Documented incontinence after stroke: a secondary analysis of a cohort study. Reducing Ethnic and Geographic Inequities to Optimise New Zealand Stroke Care (REGIONS Care)
View Article PDF
Incontinence after stroke is a common problem. Additionally, urinary incontinence and faecal incontinence are both associated with more severe stroke.[[1]] Urinary incontinence after stroke may be caused by the direct effects of the cerebral lesion on neurological bladder control physiology, but may also be caused by motor (mobility and dexterity), cognitive (apraxia and agnosia), visual, and language effects of the stroke that adversely affect toileting.[[2]] The reported prevalence of urinary and faecal incontinence after stroke varies by study but it is likely that more than half of patients may experience urinary incontinence,[[3,4]] and up to 40% have faecal incontinence in the acute stage after stroke.[[5]] Urinary incontinence after stroke is associated with poor functional outcomes,[[3,6]] lower quality of life,[[7–10]] more social isolation and depression,[[10]] greater likelihood of living in residential care,[[4,6,11]] and greater mortality.[[4,11,12]] There is evidence that faecal incontinence after stroke is associated with increased mortality,[[5,13]] and likelihood of living in residential care.[[13]] This contrasts with a New Zealand study of those without stroke, which reported that in older adults urinary but not faecal incontinence was a risk factor for admission to residential care.[[14]]
In Aotearoa New Zealand, two studies report on incontinence after stroke. A 1986 study reported the prevalence of urinary incontinence after stroke was high and depended on the time of assessment after stroke. In that study 60%, 42% and 29% of survivors after stroke, had urinary incontinence one, four and 12 weeks after stroke, respectively.[[15]] The Auckland Stroke Outcomes (ASTRO) study reported that 27% of stroke survivors had “bladder control problems” five years after stroke.[[16]] The prevalence of incontinence for different ethnic groups after stroke is unclear. The ASTRO study reported that there was no statistical evidence of a difference in bladder control problems after stroke in relation to ethnicity, although a higher proportion, 36% of Pacific peoples, reported problems with bladder control, compared to 27% of participants overall. The number of Pacific peoples in the study was small (n=30), and the study likely lacked statistical power to definitively assess whether a meaningful difference in urinary incontinence prevalence in relation to ethnicity was present or not.
The Australian and New Zealand Clinical Guidelines for Stroke Management recommend that all patients with stroke who may have continence problems should have a “structured functional assessment”, and those with incontinence should have a continence management plan implemented.[[17]] A recent survey of stroke service provision in New Zealand reported that 74% of hospitals that routinely treat patients with acute stroke had a protocol or guideline in place to manage urinary incontinence,[[18]] and 59% had one for managing faecal incontinence (unpublished data, 2020).[[18]] In Australia, a 2019 acute stroke services clinical audit reported that although 32% of patients had urinary incontinence documented, only 37% of these patients had a documented continence management plan.[[19]]
There is a lack of contemporary data about the prevalence and management of incontinence after stroke in New Zealand to inform planning and provision of services aimed to enable people with stroke to live well, for longer, in the community. This study is a secondary analysis of data collected for the Reducing Ethnic and Geographic Inequalities to Optimise New Zealand Stroke (REGIONS) Care study.[[20]] The current study aims were to: 1) estimate the prevalence of incontinence after stroke; 2) explore if the prevalence of incontinence after stroke was related to ethnicity including after adjustment for other important predictors of incontinence such as age and stroke severity; 3) estimate the strength of association, adjusted for important predictors of incontinence, between mortality and/or living in residential care with incontinence; 4) assess whether mobility status is an effect modifier to the association between living in residential care and continence; and 5) estimate the health utilities in patients with and without incontinence.
Methods
REGIONS Care is a nationwide study assessing the impact of hospital geographic location and ethnicity on stroke outcomes and access to best practice care. Full study methods and primary analyses for REGIONS Care are described elsewhere.[[20–22]]
In brief, REGIONS Care was a prospective, observational cohort study of all patients with a confirmed diagnosis of stroke who were admitted to the 28 hospitals that treat patients with acute stroke between 1 May and 31 July 2018. After this date, data collection continued until hospitals had achieved a minimum sample size of 150 (for stroke clot retrieval centres) or 100 for other centres, or until 31 October 2018—whichever occurred first. Data (baseline characteristics; acute treatment, investigations, and rehabilitation) up to and including the three months post stroke follow-up were collected by hospital staff. As registry-based data collection this process was considered audit and did not require individual patient consent. Patients were then contacted and were invited to consent to extended follow-ups at six and twelve months, data linkage with health administrative data, focus groups and surveys (Central Health and Disability Ethics Committee, reference 17/CEN/164).
Selected participant baseline characteristics, and possible predictors of incontinence (any type) after stroke were extracted from the full data set. These included: age, sex, ethnicity, stroke risk factors, primary diagnosis, pre-morbid level of function, pre-stroke living situation, and stroke severity at time of hospital admission using the six simple variables (SSV) model (a composite of age, living alone pre-stroke, independence in activities of daily living pre-stroke, Glasgow Coma Scale verbal component, arm power, and ability to walk unaided at admission).[[23]] Patient outcomes at discharge were death and discharge destination (home or residential care), and at three, six and twelve months, outcomes were death or living in residential care which is a composite measure for a poor outcome that encompasses both severe functional deficits and mortality. The composite measure is needed because a high proportion of those with stroke die, and so an outcome that just depends on residential care status will have missing data. As a potential effect modifier, mobility status (independent or dependent), was determined at each time point from the relevant modified Rankin Scale (mRS) score. The dichotomisation (mRS 0 to 3; mRS 4 or 5) was made to capture different levels of disability likely to influence the functional aspects of toileting, particularly walking to the toilet independently. People with a mRS score of 4 or 5 require external assistance for walking or are bedridden respectively. The EuroQol EQ-5D-3L quality of life measure was used to estimate Health Utility Scores which are scaled as a score from zero, representing a quality of life that is equivalent to death, to one, representing perfect health.[[24]]
Incontinence-related variables in the REGIONS Care dataset were any documented incontinence (any type), documented continence management plan, use of an indwelling catheter (IDC), urinary tract infection (UTI) and constipation post stroke, during acute care and before discharge; all these were collected as yes/no responses. If an IDC was used, a reason could be stated. Data collection did not distinguish between urinary or faecal incontinence, or existence of both. Incontinence data were collected in a non-compulsory data field related to stroke complications; incontinence may have been documented in clinical notes but inconsistently in study data.
Continuous variables are summarised by mean and standard deviation (SD), and median and interquartile (IQR) range, as appropriate. Counts are summarised by proportions and appropriate confidence intervals for a proportion. Logistic regression examines associations with dichotomous outcome variables and linear regression with continuous outcome variables. Multivariate models were adjusted for age, sex, ethnicity, and stroke severity as confounders and the association between living in residential care and incontinence main effects and mobility assessed by appropriate interaction terms. Stata/IC 16.0 was used for analysis.
The Health Research Council of New Zealand (HRC 17/037) funded the REGIONS Care study and the study received ethics approval from the Central Region Health and Disability Ethics Committee (17CEN164).
Results
The baseline characteristics of patients are summarised in Table 1. There were 2,379 patients with stroke admitted to hospital during the study period, of whom 320 had documented incontinence and 2057 did not; two did not have recorded data about incontinence. The mean (SD) age was 75 (13.7) years; 1,219/2,379 (51.2%) were male; and 2,052/2,343 (87.6%) with a recorded stroke type had ischemic stroke. Baseline characteristics of patients with and without incontinence (any type) are shown in Table 1.
The overall prevalence of incontinence problems and related issues are summarised in Table 2.
The prevalence of incontinence and related variables by ethnicity is summarised in Appendix 1. There was no evidence that the proportion of those with incontinence was associated with ethnicity in univariate (P=0.61) or multivariate analysis adjusting for age, sex and stroke severity, P=0.58. The univariate and multivariate adjusted odds ratios (OR) are shown in Table 3 with NZ European ethnicity as the reference level.
Patients with documented incontinence were more likely to die (52/320 (16.3%)) or be discharged to residential care (118 (36.9%)) than those with no documented incontinence (9.6% and 11.8% respectively). Table 4 shows the summary data, unadjusted and adjusted (for age, sex, and stroke severity) odds ratios for the individual outcomes at discharge, and the composite outcomes at the subsequent time points, in relation to documented incontinence.
Disability on discharge, for those who were discharged alive, was an effect modifier for the risk of entering residential care. In a univariate analysis 247/360 (68.6%) of those discharged to residential care had a mRS of 4 or 5 (includes lacking independent mobility) compared to 159/1,749 (9.1%) of those discharged to the community, OR (95% CI) 18.5 (13.9–24.7), P<0.001. For the same group of those discharged alive, 118 (32.8%) of those discharged to residential care had incontinence compared to 149 (8.5%) of those discharged to the community; OR (95% CI) 1.87 (1.32–2.65), P<0.001. However, the P-value for the interaction between disability and incontinence in relation to discharge status was statistically significant (P<0.001) indicating that a person with less disability was more likely to be discharged to residential care if incontinent. Table 5 shows the proportions of those with incontinence amongst those with more versus less disability discharged to residential care or the community. In those with less disability, incontinence was strongly related to discharge to residential care; but there was no association in those with more disability.
