trial team to contact them with the purpose of providing more detailed information about the trial and answering any questions. An aphasia-friendly version of the participant information sheet was made available when appropriate. This approach was made by the regional research assistant (RRA), who was a registered nurse working as part of the trial team. Under the Adults with Incapacity (Scotland) Act 2000,31 when a resident had a certificate of incapacity the local PI would identify and provide the study information to the resident’s welfare attorney (if one was appointed) or their nearest relative. If there was no welfare attorney identified or the resident did not have a relative who could be consulted then they were considered ineligible to participate in the study. The local PI would seek agreement from the welfare attorney/nearest relative for the RRA to speak to the resident. Following confirmation from the local PI of permission to approach the resident or welfare attorney/nearest relative, the RRA would arrange to meet them (or telephone them if a meeting was not possible) to provide a full explanation of the trial, ensure eligibility (see Report Supplementary Material 10) and seek written consent for the resident to participate. In accordance with the principles of the Adults with Incapacity (Scotland) Act 2000,31 verbal consent or assent was sought at every research contact (to ensure ongoing consent to participate) and welfare attorneys/nearest relatives could withdraw consent for the resident at any time.

Recruitment in England

In England, the PI compiled a log of potentially eligible residents and approached residents with capacity similar to the procedure described for Scotland. In accordance with the Mental Capacity Act (MCA) 2005,32 if the local PI believed a resident’s capacity was in question, they would identify and provide the information to the resident’s personal consultee (usually a family member or friend) or, if one was not available, a nominated consultee identified by the care home study team, and would seek their agreement for an approach from the RRA. The RRA then provided a full explanation of the study, ensured eligibility and sought the consultee’s advice on what they felt that the resident’s wishes would have been about taking part in the trial, if they had had capacity. The consultee would sign a declaration form if they believed that the resident would have chosen to agree to participate. Similar to Scotland, verbal consent or assent was sought at every research contact and consultees could withdraw consent for the resident at any time if they felt that the resident’s wishes had changed.

Recruitment of family carers and care home staff

In both Scotland and England, written consent to answer questions about residents was sought from family carers and care home staff by the RRAs. In addition, the qualitative researcher identified and approached a sample of residents, family members and staff to take part in the process evaluation interviews and focus groups. When not previously given, written consent was obtained prior to the interview taking part.

Regional research assistant training

The RRAs were trained by the chief investigator and trial managers in the following tasks: assessing eligibility, taking consent, collecting data, completing the case report form (CRF), recording adverse events (AEs)/change of status, randomisation procedures and using the trial database. This training was accompanied by a detailed standard operating procedure (SOP) for the RRAs. Each RRA underwent good clinical practice (GCP) training and 2-yearly updates as required.

Implementation support facilitator training

The implementation support facilitators (ISFs) received training from the chief investigator and trial managers on administering TPTNS in both the treatment and sham groups (to check correct use by care home staff), checking adherence and reporting AEs/change of status. A detailed SOP was developed for ISFs. The ISFs also underwent GCP training and 2-yearly updates as required.

Care home staff training

A bespoke training programme for care home staff on how to administer the intervention was developed and delivered to small groups in each home by the chief investigator and trial managers. The half-day theory and practice-based training course included:

• a background presentation, including the purpose of the study, participants’ individual roles, education about UI in care home residents, usual management strategies, common challenges to continence promotion, theory of TPTNS and how to implement it in care

• an informal discussion about what the staff found difficult and challenging around UI, what they believe ‘works’ when supporting people with dementia and how this would help implement TPTNS in usual-care pathways

• a practical demonstration of application, initiation and removal of TPTNS/sham intervention

• experiential learning, during which all staff were provided with an electrical stimulator to become familiar with the equipment and practise administering TPTNS/sham intervention to each other under supervision of the chief investigator and trial managers.

A SOP was developed for care home staff and a signed Certificate of Competence was awarded to staff who completed the ELECTRIC trial training and achieved competency in delivering the intervention.

Following the training session(s), the ISF worked with the care home staff in a period of facilitated support to build their confidence and help them implement their learning in practice. This was followed by an individual staff competency assessment during the first 1–2 weeks of the intervention, during which the ISF corrected any errors in delivery of the intervention (TPTNS or sham), provided any remedial training and issued a further Certificate of Competence.

Care home principal investigator training

The PI in each care home underwent similar training to care home staff (described above), with additional training on identifying potential participants and completing the site screening log, reporting AEs/change of status and storing trial paperwork. A SOP was developed to provide support to PIs in this role.

Training handbook and digital versatile disc

As part of the training programme, a training handbook (see Report Supplementary Material 3) and digital versatile disc (DVD) were developed to ensure that care home staff were knowledgeable and competent to administer the intervention. These resources were available electronically and were presented to staff during training.

Primary outcome

The primary outcome was the volume of urine leaked in 24 hours at 6 weeks post randomisation, measured using a 24-hour PWT. This test is based on the premise that 1 g of fluid weight is equal to 1 ml of urine and is thus an objective measure of urine leakage. To carry out a PWT, the participant (with the help of staff if required) emptied their bladder, applied a clean, dry pad at an agreed set time and then retained all used pads for the following 24 hours. Individual pads were sealed in a small plastic bag and then collected in a larger resealable bag to prevent evaporation. The RRA collected and weighed the large bag as close to the end of the 24-hour pad collection period as possible, and the dry weight of an equivalent number of pads to those collected was deducted from the total weight of the bag, to provide the 24-hour volume of urine passed as a result of UI.

Secondary outcomes

Qualitative data collection

See Chapter 5 for further details.

Original sample size justification

The original recruitment target was calculated to be 500 care home residents, based on a sample size of 344 needed to detect important clinical differences of 200 ml per 24 hours of urine leakage, with 90% power at the two-sided 5% alpha level, including an inflated attrition estimate of 30% to account for loss owing to death and other types of loss to follow-up. The standard deviation (SD) used in the original calculation came from a small, single-centre trial with a selected population in which the reported SD was 450 ml. 39 A 95% confidence interval (CI) was put around the SD estimate and the upper CI bound (570 ml) was used as a conservative measure for the sample size calculation to account for recruiting to a pragmatic multicentre trial.

Revised recruitment target

Following 1 year of recruitment to the ELECTRIC trial, a data cut was performed by CHaRT, University of Aberdeen, and the recruitment target was reviewed. Kieser and Friede40 recommend re-estimating the sample variance from observed data using the whole trial cohort and calculating the one-sample variance, and also an adjusted estimate to account for potential bias in the one-sample variance, under the alternative hypothesis. The required sample size was then recalculated without penalty to the type I error rate. The SDs using these two methods were 427 and 415, which indicated sample sizes of 194 and 184, respectively. Choosing the more conservative of these and applying the increase for 30% attrition gave a total of 278 participants randomised for 90% power to detect a 200 ml difference in the primary outcome. The observed attrition after 1 year of recruitment to the ELECTRIC trial was 15%, suggesting 278 participants to be a conservative upper bound estimate of the required sample size. However, to account for potential differences in variability and missingness of data at the beginning and end of the trial it was concluded that recruitment should continue for the duration of the planned recruitment period (18 months), with the aim of exceeding the minimum requirement of 278 randomised participants. Three blinded independent statisticians, the Data Monitoring and Ethics Committee (DMEC), the TSC, the PMG and the funders all agreed with the revision to the sample size and recruitment target.

Primary outcome

The primary outcome, total volume of urine leaked in 24 hours at 6 weeks post randomisation, was analysed using linear regression correcting for baseline 24-hour PWTs and the design covariates [severity of UI (mild/moderate/severe) and sex (male/female)]. Potential care home clustering was controlled for using a random-effects robust variance. In addition, the effects of adherence to treatment (duration, intensity and position) were explored using randomisation as the instrumental variable in a complier-average causal effect (CACE) model40 (using two-stage least squares). Compliance was determined for both the TPTNS and sham groups individually (see Figure 3). However, for the CACE analysis, compliance is defined as being compliant to the intervention (i.e. TPTNS). Consequently, in the modelling, compliance for the sham group is set to zero to reflect that the sham group is not compliant to TPTNS.

Secondary outcomes

Secondary outcomes were analysed using a similar strategy, but with models suitable for each outcome. In addition, these utilised all available follow-up data from all randomised participants using a standard time interaction model to incorporate repeated measures. These were estimated using generalised linear model linear regression for continuous data; although outcomes may be skewed, the use of baseline information as a covariate will satisfy the normal assumptions. For binary outcomes, Poisson regression models were used with a log-link function summarising the treatment effects as adjusted risk differences and adjusted relative risk ratios. All models were adjusted as described above.

All model assumptions were assessed by means of the summary statistics and/or graphical plots.

Patient-reported outcome measures

Patient-reported outcome measures (PROMs) collected using validated questionnaires were combined into an overall score and missing items were imputed using the strategies described in appendix B of the SAP (see Report Supplementary Material 8), and taking into account the level of missingness overall and within a person.

Subgroup analyses

Predefined subgroups were reported as the magnitude of the subgroup effect estimates along with their 95% CIs, and interpreted in an exploratory manner and broadly to provide recommendations for further investigations. Predefined subgroups were tested using interactions for the following:

• sex (male/female)

• UI severity [mild (0–200 ml per 24 hours); moderate (201–400 ml per 24 hours); severe (> 400 ml per 24 hours)]

• functional dependency –

  • total Barthel Index score

  • Barthel Index Mobility score (immobile, wheelchair independent, walks with help of one person, independent)

  • Barthel Toilet Use score (dependent; occasional accident; continent)

• clinical frailty scale – (≤ 5, 6, 7 or more)

• on anticholinergics for incontinence (or not)

• falls status in last 6 months

  • number residents who, at baseline, have fallen in the last 6 months [n (%)]

  • number of falls: ≤ 6 falls, > 6.

For identified post hoc subgroups, forest plots were used to illustrate possible effects of the following:

• size of the centre

• functional mobility

• restricted communication

• need for help to use the toilet

• number of helpers needed to use the toilet

• urinary urgency symptoms.

Economic evaluation

A cost–consequences analysis (CCA) approach was used to incorporate the range of costs and benefits of the TPTNS intervention (see Chapter 4). The conclusions of economic analysis were presented in a disaggregated ‘balance sheet’ form, presenting costs of providing the intervention alongside the key primary and secondary outcomes. Health-related quality of life was assessed using Dementia Quality of Life - utility index (DEMQOL-U) and Dementia Quality of Life Proxy - utility index (DEMQOL-PROXY-U) with a UK tariff. 41 Full details on how unit costs were attached to resource use are presented in Chapter 4.

Baseline characteristics

Table 2 summarises the participants’ general baseline characteristics. The two groups were comparable at baseline. Resident age ranged from 58 to 107 years and the mean age was 85 years (SD 8 years). The majority of residents in both groups were female (76.6% in the TPTNS group; 78.5% in the sham group) with moderate cognitive impairment [mean Mini Mental State Examination (MMSE) score 12.8 (SD 9.3) in the TPTNS group and 13.4 (SD 8.9) in the sham group]. Participants had similar levels of high physical dependency [mean Barthel score 7.2 (SD 3.9) in the TPTNS group and 7.9 (SD 4.0) in the sham group] and, in both groups, the majority were severely or very severely frail (54.3% in the TPTNS group; 51.2% in the sham group).

