Evaluation of Musculoskeletal Pain Management Practices in Rural Nursing Homes: Comparing to Evidence-Based Criteria

Abstract

Chronic pain, mainly associated with musculoskeletal diagnoses, is inadequately and often inappropriately treated in nursing home residents. The purpose of this descriptive study is to identify the musculoskeletal diagnoses associated with pain and to compare pain management of a sample of nursing home residents with the 1998 evidence-based guideline proposed by the American Geriatrics Society (AGS). The sample consists of 215 residents from 13 rural Iowa nursing home homes. The residents answered a series of face-to-face questions that addressed the presence/absence of pain and completed the Mini-Mental State Examination (MMSE). Data on pain were abstracted from the Minimum Data Set (MDS). Analyses included descriptive statistics, cross tabulations, and one-way analysis of variance. Residents’ responses to the face-to-face pain questions yielded higher rates of pain compared to the MDS pain data. Resident records showed that acetaminophen was the most frequently administered analgesic medication (30.9%). Propoxyphene, not an AGS recommended opioid, was also prescribed for 23 (10.7%) residents. Of the 70 (32.6%) residents expressing daily pain, 23 (32.9%) received no scheduled or pro re nata (PRN) analgesics. There was no significant difference between MMSE scores and number of scheduled analgesics. Additionally, residents’ self-reported use of topical agents was not documented in the charts. The findings suggest that the 1998 AGS evidence-based guideline for the management of chronic pain is inconsistently implemented.

Chronic pain is a significant clinical issue as 45% to 83% of older nursing home residents report they experience pain (Ferrell, Ferrell, & Rivera, 1995; Parmelee, Smith, & Katz, 1993) and 3.7% (n = 80,512) reported daily pain that was excruciating more than once the previous week (Teno, Kabumoto, Wetle, Roy, & Mor, 2004). Musculoskeletal diseases are common causes of chronic pain in nursing home residents (Buffum, Sands, Miaskowski, Brod, & Washburn, 2004; Cramer, Galer, Mendelson, & Thompson, 2000; Ferrell et al., 1995; Won et al., 2004). Under assessment and under treatment of chronic pain exists among nursing home residents negatively impacting their quality of life (Bernabei et al., 1998; Won et al., 1999; Won et al., 2004).

Pain assessments are often inconsistent (Cramer et al., 2000) and poorly documented in the resident’s medical record (Ferrell et al., 1995). Further complicating appropriate pain management in the nursing home setting is the use of inappropriate medications, inadequate dosing of analgesics, and the ordering of PRN analgesics (Bernabei et al., 1998; Buffum et al., 2004; Cramer et al., 2000; Hutt, Pepper, Vojir, Fink, & Jones, 2006; Won et al., 2004). The older adults most likely to receive inadequate doses of analgesics or no analgesics at all are likely to be 85 years of age and older, of minority status (Allen et al., 2003; Bernabei et al., 1998; Won et al., 1999), and have a cognitive impairment (Bernabei et al., 1998; Won et al., 1999). Nonpharmacologic interventions are usually limited to rehabilitation therapies (physical, occupational, and recreational) (Cramer et al., 2000; Ferrell et al., 1995; Won et al., 1999) and heating pads (Ferrell et al., 1995). There is limited evidence that nursing homes provide nonpharmacologic pain management for their residents.

The purpose of this study is to compare pain management practices in 13 rural Iowa nursing homes with the 1998 American Geriatrics Society (AGS) guideline, The Management of Chronic Pain in Older Persons. Although the AGS evidence-based guideline was developed for the management of chronic pain for all older adults and is not specific to older nursing home residents, the guideline focuses on the oldest, sickest, and most frail older adults. Therefore, the AGS guideline is appropriate for use in older adults in nursing homes. In fact, the Institute of Medicine (2000) recommends the use of guidelines to improve quality of care.

Research findings that have emerged from nursing home data following the publication of the AGS guideline (1998) are limited (Allen et al., 2003; Buffum et al., 2004; Cohen-Mansfield & Lipson, 2007; Cramer et al., 2000; Hutt et al., 2006; Won et al., 2004). Researchers have focused their evaluations on the appropriateness and effectiveness of analgesics in nursing home residents. There have not been any studies evaluating the implementation of the 1998 AGS guidelines specific to the management of chronic pain in nursing home residents.

