- Email: [email protected]
Impact of a care delivery redesign initiative for vascular surgery
From the Society for Vascular Surgery
Impact of a care delivery redesign initiative for vascular surgery Joyce Peralta, DNP, ACNP-BC,a Sung Wan Ham, MD,a Gregory A. Magee, MD, MSc,a Christianne Lane, PhD,b Cali Johnson, MD,a Alice Issai, MBA,c Lindsey Lawrence, MHA,c Chris Allen, BS,c Todd Wilson, MBA,a and Fred A. Weaver, MD, MMM,a Los Angeles, Calif
ABSTRACT Objective: A hospital-wide quality improvement process through a care delivery redesign (CDR) was initiated to improve patient care efficiency, clinical documentation, and length of stay (LOS). The impact of CDR was assessed through LOS, unplanned readmission rates, and hospital financial metrics. Methods: The CDR team consisted of the Chief of Vascular Surgery, inpatient nurse practitioner, dedicated case manager, clinical documentation improvement specialist, and vascular surgery residents and faculty. The nurse practitioner facilitated patient care coordination, resident system-based education, and multidisciplinary collaboration. Tools created to track performance and to ensure sustainability included daily discussions of patient care barriers and solutions; standardized order sets; a mobile app for residents containing resident service expectations, disease-specific resources, and vascular surgery journal links; and a weekly inpatient tracker showing real-time patient care data. Outcome measures included LOS, case mix index, contribution margin, and unplanned readmissions. Each outcome was determined for all inpatient admissions the year before and the 12 months after CDR was initiated. Outcomes were compared between the two groups. Results: Implementation of CDR resulted in a 23% decrease in LOS (P ¼ .003), reducing the gap to the Centers for Medicare and Medicaid Services geometric mean LOS from 2.1 days to 0.5 day (P < .001). Clinical documentation resulted in an increase in case mix index of 10% (P ¼ .011). The 30-day unplanned readmission rates did not change in the 12 months after CDR was initiated compared with the year before (P ¼ .92). Financial data demonstrated decreased variable cost and increased revenue resulting in a $1.89 million increase in contribution margin. Conclusions: A CDR predicated on a dedicated service line advanced practitioner, clinical documentation education, weekly service tracker review, and real-time management of system-related barriers to patient care is described. Implementation of the CDR reduced hospital LOS with no change in unplanned readmissions and provided significant financial benefit to the hospital by increasing revenue and decreasing variable cost. (J Vasc Surg 2019;-:1-10.) Keywords: Length of stay; Diagnosis-related group; Geometric mean length of stay; Case mix index; Variable cost; Contribution margin
The demands of today’s health care environment have placed tremendous pressure on hospitals nationwide to decrease length of stay (LOS) and to deliver patientcentered care while minimizing costs. In 2016, the Centers for Medicare and Medicaid Services (CMS) reported a 4.3% growth in health care expenditures to $3.3 trillion
From the Division of Vascular Surgery and Endovascular Therapy, Department of Surgery,a and Department of Preventive Medicine,b Keck School of Medicine, and Keck Medical Center,c University of Southern California. Author conflict of interest: none. Presented at the 2018 Vascular Annual Meeting of the Society for Vascular Surgery, Boston, Mass, June 20-23, 2018. Additional material for this article may be found online at www.jvascsurg.org. Correspondence: Fred A. Weaver, MD, MMM, Division of Vascular Surgery and Endovascular Therapy, Department of Surgery, Keck School of Medicine, University of Southern California, 1520 San Pablo St, Ste 4300, Los Angeles, CA 90033 (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2019 by the Society for Vascular Surgery. Published by Elsevier Inc. https://doi.org/10.1016/j.jvs.2019.03.053
or $10,348 per person. This accounted for 17.9% of the gross domestic product in the United States. The projected growth rate for health care expenditures will far outpace the rest of the economy at 5.5% per year from 2017 to 2026, reaching $5.7 trillion by 2026.1 The existing and predicted escalation of health care costs has placed increased demand on hospitals and health care organizations to decrease inpatient LOS, to reduce unplanned readmission rates, and to improve efficiency. Physician leadership is needed to accomplish these challenging goals. Current inpatient hospital reimbursement is predicated on a diagnosis-related group (DRG) classification system that is based on primary and secondary diagnoses, procedures performed, comorbidities, and complications.2 The DRG system incentivizes proper documentation to ensure that the correct DRG is applied as payment is a lump sum that is supposed to cover all hospital expenses from admission to discharge. CMS reimburses hospitals on the basis of the geometric mean LOS (GMLOS) for each individual Medicare Severity Diagnosis Related Group (MS-DRG) code. The geometric mean, which is 1
2
Journal of Vascular Surgery
Peralta et al
---
calculated by the nth root of the product of n numbers, more accurately represents the central value than a simple arithmetic mean because it is less sensitive to outliers. p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi n Geometric mean ¼ x1 x2 x3 . xn When inpatient LOS exceeds the CMS GMLOS, patient care costs can exceed reimbursement. The MS-DRG is a tiered system with increases in GMLOS if there are specific associated comorbidities and complications (CC) or major comorbidities and complications (MCC). As a quaternary medical center, Keck Medical Center of the University of Southern California (KMC) operates at a high cost structure because of patient acuity and complexity resulting in increased LOS. As a result, costs per patient discharge are often greater than reimbursement provided by the assigned MS-DRG and associated GMLOS. To address these challenges and to improve patient care efficiency, KMC started an initiative called care delivery redesign (CDR), which engaged nursing case management, a clinical documentation improvement (CDI) specialist, and the value information office (VIO). The vascular surgery service was one of five service lines (along with urology, heart failure, stroke, and spine) included in the CDR. The CDI specialist and VIO were pre-existing KMC staff, so no additional costs were incurred. The primary objectives of the CDR were to lower costs and to optimize patient care by improving care efficiency, facilitating vascular surgery service MS-DRG education, and leveraging real-time data analytics to address systematic barriers to patient care. The results of the CDR initiative and its impact on LOS and associated financial metrics were evaluated after 1 year of implementation.
METHODS CDR program structure. The vascular surgery service CDR involved the Chief of Vascular Surgery, the inpatient vascular surgery nurse practitioner (NP), a dedicated vascular surgery case manager, and vascular surgery residents and faculty. A VIO performance improvement analyst provided a weekly snapshot of vascular surgery service performance by tracking patient disposition with actual LOS compared with GMLOS for the assigned MS-DRG. The MS-DRG assignment, CC/MCC, and GMLOS were provided by the clinical documentation office. To evaluate all inpatients irrespective of insurer, nonMedicare beneficiary DRG assignments were converted to MS-DRG and associated GMLOS using 3M Coding and Reimbursement (3M, Maplewood, Minn) software. Similar CDR program structures were instituted contemporaneously for urology, heart failure, stroke, and spine service lines. Team member organization and education. The NP educated all fellows, residents, and interns on CMS terminology (GMLOS, MS-DRG) and how clinical
2019
ARTICLE HIGHLIGHTS d
d
d
Type of Research: Single-center retrospective cohort study Key Findings: Implementation of the care delivery redesign for vascular surgery patients in a hospital setting resulted in a decrease in length of stay by 23% with no change in 30-day unplanned readmission rates and a concomitant decrease in variable cost/patient, resulting in an increase of $1.89 million in contribution margin for the hospital. Take Home Message: A redesign of vascular surgery care can lead to more efficient patient care, which is associated with a significant financial benefit.