Table 3 summarises the residents’ continence status at baseline and shows that UI was a chronic condition, being present for at least 18 months in the TPTNS group and 20 months in the sham group. The majority of residents (59.4% in the TPTNS group; 55.5% in the sham group) experienced severe UI of > 400 ml leakage in 24 hours and all apart from one resident in the TPTNS group wore absorbent pads continuously to contain the leakage. The number of pads used in a 24-hour period was similar for both groups. The PVRUs were relatively small (approximately 80 ml, which is within normal limits), as was the number of residents taking anticholinergic medication to treat UI. The mean total volume of urine leaked in 24 hours (the primary outcome measure) at baseline was similar between the two groups (536 ml in the TPTNS group; 560 ml in the sham group) but the SDs indicate greater variability of leakage in the sham group (370 ml in the TPTNS group; 469 ml in the sham group).

Restricted access to independent toilet use as a result of mobility problems was found for more than three-quarters of residents (85% in the TPTNS group; 78% in the sham group) and communicating need to use the toilet was an issue for > 40% of residents in both groups (44% in the TPTNS group; 41% in the sham group). Functional skills for independent toilet use were similar in both groups for the resident-reported and the staff-reported MTSQ; however, staff reported more severe functional deficits than residents (scores range from 0 to 20, with higher scores indicating more difficulty). Ratings of PBC by each of the three respondent groups (residents, family carers and staff) were similar for the TPTNS and the sham groups; however, residents rated their PBC as less severe [mean score 2.4 (minor problems to some moderate problems)] than staff (mean score 3.3) and family carers [mean score 3.6 (some moderate problems to severe problems)].

Sensitivity analysis

Sensitivity analysis to assess the robustness of the treatment effect to missing data and non-compliance to the TPTNS gave similar results to the complete-case intention-to-treat (ITT) analysis.

Subgroup analysis

Figure 4 summarises the prespecified subgroup analyses and Figure 5 presents the post hoc subgroup analyses. There were no modifying effects on the treatment effect in a subgroup analysis.

FIGURE 4.

Forest plot of subgroup analysis for the primary outcome. Clinical frailty: ≤ 5, very fit to mildly frail; 6, moderately frail; ≥ 7, severely or very severely frail. ADL, activities of daily living.

FIGURE 5.

Forest plot of post hoc subgroup analysis for the primary outcome. Effect size of total urine leaked (after 6 weeks).

Table 6 describes total urine leakage over time. Treatment effect estimates were fairly consistent over time: incompatible with a worthwhile reduction in urine volume when receiving TPTNS. It should be noted that the treatment effect estimate at 6 weeks differs from the figure in Table 5, reflecting the longitudinal model. Secondary clinical outcomes are reported in Table 7; pad use and PVRU were similar between groups at all time points, confirming that TPTNS does not increase urinary retention in this population.

Resident- and staff-reported outcomes and treatment effect estimates are summarised in Table 8. Perception of bladder control was measured using the single item PBC (by residents, their family and care home staff). The rating was fairly constant over time (between minor and moderate problems) for residents and staff; CIs around treatment effects generally rule out any difference of more than one point on the item in either direction. The family ratings are few and should be interpreted with care.

The MTSQ scores range from 0 (no difficulties) to 20 (cannot do), rated by both residents and staff at the care home. For both outcomes at each time point there were small differences favouring the sham intervention.

Fewer than 40% of residents in each arm were able to complete the DEMQOL health-related quality-of-life questionnaire at baseline. The proxy version of the score was completed by a family member or a member of care home staff at both 6 and 18 weeks. There was a small difference (about 0.2 SD) in favour of the TPTNS treatment at 6 weeks.

Serious adverse events

There were four serious adverse events (SAEs) that required residents to be admitted to hospital: three in the sham group (one for a kidney infection, one for a chest infection and a UTI, and one for vomiting and irregular heart rhythm); and one in the TPTNS group (for a UTI and pulmonary oedema). All four events occurred during the 6-week intervention programme; none of the events was classified as being related to the trial interventions (see Appendix 4).

Non-serious adverse events

There were a total of 21 participants who had a non-SAE, mainly a UTI (15/21), three of whom also had a potential chest infection. All non-SAEs were judged to be unrelated to the intervention, other than two in the sham group: one participant withdrew because their leg felt ‘heavy’ and they felt dizzy; one participant became agitated and distressed, with no impact on intervention delivery as the contingency plan to offer the intervention at alternative times was implemented successfully (see Appendix 4).

Economic evaluation

A CCA approach was used. 42 This involves a descriptive presentation of the range of costs and benefits of an intervention. Outcomes are reported in natural units, such as the use of incontinence products or the impact on continence care pathways, which should resonate with those involved in the planning and delivery of continence care in care homes. It also enables costs and benefits to be reflected where there is a known direction of change, but insufficient data is available to quantify that change. The use of a ‘balance sheet’ reporting approach allows the reader to form their own opinion on the relevance and relative importance of the findings to their decision-making context. 42 Disaggregated costs and benefits were descriptively presented and outcomes reported in natural units.

Perspective

A public sector payer perspective (in the UK, this is the central government Treasury), specifically in terms of NHS costs and local authority costs for care homes, was used to provide a practical assessment of resource implications and consequences of interest for those involved in care home service provision and funding decisions.

Discounting

Reporting of economic analysis was in 2018–19 Great British pounds (GBP). The time horizon for the trial was 18 weeks. As costs did not extend beyond 1 year, discounting was not required.

Quality-of-life health outcomes

Data about quality of life were collected using the DEMQOL and DEMQOL-PROXY (titled DEMQOL-Carer) questionnaires. These are condition-specific measures of health-related quality of life designed specifically for use with individuals experiencing cognitive decline and dementia. 35,43 The raw scores from each of these can be used to generate health state utility values for use in economic evaluation through conversion to DEMQOL-U and DEMQOL-PROXY-U, respectively. 41,44 DEMQOL-U consists of five domains: positive emotion, memory, relationship, negative emotion and loneliness. Four possible levels of response relating to severity are available for each domain. DEMQOL-PROXY-U consists of four domains: positive emotion, memory, appearance and negative emotion. As with DEMQOL-U, each has four possible levels of response relating to severity. DEMQOL was administered at baseline and at the 6- and 18-week follow-ups, and completed by the participant. DEMQOL-PROXY was administered at baseline and at the 6- and 18-week follow-ups, and completed with a proxy resident perspective by a single, named proxy. This could be a different proxy at each time point. Individual raw scores were converted to health state utility values using the scoring algorithm for the UK. 41

Resource use

Resources to deliver the TPTNS intervention, details about the use of absorbent pads (number and type) and other equipment, and (if appropriate) the number of staff required to assist residents to use the toilet were measured. Data on resource use were collated from the RUQ designed for this study and from data collected as part of the main trial (see Report Supplementary Material 4–7). The RUQ consisted of three questions related to medication prescribed for incontinence, resources required for toilet assistance, and aids and devices for managing incontinence. In addition, at the 6- and at 18-week follow-ups, questions about the use of primary care in the preceding 6-week period were included. Information on all consultations with health-care professionals external to the care home as a result of their UI was categorised according to who the consultation was with [general practitioner (GP), nurse, physiotherapist, etc.] and where the consultation took place (in the surgery or in the care home). Stimulation diary records were used to determine the average time taken by care home staff to complete stimulations [over 12 sessions in a 6-week period (minutes)] and the staff grade of the stimulation administrator. This was recorded on the RUQ. The RUQ was administered at baseline to establish the usual continence care pathway and completed by the RRA. The RRA was a registered nurse working in trial regions in Scotland and England. The RUQ was used again at the 6- and 18-week follow-ups.

Unit costs

The unit costs and sources used to estimate total cost per participant are given in Appendix 5. Unit costs were attached to the individual resources identified in the RUQ. Unit costs were identified using Unit Costs of Health and Social Care 201937 for staff for primary care, and British National Formulary38 for prescribed medication. Incontinence products’ market prices were identified from supplier direct websites or large chain shops with an online presence. Direct costs of interest included expenditure on absorbent pads and expenditure on other protection products to manage UI.

Intervention costs

Intervention costs included staff training (training time using staff roles reported for delivery of TPTNS/sham intervention), the trainer and the materials (TPTNS electrical stimulator machines, skin electrodes, handbook and training DVD). Costs were based on the market rates for these items. The costs of developing the handbook and DVD were not included. Cost of training facilities was not included. (It was assumed that training would not be expected to be delivered off site in normal working practice.) Equipment to deliver the intervention [cost of TPTNS Neurotrac machine, consumables (skin electrodes, wipes, batteries)] were also based on market rates for these items. Time to deliver the intervention was costed using staff time.

Intervention costs

Seventy-two bespoke training courses were delivered to 425 care home staff at care home venues. Each staff member received an online handbook and DVD to support their skills development. A total of 148 participants also received a follow-up individual staff competency assessment in their workplace. The average cost of the training and support package per staff member was estimated to be £121.03, based on the assumption that normal practice would be to make use of local trainers [10-mile radius (travel time of 30 minutes and 20-mile round-trip mileage reimbursement per person)] and excluding economic cost of venue. Although this reflects trial costs for training conducted in care homes in Scotland, it should be noted that additional costs were incurred by travel and subsistence reimbursements for training conducted in care homes in England because distances to care homes were greater.

Delivery of the intervention required one staff member for each of the 12 times that the stimulation was delivered during the trial. An average salary of £38.06 per hour was used to value staff time. This was estimated from staff roles reported in trial stimulation diary records (66% care assistants, 13% care leaders, 20% nurses). Staff did not need to be present during stimulation. Data for time to set up and take off the machine for each stimulation event was available for 92% of stimulation events. Data were missing for 8% (345/4127) stimulation events. Time to set up and take off the machine took ≤ 5 minutes for 94% (3536/3782) of stimulations, with 80% of stimulations (3042/3782) taking ≤ 3 minutes. The cost of delivery of TPTNS (excluding training) in the trial was estimated to be £81.20 per participant (Table 10).

Dementia Quality of Life and Dementia Quality of Life Proxy

A total of 406 residents from 37 care homes in the UK (23 in Scotland and 14 in England) completed DEMQOL-U or proxy-completed DEMQOL-PROXY-U for at least one measurement point. Baseline utility scores could be calculated for DEMQOL-U for 141 (35%) participants, and for DEMQOL-PROXY-U for 397 (98%) participants (Table 11). At baseline, for 257 participants (126 in the TPTNS group and 131 in the sham group) only DEMQOL-PROXY-U data were available (see Appendix 6). There were 41 participants (15 in the TPTNS group and 26 in the sham group) with both DEMQOL-U and DEMQOL-PROXY-U data available at all time points. A larger range of mean utility values was observed for DEMQOL-PROXY (0.722–0.742) than for DEMQOL (0.790–0.803), as reported by the care home resident. Minimum and maximum utility values were reported for both tools at baseline (DEMQOL-U, 0.243–0.986 and DEMQOL-PROXY-U, 0.363–0.937). No significant difference was found between participants’ scores over time, or between treatment and control groups at any time point, for either DEMQOL-U or DEMQOL-PROXY-U (Table 11).

