HomeLarge Type Edition
HOME ARCHIVE SEARCH TABLE OF CONTENTS
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
PubMed
Right arrow PubMed Citation
The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 59:M281-M285 (2004)
© 2004 The Gerontological Society of America

A Minimum Data Set Prevalence of Pain Quality Indicator: Is It Accurate and Does It Reflect Differences in Care Processes?

Mary P. Cadogan1,2,, John F. Schnelle1,3,4, Noriko Yamamoto-Mitani1,2, Georgina Cabrera1 and Sandra F. Simmons1,3

1 Los Angeles Jewish Home for the Aging, UCLA Borun Center for Gerontological Research, Reseda, California.
2 UCLA School of Nursing, Los Angeles, California.
3 UCLA Department of Medicine, Division of Geriatrics, Los Angeles.
4 Veterans Administration Hospital, Sepulveda, California.

Address correspondence to Mary P. Cadogan, DrPH, RN, CGNP, UCLA School of Nursing, Factor 5-952, Box 956919, Los Angeles, CA 90095-6919. E-mail: mcadogan{at}ucla.edu


   Abstract
Top
 Abstract
Methods
 Results
 Discussion
 References
 
Background. A new prevalence of pain quality indicator based on the Minimum Data Set (MDS) has been developed for the purpose of providing nursing home (NH) consumers with information that would allow them to compare the quality of pain care between facilities. The purpose of this study was to compare pain-related care processes between NHs that scored in the lower 25th percentile (low pain prevalence) or upper 75th percentile (high pain prevalence) based on this indicator.

Methods. Care processes related to pain assessment, documentation, and treatment were independently evaluated using standardized resident interview and medical record review protocols for 255 residents in 16 NHs that reported MDS pain prevalence of less than 15% (8 NHs in lower 25th percentile) or greater than 30% (8 NHs in upper 75th percentile).

Results. A significantly greater proportion of participants in the high pain prevalence NHs reported symptoms indicative of chronic pain during interview. The standardized pain interview revealed a significantly higher prevalence of pain among participants in the lowest quartile NH group compared to the MDS pain prevalence quality indicator, but the pain prevalence according to both MDS and interview were comparable in the higher quartile NHs. Medical record review showed that a significantly greater proportion of participants in upper quartile NHs had pain assessments documented by licensed nurses and physicians, received pain medication, and had documentation of treatment response.

Conclusions. An MDS pain quality indicator accurately discriminates prevalence of pain between facilities. However, interpretation of the pain indicator requires caution. Rather than reflecting poor quality, a high prevalence of pain according to the MDS was associated with better pain assessment and treatment care processes.

SINCE 1998, nursing home (NH) staff have been required to assess the frequency and intensity of pain for all residents using the Minimum Data Set (MDS) protocol (1). The MDS is a standardized assessment required for all NH residents upon admission and quarterly thereafter, or when there is a significant change in condition. The MDS is intended to assess multiple aspects of a resident's functional status including, but not limited to, pain. A pain quality indicator is currently being introduced based on MDS data for the purpose of providing NH consumers with information to evaluate the quality of pain care in facilities (2). Specifically, the prevalence of pain documented on the MDS after admission is suggested as an indicator of new or persistent pain that may reflect differences in pain care between NHs. This new pain quality indicator is based on the assumption that lower MDS pain prevalence reported by facilities is reflective of better quality of pain care (e.g., assessment and treatment) than higher MDS pain prevalence reports. However, in view of multiple studies demonstrating a high prevalence and underdetection of pain among NH residents (3–8), one might hypothesize that different scores on this indicator may reflect differences in how well NH staff assess and document pain among residents such that a high prevalence of pain reported on the MDS may reflect better detection and pain care quality.

MDS pain data were collected for all participants as part of a larger study comparing quality of care for 7 conditions in 30 NHs. MDS pain prevalence rates ranged from a low of 4% to a high of 55% across the 30 NHs. This disparity in MDS pain prevalence among the 30 study NHs raised the following research question to be addressed in this study: Do NHs that score in the lower 25th percentile on MDS pain prevalence differ from NHs that score in the upper 75th percentile?

The goals of this study were to:


   METHODS
Top
 Abstract
 Methods
Results
 Discussion
 References
 
Subjects and Setting
Participants were part of a larger study designed to determine whether NHs that report different MDS-derived quality indicator data across 7 conditions also implement different processes of care (9). Processes of care related to pain assessment and management were included as part of the larger study but were analyzed separately because a quality indicator based on MDS pain prevalence data had not been developed at the time of the study. Pain prevalence rates were based on each participant's most recent MDS assessment, excluding the initial admission assessment.

