The PEACE Project: Identification of Quality Measures for Hospice and Palliative Care

Abstract

Context:

In 2008, the Centers for Medicare & Medicaid Services (CMS) required U.S. hospices to implement comprehensive quality improvement programs. CMS contracted with the Quality Improvement Organization in North and South Carolina to develop quality measures and instruments to assess hospice and palliative care quality.

Objectives:

To develop a set of quality measures, with complete specifications, and data collection tools for use by hospice and palliative care providers in quality improvement.

Methods:

Quality measures were identified from: published literature, the National Quality Forum, CMS measures, the National Quality Measures Clearinghouse, and measures submitted by two national hospice organizations. Available data on the quality measures were gathered and pilot data were collected for measures with no availble data. A Technical Expert Panel (TEP) rated quality measures on: importance, scientific soundness, feasibility and usability, using numeric scores for each dimension. Scores for quality measures were averaged across dimensions and across TEP members to identify measures for further development.

Results:

Of 174 measures identified, 88 were determined appropriate to the setting and were reviewed and rated by the TEP. Measures with overall scores ≥75th percentile (n = 23), measures with high importance scores (n = 7), measures for under-represented domains (n = 2), and process measures antecedent to TEP identified measures (n = 2), and were selected.

Conclusions:

Specifications and data collection tools are available for 34 PEACE quality measures that were highly rated by experts in hospice and palliative care. Future research should assess the scientific soundness and responsiveness of these measures to quality improvement.

Introduction

Each year, nearly 1 million Medicare enrollees receive hospice care from over 3000 hospice providers across the United States.^1,2 Hospice demand has grown dramatically in recent years, a trend expected to continue as American baby boomers age and seek care matched to their preferences during serious illness.³ While many patients and their families experience exceptional care for life-threatening conditions, others report areas of care that could be improved. The Institute of Medicine has recognized quality improvement⁴ and health care quality measurement and reporting⁵ as national priority areas for care of patients with serious and life-threatening illnesses.

The National Consensus Project for Palliative Care Quality identified eight key domains, and corresponding practice guidelines, for high quality care at the end of life.⁶ This framework and 38 related preferred practices were endorsed by the National Quality Forum (NQF) in 2006⁷ and updated in 2009.⁸ These guidelines targeted the following domains: structures and processes of care; physical aspects of care; psychological and psychiatric aspects of care; social aspects of care; spiritual, religious, and existential aspects of care; cultural aspects of care; care of the imminently dying patient; and, ethical and legal aspects of care. Preferred practices encompass care provided by multiple disciplines, to patients and family members, across various settings over an extended period of time. These features of hospice and palliative care create challenges for providers interested in assessing the quality of care they provide.

Several reviews have identified clinical instruments for recording patient-level information relevant to important domains of palliative care practice.^9–11 These instruments are an essential component of quality, providing the clinician with a way to evaluate patient-level status and response to treatment for symptoms and conditions in physical, psychological, and other domains. Measuring the quality of care provided to a group of patients, however, requires a well-defined process for how clinical data from individuals will be used in a quality measure. A quality measure addresses a specific aspect of care or a related outcome, and is based on an accepted standard of care. Quality measures have specifications detailing the eligible population, data collection procedures, and types of analyses needed to calculate a numerator and denominator.¹²

In 2008, the Centers for Medicare & Medicaid Services (CMS) published new Conditions of Participation for hospices instructing them to implement Quality Assessment and Performance Improvement (QAPI) programs.¹³ These regulations require hospices to use a systematic, data-driven approach to measure the quality of care they deliver, to identify areas in need of improvement, and to develop strategies to enhance care. In anticipation of QAPI requirements, CMS contracted with The Carolinas Center for Medical Excellence (CCME), the Quality Improvement Organization for North and South Carolina, to develop an instrument package and procedures for hospice and palliative care programs to assess and monitor the quality of the care they provide. This project, called the PEACE project, for “Prepare, Embrace, Attend, Communicate, Empower,” was conducted over an 18-month period ending in February 2008. The PEACE project resulted in a set of 34 recommended quality measures and data collection tools to assist hospice and palliative care programs in monitoring their care. This report describes the process used to identify potential measures and the criteria for assessment and selection of the recommended PEACE quality measures.