Resource use data

Baseline resource use assessment was completed for 406 participants, at the 6-week follow-up for 370 participants and at the 18-week follow-up for 321 participants (see Appendix 7). Assessment of trial data collected about products related to incontinence management indicated no significant differences between groups at any time point (Table 12). Three-quarters of participants used between 1 and 3 items in 24 hours at an average cost of £1.19 (SD £1.51) per participant.

At each time point, around 85% of participants required assistance from care home staff to use a toilet. Most residents needed the assistance of one or two members of staff and visited the toilet four or five times a day (Table 13). There was no evidence of a difference between the TPTNS group and the sham group that was practicably meaningful. Assuming 5 minutes per toilet visit, and a derived average hourly pay of £36.08 per staff member (based on unit costs for proportions of staff delivering intervention), the value of staff time to assist a resident to attend the toilet was estimated as £19.17 (SD £13.22) for the TPTNS group and £17.30 (SD £13.33) for the sham group (per 24-hour period) using all available data. No significant differences were observed between groups.

A total of 898 items (433 in the TPTNS group, 465 in the sham group) of special equipment provided as a result of incontinence were reported for daily use (890 items) and for infrequent use (eight items: four transfer aids, two toilet aids, one mobility aid and one incontinence protection for furniture). Daily use items were predominantly related to mobility aids (n = 402), transfer aids (n = 244) and toilet aids (n = 179). Across the treatment groups, mobility aids (≈ 40%), transfer aids (≈ 25%) and toilet aids (≈ 20%) accounted for similar proportions of special equipment required. Pressure mats (n = 20); sensor alarms (n = 20); falls protection items, such as crash mat or bed rails (n = 20); and incontinence protection for furniture (n = 5) were also reported. Costs were not estimated for infrequent use items as there were so few of these items.

Cost of health-care services used

Appointments with health services staff for continence problems were available from the resource use questionnaire at 6 weeks and 18 weeks (each covering the preceding 6 weeks). The number of NHS contacts is shown (see Appendix 8). Use of NHS services were reported for 54 participants (21 in the TPTNS group; 33 in the sham group). A total of six participants (all in the sham group) used services during both follow-up periods. No significant differences were observed between groups at the 6-week or 18-week follow-up. Average costs at the 6-week follow-up were £11.88 (95% CI £5.72 to £18.77) for TPTNS, compared with £19.40 (95% CI £11.62 to £27.93) for sham. At the 18-week follow-up, the average cost was £7.66 (95% CI £2.52 to £14.25) for the TPTNS group and £12.60 (95% CI £6.81 to £19.24) for the sham group. Statistical power to detect between-arm differences is very low, and the evidence available from this analysis may have been due to the small numbers of individuals and the low reported use of these services.

Out of 406 participants, 370 had at least one medication reported at baseline (181 in the TPTNS group; 189 in the sham group). Across the three RUQ time points, a total of 336 different medications were reported. This included seven medications used directly for treatment of UI reported for 32 participants (20 in the TPTNS group; 12 in the sham group): duloxetine (Cymbalta^®, Eli Lilly and Company) (n = 3), fesoterodine (Toviaz^®, Pfizer Inc., New York, NY USA) (n = 1), mirabegron (Betmiga^®, Astellas Pharma) (n = 7), oxybutynin (Ditropan^®, Janssen Pharmaceutica) (n = 7), solifenacin (Vesicare^®, Astellas Pharma) (n = 8), tolterodine (Detrusitol^®, Pfizer Inc) (n = 2) and trospium chloride (Regurin^®, Madaus Pharma) (n = 4) (Appendix 7). All were classed as ‘ongoing’ medication and there was very little change over the course of the 18 weeks of the study. No further analysis of cost of medications was undertaken.

Cost–consequences analysis

Table 14 presents a descriptive comparison of the costs and outcomes of TPTNS compared with usual continence care pathways. The balance sheet reports both the costs and benefits mix of qualitative and quantitative findings, including financial and non-financial outcomes, to provide a more representative reflection of the impact of TPTNS on care home residents. Different stakeholders will find different parts to be of more direct relevance. The economic analysis indicated that there was no significant change in resource use over time (pads, toileting assistance, health-care use) and there was no evidence of impact on health-related quality of life (as measured using DEMQOL-U/DEMQOL-PROXY-U). This was in line with the main findings of the study, which reported that TPTNS was not found to be clinically effective during the trial, and that toilet use routines and absorbent pad use did not change.

Participants

Analysis

The complex nature of the data generated by the ‘whole-team’ approach required use of an analysis framework that would support understanding and facilitate presentation of data in a format that would enable subsequent optimisation and operationalisation of the ELECTRIC trial.

As described above, the COM-B model of behaviour49 was used to inform development of the topic guides for care home staff to elicit understanding of current practice (behaviour) and identify mechanisms and strategies to promote and support behaviour change. The theoretical domains framework (TDF) (v2)50 comprises 14 domains (knowledge, skills, social/professional role and identity, beliefs about capabilities, optimism, beliefs about consequences, reinforcement, intentions, goals, memory, attention and decision processes, environmental context and resources, social influences, emotions and behavioural regulation), providing a comprehensive, theory-informed approach to identifying determinants of behaviour. The TDF aligns with the elements of the COM-B49,50 and is used in this study to structure the analytic framework. 50

Framework analysis allowed comparison of data at two levels, individual (interviews and focus groups) and global (across the entire data set), and enables generation of themes in and between cases. 51 The TDF was used explicitly to inform understanding of factors affecting implementation of TPTNS in a care home context. Inductive codes were also developed and used to explore the residents’ data.

Analysis was conducted following Richie and Spencer’s52 five phases of analysis:

1. familiarisation

2. identifying a thematic framework

3. indexing

4. charting

5. mapping and interpretation.

The QSR NVivo (version 11) (QSR International, Warrington, UK) data management and analysis software was used to support the analysis process. Individual frameworks were developed for each distinct set of data (residents/family members, staff, care home managers). Specific analytic intentions were associated with each data set derived from the qualitative study objectives and areas of questioning for each group of participants. All transcripts were summarised, coded and informed by key concepts. For the care home residents and family members, the framework was based on the questions about their experiences of UI, associated care practices and views about the ELECTRIC trial. For the staff the key concepts of the TDF formed the analytic framework, charted by the qualitative researcher (see Appendix 10). The range and diversity of the themes was mapped, followed by a process of interpretation in which patterns of association were investigated and possible reasons for these explored. Two reviewers (CoD and LM) independently analysed 52% of transcripts. Iterative rounds of discussion between ELECTRIC trial researchers (qualitative researcher, JB, ML and CoD) led to agreement on the coding matrix and analysis of the inductive themes. This iterative group approach enhanced reliability and reduced the risk of lone researcher bias. 53

Reporting the results

Seven TDF domains significantly captured the involvement of care home staff within the ELECTRIC trial: knowledge, skills, social/professional role and identity, beliefs about capabilities, beliefs about consequences, goals, and environmental context and resources. Themes were inductively derived from the populated TDF domains. The challenges experienced and facilitators of implementing TPTNS into routine practice and evidence of behaviour change required to achieve this are explored. Direct anonymised quotations are used to illustrate themes. To indicate the source of each quotation, a different respondent code is used, comprising the code for level of staff, care home identification number and order of recruitment to the study. For example, ‘RN, 26016’ denotes a RN from care home 260, recruitment number 16. Residents are identified by sex, the care home identification number and order of recruitment for example, ‘Female 18001’ denotes a female resident in care home 180, participant number 01. ‘FM’ before an identification number denotes a family member.

The results of the analysis are presented in two sections: the first addresses objective 1, and describes the views and experiences of care home residents, family members and care home staff on the TPTNS intervention. The second addresses objectives 2 and 3 conjointly. These data, although not directly aligned with the TDF, have been analysed thematically and integrated into this report of the findings to enhance understanding of factors affecting the intervention in the care home context, optimisation strategies, and perceptions and experiences of care home staff regarding participation in the ELECTRIC trial.

Recruitment

Participants (residents, family members and care home staff) were recruited from care homes (n = 23) participating in the ELECTRIC trial in Scotland (n = 15) and England (n = 8). Care homes were categorised as nursing, residential or local authority, that is owned and run by the local city council (Scotland only), and small (< 42 residents) or large (≥ 42 residents).

The participant sample results follow, and the framework analysis of interview and focus group data are then reported. Table 15 provides an overview of the total participation in interviews and focus groups. This includes participation by some people in more than one interview or focus group (see Participants).

Residents and family members

Residents and family members participated in one-to-one interviews; however, residents without capacity participated in ‘mixed capacity’ focus groups.

Care home staff

When small staff numbers and busy workloads compromised the release of more than one staff member at a time, care homes opted for individual interviews (n = 35) rather than the planned focus groups.

Care home managers

Care home managers usually opted to participate in face-to-face interviews (n = 14) at their care home rather than the planned telephone interviews. Some (n = 3) chose to take part with a member of staff who had first-hand knowledge of the running of the trial. One care home manager opted for a telephone interview. The findings of care home managers and all other care home staff focus groups and interviews were analysed as a complete data set, rather than separating the care home managers as had been originally planned. This accommodated the varying roles and responsibilities of care home staff.

Experiences and perceptions of residents and family members

Many of the residents could provide limited responses only. This was mainly because of decreased cognitive ability; even those deemed to have capacity sometimes had difficulty recalling the trial and/or receiving TPTNS. Others, however, were able to express their views and perceptions, and family members of residents with dementia shared opinions relating to their relatives.

The effect of urinary incontinence on residents’ health and well-being

Residents spoke openly about their UI problems. They described areas in which their quality of life was adversely affected through loss of dignity:

I had to ring for help, I didn’t much like that. It’s not what ladies do!

Female, 18001

Participants also described the adverse effect on their ability to socialise:

I can’t really say ‘Nip to the loo’, I can’t ‘nip to the loo’, I have to be there about 10 minutes [laughter] because I go and stop, and I think it’s finished. I might get up and then go ‘Oh no!’ and I have to go back again. I am conscious of me bladder all the time, it’s sore.

Female, 51025

It’s going to other people’s houses that bothers you, you are always frightened in case you get up and go to the toilet . . . in case you wet the settee or something.

Female, 44018

Family members of residents with dementia commented on the progression of their UI:

I think now she is at the stage where before she would have got up and tried to go to the toilet, she will just sit and let it come away from her.

Male FM, 22504

Now it [UI] doesn’t bother him because he is not really aware. I think he is aware when he does it because we have walked in many a time and he has pulled the pad out and done various things with it.