documentation influences MS-DRG tier and assigned GMLOS. Understanding the different vascular MS-DRGs, respective CC/MCC, and their corresponding GMLOS was the focus. In addition, all trainees received CDI education from the CDI specialist to further reinforce how accurate documentation affects MS-DRG level and patient severity of illness. Each patient’s assigned admission MS-DRG and corresponding GMLOS were displayed in a banner in the electronic medical record and updated daily to reflect the most accurate working MS-DRG and GMLOS. The NP also created standardized, procedurebased clinical pathways and order sets to standardize and to structure the patient care process. These order sets provided, among other things, for early ambulation and physical therapy evaluations for patients undergoing lower extremity bypass procedures; prompt restarting of home blood pressure medications; default physical therapy and occupational therapy orders, depending on procedure performed; and early social work consultations to recognize and to address discharge barriers, such as limited family resources, transportation issues, and insurance limitations. All faculty and trainees were required to use the standardized order sets at the onset of the CDR process. All this information was available on an institutionally designed mobile app, which has been previously described.3 The mobile app also included resident service expectations, disease-specific resources, and vascular surgery journal links. Directed by the Chief of Vascular Surgery, the NP established a daily patient care workflow involving closed-loop communication with the service line case manager. Communication included anticipated LOS, potential discharge needs, expected patient disposition, and potential discharge date. The NP created a care coordination checklist that was reviewed during every morning and afternoon rounds with the vascular surgery team. Checklist items included review of LOS compared with GMLOS, discussions on barriers to care, revisions of daily progress notes necessary to capture appropriate CC/MCC, and plans for discharge and
Journal of Vascular Surgery -,
In-house
Volume
Number
Peralta et al
Aending Physician
Admit Date
LOS
Working CMS GMLOS
Working LOS O/E
MS-DRG
ROWE MD, VINCENT L
4/10/2018
7.8
1.8
4.32
269
WEAVER MD, FRED A
4/9/2018
9.4
6.5
1.45
270
WEAVER MD, FRED A
4/16/2018
2.5
2.1
1.17
35
HAN MD, SUKGU M
4/12/2018
6.3
6.2
1.01
220
WEAVER MD, FRED A
4/16/2018
2.1
2.3
0.93
254
WEAVER MD, FRED A
4/16/2018
2.3
3.6
0.62
863
ROWE MD, VINCENT L
4/17/2018
1.1
4
0.28
299
AORTIC AND HEART ASSIST PROCEDURES EXCEPT PULSATION BALLOON W/O MCC OTHER MAJOR CARDIOVASCULAR PROCEDURES W MCC CAROTID ARTERY STENT PROCEDURE W CC CARDIAC VALVE & OTH MAJ CARDIOTHORACIC PROC W/O CARD CATH W CC OTHER VASCULAR PROCEDURES W/O CC/MCC POSTOPERATIVE & POSTTRAUMATIC INFECTIONS W/O MCC PERIPHERAL VASCULAR DISORDERS W MCC
Comments
USC-5N
USC-CC
Max GMLOS for DRG group
USC-5N USC-7W USC-5N USC-5N
Pending procedure
USC-5N
Max GMLOS for DRG group
MS-DRG
Aending Physician
Admit Date
Discharge Date
LOS
Working CMS GMLOS
Working LOS O/E
MS-DRG
ROWE MD, VINCENT L
4/2/2018
4/12/2018
10.4
4.2
2.49
253
MAGEE MD, GREGORY
4/2/2018
4/11/2018
9.0
4.8
1.88
314
MAGEE MD, GREGORY
4/1/2018
4/17/2018
16.0
9.7
1.65
957
ROWE MD, VINCENT L
4/7/2018
4/18/2018
10.4
6.5
1.59
270
HAN MD, SUKGU M
4/9/2018
4/11/2018
2.5
1.8
1.39
269
ZIEGLER MD, KENNETH
4/10/2018
4/13/2018
3.2
2.4
1.33
301
HAM MD, SUNG W
4/9/2018
4/12/2018
3.2
4
0.81
299
ROWE MD, VINCENT L
4/9/2018
4/12/2018
3.2
4.2
0.76
253
ZIEGLER MD, KENNETH
4/13/2018
4/15/2018
2.4
3.3
0.73
300
HAN MD, SUKGU M
4/8/2018
4/12/2018
4.2
6.2
0.68
220
WEAVER MD, FRED A
4/16/2018
4/17/2018
1.2
2.5
0.49
983
HAM MD, SUNG W
4/11/2018
4/13/2018
2.1
4.4
0.48
271
AVG D/C LOS: AVG GMLOS: Weekly Working O/E:
Current Unit
MS-DRG Descripon
GMLOS
Actual LOS
Discharge d Pat ie nt s
3
-
MS-DRG Descripon
Discharge Unit
OTHER VASCULAR PROCEDURES W USC-5N CC OTHER CIRCULATORY SYSTEM USC-5N DIAGNOSES W MCC OTHER O.R. PROCEDURES FOR MULTIPLE SIGNIFICANT TRAUMA W USC-5N MCC OTHER MAJOR CARDIOVASCULAR USC-7W PROCEDURES W MCC AORTIC AND HEART ASSIST PROCEDURES EXCEPT PULSATION USC-5N BALLOON W/O MCC PERIPHERAL VASCULAR DISORDERS USC-5N W/O CC/MCC PERIPHERAL VASCULAR DISORDERS USC-5N W MCC OTHER VASCULAR PROCEDURES W USC-5N CC PERIPHERAL VASCULAR DISORDERS USC-5N W CC CARDIAC VALVE & OTH MAJ CARDIOTHORACIC PROC W/O CARD USC-5N CATH W CC EXTENSIVE O.R. PROCEDURE UNRELATED TO PRINCIPAL USC-5N DIAGNOSIS W/O CC/MCC OTHER MAJOR CARDIOVASCULAR USC-7W PROCEDURES W CC
Comments Final coded Max GMLOS for DRG group Max GMLOS for DRG group Max GMLOS for DRG group
Final coded Max GMLOS for DRG group
CDI to put in Na level severity query
5.66 4.50 1.26
Fig 1. Vascular surgery weekly patient and discharge tracker. CC, Comorbidities and complications; CDI, clinical documentation improvement; CMS, Centers for Medicare and Medicaid Services; DRG, diagnosis-related group; GMLOS, geometric mean length of stay; LOS, length of stay; MCC, major comorbidities and complications; MSDRG, Medicare Severity Diagnosis Related Group; O/E, observed/expected; O.R., operating room.
patient disposition including alternatives. The NP also independently communicated with the CDI specialist to provide updates on changes in a patient’s medical condition such that the CC/MCC reflected the most accurate working MS-DRG. All CC/MCC that occurred
during a patient admission were documented in the discharge summary. In addition, the NP and dedicated CDI specialist tracked each patient after discharge to ensure assignment of the accurate final MS-DRG code.
4
Peralta et al
Journal of Vascular Surgery ---
2019
Fig 2. Graphic of mean length of stay (LOS) for all procedures vs geometric mean length of stay (GMLOS) for all procedures (y-axis, mean days; x-axis, time). CMS, Centers for Medicare and Medicaid Services.
Weekly inpatient tracker. At the weekly vascular surgery conference, the faculty, fellows, residents, and staff reviewed all current inpatients and recently discharged patients using the weekly inpatient tracker provided by the VIO specialist. The tracker table displayed patient identifiers, admission/discharge unit and date, vascular surgery attending physician, current LOS, GMLOS, current MS-DRG, and associated CC/MCC (Fig 1). The tracker MS-DRG column was color coded to indicate whether a patient’s LOS was under (green), within (yellow), or over (red) the GMLOS benchmark. A second graphic analysis displayed the trend of our actual LOS vs GMLOS over time (Fig 2). Weekly review of the tracker enabled evaluation of current management and barriers to efficient care, providing an opportunity for modification of clinical decision-making and patient disposition. In addition, the tracker provided real-time feedback on coding and associated GMLOS based on current electronic medical record documentation. CDR analysis. To evaluate the impact of CDR, inpatient admissions to the vascular surgery service at KMC during the initial 12 months of CDR (July 2016-June 2017 [CDR17]) were compared with all inpatient vascular surgery admissions in the immediately preceding fiscal year (pre-CDR 2016 [PCDR16]). Data collected for both groups included patient demographics, procedures performed, LOS, MS-DRG, GMLOS, insurance type, and unplanned readmissions. Unplanned readmissions were defined as any unanticipated admission to the KMC vascular surgery service within 30 days of the patient discharge date. Case mix index (CMI) was calculated by dividing the total MS-DRG weight of all the vascular surgery inpatient
discharges for the year by the number of hospital discharges.4 Data were also collected for hospital revenue from all inpatient reimbursements received from Medicare, Medicaid, and other third-party payers as well as for variable cost, which is defined as the cost of the inpatient hospitalization exclusive of overhead and fixed costs. Contribution margin (CM) was determined by subtracting variable cost from revenue. Financial data and metrics were provided, reviewed, and confirmed by the Chief Financial Officer of KMC. Outcomes measured. Primary outcome measures included LOS compared with GMLOS, CMI, CM, and unplanned readmission rate. To determine changes in outcome measures after institution of the CDR, the PCDR16 group was compared with the CDR17 patient group. The Institutional Review Board of KMC approved the study. Informed patient consent was waived. Statistical analysis. Outcomes were compared using longitudinal mixed-effects modeling to adjust for repeated measures across patients within the period of observation (PCDR16 and CDR17). Continuous variables were analyzed using Student t-test. The c2 test was used to evaluate categorical variables expressed as frequency and percentages. The z-test proportion of two populations was used for readmission analysis. Comparisons were made for designated DRGs as well as for the overall within designated categories and all DRGs. Given the stratification, statistical comparisons were possible between the study years. Statistical analyses were performed using SPSS version 24.0 software (IBM, Armonk, NY). A P value of < .05 was considered statistically significant.
Journal of Vascular Surgery Volume
-,
Number
Peralta et al
5
-
Table I. Demographics PCDR16 (n ¼ 405)
CDR17 (n ¼ 358)
66.5 (13.75)
67.3 (13.32)
.53
Male
63
62
.79
Female
37
38
Hypertension
23
29
.06
Coronary artery disease
18
19
.73
Chronic kidney disease
41
34
.05
Diabetes mellitus type 2
10
12
.41
Peripheral artery disease
8
6
.27
Age, years
P value
Sex
Comorbidities
CDR17, Care delivery redesign 2017; PCDR16, pre-care delivery redesign 2016. Categorical variables are presented as percentage. Continuous variables are presented as mean 6 standard deviation.
RESULTS The PCDR16 cohort included 405 patients compared with 358 patients in the CDR17 group. Patients’ demographics were similar between groups, with the exception of chronic kidney disease, which was more frequent in PCDR16 patients (Table I). There was no significant difference in procedures performed on the basis of MS-DRG or in overall procedure volume (Table II). A detailed breakdown of the levels of MS-DRG based on procedure type is provided in Table II. Carotid procedures were significantly increased in CDR17, whereas lower extremity procedures were higher in PCDR16. Payer mix remained unchanged (P ¼ .55), with Medicare beneficiaries representing 68% in PCDR16 compared with 61% in PDCR17. LOS. The overall average LOS decreased by 23%, from 6.9 days in PCDR16 to 5.4 days in CDR17 (P ¼ .003; Table III). Significant reductions in LOS occurred in carotid, endovascular abdominal and thoracic aorta, and open abdominal aorta/visceral MS-DRGs. The gap between GMLOS and LOS was reduced from 2.1 days in PCDR16 to 0.5 day in CDR17 (P < .001; Table IV) with significant reductions in all MS-DRG categories except MS-DRGs 299 to 301 (circulatory disorders), in which the LOS vs the GMLOS actually increased (P < .001). In the CDR17, 29% of patients had LOS below the GMLOS, 48% had LOS within a 1-day gap to GMLOS, and 23% had LOS 1 day longer than GMLOSda significant improvement compared with PCDR16 (P < .001). A comparison of the LOS results for the other service lines (urology, heart failure, stroke, and spine) involved in the CDR process is provided in the Supplementary Table (online only). CMI. Overall, there was a 10% increase in CMI from 3.07 in PCDR16 to 3.37 in CDR17 (P ¼ .011). The increase in CMI was manifested by an 11% increase in assignment of a CC/MCC from PCDR16 to CDR17.
Unplanned readmission. The unplanned 30-day readmission rates did not change significantly between CDR17 and PCDR16 (9.6% vs 9.8%; P ¼ .92) and were unchanged in all MS-DRG categories. Revenue, variable cost, and CM. In PCDR16, the revenue/patient was $45,817 compared with $46,654 in CDR17, a 1.8% increase of $837 per patient (Table V). There was a concomitant decrease in variable cost/ patient of 14%, from $38,278 in PCDR16 to $32,846 in CDR17, a reduction of $5432/patient. The total service CM increased 62%, from $3,053,179 in PCDR16 to $4,943,082 in CDR17, resulting in a total CM increase of $1.89 million. This represented an 83% increase in CM/patient from $7539 in PCDR16 to $13,807 in CDR17.