Prescribed analgesics in nursing homes included: nonopioids, nonselective nonsteroidal anti-inflammatory drugs (NSAIDs), and opioids. Nonopioid administration varied between 35.6% and 47% (Allen et al., 2004; Cramer et al., 2000; Won et al., 2004), with acetaminophen, a recommended pharmacologic agent for mild to moderate musculoskeletal pain (AGS, 1998) most commonly administered. The administration of 2,600 milligrams (mg.) acetaminophen over 24 hours did not provide relief in nursing home residents with degenerative joint diseases, fractures, or back pain (Buffum et al., 2004). In contrast, acetaminophen 1000mg. or 650 mg. administered four times a day provided pain relief for nursing home residents with pain of unknown etiology (Cohen-Mansfield & Lipson, 2007). NSAIDs were prescribed for anywhere from 4% to over 70% of nursing home residents (Cramer et al., 2000; Hutt et al., 2006; Won et al., 2004). High doses of NSAIDs were administered for nearly one-third of the residents. The use of nonopioids was not associated with adverse events such as gastrointestinal bleeding, congestive heart failure, peripheral edema, or renal failure (Won et al., 2004). Opioid administration varied from 23% to 67.6% (Allen et al., 2003; Cramer et al., 2000; Won et al., 2004). The combination of acetaminophen, 1000 mg. with oxycodone 2.5 mg. or 5 mg. provided pain relief for 6 residents with pain of unknown etiology (Cohen-Mansfield & Lipson, 2007). Nonpharmacologic interventions were not evaluated in these studies.

This study builds on previous research that evaluated selected analgesics for pain management by comparing current pharmacologic and nonpharmacologic interventions from multiple Iowa nursing homes to the current 1998 AGS guideline. The questions guiding this study are as follow: 1). What musculoskeletal diagnoses contributed to self-reports of pain? 2). To what extent was the AGS guideline used as a standard of care for the management of chronic pain?

Method

Design and Sample

As part of a longitudinal study on delirium among older adults in rural nursing homes (Culp et al., 2004), this study focused on evaluating the nature and management of pain associated with musculoskeletal diagnoses. The sampling technique for the parent study was to randomly select 6 of the 15 rural counties in southeast Iowa. Thirteen nursing homes in the 6 counties were then randomly stratified based on facility size, with half of the nursing homes having 75 beds or less and half having 75 or more. Residents were include in the study if they: (a) could read, write and speak English; (b) were 65 years of age or older; (c) had no admitting diagnosis of a psychosis, head trauma, brain tumor, or toxin-related neurological disorders; (d) were admitted to a skilled or intermediate care facility for at least 30 days; (e) had no implanted defibrillator; (f) had no admitting diagnosis of delirium or current delirium symptoms; (g) had no indwelling urinary catheter, and h) were not on dialysis. Residents with dementia were included in the protocols. However, the severity of cognitive impairment precluded participation by those unable to respond to the pain questions.

Procedures

Protocols were reviewed by the University of Iowa Institutional Review Board (IRB), and consent was obtained from either the participant or the legal guardian. Registered nurses, who served as research assistants (RAs), abstracted demographic information, medical diagnoses, and pain data from the most recent MDS, and medications upon admission to the study. The RAs completed the face-to-face pain assessment and the Mini Mental State Examination (MMSE) upon admission to the study. The data collection period began in 1999 and continued through 2003.

The evidence-based guideline, The Management of Chronic Pain in Older Persons (AGS, 1989), was selected as the standard for evaluating the management of chronic pain related to musculoskeletal diagnoses. The 1998 version coincided with the beginning of the data collection period. The guideline focuses on a comprehensive pain assessment and principles of pharmacologic interventions. Although the guideline notes the benefits of combining pharmacologic and nonpharmacologic interventions, nonpharmacologic interventions are not identified in the guideline (AGS, 1998). Pharmacologic recommendations include acetaminophen, not to exceed 4000 mg/24 hours, aspirin, and Nonsteroidal Anti-Inflammatory Drugs (NSAIDs), such as ibuprofen, naproxen, and trisalicylate. High doses of NSAIDs for long periods of time are discouraged. Examples of adjuvant analgesic drugs found to be useful for mainly neuropathic pain include: prednisone, antidepressants, anticonvulsants, anti-arrhythmics, local anesthetics, and baclofen. Although, the use of opioids for noncancer pain is controversial, suggested opioid analgesics include: short-acting drugs (morphine sulfate, hydrocodone, oxycodone, and hydromorphone) and long-acting drugs (sustained release morphine and oxycodone, duragesic). Analgesics are best given on a regularly scheduled basis (AGS, 1998).