Female FM, 21501

Wearing pads

Residents spoke about wearing incontinence pads; it was an important aspect of UI for many. This participant had accepted wearing pads since moving into the care home:

As soon as you come into the home you wear pads, it’s the ‘done thing’.

Male, 43014

Another participant viewed wearing pads as a consequence of growing older:

Supposing I was somewhere I couldnae get to the toilet, I wouldn’t completely wet myself but I might dribble a bit, so it is a safety thing. I put it doon [down] to getting older.

Female, 35004

However, another participant did not completely understand why they had to wear pads:

I don’t know why they [care home staff] . . . why they do it [give her a pad]. I think sometimes I get a bit loose . . . me bladder. It just comes on all of a sudden and perhaps I am a bit too slow, so they put the pads on now. [Shrugs and sighs] I don’t mind. It’s alright.

Female, 38021

Residents’ impressions of the trial

Although residents could often reflect on their UI, recalling their initial reaction to the trial varied owing to cognitive ability. However, sometimes there was wholehearted acceptance, often accompanied by enthusiasm:

As soon as somebody mentioned [the trial] I said, ‘You can put my name down for that, I definitely want to be on it, I am adamant about it’.

Female, 51025

It feels great [to be part of a trial]. It feels you are helping other folk as well as yourself.

Female, 44017

There was some understanding of how TPTNS works:

They put electric lines on your heel and it travels up your bones and to your bladder and tightens up the muscles.

Female, 51023

However, other participants perceived they were primarily helping in a more general sense:

Well it is more to help you . . . and the staff to understand what it [UI] is all about.

Female, 44019

Not very much really other than that it chartered the urinary incontinence with placement of the electrodes, so you knew you were getting them in the right place.

Female, 14005

Family members’ experiences of the trial

Family members whose relatives had dementia appeared to be most concerned about potential distress:

Anything that doesn’t freak her out I am all for it. She can get nervous of the least wee [little] thing now. I suppose it is ‘the condition’ [dementia] that causes this.

Male FM, 22504

Reassurance and information giving was also important:

She [trial manager] talked about an ankle monitor, I couldn’t understand how that [TPTNS] could have anything to do with incontinence. But she assured me that it was part and parcel of the trial, that Mum wouldn’t be forced to wear it, it would depend how well she tolerated it. So that was good, they wanted to gather data to see whether this had any benefits.

Female FM, 21503

Ultimately, family members were interested in the outcome of the trial and whether or not there could be any benefit of TPTNS to UI:

Once the trial is over, I would like to know whether there’s been ‘Yes it has been successful’ or ‘No’ or ‘There’s other work taken from it’. Just a couple of lines just to keep in touch would be nice.

Female FM, 35506

Residents’ perceptions of transcutaneous posterior tibial nerve stimulation

None of the participating residents who recalled TPTNS viewed it as an inconvenience or annoyance. They recounted how they incorporated the sessions into their everyday life:

I was sitting watching my telly so it [TPTNS] wisnae [wasn’t] any inconvenience. You know I sat and watched my telly.

Female, 35004

Oh, I can’t remember how long [I had TPTNS for]. It never troubled me or anything, I just rested for that time. I had no problems at all.

Female, 38201

This resident perceived they should receive a stronger dose (they were in the sham group), because they sensed TPTNS was not having any effect. They showed a desire to keep going with treatment:

When they put the machines on up to [stimulation strength of] 4, I need it up higher to feel the tingle. I think I need perhaps a bit stronger [dose] and for a bit longer. It doesn’t matter [which group] I was in; I am going to do it for longer to see.

Female, sham group, 51025

Others perceived a positive effect on their UI status as a result of their TPTNS:

I think it has slowed down in some ways. I think I have slowed down from what it was at the beginning. I don’t have to think about having a wee as many times and stuff like that.

Male, TPTNS group, 18002

I am lasting longer; I am not going all the time. Sometimes I forget to go. I think ‘Oh gosh I better go,’ and I go. Because I have more control.

Female, TPTNS group, 14005

Although this participant was in the sham group, they perceived that their UI had improved:

But I feel it [UI] has got better. I think I have got more control of it; I can hold it a bit better. I feel it gives me more time to get to the toilet.

Female, sham group, 44018

A family member who was asked if there were any notable change in their mother’s UI status perceived that it was the CAs (not the RNs) who would notice any changes:

Have I noticed? I think that would have to be the carers that would need to tell you; they see to her personal [UI] care. I don’t know.

Female FM, 22505

The residents and family members showed an interest in the trial, although their knowledge was limited. For family members, interest was usually focused on ensuring that, owing to the cognitive status of their relative, they would come to no harm, rather than any particular interest in the trial per se. None of the residents spoke unfavourably about the trial or TPTNS; it did not interfere with their daily life, rather they were able to continue with their routine undisturbed, and some perceived that they had noticed a positive change in their UI status.

Experiences and perceptions of care home staff and managers

For the ELECTRIC trial to be successful, changes to continence care practices in care homes were required. These changes took the form of two additions to practice, together with the need to change current routine toileting activities, to reflect individual resident continence and voiding needs. The two additions were (1) the delivery of the TPTNS/sham intervention and (2) the implementation of 24-hour pad collections. Evidence of these changes in care home staff behaviour and the factors affecting the nature and extent of the changes in the particular context of care homes is provided, based on the underpinning TDF and COM-B model and subsequent thematic analysis. Data coding and analysis in terms of the TDF domains is shown in Appendix 11 and the thematic analysis is presented here, with the relevant TDF domains indicated in brackets as numbers between 1 and 14.

Gaining knowledge (theoretical domains framework 1 – knowledge)

The desire to gain new knowledge encompasses initial impressions of the trial and the TPTNS training, and perceptions of gaining new knowledge. Scepticism, reservation, enthusiasm and curiosity in relation to their new knowledge are described. How prepared care home staff were for the ELECTRIC trial varied between care homes, with some staff being well informed prior to the ELECTRIC trial training and others having no knowledge of the trial until they were allocated to attend the ELECTRIC trial training by the care home managers. Many care home managers chose not to attend the training and cascaded responsibility for the trial to others from the outset.

Initial impressions (theoretical domains framework 12 – social influences)

I wasn’t quite sure how it was going to work.

RN, 22808

Prior to commencement of the trial, care home managers were often well informed about its purpose and the processes it would involve, having attended external regional meetings and forums and from speaking to the trial managers and chief investigator:

Our clinical director is very keen on research and making things happen, and change and innovation within care homes. She actually introduced the ELECTRIC trial to the group.

Care home manager, 43913

And the [trial manager] that we discussed how we felt and that was quite good. But it was also very interesting to get that information.

Care home manager, 29914

This SCA received information about the trial directly from the care home manager and felt well informed prior to TPTNS training:

Well, [our care home manager] came to us and told us that this trial was coming on and that it was a nationwide thing to see if they can, not get rid of incontinence, but lessen the impact that it has on [residents] through . . . stimulating a nerve in the leg, and obviously all these things need to be tested.

SCA, 41725

However, the above participant was in the minority: most care home staff described being unprepared, with some having little or no knowledge about the trial until they were sent for TPTNS training, often having received a directive from management to attend without prior warning:

When I came to do the training, I had never heard of it before.

RN, 14816

I only heard about it actually at the training. When I was told I was going to the training, that was the first time I heard about it.

SocCA, 36621

Having attended the training, care home staff views were mainly very positive, and they described how they approached the training with an attitude of curiosity in anticipation of potential benefits to residents with UI:

I was curious about it as I had never heard of TNS [TPTNS] before . . . I thought it was a good idea.

SCA 34726

My initial reaction was ‘It would be absolutely amazing if we could find something to help the residents’.

CA, 51853

Learning about research

Looking at the broad picture of the trial.

Care home manager, 37915

The ELECTRIC trial training provided learning around the purpose and processes of the trial, and specific knowledge and skills training, including learning around UI and TPTNS. Responses to the two distinct elements of the training differed, with UI education and TPTNS skills being mentioned more frequently than the research elements. Staff appeared to have knowledge relating to understanding UI and delivery of TPTNS, but not necessarily of their input and contribution to a large trial and the importance of maintaining rigour.

However, some staff enjoyed being involved in the broader aspects of the trial. For instance, this SocCA appreciated being consulted about potential participants by the trial RRA:

I think having input into who went on the trial, certainly in my unit . . . the girl that came from the University [the RRA] she said to me ‘Who would be suitable, who wouldn’t be suitable?’ and helped me pick the residents that it was suitable for . . .

SocCA, 36622

Care home staff also referred to the professionalism of the trainers and some were able to appreciate that they were participating in a wider trial because of feedback from the trial staff about their contribution:

We were one of the first homes, we were almost one of the guinea pigs in it, and it was really good to get the feedback from both [chief investigator and RRA] that a lot of the stuff that we had kind of thought about and put forward as ideas was used for other homes, so that they kind of changed things. So that was actually good to actually be part of the beginning of it and also . . . you know, help the trial go further because of things that we had identified and changed.

Care home manager, 43913

It is always good even being part of research programmes – it is really interesting.

RN, 35815

And there was a sense of understanding of the wider purpose of the trial:

. . . the [trial] staff that came down and did the training were very friendly, very professional, and it was a very enjoyable session. All the staff from here who took part said they really felt they had learned a lot around incontinence and why the trial was going to take place.

Care home manager, 14905

Finally, the training gave a more in-depth insight to UI than previous training they had received:

. . . it’s opened . . . my eyes, because the continence training we had previously was how to apply the pad and how absorbent is the pad. This is something more, it’s extending the subject, so it was interesting.

SCA, 500744

Enthusiasm (theoretical domains framework 5)

I mean if it works, it is going to be fantastic.

SCA, 44638

Many care home staff were optimistic and hopeful about participating in the trial and the possible impact it could have on UI care. On completion of training, staff recalled how they felt enthused and motivated to put their training into practice:

I think what [ELECTRIC is] doing is absolutely wonderful because a lot of our time works around incontinence, if you like. And if you can cut that wee [small] part out, it is more time with the resident to give them more activities, more outdoor things. I just think if it works it will be absolutely wonderful.

CA, 43629

Providing underpinning knowledge that was so often lacking gave staff much needed encouragement:

The carers haven’t actually got that knowledge, and I think the training gave them the underpinning knowledge [they needed] . . . it gave them a boost and I think that’s really important.

Care home manager, 43913

Scepticism (theoretical domains framework 5)

There’s no way in a million years that will work!

SCA 50746

Some care home staff expressed scepticism about the concept of TPTNS after the training. Although they had been taught the principles and simplicity of TPTNS, some could not conceive how it could work and believed that it would be time-consuming:

I didnae [did not] believe it would work. I was a non-believer . . . I thought, ‘I don’t know how can a machine stop you from peeing? [urinating]’ I just couldnae [could not] fathom that oot [out], how a machine could stop you from peeing.

SocCA, 36624

I thought it was going to be an absolute pain, if I am honest, I thought we would be putting it on, and we would have to sit there for half an hour and keep talking the resident into keeping it on. I thought it was going to be really time consuming . . .