The mean (SD) health utility scores for patients with documented incontinence were 0.52 (0.28) at three months, 0.55 (0.28) at six months and 0.53 (0.27) at 12 months; compared to 0.71 (0.23) at three months, 0.71 (0.24) at six months and 0.70 (0.25) at 12 months for patients without documented incontinence. The differences in mean health utility score by continence status in the univariate modelling was -0.19 (95% CI -0.23 to -0.15) at three months, -0.16 (95% CI -0.21 to -0.10) at six months, and -0.17 (95% CI -0.24 to -0.12) at 12 months. In the multivariate model (adjusted for age, sex, ethnicity and stroke severity), the differences were -0.11 (-0.15 to -0.07), -0.07 (-0.13 to -0.02), and -0.10 (-0.16 to -0.04) respectively. For patients who reported no problem on each of the five EQ-5D-3L dimensions, thus scoring a health utility score of 1 (meaning full health), the mean (SD) EQ-VAS scores at three, six and twelve months were 82.3 (13.7), 82.5 (15.5) and 84.2 (11.8) respectively.
View Tables 1–5.
Discussion
In this secondary analysis of documented incontinence (or not) in a large cohort of hospital admission for stroke we found a much lower incontinence prevalence than reported in previous New Zealand[[15,16]] or more recent international research.[[4,5]] Documented incontinence included any type or severity (i.e., any frequency or volume of leakage) at any stage in acute care before discharge. It is difficult to explain why there is a markedly reduced prevalence of post stroke incontinence in New Zealand compared to international data. It seems unlikely this reflects pathophysiological differences or acute care. It seems more likely that this difference reflects a lower rate of identifying patients with incontinence or at least inconsistent documentation of incontinence in the inpatient record,[[25]] rather than being a true reflection of the prevalence of incontinence post stroke in New Zealand. When incontinence was documented, nearly three quarters of patients had a continence management plan documented too. The data collection did not allow investigation of whether incontinence, or a continence plan, was more common if the incontinence was more severe, or the person had double (urinary and faecal) incontinence. If impact on nursing is greater, e.g., bed changes, or use of containment products, it is possible that documentation is more likely. The low reported prevalence may thus be focussed on the proportion of stroke survivors with severe incontinence.
Stroke guidelines recommend that all patients with suspected incontinence are assessed.[[14]] The low prevalence of documented incontinence in this cohort, compared to international prevalence reports, raises concerns that not all patients with incontinence were identified and the possibility that some may have experienced unmet needs in relation to continence management. This may represent an important area for ongoing service improvement efforts, such as those found to be useful for increasing guideline adherence in other areas of stroke care[[26]] and incontinence care[[25,27]] in medical and older persons wards.
We found that those with documented incontinence had increased risk of being discharged to residential care, and increased risk of being deceased or living in residential care three, six and twelve months after stroke. These findings align with previous research reporting that incontinence after stroke is associated with poorer outcomes.[[4,6,11,12]] Disability was an effect modifier for the association between incontinence and living in residential care. Thus, incontinence increased the risk of moving into residential care for those stroke survivors with less disability, but not for those with moderate or severe disability. It is possible those with greater dependency are already more likely to discharge to residential care. The Sentinel Stroke National Audit in the UK found that a mRS of 4 or 5 at discharge was strongly associated with first time residential placement.[[28]] In contrast to our findings Dutta et al, in a subgroup analysis of stroke patients with an mRS of 4 or 5, found incontinent patients had nearly five times the odds of discharge to residential care.[[28]]
There was no association of documented incontinence with ethnicity. A previous study, The Auckland Stroke Outcomes (ASTRO) study, did find a higher prevalence of incontinence in Pacific peoples compared to the sample population overall. Two qualitative studies, both about women’s perspectives on urinary incontinence, have specifically sought experiences of Māori and Pacific women. Lennan et al. (1999)[[29]] sought to understand the impacts of urinary incontinence on psychological, social and medical wellbeing—the study included eight Māori, nine Pacific, and 17 Pākehā women—and found that Māori and Pacific women were more likely to say there were barriers to help seeking, including language barriers and trust and confidentiality issues. Having a trust relationship and being assured of confidentiality were the most important attributes of the doctor for Māori and Pacific women. TuiSamoa et al. (2022)[[30]] carried out a single focus group with 10 Pasifika women, not all of whom had incontinence. The findings were consistent with those from Lennan et al., with the addition of understanding that culturally safe care may vary according to the woman’s generation. As continence services in New Zealand are limited and inequitably distributed,[[31]] further study of possible inequity in access, experience and outcomes for incontinence care after stroke may be warranted.
Participants with documented incontinence had consistently reduced health utility scores (closer to zero) than those without documented incontinence at three, six-and twelve months post stroke. The mean differences (after adjusting for age, sex, ethnicity and stroke severity) were consistent with minimal clinically important differences in EuroQoL utilities reported in stroke research.[[32]] This suggests additional and potentially important clinical impact, attributable to incontinence, on health-related quality of life after stroke.
The strength of this analysis is that the data came from a large prospective cohort of patients with stroke. There are several limitations. First, the REGIONS Care study was not designed to specifically investigate incontinence after stroke. Although documentation of incontinence (yes/no) was not a compulsory data collection field, the data were missing for only two of the 2,377 participants. Second, there is the lack of detail about incontinence meaning it was not possible to estimate prevalence of urinary, faecal, or double incontinence, or pre-existing incontinence symptoms, how severe incontinence symptoms were, and the influence of any of these on the associations we examined. Third, the ceiling effects of the EQ-5D-3L, may have influenced the estimated health utility scores. Most patients who reported no difficulties across any of the five dimensions of the questionnaire, thus scoring a 1 for the health utility score, corresponding to “full health”, did not score their overall health at 100 on the EQ-VAS, with mean scores ranging from 82.3 to 84.2 at three and twelve months respectively. The five level version of the EQ-5D may have increased the sensitivity and reduced some of the observed ceiling effects in the three level version.
The low prevalence of documented incontinence after stroke compared to international estimates may indicate that incontinence is not reliably identified, which means some stroke survivors may have unmet continence management needs. Given the data limitations outlined above, a more detailed investigation of post stroke incontinence is needed, distinguishing urinary and faecal incontinence, estimating severity using standardised and internationally validated continence assessment instruments, and to clarify if incontinence is new or pre-existing. A better description of care received, and unmet needs would also clarify the effect of incontinence on stroke survivors and their carers. These data are needed to understand the scale of the need continence services are trying to meet and make reasonable estimates of need with population ageing. However, collecting these data may also require attention to the many barriers health professionals face in identifying and managing incontinence.
Consistent with existing literature we found an association between continence problems and increased mortality and living in residential care. The finding that stroke survivors who have none to slight disability, and incontinence, were more likely to live in residential care than those who were moderately or severely disabled is counter-intuitive; the expectation being that those with more disability and continence problems are more likely to need residential care on discharge. This may possibly reflect the nature of the data collection, and is worth exploration as effective continence management in those with less disability may enable continued community living.
Conclusion
The prevalence of documented in-hospital post stroke incontinence in New Zealand is either much less than international estimates or was, more likely, underestimated in this secondary analysis of REGIONS Care data. Prevalence might not be associated with ethnicity, although more detailed incontinence data—such as type and severity—are needed to investigate the association more closely. Documented incontinence was associated with increased mortality and probability of living in residential care at discharge, three, six and twelve months. As might be expected, those with documented incontinence have reduced health utility scores compared to those without documented incontinence.
View Appendix 1.
Summary
Abstract
Aim
To estimate the prevalence of incontinence after stroke in Aotearoa New Zealand overall and by ethnicity, the associations between incontinence and subsequent mortality and living in residential care, and to estimate the health utilities in relation to continence.
Method
Secondary analysis of data from a prospective (1 May to 31 July 2018) cohort study (REGIONS Care study) of patients with a confirmed stroke admitted to New Zealand hospitals. Logistic and linear regression were used, and multivariate models were adjusted for age, sex, ethnicity, and stroke severity. The association between living in residential care, incontinence, and mobility was also assessed.
Results
There were 320/2,377 (13.5%) patients with documented incontinence during hospitalisation after stroke. Incontinence was not associated with ethnicity but was associated with increased mortality/living in residential care, at discharge, three, six and twelve months after stroke. Stroke survivors with independent mobility were more likely to live in residential care if incontinent. Health utility scores were lower at three, six and twelve months for those with incontinence after stroke.
Conclusion
This study likely underestimated incontinence prevalence after stroke, although incontinence was associated with increased mortality and probability of living in residential care.
Author Information
Acknowledgements
Correspondence
Correspondence Email
Competing Interests
1) Brocklehurst JC, Andrews K, Richards B, Laycock PJ. Incidence and correlates of incontinence in stroke patients. J Am Geriatr Soc. 1985;33(8):540-2.
2) Thomas LH, Coupe J, Cross LD, Tan AL, Watkins CL. Interventions for treating urinary incontinence after stroke in adults. Cochrane Database of Systematic Reviews. 2019(2).
3) Barer D. Continence after stroke: useful predictor or goal of therapy? Age Ageing. 1989;18(3):183-91.
4) Kolominsky‐Rabas PL, Hilz MJ, Neundoerfer B, Heuschmann PU. Impact of urinary incontinence after stroke: results from a prospective population‐based stroke register. Neurourol Urodyn. 2003;22(4):322-7.
