Palliative Care Scorecard

Abstract

Background:

In response to poor healthcare quality outcomes and rising costs, healthcare reform triple aim has increased requirements for providers to demonstrate value to payers, partners, and the public.

Objective:

Electronically automating measurement of the meaningful impact of palliative care (PC) programs on clinical, operational, and financial systems over time is imperative to the success of the field and the goal of development of this automated PC scorecard.

Design:

The scorecard was organized into a format of quality measures identified by the Measuring What Matters (MWM) project that are defined as important to the team, automatically extracted from the electronic health record, valid, and can be impacted over time.

Setting:

The scorecard was initially created using University of Florida Health (UF) data, a new PC program, and successfully applied and implemented at University of Colorado Anschutz Medical Campus (CU), a second institution with a mature PC program.

Measurements:

Clinical metrics are organized in the scorecard based on MWM and described in terms of the metric definition, rationale for selection, measure type (structure, process, or outcome), and whether this represents a direct or proxy measure.

Results/Conclusions:

The process of constructing the scorecard helped identify areas within both systems for potential improvement in team structure, clinical processes, and outcomes. In addition, by automating data extraction, the scorecard decreases costs associated with manual data entry and extraction, freeing clinical staff to care for patients and increasing the value of PC delivered to patients.

Introduction

Palliative care (PC) has demonstrated value by improving quality of care while reducing costs.¹ Recently, the Institute of Medicine (IOM) released a report, Dying in America, calling for action to provide all people with serious illness access to interdisciplinary PC characterized by “transparency and accountability through public reporting of aggregate quality and cost measures.”¹ In response to poor healthcare quality outcomes and rising costs, healthcare reform triple aim has increased requirements for providers to demonstrate value to payers, partners, and the public. PC programs are preparing for quality reporting requirements through the Centers for Medicare and Medicaid Services (CMS) in 2017. Currently, provider and system performance is variably defined across PC programs and research. Electronically automating measurement of the meaningful impact of PC on clinical, operational, and financial systems over time is imperative to the success of the field and the goal of this automated PC scorecard.

The business concept of a balanced scorecard (BSC) was introduced almost 25 years ago as a tool to view business and performance broadly, using both financial and value-based performance metrics.² In 2013, Hwa published the first hospital medicine BSC as part of a strategic planning initiative for an academic hospital medicine group. The BSC focused on four areas: quality, education, faculty development, and research. Those areas were further extrapolated into metrics based on the following four criteria: (1) importance, (2) measurable, (3) valid, and (4) amenable to improvement.³

Our team set out to develop a similar scorecard that reflects palliative medicine strategic mission, an endeavor that, to our knowledge, has not been published for this specialty. The University of Florida Health (UF) and The University of Colorado Anschutz Medical Campus (CU) combined efforts to create the scorecard representing inpatient consult service PC populations. UF's PC program started in 2013 and data represent programmatic inception and growth. The mature CU program was created in 2000, is Joint Commission Certified since 2013, and has a strong research and educational component. Data from CU represent more of a steady state PC system. The program directors from both institutions partnered to review, modify, and improve on the scorecard development with the support of the UF Sebastian Ferrero Office of Clinical Quality and Patient Safety and the CU Institute for Healthcare Quality, Safety, and Efficiency. Both health systems utilize EPIC electronic health record (EHR).

To measure quality and performance broadly for PC populations, we consulted numerous professional organizations, regulatory bodies, and data sources, including the American Academy of Hospice and Palliative Medicine (AAHPM), Hospice and Palliative Nurses Association (HPNA), the Center to Advance Palliative Care (CAPC), the National Quality Forum (NQF), The Joint Commission (TJC), IOM Dying in America report, Vizient (formerly University HealthSystem Consortium), evidence-based medicine, expert opinion, and local standards of care to determine what metrics would be meaningful, measurable, and amenable to improvement.^1,4–8 The goal was to focus the scorecard only on data that could be automatically extracted from the EHR to avoid effort and cost associated with manual data entry.⁹

Methods

Development

Following Hwa's hospital medicine BSC framework, our metrics met the following four criteria³:

1. Meaningful: important to the PC team, the quality department, and administration reflecting mission and goals of the PC program.

2. Measurable: through the EHR and amenable to automated extraction, without manual data entry or extraction of free text data.

