Population Health Management

Raising CMI Through a Multidisciplinary Approach in the Community Hospital Setting

June 20, 2017 2:03 pm

Reducing interhospital transfers to neighboring urban hospitals enabled one hospital to raise its case mix index, thereby helping to increase net patient services revenue and providing a means to financially sustain a population health management initiative.

As hospitals adjust to today’s new norms under healthcare reform, a chief concern is mitigating risk from alternative payment models (APMs). To succeed in the emerging value-focused healthcare environment, hospitals seek new ways to improve financial performance while managing these new APMs in delicate and dynamic balance with traditional fee-for-service payment. Of course, providing a safe and high-quality patient experience remains paramount. But hospitals also have a mission-critical need to sustain and advance their financial well-being, without which they could not begin to deliver high-quality care. An effective way for hospitals to realize their financial goals is to focus on optimizing care delivery, service, and case mix index (CMI).

The experience of Lawrence General Hospital (LGH), a 189-bed community hospital in Lawrence, Mass., exemplifies the types of considerations that are inherent in such an effort. Facing declining margins under a population health management risk contract, LGH responded with a focus on reducing interhospital transfers. LGH found that by avoiding interhospital transfers that were not medically necessary, within the context of the larger population health platform, it could increase its case mix index (CMI), leading to improved financial sustainability and success. The catalyst for this change was a reduction in margin stemming from a decline in admissions for chest paint that began in FY14-15 as the result of efforts to meet quality standards under a risk contract.

For LGH, avoiding interhospital transfers provided a means to complement its population health-based efforts by keeping care local and driving volume back into a healthcare system. There also were clear quality considerations, given that interhospital transfers have been shown to lead to worse health outcomes, including increased mortality. a   Preventable transfers to outside hospitals also can disproportionately lead to lower critical care volume and subsequently reduce downstream revenue. Moreover, improving critical care bed flow can open up hospital beds to accommodate critical care patients. b

Background: Impact of Declining Chest Pain Admissions, FY15

In FY14-15, LGH went from an estimated deficit of more than $1 million in one risk contract to a surplus partly by adopting care guidelines and addressing unnecessary hospital admissions, one of the biggest drivers in cost. However, the changes also caused a decrease in LGH’s total volume and net patient service revenue that more than negated the retained earnings from the risk contract. LGH’s leaders recognized that further improvements in the population health management and value-based payment principle would likely work against the hospital’s existing fee-for-service payment mechanisms.

Among top drivers in overall cost were admissions per 1,000 patients, and among the top diagnosis codes was chest pain (MS-DRG 313). In accordance with a newly developed protocol for low-risk chest pain, patients presenting with chest pain and assessed as “ultra-low risk” were no longer admitted to observation or inpatient status, but were discharged home with a follow-up appointment and stress test (next-day). This determination was made using risk stratification with a validated decision-making tool, the HEART score (which assesses a patient’s History, Electrocardiogram findings, Age, Risk factors, and Troponin level).

With a focus on doing what was best for all patients, emergency department (ED) providers did not reserve this protocol solely for patient populations covered under risk contracts. ED providers began to routinely use the risk-stratification protocol for all patients to rule-out heart attacks and determine whether patients could go home, eat their own dinners, sleep in their own beds, and return to the hospital for a scheduled exercise tolerance testing (ETT) appointment the following morning rather than remain for 18 to 36 hours in hospital beds surrounded by sicker patients.

Use of the protocol prompted a decline in discharges and observations amounting to 30 to 35 cases per month, and 400 cases for the year, for this one diagnosis code, all but negating the retained earnings from the one risk contract. Because chest pain was the primary reason for ED visits under the risk contract, the chest pain practice also became the biggest cause of reduced volume overall for the hospital.

Overall, inpatient discharges were down more than 5 percent in FY15 (compared with FY14), and the downward trend continued in FY16. Nonetheless, the ED providers were rightly committed to maintaining their practice with respect to chest pain, and LGH was compelled to look elsewhere to make up for the lost margin. It became clear that patient volume, especially cases with a considerable margin, would be a much more important contributor to the organization’s future success.

Transformation: Increased Critical Care Admissions to LGH From Its ED, FY16

Of the more than 70,000 annual encounters in LGH’s ED, only about 1 percent are classified under the Emergency Severity Index (ESI) as level ESI-1 and about 9 percent are classified as ESI-2, the two levels representing the highest and second highest levels of severity, respectively. Historically at LGH, a Medicare patient categorized as ESI-1 with an order to be admitted from the ED to critical care had an average CMI of 2.6446 upon inpatient discharge, which is twice that of the typical, budgeted Medicare CMI (1.3200). At LGH, such patients typically would be transferred to critical care at another facility. LGH came to realize, however, that by admitting these and other patients to its own facility for critical care, it could increase its inpatient volume and add a disproportionate share of case weight, increasing its CMI and total net patient services revenue.