5) Nakayama H, Jørgensen H, Pedersen P, Raaschou H, Olsen T. Prevalence and risk factors of incontinence after stroke: the Copenhagen Stroke Study. Stroke. 1997;28(1):58-62.
6) Ween JE, Alexander MP, D'Esposito M, Roberts M. Incontinence after stroke in a rehabilitation setting: outcome associations and predictive factors. Neurology. 1996;47(3):659-63.
7) Brittain K, Perry S, Peet S, Shaw C, Dallosso H, Assassa R, et al. Prevalence and impact of urinary symptoms among community-dwelling stroke survivors. Stroke. 2000;31(4):886-91.
8) Patel M, McKevitt C, Lawrence E, Rudd A, Wolfe C. Clinical determinants of long-term quality of life after stroke. Age Ageing. 2007;36(3):316-22.
9) Brittain K, Castleden C. Suicide in patients with stroke: Depression may be caused by symptoms affecting lower urinary tract. BMJ. 1998 Oct 10;317(7164):1016-7.
10) Brittain KR, Shaw C. The social consequences of living with and dealing with incontinence—A carers perspective. Soc Sci Med. 2007 Sep;65(6):1274-83.
11) Patel M, Coshall C, Rudd AG, Wolfe CD. Natural history and effects on 2-year outcomes of urinary incontinence after stroke. Stroke. 2001;32(1):122-7.
12) Rotar M, Blagus R, Jeromel M, Škrbec M, Tršinar B, Vodušek DB. Stroke patients who regain urinary continence in the first week after acute first‐ever stroke have better prognosis than patients with persistent lower urinary tract dysfunction. Neurourol Urodyn. 2011;30(7):1315-8.
13) Harari D, Coshall C, Rudd AG, Wolfe CD. New-onset fecal incontinence after stroke: prevalence, natural history, risk factors, and impact. Stroke. 2003;34(1):144-50.
14) Schluter PJ, Ward C, Arnold EP, Scrase R, Jamieson HA. Urinary incontinence, but not fecal incontinence, is a risk factor for admission to aged residential care of older persons in New Zealand. Neurourol Urodyn. 2017;36(6):1588-95.
15) Borrie MJ, Campbell AJ, Caradoc-Davies TH, Spears GF. Urinary incontinence after stroke: a prospective study. Age Ageing. 1986;15(3):177-81.
16) Feigin VL, Barker-Collo S, Parag V, Senior H, Lawes C, Ratnasabapathy Y, et al. Auckland Stroke Outcomes Study: Part 1: Gender, stroke types, ethnicity, and functional outcomes 5 years poststroke. Neurology. 2010 Nov 2;75(18):1597-607.
17) Stroke Foundation of New Zealand and New Zealand Guidelines Group. New Zealand clinical guidelines for stroke management 2010. Wellington: Stroke Foundation of New Zealand; 2010.
18) Thompson S, Barber PA, Fink J, Gommans J, Davis A, Harwood M, et al. New Zealand hospital stroke service provision. New Zealand hospital. 2020;133(1526).
19) Stroke Foundation. National stroke audit - acute services report 2019. 2019.
20) Ranta A, Thompson S, Harwood MLN, Cadilhac DA-M, Barber PA, Davis AJ, et al. Reducing Ethnic and Geographic Inequities to Optimise New Zealand Stroke Care (REGIONS Care): Protocol for a Nationwide Observational Study. JMIR research protocols. 2021;10(1):e25374.
21) Thompson SG, Barber PA, Gommans JH, Cadilhac DA, Davis A, Fink JN, et al. The impact of ethnicity on stroke care access and patient outcomes: a New Zealand nationwide observational study. Lancet Reg Health West Pac. 2022 Jan 3;20:100358.
22) Ranta A, Thompson S, Davis A, Barber PA, Fink J, Gommans J, et al. Urban versus Non-Urban Hospital Location Impacts on Stroke Outcomes. Stroke. 2021;52(Suppl_1):A29-A.
23) Counsell C, Dennis M, McDowall M. Predicting functional outcome in acute stroke: comparison of a simple six variable model with other predictive systems and informal clinical prediction. Journal Neurol Neurosurg Psychiatry 2004;75(3):401-405.
24) Devlin N, Parkin D, Janssen B. Methods for analysing and reporting EQ-5D data: Springer Nature; 2020.
25) Trad W, Flowers K, Caldwell J, Sousa MS, Vigh G, Lizarondo L, et al. Nursing assessment and management of incontinence among medical and surgical adult patients in a tertiary hospital: A best practice implementation project. JBI Database System Rev Implement Rep. 2019 Dec;17(12):2578-2590.
26) Tooher R, Middleton P, Pham C, Fitridge R, Rowe S, Babidge W, et al. A systematic review of strategies to improve prophylaxis for venous thromboembolism in hospitals. Ann Surg. 2005 Mar;241(3):397-415.
27) Sanai Memon G, Imam A, Datta-Chaudhuri M, Robertson E, Frain C. Improving identification and assessment of urinary incontinence in older people. Age Ageing. 2020;49.
28) Dutta D, Thornton D, Bowen E. Using population-based routinely collected data from the Sentinel Stroke National Audit Programme to investigate factors associated with discharge to care home after rehabilitation. Clin Rehabil. 2018 Aug;32(8):1108-1118.
29) Lennan M, Smalldridge J, Fa'alau F. Urinary incontinence: a qualitative approach: Department of Women's Health, South Auckland Health; 1999.
30) TuiSamoa A, Heather M, Kruger J. Urinary incontinence in Pasifika women: A pilot focus group study. TheAustralian and New Zealand Continence Journal. 2022;28(1):4-8.
31) Esplin J, Smith J, Doust E, Poynton M. Report on Good Practice of Continence Services in New Zealand Report. 2017.
32) Chen P, Lin K-C, Liing R-J, Wu C-Y, Chen C-L, Chang K-C. Validity, responsiveness, and minimal clinically important difference of EQ-5D-5L in stroke patients undergoing rehabilitation. Quality of life research. 2016;25(6):1585-96.
For the PDF of this article,
contact nzmj@nzma.org.nz
Documented incontinence after stroke: a secondary analysis of a cohort study. Reducing Ethnic and Geographic Inequities to Optimise New Zealand Stroke Care (REGIONS Care)
View Article PDF
Incontinence after stroke is a common problem. Additionally, urinary incontinence and faecal incontinence are both associated with more severe stroke.[[1]] Urinary incontinence after stroke may be caused by the direct effects of the cerebral lesion on neurological bladder control physiology, but may also be caused by motor (mobility and dexterity), cognitive (apraxia and agnosia), visual, and language effects of the stroke that adversely affect toileting.[[2]] The reported prevalence of urinary and faecal incontinence after stroke varies by study but it is likely that more than half of patients may experience urinary incontinence,[[3,4]] and up to 40% have faecal incontinence in the acute stage after stroke.[[5]] Urinary incontinence after stroke is associated with poor functional outcomes,[[3,6]] lower quality of life,[[7–10]] more social isolation and depression,[[10]] greater likelihood of living in residential care,[[4,6,11]] and greater mortality.[[4,11,12]] There is evidence that faecal incontinence after stroke is associated with increased mortality,[[5,13]] and likelihood of living in residential care.[[13]] This contrasts with a New Zealand study of those without stroke, which reported that in older adults urinary but not faecal incontinence was a risk factor for admission to residential care.[[14]]
In Aotearoa New Zealand, two studies report on incontinence after stroke. A 1986 study reported the prevalence of urinary incontinence after stroke was high and depended on the time of assessment after stroke. In that study 60%, 42% and 29% of survivors after stroke, had urinary incontinence one, four and 12 weeks after stroke, respectively.[[15]] The Auckland Stroke Outcomes (ASTRO) study reported that 27% of stroke survivors had “bladder control problems” five years after stroke.[[16]] The prevalence of incontinence for different ethnic groups after stroke is unclear. The ASTRO study reported that there was no statistical evidence of a difference in bladder control problems after stroke in relation to ethnicity, although a higher proportion, 36% of Pacific peoples, reported problems with bladder control, compared to 27% of participants overall. The number of Pacific peoples in the study was small (n=30), and the study likely lacked statistical power to definitively assess whether a meaningful difference in urinary incontinence prevalence in relation to ethnicity was present or not.
The Australian and New Zealand Clinical Guidelines for Stroke Management recommend that all patients with stroke who may have continence problems should have a “structured functional assessment”, and those with incontinence should have a continence management plan implemented.[[17]] A recent survey of stroke service provision in New Zealand reported that 74% of hospitals that routinely treat patients with acute stroke had a protocol or guideline in place to manage urinary incontinence,[[18]] and 59% had one for managing faecal incontinence (unpublished data, 2020).[[18]] In Australia, a 2019 acute stroke services clinical audit reported that although 32% of patients had urinary incontinence documented, only 37% of these patients had a documented continence management plan.[[19]]
There is a lack of contemporary data about the prevalence and management of incontinence after stroke in New Zealand to inform planning and provision of services aimed to enable people with stroke to live well, for longer, in the community. This study is a secondary analysis of data collected for the Reducing Ethnic and Geographic Inequalities to Optimise New Zealand Stroke (REGIONS) Care study.[[20]] The current study aims were to: 1) estimate the prevalence of incontinence after stroke; 2) explore if the prevalence of incontinence after stroke was related to ethnicity including after adjustment for other important predictors of incontinence such as age and stroke severity; 3) estimate the strength of association, adjusted for important predictors of incontinence, between mortality and/or living in residential care with incontinence; 4) assess whether mobility status is an effect modifier to the association between living in residential care and continence; and 5) estimate the health utilities in patients with and without incontinence.