3. Valid: consistent with national standards of quality measurement and reporting.

4. Actionable: amenable to improvement by frontline staff.

After developing the scorecard, AAHPM and HPNA published a list of 10 top-ranked quality measures for PC in 2015 called Measuring What Matters (MWM). MWM was developed through a rigorous scientific process narrowing 75 published quality measures through technical advisory, clinical review panel, and public rating.⁴ The program directors at both institutions originally went through a similar process on a much smaller scale, including literature review to identify published quality measures, quality analysts provided technical advisory, and frontline staff performed clinical review. The scorecard was revised with the MWM measures serving as the organizational foundation, supplemented by more traditional hospital-wide quality measures. We acknowledged that other important metrics exist, but were excluded if reported in separate databases outside the EHR or in free text format, balancing a comprehensive approach with efficiency and reproducibility of data acquisition and analysis. Directors focused on simplification, automation, and importance to the overall quality of care provided by the hospital system.

Results of the scorecard are reported in a simple one-page format that is updated quarterly to accurately reflect the activity and impact of the PC program over time. Specific definitions of the inclusion criteria were developed as reported in Table 1. The UF quality team completed a raw data extraction from the EHR, through Clarity database, which enabled further analysis of data. All personal identifiers were removed in accordance with the Health Insurance Portability and Accountability Act. The definitions and scorecard were shared with the CU quality team members who replicated the same process at their institution. Of note, the two EPIC EHR systems vary slightly based on institution-specific modifications further discussed hereunder. Benchmarks were created when possible, from a variety of sources reported hereunder.

Table 1.

Definitions

Measuring what matters	Metric	Definition
Measure 1: Comprehensive Assessment	1. Number of Encounters with an IP Consult to Palliative Care	Count of encounter numbers by month for discharges with an order for “Inpatient Consult to Palliative and Supportive Care” or “Inpatient consult to Palliative Care”
		Note: CU palliative care patients were identified by number of completed ESAS
		Includes: All inpatients regardless of age
		Measure: Number of encounters by month of discharge
	2. Number of Palliative Care Follow-up Visits	Count of the number of follow-up progress notes per month for patients with a completed palliative care consult
		Includes: All palliative care patients with a consult and a follow-up progress note
		Excludes: Consult notes
		Measure: Number of progress notes
	3. Percentage of Hospital Patients with a Palliative Care Consult	Percentage of hospital inpatients with a consult to palliative care per discharge month
		Includes: All hospital inpatients regardless of age
		Excludes: Observation patients
		Numerator: All inpatients encounters with a consult note from palliative care
		Denominator: All inpatient hospital encounters
	4. Percentage of IP Consult Orders to Palliative Care Written within four days of Admission	Percentage of encounters by admission month for patients with an order for “Inpatient Consult to Palliative and Supportive Care” or “Inpatient consult to Palliative Care” completed within four days of admission
		Numerator: Encounters with an order for a consult to palliative care entered within 96 hours of admission
		Denominator: All encounters with an order for “Inpatient Consult to Palliative and Supportive Care” or “Inpatient Consult to Palliative Care”
	5. Median Number of Days Until Palliative Care Consulted for ICU Encounters	Median days from entering ICU to completed palliative care consult.
Measures 2, 3, and 4: Screening and Tx/Mx of Physical Symptoms	6. ESAS
Measure 5: Discussion of Emotional or Psychological Needs	7. Number of Social Work Visits Completed for Palliative Care Patients	Total number of palliative care social work visits by note type palliative care progress note/social work
Measure 6: Discussion of Spiritual/Religious Concerns	8. Number of Chaplain Encounters for Palliative Care Patients	Total number of palliative care chaplain visits by note type palliative care progress note/chaplain