Methods

Definitions

The PEACE project defined palliative care using the World Health Organization's definition, as care with the goal of improving the quality of life for patients with life-threatening illness.¹⁴ Palliative care encompasses patient and family care by addressing physical, cognitive, emotional, social, and spiritual concerns. Hospice care was defined as a system that delivers palliative care to patients during the terminal stages of an illness or condition.⁷ Hospice care promotes comfort and support for persons at the end of life, the majority of which is provided in the location where the patients live.³ Hospice care continues after a patient's death, providing the surviving family members with bereavement care. The objective of the PEACE project was to identify quality measures that could be used by palliative care and hospice providers.

Quality measures were defined as rates of how often a care process or outcome occurred (numerator) in a population for whom the process or outcome was expected (denominator).

Input from experts

The PEACE project team convened a 14-member Technical Expert Panel (TEP) of nationally recognized experts with extensive experience in the following areas: medical or nursing expertise in hospice and palliative care, methods and instrumentation, and quality improvement. TEP members held an initial in-person meeting, followed by a series of conference calls and email communications over a one year period. Using the NQF preferred practice guidelines as a framework, the TEP discussed potential quality measures for each of the National Consensus Project domains and recommended criteria for assessing measures. Within physical and psychological domains TEP members identified key symptoms with greatest impact in hospice and palliative care populations: pain, dyspnea, nausea, constipation, depression, anxiety, and delirium.

Identification of potential measures

We began by looking for quality measures targeting care in each of the eight NCP domains published in the scientific literature. We also included quality measures used by government entities and measures identified by national hospice organizations. The goal of the review was not to be exhaustive but rather to be focused—to identify quality measures that had been, or had the potential to be, used with patients in hospice or palliative care. We accepted quality measures if they addressed a care process or outcome for one of the NCP domains, and if they could be applied in the settings of care where hospice and palliative care are provided. To be included for further review, specifications for the numerator and denominator had to be explicit. Measures that were developed solely for use in research studies were not included.

Published quality measures identified for this review included: measures from the Assessing Care for Vulnerable Elders (ACOVE) project,^15–18 National Quality Forum (NQF) consensus measures,^5,7,19 measures from an AHRQ review of cancer measures,²⁰ palliative care measures from University Health Consortium (UHC) Palliative Care Benchmarking Project,²¹ and measures from the Quality Measures Clearinghouse.²² Measures used by governmental programs included: CMS measures used with similar populations that were not already included in the NQF endorsed measures, including the Nursing Home Improvement Feedback Tool (NHIFT) measures²³ and measures from the CMS Rural Hospice Demonstration project.²⁴ We solicited measures from the National Hospice and Palliative Care Organization (NHPCO) and the National Association of Home Care and Hospice (NAHC). Measures were grouped to the NCP domain they most closely represented. When available, information about measure properties or prior use was documented.

From this pool of potential measures, investigators selected quality measures for further investigation using three criteria: (1) appropriateness for hospice and palliative care, (2) applicability across diverse health care settings, and (3) capacity to measure a unique aspect of care delivery. Measures for which no data were available were pilot tested. The pilot data, reported elsewhere, provided evidence of the ability of hospices to collect data for the measure and the means and distributions for measures that did not have data, and allowed for construction of new potential quality measures for domains that had few or no available measures (A.P. Schenck, F. Rokoske, D.D. Durham, J. Cagle, L.C. Hanson, unpublished data).