Figure 1 shows the recruitment of study participants. Of the total resident population in the 30 NHs (3095 residents), written consent was obtained from either the resident or a responsible party designated in the medical record for 907 residents (29%). Consent procedures were approved by the University of California, Los Angeles, Office for the Protection of Research Subjects. Sixteen of the 30 NHs were either in the lower 25th percentile (8 NHs; pain prevalence range 4%–14%) or the upper 75th percentile (8 NHs; pain prevalence range 31%–55%) according to the MDS pain prevalence data for participants (n = 448). Complete resident interview and medical record review data were obtained for a subsample of 255 participants (Figure 1).



View larger version (29K):
[in this window]
[in a new window]
  		Figure 1. Recruitment of participants. NHs = nursing homes; MDS = Minimum Data Set

 
The 16 study NHs ranged in size from 59 to 210 beds (mean occupancy rate per NH = 94 ± 42 residents). Resident-to-nurse-aide ratios, as reported by the directors-of-nursing across the 16 NHs, averaged 7.17 (±1.8) residents per aide on the day shift (7 AM to 3 PM) and 11.6 (±2.4) residents per aide on the evening shift (3 PM to 11 PM). Licensed nursing staff ratios averaged 14.4 (±4.3) and 33.2 (±11.4) residents per licensed nurse for day and evening shifts, respectively.

Measures
Medical Record Review.-- The medical records of 255 participants in the 16 study NHs were reviewed by a trained research physician and/or a gerontological nurse practitioner using a standardized protocol. The interrater reliability for all data elements collected through medical record review was excellent (kappa values ranged from.65 to 1.00). A maximum of 20 charts per facility was reviewed in NHs for which over 20 residents provided consent.

The presence of three pain-related diagnoses (osteoarthritis, cancer, and recent fracture) was abstracted for up to 1 year prior to the date of on-site medical record review for each participant. Pain was noted to be present on each participant's most recent MDS, excluding the initial admission assessment, if there was documentation of mild, moderate, or excruciating pain daily (MDS items j2a = 2 and j2b = 1, 2, or 3) or moderate-to-excruciating pain less than daily (MDS items j2a and j2b = 2 or 3). In addition, 4 MDS items relevant to cognitive status were also retrieved from each participant's most recent MDS (Section B: Cognitive Patterns, items 3a–3d as listed in Table 1 legend: MDS Recall Score). These 4 items comprise a memory/recall score that ranges from 0 (rated by NH staff as unable to accurately recall any items) to 4 (rated by NH staff as able to accurately recall all 4 items). An MDS-derived recall score of 2 or greater has been shown to be indicative of NH residents' ability to reliably respond to interview questions about the care that they receive (10). The quality of pain-related care processes was measured in two broad areas: detection and management. Quality of pain detection was evaluated through physician and licensed nurse documentation of pain in residents for whom there was independent evidence of pain. Independent evidence of pain was defined as either MDS documented pain or symptoms indicative of chronic pain endorsed in the resident interview.


View this table:
[in this window]
[in a new window]
  		Table 1. Demographic and Pain Characteristics of 225 Participants.

 
The frequency and content of physician and licensed nurse assessment of pain was obtained for each resident for up to 1 year or to the date of admission, if less than 1 year. Physician orders for both regularly scheduled and "as needed" (PRN) pain medications were collected for each participant. Data were obtained regarding the frequency of administration of both scheduled and PRN pain medications in the previous month and physician documentation of response to treatment for all participants who had any order for pain medication. Because of the California law enacted in 2000 requiring pain to be assessed as a fifth vital sign (11), vital sign records and/or licensed nurse weekly summaries were reviewed for the month prior to the date of on-site medical record review and resident interview. These reviews were done to determine the frequency of pain assessment as the fifth vital sign and to identify the highest level of pain recorded in the medical record during this 1-month period. Licensed NH staff in all 16 study NHs used a 0 (no pain) to 10 (severe pain) rating scale for pain as a fifth vital sign.