SCA, 22709

Often, once the care home staff had experienced delivery of TPTNS, these concerns were resolved:

But it turns out it’s not [time-consuming], and [the residents] all took it quite well.

SCA, 22709

Concern about involving people with dementia in research (theoretical domains framework 6 – beliefs about consequences)

I was kind of not sure whether I agreed with it, to be honest.

Care home manager, 29914

Trial training included discussions with staff about involving people with dementia in research and the reasons for the wide eligibility criteria for the ELECTRIC trial. This included discussion of whether or not it was ethical to not offer people with dementia the opportunity to take part in research from which they may benefit. The majority of staff were in favour of including people with dementia and those who voiced concerns were very much in the minority; no one declined to participate on these grounds:

I wasn’t that comfortable – but not as uncomfortable as not to take part.

Care home manager, 29914

However, it was important to acknowledge staff concerns:

I didn’t think it was a guid [good] thing. Because most of oor [our] residents have got dementia or whatever and I just thought, ‘I don’t get this at all.’ Because I am still in the opinion that these people [residents with dementia], they don’t even understand – do you know what I mean? So sorry I jist cannae [just cannot] get my heed [head] round it.

SCA, 450751

Ethically, I was struggling . . . yeah. There was always that part of me that was, ‘She doesn’t even really know what I am doing.’ I struggled with that a little bit.

Care home manager, 29914

Staff views and opinions of the training was generally positive. However, there was some evidence of ethics concerns and, with regards to trial processes, there was limited interest in the outcome measures; for example, interest was focused on learning how to deliver TPTNS.

Developing skills (theoretical domains framework 2)

Completing the task correctly

Just double checked we were doing it right.

SCA, 22709

Input from the ISFs was frequently commented on as valuable and helpful. Their role was multifaceted: they continued the TPTNS skills training in the practice arena and carried out competency assessments. They encouraged staff and provided support, instilling confidence where individuals had concerns about ‘getting it right’ or ‘doing it wrong’ (see Becoming confident below). Staff described the TPTNS competency training as a check to ensure that they were completing the task correctly:

. . . [the ISF] came back in to see if we were doing it right . . . and it was fine yeah.

SCA, 500745

However, although ISFs were able to oversee intervention delivery and adherence checking, they were not involved in trial data collection. Responsibility for many research aspects of the trial (e.g. completing outcomes measures accurately, pad collections, bladder diaries) fell to care home staff and the success of these activities was often dependent on professional roles, responsibilities, and care culture within the individual homes.

The importance of leadership (theoretical domains framework 3)

The complexities of the care home setting, the various grades of staff, care home culture and the importance of leadership all had an impact on staff attitudes towards UI and UI care, efficient running of the trial and, ultimately, on whether or not change in practice could take place. Care homes varied in the way they organised trial-related work. Some appointed a member of care home staff as a ‘champion’, who was wholly responsible for organising trial activity (n = 8), or the care home manager (n = 3) assumed the ‘champion’ role. Other nursing homes (n = 4) afforded shared responsibility to SCAs and CAs, with limited input from RNs. In residential care homes and local authority care homes CAs (n = 8) were sometimes completely responsible for the trial activities. Care home managers’ roles ranged from being remote to wholly engaged, enthusiastic and ‘hands on’.

Strong leadership: working as a team

We are working together to benefit the residents.

SCA, 500744

Strong leadership that actively encouraged teamwork was identified as a key factor in running the trial successfully within the care home setting. There were care homes who described a cohesive team approach, with staff of all levels being involved in the trial. Staff described a welcome lack of ‘rank’, with all care home staff being aware of the trial:

I think it is good as well that there’s been no . . . rank difference, so it is anybody from [care home manager] down to the care assistants and nurses, everybody has been involved in it.

RN, 14816

The staff are all aware that this trial has to take place and what we are doing. It is not just us as nurse assistants, [name] as a senior is doing it . . . Everybody is aware.

SCA, 41725

Care homes adopting this team approach frequently appointed a ‘champion’ who took responsibility for organising the smooth running of the trial. As this champion was often a respected member of the staff and a good communicator, this approach resulted in good teamwork. This ‘continence champion’ within a larger care home gave an encouraging response:

I am the continence champion here. [The trial chief investigator] came and spoke to me, the care home manager and deputy care home manager, you could see how excited she was about it as well, and that made me excited . . . and later we [care home staff] all sat together . . . discussing things and how this could make a difference.

RN, 22808

Sometimes, in smaller homes, the care home managers oversaw the trial work and this was readily accepted by the staff, who were clearly used to strong leadership:

The manager oversaw it all so . . . when the team were coming and saying, ‘Oh we need so many [TPTNS] doing and we have got a little behind with this . . .,’ whoever was on duty, we just rolled it out. It was very much a team effort which is really, really good.

RN, 14816

And this care home manager, also in a smaller care home, related an important lesson learnt when they initially tried a team-focused approach that did not work and they subsequently felt compelled to take charge:

I completely controlled it [laughter] . . . initially we kind of said, ‘Right these are your residents, you work out when your treatments are’. But actually, it wasn’t working, it needed more control. So, I think that one of the learnings for me was that you really need somebody to lead it, you really need a champion.

Care home manager, 43913

Enjoying the trial through feeling valued

I just sort of kind of blossomed through it!

RN, 50847

Care home staff described a sense of feeling valued not only by management but also through being part of the ELECTRIC trial, which emerged as a key factor in staff investment in the research. Differences between care homes were described. Those that adopted a cohesive team approach and/or had appointed a ‘champion’ comprised staff who perceived they were valued, and felt themselves and their care home to be contributing to a worthwhile project:

I would say that the staff who were trained, really had a feel-good factor and it made them feel very valued and that it was an important job that they were doing.

Care home manager, 21903

Absence of support or leadership

The most they will do is they will write in the diary your name and ‘ELECTRIC trial’.

SocCA, 36624

Although strong leadership was evident in some care homes and was associated with positive experiences of the trial, there were care homes in which leadership was weak or virtually absent and this resulted in more negative experiences. Here, staff described managers that were remote and uncommunicative. And, although staff had enjoyed the training and demonstrated a positive attitude towards the trial, they struggled under demanding workloads to maintain trial-related responsibilities. Staff described feeling that they were part of a ‘tick-box’ exercise and that there was lack of concern for trial processes or for their engagement in the trial from managers who communicated by telephone or through checklists in a diary only:

When they [‘seniors’] phone for the report and you get, ‘. . . and your ELECTRIC trial, did you do them [the TPTNS in the diary for that day]?’ ‘Yes.’ It is like a tick box . . . it is a tick box kind of thing for them.

SocCA, 36622

This staff member went on to recount how their senior, who was rarely involved in the trial activities, had delivered the wrong TPTNS using a dose appropriate for someone randomised to the TPTNS group to a resident randomised to the sham group (no harm came to this resident):

She had it [TPTNS] [and had documented as such] on the wrong bit of the leg [laughter] and had it up to 34 [a therapeutic stimulation strength, not sham] [laughter]. So, the Seniors no, they’re not involved enough.

SocCA, 36622

One SocCA had little or no contact with their senior, was never asked how they was getting on or whether or not they needed support in the trial and was left largely unsupervised:

No, he has no’ [not] asked me. Our senior . . . doesn’t work my shift pattern so . . . I very rarely see him.

SocCA, 44637

Whose role is urinary incontinence care? (theoretical domains framework 3)

Delivery of UI care was usually seen as the role of CAs and SCAs. In nursing homes, RNs were viewed as being too busy with other responsibilities, one of the reasons SCAs and CAs were more involved in the ELECTRIC trial than originally anticipated:

It is usually just us like, yeah just us.

CA, 14618

As previously described, in the nursing homes senior staff (RNs and care home managers) were not as involved in the trial as anticipated, one reason being that they did not view UI care as part of their role. In many care homes, senior staff described delegating UI care, describing CAs as the main UI caregivers. This RN was asked whether or not they had noticed any change in residents’ UI status since starting TPTNS:

As I say . . . we maybe don’t notice it [UI care] day-to-day because obviously we are not as ‘hands on’ . . . as the girls are.

RN, 26832

And this care home manager:

I think probably . . . the carers benefited most [from participating in the trial] because they are the ones that are doing the . . . continence work.

Care home manager, 45909

This CA corroborated that understanding:

[junior members of staff] . . . [are] the ones that can do it [UI care] and sometimes the seniors [SCAs] as well, but then not so much the nurses . . . they have to do their own thing [laughter].

CA, 14618

This care home manager commented that in their care home the RNs were not enthusiastic about participating in the trial, and it was agreed that the SCAs and CAs should mainly carry out trial activities:

The nurses heaved a sigh of relief . . . ‘Oh, I can go and get so and so to do it?’ [laughter] . . . we didn’t see the point in it only being nurses, you know, because it wasn’t something that particularly needed to be done by the nurses.

Care home manager, 22907

They went on to say that being part of the trial was a good thing for the care staff because it gave them a sense of self-worth:

But then as they [RNs] were able to pass that duty down to the care staff I think it gave the care staff a sense of self-worth that they were able to do it.

Care home manager, 22907

Disinclined to participate in the trial

The care staff did more of the trial because I was just doing other things.

RN, 21806

The role of UI care frequently fell to junior care home staff. Often, if SCAs and CAs took on UI care, it seemed to follow that they would also be responsible for the trial activities. This RN, a ‘champion’, comments that although they appreciated that other RNs had many other tasks to carry out, they also sensed that they preferred not to be involved:

It was actually the nurses were not too sure. We had senior carers and nurses trained up, and the senior carers seem to just run with it, maybe because they are ‘at the coal face’ [carrying out UI care] more than the nurses, but they seem to just grab the idea and run with it. But the nurses . . . it was just another added thing onto their role, it was another thing to do.

RN, 50847

This SCA was quite clear about who engaged with the trial:

I haven’t seen one nurse do it. I don’t really know I have never really heard any of them [RNs] say anything about [the trial]. They are always too busy [with] meds and dressings and stuff like that but . . . pretty much yes, it’s down to the care staff.

SCA, 18702

Clearly, junior staff were frequently responsible for most of the trial-related work, including co-ordinating TPTNS, completing pad collections and maintaining bladder diaries within their routine practice. The impact of this varied; some junior staff were able to cope well whereas others found it stressful.

In summary, most care homes perceived UI care as the role of junior staff (CAs and SCAs) and senior staff were better placed continuing with their normal work, allowing the CAs/SCAs to manage the trial activity. However, although TPTNS training was well received and implementation was efficiently monitored by the ISFs, junior staff may have required more support in managing the other trial activities (see Making the trial work).

Becoming confident (theoretical domains framework 4)

An important feature of participating in the trial from the perspective of care staff was their new learning, including knowledge about UI and skills to implement TPTNS, but their confidence to implement this new learning in practice was also essential. Engaging with the ELECTRIC trial was seen as ‘more than just learning a new task’ and:

It’s not just a physical thing, it is also a mental thing.