DISCUSSION The CDR initiative at KMC improved patient care efficiency and more effectively aligned LOS and cost with the required MS-DRG GMLOS. With the NP serving as the CDR “quarterback,” changes in the process of care, education, and real-time data availability improved overall patient care efficiency, documentation, and cost structure. The CDR resulted in a 23% decrease in LOS and an increase in revenue coupled with a reduction in variable costs. This led to a 62% increase in CM and 83% increase in CM/patient without an increase in unplanned readmissions. This suggests a significant improvement in care efficiency with no decrease in quality of care. The literature contains little information about the impact of allied health care practitioners on the efficiency and quality of care for vascular surgery patients. A study by Reed et al5 demonstrated a decrease in LOS for vascular surgery patients when a physician assistant was empowered to be the LOS officer focused on discharge planning. This construct resulted in a 33% decrease in LOS with an overall reduction in patient care costs and an estimated financial benefit of $616,200 in the first year and $847,500 in the second year. The LOS reduction and financial benefit were promptly lost when the position was later terminated. This concept was also used at El Camino Hospital, where an LOS officer along with changes in discharge coordination and innovative data analytics reduced LOS by 7.8%, readmission rate by 14.8%, hospital-acquired conditions by 55%, and patient safety indicator occurrences by 32%, resulting in an annual cost savings of $2 million.6 The designation of an LOS officer or quarterback was essential to the success reported in both of these studies as well as in the CDR. Another aspect of CDR was addressing process of care barriers that hindered the efficiency of patient care and increased costs. Creation of standardized order sets and the mobile app along with a checklist to organize daily workflow changed the patient care process in the CDR, lowering LOS and cost. Other groups have reported similar efforts but with different approaches. For
6
Journal of Vascular Surgery
Peralta et al
---
2019
Table II. Patients in each Medicare Severity Diagnosis Related Group (MS-DRG) and comorbidities and complications (CC) or major comorbidities and complications (MCC) category MS-DRG
MS-DRG description
PCDR16, No. (%)
CDR17, No. (%)
8 (9)
P value
Lower extremity bypass, thrombectomy, endarterectomy, graft angioplasty 254
Other vascular procedures
24 (22)
253
w/ CC
49 (46)
41 (47)
252
w/ MCC
34 (32)
39 (44)
107 (100)
88 (100)
All
.03
Carotid artery stenosis; carotid endarterectomy 39
Extracranial procedures
38
w/ CC
37
w/ MCC All
10 (53)
5 (22)
5 (26)
16 (70)
4 (21)
2 (8)
19 (100)
23 (100)
.02
Endovascular abdominal aortic aneurysm repair 269
Aortic/heart assist procedures except pulsation balloon
36 (79)
42 (62)
268
w/ MCC
15 (21)
26 (38)
All
51 (100)
68 (100)
1 (4)
4 (11)
.32
Open and endovascular repair of thoracic aortic aneurysm, dissection, and thoracoabdominal aortic aneurysm 221
Cardiac valve/other major cardiothoracic procedures
220
w/ CC
10 (38)
11 (31)
219
w/ MCC
15 (58)
21 (58)
All
26 (100)
36 (100)
.53
Open aorta repair of abdominal aortic aneurysm, aortoiliac disease, visceral and aortoiliac/femoral bypass 272
Other major cardiovascular procedures
271
w/ CC
270
w/ MCC All
2 (9)
3 (14)
10 (43)
13 (62)
11 (48)
5 (24)
23 (100)
21 (100)
.25
Circulatory disorders except coronaries; admission with no procedures performed 301
Peripheral vascular disorder
300
w/ CC
299
w/ MCC All a
Other MS-DRGs
7 (24)
7 (27)
13 (45)
10 (38)
9 (31)
9 (35)
29 (100)
26 (100)
.89
150
119
.27
CDR17, Care delivery redesign 2017; PCDR16, pre-care delivery redesign 2016. P values were computed from Pearson c2 test looking at the distribution of the DRGs within the category. Boldface P values indicate statistical significance. a Other MS-DRGs occurred in three or fewer inpatients.
instance, Collier7 used preoperative optimization, outpatient preanesthesia assessment, patient education, early discharge planning, and selective use of the intensive care unit to lower LOS and to reduce costs by $1.256 million. Using a different approach, Itoga et al8 focused on modifications in imaging utilization, preoperative care, and medications to reduce the cost of imaging by 93%, the cost of preoperative care by 35%, and the cost of medications by 38%. Implementation of initiatives empowering an individual to quarterback change requires frequent and repetitive feedback to assess performance and to make
adjustments. Data to measure the real-time impact of any initiative are thus critical. In CDR, the VIO provided the data through the weekly tracking of all inpatients on the vascular surgery service. Review of the tracker at the weekly vascular surgery conference provided consistent feedback to the clinical team regarding patient care deficiencies, resource allocation, and clinical documentation. The tracker presented by the senior resident or fellow with the faculty and junior residents as the audience enhanced resident system-based education along with ownership of patient care and the importance of appropriate documentation.
Journal of Vascular Surgery Volume
-,
Number
Peralta et al
7
-
Table III. Mean length of stay (LOS) by Medicare Severity Diagnosis Related Group (MS-DRG) LOS, days MS-DRG
MS-DRG description
PCDR16
CDR17
2.0 (1.6)
P value
Lower extremity bypass, thrombectomy, endarterectomy, graft angioplasty 254
Other vascular procedures
3.0 (2.0)
253
w/ CC
6.3 (8.7)
4.4 (2.7)
252
w/ MCC
6.9 (5.7)
6.5 (6.6)
All
5.7 (6.9)
5.1 (5.0)
.846
Carotid artery stenosis; carotid endarterectomy 39
Extracranial procedures
1.5 (0.5)
1.0 (0)
38
w/ CC
3.8 (4.2)
2.3 (1.3)
37
w/ MCC
3.0 (2.8)
4.0 (4.2)
All
2.4 (2.5)
2.1 (1.6)
.003
Endovascular abdominal aortic aneurysm repair 269
Aortic/heart assist procedures except pulsation balloon
268
w/ MCC All
3.8 (3.4)
2.4 (2.5)
17.7 (16.1)
9.7 (8.1)
7.9 (11.0)
5.2 (6.4)
2 (0)
4 (1)
.003
Endovascular repair of thoracic aortic aneurysm, dissection, and thoracoabdominal aortic aneurysm 221
Cardiac valve/other major cardiothoracic procedures no cardiac catheterization
220
w/ CC
6.9 (4.7)
6.0 (2.9)
219
w/ MCC
13.2 (8.4)
11.9 (8.5)
All
10.3 (7.7)
9.3 (7.7)
6.0 (1.4)
2.0 (1.7)
.003
Open aorta repair of abdominal aortic aneurysm, aortoiliac disease, visceral, and aortoiliac/femoral bypasses 272
Other major cardiovascular procedures
271
w/ CC
270
5.1 (1.9)
3.8 (2.5)
w/ MCC
12.2 (9.8)
12.4 (15.8)
All
8.6 (7.6)
5.6 (8.3)
3.3 (2.8)
1.7 (0.8)
.024
Circulatory disorders except coronaries; admission with no procedures performed 301
Peripheral vascular disorder
300
w/ CC
4.5 (3.3)
4.1 (2.7)
299
w/ MCC
4.6 (1.9)
6.4 (5.0)
All Other MS-DRGsa b
All MS-DRGs
4.2 (2.8)
4.3 (3.8)
7.64 (7.7)
5.5 (4.6)
.018
6.9 (7.8)
5.4 (5.7)
.003
CC, Comorbidities and complications; CDR17, care delivery redesign 2017; MCC, major comorbidities and complications; PCDR16, pre-care delivery redesign 2016. Values are reported as mean 6 standard deviation. Boldface P values indicate statistical significance. a Other MS-DRGs occurred in three or fewer inpatients. b Calculation of P value for all MS-DRGs includes other MS-DRGs.
As can be seen, a variety of approaches can be used to increase care efficiency and to decrease LOS with financial benefit as a byproduct. Although those mentioned in other reports differ in many respects from the CDR, certain themes emerge. First, there needs to be a designated individual or individuals (physician, nurse, or allied health care provider) tasked to oversee and to implement the process. Second, specific strategies need to be formulated to optimize the process of care. Thus, in
the CDR, the standardized order sets emphasizing early social worker and case management consultations facilitated discharge planning and early ambulation, whereas in other reports, preoperative optimization, limitation of intensive care unit use, or modifications in imaging utilization were made. Third, continuous real-time feedback using hospital-derived data and analytics is essential to refining strategies of care and making adjustments when necessary.
8
Journal of Vascular Surgery
Peralta et al
---
2019
Table IV. Mean length of stay (LOS) e geometric mean LOS (GMLOS) benchmark LOS, delta days MS-DRG
PCDR16
CDR17
P value
Lower extremity bypass, thrombectomy, endarterectomy, graft angioplasty 0.04 (1.6)
254
0.7 (1.9)
253
2.1 (8.74)
0.1 (2.7)
252
1.4 (5.7)
0.9 (6.5)
All
1.5 (6.7)
0.4 (4.7)
<.001
Carotid artery procedures; carotid endarterectomy 39
0.2 (0.5)
38
1.5 (4.2)
37
0.3 (0) 0.1 (1.3) 1.3 (4.2)
2.4 (2.9) 0.02
All
0.2 (1.4)
<.001
Endovascular abdominal aortic aneurysm repair 269
1.9 (3.3)
0.5 (2.4)
268
10.8 (16.1)
3.0 (8.1)
All
4.5 (9.8)
1.5 (5.5)
<.001
Open and endovascular repair of thoracic aortic aneurysm, dissection, thoracoabdominal aortic aneurysm 2.9 (**)
221
0.2 (**)
220
0.4 (4.6)
0.4 (2.9)
219
3.6 (8.3)
4.2 (8.4)
All
2.1 (7.1)
1.3 (6.7)
<.001
Open repair of abdominal aortic aneurysm, aortoiliac disease, visceral, aortoiliac/femoral bypasses 272
3.7 (1.4)
0.3 (1.7)
271
0.6 (1.9)
0.5 (2.5)
270
5.7 (9.7)
5.9 (15.7)
All
3.3 (7.14)
1.0 (7.8)
<.001
Circulatory disorders except coronaries; admission with no procedure performed 301
0.6 (2.7)
0.9 (0.7)
300
0.9 (3.2)
0.6 (2.7)
299
0.3 (1.9)
2.3 (4.9)
All
0.6 (2.7)
0.8 (3.5)
<.001
Other MS-DRGs
2.1 (6.3)
0.3 (3.9)
<.001
All MS-DRGsb
2.1 (6.8)
0.5 (4.9)
<.001
a
CDR17, Care delivery redesign 2017; MS-DRG, Medicare Severity Diagnosis Related Group; PCDR16, pre-care delivery redesign 2016. Values are reported as mean 6 standard deviation. Boldface P values indicate statistical significance. **Indicates standard deviation not valid due to small sample size. a Other MS-DRGs occurred in three or fewer inpatients. b Calculation of P value for all MS-DRGs includes other MS-DRGs.