Methods
REGIONS Care is a nationwide study assessing the impact of hospital geographic location and ethnicity on stroke outcomes and access to best practice care. Full study methods and primary analyses for REGIONS Care are described elsewhere.[[20–22]]
In brief, REGIONS Care was a prospective, observational cohort study of all patients with a confirmed diagnosis of stroke who were admitted to the 28 hospitals that treat patients with acute stroke between 1 May and 31 July 2018. After this date, data collection continued until hospitals had achieved a minimum sample size of 150 (for stroke clot retrieval centres) or 100 for other centres, or until 31 October 2018—whichever occurred first. Data (baseline characteristics; acute treatment, investigations, and rehabilitation) up to and including the three months post stroke follow-up were collected by hospital staff. As registry-based data collection this process was considered audit and did not require individual patient consent. Patients were then contacted and were invited to consent to extended follow-ups at six and twelve months, data linkage with health administrative data, focus groups and surveys (Central Health and Disability Ethics Committee, reference 17/CEN/164).
Selected participant baseline characteristics, and possible predictors of incontinence (any type) after stroke were extracted from the full data set. These included: age, sex, ethnicity, stroke risk factors, primary diagnosis, pre-morbid level of function, pre-stroke living situation, and stroke severity at time of hospital admission using the six simple variables (SSV) model (a composite of age, living alone pre-stroke, independence in activities of daily living pre-stroke, Glasgow Coma Scale verbal component, arm power, and ability to walk unaided at admission).[[23]] Patient outcomes at discharge were death and discharge destination (home or residential care), and at three, six and twelve months, outcomes were death or living in residential care which is a composite measure for a poor outcome that encompasses both severe functional deficits and mortality. The composite measure is needed because a high proportion of those with stroke die, and so an outcome that just depends on residential care status will have missing data. As a potential effect modifier, mobility status (independent or dependent), was determined at each time point from the relevant modified Rankin Scale (mRS) score. The dichotomisation (mRS 0 to 3; mRS 4 or 5) was made to capture different levels of disability likely to influence the functional aspects of toileting, particularly walking to the toilet independently. People with a mRS score of 4 or 5 require external assistance for walking or are bedridden respectively. The EuroQol EQ-5D-3L quality of life measure was used to estimate Health Utility Scores which are scaled as a score from zero, representing a quality of life that is equivalent to death, to one, representing perfect health.[[24]]
Incontinence-related variables in the REGIONS Care dataset were any documented incontinence (any type), documented continence management plan, use of an indwelling catheter (IDC), urinary tract infection (UTI) and constipation post stroke, during acute care and before discharge; all these were collected as yes/no responses. If an IDC was used, a reason could be stated. Data collection did not distinguish between urinary or faecal incontinence, or existence of both. Incontinence data were collected in a non-compulsory data field related to stroke complications; incontinence may have been documented in clinical notes but inconsistently in study data.
Continuous variables are summarised by mean and standard deviation (SD), and median and interquartile (IQR) range, as appropriate. Counts are summarised by proportions and appropriate confidence intervals for a proportion. Logistic regression examines associations with dichotomous outcome variables and linear regression with continuous outcome variables. Multivariate models were adjusted for age, sex, ethnicity, and stroke severity as confounders and the association between living in residential care and incontinence main effects and mobility assessed by appropriate interaction terms. Stata/IC 16.0 was used for analysis.
The Health Research Council of New Zealand (HRC 17/037) funded the REGIONS Care study and the study received ethics approval from the Central Region Health and Disability Ethics Committee (17CEN164).
Results
The baseline characteristics of patients are summarised in Table 1. There were 2,379 patients with stroke admitted to hospital during the study period, of whom 320 had documented incontinence and 2057 did not; two did not have recorded data about incontinence. The mean (SD) age was 75 (13.7) years; 1,219/2,379 (51.2%) were male; and 2,052/2,343 (87.6%) with a recorded stroke type had ischemic stroke. Baseline characteristics of patients with and without incontinence (any type) are shown in Table 1.
The overall prevalence of incontinence problems and related issues are summarised in Table 2.
The prevalence of incontinence and related variables by ethnicity is summarised in Appendix 1. There was no evidence that the proportion of those with incontinence was associated with ethnicity in univariate (P=0.61) or multivariate analysis adjusting for age, sex and stroke severity, P=0.58. The univariate and multivariate adjusted odds ratios (OR) are shown in Table 3 with NZ European ethnicity as the reference level.
Patients with documented incontinence were more likely to die (52/320 (16.3%)) or be discharged to residential care (118 (36.9%)) than those with no documented incontinence (9.6% and 11.8% respectively). Table 4 shows the summary data, unadjusted and adjusted (for age, sex, and stroke severity) odds ratios for the individual outcomes at discharge, and the composite outcomes at the subsequent time points, in relation to documented incontinence.
Disability on discharge, for those who were discharged alive, was an effect modifier for the risk of entering residential care. In a univariate analysis 247/360 (68.6%) of those discharged to residential care had a mRS of 4 or 5 (includes lacking independent mobility) compared to 159/1,749 (9.1%) of those discharged to the community, OR (95% CI) 18.5 (13.9–24.7), P<0.001. For the same group of those discharged alive, 118 (32.8%) of those discharged to residential care had incontinence compared to 149 (8.5%) of those discharged to the community; OR (95% CI) 1.87 (1.32–2.65), P<0.001. However, the P-value for the interaction between disability and incontinence in relation to discharge status was statistically significant (P<0.001) indicating that a person with less disability was more likely to be discharged to residential care if incontinent. Table 5 shows the proportions of those with incontinence amongst those with more versus less disability discharged to residential care or the community. In those with less disability, incontinence was strongly related to discharge to residential care; but there was no association in those with more disability.
The mean (SD) health utility scores for patients with documented incontinence were 0.52 (0.28) at three months, 0.55 (0.28) at six months and 0.53 (0.27) at 12 months; compared to 0.71 (0.23) at three months, 0.71 (0.24) at six months and 0.70 (0.25) at 12 months for patients without documented incontinence. The differences in mean health utility score by continence status in the univariate modelling was -0.19 (95% CI -0.23 to -0.15) at three months, -0.16 (95% CI -0.21 to -0.10) at six months, and -0.17 (95% CI -0.24 to -0.12) at 12 months. In the multivariate model (adjusted for age, sex, ethnicity and stroke severity), the differences were -0.11 (-0.15 to -0.07), -0.07 (-0.13 to -0.02), and -0.10 (-0.16 to -0.04) respectively. For patients who reported no problem on each of the five EQ-5D-3L dimensions, thus scoring a health utility score of 1 (meaning full health), the mean (SD) EQ-VAS scores at three, six and twelve months were 82.3 (13.7), 82.5 (15.5) and 84.2 (11.8) respectively.
View Tables 1–5.
Discussion
In this secondary analysis of documented incontinence (or not) in a large cohort of hospital admission for stroke we found a much lower incontinence prevalence than reported in previous New Zealand[[15,16]] or more recent international research.[[4,5]] Documented incontinence included any type or severity (i.e., any frequency or volume of leakage) at any stage in acute care before discharge. It is difficult to explain why there is a markedly reduced prevalence of post stroke incontinence in New Zealand compared to international data. It seems unlikely this reflects pathophysiological differences or acute care. It seems more likely that this difference reflects a lower rate of identifying patients with incontinence or at least inconsistent documentation of incontinence in the inpatient record,[[25]] rather than being a true reflection of the prevalence of incontinence post stroke in New Zealand. When incontinence was documented, nearly three quarters of patients had a continence management plan documented too. The data collection did not allow investigation of whether incontinence, or a continence plan, was more common if the incontinence was more severe, or the person had double (urinary and faecal) incontinence. If impact on nursing is greater, e.g., bed changes, or use of containment products, it is possible that documentation is more likely. The low reported prevalence may thus be focussed on the proportion of stroke survivors with severe incontinence.
Stroke guidelines recommend that all patients with suspected incontinence are assessed.[[14]] The low prevalence of documented incontinence in this cohort, compared to international prevalence reports, raises concerns that not all patients with incontinence were identified and the possibility that some may have experienced unmet needs in relation to continence management. This may represent an important area for ongoing service improvement efforts, such as those found to be useful for increasing guideline adherence in other areas of stroke care[[26]] and incontinence care[[25,27]] in medical and older persons wards.
