Fecal incontinence with loose stools is a risk factor for skin damage that may develop into a pressure ulcer (PU). The aim of this study was to determine the feasibility of applying the Incontinence-Associated Dermatitis and its Severity (IADS) instrument to patients with fecal incontinence as a tool to predict PU development. This prospective study enrolled 120 intensive care unit patients with bowel incontinence of Bristol Stool type 5, 6, and 7. Trained nurses evaluated IADS scores and the occurrence of PUs daily for 7 days. Patients with higher IADS scores were significantly more likely to develop a PU (odds ratio = 1·22, 95% confidence interval = 1·12–1·33). The receiver operating characteristic curve analysis revealed the area under curve to be 0·790, suggesting that higher IADS scores are associated with an increased likelihood of developing a PU (sensitivity 72·5%, specificity 71·2%, using a cut-off value of 8/9). Our results suggest that the IADS instrument can serve as a tool for predicting the occurrence of PUs in patients with fecal incontinence. Patients with IADS scores that exceed eight points should be classified as being at risk of developing a PU, and placed under intensive care as a proactive measure to prevent PU development.
The development of a pressure ulcer (PU) is the most serious manifestation of impaired skin integrity; therefore, its occurrence is often used as an indicator of the quality of nursing care . PUs can contribute to an increase in length of hospital stay, mortality, morbidity, and the incidence of complications. In addition, the Joint Commission International reported that the number of cases of PU in the US increases by 25 million annually , suggesting a negative impact on the global economy, as well as decreasing the quality of life of patients.
A Korean study, targeting 796,856 patients, showed an average incidence rate of 3·2%: 2·7% for general hospitals, 1·7% for hospitals, and 8·2% for long-term care facilities . Padula et al.  reported that in 2009 in the US, the daily cost of treatment for hospital-acquired Stage I and Stage II PUs and Stage III and Stage IV PUs was $12,770·54 and $5,622·98 per day, respectively, suggesting that PU prevention is more cost-effective than treatment. If it is possible to predict PU development, it is likely to reduce the cost of medical treatment, as well as greatly improve the patient's quality of life.
In addition to managing pressure and shearing forces, PU prevention requires the strict management of incontinence, which poses an important risk factor in PU development. Over half (56·7%) of PU patients suffer from fecal incontinence; studies suggest that patients with fecal incontinence are 22 times more likely to develop a PU compared to patients without fecal incontinence .
Incontinence-associated dermatitis (IAD), also known as perineal dermatitis, is the inflammation of the skin associated with exposure to urine or stool . In patients with IAD, the skin tends to be so vulnerable to pressure and shearing forces that it can easily lead to the development of a PU [7, 8]. Many studies have verified that incontinence control helps to reduce the incidence of PUs [9, 10], demonstrating that IAD is closely correlated to PU occurrence.
In many cases, intensive care units (ICU) fail to manage patients with urine and stool incontinence appropriately, either because of the severity of the patient's physical condition or because of the method of treatment within the unit. In an ICU, priority is given for the treatment of the underlying disease, rather than the IAD; the IAD subsequently becomes aggravated, causing the development of a PU .
Fecal incontinence is a greater risk factor for PU development than urinary incontinence because the bacteria and enzymes contained in stool can be caustic to the skin. In the presence of both urinary and fecal incontinence, fecal enzymes convert urea to ammonia, raising the skin pH, and allowing it to become more permeable to other irritants (Figure 1) . ICU patients with moisture issues are four times more likely to develop a PU than patients without . Some studies reported one-third of the ICU patients had major fecal incontinence and 42·5% among them showed incidence of IAD [14, 15].
In addition, incontinent patients with loose stool reported a greater severity of IAD . Incontinence, therefore, has a large and independent influence on PU development.
The Braden Scale is one of the most intensively studied risk assessment scales used to identify the risk of developing a PU . One of the six risk factors assessed on this scale is moisture (exposure to incontinence, drainage, and sweat). The Braden Scale, however, appears to underestimate the influence of incontinence on PU incidence. Therefore, in the case of incontinent patients with loose stools, it can be helpful to use an IAD-related measurement in addition to the Braden Scale to predict PU occurrence.
This study used the Incontinence-Associated Dermatitis and its Severity (IADS) instrument  to predict PU occurrence in patients with fecal incontinence to determine if there is a correlation between IAD and PU occurrence, and whether the IADS instrument is able to discriminate PU occurrence.
Materials and methods
Setting and study population
Patients were recruited from five ICUs within the Samsung Medical Center, located in Seoul, South Korea, between September 2011 and April 2012; study procedures were approved by the institutional review board.
