Geriatrics, Pain Behaviors, MDS-RAI 2.0, Cognition, Concomitance, Cognitive Impairment, Structural Equation Modeling, Theoretical Modeling


An integral concern across care settings is the prompt intervention for patients suffering with pain. Long-term care (LTC) settings present with unique challenges to assess and manage pain in resident populations. Pain assessment is especially challenging, because residents have varying degrees of cognition to communicate their pain, and clinician/staff knowledge of pain symptoms may be lacking. The purpose of this research was to improve the measurement of pain and outcomes of care for the elderly residing in skilled nursing care, especially those with cognitive-impairment. The specific aims of this study were to: 1) Determine the magnitude of the relationship between pain behaviors and a measurement model hypothesized for pain; 2) Test the construct validity of a pain measurement model; 3) Examine the concomitance of pain and cognition in a three-year longitudinal analysis. The research questions answered: 1) Is there a difference in the prevalence of pain in cognitively intact versus cognitively-impaired residents; 2) Can a theoretically derived model of pain aid in detecting pain across all cognitive levels; and 3) Do pain and cognitive status concomitantly correlate? The goal was to examine the covariance model of concomitance of pain and cognition to more accurately construct theoretical models of pain to then include additional resident care factors in future research. Traditional self-reports of pain are often under-assessed and under-treated in the cognitively-impaired (CI) elderly resident. Having additional measures to detect pain beyond self-reports of pain intensity and frequency increases the likelihood of detecting pain in populations with complex symptom presentation. Data collected from skilled nursing facilities offer exceptional opportunities to study resident demographics, characteristics, symptoms, medication use, quality indicators, and care outcomes. The Minimum Data Set-Resident Assessment Instrument (MDS-RAI) 2.0, a nationally required resident assessment tool, must be completed on every resident in a Medicare LTC facility within 14 days of admission, quarterly, annually and with significant changes in resident status. Because the MDS is widely used and recognized in LTC settings, core items from MDS [i.e., pain frequency (J2a) and pain intensity (J2b)] along with additional MDS items hypothesized to signify pain were analyzed in the pilot measurement model. Ten core items from MDS were used: 1) Inappropriate behavior frequency (E4da); 2) Repetitive physical movements; 3) Repetitive verbalizations (E1c); 4) Sad facial expressions (E1l); 5) Crying (E1m); 6) Change in mood (E3); 7) Negative statements (E1a); 8) Pain frequency (J2a); 9) Pain intensity (J2b); and 10) Cumulative pain sites scores. All indicators of pain were significant at the <.01 level. A longitudinal cohort design was used to answer if a concomitance exists between pain and cognition. Data were collected from MDS annual assessments from 2001, 2002 and 2003 for residents across the United States. The sample consisted of 56,494 residents age 65 years and older with an average age of 83 [plus or minus] 8.2 years. Descriptive statistics, ANOVA and a covariance model were used to evaluate cognition and pain at the three time intervals. ANOVA indicated a significant effect (<.01) for pain and cognition with protected t-tests indicating scores decreased significantly over time with resident measures of pain and cognition. Results from this study suggest that: 1) Using only pain intensity and frequency, pain prevalence was found in 30% of the pilot population, while 47.7% of cognitively intact residents had documented pain and only 18.2% of the severely CI had documented pain, supporting previous research that pain is potentially under-reported in the CI; 2) Parsimonious measurements models of pain should include dimensions beyond self-reports of pain (i.e., cognitive, affective, behavioral and inferred pain indicators); 3) Model fit was improved by using specific MDS items in the pain construct; 4) Longitudinal analysis revealed relative stability for pain and cognition measures over time (e.g., larger stability or consistency was found in cognitive measures than the measures of pain over the three-year period); 5) Crossed-legged effects between pain and cognition were not consistent; 6) A concomitant relationship was not found between pain and cognition. The relationship was significant (<.01), but associations were weak (r=0.03 to 0. 08). Pain or cognition should not be used as a predictor of the other in theoretical models for similar populations. The MDS is a reliable instrument to follow resident attributes, quality of care, and patient outcomes over time. The development of more accurate assessments of pain may improve resident care outcomes. Ineffectively intervening on the pain cycle is posited to cause secondary unmet needs that affect the resident's quality of life. Findings support the importance of improving clinical outcomes in the management of pain in the elderly residing in long-term care. Deficits in the treatment of pain highlight the impetus to support health policy change that includes pain treatment as a top health priority and a quality indicator for federally funded programs supporting eldercare.


If this is your thesis or dissertation, and want to learn how to access it or for more information about readership statistics, contact us at

Graduation Date



Sole, Mary Lou


Doctor of Philosophy (Ph.D.)


College of Nursing



Degree Program









Release Date

March 2010

Length of Campus-only Access


Access Status

Doctoral Dissertation (Open Access)

Included in

Nursing Commons