We found that those with documented incontinence had increased risk of being discharged to residential care, and increased risk of being deceased or living in residential care three, six and twelve months after stroke. These findings align with previous research reporting that incontinence after stroke is associated with poorer outcomes.[[4,6,11,12]] Disability was an effect modifier for the association between incontinence and living in residential care. Thus, incontinence increased the risk of moving into residential care for those stroke survivors with less disability, but not for those with moderate or severe disability. It is possible those with greater dependency are already more likely to discharge to residential care. The Sentinel Stroke National Audit in the UK found that a mRS of 4 or 5 at discharge was strongly associated with first time residential placement.[[28]] In contrast to our findings Dutta et al, in a subgroup analysis of stroke patients with an mRS of 4 or 5, found incontinent patients had nearly five times the odds of discharge to residential care.[[28]]
There was no association of documented incontinence with ethnicity. A previous study, The Auckland Stroke Outcomes (ASTRO) study, did find a higher prevalence of incontinence in Pacific peoples compared to the sample population overall. Two qualitative studies, both about women’s perspectives on urinary incontinence, have specifically sought experiences of Māori and Pacific women. Lennan et al. (1999)[[29]] sought to understand the impacts of urinary incontinence on psychological, social and medical wellbeing—the study included eight Māori, nine Pacific, and 17 Pākehā women—and found that Māori and Pacific women were more likely to say there were barriers to help seeking, including language barriers and trust and confidentiality issues. Having a trust relationship and being assured of confidentiality were the most important attributes of the doctor for Māori and Pacific women. TuiSamoa et al. (2022)[[30]] carried out a single focus group with 10 Pasifika women, not all of whom had incontinence. The findings were consistent with those from Lennan et al., with the addition of understanding that culturally safe care may vary according to the woman’s generation. As continence services in New Zealand are limited and inequitably distributed,[[31]] further study of possible inequity in access, experience and outcomes for incontinence care after stroke may be warranted.
Participants with documented incontinence had consistently reduced health utility scores (closer to zero) than those without documented incontinence at three, six-and twelve months post stroke. The mean differences (after adjusting for age, sex, ethnicity and stroke severity) were consistent with minimal clinically important differences in EuroQoL utilities reported in stroke research.[[32]] This suggests additional and potentially important clinical impact, attributable to incontinence, on health-related quality of life after stroke.
The strength of this analysis is that the data came from a large prospective cohort of patients with stroke. There are several limitations. First, the REGIONS Care study was not designed to specifically investigate incontinence after stroke. Although documentation of incontinence (yes/no) was not a compulsory data collection field, the data were missing for only two of the 2,377 participants. Second, there is the lack of detail about incontinence meaning it was not possible to estimate prevalence of urinary, faecal, or double incontinence, or pre-existing incontinence symptoms, how severe incontinence symptoms were, and the influence of any of these on the associations we examined. Third, the ceiling effects of the EQ-5D-3L, may have influenced the estimated health utility scores. Most patients who reported no difficulties across any of the five dimensions of the questionnaire, thus scoring a 1 for the health utility score, corresponding to “full health”, did not score their overall health at 100 on the EQ-VAS, with mean scores ranging from 82.3 to 84.2 at three and twelve months respectively. The five level version of the EQ-5D may have increased the sensitivity and reduced some of the observed ceiling effects in the three level version.
The low prevalence of documented incontinence after stroke compared to international estimates may indicate that incontinence is not reliably identified, which means some stroke survivors may have unmet continence management needs. Given the data limitations outlined above, a more detailed investigation of post stroke incontinence is needed, distinguishing urinary and faecal incontinence, estimating severity using standardised and internationally validated continence assessment instruments, and to clarify if incontinence is new or pre-existing. A better description of care received, and unmet needs would also clarify the effect of incontinence on stroke survivors and their carers. These data are needed to understand the scale of the need continence services are trying to meet and make reasonable estimates of need with population ageing. However, collecting these data may also require attention to the many barriers health professionals face in identifying and managing incontinence.
Consistent with existing literature we found an association between continence problems and increased mortality and living in residential care. The finding that stroke survivors who have none to slight disability, and incontinence, were more likely to live in residential care than those who were moderately or severely disabled is counter-intuitive; the expectation being that those with more disability and continence problems are more likely to need residential care on discharge. This may possibly reflect the nature of the data collection, and is worth exploration as effective continence management in those with less disability may enable continued community living.
Conclusion
The prevalence of documented in-hospital post stroke incontinence in New Zealand is either much less than international estimates or was, more likely, underestimated in this secondary analysis of REGIONS Care data. Prevalence might not be associated with ethnicity, although more detailed incontinence data—such as type and severity—are needed to investigate the association more closely. Documented incontinence was associated with increased mortality and probability of living in residential care at discharge, three, six and twelve months. As might be expected, those with documented incontinence have reduced health utility scores compared to those without documented incontinence.
View Appendix 1.
Summary
Abstract
Aim
To estimate the prevalence of incontinence after stroke in Aotearoa New Zealand overall and by ethnicity, the associations between incontinence and subsequent mortality and living in residential care, and to estimate the health utilities in relation to continence.
Method
Secondary analysis of data from a prospective (1 May to 31 July 2018) cohort study (REGIONS Care study) of patients with a confirmed stroke admitted to New Zealand hospitals. Logistic and linear regression were used, and multivariate models were adjusted for age, sex, ethnicity, and stroke severity. The association between living in residential care, incontinence, and mobility was also assessed.
Results
There were 320/2,377 (13.5%) patients with documented incontinence during hospitalisation after stroke. Incontinence was not associated with ethnicity but was associated with increased mortality/living in residential care, at discharge, three, six and twelve months after stroke. Stroke survivors with independent mobility were more likely to live in residential care if incontinent. Health utility scores were lower at three, six and twelve months for those with incontinence after stroke.
Conclusion
This study likely underestimated incontinence prevalence after stroke, although incontinence was associated with increased mortality and probability of living in residential care.
Author Information
Acknowledgements
Correspondence
Correspondence Email
Competing Interests
1) Brocklehurst JC, Andrews K, Richards B, Laycock PJ. Incidence and correlates of incontinence in stroke patients. J Am Geriatr Soc. 1985;33(8):540-2.
2) Thomas LH, Coupe J, Cross LD, Tan AL, Watkins CL. Interventions for treating urinary incontinence after stroke in adults. Cochrane Database of Systematic Reviews. 2019(2).
3) Barer D. Continence after stroke: useful predictor or goal of therapy? Age Ageing. 1989;18(3):183-91.
4) Kolominsky‐Rabas PL, Hilz MJ, Neundoerfer B, Heuschmann PU. Impact of urinary incontinence after stroke: results from a prospective population‐based stroke register. Neurourol Urodyn. 2003;22(4):322-7.
5) Nakayama H, Jørgensen H, Pedersen P, Raaschou H, Olsen T. Prevalence and risk factors of incontinence after stroke: the Copenhagen Stroke Study. Stroke. 1997;28(1):58-62.
6) Ween JE, Alexander MP, D'Esposito M, Roberts M. Incontinence after stroke in a rehabilitation setting: outcome associations and predictive factors. Neurology. 1996;47(3):659-63.
7) Brittain K, Perry S, Peet S, Shaw C, Dallosso H, Assassa R, et al. Prevalence and impact of urinary symptoms among community-dwelling stroke survivors. Stroke. 2000;31(4):886-91.
8) Patel M, McKevitt C, Lawrence E, Rudd A, Wolfe C. Clinical determinants of long-term quality of life after stroke. Age Ageing. 2007;36(3):316-22.
9) Brittain K, Castleden C. Suicide in patients with stroke: Depression may be caused by symptoms affecting lower urinary tract. BMJ. 1998 Oct 10;317(7164):1016-7.
10) Brittain KR, Shaw C. The social consequences of living with and dealing with incontinence—A carers perspective. Soc Sci Med. 2007 Sep;65(6):1274-83.
11) Patel M, Coshall C, Rudd AG, Wolfe CD. Natural history and effects on 2-year outcomes of urinary incontinence after stroke. Stroke. 2001;32(1):122-7.
12) Rotar M, Blagus R, Jeromel M, Škrbec M, Tršinar B, Vodušek DB. Stroke patients who regain urinary continence in the first week after acute first‐ever stroke have better prognosis than patients with persistent lower urinary tract dysfunction. Neurourol Urodyn. 2011;30(7):1315-8.
13) Harari D, Coshall C, Rudd AG, Wolfe CD. New-onset fecal incontinence after stroke: prevalence, natural history, risk factors, and impact. Stroke. 2003;34(1):144-50.
14) Schluter PJ, Ward C, Arnold EP, Scrase R, Jamieson HA. Urinary incontinence, but not fecal incontinence, is a risk factor for admission to aged residential care of older persons in New Zealand. Neurourol Urodyn. 2017;36(6):1588-95.
15) Borrie MJ, Campbell AJ, Caradoc-Davies TH, Spears GF. Urinary incontinence after stroke: a prospective study. Age Ageing. 1986;15(3):177-81.
16) Feigin VL, Barker-Collo S, Parag V, Senior H, Lawes C, Ratnasabapathy Y, et al. Auckland Stroke Outcomes Study: Part 1: Gender, stroke types, ethnicity, and functional outcomes 5 years poststroke. Neurology. 2010 Nov 2;75(18):1597-607.
17) Stroke Foundation of New Zealand and New Zealand Guidelines Group. New Zealand clinical guidelines for stroke management 2010. Wellington: Stroke Foundation of New Zealand; 2010.
18) Thompson S, Barber PA, Fink J, Gommans J, Davis A, Harwood M, et al. New Zealand hospital stroke service provision. New Zealand hospital. 2020;133(1526).
19) Stroke Foundation. National stroke audit - acute services report 2019. 2019.
20) Ranta A, Thompson S, Harwood MLN, Cadilhac DA-M, Barber PA, Davis AJ, et al. Reducing Ethnic and Geographic Inequities to Optimise New Zealand Stroke Care (REGIONS Care): Protocol for a Nationwide Observational Study. JMIR research protocols. 2021;10(1):e25374.