Measure 7: Documentation of Surrogate	9. Percentage of Palliative Care Patients with a Completed Advance Directive	Percentage of palliative care patients by discharge month with an order for “Inpatient Consult to Palliative and Supportive Care” or “Inpatient consult to Palliative Care” with an advanced directive on file.
		Includes: Any “advance directive” document located in the media tab for encounters where a palliative care consult has occurred
	10. Percentage of Hospital Patients with a Completed Advance Directive (Includes Palliative Care Patients)	Percentage of hospital inpatients by discharge month with an advanced directive on file.
		Note: CU #10.a: All hospital inpatient encounters with a documented healthcare surrogate decision maker defined as MDPOA—document present, medical proxy inpatient who lacks capacity chosen by interested parties, or healthcare court appointed guardian
		Includes: All hospital inpatient encounters where any “advanced directive” document is indicated in the media tab of the EPIC
		Excludes: Pediatrics
Measure 8: Treatment Preferences	11. Percentage of Palliative Care Patients with a DNAR Order	Percentage of palliative care patients by discharge month with an order for “Inpatient Consult to Palliative and Supportive Care” or “Inpatient Consult to Palliative Care” with a DNAR order on file
		Numerator: Palliative care encounters with DNAR orders
		Denominator: All encounters with an order for “Inpatient Consult to Palliative and Supportive Care” or “Inpatient Consult to Palliative Care”
	12. Percentage of Expired Patients with DNAR Status Following a Palliative Care Consult	Percentage of expired palliative care patients with a DNAR order during the consult encounter, where the DNAR order time is greater than the initial consult order time
		Numerator: Expired palliative care patients with a DNAR order date/time greater than the palliative care consult order date/time
		Denominator: Expired patients with a completed palliative care consult
Measure 9: Care Consistency with Documented Care Preferences	13. Percentage Palliative Care Encounters with a Referral to Hospice	Percentage of palliative care encounters by discharge month with an order for a referral to hospice
		Numerator: Palliative care encounters with an order for “IP Consult to Social Work for Hospice”
		Denominator: Encounters with a completed palliative care consult
	14. Number of Patients with Hospice as Primary Payer	Count of patients, by discharge month, with a primary insurance plan of “Haven Hospice” or “Hospice Other/Out of Area”
		Includes: All hospital inpatients with or without a palliative care consult Excludes: Observation patients
	15. Percentage of Expired Patients with a Palliative Care Consult	Percentage of patients by discharge month with a discharge disposition of expired who were seen by palliative care
		Numerator: Palliative care patients with a discharge disposition of expired Denominator: All discharged patients with a discharge disposition of expired
Measure 10: Global Measure	16. Joint Commission Advanced Certification in Palliative Care
Operational Metrics	17. Percentage Mortality for Palliative Care Patients	Percentage of palliative care patients who expired by month of discharge
		Numerator: Patients with a completed palliative care consult and discharge disposition of expired
		Denominator: All patients discharged with a completed palliative care consult
	18. Percentage of Palliative Care Patients with Readmission in 30 days	Percentage of patients with a completed palliative care consult during the index admissions that are readmitted within 30 days of discharge, reported as month of discharge for the index admission
		Numerator: Palliative care patients returning in 30 days
		Denominator: Palliative care discharges during the index discharge month
		Denominator excludes: Patients with palliative care consult that expire during index admission
	19. ICU LOS for Patients with a Palliative Care Consult (Median Days)	Median number of days spent in the ICU for discharged patients with a palliative care consult
	20. LOS for Patients with a Palliative Care Consult (Median Days)	Median number of days in the hospital for discharged patients with a palliative care consult
	21. Percentage of Palliative Care Patients with Palliative Care ICD9 Code (V66.7)	Percentage of palliative care encounters with a palliative care ICD9 Vcode (V667) at discharge on final coding summary
		Numerator: Palliative care encounters with a Vcode for palliative careDenominator: Encounters with a completed palliative care consult