In FY14-15, LGH was transferring five ESI-1 and 13 ESI-2 patients per week from the ED to other acute care hospitals. Yet the reasons for these transfers were not always clearly explained or understood. Data therefore were collected to identify causes of transfers and opportunities for retaining the patients as inpatient admissions at LGH, if appropriate. It was clear that some of the transfers were necessitated by specific patient preference or clinical indications for a higher level of care than LGH was capable of delivering. Among the many other documented reasons for transfers, an often-cited one was lack of critical care bed availability at LGH.

Interhospital transfers resulted in a lengthy and complex transfer process involving multiple clinicians and arrangements for available space and transport across two hospitals. Patients often ended up with more expensive stays at tertiary medical centers, with extended times away from their communities and family support networks.

The findings of the analysis were presented to senior leadership at the end of FY15 as budgets were being presented for FY16. In addition to creating a strategic plan with new surgeons and primary care practitioners, LGH had to recapture this lost business and realize the full opportunity of these highest-paying inpatient discharges.

The population health team supported the effort by making preventable transfers one of its top 10 initiatives starting September 2015, with a focus on decreasing avoidable admissions to make room for the increase in admissions from the reduced interhospital critical care transfers. A 24-hour real-time rounding intervention was created to identify and facilitate transfers out of critical care throughout the day. Patient plans were updated regularly and critical care capacity was communicated to the ED.

LGH leaders engaged in discussions with physicians who were admitting and managing patients in critical care regarding appropriate use of critical care beds. The hospital had a long-standing policy of keeping a spare critical care bed open in case a patient on the hospital floor were to become critically ill, even if that necessitated transferring a patient out of the hospital. It was decided that, instead, the hospital should adopt a new policy of maintaining as many critical care beds as possible in anticipation of the next emergency.

LGH’s leaders began to look closely at the hospital’s interhospital transfer policy. Senior LGH leaders who had previously been meeting weekly to review all transfers formed a “Prevent-Transfer Team,” which was charged with collecting and reviewing transfer data from the ED. Within one to two days of each transfer from the ED, case details—including the reason for transfer, who made the decision to transfer, and the receiving facility—were sent to the team.

The team instituted a policy of calling upon a physician to weigh in on each decision that involved the possibility of transferring a patient to an urban hospital and determine whether such a transfer was necessary. Bed availability, staffing, consults, and other issues were addressed as they arose and managed when possible. More important, the new critical care rounding cycle, communication, and attention to beds created a cohesive team between ED staff, hospitalist physicians, and critical care nurses to retain patients requiring critical care.

All transfers were reviewed to ensure appropriateness of care regardless of whether it was related to critical care. For all requested transfers that were deemed avoidable and within LGH’s capabilities, the Prevent-Transfer Team implemented a process for examining the reasons for the requests, which included conversations with individual providers when it seemed likely the reason for the request was provider convenience or availability, or to access equipment that would mitigate the chance of a future transfer. These efforts helped create a culture of accountability.


The increase in critical care capacity has led to an almost threefold increase in the number of ED ESI-1 admit to critical care orders per month, from six to 17 inpatients. Total transfers from the ED to other hospitals went from 140 to 105 per month, on average. And transfers of ESI-1 and ESI-2 patients declined from 70 to 52 per month.

After two years of decreased Medicare inpatient volume, FY16 Medicare inpatient discharges were slightly ahead of FY15 numbers, having increased from 343 to 345 per month (less than 1 percent) in that time frame. Monthly Medicare inpatient discharges associated with ED admissions to critical care increased from nine to 27. The increase in these high-CMI admissions represent avoided transfers to more expensive tertiary care centers and mainly account for LGH Medicare CMI going from 1.3220 to 1.3841 from FY15 to FY16.

From this specific population, the estimated net impact from Medicare CMI reaching 1.3620 as a result of keeping care local amounted to a greater than $1.25 million increase in net patient service revenue. This population of Medicare inpatients admitted to critical care from the ED went from 3 percent of total discharges in FY15 to 8 percent in FY16. The critical nature of these patients and elevated CMI creates a disproportionate effect on total case weight, given that the 3 percent of patients represented 4 percent of total case weight in FY15, whereas 8 percent of patients represented 14 percent of total case weight in FY16. The result was a higher CMI and total case weight in FY16, resulting in an increase in net patient service revenue, despite near-flat total discharges.