21) Thompson SG, Barber PA, Gommans JH, Cadilhac DA, Davis A, Fink JN, et al. The impact of ethnicity on stroke care access and patient outcomes: a New Zealand nationwide observational study. Lancet Reg Health West Pac. 2022 Jan 3;20:100358.
22) Ranta A, Thompson S, Davis A, Barber PA, Fink J, Gommans J, et al. Urban versus Non-Urban Hospital Location Impacts on Stroke Outcomes. Stroke. 2021;52(Suppl_1):A29-A.
23) Counsell C, Dennis M, McDowall M. Predicting functional outcome in acute stroke: comparison of a simple six variable model with other predictive systems and informal clinical prediction. Journal Neurol Neurosurg Psychiatry 2004;75(3):401-405.
24) Devlin N, Parkin D, Janssen B. Methods for analysing and reporting EQ-5D data: Springer Nature; 2020.
25) Trad W, Flowers K, Caldwell J, Sousa MS, Vigh G, Lizarondo L, et al. Nursing assessment and management of incontinence among medical and surgical adult patients in a tertiary hospital: A best practice implementation project. JBI Database System Rev Implement Rep. 2019 Dec;17(12):2578-2590.
26) Tooher R, Middleton P, Pham C, Fitridge R, Rowe S, Babidge W, et al. A systematic review of strategies to improve prophylaxis for venous thromboembolism in hospitals. Ann Surg. 2005 Mar;241(3):397-415.
27) Sanai Memon G, Imam A, Datta-Chaudhuri M, Robertson E, Frain C. Improving identification and assessment of urinary incontinence in older people. Age Ageing. 2020;49.
28) Dutta D, Thornton D, Bowen E. Using population-based routinely collected data from the Sentinel Stroke National Audit Programme to investigate factors associated with discharge to care home after rehabilitation. Clin Rehabil. 2018 Aug;32(8):1108-1118.
29) Lennan M, Smalldridge J, Fa'alau F. Urinary incontinence: a qualitative approach: Department of Women's Health, South Auckland Health; 1999.
30) TuiSamoa A, Heather M, Kruger J. Urinary incontinence in Pasifika women: A pilot focus group study. TheAustralian and New Zealand Continence Journal. 2022;28(1):4-8.
31) Esplin J, Smith J, Doust E, Poynton M. Report on Good Practice of Continence Services in New Zealand Report. 2017.
32) Chen P, Lin K-C, Liing R-J, Wu C-Y, Chen C-L, Chang K-C. Validity, responsiveness, and minimal clinically important difference of EQ-5D-5L in stroke patients undergoing rehabilitation. Quality of life research. 2016;25(6):1585-96.
For the PDF of this article,
contact nzmj@nzma.org.nz
Documented incontinence after stroke: a secondary analysis of a cohort study. Reducing Ethnic and Geographic Inequities to Optimise New Zealand Stroke Care (REGIONS Care)
View Article PDF
Incontinence after stroke is a common problem. Additionally, urinary incontinence and faecal incontinence are both associated with more severe stroke.[[1]] Urinary incontinence after stroke may be caused by the direct effects of the cerebral lesion on neurological bladder control physiology, but may also be caused by motor (mobility and dexterity), cognitive (apraxia and agnosia), visual, and language effects of the stroke that adversely affect toileting.[[2]] The reported prevalence of urinary and faecal incontinence after stroke varies by study but it is likely that more than half of patients may experience urinary incontinence,[[3,4]] and up to 40% have faecal incontinence in the acute stage after stroke.[[5]] Urinary incontinence after stroke is associated with poor functional outcomes,[[3,6]] lower quality of life,[[7–10]] more social isolation and depression,[[10]] greater likelihood of living in residential care,[[4,6,11]] and greater mortality.[[4,11,12]] There is evidence that faecal incontinence after stroke is associated with increased mortality,[[5,13]] and likelihood of living in residential care.[[13]] This contrasts with a New Zealand study of those without stroke, which reported that in older adults urinary but not faecal incontinence was a risk factor for admission to residential care.[[14]]
In Aotearoa New Zealand, two studies report on incontinence after stroke. A 1986 study reported the prevalence of urinary incontinence after stroke was high and depended on the time of assessment after stroke. In that study 60%, 42% and 29% of survivors after stroke, had urinary incontinence one, four and 12 weeks after stroke, respectively.[[15]] The Auckland Stroke Outcomes (ASTRO) study reported that 27% of stroke survivors had “bladder control problems” five years after stroke.[[16]] The prevalence of incontinence for different ethnic groups after stroke is unclear. The ASTRO study reported that there was no statistical evidence of a difference in bladder control problems after stroke in relation to ethnicity, although a higher proportion, 36% of Pacific peoples, reported problems with bladder control, compared to 27% of participants overall. The number of Pacific peoples in the study was small (n=30), and the study likely lacked statistical power to definitively assess whether a meaningful difference in urinary incontinence prevalence in relation to ethnicity was present or not.
The Australian and New Zealand Clinical Guidelines for Stroke Management recommend that all patients with stroke who may have continence problems should have a “structured functional assessment”, and those with incontinence should have a continence management plan implemented.[[17]] A recent survey of stroke service provision in New Zealand reported that 74% of hospitals that routinely treat patients with acute stroke had a protocol or guideline in place to manage urinary incontinence,[[18]] and 59% had one for managing faecal incontinence (unpublished data, 2020).[[18]] In Australia, a 2019 acute stroke services clinical audit reported that although 32% of patients had urinary incontinence documented, only 37% of these patients had a documented continence management plan.[[19]]
There is a lack of contemporary data about the prevalence and management of incontinence after stroke in New Zealand to inform planning and provision of services aimed to enable people with stroke to live well, for longer, in the community. This study is a secondary analysis of data collected for the Reducing Ethnic and Geographic Inequalities to Optimise New Zealand Stroke (REGIONS) Care study.[[20]] The current study aims were to: 1) estimate the prevalence of incontinence after stroke; 2) explore if the prevalence of incontinence after stroke was related to ethnicity including after adjustment for other important predictors of incontinence such as age and stroke severity; 3) estimate the strength of association, adjusted for important predictors of incontinence, between mortality and/or living in residential care with incontinence; 4) assess whether mobility status is an effect modifier to the association between living in residential care and continence; and 5) estimate the health utilities in patients with and without incontinence.
Methods
REGIONS Care is a nationwide study assessing the impact of hospital geographic location and ethnicity on stroke outcomes and access to best practice care. Full study methods and primary analyses for REGIONS Care are described elsewhere.[[20–22]]
In brief, REGIONS Care was a prospective, observational cohort study of all patients with a confirmed diagnosis of stroke who were admitted to the 28 hospitals that treat patients with acute stroke between 1 May and 31 July 2018. After this date, data collection continued until hospitals had achieved a minimum sample size of 150 (for stroke clot retrieval centres) or 100 for other centres, or until 31 October 2018—whichever occurred first. Data (baseline characteristics; acute treatment, investigations, and rehabilitation) up to and including the three months post stroke follow-up were collected by hospital staff. As registry-based data collection this process was considered audit and did not require individual patient consent. Patients were then contacted and were invited to consent to extended follow-ups at six and twelve months, data linkage with health administrative data, focus groups and surveys (Central Health and Disability Ethics Committee, reference 17/CEN/164).
Selected participant baseline characteristics, and possible predictors of incontinence (any type) after stroke were extracted from the full data set. These included: age, sex, ethnicity, stroke risk factors, primary diagnosis, pre-morbid level of function, pre-stroke living situation, and stroke severity at time of hospital admission using the six simple variables (SSV) model (a composite of age, living alone pre-stroke, independence in activities of daily living pre-stroke, Glasgow Coma Scale verbal component, arm power, and ability to walk unaided at admission).[[23]] Patient outcomes at discharge were death and discharge destination (home or residential care), and at three, six and twelve months, outcomes were death or living in residential care which is a composite measure for a poor outcome that encompasses both severe functional deficits and mortality. The composite measure is needed because a high proportion of those with stroke die, and so an outcome that just depends on residential care status will have missing data. As a potential effect modifier, mobility status (independent or dependent), was determined at each time point from the relevant modified Rankin Scale (mRS) score. The dichotomisation (mRS 0 to 3; mRS 4 or 5) was made to capture different levels of disability likely to influence the functional aspects of toileting, particularly walking to the toilet independently. People with a mRS score of 4 or 5 require external assistance for walking or are bedridden respectively. The EuroQol EQ-5D-3L quality of life measure was used to estimate Health Utility Scores which are scaled as a score from zero, representing a quality of life that is equivalent to death, to one, representing perfect health.[[24]]
Incontinence-related variables in the REGIONS Care dataset were any documented incontinence (any type), documented continence management plan, use of an indwelling catheter (IDC), urinary tract infection (UTI) and constipation post stroke, during acute care and before discharge; all these were collected as yes/no responses. If an IDC was used, a reason could be stated. Data collection did not distinguish between urinary or faecal incontinence, or existence of both. Incontinence data were collected in a non-compulsory data field related to stroke complications; incontinence may have been documented in clinical notes but inconsistently in study data.
