Inequities in pain assessment and management across demographic and geriatric-related variables in older, hospitalized adults

Inequities in pain assessment and management across demographic and geriatric-related variables in older, hospitalized adults

Background:

Assessing and treating pain in older, diverse adults is challenging. Although having patients self-report pain on a numeric scale is often the recommended approach, age-associated conditions such as dementia, impaired vision, and hearing can prevent patients from effectively communicating. Older patients are also at risk of adverse effects from analgesics such as opioids, making clinicians cautious. Additionally, there are documented inequities in pain management for minoritized patients. Pain is assessed by nurses and physicians around the clock for hospitalized patients. Clinicians can use one of many validated pain assessment tools. Self-Report Tools are where patients report their own pain level through the Numeric Rating Scale, the FACES Pain Scale, or the Verbal Descriptor Scale. Behavioral Tools are used when patients cannot self-report and include the Checklist of Nonverbal Pain Indicators (CNPI) or the Critical Care Pain Observation Tool (CPOT), based on assessment of behaviors like grimacing, body movements, and vocalization.  

Purpose/Objectives:

There are key gaps in the literature. We do not know how the use of these different pain assessment tools vary in practice for hospitalized, older adults across diverse patient demographics and geriatric-related conditions. The purpose of this analysis is to characterize the relationship between pain assessment tools, medications administered, and patient-level factors, focusing on demographic and geriatric-related variables, in hospitalized adults older than 65. First, we describe the differential use of pain assessment tools across self-report and behavioral tools. Second, we report the adjusted odds of a patient receiving a numeric pain score, the comparative "gold standard", compared to other pain assessment tools. Third, within the subset of numeric assessments, we calculate the adjusted numeric score across patients and the adjusted average opioids administered per day.

Method:

This is a retrospective electronic health record cohort study of all adults, 65+, discharged from the general medicine service between January 2013 and September 2021 from an urban academic medical center. Primary predictors included the patient’s self-reported race/ethnicity and limited English proficiency (LEP) status and geriatric related variables, including dementia or mild cognitive impairment, hearing or visual impairment, end-of-life clinical status, and involvement of the geriatrics consult service. Primary outcomes include: (1) The type of pain assessment tool used by nurses, including Self-Report Tools (Numeric, FACES, Verbal Descriptor) and Behavioral Tools (CNPI, CPOT) and (2) The adjusted odds of a patient receiving a numeric assessment, compared to all other types of pain assessment. Exploratory analyses focused on only numeric assessments, and included adjusted morphine milligram equivalents administered to patients, the adjusted numeric pain score, and the odds of a patient reporting a “0” as their pain score.

Results:

There were 1,378,215 pain assessments across 27,857 patient hospitalizations. Outcome 1: The most common tool was the Numeric Rating Scale. White patients, English speaking patients, and those without geriatric-related conditions had higher use rates of the Numeric Rating Scale.  Outcome 2: Compared to White and English-speaking patients, Asian (OR 0.72) Latino (OR 0.89) and LEP patients (OR 0.64) had significantly lower odds of receiving a numeric assessment. Patients at the end-of-life (OR 0.51), with dementia (OR 0.59), those aged 75-84 (OR 0.87) and 85+ (OR 0.71) were all significantly less likely to receive numeric assessments.

Conclusions: Numeric pain scores, which are widely recommended, frequently documented, and used to guide treatment of pain, are less likely to be used in racial and ethnic minority patients, patients with LEP, older patients, those with dementia, or at the end-of-life. These findings reflect a clear inequity in patient care. Differences in reported pain scores, communication differences between clinicians and patients, and the willingness to accept a prescribed dose of pain medication are all susceptible to cultural differences, bias, and racism. Pain assessment variation for patients with geriatric related conditions is likely due to ineffective communication from visual, hearing, and/or cognitive impairment. There may also be bias in providing more potent medications for older patients due to concerns for over-sedation, delirium, and respiratory depression. Our findings are novel for the field of pain management, geriatric general medicine, and health equity. We need to further collect data through an equity lens, track the use of medical interpreters, reasons for the use of different pain assessment tools, and specific indications for pain medication administration. We must also test and validate new, multidimensional tools, that consider function and quality of life, specifically for minority groups and those with geriatric-related conditions.

References: Mathur VA, Trost Z, Ezenwa MO, Sturgeon JA, Hood AM. Mechanisms of injustice: what we (do not) know about racialized disparities in pain. PAIN. 2022;163(6):999. doi:10.1097/j.pain.0000000000002528

Deng LX, Patel K, Miaskowski C, et al. Prevalence and Characteristics of Moderate to Severe Pain among Hospitalized Older Adults. J Am Geriatr Soc. 2018;66(9):1744-1751. doi:10.1111/jgs.15459

Hadjistavropoulos T, Herr K, Turk DC, et al. An interdisciplinary expert consensus statement on assessment of pain in older persons. Clin J Pain. 2007;23(1 Suppl):S1-43. doi:10.1097/AJP.0b013e31802be869