Measure rating and selection processes

TEP members received an electronic spreadsheet containing: the quality measure description, information on the developer of the measure, measure means and distributions, the sample size from which the data were gathered, and information about reliability and the extent of missing data when available. Using criteria provided by the CCME study team (Table 1), TEP members rated each potential quality measure from 1 (low) to 5 (high) on four criteria: importance, scientific soundness, feasibility and usability. The rating criteria mirrored those used by the National Quality Forum and the CMS Measures Management System. TEP members entered their ratings directly into spreadsheets and returned them to the CCME. To identify the measures with the most favorable ratings, we created a summary measure. For each quality measure, we calculated the average TEP rating for each criterion and then tabulated an overall Average Measure Rating (AMR), weighting each the criteria equally:

Table 1.

Guidelines for Ratings Use by Technical Expert Panel ^a

Criteria	A low rating would indicate:	A medium rating would indicate:	A high rating would indicate:
Importance	You feel the measure is not important	You feel the measure is important but this assessment is not based on external evidence.	There is external evidence of importance, such as consensus standards.
Scientific soundness	The measure is not grounded in evidence or consensus Reliability and validity are questionable Risk adjustment would be absolutely necessary for the measure to be meaningful^b	The measure is based on preferred practice that has an intuitive but not proven link to outcome The measure is likely to be reasonably reliable and valid; risk adjustment may be valuable^b	A link between intervening and improvement in the target outcome, convincing evidence of reliability and validity of the measure; no need for risk adjustment^b
Usability	The actions a provider would take to improve are unknown; Measure results are difficult to interpret even with education and training; The measure is highly influenced by forces outside of the providers' control; There is no room for improvement and/or variability across providers	The actions a provider would take to improve are not entirely clear when looking at the results of the measure; Measure results cannot be understood without education and training; The measure can be influenced by forces outside of the providers' control; There is limited room for improvement and/or variability across providers	The actions a provider would take to improve are clear when looking at the results of the measure Those who use the measure will understand the results easily The measure reflects a process under the control of the provider The measure shows room for improvement (no ceiling effect) and/or variability across providers
Feasibility	Data is very difficult or impossible to collect Distortion in the data is likely to be a major problem	Data can be collected but would require effort There are minor problems with distortion in the data	Administrative data or chart review data can be collected with little effort There is no apparent distortion in the data (e.g., missing data, non-response)

TEP members were asked to rate each potential measure using a scale from 1-5 (where 1 was the lowest and 5 was the highest score). Although there were multiple dimensions under some of the criteria, TEP were asked to use their best judgment to come up with a single score for the category.

Since our project was tasked with identifying quality measures for hospices to use for internal quality improvement, priority was given to measures that were capable of being used without risk adjustment.

The 75th percentile of the average measure ratings was used as the cut-point for identifying which measures were recommended by the TEP. In addition to selecting measures with a high rating, the study team and the CMS contract and medical officers reached consensus on a small number of additional quality measures with high importance ratings (≥4.0), and quality measures designed to address underrepresented domains.

Results

Recommended measures

Our initial review identified 174 potential quality measures. Quality measures were most abundant for the domain of physical aspects of care (Fig. 1), for which 82 measures were identified. Within the physical domain, the majority of measures (70%) targeted pain. Measures for physical symptoms other than pain were notably rare. For example, we identified no measures that targeted nausea. Quality measures for psychological aspects of care (n = 19) addressed depression, but not anxiety or delirium. Quality measures for spiritual and cultural aspects of care were also rare (n = 3 and 2, respectively). Eight quality measures addressed ethical and legal aspects of care, but few had been applied in practice and tended to emphasize communication about, and documentation of, stated preferences, rather than the completion of formal advance directives.

FIG. 1.

Hospice and palliative care quality measures reviewed, by domain.

From the 174 measures originally identified, investigators excluded 86 as inappropriate for the target population, not applicable across settings, or duplicative. The remaining 88 quality measures with data from pilot testing (n = 38) and other sources (n = 50) were shared with the TEP. Many of the measures captured similar aspects of care quality, varying by the time specification or data collection method. For example, three measures assessed use of a comprehensive assessment on admission: one measure captured whether the assessment was completed in one day, another looked for assessment completion in two days and the final measure examined assessment completion within 5 days. Similarly, measures of treatment of pain had differing time frames for the satisfaction with pain control, ranging from 4 to 48 hours.