Lawrence General Hospital: Impact on Case Mix Index From Increased Medicare Admissions to Critical Care From Emergency Department, By Emergency Severity Index (ESI) Level, FY15 and FY16

Further analysis was unable to detect any pattern among specific surgeons, providers, case types, or other factors. High-CMI DRG groupings such as DRGs 853, 854, and 855—Infectious and Parasitic Diseases with OR Procedure (with MCC, with CC, and w/o CC/MCC, respectively), and DRGs 981, 982, and 983—Extensive OR Procedure Unrelated to Principal Diagnosis (with MCC, with CC, and w/o CC/MCC, respectively) through the ED doubled from FY14-15 into FY16. The 70 operating room encounters within DRGs 981-983 were performed by 27 different surgeons in FY16, and 25 of the 27 surgeons each accounted for five or fewer discharges over 12 months from FY15 to FY16.

Inherently, cases admitted from the ED to inpatient status through critical care have complexities that effect not only the DRG, but also the level within the DRG in terms of complications and comorbidities (CC), given that a level “without CC” is less common among this population of patients than are the levels “with CC” and “with major CC.” The competence and ongoing management of this sicker population has enabled the return of these discharges, elevating LGH’s CMI to historic highs.

Lawrence General Hospital: Composite Case Mix Index (CMI), FY15 and FY16

The previously cited estimated $1.25 million improvement in net patient services revenue is from Medicare CMI alone. Additional revenue improvement can be expected from other health plans, both government and commercial, using a relative weight for acuity and resources. For Medicare and other inpatient discharges, this effort likely has helped decrease overall costs, produce safer outcomes, and improve patient and family satisfaction from avoidance of ambulance transfers through traffic to urban tertiary care centers.

Challenges, Opportunities, and Next Steps

To deliver state-of-the-art critical care, construction limited critical care beds by approximately one quarter to one half between April to July 2016. In FY15, 128 of LGH’s 189 licensed beds were reserved for adult patients requiring acute care. However, because of recent renovations made by LGH focused on the creation of more single-occupancy rooms, the number of adult acute care beds has been reduced to 119. Although these 119 beds are more usable, given that a higher mix of single occupancy rooms provides flexibility relative to gender and precaution status, the tighter supply of adult acute care beds also has affected critical care bed capacity, because patients with orders to transfer off the critical care unit must remain in the unit if there is no available acute care bed.

Thus, LGH’s increased critical care capacity poses a challenge that the hospital still must address: That new capacity must be balanced with an appropriate acute care bed capacity to avoid the unnecessarily higher cost of the critical care bed for a patient who could be in an acute bed. All too often, one critical care patient arriving in the ED can be followed by multiple other critically ill patients. The critical care unit and capacity is best managed with a fluid process that involves reassessing needs constantly rather than at fixed times during the day. From a lean perspective, this approach avoids instances of batching and trying to move three to four patients at the same time as a reactive process.

In FY17, LGH is seeking to optimize critical care operations by managing the new staffing demands from a nursing, mid-level provider (nurse practitioner), and intensivist physician perspective. Given higher acuity, the hospital transitioned from a model focused on generalist hospitalists and community pulmonary physicians to one focused on nurse practitioners and intensivist physicians. This model promises to allow LGH’s CMI to go even higher for 2017.

The key lesson other hospitals can take away from LHG’s experience: Optimizing critical care bed availability can support a higher retention of critically ill patients, thereby enhancing a hospital’s revenue streams, including CMI.

Brian Collins 
is director of decision support, Lawrence General Hospital, Lawrence, Mass.

George Kondylis, MD, 
chief of emergency medicine and EMS medical director, Lawrence General Hospital, Lawrence, Mass.

Schawan Kunupakaphun 
is a population health data analyst, Lawrence General Hospital, Lawrence, Mass.

Pracha Eamranond, MD, 
senior vice president of medical affairs and population health, Lawrence General Hospital, Lawrence, Mass., and assistant professor, Harvard Medical School, Boston.


a. Sokol-Hessner, L., Andrew A White, A.A., Davis, K.F.,3 Shoshana J. Herzig, S.J., MD, MPH, and Samuel F Hohmann, S.F., “Inter-hospital Transfer Patients Discharged by Academic Hospitalists and General Internists: Characteristics and Outcomes,” Journal of Hospital Medicine, April 2016.

b. Howell, E., Bessman, E., Marshall, R., Wright, S., “Hospitalist Bed Management Effecting Throughput From the Emergency Department to the Intensive Care Unit,” Journal of Critical Care, June 2010.


googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text1' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text2' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text3' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text4' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text5' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text6' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-text7' ); } );
googletag.cmd.push( function () { googletag.display( 'hfma-gpt-leaderboard' ); } );