Data extracted from EPIC.

DNAR, do not attempt resuscitate; ESAS, Edmonton Symptom Assessment Scale; ICU, intensive care unit; LOS, length of stay.

Permissions

Institutional Review Board applications were submitted to both UF and CU and determined to be exempt from human subject research. Copyright permission was obtained to reprint a modified version of the MWM clinical measure set.

Results

Of the metrics considered, 21 metrics were selected as most important and manageable on a one-page summary sheet. Data for fiscal year 2014 and 2015 were extracted from the EHR. Results are reported in monthly and yearly aggregate data. The scorecard was updated quarterly and modified based on input from stakeholders as seen in UF Table 2 and CU data reported in Table 3.

Table 2.

University of Florida Health Palliative Care Scorecard

		Monthly performance—FY2015												Year-to-date performance
Measuring what matters	Key measures	Jul 14	Aug 14	Sep 14	Oct 14	Nov 14	Dec 14	Jan 15	Feb 15	Mar 15	Apr 15	May 15	Jun 15	Baseline FY2014	Average FY2015 to date	Trending (baseline to YTD)	Benchmarks
Measure 1: Comprehensive Assessment	1. Number of Encounters with an IP Consult to Palliative Care	79	58	82	109	69	102	118	93	99	101	117	94	42	93	+	150
	2. Number of Palliative Care Follow-up Visits	157	174	239	369	168	230	288	222	221	184	222	176	76	221	+	450
	3. Percentage of Hospital Patients with a Palliative Care Consult	2.7%	2.0%	2.9%	3.7%	2.5%	3.5%	4.0%	3.5%	3.3%	3.5%	4.2%	3.4%	1.5%	3.3%	+	7.5%
	4. Percentage of IP Consult Orders to Palliative Care Written within four days of Admission	53.0%	59.0%	49.0%	52.0%	58.0%	54.0%	60.0%	57.0%	59.0%	45.0%	53.0%	51.0%	51.3%	54.2%	+	NA
	5. Median Number of Days Until Palliative Care Consulted for ICU Encounters	5.0	4.0	4.0	1.0	5.0	5.0	1.0	2.5	3.0	3.0	3.0	3.0	4.8	3.3	+	NA
Measures 2, 3, and 4: Screening and Tx/Mx of Physical Symptoms	6. ESAS																100%
Measure 5: Discussion of Emotional or Psychological Needs	7. Number of Social Work Visits Completed for Palliative Care Patients		34	51	72	32	36	76	20	40	73	59	60	NA	50	NA	NA
Measure 6: Discussion of Spiritual/Religious Concerns	8. Number of Chaplain Encounters for Palliative Care Patients																NA
Measure 7: Documentation of Surrogate	9. Percentage of Palliative Care Patients with a Completed Advance Directive	51%	67%	71%	61%	59%	71%	67%	68%	67%	63%	72%	61%	57%	65%	+	85%
	10. Percentage of Hospital Patients with a Completed Advance Directive (Includes Palliative Care Patients)	30%	30%	31%	32%	30%	30%	32%	31%	31%	32%	30%	31%	29%	31%	+	85%
Measure 8: Treatment Preferences	11. Percentage of Palliative Care Patients with a DNAR Order	70%	83%	70%	82%	80%	77%	80%	74%	79%	70%	78%	74%	72%	76%	+	NA
	12. Percentage of Expired Patients with DNAR Status Following a Palliative Care Consult	47%	50%	29%	45%	48%	41%	43%	35%	32%	46%	50%	48%	44%	43%	−	NA
Measure 9: Care Consistency with Documented Care Preferences	13. Percentage Palliative Care Encounters with a Referral to Hospice	53%	50%	52%	55%	43%	54%	46%	44%	63%	52%	60%	59%	55%	53%	−	NA
	14. Number of Patients with Hospice as Primary Payer	4	8	9	7	1	5	1	3	7	4	12	7	5	6	+	NA
	15. Percentage of Expired Patients with a Palliative Care Consult	26%	19%	23%	39%	29%	25%	41%	31%	29%	29%	36%	29%	12%	30%	+	X
Measure 10: Global Measure	16. Joint Commission Advanced Certification in Palliative Care																+
Operational Metrics	17. Percentage Mortality for Palliative Care Patients	24%	24%	21%	30%	36%	22%	31%	28%	19%	24%	27%	22%	25%	26%	−	X
	18. Percentage of Palliative Care Patients with Readmission in 30 days	9%	5%	16%	5%	10%	12%	13%	14%	9%	14%	5%	9%	14%	10%	+	X
	19. ICU LOS for Patients with a Palliative Care Consult (Median Days)	6.5	6.0	4.0	5.0	6.5	5.0	5.0	5.0	5.0	4.0	3.0	4.0	5.1	4.9	+	NA
	20. LOS for Patients with a Palliative Care Consult (Median Days)	7.0	8.0	8.0	8.0	8.0	7.0	7.0	7.0	7.0	8.0	7.0	9.0	8.2	7.6	+	X
	21. Percentage of Palliative Care Patients with Palliative Care ICD9 Code (V66.7)	43%	31%	38%	44%	71%	58%	60%	58%	57%	52%	65%	60%	44%	53%	+	100%

NA, not applicable.

Table 3.