Completed ratings were received from 13 of the 14 TEP members. The 75th percentile cut-point translated into an AMR ≥3.73 (on a scale of 1 to 5 where 5 is highest). This process resulted in the identification of 23 measures with the highest TEP ratings (Table 2). The quality measure with the highest overall TEP rating, and highest importance rating was “Percent of patients on opiates for whom a bowel regimen has been established.”

Table 2.

Technical Expert Panel Ratings of Quality Measures for Measures above the 75th Percentile (≥3.73)^a

Domain	Measure	Importance	Scientific soundness	Usability	Feasibility	Overall average
Structure and process	Percent of agencies that use standard questions to assess patient depression	3.92	3.92	3.38	4.08	3.83
	Percent of patients admitted to hospice or palliative care who have a targeted screening for symptoms on admission	4.23	3.62	4.15	3.92	3.98
Physical	Percent of patients who have pain identified on initial screening that have a targeted comprehensive assessment within 24 hours of admission	4.67	4.00	4.00	3.92	4.15
	Percent of patients with pain on admission who had an order for regularly scheduled (not PRN) pain medication in 24 hours	3.77	3.46	3.85	3.85	3.73
	Percent of patients who have pain identified on initial screening that receive treatment within 24 hours of admission	4.08	3.69	4.08	4.00	3.96
	Percent of patients with moderate to severe pain who report their pain was improved within 24 hours of admission	4.62	4.38	4.38	4.15	4.38
	Percent of patients with cognitive and language problems receiving targeted pain assessment	4.31	4.00	3.15	3.77	3.81
	Percent of patients whose pain was at comfortable level within 48 hours of admission	4.69	4.00	4.31	4.08	4.27
	Percent of patients who have constipation identified on initial screening that receive treatment within 24 hours of admission	4.31	4.00	4.15	3.92	4.10
	Percent of patients with bowel regimen initiated within 24 hours of opiates among those treated with narcotics	4.62	4.31	4.46	4.15	4.38
	Percent of patients who have nausea who report their nausea was improved within 24 hours of admission	4.00	3.85	3.77	3.77	3.85
	Percent of patients with bowel function assessed at least weekly	4.38	4.25	4.13	3.75	4.13
	Percent of patients on opiates for whom bowel regimen established	4.88	4.63	4.50	3.88	4.47
Psychological	Percent of patients who screen positively for depression for whom follow-up was initiated	4.31	4.00	4.00	3.92	4.06
	Percent of patient with diagnosis of depression who receive treatment within two weeks	4.23	3.62	3.54	3.54	3.95
Social	Attending to family needs for information about medication, treatment and symptoms	4.20	3.60	3.40	3.90	3.78
Spiritual	Percent of patients who have documented discussion to identify spiritual or religious concerns as they affect care	4.15	3.38	3.46	4.00	3.75
Care of the imminently dying	Percent of patients who had moderate to severe pain in last week of life	4.15	3.69	3.85	3.46	3.79
Cultural	Availability of hospice to provide interpreter or translators for non-English or deaf patients	4.38	3.54	3.92	4.00	3.96
Ethical & legal	Percent of patients with written documentation of their preference for life-sustaining treatments prominently displayed in chart	4.38	3.85	3.85	4.08	4.04
	Percent of patients who have advanced directive or documentation of discussion	4.31	3.46	3.69	4.00	3.87
	Percent of patients who have documentation of surrogate decision maker	4.38	3.62	3.85	3.85	3.92
	Percent of patients with dementia, coma or other altered mental status that have documentation of surrogate decision maker in chart in 48 hours	4.77	3.85	4.15	4.31	4.27

On a scale of 1 (low) to 5 (high).

Eleven additional measures were added to the recommended list after reviewing the subset of measures selected by the TEP rating process. These included seven measures that received ratings of “high importance” from the TEP (average over 4.0) but missed the 75th percentile AMR and four measures identified by investigators. The four measures added by the study team included two measures targeting underrepresented domains (a psychological care measure and a social care measure) and two care process measures considered antecedent to TEP-identified quality measures. The latter two measures could be constructed from data collected for other measures identified by the TEP (screening on admission for pain and dyspnea).

