INTRODUCTION
The incidence of pressure ulcers (PrUs) is 2.2% to 23.9% in long-term-care settings.1 Pressure ulcers are associated with poorer health outcomes and can contribute to death in some individuals.1,2 The presence of PrUs can interfere with activity and mobility and may be complicated by pain. Gorecki et al3 indicated that pain was a common problem for people with PrUs. Teno et al4,5 reported that nursing home (NH) residents with PrUs are 60% more likely to have daily excruciating pain than those without a PrU.
Pain is a complex phenomenon derived from sensory stimuli or neurologic injury and modified by individual memory, expectation, and emotions.6 Pain is a common condition among NH residents.7-10 There is no empirical evidence that persons with dementia (PWD) experience more or less pain, but the prevalence of pain among PWD is reported to be lower than their cognitively intact counterparts.6 Nursing home residents with dementia have cognitive and communicative impairments as symptoms of dementia, thereby hampering effective verbalization of their pain intensity. Thus, pain is often underreported in PWD, even when there is a probable cause for pain and management of the pain may be compromised.11,12
The purpose of this study was to examine the relationship between level of cognitive impairment and pain severity in NH residents with PrUs using the Minimum Data Set (MDS) 2.0 data set. The authors hypothesized that NH residents with severe cognitive impairment would exhibit less severe pain compared with the residents with mild cognitive impairment across all stages of PrU. The specific aims of this study were to (a) describe the prevalence of PrUs and pain in NH residents with dementia, (b) examine the relationship between stages of PrU and severity of pain, and (c) investigate the presence of a PrU on pain severity, after controlling for covariates, based on the severity of cognitive impairment. After reading this article, clinicians will have a better understanding of how cognitive impairment affects the severity of PrU pain in NH residents.
METHODS
Participants (N = 56,577) in this study were residents with dementia in Medicare- or Medicaid-certified NHs in the state of Florida between January 1, 2009, and December 31, 2009, who had an MDS comprehensive assessment on file. The study was approved by the institutional review board of the affiliated university. The study used a 3-group, cross-sectional, comparative design. The groups were determined by cognitive status and stage of PrU: group 1 consisted of NH residents with a PrU (Stages I-IV) and mild cognitive impairment; group 2 consisted of NH residents with a PrU (Stages I-IV) and moderate cognitive impairment; and group 3 consisted of NH residents with a PrU (Stages I-IV) and severe cognitive impairment. Selection criteria were applied to ascertain data from noncomatose NH residents older than 65 years with Alzheimer disease or other dementia.
Measurements
All the measures were collected from the MDS 2.0 data set. The MDS is a multidimensional questionnaire used to assess NH residents in Medicare- or Medicaid-certified long-term-care nursing facilities in the United States. An MDS assessment is routinely completed upon admission and on a quarterly basis thereafter and contains standardized physical, psychological, and psychosocial assessment data of NH residents. The MDS data include about 300 items, which can be converted to ordinal or continuous measures that generally meet the psychometric criteria of reliability and validity.13,14 Although the MDS is primarily used for clinical purposes, several subscales (eg, pain and cognition) have also been validated and used for research.