Subjects included in our prospective study, were ICU patients aged over 40 years; exhibiting fecal incontinence with a Bristol Stool type 5 (soft blobs with clear-cut edges, passed easily), 6 (fluffy pieces with ragged edges, a mushy stool), or 7 (watery, no solid pieces, entirely liquid); no evidence of IAD or PU on the baseline skin evaluation ; and a Braden Scale score of 16 or less. Informed consent was obtained from all patients or legal representatives. One hundred and thirty one patients were enrolled from the medical ICU, the surgical ICU, the cardiac care unit, the cardiovascular ICU, or the neurosurgical ICU. A sample size of at least 97 patients was determined using a two-sided 5% significance level; PU occurrence rate was expected to be 35 ± 10% , with an anticipated dropout rate of 10%. Eleven of the 131 subjects (8·4%) dropped out of the study because of transfer to other units in the same hospital or discharge from the hospital (n = 6), skin disease (n = 2), perianal ulcer (n = 1), or withdrawal from the study (n = 2; Figure 2).
The IADS instrument was used to evaluate IAD . This tool requires a nurse to assess skin damage using the 5-point Likert scale, with each of 13 body locations (perianal skin, crease between buttocks, left lower buttock, right lower buttock, left upper buttock, right upper buttock, genitalia, lower abdomen/suprapubic, crease between genitalia and thigh, left inner thigh, right inner thigh, left posterior thigh, and right posterior thigh) assigned a value according to the most severe attribute noted in that location. A score of 1 was assigned to those areas identified as pink in color; a score of 2 was assigned to those areas observed to be red in color, without a rash or skin loss; locations noted as having a fungal rash received a score of 3; and locations with any degree of skin loss received a score of 4. The scoring algorithm, therefore, had a range of 0 (no degree of redness, rash, or skin loss) to 52; a lower score indicated better skin condition. Criterion validity using intra-class correlations was r = 0·98, P = 0·006 indicating the agreement between the 347 raters and 3 experts, and r = 0·91, P = 0·008 among the 3 experts .
The Bates-Jensen Wound Assessment Tool (BWAT)  was used as the gold standard to test the ability to discriminate PUs using IADS. BWAT comprises 15 items: location, shape, size, depth, edges, undermining or pockets, necrotic tissue type, necrotic tissue amount, exudate type, exudate amount, surrounding skin color, peripheral tissue edema, peripheral tissue induration, granulation tissue, and epithelialization. Location and shape are not scored; the remaining 13 items are scored with characteristic descriptors rated on a scale of 1 (best attribute) to 5 (worst attribute). Wounds were scored initially for a base-line assessment and then at regular intervals. The content validity of the tool was established using a nine-member expert judge panel (mean overall content validity index = 0·91, P = 0·05). lntra-rater reliability estimates yielded a mean of 0·975 .
The Bristol Stool Scale was used to assess stool consistency and to accurately identify loose or liquid stool. The scale classifies stool consistency based on seven categories: 1 and 2 indicate hard stools; 3 and 4 indicate normal stools; and 5, 6, and 7 indicate liquid stool with or without smaller formed or semi-formed fecal content.
The Braden Scale was used to evaluate the eligibility of participants. It is composed of six subscales that conceptually reflect degrees of sensory perception, moisture, activity, mobility, nutrition, and friction and shear. All subscales are rated from 1 to 4, except for friction and shear, which is rated from 1 to 3. The subscales are then summed for a total score, ranging from 6 to 23; lower scores indicate lower function and higher risk of developing a PU.
Data were collected by primary wound care nurses (PWNs), institutionally qualified nurses who have completed middle-level courses on wound care and who provide primary wound care within their ICUs under the supervision of wound ostomy continence nurses. Five PWNs, one each from the medical ICU, surgical ICU, cardiac care unit, cardiovascular ICU, and neurosurgical ICU, who had worked in the ICU for over 3 years, had undergone a 1-hour instruction on how to score the IADS instrument and the BWAT, and to assess feces based on the Bristol Stool Scale. Educational techniques included a Power Point presentation and oral explanations. The knowledge of the ICU PWNs was then tested using five computer-simulated cases. Intra-class correlations of IADS scores and BWAT scores were 0·96 and 0·92, respectively.
The trained PWNs then completed the IADS and BWAT, and checked for development of PUs daily for 7 days. The highest IADS scores and PU stage documented during the 7-day period were used for data analysis.
Data were analyzed using Predictive Analytics Software version 18.0 (SPSS Inc., Chicago, IL). General, incontinence- and PU-related characteristics of the participants were analyzed using descriptive statistics, such as, frequency, percentage, mean, and standard deviation. The intra-class correlation analysis was used to test reliability of the IADS instrument. The relationship between IADS scores and PU development was analyzed using logistic regression on the premise that patients with higher IADS scores had more developed PUs.