Resident interview.-- Research staff attempted to conduct a standardized 7-item pain interview (yes/no response format for each question) with all 448 participants, and 255 completed the pain interview. Of these 255, 181 had an MDS recall score of 2 or greater (Figure 1) and data from this subsample were used for selected analyses. The first 4 interview questions based on the Geriatric Pain Measure (12) elicited information about the frequency and functional impact of pain symptoms (listed in Table 2 legend: Do you have pain now?, Do you have pain every day?, Does your pain keep you from sleeping?, Does your pain keep you from participating in activities?). Participant responses to each of these 4 questions were scored as 1 (yes) or 0 (no) with a possible total pain score range from 0–4. Participants with a total pain score of 3 or higher and/or the report of daily pain alone (i.e., a "yes" response to the question "Do you have pain every day?") were classified as having chronic pain.


View this table:
[in this window]
[in a new window]
  		Table 2. Pain Self-Reports, Assessment, Treatment, and Documentation in 181 Participants With MDS Recall Scores of 2 or Greater.

 
The remaining 3 pain interview questions (Table 2) were related to communication about pain between residents and NH staff (Do you tell the nurse about your pain?, Does the nursing staff ask you about your pain?) and treatment preferences (Would you prefer to take medication for your pain?). Interrater reliability data showed excellent agreement for all 7 pain interview questions (kappa value range =.83-1.0, p <.01). In addition, repeat interviews were conducted within 24 hours for a sample of 43 participants, and results showed good stability of chronic pain status (kappa =.65, p <.01).

Data analyses.-- Demographic and clinical characteristics of the participants within each group of NHs were compared with t tests for independent samples for continuous variables (e.g., age, length of residency) and chi-square ({chi}2) analyses for categorical variables (e.g., gender, insurance status, ethnicity, proportion with an MDS recall score of 2 or greater, presence of any pain-related diagnosis). Chi-square analyses were also used to compare medical record care process measures of pain assessment and treatment. Data are reported only for participants with chronic pain (endorsed 3 or more symptoms or daily pain during interview) for some comparisons. The rationale for this analysis was that, at a minimum, pain should be detected by NH care providers for residents who endorse multiple pain symptoms and who are rated by NH staff as cognitively aware.


   RESULTS
Top
 Abstract
 Methods
 Results
Discussion
 References
 
Demographic Characteristics
Demographic characteristics of the 255 participants are displayed in Table 1 separately for the 2 groups of NHs. A significantly greater proportion of participants in the upper quartile NHs (high MDS pain prevalence) were Caucasian (Table 1: 76% vs 43%; {chi}2 = 41.9, p <.01). In addition, a significantly smaller proportion of participants in the upper quartile NHs had MediCal payment status (Table 1: 58% vs 69%; {chi}2 = 41.3, p <.01). There were no other demographic characteristics that showed significant differences in the 2 consented samples, including the proportion of participants with an MDS recall score of 2 or greater in each group.

Pain Assessment and Treatment
Because the 16 NHs were explicitly selected based on their extreme (quartiles) MDS pain prevalence scores, there was a significant difference between the 2 groups of NHs on the proportion of participants who had MDS documentation of pain (Table 1: 10% vs 40% in the lowest and highest quartile NHs, respectively; {chi}2 = 33.5, p <.01). There was also a significant difference between participants in the 2 groups of NHs based on the highest pain scores documented by licensed nurses as the fifth vital sign during the month immediately preceding medical record review and resident interview. Based on the 0 (no pain) to 10 (severe pain) rating scale, 11% percent of participants in the lowest quartile NHs had pain scores greater than zero compared to 29% in the highest quartile NHs (Table 1: {chi}2 = 9.92, p <.01). In addition, a significantly smaller proportion of participants in the lowest quartile NHs had a physician assessment of their pain at any time during the year prior to medical record review (Table 1: 47% vs 68%; {chi}2 = 12.28, p <.001). There was no difference between the 2 groups in the proportion of participants with any pain-related diagnosis, but a significantly smaller proportion of participants in the lowest quartile NHs received pain medication (Table 1: 24% vs 39%; {chi}2 = 6.38, p <.01), and significantly fewer participants receiving pain medication in the lowest quartile NHs had a physician note documenting response to treatment (Table 1: 26% vs 57%; {chi}2 = 8.54, p <.01).