University of Colorado Palliative Care Scorecard

		Monthly performance—FY2015												Year-to-date performance
Measuring what matters	Key measures	Jul 14	Aug 14	Sep 14	Oct 14	Nov 14	Dec 14	Jan 15	Feb 15	Mar 15	Apr 15	May 15	Jun 15	Baseline FY2014	Average FY2015 to Date	Trending (baseline to YTD)	Benchmarks
Measure 1: Comprehensive Assessment	1. Number of Encounters with an IP Consult to Palliative Care	57	58	56	69	53	80	94	63	89	74	72	80	52	70	+	94
	2. Number of Palliative Care Follow-up Visits	91	136	159	142	115	194	227	143	188	157	169	147	108	156	+	282
	3. Percentage of Hospital Patients with a Palliative Care Consult	2.0%	3.0%	2.0%	3.0%	2.0%	3.0%	4.0%	3.0%	3.0%	3.0%	2.0%	3.0%	2.0%	3.0%	+	7.5%
	4. Percentage of IP Consult Orders to Palliative Care Written within four days of Admission	61.4%	72.4%	73.2%	66.7%	71.7%	66.3%	67.0%	66.7%	73.0%	71.6%	63.9%	63.6%	65.0%	68.0%	+	NA
	5. Median Number of Days Until Palliative Care Consulted for ICU Encounters	4.0	4.0	24.0	4.5	0.9	6.6	1.2	8.8	35.7	2.6	5.9	7.6	12.8	8.8	+	NA
Measures 2, 3, and 4: Screening and Tx/Mx of Physical Symptoms	6. ESAS	98%	100%	100%	97%	98%	100%	99%	99%	100%	100%	100%	98%	98%	99%	+	100%
Measure 5: Discussion of Emotional or Psychological Needs	7. Number of Social Work Visits Completed for Palliative Care Patients	46	50	35	50	41	52	66	38	55	41	45	51	27	47	+	NA
Measure 6: Discussion of Spiritual/Religious Concerns	8. Number of Chaplain Encounters for Palliative Care Patients	43	45	38	39	31	36	50	34	54	41	39	44	26	41	+	NA
Measure 7: Documentation of Surrogate	9. Percentage of Palliative Care Patients with a Completed Advance Directive																85%
	10. Percentage of Hospital Patients with a Completed Advance Directive(Includes Palliative Care Patients)	39%	37%	38%	45%	45%	38%	37%	43%	38%	43%	39%	41%	28%	40%	+	85%
	10a. Percentage of Hospital Patients with Documentation of Surrogate Healthcare Decision Maker							23%	21%	22%	24%	24%	26%		23%	+	85%
Measure 8: Treatment Preferences	11. Percentage of Palliative Care Patients with a DNAR Order	74%	71%	68%	72%	77%	71%	72%	73%	72%	70%	69%	73%	73%	72%	−	NA
	12. Percentage of Expired Patients with DNAR Status Following a Palliative Care Consult	10%	17%	30%	25%	31%	26%	32%	11%	18%	30%	13%	67%	33%	26%	−	NA
Measure 9: Care Consistency with Documented Care Preferences	13. Percentage Palliative Care Encounters with a Referral to Hospice	54%	43%	52%	54%	43%	43%	50%	43%	35%	41%	56%	32%	45%	45%	=	NA
	14. Number of Patients with Hospice as Primary Payer																NA
	15. Percentage of Expired Patients with a Palliative Care Consult	9%	6%	13%	5%	6%	7%	11%	15%	7%	8%	4%	0%	7%	8%	+	X
Measure 10: Global Measure	16. Joint Commission Advanced Certification in Palliative Care	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+	+
Operational Metrics	17. Percentage Mortality for Palliative Care Patients	18%	10%	18%	12%	25%	24%	20%	14%	12%	14%	11%	14%	16%	16%	=	X
	18. Percentage of Palliative Care Patients with Readmission in 30 days	11%	17%	13%	14%	8%	15%	11%	16%	18%	16%	15%	14%	11%	14%	−	X
	19. ICU LOS for Patients with a Palliative Care Consult(Median Days)	6.0	4.0	3.9	6.5	2.1	5.2	4.7	5.3	5.2	5.2	4.6	16.5	5.1	5.8	−	NA
	20. LOS for Patients with a Palliative Care Consult(Median Days)	8.3	8.8	7.7	8.0	7.7	8.5	9.0	7.6	7.8	7.1	10.3	26.3	9.1	9.8	−	X
	21. Percentage of Palliative Care Patients with Palliative Care ICD9 Code (V66.7)	79%	91%	95%	87%	94%	88%	85%	92%	94%	92%	90%	100%	83%	91%	+	100%

Clinical metrics

Clinical metrics are organized in the scorecard and summarized hereunder based on MWM. The metrics are described as follows: measure type (structure, process, or outcome), direct or proxy measure, definition, and rationale for selection.⁴

MWM Measure 1: Comprehensive assessment completed .

Measure type: process; direct

Metric 1: Number of encounters with an inpatient consult to PC

Metric 2: Number of PC follow-up visits

Metric 3: Percentage of hospital patients with a PC consult

Metric 4: Percentage of inpatient consult orders to PC written within four days of admission

Metric 5: Median number of days until PC consulted for intensive care unit (ICU) encounters

Rationale: Both UF and CU PC programs created similar standardized consult and follow-up note templates. The templates outline a comprehensive assessment based on the NQF: National Framework and Preferred Practices for Palliative and Hospice Care, and the National Consensus Project: Clinical Practice Guidelines for Quality Palliative Care.^6,10 Thus metrics 1 and 2, consult and follow-up volume, represent the number of patients who received a comprehensive assessment. By monitoring volume over time, a program can determine if it is operationally meeting predicted demand as well as plan for increased staffing needs. Benchmarks for recommended consult volume were calculated from CAPC: Calculating Staffing Needs predicted based on hospital size.^10–14 Metric 3 reports PC consult volume as a percentage of total annual admissions. The goal benchmark of metric 3 is 5–10% of admissions being seen by the PC team, with most programs ranging in the 2–4% percentile. When program growth reaches 10%, expert opinion recommends focusing efforts not on expansion of the inpatient program but rather on improving generalist skills in primary PC as well as providing outpatient PC services.^12–15 Timing of PC consultation is represented in metrics 4 and 5.^4,16 The goal is to have PC consultation earlier in the hospital course for improved benefit to the patient, families, and the system without a specific benchmark.^17–19

MWM Measures 2, 3, and 4: Screening for physical symptoms, pain treatment, and dyspnea screening and management .

