The Perception Gap in Prison Healthcare: Correlates of Inter-Rater Agreement of Patient Quality of Life

• Other Authors: Prost, Stephanie Grace (authoraut)
• Format: Others
• Language: English; English
• Published: Florida State University
• Subjects: Social service
• Online Access: http://purl.flvc.org/fsu/fd/FSU_2017SP_Prost_fsu_0071E_13678

The current study sought to examine inter-rater agreement between patients and their peer caregiver's assessments of patient quality of life (QoL) and to identify which caregiver service provision characteristics were associated with inter-rater agreement of patient QoL within prison healthcare settings. In light of the growing proportion of older adult offenders, many of whom face complex acute, chronic, and terminal illnesses, the current study sought to contribute to the social work literature by examining one component of prison healthcare service provision, peer caregiving and proxy ratings of patient QoL, and the potential factors associated with inter-rater agreement. Patient QoL is a multi-dimensional, holistic, and subjective barometer of well-being. It serves as an important indicator of patient sign, symptom, and pain management in both research and practice settings. A growing body of literature has revealed that proxy-ratings of QoL are used when barriers to self-assessment exist. However, the validity of such proxy ratings is unknown in as previous head-to-head studies have shown such assessments are often inaccurate, most commonly related to overestimation of negative symptomology when compared to patient self-reports. Using primary data collection strategies involving both face to face interviews with patients and peer caregivers (i.e., other offenders) and medical record reviews, the current study describes a small sample of patient-caregiver dyads within several prison healthcare settings in the Louisiana Department of Public Safety and Corrections (n=52). Univariate and bivariate statistics were used to describe the sample and measures. Bivariate statistics were used to identify meaningful, moderate-effects among patient, caregiver, and prison characteristics. Inter-rater agreement of patient QoL ratings was calculated at both the individual-level (dyad) and group-level (sample). Individual-level measures included raw differences, absolute differences, dichotomous agreement, and identified symptom agreement. Group-level measures included measures of association, rater consistency, and standardized mean differences of patient QoL. Bivariate statistics were used to identify potential correlates of inter-rater agreement prior to the use of inferential models and multiple OLS linear and binary logistic regressions were used to examine the capability of caregiver palliative care self-efficacy and understanding of preparatory grief to account for variation in patient-caregiver inter-rater agreement. Patients, on average, were roughly 61 years old. Patients most often self-identified as Black or African American (n=27, 51.9%), male (n=37, 71.2%), divorced or separated (n=23, 44.2%), and had a highest completed education at the high school level (n=27; 51.9%). On average, patients self-reported relatively low levels of physical symptom severity (M=2.53), moderate existential QoL (M=5.54), and moderate to high global and psychological QoL (M=6.12, M=6.70, respectively). Caregivers were markedly younger, approximately 44 years old and most often self-reported as Black or African American (n=14, 66.67%), male (n=14, 66.67%), divorced or separated (n=10, 47.60%), and had a highest completed education at the high school level (n=17, 81.00%). Caregivers average ratings of patient QoL were as follows: physical symptom severity (M=2.66), existential QoL (M=6.18), global QoL (M=6.91), and psychological QoL (M=6.51). More than 40% of both patients and caregivers were serving life without the possibility of parole, due to violent offenses, a lifetime length of incarceration hovering between 17 and 37 years, and an average age of first incarceration between 25 and 30 years old. Inter-rater agreement statistics revealed mixed-findings with individual- and group-level measures indicating caregivers were best able to approximate patients' ratings of physical symptom severity and psychological QoL and larger disparities (viz. patient-caregiver perception gaps) related to global and existential QoL. Group-level measures revealed similar findings as caregivers were again most adept at matching patient physical symptom severity and psychological QoL ratings. Though physical symptom severity was hypothesized as most likely to demonstrate increased inter-rater agreement across individual- and group-level measures, psychological QoL ratings were not anticipated to be closely matched by caregivers. Examination of correlates of inter-rater agreement revealed few bivariate relationships as both moderate and meaningful in effect, though patient characteristics accounted for 17 of the 23 moderate relationships (74%)—findings, again, in contrast to the larger body of perception gap literature indicating that caregiver characteristics are more commonly associated with rating disparities. Results of both linear and logistic regression models emerged as null indicating that neither of the hypothesized caregiver service provision characteristics accounted for variation in rater-agreement or perception gaps of the single-item indicator of QoL and MQOL-CSF subscales. Limitations of the current study were numerous and included concerns related to bias associated with self-report, cross-sectional data collection, small sample size and subsequent power, and measurement. Limitations notwithstanding, the current study reveals that inter-rater agreement among patients and peer caregivers within the sample prison healthcare settings is greater than similar studies completed with both professional and significant-other proxies in community healthcare settings, most notably for psychological QoL ratings. Three reasons for the increased rater-agreement globally and for physical and psychological QoL, more narrowly, are described related to social role-taking competency, homophily, and shared negative attitudes. Directions for future research including the use of larger, random samples, longitudinal data collection strategies, multi-level modeling techniques, and additional correlates of inter-rater agreement and rater-disparities (e.g., patient education) are noted. Discussed are implications for social work practice within prison healthcare settings, as well as areas for caregiver training and continuing education to reduce of the perception gap. === A Dissertation submitted to the College of Social Work in partial fulfillment of the Doctor of Philosophy. === Spring Semester 2017. === March 7, 2017. === Caregiving, Healthcare, Prison, Quality of life === Includes bibliographical references. === Stephen J. Tripodi, Professor Directing Dissertation; Kevin Beaver, University Representative; Jeffrey R. Lacasse, Committee Member; Jean C. Munn, Committee Member.