The final set included 34 recommended quality measures (Table 3). More measures were recommended for the physical domain (n = 16) than any other domain, followed by the domains of ethical and legal aspects of care (n = 4), and psychological aspects of care (n = 3). The final list of recommended measures contained at least one measure for each targeted domain.

Table 3.

PEACE Recommended Quality Measures

Domain	Quality measures	Rationale for inclusion	Data collection approach
Structure and Process	Use of standard questions to assess patient depression	TEP overall rating	Organizational checklist
	Percent of patients admitted to hospice or palliative care who have a screening for symptoms during the admission visit	TEP overall rating	Patient chart review
	Policy/procedure specifying frequency with which pain & dyspnea should be assessed	TEP importance rating	Organizational checklist
	Percent of patients who had comprehensive assessment completed within 5 days of admission	TEP importance rating	Patient chart review
Physical—pain	Percent of patients screened for pain during the admission visit	Antecedent measure	Patient chart review
	For patients who screened positive for pain, the percent with a clinical assessment within 1 day of screening	TEP overall rating	Patient chart review
	For patients who screened positive for pain, the percent with any treatment within 1 day of screening	TEP overall rating	Patient chart review
	For patients who screened positive for pain, the percent who had an order for regularly scheduled (not as needed) pain medication within 1 day of screening	TEP overall rating	Patient chart review
	For patients who screened positive for pain, the percent with improvement within 1 day of screening	TEP overall rating	Patient chart review
	For patients who screened positive for pain, the percent whose pain was at comfortable level within 2 days of screening	TEP overall rating	Patient chart review
	Percent of patient with cognitive and language problems receiving pain assessment appropriate to communication needs	TEP overall rating	Patient chart review
Physical—dyspnea	Percent of patients who were screened for shortness of breath during the admission visit	Antecedent measure	Patient chart review
	For patients who screened positive for dyspnea, the percent who received treatment within 1 day of screening	TEP importance rating	Patient chart review
	For patients who screened positive for dyspnea, the percent who improved within 1 day of screening	TEP importance rating	Patient chart review
	For new ratings of shortness of breath ≥4 (on a 10 point scale), the percent with treatment (or satisfied) within 4 hours	TEP importance rating	Patient chart review
Physical—other symptoms	For patients who screened positive for nausea, the percent of patients who received treatment within 1 day of screening	TEP overall rating	Patient chart review
	Percent of patients with bowel function assessed at least weekly	TEP overall rating	Patient chart review
	For patients who screen positive for constipation, the percent who receive treatment within 1 day of screening	TEP overall rating	Patient chart review
	Percent of patients on opioids for whom a bowel regimen is established	TEP overall rating	Patient chart review
	Percent of patients on opioids who have a bowel regimen initiated within 1 day of opioid initiation	TEP overall rating	Patient chart review
Psychological symptoms	For patients who screen positive for depression, the percent who receive further assessment, counseling or medication treatment	TEP overall rating	Patient chart review
	For patients diagnosed with depression, the percent who receive interpersonal or medication treatment within two weeks of diagnosis	TEP overall rating	Patient chart review
	For patients who screened positive for anxiety, the percent who received treatment within two weeks of diagnosis	Underrepresented domain	Patient chart review
Social	Percent of families reporting the hospice attended to family needs for information about medication, treatment and symptoms	TEP overall rating	After death survey with patient families
	Percent of families who respond “Always” to question on Family/Friends Evaluation of how often family was kept informed about patient's condition	Underrepresented domain	After death survey with patient families
Spiritual	Percent of patients with chart documentation of a discussion of spiritual or religious concerns	TEP overall rating	Patient chart review
Cultural	Provision of interpreter or translators for non-English–speaking or deaf patients	TEP overall rating	Organizational checklist
Care of the imminently dying	Percent of patients who had moderate to severe pain on a standard rating scale at any time in the last week of life	TEP overall rating	Patient chart review
	Percent of families reporting they were informed of what to expect around the time of death	TEP importance rating	After death survey with patient families
Ethical and legal	Percent of patients with chart documentation of their preference for life-sustaining treatments	TEP overall rating	Patient chart review
	Percent of patients with chart documentation of an advanced directive or discussion that there is no advanced directive	TEP overall rating	Patient chart review
	Percent of patients with contact information for surrogate decision-maker in chart	TEP overall rating	Patient chart review
	Percent of patients with impaired decision-making (dementia, coma or other altered mental status) with documentation of surrogate decision maker in chart within 2 days of recognition of impaired decision-making	TEP overall rating	Patient chart review
Adverse events	Selected number of occurrences per 100 patient daysNote: 4 types of issues are tracked: falls, medication errors, DME issues (complaint, malfunction or error), and patient/family complaints)	TEP importance rating	Agency level incident reports