Measure type: process; proxy

Metric 6: Edmonton Symptom Assessment Scale (ESAS)

Rationale: The ESAS is a validated symptom assessment tool in the PC population.^20–22 The UF program excluded symptom screening on the scorecard because symptom assessment was reported in text format and could not be extracted without manual chart review. Building the ESAS into the UF EHR has been identified as a strategic priority going forward.^18,19,23 At CU, individual ESAS symptom scores are tracked over time through EHR extraction, but not included in the scorecard for simplicity other than to note that 99% of patients had a completed ESAS. If scorecard space allowed, inclusion of outcome data on individual symptom improvement over time would be helpful for quality improvement and reporting of performance.

MWM Measures 5 and 6: Discussion of emotional/psychological needs and spiritual/religious concerns .

Measure type: structure/process; direct

Metric 7: Number of social work visits completed for PC patients

Metric 8: Number of chaplain encounters for PC patients

Rationale: Psychosocial and spiritual assessment would ideally be completed by dedicated PC social work and chaplain, in addition to the assessment performed by PC physicians and advanced practice providers. At UF, a certified PC social worker started in September of 2015 and visits were captured over time. To date, a dedicated team chaplain does not exist for the UF team, and is part of the long-term strategic goals reflected in the scorecard metric 8 as a darkened box. At CU, a certified PC social worker and a dedicated PC chaplain became part of the interdisciplinary team in 2013.^6,10,18,19

MWM Measure 7: Documentation of a surrogate .

Measure type: outcome; proxy

Metric 9: Percentage of PC patients with a completed advance directive (AD)

Metric 10: Percentage of all hospital patients with a completed AD

Rationale: At UF, AD is defined as any of the following: healthcare surrogate or medical durable power of attorney, medical living will, physician order for life sustaining treatment, and/or do not attempt resuscitate (DNAR) forms. UF created an EHR “checkbox” that would categorize for easy retrieval, AD documents, allowing extraction of numeric data as reflected in metrics 9 and 10. CU developed a “Healthcare Directive Page” that provides an area specifically dedicated to the documentation and retrieval of ADs. As CU already tracks data on documentation of the surrogate decision maker, this direct outcome metric was added in as 10a in Table 3. PC team AD completion rate was excluded for CU as these data are collected only for outpatients. In 1998, Gunderson Health System published an AD completion rate of 85% obtained through a community-wide advance care planning program, which is held as the benchmark.^24,29

MWM Measure 8: Documentation of treatment preferences.

Measure type: outcome; direct

Metric 11: Percentage of PC patients with a DNAR order

Metric 12: Percentage of expired patients with DNAR status after a PC consult

TableRationale: Each DNAR order ideally represents a discussion with the patient/surrogate about life-sustaining treatment preferences, with preferences summarized in the form of an order. EHR functionality allows the ordering of DNAR status, defaulting back to “full code” each admission, and saving historical code information. The palliative population represents a spectrum of patients with serious illness that may be curable or chronic, thus no benchmark exists as noted by the “not applicable.”

MWM Measure 9: Care consistency with documented care preferences .

Measure type: outcome; proxy

Metric 13: Percentage of PC encounters with a referral to hospice

Metric 14: Number of patients with hospice as primary payer

Metric 15: Percentage of expired patients with a PC consult

Rationale: MWM measure 9 promotes that care preferences not only be documented but also honored by care received concordant with patient wishes. Metrics 13 and 14 are extracted from the EHR by “hospice referral” order or primary payer source of hospice. Evidence exists that with increased PC consultation, hospice referrals increase as well. UF provides general inpatient care for hospice patients under a contract with local hospice agencies. CU does not contract for general inpatient hospice care, but refers to community hospice care centers.^17,26 The benchmark for number of expired patients with a PC consult comes from reported Vizient data for 2015, and is not included in the published scorecard because of publishing restrictions associated with proprietary reports.^19,26,27

MWM Measure 10: Global measure .

Measure type: structure/process; proxy

Metric 16: TJC Advanced Certification in PC

Rationale: No specific global measure of PC currently exists. TJC Advanced Certification in PC represents achievement of the highest national standards of quality PC. TJC certification includes, but is not limited to, 24-hour availability of quality PC delivered by a specialized interdisciplinary team by defined national standards, patient and family satisfaction surveys, continuous quality improvement, and measurement of impact over time.⁷ The certification is a continuous process with monthly reported quality measures and ongoing recertification requirements. UF is not certified by TJC, and this goal is represented in scorecard metric 16 with a shaded box. CU received TJC certification in 2013.