Continuous variables are summarised by mean and standard deviation (SD), and median and interquartile (IQR) range, as appropriate. Counts are summarised by proportions and appropriate confidence intervals for a proportion. Logistic regression examines associations with dichotomous outcome variables and linear regression with continuous outcome variables. Multivariate models were adjusted for age, sex, ethnicity, and stroke severity as confounders and the association between living in residential care and incontinence main effects and mobility assessed by appropriate interaction terms. Stata/IC 16.0 was used for analysis.
The Health Research Council of New Zealand (HRC 17/037) funded the REGIONS Care study and the study received ethics approval from the Central Region Health and Disability Ethics Committee (17CEN164).
Results
The baseline characteristics of patients are summarised in Table 1. There were 2,379 patients with stroke admitted to hospital during the study period, of whom 320 had documented incontinence and 2057 did not; two did not have recorded data about incontinence. The mean (SD) age was 75 (13.7) years; 1,219/2,379 (51.2%) were male; and 2,052/2,343 (87.6%) with a recorded stroke type had ischemic stroke. Baseline characteristics of patients with and without incontinence (any type) are shown in Table 1.
The overall prevalence of incontinence problems and related issues are summarised in Table 2.
The prevalence of incontinence and related variables by ethnicity is summarised in Appendix 1. There was no evidence that the proportion of those with incontinence was associated with ethnicity in univariate (P=0.61) or multivariate analysis adjusting for age, sex and stroke severity, P=0.58. The univariate and multivariate adjusted odds ratios (OR) are shown in Table 3 with NZ European ethnicity as the reference level.
Patients with documented incontinence were more likely to die (52/320 (16.3%)) or be discharged to residential care (118 (36.9%)) than those with no documented incontinence (9.6% and 11.8% respectively). Table 4 shows the summary data, unadjusted and adjusted (for age, sex, and stroke severity) odds ratios for the individual outcomes at discharge, and the composite outcomes at the subsequent time points, in relation to documented incontinence.
Disability on discharge, for those who were discharged alive, was an effect modifier for the risk of entering residential care. In a univariate analysis 247/360 (68.6%) of those discharged to residential care had a mRS of 4 or 5 (includes lacking independent mobility) compared to 159/1,749 (9.1%) of those discharged to the community, OR (95% CI) 18.5 (13.9–24.7), P<0.001. For the same group of those discharged alive, 118 (32.8%) of those discharged to residential care had incontinence compared to 149 (8.5%) of those discharged to the community; OR (95% CI) 1.87 (1.32–2.65), P<0.001. However, the P-value for the interaction between disability and incontinence in relation to discharge status was statistically significant (P<0.001) indicating that a person with less disability was more likely to be discharged to residential care if incontinent. Table 5 shows the proportions of those with incontinence amongst those with more versus less disability discharged to residential care or the community. In those with less disability, incontinence was strongly related to discharge to residential care; but there was no association in those with more disability.
The mean (SD) health utility scores for patients with documented incontinence were 0.52 (0.28) at three months, 0.55 (0.28) at six months and 0.53 (0.27) at 12 months; compared to 0.71 (0.23) at three months, 0.71 (0.24) at six months and 0.70 (0.25) at 12 months for patients without documented incontinence. The differences in mean health utility score by continence status in the univariate modelling was -0.19 (95% CI -0.23 to -0.15) at three months, -0.16 (95% CI -0.21 to -0.10) at six months, and -0.17 (95% CI -0.24 to -0.12) at 12 months. In the multivariate model (adjusted for age, sex, ethnicity and stroke severity), the differences were -0.11 (-0.15 to -0.07), -0.07 (-0.13 to -0.02), and -0.10 (-0.16 to -0.04) respectively. For patients who reported no problem on each of the five EQ-5D-3L dimensions, thus scoring a health utility score of 1 (meaning full health), the mean (SD) EQ-VAS scores at three, six and twelve months were 82.3 (13.7), 82.5 (15.5) and 84.2 (11.8) respectively.
View Tables 1–5.
Discussion
In this secondary analysis of documented incontinence (or not) in a large cohort of hospital admission for stroke we found a much lower incontinence prevalence than reported in previous New Zealand[[15,16]] or more recent international research.[[4,5]] Documented incontinence included any type or severity (i.e., any frequency or volume of leakage) at any stage in acute care before discharge. It is difficult to explain why there is a markedly reduced prevalence of post stroke incontinence in New Zealand compared to international data. It seems unlikely this reflects pathophysiological differences or acute care. It seems more likely that this difference reflects a lower rate of identifying patients with incontinence or at least inconsistent documentation of incontinence in the inpatient record,[[25]] rather than being a true reflection of the prevalence of incontinence post stroke in New Zealand. When incontinence was documented, nearly three quarters of patients had a continence management plan documented too. The data collection did not allow investigation of whether incontinence, or a continence plan, was more common if the incontinence was more severe, or the person had double (urinary and faecal) incontinence. If impact on nursing is greater, e.g., bed changes, or use of containment products, it is possible that documentation is more likely. The low reported prevalence may thus be focussed on the proportion of stroke survivors with severe incontinence.
Stroke guidelines recommend that all patients with suspected incontinence are assessed.[[14]] The low prevalence of documented incontinence in this cohort, compared to international prevalence reports, raises concerns that not all patients with incontinence were identified and the possibility that some may have experienced unmet needs in relation to continence management. This may represent an important area for ongoing service improvement efforts, such as those found to be useful for increasing guideline adherence in other areas of stroke care[[26]] and incontinence care[[25,27]] in medical and older persons wards.
We found that those with documented incontinence had increased risk of being discharged to residential care, and increased risk of being deceased or living in residential care three, six and twelve months after stroke. These findings align with previous research reporting that incontinence after stroke is associated with poorer outcomes.[[4,6,11,12]] Disability was an effect modifier for the association between incontinence and living in residential care. Thus, incontinence increased the risk of moving into residential care for those stroke survivors with less disability, but not for those with moderate or severe disability. It is possible those with greater dependency are already more likely to discharge to residential care. The Sentinel Stroke National Audit in the UK found that a mRS of 4 or 5 at discharge was strongly associated with first time residential placement.[[28]] In contrast to our findings Dutta et al, in a subgroup analysis of stroke patients with an mRS of 4 or 5, found incontinent patients had nearly five times the odds of discharge to residential care.[[28]]
There was no association of documented incontinence with ethnicity. A previous study, The Auckland Stroke Outcomes (ASTRO) study, did find a higher prevalence of incontinence in Pacific peoples compared to the sample population overall. Two qualitative studies, both about women’s perspectives on urinary incontinence, have specifically sought experiences of Māori and Pacific women. Lennan et al. (1999)[[29]] sought to understand the impacts of urinary incontinence on psychological, social and medical wellbeing—the study included eight Māori, nine Pacific, and 17 Pākehā women—and found that Māori and Pacific women were more likely to say there were barriers to help seeking, including language barriers and trust and confidentiality issues. Having a trust relationship and being assured of confidentiality were the most important attributes of the doctor for Māori and Pacific women. TuiSamoa et al. (2022)[[30]] carried out a single focus group with 10 Pasifika women, not all of whom had incontinence. The findings were consistent with those from Lennan et al., with the addition of understanding that culturally safe care may vary according to the woman’s generation. As continence services in New Zealand are limited and inequitably distributed,[[31]] further study of possible inequity in access, experience and outcomes for incontinence care after stroke may be warranted.
Participants with documented incontinence had consistently reduced health utility scores (closer to zero) than those without documented incontinence at three, six-and twelve months post stroke. The mean differences (after adjusting for age, sex, ethnicity and stroke severity) were consistent with minimal clinically important differences in EuroQoL utilities reported in stroke research.[[32]] This suggests additional and potentially important clinical impact, attributable to incontinence, on health-related quality of life after stroke.
The strength of this analysis is that the data came from a large prospective cohort of patients with stroke. There are several limitations. First, the REGIONS Care study was not designed to specifically investigate incontinence after stroke. Although documentation of incontinence (yes/no) was not a compulsory data collection field, the data were missing for only two of the 2,377 participants. Second, there is the lack of detail about incontinence meaning it was not possible to estimate prevalence of urinary, faecal, or double incontinence, or pre-existing incontinence symptoms, how severe incontinence symptoms were, and the influence of any of these on the associations we examined. Third, the ceiling effects of the EQ-5D-3L, may have influenced the estimated health utility scores. Most patients who reported no difficulties across any of the five dimensions of the questionnaire, thus scoring a 1 for the health utility score, corresponding to “full health”, did not score their overall health at 100 on the EQ-VAS, with mean scores ranging from 82.3 to 84.2 at three and twelve months respectively. The five level version of the EQ-5D may have increased the sensitivity and reduced some of the observed ceiling effects in the three level version.
The low prevalence of documented incontinence after stroke compared to international estimates may indicate that incontinence is not reliably identified, which means some stroke survivors may have unmet continence management needs. Given the data limitations outlined above, a more detailed investigation of post stroke incontinence is needed, distinguishing urinary and faecal incontinence, estimating severity using standardised and internationally validated continence assessment instruments, and to clarify if incontinence is new or pre-existing. A better description of care received, and unmet needs would also clarify the effect of incontinence on stroke survivors and their carers. These data are needed to understand the scale of the need continence services are trying to meet and make reasonable estimates of need with population ageing. However, collecting these data may also require attention to the many barriers health professionals face in identifying and managing incontinence.