SCA, 51752

Some staff admitted to feeling somewhat uncertain moving forward in delivering the intervention. This care home manager recognised the important role training had in teaching staff the right skills and enabling them to have confidence in delivering the intervention:

Obviously, it was a wee [little] bit fearful at first because you have always got this thing, ‘Oh, what if I do something wrong?’ And to actually have the training and be reassured that basically, you know, that the staff are comfortable and confident using this, that was probably the best thing.

Care home manager, 35917

The ISFs played a key role in instilling confidence and reinforcing new skills, being available for care home staff as they implemented TPTNS into everyday practice. Many care home staff had not learnt a new skill for some time, nor had they participated in research before. Having support from the ISFs reinforced the learning from the training sessions and ensured that staff were competent and confident to deliver the intervention. The ISFs were often present in the care home during that time, ready with advice and guidance:

Because sometimes you worry, and you are not sure, ‘Is that right? Was it three fingers I was doing or one?’ So it was good to have somebody come in and go, ‘No’ and put us straight. So that was good there was a wee [little] bit of reassurance so that we got it right. I appreciated that [laughter].

SCA, 22709

. . . [the ISF] would say, ‘You are very good, this is great, you have got it all right.’ It was good, we were well supported.

Care home manager, 450748

The training and subsequent support from trial staff visiting the care homes gave care home staff the confidence to put new learning into practice and engage in TPTNS/sham provision.

This SCA had also been able to put their learning into practice, recognising that this helped not only the resident, but also their own practice, knowing that they was providing effective treatment for UI:

I have learned. . . that a wee [little] lady gets ‘functionally incontinent’, so [that] is helping me. [Residents] get really distressed because they still know they need to go to the toilet; they just can’t locate where the toilet is. So now that you can pre-empt that happening, it’s much better for them and for us, you are stopping somebody from being incontinent in a corridor, their dignity is not compromised . . .

SocCA, 36622

Potential consequences of the ELECTRIC trial (theoretical domains framework 6)

Care home staff views about what might be the outcomes of engaging with the ELECTRIC trial were varied and seemed to depend on the level of reflection of each particular staff member following training.

Potential consequences were both negative and positive, particularly with regard to attitudes towards UI care in a care home context.

Proactive treatment for urinary incontinence

There’s some hope out there; there’s something that will help, because UI has never been a high priority. You know ‘I have got this and this is me for life.’ But you know that’s maybe not the case then?

Care home manager, 26911

Well because there isn’t anything out there for a start other than just popping a pad on them. Is there? You know.

Care home manager, 26911

This is the first I have ever heard of any kind of continence trials or anything like that and like I say I have been nursing for years and I have never heard of anything like this. So this is why again I jumped at the chance of being involved in it.

SCA, 510752

Care home staff commented on the training and the impact their learning had on their UI practice. This care home manager talked about the key message and the change in practice they had implemented:

It is not the fact that ‘I am incontinent there’s nothing I can do’ it is the fact that ‘Yes you are incontinent but we are going to try and help you, we are going to try and do something else’.

Care home manager, 14905

If the residents ask for the toilet themselves – brilliant, go to the toilet and empty their bladder. Less with pads, less with clothing, less work to change the whole resident. So it’s working for them and for us as well.

SCA, 50744

A positive impact on some residents’ UI status was perceived following TPTNS, and this RN describes the ‘light bulb moment’ when CAs realised the change in one resident:

The staff were coming to me and going, ‘There’s a problem with this lady, her pads are not wet, her pads are just not wet!’ and I said, ‘Are you taking her to the toilet, is she actually passing urine in the toilet?’ ‘Yes, but her pads are not wet,’ and I said, ‘Well, why do you think that might be?’ and they were like, ‘OH!’ [laughter], like a light bulb moment for them.

RN, 50847

Care home staff acknowledged that improving residents’ UI status would have an impact on their quality of life, both socially and psychologically. This RN identifies the potential effects on residents’ psychological well-being:

Reduced anxiety for them . . . because I think that is what I see in a lot of the residents, is that anxiety about going to the toilet or needing to go to the toilet.

RN, 22808

There were also accounts of residents having an improvement socially, such as this resident who was able to go on holiday and be less focused on their UI:

She actually went on holiday just after [TPTNS] and we did notice a change in her going to the bathroom and that as well. Whereas before she would be more toilet focused but she didn’t have to be. It meant that she could go on holiday and enjoy it a wee [little] bit better.

Care home manager, 35917

Finally, this RN summed up their views that this research could help residents return to being able to manage their own UI care:

I thought it was great, yes. I think that anything that can give them that wee [little] bit of self-respect and dignity back is always worth researching and giving them that hope that, ‘Yes, we can help you with this wee [little] thing.’ That they can self-manage their personal care again. It is a small thing, but it is a big issue in their lives.

RN, 35815

Person-centred changes to practice

Let’s have an improved quality of life instead of ‘Just let’s do toileting’.

RN, 44733

Initiating a move from task orientation to person-centred care is summarised by this care home manager, who had noticed positive changes in residents’ UI status:

Now we are trying to get away from task orientation and [TPTNS] is helping because the residents who were badly incontinent . . . we know when they need to go. So, they are not getting taken to the toilet because it’s time you know, it’s part of the task. They are going because it is time for them to go.

Care home manager, 45909

The knowledge gained from the training had brought about change in practice to a more person-centred approach according to this care home manager:

Even if it means that you might still be incontinent but more aware . . . so we can . . . take people to the toilet at specific times. Or they may have a bit more warning so that they can say, ‘I think I need to go to the toilet now’.

Care home manager, 14905

Making the trial work in care homes (theoretical domains framework 11, theoretical domains framework 12)

In terms of the trial-related activities the smooth running of the research varied across care homes, with some reporting an easy adjustment and others experiencing difficulties:

You just organise yourself at the end of the day.

SCA, 51752

Care homes that were well organised, with a ‘champion’ and/or strong team approach, managed day-to-day trial-related activities well. This was not only delivering the TPTNS, but also completing pad collections, maintaining bladder diaries and assigning trial tasks to staff:

I used to put it in the diary, ‘Tuesday, the ELECTRIC trial day’. You know, you just organise yourself and then you don’t get stressed. And if you did have a busy day on that day, you just think, ‘Well, that’s OK, I will just do it tomorrow then.’ It was never stressful for me and it wasn’t a big deal.

SCA, 51752, appointed ‘champion’

Other staff were left to organise their own work, without supervision:

My thing then was finding the time because I had five [residents] on it. So, it’s quite a lot. But I found if I did them all at the one time it was so much better.

SocCA, 36622

Communication breakdowns regularly occurred, which could cause distress and make the organisation of the trial activities challenging for staff. For example, a failure to pass on information in the diary about starting stimulation undermined this socCA’s confidence:

We have got three shift patterns . . .and it [communication] sort of broke down with the other shift, it jist didnae [just did not] seem to follow through . . . I usually follow this other team, I don’t follow [name’s] team . . . so this is where it wisnae [was not] left . . . there were no notes [in the diary] . . .

SocCA, 44637

Perceived ‘excuse-making’ was a frustration for other CAs:

They say, ‘Oh, I have never done it,’ [TPTNS] and I am like, ‘Well why? Why? There’s no excuse! Why?’ ‘I am too busy’. But . . . please excuse my French, but it is crap because all it takes is one minute to put on and then they can walk away for half an hour!

CA, 18601

Prioritising trial outcomes measures: the role of good leadership and teamwork

This is important, it is important, and it could have an impact.

Care home manager, 50918

The organisation of trial-related tasks worked well with good leadership and the engagement of the whole team. This care home manager recognised the importance of accurate recording of research data and successfully conveyed this to staff:

The staff have got involved with the pad changes, including the staff on nights. . . . the important thing was recording on the charts, you have got to record everything and not miss anything out. And . . . I am proud to say, I think there has only been one session that was missed.

Care home manager, 51904

And some care home staff organised efficient pad collections:

We were remembering [pad collections] because we were putting them in the residents’ rooms. The bags were in their rooms, so it was a case of when you took [a pad] off the certain person you went straight to their room, so we all knew how to do that.

CA, 150643

In some care homes, staff worked in partnership with residents who had capacity, so that a small number were able to self-manage their pad collections:

The staff tell me when . . . aye, they leave me a bag and I put the pads in.

Female, 35004

I do a check sheet and save my front pads.

Male, 18002

This resident corrected staff when they almost used the wrong pad collection bag:

The first time we lost the bag. This time they wanted to put the pad in a yellow bag. I said, ‘But they wanted it in an orange bag.’ . . . [laughter]!’

Female, 18001

Organisation of the trial also included recording bladder diaries, and again, in some care homes, this was conducted efficiently:

. . . [The bladder diaries] are all in their little packets, so the residents all have their own little packets that go into the [bed]room and then we used to sit and write it in there with them.

CA, 14618

We got really good feedback from [the ISF] that the bladder diaries were perfect [laughter].

Care home manager, 29914

In care homes where organisation and leadership were less clear, uncertainty with bladder diaries and pad collection processes was apparent:

I don’t know about that [bladder diaries]. No, I think . . . were they just popped in the, em . . . that’s interesting, I don’t know then?

RN, 35815

This SocCA, who delivered TPTNS and was responsible for the trial in their care home section, had not heard of the bladder diaries either:

No we have not done that. We don’t do that.

SocCA, 44638

Bladder diaries were sometimes difficult to maintain accurately because they were moved from room to room and residents would go to the toilet without the care home staffs’ knowledge:

The [bladder] diary is more of a nightmare because it can be in like a dining room because we are all having to take them back and forward. Trying to keep track because a couple [of residents] obviously go [to the toilet] themselves and then trying to remember all the ones that [have bladder diaries].

CA, 15643

Uncertainty as to who was responsible for completing the bladder diaries could occur:

I think the care staff probably did the bulk of them [bladder diaries]. Obviously, they can go back and ask certain ones but it’s the [care] staff did them.

RN, 26832

And sometimes the bladder diaries were completed all at one time:

Yeah I jist [just] sat and done it . . . we used to bring the packs out and then jist [just] sit and fill them in . . . jist [just] sit and fill it all in and that was that.

Social care worker, 44735

However, the method of completing the diaries ‘all at once’ described in the quotation above was contrary to the directions for bladder diary completion, which were designed to capture the individual’s information at the time it occurred to accurately reflect toileting activities and times for each trial participant.

It was often difficult to record the bladder diary accurately when residents used the toilet independently and may leave the care home for periods of the day:

It impacted . . . it is a wee bit time consuming the bladder diary, and plus the difficulty with that was two ladies who have capacity, they toilet independently. So you didn’t want to interfere with their daily routine so it’s very difficult to get a true reflection recorded. Also her family take her out for the best part of the day. It wasn’t a true record of how she was managing and unfortunately she is the lady who is [in the TPTNS group].

Social care worker, 44734

Recording outcome measures could be problematic

Probably the most onerous part for us was the pad weighs – we tried our best.