Operational:

Measure type: outcome; direct

Metric 17: Percentage mortality for PC patients

Metric 18: Percentage of PC patients with readmission in 30 days

Metric 19: ICU length of stay (LOS) for patients with a PC consult

Metric 20: LOS for patients with a PC consult

Metric 21: Percentage of PC patients with PC ICD-9 code (V66.7)

Rationale: The financial impact of PC consults has been important to hospital systems. Nationally, PC consultation has shown to both decrease 30-day readmissions and LOS, decreasing costs and increasing potential bed capacity for hospitals.^16,25,28 Data were extracted for palliative populations through Vizient with the palliative care ICD-9 code V66.7 (now Z51.5 for ICD-10). Use of V66.7 may favorably improve the mortality index, but may not represent only PC consult patients. A program goal is for PC patients to be better defined and tracked through the Vizient system with a benchmark of 100% of PC patients having a V66.7 code. Benchmarks were defined, in part, as metrics reported through Vizient, which compares similar academic hospitals.^8,12 Benchmarks are used internally and not published because of proprietary restrictions.

Discussion

For PC programs, efforts abound to identify nationally agreed-upon metrics, establish benchmarks, and develop national registries for tracking data across institutions. The hope for developing the scorecard is that palliative teams will be better able to care for patients and families by tracking quality metrics. The scorecard provides the opportunity to track team progress and impact over time and strive for the highest quality care. Hospices were recently required to report quality data starting in 2014 in a “pay for reporting” effort that will transition to “pay for performance” and public reporting analogous to the IOM call for “transparency and accountability through public reporting of aggregate quality and cost measures.”¹ The expectation is that PC will soon follow and be required to gather quality data metrics, report outcomes both privately and publicly, and improve performance.

To our knowledge, no automated PC scorecard exists in the literature. The scorecard was initially created using UF data gleaned from a relatively new PC program, and successfully applied and implemented at CU, a second institution with a mature PC program. Electronic creation of the initial scorecard through data extraction, analysis, and verification was a time-intensive process completed by each respective quality department data analyst team, a resource that may not be available at other institutions. The scorecard is automated now with quarterly updates. Limitations of the scorecard include exclusion of potentially important metrics that are not readily extractable from the EHR, including patient/family satisfaction and quality-of-life survey outcome data and calculated cost avoidance. In some instances, scorecard data varied slightly from manual entry comparison data. In addition, both institutions utilize the EPIC EHR, and replication with a different EHR may be more challenging. The hope is this scorecard can provide a starting point for other programs interested in developing their own institution-specific automated data reporting system, regardless of EHR type, based on the MWM measures. By supporting quality assessment and measurement over time, the MWM quality indicators hope to improve patient care and have been recently incorporated into TJC mandatory performance measures for Advanced Certification in Palliative Care. However, agreed-upon definitions of the numerator denominators have yet to be developed, and some measures, for example, care consistency with documented care preferences and global measure of quality, do not have direct measurements that are readily extractable from the EHR. This limits the scorecard by use of proxy measures when direct measurement is not feasable.⁴

Publishing these results may inform further development and discussion of both meaningful and quantifiable metrics with automated EHR extraction—driving the field of palliative medicine forward using technology. Future work could include incorporating the MWM measures into existing national PC registries for development of national benchmarks and a unified national database. The process of constructing the scorecard helped identify areas within both systems for potential improvement in team structure, clinical processes, and outcomes. In addition, by automating data extraction, the scorecard decreases costs associated with manual data entry and extraction, freeing clinical staff to care for patients and increasing the value of PC delivered to patients. Given the significant overlap of the MWM and scorecard metrics with the Affordable Care Act mandated CMS hospice quality reporting program metrics and the recently released JC Advanced Certification in Palliative Care mandatory performance measures, it may be possible for scorecard metrics to be tied to healthcare reporting and reimbursement.

Footnotes

Acknowledgment

Salary support for University of Florida Palliative Care Data Manager Keri Merchant through Gatorade funding.

Author Disclosure Statement

No competing financial interests exist.

The views expressed in the submitted article are our own and not an official position of the institution or funder.

References

Committee on Approaching Death: Addressing Key End of Life Issues; Institute of Medicine:. Dying in America: Improving Quality and Honoring Individual Preferences Near the End of Life. Washington, DC: National Academies Press, 2015.

Kaplan

, Norton

: Using the balanced scorecard as a strategic management system. Harv Bus Rev, 1992; 74:71–79.