Measures

Diseases of the musculoskeletal system and fractures, known to cause pain, were of interest for this study. Medications of interest were acetaminophen, NSAIDs, Cox-2 inhibitors, opioids, salicyates, and anticonvulsants. Uric-acid lowering medications to manage gout were not included in the analysis.

The face-to-face pain assessment included four questions with a yes/no response to indicate presence/absence of pain. The questions asked: “Have you experienced any pain or discomfort in the last 24 hours?”; “Would you say you have pain or discomfort every day?”; “Would you say the pain or discomfort clouds your thinking?”; “Would you say the pain medication clouds your thinking?” Residents were also asked to report the location of their pain and ways used to ease their pain or discomfort, outside of pain medications.

The alpha reliability for the four pain questions was 0.69 (n = 206). To establish content validity, the authors sought input from five nurse experts, all selected because of their clinical and research experience about pain in nursing home residents. The nurse experts rated the four pain questions using a 1 to 4 scale with 1 meaning irrelevant and 4 meaning extremely relevant. The content validity ranged from 0.80 to 1.00.

The pain section from the MDS provided the second source of pain assessment data. This assessment included pain frequency, intensity, and location. The code for pain frequency is 0 (no pain), 1 (pain less than daily), and 2 (pain daily). The code for pain intensity is 1 (mild pain), 2 (moderate pain), and 3 (horrible or excruciating). Pain sites were limited to nine body locations. The kappa statistic for the MDS pain section and the Visual Analogue Scale among nursing home residents was 0.70 (Fries, Simon, Morris, Flodstrom, & Bookstein, 2001).

The Mini-Mental State Examination (MMSE) was used to assess the residents’ mental status. The MMSE is a standard assessment procedure with established reliability and validity for use with older adults in nursing homes (Nadler et al., 1995). The MMSE contains questions that assess orientation, registration of information, attention and calculation, recall, language, and visual construction. A score of 23 or less indicates a cognitive impairment (Tombaugh & McIntyre, 1992).

Analysis

Descriptive statistics were used to describe the sample, their musculoskeletal diagnoses, pain assessments, and pharmacologic and nonpharmacologic management of their pain. Cross tabulations were completed for the diagnoses of pain in musculoskeletal conditions with scheduled and PRN pharmacological interventions. A one-way analysis of variance (ANOVA) was computed to determine any difference between the number of analgesics administrated for cognitively impaired and cognitively intact nursing home residents.

Results

Sample

The sample, 215 nursing home residents, was predominately female (75.8%, n = 163), widowed (69.8%, n = 150), and White (99.5%, n = 214) with a mean age of 86.4 years (SD = 7.27). The mean MMSE score was 22 (SD = 5.53; n = 210) with a range of 9 to 30. Of those who participated, 51.6% (n = 111) had a documented diagnosis of dementia.

Etiology of Pain

Among the 215 residents, 166 (77 %) had pain associated with a musculoskeletal etiology. Among the residents self-reporting pain, 49 (29.5%) did not have a documented musculoskeletal diagnosis, yet 47 (95.9%) reported musculoskeletal pain. When a musculoskeletal diagnosis was documented, the most common diagnosis contributing to pain was osteoarthritis (n = 102, 47.5%). Eighty-one (37.7%) residents had at least one musculoskeletal diagnosis associated with pain, 85 (39.5%) had two, and 33 (15.3%) had three. The most common combination of musculoskeletal diagnoses was degenerative joint disease or osteoarthritis or arthropathy (n = 143, 66.5%) followed by osteoporosis and osteoarthritis (n = 139, 64.7%). Additionally, 20 (9.3%) residents with musculoskeletal diagnoses also had neoplasms.

Pain Assessments

All residents (n = 215, 100%) completed the face-to-face pain assessment question about the presence of pain now. Eighty-five (39.5%) residents indicated pain now and 70 (32.6%) indicated pain daily. Twenty-two residents (10.2%) indicated that pain clouds their thinking and 9 residents (4.2%) reported that medications cloud their thinking. In contrast, pain assessment data abstracted from the most recent MDS indicated that only 46 residents (21.4%) had daily pain with over half of these residents rating their pain as mild or moderate (Table 1).

Table 1.