Receiver operating characteristic (ROC) curve analysis was performed , with the BWAT score as the gold standard, to determine the ability of the IADS instrument to discriminate PU occurrence. Area under the receiver operating characteristic curve (AUROC) was extracted to estimate sensitivity (the ability of the test to correctly identify the discrimination of the PU occurrence with IADS instrument) and specificity (the ability of the test to correctly identify those patients without PU).
The age of the study population ranged 40–92 years, with a mean age of 67·5 ± 13·1 years (±SD); 64·2% of patients were aged 65 years or older. More than half of the patients were male (n = 71; 59·2%). The most frequent Bristol Stool type observed was 7 (n = 52; 43·3%), and the most common average frequency of fecal incontinence was less than 4 hours (six times or over/day), observed in 63 patients (52·5%). All of the patients with urinary incontinence (n = 111, 92·5%) had indwelling catheter. The average Braden Scale score and body mass index (BMI) were 11·9 ± 1·7 and 22·0 ± 3·7 kg/m2, respectively (Table 1).
Table 1. General, incontinence-associated dermatitis-related, and pressure ulcer-related characteristics of patients (N = 120)
n (%) or Mean ± SD
BMI, body mass index; PU, pressure ulcer.
*Bristol Stool Scale.
†Septic, hypovolemic, cardiogenic shock.
‡Pre-albumin < 20 (mg/dL), albumin < 2·5 (g/dL) or NPO more than 3 days.
Characteristics of incontinence-associated dermatitis and pressure ulcer
We observed that all patients with fecal incontinence had some evidence of IAD based on the IADS instrument; the average IADS score was 9·30 ± 7·42 (Table 2). One-third of the patients (n = 40) developed a PU, and the mean BWAT score was 23·30 ± 3·84 (Table 3).
Table 2. Characteristics of incontinence-associated dermatitis (N = 120)
There was a significant relationship between IADS scores and PU development (P < 0·001). Patients with higher IADS scores had a greater likelihood of developing a PU [odds ratio (OR) = 1·22. 95% confidence interval (CI) = 1·12–1·33; Table 4).
Table 4. Relationship between the IADS score and pressure ulcer development (N = 120)
95% CI for OR
CI, confidence interval; IADS, Incontinence-Associated Dermatitis and its Severity; OR, odds ratio; SE, standard error.
Discriminant ability of IADS instrument for PU occurrence
Based on the ROC curve analysis, the AUROC was 0·790 (95% CI: 0·701–0·879; Table 5). The optimal cut-off point of the IADS instrument was 8/9 indicating that the person who scored 9 and higher has a high probability of having a PU (sensitivity 72·5%, specificity 71·2%, using a cut-off value of 8/9; Figure 3).
Table 5. Receiver operating characteristic curve analysis of the Incontinence-Associated Dermatitis and its Severity instrument
The aim of the study was to determine the feasibility of applying the IADS instrument to patients with fecal incontinence, as a tool to predict PU development.
This study revealed that the greater the severity of IAD, the greater the likelihood that a PU will occur. This result indicated that concurrent validity was supported. We concluded that fecal incontinence constitutes a risk factor that leads to the development of a PU, as observed in a previous study , because chemical irritants destroy the acidic barrier of the skin and damage the skin when exposed to liquid stool , despite the mechanism of IAD and PU being different . Our results show that a 10-point increase in the IADS score constitutes a 7 times greater risk of PU development.
This study was targeted at ICU patients with fecal incontinence in an acute setting and may, therefore, have limitations if universally applying its findings to patients hospitalized in long-term care facilities.
Another limitation of this study was that the IADS score itself cannot distinguish how many locations are involved and the severity in each region. Because the severity of a dermatitis region has more effect on the occurrence of PU than the number of skin regions, clinicians should consider it for prediction and prevention of PU. Repetitive study will be needed.
In addition, both the IADS instrument and BWAT used in the current study are tools for nurses to check patients' skin status. The most important thing is nurses' skill in assessing a patient's skin status; therefore, training nurses before applying these instruments may be essential. In this study, 5 PWNs who have cared for patients' wounds rated; therefore, it might be possible to get good inter-rater reliability with only 1-hour instruction.
In conclusion, the calculated AUROC indicated that the IADS instrument can be used as a tool to predict PU development in patients with fecal incontinence; patients with an IADS 9 and higher score are most likely to develop a PU. We suggest that patients with IAD could be classified as high risk for PU occurrence, and that they should be monitored appropriately with the IADS instrument to prevent the development of a PU.
We thank K Borchert, RN, CWOCN, ACNS-BC, for permitting the use of the IADS instrument for this study. We thank J. M. Hwang, RN, for the statistical analysis. We also thank PWNs and RNs in the Samsung Medical Center for assisting with data collection.