Care home manager, 21903

Some care homes found organising the trial difficult, with the most commonly recounted problems being forgetting and/or throwing away pads during collection times:

I have to confess I have thrown a couple of the pads out; I totally forgot.

SocCA, 36624

The stress for me was just when we were doing the pad collections in case there were any mistakes or people were forgetting . . .

Care home manager, 31910

Senior staff were sometimes aware that junior staff had thrown pads away:

Sometimes when you go to the toilet with somebody the pad goes straight into the bin [rather than being saved for pad collection]. Sometimes the carers might have forgotten to collect the pads . . .

RN, 21807

Staff also described how residents could be unpredictable and either throw pads away or hide them, making accurate pad collections problematic:

The hardest part was just the pad collection, because you never know when residents are going to take a pad off or where they are going to put it. So keeping track of all the people who were [in ELECTRIC] was . . . interesting.

SCA, 37740

There are a few times we have gone in [the resident’s room] and the pads have been removed in the most obscure places, so it’s obvious that he has removed them himself. So that’s been quite frustrating, in the fact he doesn’t want to take part, he will say, ‘Yes’ and then he will sneak the pad off. So, we have not really got fantastic readings.

SCA, 41725

Other factors also disrupted smooth running of pad collections, such as having to rely on agency staff:

Obviously if you have last minute sickness and you were doing a pad collection and you have agency staff on, and they don’t know . . . or they are not as aware, they will just whizz it [the pad] in the bin . . . it kind of put us on the back foot a little bit, I felt.

Care home manager, 37915

Implementing essential aspects of the research (e.g. completing outcomes measures), although managed efficiently in some care homes, was problematic for others. Challenging, sometimes complex, care home working conditions meant that research activities proved onerous, often because staff forgot, or were not told, when to begin pad collections/bladder diaries, or were not sufficiently aware of the importance of outcomes measures and focused their efforts on effective delivery of the TPTNS, to the detriment of the outcomes collection activities. The heavy workload, and sometimes a lack of support, confounded the efficacy of the research for some, but were dealt with more easily by other care homes.

Impact of the research on staff time and resources

The trial has become part of our routine; you quickly adapt to routine.

SCA, 41725

Care homes that adapted well to the trial had good leadership and were well organised, with staff who understood their roles and tended not to feel compromised by conflicting priorities. Staff expressed a willingness to participate and did not feel under pressure of time:

Over time the staff got more confident with the [TPTNS] and it didn’t take them as long, and the same with pad collections, they were a bit . . . time consuming but staff were making sure that they had the right pads, the right times and things. But we got more confident with it over time.

RN, 22808

No, it’s not been stressful in the slightest. It’s all just quick to do and [TPTNS] is easy obviously because you can leave the machine sitting with [residents] for half an hour and then just go back and take it off.

SCA, 50746

Residents also commented that the trial did not impact on their daily routine:

I can’t remember much about it quite honestly, it [TPTNS] was just a routine thing we had to do.

Female, 34008

However, some staff members described finding difficulty in coping with their heavy workload:

One thing that went against us was because we are so busy in here and we only have two staff. And some of the residents who were in the trial . . . maybe couldn’t be left alone for the half an hour, so finding the time has been the main issue.

SocCA, 44638

The same member of staff went on to recount how they had missed a TPTNS session:

I was under pressure because I thought, ‘I need to get this done before the time is up,’ but . . . I felt guilty about missing it, but there was really nothing I could have done.

SocCA, 44638

Being the only member of staff who had done the training became burdensome for this SCA:

I will try and get the night staff to do it, like, ‘Can you do this?’ and it’s like, ‘Well, we have not had the training.’ That [lack of other staff being trained] then does put pressure on me to try and get [TPTNS] on then and I will be leaving work late.

SCA, 18702

Although not the only member of staff who had been trained, being solely responsible for delivering the intervention during a shift was perceived as challenging to the point that intervention delivery could not be completed. The practice of non-completion because of lack of time and potential deferment to the following day was context dependent, being acceptable in some care homes and not others:

At times I have been like, ‘Oh I have got this bloody ELECTRIC trial to do!’ [laughter] steam coming out my ears. Oh, the residents are all going off their nuts and I am like, ‘Oh, my good God in Govan! 7 o’clock at night I am running about sticking pads to people’s feet!’ [laughter]. The most management will do is write in the diary . . . but I have just been scoring through it and saying, ‘Too busy; far too busy!’

SocCA, 35613

Potential benefits to care home resources

Although care home managers observed that the most likely positive impact on resources would be a reduction in the quantity of incontinence aids/pads used, they did not anticipate reducing their orders in the near future:

Yes, less [sic] pads, I would think so, yeah . . . And maybe [reduced] pad absorbencies . . . as well, in the future.

Care home manager, 50918

We daren’t reduce the amount of continence aids or pads that we get from the company because trying to get them increased again – it’s like gold dust.

Care home manager, 21903

Another care home manager identified laundry as another resource impact, as residents who were perceived to have benefited from TPTNS were not requiring such frequent changes of clothing:

It’s affected the laundry staff as well because they are not going through as many clothes for the residents . . . one of the [residents], it could have been twice or three times a day – a full set of clothes, and that’s reduced completely.

Care home manager, 45909

What happens next? (theoretical domains framework 9 – goals)

Care home staff spoke about their intention to carry on with TPTNS treatment after the trial finished. They specifically identified residents who had been in the sham group, initially refused to participate or not been able to provide consent in time for the trial:

[Name of resident] is always trotting to the loo and she refused to go on the trial. But I think she’s got most of her faculties and she will be looking to see how the others get on and then she will come forward and it’s good to have a machine because then we can try it on her.

Care home manager, 51904

For residents who were allocated to the sham group, staff posited that offering them TPTNS would lessen the disappointment of discovering they were not in the treatment group:

She might be a bit disappointed when I tell her that it was the sham that she was on, but I think that, you know, when she was so excited about trying it anyway, I think she would be quite keen to try . . .

RN, 26832

Did care homes continue with transcutaneous posterior tibial nerve stimulation?

Care homes that had received Neurotrac machines following trial completion gave a mixed response. One care home manager who had been enthusiastic about participating in the trial admitted that the machines had not been used since the trial had finished, and that they felt ‘a bit disappointed in myself . . . that I didn’t continue with it’ (care home manager, 43913):

Do you know that way, we did the trial, it’s done, it’s ticked? We have got two machines which are probably [still] in the box. We shouldn’t have seen it as a beginning and an end, we should have seen it as a continuing thing, and we didn’t. I think we need to actually think about this, and actually go and try . . . even if it’s two residents that we identify.

Care home manager, 43913

However, another care home had continued for a number of weeks after the trial finished:

We had amazing results from the trial. We are using the machines on a couple of residents that have got capacity, but we haven’t had much of a success with them yet though, but we will keep going.

Care home manager, 45909

Resident-related factors

Not unexpectedly for a care home population, our participants were highly dependent (mean Barthel Index score 7.6) and more than half were severely or very severely frail, with moderate to severe dementia as indicated by the mean MMSE score of 13 (12–20 indicates moderate dementia; < 12 indicates severe dementia). More than 80% of participants had restricted access to the toilet because of mobility reasons and 42% were unable to communicate their need to use the toilet. Thus, these participants were almost exclusively dependent on staff for their toileting and continence care and were unlikely to initiate toileting, as indicated in the qualitative data in which regular toileting routines, rather than individualised routines, were highlighted as the norm for this care context.

The mean number of toilet visits per resident for the total sample across all time points was 4.15 (SD 2.68) in 24 hours (CCA data), and Figures 6 and 7 show that, for both groups, there was no change in routine practice in the care homes regarding toileting throughout the study. Thus, even if residents had experienced any reduced sensation of urinary urgency and increased warning time for the need to void (which should occur if TPTNS is effective), no additional toilet visits were provided to enable residents to void in a toilet, as opposed to using the absorbent pads to void. Without associated changes to toileting practices, any effects produced by TPTNS would be unlikely to be recognised. Furthermore, fewer than five toilet visits per day is insufficient for frail older adults to maintain independent continence, as a normal adult voiding frequency ranges between six and eight voids per 24 hours. 63 Participants would thus be forced to void in their pads on at least some occasions.

This situation contrasts markedly with participants in all other TPTNS studies, who are fully independent regarding their toilet use and continence management and can therefore adjust to any changes in their bladder sensation and function accordingly, without having to rely on anyone else for support.

In addition, participants in other TPTNS studies have full mental capacity and cognitive abilities to understand, respond to and complete standardised PROMs that are routinely used in continence trials. The ELECTRIC trial participants were largely unable to do this, and a number of outcome measures were completed by proxies who were family members or, in the majority of cases, members of staff who knew the resident well. For example, only 150 (37%) residents completed the DEMQOL and 176 (43%) residents completed the MTSQ. Reliance on proxies for judgements about personal functions, such as PBC or quality of life, is less than ideal and there is no way to determine degree of accuracy; however, there was no alternative for this frail, dependent care home population. This feature of research in care homes was understood from the outset and influenced the design of the study, particularly the choice of primary outcome, which, for the reasons outlined previously, could not be a PROM. It needed to be objectively measurable, hence the use of absorbent pad weights to determine 24-hour urinary leakage. However, there were challenges associated with this outcome, which are discussed in Methodological considerations.

Poor information about the type of UI is a further resident-related factor that may help to explain the ELECTRIC trial results. TPTNS specifically targets symptoms of urinary urgency, frequency and nocturia, and was used in this trial because these symptoms occur in mixed UI, the most common type of UI in older adults. 3,16 Type of incontinence was recorded in the residents’ notes for only 135 residents, 33% of the sample, and there was no way to determine whether or not these residents did indeed experience overactive bladder symptoms of urgency, frequency or nocturia. Furthermore, the levels of dependency suggest that regardless of any other urinary symptoms, functional UI was the most prevalent clinical diagnosis in this sample. Earlier research supports this conclusion,64,65 in which 60–90% of care home residents with incontinence had significant mobility issues and severe cognitive impairment, similar to our results. The difficulties associated with enabling people with functional UI to use a toilet reinforces the previous interpretation that a lack of change to toileting practice will mask any effects of TPTNS that may have occurred.

Although we were keen to be inclusive and show that a dementia diagnosis should not exclude care home residents from participating in continence research, we should consider that many of these participants may have been too cognitively and physically frail to benefit from any effects of TPTNS on their UI because of their total dependence on staff for all continence activities. Such dependence makes sustainable changes to toileting practices unrealistic for large numbers of highly dependent residents in the majority of care homes. Nevertheless, TPTNS was shown to be well tolerated and safe, even for this frail population. Given the evidence from a number of other published trials and the divergence between these and the ELECTRIC trial results, we conclude that TPTNS is not effective in the care home context partly because of residents’ inability to independently use/express their need to use a toilet and their reliance on staff for all continence care. Such high levels of care input cannot be consistently provided or sustained in practice. In addition, the lack of effectiveness of TPTNS was shown across the entire resident population and reflected in the subgroup analyses, in which no evidence for any prespecified or post hoc grouping variable was found.

