Candida albicans colonisation, continence status and incontinence‐associated dermatitis in the acute care setting: a pilot study

Abstract

Candida albicans is the most prevalent human fungal commensal organism and is reported to be the most frequent aetiological organism responsible for infection associated with incontinence‐associated dermatitis. However, it remains unclear whether incontinence predisposes a patient to increased Candida colonisation or whether incontinence acts as a trigger for Candida infection in those already colonised. The purpose of this observational cross‐sectional study was to estimate colonisation rates of C. albicans in continent, compared to incontinent patients, and patients with incontinence‐associated dermatitis. Data were collected on 81 inpatients of a major Australian hospital and included a pelvic skin inspection and microbiological specimens to detect C. Albicans at hospital admission. The mean age of the sample was 76 years (SD = 12.22) with 53% being male. Incontinent participants (n = 53) had a non‐significant trend towards greater Candida colonisation rates at the perianal site (43% versus 28%) χ ²(1, N = 81) = 4·453, p = ·638 and the inguinal site (24% versus 14%) χ ²(1, N = 81) = 6·868, p = ·258 compared to continent patients (n = 28). The incontinent subgroup with incontinence‐associated dermatitis (n = 22) showed no difference in colonisation rates compared to those without incontinence‐associated dermatitis. Understanding the epidemiology of colonisation may have implications for the prevention of Candida infection in these patients.

Introduction

Candida, a genus of yeasts, is the most common fungal commensal organism in humans, and is responsible for a wide range of mucosal and invasive diseases 1, 2, 3. Systemic infection by Candida (predominately Candida albicans) is the leading cause of nosocomial fungal bloodstream infection, with a mortality rate of 30–40% 1, 4, 5. Candida albicans is the most common fungal human colonizer, with oropharangeal and vulvovaginal the most frequent Candida mucosal infections. These mucosal Candida infections are responsible for considerable morbidity, discomfort and decreased quality of life for affected individuals 6. For example, 65% of denture wearers will experience Candida‐associated denture stomatitis, while up to 75% of women will experience an episode of vulvovaginal Candida infection 2, 6.

Candida species (spp.) form part of the normal mucosal microflora in up to 80% of individuals 7. Candida albicans can colonise several anatomical sites, with the organism being present in the gastrointestinal and genito‐urinary tracts of up to 70% of individuals 3 Under certain conditions, Candida can transform from an innocuous commensal organism to a harmful pathogen 2. The primary source of both invasive and superficial Candida infection is endogenous, with colonisation being a necessary antecedent to infection 7, 8, 9. The high frequency of occurrence, morbidity and mortality associated with systemic and mucosal Candida infections, coupled with the ability of the organism to grow in a wide range of conditions means that Candida infections are important human diseases 2, 7.

In adults, Candida infections have been associated with a condition known as incontinence‐associated dermatitis (IAD) 10, 11, 12, 13. IAD is an irritant contact dermatitis that can occur in adults with urinary and/or faecal incontinence. The condition is characterised by inflammation and erythema and may present with blisters, erosion or serous exudate 10, 11. IAD is a painful, complex condition often resulting in increased morbidity and length of hospital stay 11, 14. Commonly, IAD is complicated by a fungal infection, with C. albicans being reported as the most frequent aetiological organism 11, 12. Candida infections associated with incontinence or IAD can either manifest as a bright red maculopapular rash spreading from a central area, with characteristic satellite lesions at the margins of the rash extending into normal skin, or non‐specific confluent papules, making clinical diagnosis difficult 10, 15.

It has been reported that Candida infections play a central role in the pathogenesis of diaper dermatitis in infants 16, 17, 18, 19, 20, 21, 22. Several studies have shown significantly higher rates of Candida colonisation in oral, inguinal and perianal sites in infants with diaper dermatitis than those without, and colonisation rates correlated with the severity of diaper dermatitis 17, 20. However, the role of Candida colonisation or infection has not been widely considered in the pathogenesis of adult IAD, despite the reported frequency of Candida infections as a complication of the condition 10, 11, 12, 13. It remains unclear whether incontinence predisposes to increased Candida colonisation or whether incontinence acts as a trigger for Candida to transform from a commensal to infective pathogen in those already colonised. Furthermore, it is unclear whether Candida plays a role in the pathogenesis of IAD in incontinent individuals. Understanding the epidemiology of incontinence‐associated Candida infections requires accurate estimates of Candida colonisation in continent compared to incontinent patients, as well as those with IAD. Knowledge of the epidemiology of Candida colonisation in these situations may guide assessment and prevention strategies for these patients.

