Study design, materials and methods
Design: A cross-sectional retrospective review of medical records of residents.
Participants: Residents residing at three publicly funded RACFs within a regional health service in Australia as of 1 June 2018.
Inclusion Criteria: Had lived at the facility on a permanent basis for at least 4 weeks prior to 1 June 2018.
Exclusion Criteria: Individuals with stomas.
Ethics / Consent. Obtained from the health service’s Human Research Ethics Committee. Individual consent was deemed not necessary.
Data collection: The principal investigator collected data from a computerised commercial-in-confidence software program used by the RACFs. Data collected included age, gender, date of arrival to the facility, body mass index (BMI), incident FI after admission to the RACF, history of FI, bowel frequency and consistency using the Bristol Stool Chart , history of urinary incontinence (UI), medical comorbidities, mobility and transfer capabilities, dementia status, current cognitive function, history of constipation, diet, functional activities of daily living and the ability to communicate the urge to defaecate.
Definition of FI: the involuntary loss of faeces, liquid or solid. (1)
Statistical analysis: Data were analysed using Stata Statistical Software. Categorical data were reported as frequencies with percentages whilst continuous data were summarised using medians with the interquartile range (IQR; 25th, 75th percentiles). The prevalence of FI was estimated as a percentage and reported together with the 95% confidence interval (CI). The relationship between FI and resident clinical/demographic characteristics was established using Poisson regression and the results were summarised using risk ratios (RR) with the 95% CI.
Results
Medical records of 107 residents were extracted and reviewed, 58 female and 49 male residents, with a mean age of 78.9 yo.
The prevalence of FI was 62.6 % (95% CI 52.7% - 71.8%). About two thirds (65.1 %) had documented FI at admission. Amongst those with FI the median number of FI episodes during the one-month study period was 7 (range 2-14).The median number of bowel opening episodes among residents without FI was 12 (IQR = 9, 17) over a 4-week period, compared with 14 (IQR = 10, 21) among those with FI. The most common type of stool passed in those with FI was Bristol Stool Form Chart types 4, 5 and 6. The majority (n=92, 87.6%) of residents had experienced constipation, of which 55 (59.8%) had FI (RR 0.78, 95% CI 0.55 – 1.10, p=0.150).
UI was present in 103 residents and 65.1% (n=67) had FI. All residents with FI had UI.
The median weight of residents with FI was an average of 12 kg lower than that of residents without FI. Consequently, BMI was significantly associated with presence of FI (p=0.005). The median BMI among residents with FI 28kg/m2 (IQR: 24.7 – 31.4) was significantly lower than that of residents without FI (median BMI = 29.8kg/m2; IQR 27.3 – 40.5).
Many of the residents (n=92, 86%) were cognitively impaired or had a diagnosis of dementia. About two-thirds of these (n=59, 64%) had FI. However, cognitive status was significantly associated with FI incidence. Three residents had active inflammatory bowel disease (IBD), three had irritable bowel syndrome (IBS) and they all had FI. The only other medical condition significantly associated with FI was osteoporosis. FI was not significantly associated with stroke, neurological disorders, hypertension, dementia, diabetes, depression and hypercholesterolaemia. Pelvic surgery was documented for 17 residents, and 10 had FI (p =0.736).
The residents’ median length of stay in years for non FI was 0.83 (0.46, 1.96) versus FI 1.92 (0.75, 4.83). FI increased the relative risk to 1.06 (95% CI 1.03 – 1.09) p<0.001.
FI was significantly associated with needing help with initiation, prompting to toilet, assistance with transfers, ambulating and needing help with toileting and hygiene.
Interpretation of results
The incidence of FI was high compared to previous studies of FI in RACF in other countries (2). This may correlate with RACF residents in Australia generally becoming more dependent and frail on admission. The majority of FI stool consistencies were of types 4 to 6 on the Bristol Stool Chart which may suggest this lower consistency being a risk factor in these residents consistent previous finding of diarrhoea associated with FI.
The main factors that increased the likelihood of FI were older age, longer length of time residing in the RACF, lower BMI and lower weight, and the need for assistance with toileting. The only medical conditions associated with FI were IBS and acute IBD and osteoporosis. IBS and IBD may cause increased stool frequency and reduced consistency increasing risk of FI. Osteoporosis does not directly cause FI but may reflect age and frailty.
The association between length of residence and the prevalence of FI aligns with prior research. A possible explanation for this could be accumulation of comorbidities or frailty impacting on mobility and ability to manage self-care.
Other studies demonstrated findings of cognitive impairment/dementia, constipation and urinary incontinence as major risk factors of FI but these were not found to be statistically significant in this study.(2)