The flip side of cost is revenue, which in the CDR was addressed by focusing on education in clinical documentation and coding. CMS pays hospitals using a bundle payment system for inpatient care services by a predetermined number of allowed days based on MS-DRG category.2,9 The predetermined number of days (GMLOS) frequently undervalues vascular surgery services when clinical documentation is inaccurate or not sufficiently granular to capture actual CC/MCC. Studies by Aiello et al10 and Ayub et al11 documented that poor clinical documentation and coding mistakes lead to less than optimal reimbursement for vascular surgery patients. Ayub found that simple coding errors and undercoding at her institution
for patients undergoing endovascular aortic aneurysm repair resulted in a substantial ($500,000) loss of revenue. Aiello remedied a similar finding at his institution for patients undergoing carotid endarterectomy by using physician education to elevate awareness on the importance of clinical documentation. Thus, physician awareness of the impact that clinical documentation makes on MSDRG assignment is essential for reasonable institutional reimbursement. Furthermore, coding of vascular surgery procedures is often complex, leading to inaccurate coding and undervaluing of vascular surgery services. The ultimate result is fewer financial resources being available for the care of the vascular patient.
Journal of Vascular Surgery Volume
-,
Number
Peralta et al
Table V. Financial metrics
Total revenue Revenue/patient Variable cost Variable cost/patient CM CM/patient
9
-
PCDR16 (n ¼ 405)
CDR17 (n ¼ 358)
Delta
$18,555,789
$16,701,977
$1,853,812
$45,817
$46,654
$837
$15,502,610
$11,758,896
$3,743,714
$38,278
$32,846
$5432
$3,053,179
$4,943,082
$1,889,903
$7539
$13,807
$6269
CDR17, Care delivery redesign 2017; CM, contribution margin; PCDR16, pre-care delivery redesign 2016.
The CDR process also addressed an important gap in residents’ education regarding inpatient documentation, coding, and reimbursement. This educational process informed their understanding of MS-DRGs for each surgical procedure, the importance of the CC/MCC documentation, and the corresponding GMLOS. Furthermore, residents became familiar with the most common vascular CC/MCC that have an impact on GMLOS and how these items move the MS-DRG to the level of reimbursement that is appropriately aligned with the principal diagnosis, procedure, and patient course in the hospital. Before the CDR, vascular surgery residents were unaware of how suboptimal documentation can negatively affect hospital patient reimbursement for vascular surgery care. Critical to this education was the mobile app developed specifically for the residents.3 The app provided electronic information on the iPhone or Android device concerning coding as well as vascular surgery service elements of the CDR. The app coupled with a designated vascular service line CDI specialist minimized coding variability and facilitated discussion of patient clinical documentation before the final DRG coding. The Affordable Care Act created the Hospital Readmissions Reduction Program in 2012, implementing a penalty system for hospitals with high 30-day readmission rates for myocardial infarction, heart failure, pneumonia, and chronic lung conditions.12-15 Unplanned readmissions for vascular surgery patients are common and costly because of the high prevalence of associated comorbidities and the acuity of many vascular disorders. Glebova et al16 analyzed the 2012-2013 American College of Surgeons National Surgical Quality Improvement Program for unplanned readmissions in 86,238 vascular patients and documented an overall 9.3% readmission rate similar to this study’s. The absence of an increase in unplanned readmissions despite a decrease in LOS provides confirmation of our initial impression that before CDR, the extended LOS of vascular surgery patients was a result of inefficient systematic processes rather than patient need. Our CDR was able to improve care efficiency with no change in readmissions. There are certain limitations to the study, including its retrospective nature. All elements of the CDR were
instituted simultaneously, making it impossible to specifically identify which elements were most critical and contributed directly to the effectiveness of the CDR. Identifying the essential, most pivotal elements is an important task and worthy of a follow-up study. In addition, we were unable to capture unplanned readmissions when patients were admitted to another facility. The assumption was that the incidence of this occurrence would be the same before and after the CDR; however, we do not have the data to confirm or to refute this assumption. Finally, the elements of CDR were able to address identified institutional inefficiencies; however, other institutions may need to address different aspects of care to have a similar result. Nevertheless, the principles of an inpatient clinician quarterback, education directed at clinical documentation for correct coding, and weekly review of patient hospital course, LOS, and DRG assignments using real-time data are generalizable and are strongly recommended in implementing a CDRlike process at other institutions.
CONCLUSIONS This study evaluated a specific redesign in vascular surgery inpatient care at a single quaternary care institution. The design is predicated on the critical role of the allied health care professional, the need for precise and timely clinical documentation, the importance of multidisciplinary collaboration by all stakeholders, and the value of hospital data/analytics. As described, the redesign resulted in reductions in LOS and patient care costs, resulting in a substantial increase in the CM of the vascular surgery inpatient service.
AUTHOR CONTRIBUTIONS Conception and design: JP, CJ, AI, LL, FW Analysis and interpretation: JP, SH, GM, CL, LL, CA, TW, FW Data collection: JP, AI, LL, CA, TW Writing the article: JP, SH, AI, LL, CA, TW, FW Critical revision of the article: JP, SH, GM, CL, CJ, FW Final approval of the article: JP, SH, GM, CL, CJ, AI, LL, CA, TW, FW Statistical analysis: JP, GM, CL Obtained funding: Not applicable Overall responsibility: FW
REFERENCES 1. National Health Expenditure projections 2017-2026d forecast summary. Available at: https://www.cms.gov/Re search-Statistics-Data-and-Systems/Statistics-Trends-and-Re ports/NationalHealthExpendData/NationalHealthAccounts Projected.html. Accessed December 10, 2018. 2. Acute inpatient PPS. Available at: https://www.cms.gov/ Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatient PPS/index.html. Accessed December 10, 2018. 3. Johnson CE, Peralta J, Lawrence L, Issai A, Weaver FA. Focused resident education and engagement in quality improvement enhances documentation, shortens length of
10
Journal of Vascular Surgery
Peralta et al
---
4.
5.
6.
7. 8.
9. 10.
11.
stay, and creates a culture of continuous improvement. J Surg Ed 2019;76:771-8. Case mix index. Available at: http://www.healthand hospitalcommission.com/docs/May26Meeting/CasemixIndex Defintion.pdf. Accessed December 10, 2018. Reed T Jr, Veith FJ, Garguilo NJ III, Timaran CH, Ohki T, Lipsitz EC, et al. System to decrease length of stay for vascular surgery. J Vasc Surg 2004;39:395-9. Patient-centered LOS reduction initiative improves outcomes, saves costs. Available at: https://www.healthcatalyst. com/clients/el-camino-hospital/. Accessed December 10, 2018. Collier PE. Do clinical pathways for major vascular surgery improve outcomes and reduce cost? J Vasc Surg 1997;26:179-85. Itoga NK, Tang N, Patterson D, Ohkuma R, Lew R, Well MW, et al. Episode-based cost reduction for endovascular aneurysm repair. J Vasc Surg 2019;69:219-25.e1. Aiello FA, Reddy SP. Inpatient coding and the diagnosisrelated group. J Vasc Surg 2017;66:1621-3. Aiello FA, Judelson DR, Durgin JM, Doucet DR, Simons JP, Durocher DM, et al. A physician-led initiative to improve clinical documentation results in improved health care documentation, case mix index, and increased contribution margin. J Vasc Surg 2018;68:1524-32. Ayub SS, Scali ST, Richter JA, Huber TS, Beck A, Fatima J, et al. Financial implications of coding inaccuracies in
12.
13.
14. 15.
16.
2019
patients undergoing elective endovascular abdominal aortic aneurysm repair. J Vasc Surg 2019;69:210-8. Hospital Readmissions Reduction Program (HRRP). Available at: https://www.cms.gov/Medicare/Quality-InitiativesPatient-Assessment-Instruments/Value-Based-Programs/HRRP/ Hospital-Readmission-Reduction-Program.html. Accessed December 10, 2018. Boccuti C, Casillas G. Aiming for fewer hospital U-turns: the Medicare hospital readmission reduction program. Kaiser Family Foundation Issue Brief; March 2017. Mcllvennan CK, Eapen ZJ, Allen LA. Hospital readmissions reduction program. Circulation 2015;131:1796-803. Rau J. A guide to Medicare’s reimbursement penalties and data. Available at: https://khn.org/news/a-guide-to-medicarereadmissions-penalties-and-data/. Accessed December 10, 2018. Glebova NO, Bronsert M, Hammermeister KE, Nehler MR, Black JH III, Henderson WG. Drivers of readmissions in vascular surgery patients. J Vasc Surg 2016;64:185-94.
Submitted Dec 17, 2018; accepted Mar 15, 2019.
Audio discussion from the 2018 Vascular Annual Meeting of the Society for Vascular Surgery available online at www.jvascsurg.org.