The MDS-Pain Severity Scale,15,16 combining both pain intensity and frequency, was used to assess pain severity in NH residents with dementia. The MDS-Pain Severity Scale16 is recorded on a 4-point ordinal scale: 0 (no pain), 1 (mild pain), 2 (moderate pain), and 3 (excruciating pain). Mild pain is characterized as mild intensity pain that occurs less than daily; moderate pain reflects daily pain in MDS-Pain Frequency and mild or moderate pain in MDS-Pain Intensity; and excruciating pain reflects daily pain in MDS-Pain Frequency and horrible or excruciating pain in MDS-Pain Intensity. The MDS-Pain Severity Scale has been reported to have an interrater reliability coefficient of 0.73, and [kappa] coefficient of 0.70 with a visual analog scale.16 The original MDS-Pain Intensity and Frequency Scales, which comprise the Pain Severity Scale, are reported to have an internal consistency of 0.69 to 0.88 and test-retest reliability of 0.84 to 0.87.17,18
The MDS-Cognitive Performance Scale (MDS-CPS)19 is used to measure cognitive impairment. It is a 7-point ordinal scale, with 0 (intact), 1 (borderline intact), 2 (mild impairment), 3 (moderate impairment), 4 (moderate severe impairment), 5 (severe impairment), and 6 (very severe impairment).20-22 The MDS-CPS has high test-retest reliability, a criterion validity coefficient of 0.45 to 0.75 with Mini-Mental State Examination, a criterion validity coefficient of 0.41 to 0.77 against Global Deterioration Scale, a criterion validity coefficient of 0.66 against Psychogeriatric Dependency Rating Scale, a criterion validity coefficient of 0.45 against Mattis Dementia Rating Scale, a sensitivity of 0.94, and a specificity of 0.94.21-24
The MDS-PrU item was used to assess presence and stage of PrUs. This PrU item is a 5-point ordinal scale, with 0 (none), 1 (Stage I), 2 (Stage II), 3 (Stage III), and 4 (Stage IV). The MDS-PrU item has been reported to have a reliability coefficient of 0.62.25
The MDS 2.0 was also used to measure activities of daily living (ADL) impairments, Charlson Comorbidity Index (CCI), and sociodemographic characteristics (eg, age, gender, marital status, education, and race/ethnicity).
Statistical Procedures
Analyses were performed using SPSS, version 20 (IBM Inc, Armonk, New York). The MDS-CPS was coded categorically and numerically to quantify the severity of cognitive impairment. The NH residents were grouped based on the severity of cognitive impairment: residents with mild impairment (CPS = 0-2), residents with moderate impairment (CPS = 3), and residents with severe impairment (CPS = 4-6). This grouping method has commonly been used in the literature.20-22 The MDS PrU staging was coded categorically and numerically to quantify the severity of PrU. The MDS Pain Severity Scale was coded numerically to quantify the severity of pain.
Bivariate analyses were conducted for cognition using the Pearson [chi]2 Test. Multivariate ordinal logistic regression analyses were conducted to explore the relationship between PrUs and pain severity, after controlling for covariates (eg, ADL impairments, comorbidities, and sociodemographics).
RESULTS
The sample (N = 56,577) was mostly female (67.7%), unmarried (69.7%), and with a high school education or less (72.5%) (Figure 1). The majority of participants were white (78.7%), with a mean age of 84 years (range, 65-109 years), and had an average CCI of 3 years (range, 1-16 years). There were 15,955 residents (28.2%) with mild cognitive impairment; 21,657 residents (38.3%) with moderate cognitive impairment; and 18,931 residents (33.5%) with severe cognitive impairment.
The presence of a PrU was reported for 18.1% of residents with dementia (n = 10,227) (Table 1). Among NH residents with PrUs, the prevalence of each stage of PrU is as follows: Stage I (n = 4351; 42.5%), Stage II (n = 3307; 32.3%), Stage III (n = 649; 6.3%), and Stage IV (n = 1920; 18.8%).
Pain was reported for 36.9% of residents with dementia (n = 20,858). Among NH residents with pain, the frequency distribution for the severity of pain is as follows: mild pain (n = 15,139; 72.6%), moderate pain (n = 5,227; 25.1%), and excruciating pain (n = 492; 2.4%).
The stage of PrU was significantly correlated with pain severity ([chi]2 12 = 775.74, P < .001). The bivariate relationship between PrU and pain severity is summarized in Table 2. In general, NH residents with dementia with more severe PrUs exhibited more severe pain.