Pain Reports, Assessment, and Treatment in 181 Participants With MDS Recall Scores >=2
Table 2 shows the comparison of MDS pain prevalence (9% vs 47% in the lower and upper pain quartiles, respectively; {chi}2 = 31.11, p <.01), and results from the pain interview for participants with an MDS recall score of 2 or higher (n = 181). There were no significant differences between groups on the proportion of participants who reported that the nursing staff asks them about pain (Table 2: 45% vs 56% in the lower and upper quartile groups, respectively). However, a significantly greater proportion of participants in the upper quartile NHs reported that they tell the nursing staff about their pain (Table 2: 87% vs 70%, {chi}2 = 7.05, p <.01). More than half of all participants with an MDS recall score of 2 or greater in both lower and upper quartile NHs responded "yes" to at least 1 of the 4 pain interview questions (Table 2: 52% vs 64%, respectively). However, a significantly greater proportion of participants in the upper quartile NHs reported symptoms indicative of chronic pain (Table 2: 47% vs 27%; {chi}2 = 9.79, p <.01). Furthermore, in the upper quartile NH group, the exact pain prevalence of 47% was documented both on the MDS and the independent interview (although residents were not identical in both groups). In contrast, in the lower quartile NH group, MDS pain prevalence was significantly lower than the prevalence estimate based on resident interview (9% vs 27%; {chi}2 = 5.62, p <.05). Among participants reporting symptoms of chronic pain during the interview, detection of pain by NH staff according to MDS documentation indicated that 82% in the lowest quartile group had "no pain" versus 31% in the highest quartile group (Table 2: {chi}2 = 15.88, p <.01). There was not a significant difference between the lower and upper quartile groups on the proportion of participants with chronic pain symptoms who reported a preference for being treated with pain medication (Table 2: 83% vs 73%, respectively). However, a significantly greater proportion of participants with chronic pain in the upper quartile NHs received pain medication (Table 2: 67% vs 43%; {chi}2 = 4.06, p <.05).


   DISCUSSION
Top
 Abstract
 Methods
 Results
 Discussion
References
 
This is the first study to independently evaluate the accuracy and validity of an MDS pain prevalence quality indicator. The results of this study showed that the MDS prevalence of pain indicator discriminated between NHs in the lowest versus the highest pain quartiles. Specifically, there was a significantly greater proportion of participants who endorsed symptoms of chronic pain during interview in homes with high MDS pain prevalence indicator scores as compared to homes with low MDS pain prevalence indicator scores. In fact, among the subgroup of participants with MDS recall score of 2 or higher in the upper quartile NH group, pain prevalence documented on the MDS was identical to that obtained through independent interview, and consistent with prevalence reports from other studies (4–6,13–15). However, there were some differences in classification of participants, because not all of those with chronic pain on interview were identified on the MDS. In contrast, in the lower quartile NHs, the MDS pain prevalence of 9% was significantly lower than the 27% pain prevalence obtained through resident interview and lower than the pain prevalence reported in other studies (4–6,13–15). The pain prevalence estimates obtained through independent interview in this study should be viewed conservatively because classification of pain was limited to participants with severe and/or daily pain and was based on reports from participants who had MDS recall scores of 2 or higher.

On every measure of pain-related care quality independently evaluated in this study (detection, assessment, treatment, and documentation of treatment response), NHs with a high reported prevalence of pain on the MDS performed better than NHs with low MDS pain prevalence. Specifically, more participants with symptoms of chronic pain in the highest quartile NHs also had pain documented by indigenous, licensed NH staff, documentation of a physician pain assessment, and received pain medications. These results suggest that an MDS-derived pain quality indicator could be informative because it detects both differences in pain prevalence between NHs and differences in care related to pain assessment and treatment. However, the interpretation of an MDS pain prevalence quality indicator requires caution because NHs with high MDS pain prevalence in this study demonstrated better pain-related care processes according to multiple measures than NHs with low pain prevalence. This interpretation is counter-intuitive to the interpretation of most MDS-derived quality indicators where a high prevalence is assumed to be associated with poor care quality (e.g., incontinence, pressure ulcers, weight loss).

It is clear from the results of this study, which included relatively small sample sizes, that a larger study should be conducted to further assess the accuracy and meaning of the MDS pain prevalence quality indicator prior to widespread use of the indicator for improvement purposes or making the information available to NH consumers. Why NHs with higher reported rates of MDS pain did a better job of detection and management remains unanswered. One explanation may be that higher prevalence of pain according to both interview and MDS measures could sensitize NH staff to the need for better overall care for pain. An additional finding from this study also warrants further evaluation. The striking demographic differences noted between participants in the extreme quartile NHs raise concerns that the quality of pain care may differ based on insurance status and/or ethnicity. In this study, non-Caucasians and participants with MediCal payment status were over-represented in the lowest quartile NHs, which demonstrated poorer quality of pain care as defined in this study and compared to NHs in the highest quartile. Clearly, future efforts to understand differences in the quality of pain care must also carefully consider ethnicity and payment status in the research plan.