Aims and research questions

The purpose of this study was to identify the rate of C. albicans colonisation in incontinent patients compared to continent patients and those with IAD. Specific research questions were:

1. Is colonisation with C. albicans more common at hospital admission in incontinent patients compared with continent patients?

2. In incontinent patients (urinary or urinary and faecal), is there an association between C. albicans colonisation and clinical presentation of IAD at admission to hospital?

Methods

Study design

An observational cross‐sectional study design was used.

Study setting and sample

This study was conducted in the Internal Medicine Service (IMS) wards of a 929‐bed major acute care metropolitan teaching hospital in Australia. The facility provides a comprehensive range of specialities, delivering care in 2012 to almost 94 000 inpatients, 520 000 outpatients and 72 000 emergency department presentations. The IMS has approximately 240 beds and is comprised of over 16 specialty medical areas, including general medicine, cardiology, dermatology, endocrinology and diabetes, infectious diseases, renal, rheumatology and thoracic medicine.

This study used a purposive sample of participants in three groups: (i) continent, (ii) incontinent of urine, but continent of faeces and (iii) incontinent of both urine and faeces.

Inclusion and exclusion criteria

Hospitalised adults aged ≥18 years admitted consecutively to all wards in the IMS who had an expected hospital stay of ≥3 days and where data could be collected within 48 hours of admission were eligible for inclusion in the study. Excluded patients were those who had current treatment with systemic or topical antifungal agents, a pre‐existing skin condition, those receiving radiation treatment to the pelvic area or those who were pregnant or breastfeeding.

Procedures

Patient recruitment and enrolment took place from April–October 2014. Prior to data collection, all research nurses (n = 3) were trained in the study procedures, use of the data collection instrument, pelvic skin inspection, identification of IAD, clinical presentation of Candida infection and pressure ulcers as well as the collection of microbiological specimens. Research nurses screened and enrolled patients admitted consecutively to the IMS. The research nurses visited each medical ward Monday to Friday, where the Nurse Manager or shift co‐ordinator identified patients admitted within the last 48 hours. If the newly admitted patient met the study inclusion criteria (this was determined by the bedside nurse together with the research nurse), the patient was asked by the bedside nurse if they were willing to receive information about the study from the research nurse. Data were collected within 48 hours of admission and extracted from medical records, with supplementary information sought from the patient and the bedside nurse, a pelvic skin inspection (perianus, perineum, groins, buttocks, labia/scrotum and lower abdomen) and collection of microbiological specimens.

Each patient had two microbiological specimens collected with a dry flocked swab (without transport medium) pre‐moistened in an ampoule of saline (as per laboratory protocol). The inguinal and perianal skin was sampled using separate swabs. The perianal specimen was collected from the anal verge, with any fecal matter removed from the skin prior to sampling.

Specimens were cultured directly in Sabouraud's medium for yeasts and incubated at 37°C (98·6°Fahrenheit) for 48–72 hours. Chloramphenicol and gentamycin were added to the medium to inhibit bacterial growth on the plate. A germ tube test was performed to identify C. albicans. The ability to form germ tubes after incubation in serum for 2 hours at 37°C (98·6°Fahrenheit) is presumptive evidence of C. albicans 7, 19. The laboratory reports specified either C. albicans or non‐C. albicans spp. Semi‐quantitative results were reported as follows: scant, 1+, 2+ and 3+. The laboratory did not culture for other pathogens. Results were entered into the facility pathology reporting system. Routine hospital reporting procedures were followed for those patients identified with IAD, Candida infection or any adverse findings.

Measures

Candida Colonisation

In this study, the presence of any C. albicans or Candida spp. at either the perianal or inguinal sites was classified as colonised 2, 23, 24.