Journal of Vascular Surgery Volume
-,
Number
Peralta et al
-
Supplementary Table (online only). Length of stay (LOS) pre-care delivery redesign 2016 (PCDR16) e LOS care delivery redesign 2017 (CDR17) for service lines LOS, mean days Service line
PCDR16
CDR17
Delta, mean days
Urology
3.6
3.8
0.2
Heart failure
6.2
5.9
0.3
10.4
8.8
1.6
Spine
5.3
4.5
0.8
Vascular
6.9
5.4
1.5
Stroke
10.e1
Impact of a care delivery redesign initiative for vascular surgery Joyce Peralta, DNP, ACNP-BC,a Sung Wan Ham, MD,a Gregory A. Magee, MD, MSc,a Christianne Lane, PhD,b Cali Johnson, MD,a Alice Issai, MBA,c Lindsey Lawrence, MHA,c Chris Allen, BS,c Todd Wilson, MBA,a and Fred A. Weaver, MD, MMM,a Los Angeles, Calif
ABSTRACT Objective: A hospital-wide quality improvement process through a care delivery redesign (CDR) was initiated to improve patient care efficiency, clinical documentation, and length of stay (LOS). The impact of CDR was assessed through LOS, unplanned readmission rates, and hospital financial metrics. Methods: The CDR team consisted of the Chief of Vascular Surgery, inpatient nurse practitioner, dedicated case manager, clinical documentation improvement specialist, and vascular surgery residents and faculty. The nurse practitioner facilitated patient care coordination, resident system-based education, and multidisciplinary collaboration. Tools created to track performance and to ensure sustainability included daily discussions of patient care barriers and solutions; standardized order sets; a mobile app for residents containing resident service expectations, disease-specific resources, and vascular surgery journal links; and a weekly inpatient tracker showing real-time patient care data. Outcome measures included LOS, case mix index, contribution margin, and unplanned readmissions. Each outcome was determined for all inpatient admissions the year before and the 12 months after CDR was initiated. Outcomes were compared between the two groups. Results: Implementation of CDR resulted in a 23% decrease in LOS (P ¼ .003), reducing the gap to the Centers for Medicare and Medicaid Services geometric mean LOS from 2.1 days to 0.5 day (P < .001). Clinical documentation resulted in an increase in case mix index of 10% (P ¼ .011). The 30-day unplanned readmission rates did not change in the 12 months after CDR was initiated compared with the year before (P ¼ .92). Financial data demonstrated decreased variable cost and increased revenue resulting in a $1.89 million increase in contribution margin. Conclusions: A CDR predicated on a dedicated service line advanced practitioner, clinical documentation education, weekly service tracker review, and real-time management of system-related barriers to patient care is described. Implementation of the CDR reduced hospital LOS with no change in unplanned readmissions and provided significant financial benefit to the hospital by increasing revenue and decreasing variable cost. (J Vasc Surg 2019;-:1-10.) Keywords: Length of stay; Diagnosis-related group; Geometric mean length of stay; Case mix index; Variable cost; Contribution margin
The demands of today’s health care environment have placed tremendous pressure on hospitals nationwide to decrease length of stay (LOS) and to deliver patientcentered care while minimizing costs. In 2016, the Centers for Medicare and Medicaid Services (CMS) reported a 4.3% growth in health care expenditures to $3.3 trillion
From the Division of Vascular Surgery and Endovascular Therapy, Department of Surgery,a and Department of Preventive Medicine,b Keck School of Medicine, and Keck Medical Center,c University of Southern California. Author conflict of interest: none. Presented at the 2018 Vascular Annual Meeting of the Society for Vascular Surgery, Boston, Mass, June 20-23, 2018. Additional material for this article may be found online at www.jvascsurg.org. Correspondence: Fred A. Weaver, MD, MMM, Division of Vascular Surgery and Endovascular Therapy, Department of Surgery, Keck School of Medicine, University of Southern California, 1520 San Pablo St, Ste 4300, Los Angeles, CA 90033 (e-mail: [email protected]). The editors and reviewers of this article have no relevant financial relationships to disclose per the JVS policy that requires reviewers to decline review of any manuscript for which they may have a conflict of interest. 0741-5214 Copyright Ó 2019 by the Society for Vascular Surgery. Published by Elsevier Inc. https://doi.org/10.1016/j.jvs.2019.03.053
or $10,348 per person. This accounted for 17.9% of the gross domestic product in the United States. The projected growth rate for health care expenditures will far outpace the rest of the economy at 5.5% per year from 2017 to 2026, reaching $5.7 trillion by 2026.1 The existing and predicted escalation of health care costs has placed increased demand on hospitals and health care organizations to decrease inpatient LOS, to reduce unplanned readmission rates, and to improve efficiency. Physician leadership is needed to accomplish these challenging goals. Current inpatient hospital reimbursement is predicated on a diagnosis-related group (DRG) classification system that is based on primary and secondary diagnoses, procedures performed, comorbidities, and complications.2 The DRG system incentivizes proper documentation to ensure that the correct DRG is applied as payment is a lump sum that is supposed to cover all hospital expenses from admission to discharge. CMS reimburses hospitals on the basis of the geometric mean LOS (GMLOS) for each individual Medicare Severity Diagnosis Related Group (MS-DRG) code. The geometric mean, which is 1
2
Journal of Vascular Surgery
Peralta et al
---
calculated by the nth root of the product of n numbers, more accurately represents the central value than a simple arithmetic mean because it is less sensitive to outliers. p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi n Geometric mean ¼ x1 x2 x3 . xn When inpatient LOS exceeds the CMS GMLOS, patient care costs can exceed reimbursement. The MS-DRG is a tiered system with increases in GMLOS if there are specific associated comorbidities and complications (CC) or major comorbidities and complications (MCC). As a quaternary medical center, Keck Medical Center of the University of Southern California (KMC) operates at a high cost structure because of patient acuity and complexity resulting in increased LOS. As a result, costs per patient discharge are often greater than reimbursement provided by the assigned MS-DRG and associated GMLOS. To address these challenges and to improve patient care efficiency, KMC started an initiative called care delivery redesign (CDR), which engaged nursing case management, a clinical documentation improvement (CDI) specialist, and the value information office (VIO). The vascular surgery service was one of five service lines (along with urology, heart failure, stroke, and spine) included in the CDR. The CDI specialist and VIO were pre-existing KMC staff, so no additional costs were incurred. The primary objectives of the CDR were to lower costs and to optimize patient care by improving care efficiency, facilitating vascular surgery service MS-DRG education, and leveraging real-time data analytics to address systematic barriers to patient care. The results of the CDR initiative and its impact on LOS and associated financial metrics were evaluated after 1 year of implementation.
METHODS CDR program structure. The vascular surgery service CDR involved the Chief of Vascular Surgery, the inpatient vascular surgery nurse practitioner (NP), a dedicated vascular surgery case manager, and vascular surgery residents and faculty. A VIO performance improvement analyst provided a weekly snapshot of vascular surgery service performance by tracking patient disposition with actual LOS compared with GMLOS for the assigned MS-DRG. The MS-DRG assignment, CC/MCC, and GMLOS were provided by the clinical documentation office. To evaluate all inpatients irrespective of insurer, nonMedicare beneficiary DRG assignments were converted to MS-DRG and associated GMLOS using 3M Coding and Reimbursement (3M, Maplewood, Minn) software. Similar CDR program structures were instituted contemporaneously for urology, heart failure, stroke, and spine service lines. Team member organization and education. The NP educated all fellows, residents, and interns on CMS terminology (GMLOS, MS-DRG) and how clinical
2019
ARTICLE HIGHLIGHTS d
d
d
Type of Research: Single-center retrospective cohort study Key Findings: Implementation of the care delivery redesign for vascular surgery patients in a hospital setting resulted in a decrease in length of stay by 23% with no change in 30-day unplanned readmission rates and a concomitant decrease in variable cost/patient, resulting in an increase of $1.89 million in contribution margin for the hospital. Take Home Message: A redesign of vascular surgery care can lead to more efficient patient care, which is associated with a significant financial benefit.