TEP, Technical Expert Panel; DME, durable medical equipment.

Data collection recommendations and PEACE tools

Each quality measure in the PEACE set was developed with specifications to facilitate reliable and valid data collection by different hospice and palliative care providers. The measures and specifications are publicly available on the CCME website (www.thecarolinascenter.org/research). As part of the PEACE package of instruments, a paper-based tool and data dictionary were created to allow providers to collect the data needed to calculate the 27 quality measures that use data from patient charts. An electronic data entry tool was also created to allow providers to enter patient-level data and generate graphs to monitor their progress in improving the quality measures. The 7 quality measures not calculated from patient charts come from three other sources. One measure is calculated from agency administrative and incident data—data collection and measure scoring instructions are provided in the PEACE materials. Three quality measures are based on family assessments of care. These measures can be collected during an after-death interview with family members, such as the Family Evaluation of Hospice Care.³ Three organizational level measures are included, which can be captured through an organizational readiness assessment, created by the PEACE project for providers to use to determine whether their policies and procedures would facilitate successful quality improvement. An overview of the PEACE project, the recommended measures, specifications and the readiness assessment and patient data collection tools can be downloaded from The Carolinas Center for Medical Excellence website.

Discussion

Numerous measures of quality of care are potentially relevant to palliative care and hospice patients. The PEACE Project provides a critical review and synthesis of available measures in order to create a parsimonious list of quality measures covering all National Consensus Project care domains. The PEACE measures are well specified and all have been successfully collected by hospices. These quality measures give providers important new tools to expand their capacity for quality measurement and quality improvement.

This list of recommended measures omits some aspects of quality of care previously suggested or used by others in the field. Notably absent are measures of access to hospice and palliative care, measures of undesired hospitalization or treatments, and measures of length of stay in hospice.^25,26 The omission of these and other measures does not indicate they are unimportant. On the contrary, access to hospice and palliative care is arguably the first step in improving choice, and potentially quality of care for seriously ill patients. However, measuring access requires data collection in multiple settings and was outside the scope of the PEACE project. Hospitalization during hospice care, death in an acute care setting, or receiving intensive treatments could be considered poor quality care only when clearly against the patient and family wishes. TEP members expressed concern about the ability to reconcile differences in preferences expressed by the patient and family regarding hospitalization and location of death. Hospitalization and location of death might be viewed very differently by parents of terminally ill children or by patients from minority ethnic or cultural backgrounds. Similarly, monitoring length of stay in hospice may be useful, but clinically meaningful cut-points for long or short stays have yet to be established.

The study results highlight three implications for the field of quality improvement in hospice and palliative care. First, there is a preponderance of measures targeting care for physical symptoms, often varying only in the timeframe with which the care process is to be accomplished. NQF preferred practices emphasize the importance of assessing and managing symptoms in a timely manner. However, patient acuity, and a staff capacity to respond to symptoms, differ based on the care setting. It may be reasonable to expect patients in intensive care units to have pain medication within 4 hours. However, the same response time for patients served by rural hospices may not be feasible. Care provided in different locations may, appropriately, require differing response times for physical symptoms. Quality measures may need modification to be relevant to varied health care settings.