Incontinence

Incontinence is the primary risk factor for the development of IAD 10, 11. It is defined as the inability to control the flow of urine and or faeces at any time in the previous 24 hours. In this study, patients with an indwelling urinary catheter were deemed continent for the purpose of data analysis. The presence of an indwelling catheter negates the irritant effects of urine on the skin. This categorisation is consistent with previous studies 25, 26, 27. Patients were recruited according to continence status, that is, continent, incontinent of urine only or incontinent of both urine and faeces (double incontinence). Continence status was determined by a combination of patient self‐report, clinical documentation and information from the bedside nurse.

Incontinence‐associated dermatitis (IAD)

IAD is defined as the presence of any skin redness and/or erosion caused by skin contact with urine and /or faeces (rather than other sources of moisture) on the buttocks, coccyx, rectal area, labia, scrotum, lower abdomen, upper thighs, gluteal cleft or groins in an incontinent patient 10. IAD severity was determined using the skin assessment tool (SAT) 28. The SAT derives a severity score (maximum score = 10) by adding scores within categories of skin redness, area of skin breakdown (cm²) and erosion. Word descriptors were designated to match IAD severity categories 27.

Clinical presentation of a Candida infection

Clinical presentation of Candida infection was determined by a physical assessment conducted by the research nurse. Candida infections associated with incontinence or IAD commonly present with a central maculopapular rash, with characteristic satellite lesions at the margins of the rash extending into normal skin. These rashes may also present as non‐specific confluent papules in patients with IAD 10, 15. In darker skin tones, the central area of fungal infection may be darker 10.

Incontinence‐associated Candida infection

The term incontinence‐associated Candida infection has been used to ensure clarity with reference to incontinence, IAD and Candida infections. Terms commonly used in the literature, such as secondary Candida infection or superimposed Candida infection, imply a possible causal relationship between IAD and Candida infection; however, this relationship is uncertain.

Patient descriptors

Demographic data (age, gender, ward and length of stay), admission diagnosis, comorbidities and current medications as well as height, weight, continence status, incontinence containment product and the presence of faecal or urinary stoma were recorded on a study specific data collection instrument. Pressure ulcers were identified and staged according to current international guidelines for staging pressure ulcers 29. Malnutrition was measured using the Subjective Global Assessment (SGA), a valid and reliable tool, which assesses nutritional status based on patient history and physical examination 30, 31, 32. Results of both of these assessments were combined to produce an overall global rating: well nourished (SGA‐A), moderately malnourished or suspected of being malnourished (SGA‐B) and severely malnourished (SGA‐C) 31, 32. Mobility status was measured as immobility has been associated with incontinence 33. The immobility subscale from the Braden Scale for predicting pressure sore risk was used and administered by the research nurse. Convergent construct validity has been reported for the mobility subscale of the instrument 34. Mobility categories were: immobile, very limited, slightly limited and no limitations 35.

Ethical considerations

The study was approved by the Human Research Ethics Committees for both the research site and the university. The study was conducted in accordance with ethical standards set forth in the 1975 Helsinki Declaration.

Data analysis

All data were entered into the IBM SPSS Statistics for Windows (Version 22.0, Armonk, NY). A random 10% of data entry was cross‐checked for accuracy by the first author. Descriptive statistics were used to describe sample characteristics (means and standard deviations for continuous variables; frequencies and percentages for categorical variables). Characteristics were compared between the three groups. Bivariate analyses using descriptive correlational statistics (χ ² test for independence with Yates continuity correction and Fisher's exact test where expected cell counts were less than five for categorical variables and one way between groups analysis of variance (ANOVA) with post‐hoc analysis using the Tukey test for continuous variables). Candida albicans colonisation was compared between continent and incontinent participants and between incontinent participants with and without IAD. P values less than 0·05 were considered statistically significant.

Results

Over 6 months, 1777 adult inpatients were assessed for eligibility (Figure 1), with 362 patients not meeting inclusion criteria, while 1333 patients were eligible but not recruited. Recruitment targets for each group took varying times to meet. The recruitment target was for 28 participants in each group. The recruitment target for the continent group was reached within 2 months; the target for the urinary incontinent group was reached within 3 months, while the recruitment target for the doubly incontinent group took 6 months to reach 25 participants. This meant that when recruiting for the doubly incontinent group only, it was still necessary to screen all new admissions and record those who met eligibility criteria for the study (n = 1083). However, whilst these patients were eligible, they were not enrolled as they belonged to groups whose recruitment targets had already been met (that is, the continent or incontinent of urine only groups) (Figure 1). A total of 82 patients were recruited to the study. Data for 81 patients were available for analysis as one patient withdrew consent prior to data collection.