   Acknowledgments
 
Prepared for the California HealthCare Foundation #99-504. The views expressed in this paper are those of the authors and may not reflect those of the Foundation.

Received November 7, 2002

Accepted February 13, 2003


   References
Top
 Abstract
 Methods
 Results
 Discussion
 References
 

  1. Health Care Financing Administration. Long Term Care Facility Resident Assessment Instrument (RAI) User's Manual, Minimum Data Set version 2.0. Natick, MA: Eliot Press; 1999.
  2. Harris Y, Clauser S. Achieving improvement through nursing home quality measurement. Health Care Financ Rev.. 2002;23:5-18.[Medline]
  3. Feldt K. Improving assessment and treatment of pain in cognitively impaired nursing home residents. Ann Long-Term Care.. 2000;8:26-42.
  4. Ferrell BA. Pain management in elderly people. J Am Geriatr Soc.. 1991;39:64-73.[Medline]
  5. Ferrell BA, Ferrell BR, Rivera L. Pain in cognitively impaired nursing home residents. J Pain Sympt Manage.. 1995;10:591-598.[Medline]
  6. Sengstaken E, King S. The problems of pain and its detection among geriatric nursing home residents. J Am Geriatr Soc.. 1993;41:541-544.[Medline]
  7. Tait R. Assessing pain in older adults. Int Psychogeriatrics.. 2000;12:421-424.
  8. Fries B, Simon S, Morris J, Flodstrom C, Bookstein F. Pain in U.S. nursing homes: validating a pain scale for the minimum data set. Gerontologist.. 2001;41:173-179.[Abstract/Free Full Text]
  9. Schnelle JF, Cadogan MP, Bates-Jensen B, Simmons SF, Osterweil D. A Field Test Evaluation of Clinical and Staffing Quality Indicators for Use in a Consumer Reporting System. Report to California Healthcare Foundation; October 2002.
  10. Simmons S, Schnelle J. The identification of residents capable of accurately describing daily care: implications for evaluating nursing home quality. Gerontologist.. 2001;41:605-611.[Abstract/Free Full Text]
  11. California Assembly Bill 791 (2000).
  12. Ferrell B, Stein W, Beck J. The Geriatric Pain Measure: validity, reliability, and factor analysis. J Am Geriatr Soc.. 2000;48:1669-1673.[Medline]
  13. Parmalee P, Smith B, Katz I. Pain complaints and cognitive status among elderly institution residents. J Am Geriatr Soc.. 1993;41:517-522.[Medline]
  14. Won A, Lapane K, Gambassi G, Bernabei R, Mor V, Lipsitz L. Correlates and management of nonmalignant pain in the nursing home. SAGE Study Group. Systematic Assessment of Geriatric drug use via Epidemiology. J Am Geriatr Soc.. 1999;47:936-942.[Medline]
  15. Fox P, Raina P, Jadad A. Prevalence and treatment of pain in older adults in nursing homes and other long-term care institutions: a systematic review. Can Med Assoc J.. 1999;160:329-333.[Abstract]



This article has been cited by other articles:


Home page
 GerontologistHome page
H. B. Degenholtz, J. Rosen, N. Castle, V. Mittal, and D. Liu
The Association Between Changes in Health Status and Nursing Home Resident Quality of Life
Gerontologist, October 1, 2008; 48(5): 584 - 592.
[Abstract] [Full Text] [PDF]

Home page
 GerontologistHome page
J. Zinn, W. Spector, L. Hsieh, and D. B. Mukamel
Do Trends in the Reporting of Quality Measures on the Nursing Home Compare Web Site Differ by Nursing Home Characteristics?
Gerontologist, December 1, 2005; 45(6): 720 - 730.
[Abstract] [Full Text] [PDF]

Home page
 GerontologistHome page
B. M. Bates-Jensen, S. F. Simmons, J. F. Schnelle, and C. Alessi
Evaluating the Accuracy of Minimum Data Set Bed-Mobility Ratings Against Independent Performance Assessments: Systematic Error and Directions for Improvement
Gerontologist, December 1, 2005; 45(6): 731 - 738.
[Abstract] [Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Services
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
PubMed
Right arrow PubMed Citation

HOME ARCHIVE SEARCH TABLE OF CONTENTS