An ordinal logistic regression model was conducted to predict pain severity among 3 groups: residents with mild cognitive impairment, moderate impairment, and severe impairment. The results of the logistic regression models are shown in Table 3. Among NH residents with mild cognitive impairment, the pain severity of persons with a Stage II PrU is 29% greater and is 50% greater for those with a Stage IV PrU, compared with persons without PrUs. Among NH residents with moderate cognitive impairment, those with any stage of PrU were found to likely have 24% to 68% more severe pain relative to persons without a PrU (odds ratio: Stage I = 1.24, Stage II = 1.53, Stage III = 1.40, and Stage IV = 1.68). Among NH residents with severe cognitive impairment, those with any stage of PrU were found to have approximately double the risk of being in severe pain (odds ratio: Stage I = 1.71, Stage II = 1.96, Stage III = 2.19, and Stage IV = 2.36).
DISCUSSION
The major finding in this study was that NH residents with severe cognitive impairment were assessed to have more severe pain compared with NH residents with mild or moderate cognitive impairment with any stage of PrU. Also, the authors found that more severe pain was associated with Stages III and IV PrUs in the moderately or severely cognitive-impaired NH residents. Among NH residents with moderate cognitive impairment, those with any stage of PrU were reported to have 24% to 68% more severe pain than those without PrUs. Among NH residents with severe cognitive impairment, those with any stage of PrU were reported to have 2 times more severe pain relative to persons without PrUs.
Having a Stage III or IV PrU is associated with more severe pain, a finding that is consistent with other studies in cognitively intact older adults.26-28 However, in older adults with dementia, research has shown that more advanced dementia is associated with less self-reported pain.29 The authors' findings from NH residents with severe dementia and PrUs indicate the opposite relationship-more advanced dementia is associated with more severe pain. This finding may highlight a discrepancy in how pain is assessed and recorded in severely cognitively impaired NH residents using the MDS 2.0 tool. Because PWD verbally express less pain than do those who are cognitively intact,11,12 it is possible that nurses who provide input about patients to the MDS coordinator are relying on other indicators of pain besides verbal report (eg, behavioral indicators) or are making assumptions about pain based on the presence of wounds. It is also interesting to note that, in this study, more advanced PrUs are associated with more severe pain. Taken together, these findings suggest that PrU wounds are painful, even in PWD, and that they may be associated with more pain than previously indicated in the literature. These conclusions may reflect new knowledge in this previously underinvestigated population, or they may reflect measurement discrepancies. Nonetheless, these findings warrant further investigation into the pain assessment process in this population, and the forthcoming MDS 3.0 data (which include more detailed pain indicators) are anticipated to be an important source of data to explore these questions further.
The authors acknowledge several study limitations. First, this study is inherently limited by secondary analysis of federally mandated MDS 2.0 assessment data. MDS assessment data may have some variability because of different styles and skills of MDS coordinators in each facility. In addition, the MDS 3.0 assessment data, which have better measurement of key study variables, are not yet available for analysis. This study raises some important issues that can be followed up in the next wave of data. Second, the study design is descriptive and cross-sectional, so this study is not able to examine causal relationships. Finally, the role of pain medications is not considered in this study. Pain medications are usually given when nursing staff note residents' pain. The highest level of pain could have been managed by pain medications during the last 7 days, but it is not possible to discern this in the MDS assessment data. Because PrUs are associated with pain, analgesic medications may have been given before and/or after PrU treatment for NH residents with mild cognitive impairments-those who could verbally express their pain level sufficiently. This might account for why pain levels were lower in this group of residents than among those with more advanced dementia.
SUMMARY
In summary, the NH residents with severe cognitive impairment were recorded as having more severe pain compared with residents with mild or moderate cognitive impairment, across all stages of PrUs. Nursing home residents with advanced dementia have cognitive and communicative impairments as symptoms of dementia, thereby hampering effective verbalization of their pain intensity. Thus, they may not have had adequate pain treatment.
Healthcare providers should assess pain levels in NH residents with PrUs and consider managing pain, especially before and/or after PrU dressing changes and debridement treatments.
REFERENCES