Figure 1.

Patient recruitment and enrolment diagram.

The mean age of the sample was 76 years (SD = 12·22), with 53% (n = 43) being male. The mean number of comorbidities and regular medications for each participant were five (SD = 2·02) and eight (SD = 3·10), respectively. Of the total sample, 35% (n = 28) of participants were continent; 35% (n = 28) of participants were incontinent of urine but continent of faeces; and 31% (n = 25) of participants were incontinent of both urine and faeces. Table 1 provides an overview of the characteristics of these three groups. Overall, the urinary incontinence group was significantly associated with a higher BMI F(2, 67) = 3·9, p = ·025 compared to the continent and doubly incontinent groups. Double incontinence was significantly associated with older age F(2, 78) = 5·3, p = ·007, immobility or very limited mobility χ ²(2, N = 81) = 25·9, p < 0·001 as well as risk of malnutrition or malnutrition χ ²(2, N = 81) = 15·2, p = ·003.

Table 1.

Sample characteristics (N = 81)

	Continent (n = 28)	Incontinent of urine (n = 28)	Incontinent of urine and faeces (n = 25)	Statistic	P
Age, mean (SD)	75 (13)	72 (11·3)	82 (7·6)	F(2, 78) 5	0·007
Number of comorbid conditions, mean (SD)	5 (2·8)	6 (2·2)	6 (1·7)	F(2, 78) 2·1	0·134
Number of concurrent medications, mean (SD)	9 (5·2)	9 (3·0)	7 (3·0)	F(2, 77) 2·6	0·083
Body mass index,* mean (SD)	28 (8·7)	33 (12·2)	25 (7·3)	F(2, 67) 3·9	0·025
Male gender, number (%) of patients	18 (64·3)	10 (35·7)	15 (60)	χ 2 = 5·3	0·071
Patients commenced on antibiotics upon admission (oral or intravenous), number (%) of patients	7 (25)	14 (50·0)	7 (28)	χ 2 = 4·3	0·117
Subjective global assessment score, number (%) of patients
A – Normal	20 (71·4)	18 (64·3)	6 (24)12	χ 2 = 15·2	0·003
B – At risk of malnutrition	4 (14·3)	8 (28·6)	(48)
C – Malnourished	4 (14·3)	2 (7·1)	7 (28)
Mobility, number (%) of patients
No limitation	12 (42·9)	5 (17·9)	0 (0)	χ 2 = 25·9	<0·001
Slightly limited	10 (35·7)	13 (46·4)	5 (20)
Very limited	5 17·9)	9 (32·1)	15 (60)
Completely immobile	1 (3·6)	1 (3·6)	5 (20)
Pressure ulcer number (%) of patients	1 (3·6)	1 (3·6)	4 (16)	χ 2 = 3·17	0·236
Diabetes‐type 1 and type 2 number (%) of patients	11 (39)	12 (43)	10 (40)	χ 2 = 1·26	1·0

F, One way between groups analysis of variance (ANOVA); χ ², Chi‐square test for independence.

^*Calculated as weight in kilograms divided by height in meters squared.

Candida albicans was present at the perianal site in 31% (n = 25) and the inguinal site for 15% (n = 12) of the sample, while non‐albicans Candida spp. was present at the perianal site in 7% (n = 8) and in the inguinal site for 6% (n = 5) of the sample. Candida albicans or Candida spp. were present at the perianal site in 43% (n = 23) of incontinent participants compared to 28% (n = 8) of continent participants χ²(1, N = 81) = 4·453, p = ·638. At the inguinal site, 24% (n = 13) of incontinent participants compared to 14% (n = 4) of continent participants had C. albicans or Candida spp. present χ ²(1, N = 81) = 6·868, p = ·258. Figures 2 and 3 show the frequency of C. albicans and Candida spp. according to continence status at the perianal and inguinal sites, respectively.

Figure 2.

Candida albicans colonisation rates at perianal site and continence status (N = 81).

Figure 3.

Candida albicans colonisation rates at inguinal site and continence status (N = 81).