Second, measures are frequently proposed or published without data. Without evidence on which to judge feasibility and scientific soundness, the clinical utility of these measures remains uncertain. Only 65 (37%) measures of the original 174 had any available data and none of the measures had undergone rigorous testing to establish interrater reliability or the responsiveness of the quality measure to interventions to improve care practices. The PEACE recommended measures all had some data available for review, but they, too, require further testing to establish reliability, and sensitivity to quality improvement interventions.

Third, future research is needed in the under-represented domains of quality care. Our study found several important palliative care domains have received little attention in the way of quality measures development—most notably social, cultural, and spiritual care. As these domains of care have been identified as clinically important by the NCP and NQF, additional research is needed to define best practices and effective care interventions. Only then can quality measures be designed that best capture these practices within hospice and palliative care.

It is possible our review missed quality measures being used by hospices for quality improvement. Providers who have such measures can submit them to the Quality Measures Clearinghouse so they may be more easily identified and used by others. For some domains, such as the physical domain with 82 identified measures, additional measures would have been unlikely to have influenced our results. For other domains, a wider pool of candidate measures may have resulted in different measures being recommended. Only when quality measures are shared across providers can the utility of the measures be assessed. The PEACE measures provide an opportunity for progress in that direction.

The objective of the PEACE Project was to identify quality measures for use by providers for quality improvement. As such, these measures focus primarily on processes of care rather than clinical outcomes. Future research is needed to further assure the scientific soundness, feasibility and responsiveness of these quality measures and their capacity to be used to improve care across a variety of provider settings. CMS is currently testing some of the PEACE measures in a single state. Results from this project may lead to recommendations that will further improve our ability to measure the quality of hospice and palliative care. Additional research is also needed to confirm the link between the structure and care process and improved clinical outcomes.

Footnotes

Acknowledgments

The analyses upon which this publication is based were performed under Contract Number HHSM-500-2006-NC003C (“Utilization and Quality Control Peer Review Organization for the State of North Carolina”), funded by the Centers for Medicare & Medicaid Services, an agency of the U.S. Department of Health and Human Services. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the U.S. Government. The author assumes full responsibility for the accuracy and completeness of the ideas presented. 8SOW-NC-PEACE-10-3.

The study team wishes to express their appreciation to the organizations and individuals who graciously shared their time, expertise and data with us. This project would not have been possible without their help.

Author Disclosure Statement

No competing financial interests exist.

References

Centers for Medicare & Medicaid Services: Data Compendium 2008 Edition, Table V.14. Baltimore, MD, 2008.

Hospice Association of America: Hospice Facts and Statistics 2008. Washington, D.C., 2008. www.nahc.org/haa/attachments/facts_stats2008.pdf. 2010 May 27.

National Hospice and Palliative Care Organization: NHPCO Facts and Figures: Hospice Care in America. 2009. www.nhpco.org/i4a/pages/index.cfm?pageid=5994. 2010 May 27.

Committee on Identifying Priority Areas for Quality Improvement. Priorities Areas for National Action: Transforming Health Care Quality. Washington, D.C.: National Academies Press, 2003.

National Quality Forum. National Priorities for Health Care Quality Measurement and Reporting: A Consensus Report. Washington, D.C., 2004. www.qualityforum.org/Publications/2004/12/National_Priorities_for_Healthcare_Quality_Measurement_and_Reporting.aspx. 2010 May 27.

National Consensus Project for Quality Palliative Care. Clinical Practice Guidelines for Quality Palliative Care, Second Edition, 2009. www.nationalconsensusproject.org/. 2009 January 26.

National Quality Forum. A National Framework and Preferred Practices for Hospice and Palliative Care: A Consensus Report, 2007. www.qualityforum.org/Publications/2006/12/A_National_Framework_and_PPreferre_Practices_for_Palliative_and_Hospice_Care_Quality.aspx. 2010 May 27.

National Quality Forum. Quality of Cancer Care Measure Report and Summary of the Cancer Data Methods Panel, 2009. www.qualityforum.org/. 2010 May 27.