Care home-/organisation-related factors

In these care homes, continence care was seen as part of the everyday working routines and not something that necessarily needed to change. Qualitative findings suggested that incontinence was not seen as a health condition or a problem to be dealt with, but more an inherent characteristic of the resident. They were considered to be ‘incontinent of urine’ and this was their established bladder status, which was seen by some as inevitable with ageing, particularly in those residents with dementia. As such, it was accepted, indeed expected in some care homes, that residents wore absorbent pads to contain their incontinence, despite some residents, as described in Chapter 5, not understanding the reasons for this or agreeing with the practice. Containing the urine leakage, together with regular toilet visits, formed the bulk of the continence care practices in the care homes. This is in line with evidence from other care home studies. 66–68

Because managing UI was seen as ‘care’ it follows that it was considered the role of the CAs to manage it and this was not the realm of the RNs. This offers one explanation as to why the RNs did not engage with the ELECTRIC trial as originally anticipated. It also potentially explains why the continence care and toileting routines did not change during the trial. As CAs are not equipped or expected to make changes to practice, and as identified in the qualitative study, (1) the care was often entrenched and so much a fundamental part of daily work for the CAs as to be habitual and, therefore, subconsciously executed; (2) there was a lack of time and resources to increase the number of toilet visits should the need have been recognised by the residents or the care staff; and (3) there was no expectation of changes to continence status for the residents.

However, the qualitative study also found a high level of appreciation among the CAs and SCAs for the education about incontinence that was provided as part of the ELECTRIC trial training and enthusiasm to apply their new knowledge to improve their residents’ individual continence condition. For this to happen universally, changes to practice would be required that would need a person with the authority to effect these changes to engage with the project and lead it. The qualitative study raised the key role of leadership to the ELECTRIC trial, highlighting the impact of strong leadership on successful delivery of not only the TPTNS and sham interventions, but also the completion of the trial outcomes. For those care homes where the manager and/or senior nurse championed the ELECTRIC trial, the research process was generally smooth and all trial activities were absorbed into the care home working routines with little apparent difficulty. However, in situations where the care staff were ‘left to get on with it’, there was evidence of more negative experiences for staff and no sense of any comprehensive approach to implementation – ‘doing ELECTRIC’ meant demonstrably completing the stimulations, with little attention paid to toileting practices or outcomes collection activities. This was particularly the case for the local authority care homes where managers did not actively engage with the research processes and responsibility for all ELECTRIC trial activities was delegated to a senior carer, some of whom struggled to cope.

Such disparity in leadership may reveal fundamental beliefs about the status of continence care in this context. In some care homes, the ELECTRIC trial was an opportunity for the care staff to have something of their own, and this was reflected in the carers’ expressed sense of feeling valued, but whether there was real belief or confidence in any potential changes was questionable, and may not have translated into necessary changes to practice. Care home managers were not prepared to reduce their pad orders even if there was evidence of effect from the ELECTRIC trial. This was a point of interest in the qualitative findings and may reflect underlying doubt about any likely changes to residents’ continence status, as well as an understandable reluctance to decrease orders, based on pragmatic recognition that once given up, absorbent pad supplies may be difficult to reinstate. Regardless of any reductions in absorbent pad usage that may or may not have been observed during the ELECTRIC trial, care home managers stated that they would not be making any changes to their continence product orders. This is reasonable given the different systems and challenges associated with accessing continence products that many homes experience. The inflexibility and poor responsiveness of supply systems, together with limited individual resident product prescriptions, mean that care homes are unlikely to make any changes that may reduce the availability of products for their residents in the longer term.

Strengths and limitations

The ELECTRIC trial was a multicentre, pragmatic, participant- and outcome assessor-blind randomised placebo-controlled trial with a high follow-up rate for the primary outcome. The design included an extensive qualitative study involving a ‘whole-team’ approach that enabled a comprehensive understanding of the trial results from different perspectives.

The conduct and reporting of the trial followed the CONSORT recommendations (www.consort-statement.org/). We used a central web-based computer randomisation application developed by CHaRT, accessed remotely by trial staff, to assign group allocation and delivery to care home sites electronically, thus minimising the risk of selection bias. Analyses were conducted on an ITT basis and in accordance with a prespecified SAP (see Report Supplementary Material 8) agreed by the TSC.

We recruited participants from a range of different types and sizes of nursing and residential care homes in England and Scotland and achieved real diversity, adding to the generalisability of the findings. Ultimate variability in care home sites was greater than originally planned because of the early withdrawal of our main partner, BUPA UK, before recruitment began as a result of their decision to sell their care home business. Thus, alternative care home sites had to be found to replace the 14 large (> 100 beds) care homes originally proposed. The multiple sites recruited were so diverse that we are confident of representativeness.

At baseline, the two groups were highly comparable indicating the success of the allocation process. Follow up was high, with primary outcome data for 85% of participants at 6 weeks, 76% of participants at 12 weeks and 71% of participants at 18 weeks, and missing data were well balanced, suggestive of a low risk of attrition bias. Sensitivity analyses exploring potential effects of missingness on the primary outcome did not affect our conclusions.

To minimise the risk of performance bias, elements of the trial delivery were standardised as far as possible. This included the ELECTRIC trial training, which was delivered to all staff by two out of three trainers (the chief investigator and two trial managers) and used a standard training pack, which was delivered face to face and then provided in paper format to each member of staff who attended. Every care home was also given copies of DVDs demonstrating the techniques for TPTNS and sham stimulation.

The intervention was standardised as far as is possible. Identical equipment was used in a similar pattern over the 6-week intervention period for both groups. The only differences between groups were the electrode position on the ankle (medial malleolus in the TPTNS group, lateral malleolus in the sham group) and intensity of stimulation (highest comfortable intensity > 10 mA in the TPTNS group; 4 mA in the sham group).

The participants and outcomes assessors (ELECTRIC trial RRAs) were blind to the allocated group and thus the risk of detection bias was low. However, it was not possible to blind those delivering the intervention as they needed to deliver a different stimulation protocol depending on the participants’ allocated group. To ensure adherence to the relevant protocol, a minimum of two adherence checks were undertaken by the ISFs, who were part of the research trial staff, not the care home team, and whose roles were solely to support those delivering the intervention to do so correctly and to complete trial outcome measures correctly. Adherence to the intervention was 78% in the TPTNS group and 71% in the sham group, which was considered good when accounting for the number of sessions required and the complexity of delivery in the care home context.

Overall, the attrition rates in the ELECTRIC trial, including the number of deaths, were lower than anticipated and formed the basis of a recruitment target revision that took place after 12 full months of recruitment had been completed. There were fewer deaths in the trial cohort than expected. The original sample size calculation included a potential attrition of 30%, based on the experience of the EPIC trial69 [National Institute for Health Research (NIHR) Health Technology Assessment (HTA) 11/15/13] led by Professor Claire Surr, an ELECTRIC trial co-investigator. The revised sample size required for the ELECTRIC trial to be sufficiently powered reduced the sample size from 500 participants to having primary outcome data for a minimum of 278 participants. However, to allow the sample size to be adequate to conduct meaningful subgroup analyses, recruitment was continued according to the original planned timelines, which resulted in our final sample of 406 participants.

Potential limitations are recognised in the ELECTRIC trial. Despite the relatively low rates of missing data at each outcomes time point, detailed reasons were not recorded, other than the number of withdrawals, which were 27 (6.7%), 44 (10.8%) and 50 (12.3%) participants at 6, 12 and 18 weeks, respectively. Primary outcome data were missing for 23 (5.7%), 35 (8.6%) and 38 (9.4%) participants at 6, 12 and 18 weeks, respectively, and, although balanced across the two groups, a detailed breakdown of specific reasons would have been informative. Nevertheless, feedback from the research assistants, together with qualitative findings that highlighted the challenges associated with pad collections, suggests that the majority of missing data can be expected to be the result of failure to complete the pad collections by care homes. This was an issue for a number of care homes, as discussed previously, particularly where leadership from senior staff was limited or absent and communication between care team members was inadequate for instructions to cross shifts for 24 hours.

The 24-hour PWT is considered the most accurate and reliable way to objectively assess UI. 70 It requires collection of the total pad usage over the allotted time period, which can be easily attainable for a competent individual; however, for the ELECTRIC trial participants it meant depending on whole teams of staff spanning two or three care shifts to retain pads, often for a number of individuals, and put them in the correct collection bag. Co-operation and good communication between staff were essential to complete these PWTs; however, evidence from the qualitative interviews suggested that this could be challenging for some care homes. Discrepancies between bladder diary information and numbers of pads in some collection bags corroborated the difficulties encountered, which included forgetting to retain all pads for a participant, inadvertently throwing some pads in the clinical waste and failing to locate pads that were discarded by the resident. We also do not know what happened to soiled pads as this information was rarely provided on the bladder diaries. Staff were asked to discard solid material but include the pads in the collection; however, we do not know if this happened, and it may explain some of the small pad numbers in the collections. The care home environment itself could also present challenges for accurate pad collections. The collection bags were usually stored in the participant’s bathroom; however, the participant was not always taken back to their own room to use the toilet. Therefore, some pads were removed elsewhere in the care home, which could cause difficulties for staff in ensuring that the used pad was added to the correct collection bag in the correct room.

A diminishing interest in the trial may have occurred over the 18-week course, as indicated by the balanced reduction in pads collected at the 12- and 18-week time points, although it was not accompanied by a reduced 24-hour leakage or increased number of toilet visits. This suggests that participants had fewer pad changes over time.

The qualitative findings suggest that some staff did not understand the outcomes measurement aspects of the ELECTRIC trial. Although large numbers of care home staff received training, it was not possible to train all staff who were dealing with ELECTRIC trial participants; therefore, understanding and interest in the study processes was likely to be variable across staff. Equally, although the ELECTRIC trial training focused on the skills to both apply TPTNS/sham and complete the outcomes measures, the care home staff focused more on the intervention delivery than the outcomes assessment. It seemed from the qualitative interviews that many did not understand the purpose of the pad weights or the methods for accurate pad collection. This happened despite clear verbal and written instructions on each occasion and individual support from the research assistants, who confirmed what needed to be done and by whom when they delivered the equipment to complete the pad collection the day before it was due to begin.

We acknowledge that completion rates by care home residents were low for the self-reported secondary outcomes of PBC, MTSQ and DEMQOL. Although a limitation in terms of interpreting the responses, the low rates were not unexpected and were offset by proxy completion by participants’ family members or care home staff. In addition, information on use of these measures in people with moderate to severe cognitive impairment has been provided, which will be of potential benefit for future research in care home contexts.

A limitation of the qualitative study was the smaller than expected number of interviews undertaken with residents who had received TPTNS treatment. We aimed to recruit 75% of residents from the TPTNS group; however, only 57% were recruited from this group. This was because there were fewer residents with cognitive capacity in the treatment group. Therefore, the views of residents on their subjective experiences with TPTNS were more limited than planned.