documentation influences MS-DRG tier and assigned GMLOS. Understanding the different vascular MS-DRGs, respective CC/MCC, and their corresponding GMLOS was the focus. In addition, all trainees received CDI education from the CDI specialist to further reinforce how accurate documentation affects MS-DRG level and patient severity of illness. Each patient’s assigned admission MS-DRG and corresponding GMLOS were displayed in a banner in the electronic medical record and updated daily to reflect the most accurate working MS-DRG and GMLOS. The NP also created standardized, procedurebased clinical pathways and order sets to standardize and to structure the patient care process. These order sets provided, among other things, for early ambulation and physical therapy evaluations for patients undergoing lower extremity bypass procedures; prompt restarting of home blood pressure medications; default physical therapy and occupational therapy orders, depending on procedure performed; and early social work consultations to recognize and to address discharge barriers, such as limited family resources, transportation issues, and insurance limitations. All faculty and trainees were required to use the standardized order sets at the onset of the CDR process. All this information was available on an institutionally designed mobile app, which has been previously described.3 The mobile app also included resident service expectations, disease-specific resources, and vascular surgery journal links. Directed by the Chief of Vascular Surgery, the NP established a daily patient care workflow involving closed-loop communication with the service line case manager. Communication included anticipated LOS, potential discharge needs, expected patient disposition, and potential discharge date. The NP created a care coordination checklist that was reviewed during every morning and afternoon rounds with the vascular surgery team. Checklist items included review of LOS compared with GMLOS, discussions on barriers to care, revisions of daily progress notes necessary to capture appropriate CC/MCC, and plans for discharge and
Journal of Vascular Surgery -,
In-house
Volume
Number
Peralta et al
Aending Physician
Admit Date
LOS
Working CMS GMLOS
Working LOS O/E
MS-DRG
ROWE MD, VINCENT L
4/10/2018
7.8
1.8
4.32
269
WEAVER MD, FRED A
4/9/2018
9.4
6.5
1.45
270
WEAVER MD, FRED A
4/16/2018
2.5
2.1
1.17
35
HAN MD, SUKGU M
4/12/2018
6.3
6.2
1.01
220
WEAVER MD, FRED A
4/16/2018
2.1
2.3
0.93
254
WEAVER MD, FRED A
4/16/2018
2.3
3.6
0.62
863
ROWE MD, VINCENT L
4/17/2018
1.1
4
0.28
299
AORTIC AND HEART ASSIST PROCEDURES EXCEPT PULSATION BALLOON W/O MCC OTHER MAJOR CARDIOVASCULAR PROCEDURES W MCC CAROTID ARTERY STENT PROCEDURE W CC CARDIAC VALVE & OTH MAJ CARDIOTHORACIC PROC W/O CARD CATH W CC OTHER VASCULAR PROCEDURES W/O CC/MCC POSTOPERATIVE & POSTTRAUMATIC INFECTIONS W/O MCC PERIPHERAL VASCULAR DISORDERS W MCC
Comments
USC-5N
USC-CC
Max GMLOS for DRG group
USC-5N USC-7W USC-5N USC-5N
Pending procedure
USC-5N
Max GMLOS for DRG group
MS-DRG
Aending Physician
Admit Date
Discharge Date
LOS
Working CMS GMLOS
Working LOS O/E
MS-DRG
ROWE MD, VINCENT L
4/2/2018
4/12/2018
10.4
4.2
2.49
253
MAGEE MD, GREGORY
4/2/2018
4/11/2018
9.0
4.8
1.88
314
MAGEE MD, GREGORY
4/1/2018
4/17/2018
16.0
9.7
1.65
957
ROWE MD, VINCENT L
4/7/2018
4/18/2018
10.4
6.5
1.59
270
HAN MD, SUKGU M
4/9/2018
4/11/2018
2.5
1.8
1.39
269
ZIEGLER MD, KENNETH
4/10/2018
4/13/2018
3.2
2.4
1.33
301
HAM MD, SUNG W
4/9/2018
4/12/2018
3.2
4
0.81
299
ROWE MD, VINCENT L
4/9/2018
4/12/2018
3.2
4.2
0.76
253
ZIEGLER MD, KENNETH
4/13/2018
4/15/2018
2.4
3.3
0.73
300
HAN MD, SUKGU M
4/8/2018
4/12/2018
4.2
6.2
0.68
220
WEAVER MD, FRED A
4/16/2018
4/17/2018
1.2
2.5
0.49
983
HAM MD, SUNG W
4/11/2018
4/13/2018
2.1
4.4
0.48
271
AVG D/C LOS: AVG GMLOS: Weekly Working O/E:
Current Unit
MS-DRG Descripon
GMLOS
Actual LOS
Discharge d Pat ie nt s
3
-
MS-DRG Descripon
Discharge Unit
OTHER VASCULAR PROCEDURES W USC-5N CC OTHER CIRCULATORY SYSTEM USC-5N DIAGNOSES W MCC OTHER O.R. PROCEDURES FOR MULTIPLE SIGNIFICANT TRAUMA W USC-5N MCC OTHER MAJOR CARDIOVASCULAR USC-7W PROCEDURES W MCC AORTIC AND HEART ASSIST PROCEDURES EXCEPT PULSATION USC-5N BALLOON W/O MCC PERIPHERAL VASCULAR DISORDERS USC-5N W/O CC/MCC PERIPHERAL VASCULAR DISORDERS USC-5N W MCC OTHER VASCULAR PROCEDURES W USC-5N CC PERIPHERAL VASCULAR DISORDERS USC-5N W CC CARDIAC VALVE & OTH MAJ CARDIOTHORACIC PROC W/O CARD USC-5N CATH W CC EXTENSIVE O.R. PROCEDURE UNRELATED TO PRINCIPAL USC-5N DIAGNOSIS W/O CC/MCC OTHER MAJOR CARDIOVASCULAR USC-7W PROCEDURES W CC
Comments Final coded Max GMLOS for DRG group Max GMLOS for DRG group Max GMLOS for DRG group
Final coded Max GMLOS for DRG group
CDI to put in Na level severity query
5.66 4.50 1.26
Fig 1. Vascular surgery weekly patient and discharge tracker. CC, Comorbidities and complications; CDI, clinical documentation improvement; CMS, Centers for Medicare and Medicaid Services; DRG, diagnosis-related group; GMLOS, geometric mean length of stay; LOS, length of stay; MCC, major comorbidities and complications; MSDRG, Medicare Severity Diagnosis Related Group; O/E, observed/expected; O.R., operating room.
patient disposition including alternatives. The NP also independently communicated with the CDI specialist to provide updates on changes in a patient’s medical condition such that the CC/MCC reflected the most accurate working MS-DRG. All CC/MCC that occurred
during a patient admission were documented in the discharge summary. In addition, the NP and dedicated CDI specialist tracked each patient after discharge to ensure assignment of the accurate final MS-DRG code.
4
Peralta et al
Journal of Vascular Surgery ---
2019
Fig 2. Graphic of mean length of stay (LOS) for all procedures vs geometric mean length of stay (GMLOS) for all procedures (y-axis, mean days; x-axis, time). CMS, Centers for Medicare and Medicaid Services.
Weekly inpatient tracker. At the weekly vascular surgery conference, the faculty, fellows, residents, and staff reviewed all current inpatients and recently discharged patients using the weekly inpatient tracker provided by the VIO specialist. The tracker table displayed patient identifiers, admission/discharge unit and date, vascular surgery attending physician, current LOS, GMLOS, current MS-DRG, and associated CC/MCC (Fig 1). The tracker MS-DRG column was color coded to indicate whether a patient’s LOS was under (green), within (yellow), or over (red) the GMLOS benchmark. A second graphic analysis displayed the trend of our actual LOS vs GMLOS over time (Fig 2). Weekly review of the tracker enabled evaluation of current management and barriers to efficient care, providing an opportunity for modification of clinical decision-making and patient disposition. In addition, the tracker provided real-time feedback on coding and associated GMLOS based on current electronic medical record documentation. CDR analysis. To evaluate the impact of CDR, inpatient admissions to the vascular surgery service at KMC during the initial 12 months of CDR (July 2016-June 2017 [CDR17]) were compared with all inpatient vascular surgery admissions in the immediately preceding fiscal year (pre-CDR 2016 [PCDR16]). Data collected for both groups included patient demographics, procedures performed, LOS, MS-DRG, GMLOS, insurance type, and unplanned readmissions. Unplanned readmissions were defined as any unanticipated admission to the KMC vascular surgery service within 30 days of the patient discharge date. Case mix index (CMI) was calculated by dividing the total MS-DRG weight of all the vascular surgery inpatient
discharges for the year by the number of hospital discharges.4 Data were also collected for hospital revenue from all inpatient reimbursements received from Medicare, Medicaid, and other third-party payers as well as for variable cost, which is defined as the cost of the inpatient hospitalization exclusive of overhead and fixed costs. Contribution margin (CM) was determined by subtracting variable cost from revenue. Financial data and metrics were provided, reviewed, and confirmed by the Chief Financial Officer of KMC. Outcomes measured. Primary outcome measures included LOS compared with GMLOS, CMI, CM, and unplanned readmission rate. To determine changes in outcome measures after institution of the CDR, the PCDR16 group was compared with the CDR17 patient group. The Institutional Review Board of KMC approved the study. Informed patient consent was waived. Statistical analysis. Outcomes were compared using longitudinal mixed-effects modeling to adjust for repeated measures across patients within the period of observation (PCDR16 and CDR17). Continuous variables were analyzed using Student t-test. The c2 test was used to evaluate categorical variables expressed as frequency and percentages. The z-test proportion of two populations was used for readmission analysis. Comparisons were made for designated DRGs as well as for the overall within designated categories and all DRGs. Given the stratification, statistical comparisons were possible between the study years. Statistical analyses were performed using SPSS version 24.0 software (IBM, Armonk, NY). A P value of < .05 was considered statistically significant.
Journal of Vascular Surgery Volume
-,
Number
Peralta et al
5
-
Table I. Demographics PCDR16 (n ¼ 405)
CDR17 (n ¼ 358)
66.5 (13.75)
67.3 (13.32)
.53
Male
63
62
.79
Female
37
38
Hypertension
23
29
.06
Coronary artery disease
18
19
.73
Chronic kidney disease
41
34
.05
Diabetes mellitus type 2
10
12
.41
Peripheral artery disease
8
6
.27
Age, years
P value
Sex
Comorbidities
CDR17, Care delivery redesign 2017; PCDR16, pre-care delivery redesign 2016. Categorical variables are presented as percentage. Continuous variables are presented as mean 6 standard deviation.
RESULTS The PCDR16 cohort included 405 patients compared with 358 patients in the CDR17 group. Patients’ demographics were similar between groups, with the exception of chronic kidney disease, which was more frequent in PCDR16 patients (Table I). There was no significant difference in procedures performed on the basis of MS-DRG or in overall procedure volume (Table II). A detailed breakdown of the levels of MS-DRG based on procedure type is provided in Table II. Carotid procedures were significantly increased in CDR17, whereas lower extremity procedures were higher in PCDR16. Payer mix remained unchanged (P ¼ .55), with Medicare beneficiaries representing 68% in PCDR16 compared with 61% in PDCR17. LOS. The overall average LOS decreased by 23%, from 6.9 days in PCDR16 to 5.4 days in CDR17 (P ¼ .003; Table III). Significant reductions in LOS occurred in carotid, endovascular abdominal and thoracic aorta, and open abdominal aorta/visceral MS-DRGs. The gap between GMLOS and LOS was reduced from 2.1 days in PCDR16 to 0.5 day in CDR17 (P < .001; Table IV) with significant reductions in all MS-DRG categories except MS-DRGs 299 to 301 (circulatory disorders), in which the LOS vs the GMLOS actually increased (P < .001). In the CDR17, 29% of patients had LOS below the GMLOS, 48% had LOS within a 1-day gap to GMLOS, and 23% had LOS 1 day longer than GMLOSda significant improvement compared with PCDR16 (P < .001). A comparison of the LOS results for the other service lines (urology, heart failure, stroke, and spine) involved in the CDR process is provided in the Supplementary Table (online only). CMI. Overall, there was a 10% increase in CMI from 3.07 in PCDR16 to 3.37 in CDR17 (P ¼ .011). The increase in CMI was manifested by an 11% increase in assignment of a CC/MCC from PCDR16 to CDR17.
Unplanned readmission. The unplanned 30-day readmission rates did not change significantly between CDR17 and PCDR16 (9.6% vs 9.8%; P ¼ .92) and were unchanged in all MS-DRG categories. Revenue, variable cost, and CM. In PCDR16, the revenue/patient was $45,817 compared with $46,654 in CDR17, a 1.8% increase of $837 per patient (Table V). There was a concomitant decrease in variable cost/ patient of 14%, from $38,278 in PCDR16 to $32,846 in CDR17, a reduction of $5432/patient. The total service CM increased 62%, from $3,053,179 in PCDR16 to $4,943,082 in CDR17, resulting in a total CM increase of $1.89 million. This represented an 83% increase in CM/patient from $7539 in PCDR16 to $13,807 in CDR17.