Painful Symptoms Identified in the Last 7 days from the MDS

Pain Frequency	Number	Percent
None	119	55.3
Pain Less than Daily	50	23.5
Pain Daily	46	21.4
Pain Intensity	Number	Percent
Mild	43	19.5
Moderate	50	23.3
Horrible or Excruciating	3	1.4
Pain Location	Number	Percent
Joint (other than hip)	43	20.0
Back	32	14.9
Hip	18	8.4
Headache	11	5.1
Soft Tissue	9	4.2
Bone	3	1.4
Incisional	3	1.4
Stomach	2	0.9
Chest pain while doing activity	2	0.9

Percents based on total sample size (n = 215)

Pharmacologic Management

Nonopioids and opioids were administered to the nursing home residents. Of the nonopioids, acetaminophen was administered most frequently, followed by NSAIDs and COX-2 inhibitors. The specific opioids included propoxyphene, hydrocodone (2.5 mg./ 500 mg.) and hydrocodone (5 mg./500 mg.) (Table 2). Ninety-six (44.7%) residents had at least one analgesic prescribed and administered for pain, 24 (11.1%) residents had two analgesics, and one resident (0.5%) had three analgesics.

Table 2.

Musculoskeletal Diagnoses and Residents with Prescriptions

Diagnoses	Acetaminophen	NSAIDs	Cox-2	Opioids	Salicyates	Anti- convulsants	Analgesics Total (%)	No Analgesics Total (%)
DJD or OA or Artho	30	5	8	13	1	0	57 (66.3)	29 (33.7%)
Gout	1	0	0	0	1	1	3 (70.0)	2 (30.0%)
OP	3	1	1	1	0	1	7 (43.7)	9 (56.3%)
Gout & DJ & OP	0	0	0	1	0	0	1 (100.0)	0 (0%)
S/P Fx	1	0	1	1	0	0	3 (42.9)	4 (57.1%)
Fx & [OA or DJD] & OP	3	1	1	1	0	1	7 (87.5)	1 (12.5%)
OP & OA	10	0	0	0	2	0	12 (85.7)	2 (14.3%)
DJD & Gout	0	2	0	1	0	1	4 (66.7)	2 (33.3%)
LBP & [OA or DJD or OP]	1	1	1	1	0	0	4 (80.0)	1 (20.0%)

Abbreviations: OA = osteoarthritis, OP = osteoporosis, Arthro = arthropathy DJD = degenerative joint disease, S/P = status post, Fx = fracture, and LBP = low back pain.

Forty-five percent (n = 98) of the residents had at least one scheduled analgesic and 32.3% (n = 69) received analgesics on a PRN basis. For both the scheduled (n = 53, 24.7%) and PRN administration (n = 55, 25.6%), acetaminophen was administered most often. Thirteen (6%) residents had scheduled and PRN acetaminophen. Propoxyphene was prescribed for 23 (10.7%) residents and administered to nine (4.2%) residents on a PRN basis. PRN analgesics were administered for reasons of general discomfort and aches; pain in the upper and lower extremities, hip and back; headache; and dental pain.

The total number of scheduled and PRN analgesics was tabulated and compared with the face-to-face pain assessment. Of those reporting “pain now” (n = 85), 29 residents (34.1%) were not receiving any scheduled or PRN analgesics. In contrast, of those who reported “no pain now,” 37 (28.4%) residents received a scheduled or PRN analgesic. Of those who reported having pain on a daily basis (n = 70), 23 (32.9%) nursing home residents did not receive any scheduled or PRN analgesics.

The percent of residents who received or did not receive analgesics for their pain was calculated. Residents with degenerative joint disease, osteoarthritis, or arthropathy (n = 57, 66.3%) received analgesics for pain management. Residents who were status post fracture (n = 4, 57.1%), osteoporosis (n = 9, 56.3%), or gout (n = 2, 30%) were least likely to receive analgesics (Table 2).

Nonpharmacologic Management

Of the nonpharmacologic interventions, topical agents were most commonly used for management of pain. Twenty-nine (13.5%) residents reported using agents such as Aspercream or Bengay, unspecified therapy, or topical heat. While the residents reported the use of the topical agents to the RAs, these interventions were not documented in their medical records.

Analgesic Administration and Cognition

Although 111 (51.6%) residents had a diagnosis of dementia in their medical records, all were able to respond to the face-to-face pain question about having pain now. Although not statistically significant, residents with lower MMSE scores (n=98, 46%) had one or more scheduled analgesic. PRN analgesics were more likely ordered for nursing home residents with dementia (n = 39, 35.1%) compared to cognitively intact nursing home residents (n = 25, 24.1%). Acetaminophen was the most commonly administered analgesic for residents with dementia (n = 19, 21.3%) and cognitively intact residents (n = 32, 26.9%).