The highest semi‐quantitative level of C. albicans at the perianal site was 1+ (8·6%, n = 7) and at the inguinal site 1+ (2·5%, n = 2). The highest semi‐quantitative level of Candida spp. at both perianal and inguinal sites was 3+ (1·2%, n = 1).

IAD was present in 41% (n = 22) of incontinent patients. There was no significant difference between the type of incontinence (that is urinary only or double incontinence) in those with IAD χ ²(2, N = 53) = ·044, p = ·833. IAD severity ranged from mild 73% (n = 16) to moderate (27%, n = 6). Of the patients with IAD, 45% (n = 10) had C. albicans or Candida spp. present at the perianal site compared to 42% (n = 13) of incontinent participants without IAD χ ²(1, N = 53) = ·000, p = 1·000. Inguinal Candida or Candida spp. colonization rates showed no difference in participants with IAD compared to those without IAD (27%, n = 6, compared to 23%, n = 7), χ ²(1, N = 53) = ·000, p = =1·000. Clinical signs of a Candida infection were observed in 18% (n = 4) of participants with IAD, with the presence of Candida confirmed with microbiological testing in 75% (n = 3) of those participants.

Discussion

This study is the first to provide surveillance data indicating the extent of C. albicans colonisation at perianal and inguinal sites in continent and incontinent acute care patients and in those patients with IAD. We found no significant difference in colonisation rates between continent and incontinent patients at either the perianal or inguinal sites, although there was a non‐significant trend towards higher Candida colonisation rates in incontinent patients. Candida colonisation rates between incontinent patients with and without IAD showed no difference. The rate of incontinence‐associated Candida infection in patients with IAD was lower than anticipated. While there is general acceptance that Candida infection is a common complication of IAD 10, 11, 25, 36, 37, data presenting microbiological confirmation of Candida infections in patients with a clinical presentation of IAD is scant.

Furthermore, despite the ubiquitous presence of Candida albicans as a human commensal 2, 7, there is a paucity of data that identifies Candida colonisation rates in continent compared to incontinent acute care adults. Candida colonisation varies greatly depending on the anatomical site. The highest colonisation rates are in the genitourinary and gastrointestinal tracts 3, 7, with vaginal colonisation rates reported to range between 10% and 20% in asymptomatic women of childbearing age 6, 24. Rectal colonisation is reported to range from 8% to 29% in hospital patients 38, 39, 40, 41, with no difference found between faecal and rectal colonisation rates 3. One study reported inguinal colonisation of hospital patients to be 3% 42. Odds (1988) 3 found oral Candida colonisation to range between 6% and 70%, with a mean of 41% in a compilation of studies. Overall, our study was unable to demonstrate a significant association between continence status and Candida colonisation, however there was a trend towards higher colonization in incontinent participants. In addition, we found no association between faecal frequency and/or quality and Candida colonisation.

It has been reported that patients with double incontinence and/or liquid faeces are more likely to develop IAD 10, 11. It is proposed that the association between double incontinence or frequent liquid faeces and IAD is linked to the presence of digestive enzymes in faeces (higher in liquid faeces) capable of damaging the stratum corneum 10. The higher water content in liquid faeces is also damaging to the skin barrier 11. In addition to a higher risk of developing IAD, Candida infection is frequently cited as a common complication of IAD 10. Therefore, it is reasonable to expect an association between Candida colonisation and IAD. However, quantitative data to support this is limited. Amongst our patients with IAD, 45% were colonised with C. albicans or spp., no different to the findings in incontinent, non‐IAD patients. Limited data is available to enable the comparison of Candida colonisation between adult patients with and without IAD. In a study by Foureur and colleagues (2006) 43 Candida infection prevalence was found to be 63% in a group of elderly patients (mean age 85 years), with IAD, in a long‐term care unit. Our study found no association between the type of incontinence and IAD, and furthermore, patients with urinary incontinence had higher Candida colonisation rates at the perianus than doubly incontinent patients (43% versus 24%, respectively).