Hwa

, Sharp

, Wachter

: Development and implementation of a balanced scorecard in an academic hospitalist group. J Hosp Med, 2013; 8:148–153.

Morss Dy

, Kiley

, Lupu

, et al.: Measuring What Matters: Top-ranked quality indicators for hospice and palliative care from the American Academy of Hospice and Palliative Medicine and Hospice and Palliative Nurses Association. J Pain Symptom Manage, 2015; 49:773–781.

Weissman

, Meier

, Spragens

: Center to Advance Palliative Care palliative care consultation service metrics: Consensus recommendations. J Palliat Med, 2008; 11:1294–1298.

National Quality Forum: A National Framework and Preferred Practices for Palliative and Hospice Care Quality. Consensus Report. Washington, DC: National Quality Forum, 2006.

The Joint Commission: Certification for Palliative Care Programs. 2016. www.jointcommission.org (last accessed December 29, 2016).

Vizient: Palliative Care Report: Vizient; 2015. www.vizientine.com (last accessed December 29, 2016).

Centers for Medicare and Medicaid Services: www.cms.gov. 2016 (last accessed March 1, 2016).

10.

National Consensus Project: Clinical Practice Guidelines for Quality Palliative Care, 3rd Edition. Pittsburg, PA: National Consensus Project for Quality Palliative Care, 2013.

11.

Weissman

: Calculating Staffing Needs: Center to Advanced Palliative Care; 2008. www.capc.org (last accessed December 29, 2016).

12.

Morrison

, Meier

: Clinical practice. Palliative care. NEJM, 2004:2582–2590.

13.

Quill

, Norton

, Shah

, et al.: What is most important for you to achieve: An analysis of patient responses when receiving palliative consultation. J Palliat Med, 2006; 9:382–388.

14.

von Gunten

: Secondary and tertiary palliative care in US hospitals. JAMA, 2002; 7:875–881.

15.

Realizing the Full Benefit of Palliative Care: Service Optomization and Strategic Growth: The Advisory Board Company. The Advisory Board Company, Washington, DC, 2013.

16.

Morrison

, Penrod

, Cassel

, et al.: Cost savings associated with US hospital palliative care consultation programs. Arch Intern Med, 2008; 168:1783–1790.

17.

Morrison

, Dietrich

, Ladwig

, et al.: Palliative care consultation teams cut hospital costs for Medicaid beneficiaries. Health Aff, 2011; 30:454–463.

18.

Schenck

, Rokoske

, Durham

, et al.: Quality measures for hospice and palliative care: Piloting the PEACE measures. J Palliat Med, 2014; 17:769–775.

19.

Hanson

, Rowe

, Wessell

, et al.: Measuring palliative care quality for seriously ill hospitalized patients. J Palliat Med, 2012; 15:798–804.

20.

Bruera

, Kuehn

, Miller

, et al.: The Edmonton Symptom Assessment System (ESAS): A simple method for the assessment of palliative care patients. J Palliat Care, 1991; 7:6–9.

21.

Philip

, Smith

, Craft

, et al.: Concurrent validity of the modified Edmonton Symptom Assessment System with the Rotterdam Symptom Checklist and the Brief Pain Inventory. Support Care Cancer, 1998; 6:539–541.

22.

Nekolaichuk

, Watanabe

, Beaumont

: The Edmonton Symptom Assessment System: A 15-year retrospective review of validation studies (1991–2006). Palliat Med, 2008; 41:111–122.

23.

Wenger

, Casarett

: American Medical Association-Physician Consortium for Performance Improvement Palliative and End of Life. 2014. www.thepcpi.org (last accessed December 29, 2016).

24.

Rao

, Anderson

, Lin

, et al.: Completion of advance directives among U.S. consumers. Am J Prev Med, 2014; 46:65–70.

25.

Adelson

, Paris

, Smith

, et al.: Standardized criteria for required palliative care consultation on the solid tumor oncology service. J Clin Oncol, 2013; suppl 31:abstr 37.

26.

Ahluwalia

, Chuang

, Antonio

, et al.: Documentation and discussion of preferences for care among patients with advanced cancer. J Oncol Pract, 2011; 6:361–366.

27.

Lorenz

, Rosenfeld

, Wegner

: Quality indicators for palliative and end-of-life care in vulnerable elders. J Am Geriatr Soc, 2007; 55:318–326.

28.

Norton

, Hogan

, Holloway

, et al.: Proactive palliative care in the medical intensive care unit: Effects on length of stay for selected high-risk patients. Critical Care Med, 2007; 35:1530–1535.

29.

Hammes

, Rooney

: Death and end-of-life planning in one midwestern community. Arch Intern Med, 1998; 158:383–390.