Teno

. Time:Toolkit of Instruments to Measure End-of-Life Care Center for Gerontology and Health Care Research. 2004. www.chcr.brown.edu/pcoc/toolkit.htm. 2010 January 30.

10.

Mularski

, Dy

, Shugarman

, Wilkinson

, Lynn

, Shekelle

, Morton

, Sun

, Hughes

, Hilton

, Maglione

, Rhodes

, Rolon

, Lorenz

. A systematic review of measures of end-of-life care and its outcomes. Health Serv Res, 2007; 42:1848–1870.

11.

Hanson

, Scheunemann

, Zimmerman

, Rokoske

, Schenck

. The PEACE project review of clinical instruments for hospice and palliative care. J Palliat Med, 2010; 13,10:1253–1260.

12.

McGlynn

, Asch

. Developing a clinical performance measure. Am J Prev Med, 1998; 14,3 Suppl:14–21.

13.

Centers for Medicare & Medicaid Services. Hospice Conditions of Participation Final RuleFederal Register Vol. 73 No. 1092008. http://edocket.access.gpo.gov/2008/pdf/08-1305.pdf. 2010 March 22.

14.

WHO Definition of Palliative Care. World Health Organization (WHO), 2009. www.who.int/cancer/palliative/definition/en/. 2009 May 4.

15.

Wenger

, Shekelle

. Assessing care of vulnerable elders: ACOVE project overview. Ann Intern Med, 2001; 135,8 Pt 2:642–646.

16.

Shekelle

, MacLean

, Morton

, Wenger

. Acove quality indicators. Ann Intern Med, 2001; 135,8 Pt 2:653–667.

17.

Wenger

, Rosenfeld

. Quality indicators for end-of-life care in vulnerable elders. Ann Intern Med, 2001; 135,8 Pt 2:677–685.

18.

Chodosh

, Ferrell

, Shekelle

, Wenger

. Quality indicators for pain management in vulnerable elders. Ann Intern Med, 2001; 135,8 Pt 2:731–735.

19.

National Quality Forum. Compendium 2000–2005, 2006. www.qualityforum.org/Publications/2006/05/Compendium_2000-2005.aspx. 2010 March 22.

20.

Lorenz

, Lynn

, Dy

, Hughes

, Mularski

, Shugarman

, Wilkinson

. Cancer care quality measures: Symptoms and end-of-life care. Evid Rep Technol Assess, 2006; 137:1–77.

21.

Twaddle

, Maxwell

, Cassel

, Liao

, Coyne

, Usher

, Amin

, Cuny

. Palliative care benchmarks from academic medical centers. J Palliat Med, 2007; 10:86–98.

22.

National Quality Measures Clearinghouse. Agency for Healthcare Research and Quality, 2010. www.qualitymeasures.ahrq.gov/. 2010 March 22.

23.

Quality Partners of Rhode Island Quality Net. Nursing Home Improvement Feedback Tool (NHIFT), 2010. www.qualitynet.org/dcs/ContentServer?cid=1116947632514&pagename=Medqic%2FOtherResource%2FOtherResourcesTemplate&c=OtherResource. 2009 March 20.

24.

Centers for Medicare & Medicaid Services. CMS Hospice Rural Demonstration Domains, Indicators, and Resources for Measurement, 2010. www.cms.hhs.gov/DemoProjectsEvalRpts/downloads/RMBH_QA.pdf. 2010 March 13.

25.

Earle

, Neville

, Landrum

, Souza

, Weeks

, Block

, Grunfeld

, Ayanian

. Evaluating claims-based indicators of the intensity of end-of-life cancer care. Int J Qual Health Care, 2005; 17:505–509.

26.

Grunfeld

, Lethbridge

, Dewar

, Lawson

, Paszat

, Johnston

, Burge

, McIntyre

, Earle

. Towards using administrative databases to measure population-based indicators of quality of end-of-life care: Testing the methodology. Palliat Med, 2006; 20:769–777.