Discussion

Findings from this study suggest that the 1998 guideline was inconsistently implemented in multiple rural nursing homes in which residents had multiple musculoskeletal diagnoses that contribute to high levels of pain (Cohen-Mansfield & Lipson, 2002). The underestimation of MDS-based pain experienced by nursing home residents compared to face-to-face assessments is similar to other research results (Cohen-Mansfield, 2004; Fisher et al., 2002; Teno et al., 2002; Won et al., 2004). One explanation for the underestimation is a decrease in communication skills by elders with cognitive impairments. Although over half of the residents in this study have a diagnosis of dementia, every resident could answer the question about pain now. This discrepancy emphasizes the need to use multiple pain assessment sources, to evaluate the usefulness of the MDS for pain assessment, and to consistently document pain.

The common pharmacologic intervention in this study is acetaminophen, and its administration is consistent with recommendations from the AGS guideline (1998) and previous research. The range of acetaminophen providing pain relief varied in two samples, one with musculoskeletal diagnoses (Buffum et al., 2004) and one with unidentified pain etiology (Cohen-Mansfield & Lipson, 2007). Thus a comparison between the two studies was limited due to the pain etiology.

The use of NSAIDs for the Iowa nursing home residents was found to be less than other studies. NSAIDs were routinely prescribed and administered to nearly one-fourth of the nursing home residents (Cramer et al., 2000) and one-third of the nursing home residents received high doses of NSAIDs (Won et al., 2004). The AGS guideline (1998) recommends that high doses of NSAIDs should be avoided for long-term use. A small number of Iowa residents received COX-2 inhibitors, a drug not included in the 1998 guideline, but listed in the 2002 AGS guideline. As with all NSAIDs, COX-2 selective NSAIDs must be used with caution in residents with high cardiovascular risk.

Based on the 1998 guideline, opioids are controversial for chronic pain management. Fewer opioids were administered to residents in this study compared to previous studies (Cramer et al., 2000; Won et al., 2004). One study involving opioids specified in the AGS guideline reports that acetaminophen with oxycodone provided pain relief (Cohen-Mansfield & Lipson, 2007). Propoxyphene was a drug of choice in this study as well as others (Cramer et al., 2000; Hutt et al., 2006; Won et al., 2004). One study involving opioids specified in the AGS guideline reports that acetaminophen with oxycodone provided pain relief (Cohen-Mansfield & Lipson, 2007).

A combination of pharmacologic and nonpharmacologic interventions is recommended for the management of chronic pain. The limited use of topical applications by these residents is consistent with previous research findings (Cramer et al., 2000, Ferrell et al., 1995, Won et al., 1999). Also contributing to the lack of use of topical agents was the limited nursing home staff knowledge about nonpharmacologic pain management (Jones et al., 2004).

This study identified nine residents who reported that the administration of pain medications caused “clouding my thinking.” Clinicians may be hesitant to prescribe analgesics, especially opioids, for the elderly because of possible cognitive effects. The completion of a history and physical examination is necessary to determine the cause of the “clouding my thinking,” as a change in mental status could indicate pain (AGS, 2002).

PRN administration of analgesics exists in clinical practice, with rates ranging from 32.3% in this study to 32% to 63.2% (Allen et al., 2003; Cramer et al., 2000; Won et al., 2004). Unfortunately, the use of PRN analgesics is dependent upon the cognitively impaired resident requesting pain medication or the nurse asking if the resident is in pain or being able to recognize cues of pain in a cognitively impaired resident. The nurse must also have an order to administer an analgesic, as one third of the residents who reported pain were lacking an order for analgesics. In other samples, the percentage of residents who do not receive any analgesics varied from 25% to 30% (Allen et al., 2003; Won et al., 2004).

Limitations

A limitation for this study is the lack of pain intensity from the face-to-face pain assessments. Furthermore, the effectiveness of the pharmacologic interventions was not evaluated and no reassessment followed the administration of analgesics. This study was conducted at the time of the 1998 AGS publication and reflects the state of pain management practices at that time, rather than the actual compliance with the guideline.

In summary, these research findings suggest that nursing home residents experience unresolved pain or discomfort. Research is needed to evaluate the implementation of the 2002 AGS pain guideline to evaluate pain assessment and management of nursing home residents, as there are differences with the 1998 guideline. For example, the 1998 guideline lacks information on COX-2 inhibitors, nonpharmacologic interventions, and a section on opioids of concern. Specifically, the 2002 guideline includes a specific precaution about the use of propoxyphene in older adults. Future research needs to include the use of multiple pain assessment instruments and to evaluate the effectiveness of pharmacologic and nonpharmacologic interventions for the management of chronic pain in older adult nursing home residents.