In contrast to the limited Candida colonization data available for adults, there is a multiplicity of data reporting associations with Candida colonization and diaper dermatitis in infants (17 18,  20). Two paediatric studies 44, 45 found no difference in Candida colonisation between infants with and without diaper dermatitis. On the other hand, several paediatric studies found higher Candida colonisation rates in infants with diaper dermatitis 17, 18, 20. Associations between Candida colonisation and Candida infection in infants with diaper dermatitis raises questions about the possible role of Candida in the aetiology of the condition in adults. Paediatric data should be extrapolated to adults with caution because of differences between the subjects, such as skin physiology, health status and medications However, unlike paediatric data regarding diaper dermatitis and associated Candida colonisation and infection, there is very little data available investigating these associations in adults with IAD.

Data reporting microbiological confirmation of incontinence‐associated Candida infection is scarce. We found that 18% of participants with IAD showed clinical signs of a Candida infection, which was subsequently confirmed with microbiological testing in 75% (n = 3) of cases. The study by Fourer and colleagues (2006) 43 found that only 60% of the initial Candida infection diagnoses were confirmed on mycological swab cultures or histological examination. One reason for the scarcity of epidemiological data may be that in the clinical setting, microbiological testing for the presence of Candida associated with IAD is not routine practice. This may be because cultures require a 48–72‐hour period for a reliable result 6. Usually, a presumptive diagnosis is made based on clinical presentation, which is then followed by empirical treatment. According to Odds 3, the accuracy of diagnosis of superficial forms of Candida infections is contentious as symptoms of superficial infections may be characteristic of Candida but are not unique. Microbiological confirmation of Candida presence is required, combined with careful evaluation of the clinical presentation for a diagnosis of Candida infection to be made. Conversely, the presence of microbiologically demonstrated Candida at superficial sites is not indicative of infection without due consideration of the clinical presentation 3.

The differences in characteristics between the urinary incontinence group and the double incontinence groups in this study are worthy of comment (see Table 1). In comparison, those with double incontinence were older, malnourished, or at risk of malnutrition, and were either completely immobile, or had very limited mobility. In comparison, the participants who had dual incontinence were older, malnourished or at risk of malnutrition and were either completely immobile or had very limited mobility. These characteristics may reflect differences in the underlying cause and severity of incontinence. Double incontinence appeared to be a manifestation of frailty, while the urinary incontinence group was likely to be related to obstetric/gynaecological risk factors (for example parity), menopause and obesity.

Limitations

Several limitations need to be acknowledged. This was a small pilot study intended to provide initial estimates of colonisation rates in this population. Post hoc power calculations show that the study had 80% power to detect a difference of 35% (e.g. 45% in incontinent group versus 10% in continent group), and the observed differences were much smaller (10% at inguinal site and 15% at perianal site), which may help to inform sample size calculations for further studies in this area. The study was conducted in a single centre and single admitting service; therefore, the data may not be generalisable to other acute care settings or admitting services. Larger multi‐site studies, including a variety of acute care admitting services, would be of value in the future. While every effort was made to recruit all eligible patients, respondents who were older tended to consent to participation more frequently, reflected in the average age of the sample being 76 years. The majority of patients who declined participation were under 50 years of age. Many of those respondents made comments that reflected the belief that incontinence was an old person's problem and therefore not relevant to them. Furthermore, the nature of the skin inspection and collection of microbiological specimens can be intrusive and may have influenced the decision not to participate.

Microbiological testing for C. albicans only was based on the extensive literature identifying C. albicans as the most common human fungal coloniser and pathogen 2, 7. However, we found non‐albicans spp. constituted higher colonisation severity compared to C. albicans. Future studies would benefit from identifying all Candida species that may be present.

Conclusion

This is the first study to report the colonisation rates of C. albicans in continent and incontinent hospitalised adults. Candida colonisation was common at the perianal and inguinal sites at admission to hospital. We found no significant difference in Candida colonisation between continent and incontinent patients or those with IAD, although there was a non‐significant trend towards Candida colonisation in those who were incontinent.

Candida albicans is reported as the most common aetiological agent in incontinence‐associated Candida infection, with the primary source of pathogen being an individual's endogenous commensal organisms 1, 2. Therefore, awareness of the patterns and frequencies of Candida colonisation in these patients is imperative for improving understanding of factors that regulate the colonisation and potential transformation to Candida infection, particularly in incontinent patients. Further research is warranted to understand specific factors that precipitate Candida infections in incontinent adults.