DISCUSSION The CDR initiative at KMC improved patient care efficiency and more effectively aligned LOS and cost with the required MS-DRG GMLOS. With the NP serving as the CDR “quarterback,” changes in the process of care, education, and real-time data availability improved overall patient care efficiency, documentation, and cost structure. The CDR resulted in a 23% decrease in LOS and an increase in revenue coupled with a reduction in variable costs. This led to a 62% increase in CM and 83% increase in CM/patient without an increase in unplanned readmissions. This suggests a significant improvement in care efficiency with no decrease in quality of care. The literature contains little information about the impact of allied health care practitioners on the efficiency and quality of care for vascular surgery patients. A study by Reed et al5 demonstrated a decrease in LOS for vascular surgery patients when a physician assistant was empowered to be the LOS officer focused on discharge planning. This construct resulted in a 33% decrease in LOS with an overall reduction in patient care costs and an estimated financial benefit of $616,200 in the first year and $847,500 in the second year. The LOS reduction and financial benefit were promptly lost when the position was later terminated. This concept was also used at El Camino Hospital, where an LOS officer along with changes in discharge coordination and innovative data analytics reduced LOS by 7.8%, readmission rate by 14.8%, hospital-acquired conditions by 55%, and patient safety indicator occurrences by 32%, resulting in an annual cost savings of $2 million.6 The designation of an LOS officer or quarterback was essential to the success reported in both of these studies as well as in the CDR. Another aspect of CDR was addressing process of care barriers that hindered the efficiency of patient care and increased costs. Creation of standardized order sets and the mobile app along with a checklist to organize daily workflow changed the patient care process in the CDR, lowering LOS and cost. Other groups have reported similar efforts but with different approaches. For
6
Journal of Vascular Surgery
Peralta et al
---
2019
Table II. Patients in each Medicare Severity Diagnosis Related Group (MS-DRG) and comorbidities and complications (CC) or major comorbidities and complications (MCC) category MS-DRG
MS-DRG description
PCDR16, No. (%)
CDR17, No. (%)
8 (9)
P value
Lower extremity bypass, thrombectomy, endarterectomy, graft angioplasty 254
Other vascular procedures
24 (22)
253
w/ CC
49 (46)
41 (47)
252
w/ MCC
34 (32)
39 (44)
107 (100)
88 (100)
All
.03
Carotid artery stenosis; carotid endarterectomy 39
Extracranial procedures
38
w/ CC
37
w/ MCC All
10 (53)
5 (22)
5 (26)
16 (70)
4 (21)
2 (8)
19 (100)
23 (100)
.02
Endovascular abdominal aortic aneurysm repair 269
Aortic/heart assist procedures except pulsation balloon
36 (79)
42 (62)
268
w/ MCC
15 (21)
26 (38)
All
51 (100)
68 (100)
1 (4)
4 (11)
.32
Open and endovascular repair of thoracic aortic aneurysm, dissection, and thoracoabdominal aortic aneurysm 221
Cardiac valve/other major cardiothoracic procedures
220
w/ CC
10 (38)
11 (31)
219
w/ MCC
15 (58)
21 (58)
All
26 (100)
36 (100)
.53
Open aorta repair of abdominal aortic aneurysm, aortoiliac disease, visceral and aortoiliac/femoral bypass 272
Other major cardiovascular procedures
271
w/ CC
270
w/ MCC All
2 (9)
3 (14)
10 (43)
13 (62)
11 (48)
5 (24)
23 (100)
21 (100)
.25
Circulatory disorders except coronaries; admission with no procedures performed 301
Peripheral vascular disorder
300
w/ CC
299
w/ MCC All a
Other MS-DRGs
7 (24)
7 (27)
13 (45)
10 (38)
9 (31)
9 (35)
29 (100)
26 (100)
.89
150
119
.27
CDR17, Care delivery redesign 2017; PCDR16, pre-care delivery redesign 2016. P values were computed from Pearson c2 test looking at the distribution of the DRGs within the category. Boldface P values indicate statistical significance. a Other MS-DRGs occurred in three or fewer inpatients.
instance, Collier7 used preoperative optimization, outpatient preanesthesia assessment, patient education, early discharge planning, and selective use of the intensive care unit to lower LOS and to reduce costs by $1.256 million. Using a different approach, Itoga et al8 focused on modifications in imaging utilization, preoperative care, and medications to reduce the cost of imaging by 93%, the cost of preoperative care by 35%, and the cost of medications by 38%. Implementation of initiatives empowering an individual to quarterback change requires frequent and repetitive feedback to assess performance and to make
adjustments. Data to measure the real-time impact of any initiative are thus critical. In CDR, the VIO provided the data through the weekly tracking of all inpatients on the vascular surgery service. Review of the tracker at the weekly vascular surgery conference provided consistent feedback to the clinical team regarding patient care deficiencies, resource allocation, and clinical documentation. The tracker presented by the senior resident or fellow with the faculty and junior residents as the audience enhanced resident system-based education along with ownership of patient care and the importance of appropriate documentation.
Journal of Vascular Surgery Volume
-,
Number
Peralta et al
7
-
Table III. Mean length of stay (LOS) by Medicare Severity Diagnosis Related Group (MS-DRG) LOS, days MS-DRG
MS-DRG description
PCDR16
CDR17
2.0 (1.6)
P value
Lower extremity bypass, thrombectomy, endarterectomy, graft angioplasty 254
Other vascular procedures
3.0 (2.0)
253
w/ CC
6.3 (8.7)
4.4 (2.7)
252
w/ MCC
6.9 (5.7)
6.5 (6.6)
All
5.7 (6.9)
5.1 (5.0)
.846
Carotid artery stenosis; carotid endarterectomy 39
Extracranial procedures
1.5 (0.5)
1.0 (0)
38
w/ CC
3.8 (4.2)
2.3 (1.3)
37
w/ MCC
3.0 (2.8)
4.0 (4.2)
All
2.4 (2.5)
2.1 (1.6)
.003
Endovascular abdominal aortic aneurysm repair 269
Aortic/heart assist procedures except pulsation balloon
268
w/ MCC All
3.8 (3.4)
2.4 (2.5)
17.7 (16.1)
9.7 (8.1)
7.9 (11.0)
5.2 (6.4)
2 (0)
4 (1)
.003
Endovascular repair of thoracic aortic aneurysm, dissection, and thoracoabdominal aortic aneurysm 221
Cardiac valve/other major cardiothoracic procedures no cardiac catheterization
220
w/ CC
6.9 (4.7)
6.0 (2.9)
219
w/ MCC
13.2 (8.4)
11.9 (8.5)
All
10.3 (7.7)
9.3 (7.7)
6.0 (1.4)
2.0 (1.7)
.003
Open aorta repair of abdominal aortic aneurysm, aortoiliac disease, visceral, and aortoiliac/femoral bypasses 272
Other major cardiovascular procedures
271
w/ CC
270
5.1 (1.9)
3.8 (2.5)
w/ MCC
12.2 (9.8)
12.4 (15.8)
All
8.6 (7.6)
5.6 (8.3)
3.3 (2.8)
1.7 (0.8)
.024
Circulatory disorders except coronaries; admission with no procedures performed 301
Peripheral vascular disorder
300
w/ CC
4.5 (3.3)
4.1 (2.7)
299
w/ MCC
4.6 (1.9)
6.4 (5.0)
All Other MS-DRGsa b
All MS-DRGs
4.2 (2.8)
4.3 (3.8)
7.64 (7.7)
5.5 (4.6)
.018
6.9 (7.8)
5.4 (5.7)
.003
CC, Comorbidities and complications; CDR17, care delivery redesign 2017; MCC, major comorbidities and complications; PCDR16, pre-care delivery redesign 2016. Values are reported as mean 6 standard deviation. Boldface P values indicate statistical significance. a Other MS-DRGs occurred in three or fewer inpatients. b Calculation of P value for all MS-DRGs includes other MS-DRGs.
As can be seen, a variety of approaches can be used to increase care efficiency and to decrease LOS with financial benefit as a byproduct. Although those mentioned in other reports differ in many respects from the CDR, certain themes emerge. First, there needs to be a designated individual or individuals (physician, nurse, or allied health care provider) tasked to oversee and to implement the process. Second, specific strategies need to be formulated to optimize the process of care. Thus, in
the CDR, the standardized order sets emphasizing early social worker and case management consultations facilitated discharge planning and early ambulation, whereas in other reports, preoperative optimization, limitation of intensive care unit use, or modifications in imaging utilization were made. Third, continuous real-time feedback using hospital-derived data and analytics is essential to refining strategies of care and making adjustments when necessary.
8
Journal of Vascular Surgery
Peralta et al
---
2019
Table IV. Mean length of stay (LOS) e geometric mean LOS (GMLOS) benchmark LOS, delta days MS-DRG
PCDR16
CDR17
P value
Lower extremity bypass, thrombectomy, endarterectomy, graft angioplasty 0.04 (1.6)
254
0.7 (1.9)
253
2.1 (8.74)
0.1 (2.7)
252
1.4 (5.7)
0.9 (6.5)
All
1.5 (6.7)
0.4 (4.7)
<.001
Carotid artery procedures; carotid endarterectomy 39
0.2 (0.5)
38
1.5 (4.2)
37
0.3 (0) 0.1 (1.3) 1.3 (4.2)
2.4 (2.9) 0.02
All
0.2 (1.4)
<.001
Endovascular abdominal aortic aneurysm repair 269
1.9 (3.3)
0.5 (2.4)
268
10.8 (16.1)
3.0 (8.1)
All
4.5 (9.8)
1.5 (5.5)
<.001
Open and endovascular repair of thoracic aortic aneurysm, dissection, thoracoabdominal aortic aneurysm 2.9 (**)
221
0.2 (**)
220
0.4 (4.6)
0.4 (2.9)
219
3.6 (8.3)
4.2 (8.4)
All
2.1 (7.1)
1.3 (6.7)
<.001
Open repair of abdominal aortic aneurysm, aortoiliac disease, visceral, aortoiliac/femoral bypasses 272
3.7 (1.4)
0.3 (1.7)
271
0.6 (1.9)
0.5 (2.5)
270
5.7 (9.7)
5.9 (15.7)
All
3.3 (7.14)
1.0 (7.8)
<.001
Circulatory disorders except coronaries; admission with no procedure performed 301
0.6 (2.7)
0.9 (0.7)
300
0.9 (3.2)
0.6 (2.7)
299
0.3 (1.9)
2.3 (4.9)
All
0.6 (2.7)
0.8 (3.5)
<.001
Other MS-DRGs
2.1 (6.3)
0.3 (3.9)
<.001
All MS-DRGsb
2.1 (6.8)
0.5 (4.9)
<.001
a
CDR17, Care delivery redesign 2017; MS-DRG, Medicare Severity Diagnosis Related Group; PCDR16, pre-care delivery redesign 2016. Values are reported as mean 6 standard deviation. Boldface P values indicate statistical significance. **Indicates standard deviation not valid due to small sample size. a Other MS-DRGs occurred in three or fewer inpatients. b Calculation of P value for all MS-DRGs includes other MS-DRGs.