Consistent with existing literature we found an association between continence problems and increased mortality and living in residential care. The finding that stroke survivors who have none to slight disability, and incontinence, were more likely to live in residential care than those who were moderately or severely disabled is counter-intuitive; the expectation being that those with more disability and continence problems are more likely to need residential care on discharge. This may possibly reflect the nature of the data collection, and is worth exploration as effective continence management in those with less disability may enable continued community living.
Conclusion
The prevalence of documented in-hospital post stroke incontinence in New Zealand is either much less than international estimates or was, more likely, underestimated in this secondary analysis of REGIONS Care data. Prevalence might not be associated with ethnicity, although more detailed incontinence data—such as type and severity—are needed to investigate the association more closely. Documented incontinence was associated with increased mortality and probability of living in residential care at discharge, three, six and twelve months. As might be expected, those with documented incontinence have reduced health utility scores compared to those without documented incontinence.
View Appendix 1.
Summary
Abstract
Aim
To estimate the prevalence of incontinence after stroke in Aotearoa New Zealand overall and by ethnicity, the associations between incontinence and subsequent mortality and living in residential care, and to estimate the health utilities in relation to continence.
Method
Secondary analysis of data from a prospective (1 May to 31 July 2018) cohort study (REGIONS Care study) of patients with a confirmed stroke admitted to New Zealand hospitals. Logistic and linear regression were used, and multivariate models were adjusted for age, sex, ethnicity, and stroke severity. The association between living in residential care, incontinence, and mobility was also assessed.
Results
There were 320/2,377 (13.5%) patients with documented incontinence during hospitalisation after stroke. Incontinence was not associated with ethnicity but was associated with increased mortality/living in residential care, at discharge, three, six and twelve months after stroke. Stroke survivors with independent mobility were more likely to live in residential care if incontinent. Health utility scores were lower at three, six and twelve months for those with incontinence after stroke.
Conclusion
This study likely underestimated incontinence prevalence after stroke, although incontinence was associated with increased mortality and probability of living in residential care.
Author Information
Acknowledgements
Correspondence
Correspondence Email
Competing Interests
1) Brocklehurst JC, Andrews K, Richards B, Laycock PJ. Incidence and correlates of incontinence in stroke patients. J Am Geriatr Soc. 1985;33(8):540-2.
2) Thomas LH, Coupe J, Cross LD, Tan AL, Watkins CL. Interventions for treating urinary incontinence after stroke in adults. Cochrane Database of Systematic Reviews. 2019(2).
3) Barer D. Continence after stroke: useful predictor or goal of therapy? Age Ageing. 1989;18(3):183-91.
4) Kolominsky‐Rabas PL, Hilz MJ, Neundoerfer B, Heuschmann PU. Impact of urinary incontinence after stroke: results from a prospective population‐based stroke register. Neurourol Urodyn. 2003;22(4):322-7.
5) Nakayama H, Jørgensen H, Pedersen P, Raaschou H, Olsen T. Prevalence and risk factors of incontinence after stroke: the Copenhagen Stroke Study. Stroke. 1997;28(1):58-62.
6) Ween JE, Alexander MP, D'Esposito M, Roberts M. Incontinence after stroke in a rehabilitation setting: outcome associations and predictive factors. Neurology. 1996;47(3):659-63.
7) Brittain K, Perry S, Peet S, Shaw C, Dallosso H, Assassa R, et al. Prevalence and impact of urinary symptoms among community-dwelling stroke survivors. Stroke. 2000;31(4):886-91.
8) Patel M, McKevitt C, Lawrence E, Rudd A, Wolfe C. Clinical determinants of long-term quality of life after stroke. Age Ageing. 2007;36(3):316-22.
9) Brittain K, Castleden C. Suicide in patients with stroke: Depression may be caused by symptoms affecting lower urinary tract. BMJ. 1998 Oct 10;317(7164):1016-7.
10) Brittain KR, Shaw C. The social consequences of living with and dealing with incontinence—A carers perspective. Soc Sci Med. 2007 Sep;65(6):1274-83.
11) Patel M, Coshall C, Rudd AG, Wolfe CD. Natural history and effects on 2-year outcomes of urinary incontinence after stroke. Stroke. 2001;32(1):122-7.
12) Rotar M, Blagus R, Jeromel M, Škrbec M, Tršinar B, Vodušek DB. Stroke patients who regain urinary continence in the first week after acute first‐ever stroke have better prognosis than patients with persistent lower urinary tract dysfunction. Neurourol Urodyn. 2011;30(7):1315-8.
13) Harari D, Coshall C, Rudd AG, Wolfe CD. New-onset fecal incontinence after stroke: prevalence, natural history, risk factors, and impact. Stroke. 2003;34(1):144-50.
14) Schluter PJ, Ward C, Arnold EP, Scrase R, Jamieson HA. Urinary incontinence, but not fecal incontinence, is a risk factor for admission to aged residential care of older persons in New Zealand. Neurourol Urodyn. 2017;36(6):1588-95.
15) Borrie MJ, Campbell AJ, Caradoc-Davies TH, Spears GF. Urinary incontinence after stroke: a prospective study. Age Ageing. 1986;15(3):177-81.
16) Feigin VL, Barker-Collo S, Parag V, Senior H, Lawes C, Ratnasabapathy Y, et al. Auckland Stroke Outcomes Study: Part 1: Gender, stroke types, ethnicity, and functional outcomes 5 years poststroke. Neurology. 2010 Nov 2;75(18):1597-607.
17) Stroke Foundation of New Zealand and New Zealand Guidelines Group. New Zealand clinical guidelines for stroke management 2010. Wellington: Stroke Foundation of New Zealand; 2010.
18) Thompson S, Barber PA, Fink J, Gommans J, Davis A, Harwood M, et al. New Zealand hospital stroke service provision. New Zealand hospital. 2020;133(1526).
19) Stroke Foundation. National stroke audit - acute services report 2019. 2019.
20) Ranta A, Thompson S, Harwood MLN, Cadilhac DA-M, Barber PA, Davis AJ, et al. Reducing Ethnic and Geographic Inequities to Optimise New Zealand Stroke Care (REGIONS Care): Protocol for a Nationwide Observational Study. JMIR research protocols. 2021;10(1):e25374.
21) Thompson SG, Barber PA, Gommans JH, Cadilhac DA, Davis A, Fink JN, et al. The impact of ethnicity on stroke care access and patient outcomes: a New Zealand nationwide observational study. Lancet Reg Health West Pac. 2022 Jan 3;20:100358.
22) Ranta A, Thompson S, Davis A, Barber PA, Fink J, Gommans J, et al. Urban versus Non-Urban Hospital Location Impacts on Stroke Outcomes. Stroke. 2021;52(Suppl_1):A29-A.
23) Counsell C, Dennis M, McDowall M. Predicting functional outcome in acute stroke: comparison of a simple six variable model with other predictive systems and informal clinical prediction. Journal Neurol Neurosurg Psychiatry 2004;75(3):401-405.
24) Devlin N, Parkin D, Janssen B. Methods for analysing and reporting EQ-5D data: Springer Nature; 2020.
25) Trad W, Flowers K, Caldwell J, Sousa MS, Vigh G, Lizarondo L, et al. Nursing assessment and management of incontinence among medical and surgical adult patients in a tertiary hospital: A best practice implementation project. JBI Database System Rev Implement Rep. 2019 Dec;17(12):2578-2590.
26) Tooher R, Middleton P, Pham C, Fitridge R, Rowe S, Babidge W, et al. A systematic review of strategies to improve prophylaxis for venous thromboembolism in hospitals. Ann Surg. 2005 Mar;241(3):397-415.
27) Sanai Memon G, Imam A, Datta-Chaudhuri M, Robertson E, Frain C. Improving identification and assessment of urinary incontinence in older people. Age Ageing. 2020;49.
28) Dutta D, Thornton D, Bowen E. Using population-based routinely collected data from the Sentinel Stroke National Audit Programme to investigate factors associated with discharge to care home after rehabilitation. Clin Rehabil. 2018 Aug;32(8):1108-1118.
29) Lennan M, Smalldridge J, Fa'alau F. Urinary incontinence: a qualitative approach: Department of Women's Health, South Auckland Health; 1999.
30) TuiSamoa A, Heather M, Kruger J. Urinary incontinence in Pasifika women: A pilot focus group study. TheAustralian and New Zealand Continence Journal. 2022;28(1):4-8.
31) Esplin J, Smith J, Doust E, Poynton M. Report on Good Practice of Continence Services in New Zealand Report. 2017.
32) Chen P, Lin K-C, Liing R-J, Wu C-Y, Chen C-L, Chang K-C. Validity, responsiveness, and minimal clinically important difference of EQ-5D-5L in stroke patients undergoing rehabilitation. Quality of life research. 2016;25(6):1585-96.
Contact diana@nzma.org.nz
for the PDF of this article
Documented incontinence after stroke: a secondary analysis of a cohort study. Reducing Ethnic and Geographic Inequities to Optimise New Zealand Stroke Care (REGIONS Care)
Urinary incontinence after stroke may be caused by the direct effects of the cerebral lesion on neurological bladder control physiology, but may also be caused by motor (mobility and dexterity), cognitive (apraxia and agnosia), visual, and language effects of the stroke that adversely affect toileting.
Subscriber Content
The full contents of this pages only available to subscribers.
Login, subscribe or email nzmj@nzma.org.nz to purchase this article.