The flip side of cost is revenue, which in the CDR was addressed by focusing on education in clinical documentation and coding. CMS pays hospitals using a bundle payment system for inpatient care services by a predetermined number of allowed days based on MS-DRG category.2,9 The predetermined number of days (GMLOS) frequently undervalues vascular surgery services when clinical documentation is inaccurate or not sufficiently granular to capture actual CC/MCC. Studies by Aiello et al10 and Ayub et al11 documented that poor clinical documentation and coding mistakes lead to less than optimal reimbursement for vascular surgery patients. Ayub found that simple coding errors and undercoding at her institution
for patients undergoing endovascular aortic aneurysm repair resulted in a substantial ($500,000) loss of revenue. Aiello remedied a similar finding at his institution for patients undergoing carotid endarterectomy by using physician education to elevate awareness on the importance of clinical documentation. Thus, physician awareness of the impact that clinical documentation makes on MSDRG assignment is essential for reasonable institutional reimbursement. Furthermore, coding of vascular surgery procedures is often complex, leading to inaccurate coding and undervaluing of vascular surgery services. The ultimate result is fewer financial resources being available for the care of the vascular patient.
Journal of Vascular Surgery Volume
-,
Number
Peralta et al
Table V. Financial metrics
Total revenue Revenue/patient Variable cost Variable cost/patient CM CM/patient
9
-
PCDR16 (n ¼ 405)
CDR17 (n ¼ 358)
Delta
$18,555,789
$16,701,977
$1,853,812
$45,817
$46,654
$837
$15,502,610
$11,758,896
$3,743,714
$38,278
$32,846
$5432
$3,053,179
$4,943,082
$1,889,903
$7539
$13,807
$6269
CDR17, Care delivery redesign 2017; CM, contribution margin; PCDR16, pre-care delivery redesign 2016.
The CDR process also addressed an important gap in residents’ education regarding inpatient documentation, coding, and reimbursement. This educational process informed their understanding of MS-DRGs for each surgical procedure, the importance of the CC/MCC documentation, and the corresponding GMLOS. Furthermore, residents became familiar with the most common vascular CC/MCC that have an impact on GMLOS and how these items move the MS-DRG to the level of reimbursement that is appropriately aligned with the principal diagnosis, procedure, and patient course in the hospital. Before the CDR, vascular surgery residents were unaware of how suboptimal documentation can negatively affect hospital patient reimbursement for vascular surgery care. Critical to this education was the mobile app developed specifically for the residents.3 The app provided electronic information on the iPhone or Android device concerning coding as well as vascular surgery service elements of the CDR. The app coupled with a designated vascular service line CDI specialist minimized coding variability and facilitated discussion of patient clinical documentation before the final DRG coding. The Affordable Care Act created the Hospital Readmissions Reduction Program in 2012, implementing a penalty system for hospitals with high 30-day readmission rates for myocardial infarction, heart failure, pneumonia, and chronic lung conditions.12-15 Unplanned readmissions for vascular surgery patients are common and costly because of the high prevalence of associated comorbidities and the acuity of many vascular disorders. Glebova et al16 analyzed the 2012-2013 American College of Surgeons National Surgical Quality Improvement Program for unplanned readmissions in 86,238 vascular patients and documented an overall 9.3% readmission rate similar to this study’s. The absence of an increase in unplanned readmissions despite a decrease in LOS provides confirmation of our initial impression that before CDR, the extended LOS of vascular surgery patients was a result of inefficient systematic processes rather than patient need. Our CDR was able to improve care efficiency with no change in readmissions. There are certain limitations to the study, including its retrospective nature. All elements of the CDR were
instituted simultaneously, making it impossible to specifically identify which elements were most critical and contributed directly to the effectiveness of the CDR. Identifying the essential, most pivotal elements is an important task and worthy of a follow-up study. In addition, we were unable to capture unplanned readmissions when patients were admitted to another facility. The assumption was that the incidence of this occurrence would be the same before and after the CDR; however, we do not have the data to confirm or to refute this assumption. Finally, the elements of CDR were able to address identified institutional inefficiencies; however, other institutions may need to address different aspects of care to have a similar result. Nevertheless, the principles of an inpatient clinician quarterback, education directed at clinical documentation for correct coding, and weekly review of patient hospital course, LOS, and DRG assignments using real-time data are generalizable and are strongly recommended in implementing a CDRlike process at other institutions.
CONCLUSIONS This study evaluated a specific redesign in vascular surgery inpatient care at a single quaternary care institution. The design is predicated on the critical role of the allied health care professional, the need for precise and timely clinical documentation, the importance of multidisciplinary collaboration by all stakeholders, and the value of hospital data/analytics. As described, the redesign resulted in reductions in LOS and patient care costs, resulting in a substantial increase in the CM of the vascular surgery inpatient service.
AUTHOR CONTRIBUTIONS Conception and design: JP, CJ, AI, LL, FW Analysis and interpretation: JP, SH, GM, CL, LL, CA, TW, FW Data collection: JP, AI, LL, CA, TW Writing the article: JP, SH, AI, LL, CA, TW, FW Critical revision of the article: JP, SH, GM, CL, CJ, FW Final approval of the article: JP, SH, GM, CL, CJ, AI, LL, CA, TW, FW Statistical analysis: JP, GM, CL Obtained funding: Not applicable Overall responsibility: FW
REFERENCES 1. National Health Expenditure projections 2017-2026d forecast summary. Available at: https://www.cms.gov/Re search-Statistics-Data-and-Systems/Statistics-Trends-and-Re ports/NationalHealthExpendData/NationalHealthAccounts Projected.html. Accessed December 10, 2018. 2. Acute inpatient PPS. Available at: https://www.cms.gov/ Medicare/Medicare-Fee-for-Service-Payment/AcuteInpatient PPS/index.html. Accessed December 10, 2018. 3. Johnson CE, Peralta J, Lawrence L, Issai A, Weaver FA. Focused resident education and engagement in quality improvement enhances documentation, shortens length of
10
Journal of Vascular Surgery
Peralta et al
---
4.
5.
6.
7. 8.
9. 10.
11.
stay, and creates a culture of continuous improvement. J Surg Ed 2019;76:771-8. Case mix index. Available at: http://www.healthand hospitalcommission.com/docs/May26Meeting/CasemixIndex Defintion.pdf. Accessed December 10, 2018. Reed T Jr, Veith FJ, Garguilo NJ III, Timaran CH, Ohki T, Lipsitz EC, et al. System to decrease length of stay for vascular surgery. J Vasc Surg 2004;39:395-9. Patient-centered LOS reduction initiative improves outcomes, saves costs. Available at: https://www.healthcatalyst. com/clients/el-camino-hospital/. Accessed December 10, 2018. Collier PE. Do clinical pathways for major vascular surgery improve outcomes and reduce cost? J Vasc Surg 1997;26:179-85. Itoga NK, Tang N, Patterson D, Ohkuma R, Lew R, Well MW, et al. Episode-based cost reduction for endovascular aneurysm repair. J Vasc Surg 2019;69:219-25.e1. Aiello FA, Reddy SP. Inpatient coding and the diagnosisrelated group. J Vasc Surg 2017;66:1621-3. Aiello FA, Judelson DR, Durgin JM, Doucet DR, Simons JP, Durocher DM, et al. A physician-led initiative to improve clinical documentation results in improved health care documentation, case mix index, and increased contribution margin. J Vasc Surg 2018;68:1524-32. Ayub SS, Scali ST, Richter JA, Huber TS, Beck A, Fatima J, et al. Financial implications of coding inaccuracies in
12.
13.
14. 15.
16.
2019
patients undergoing elective endovascular abdominal aortic aneurysm repair. J Vasc Surg 2019;69:210-8. Hospital Readmissions Reduction Program (HRRP). Available at: https://www.cms.gov/Medicare/Quality-InitiativesPatient-Assessment-Instruments/Value-Based-Programs/HRRP/ Hospital-Readmission-Reduction-Program.html. Accessed December 10, 2018. Boccuti C, Casillas G. Aiming for fewer hospital U-turns: the Medicare hospital readmission reduction program. Kaiser Family Foundation Issue Brief; March 2017. Mcllvennan CK, Eapen ZJ, Allen LA. Hospital readmissions reduction program. Circulation 2015;131:1796-803. Rau J. A guide to Medicare’s reimbursement penalties and data. Available at: https://khn.org/news/a-guide-to-medicarereadmissions-penalties-and-data/. Accessed December 10, 2018. Glebova NO, Bronsert M, Hammermeister KE, Nehler MR, Black JH III, Henderson WG. Drivers of readmissions in vascular surgery patients. J Vasc Surg 2016;64:185-94.
Submitted Dec 17, 2018; accepted Mar 15, 2019.
Audio discussion from the 2018 Vascular Annual Meeting of the Society for Vascular Surgery available online at www.jvascsurg.org.
Journal of Vascular Surgery Volume
-,
Number
Peralta et al
-
Supplementary Table (online only). Length of stay (LOS) pre-care delivery redesign 2016 (PCDR16) e LOS care delivery redesign 2017 (CDR17) for service lines LOS, mean days Service line
PCDR16
CDR17
Delta, mean days
Urology
3.6
3.8
0.2
Heart failure
6.2
5.9
0.3
10.4
8.8
1.6
Spine
5.3
4.5
0.8
Vascular
6.9
5.4
1.5
Stroke
10.e1