Journal of Emergency Medicine and Intensive Care

Medicare Payment for Extended Stay Services in Frontier Clinics: Impact on the Use and Cost of Emergency Care in Alaska

Download PDF

Published Date: March 07, 2017

Medicare Payment for Extended Stay Services in Frontier Clinics: Impact on the Use and Cost of Emergency Care in Alaska

Boyd Gilman*, Laura Ruttner, and Michaella Morzuch

Mathematica Policy Research, 955 Massachusetts Avenue, Suite 801, Cambridge, MA 02139, USA

*Corresponding author: Boyd Gilman, Mathematica Policy Research, Health Division, 955 Massachusetts Avenue, Suite 801, Cambridge, MA 02139, USA, Tel: 617-301-8974; E-mail: bgilman@mathematica-mpr.com

Citation: Gilman B, Ruttner L, Morzuch M (2016) Medicare Payment for Extended Stay Services in Frontier Clinics: Impact on the Use and Cost of Emergency Care in Alaska. J Eme Med Int Care 3(1): 112

 

Abstract


Objectives: The goal of this study was to assess the impact of providing enhanced Medicare payments to frontier clinics for extended stay services on the use and cost of emergency care.

Study Design: The study is based on a difference-in-differences design in which we compare changes in emergency transfers, hospitalizations, and payments among beneficiaries treated at participating clinics after the introduction of enhanced payments compared with changes among beneficiaries from nonparticipating clinics.

Data Sources/Study Setting: The study is based on an analysis of Medicare enrollment and claims data, as well as extended stay encounter data for beneficiaries treated at four ambulatory health clinics in frontier regions of Alaska from 2008 through 2013.

Methods: We used a linear probability regression model to estimate the impact of the demonstration on the probability of having an emergency transfer or an inpatient admission within seven days of a clinic visit. Our difference-in-differences study design controls for unobservable differences in transfer and hospitalization rates between demonstration and comparison clinics unrelated to the demonstration and external trends in transfer rates affecting all clinics. We control directly for beneficiary characteristics (age, gender, race/ethnicity, dual eligibility, disability, and mortality) likely to be associated with being transferred and hospitalized.

Results: Enhanced payments for extended stay services led to a 21.4% reduction in emergency transfers and a 23.9% reduction in hospitalizations. Reimbursement for extended stay services at participating clinics also resulted in a net reduction in Medicare payments of $7,707 per extended stay episode of care.

Conclusions: Coverage for extended stays has the potential for reducing avoidable transfers and hospitalizations and lowering costs for patients treated for potentially serious illness or injury in frontier communities. However, the savings might not be sufficient to offset the investments needed to build and maintain extended stay capacity at frontier clinics without higher payment rates or adoption by other payers.

Keywords: Medicare; Rural Health; Ambulatory/Outpatient Care; Health Care Costs

 

List of Abbreviations

 

APN: Advanced Practice Nurse; CMS: Centers for Medicare & Medicaid Services; FESC: Frontier Extended Stay Clinic; FQHC: Federally Qualified Health Center; HRSA: Health Resources and Services Administration; IHS: Indian Health Service; MD: Medical Doctor; NP: Nurse Practitioner; ORHP: Office of Rural Health Policy; RHC: Rural Health Center; RN: Registered Nurse.

Top ↑

Introduction


In frontier areas of the country, weather and distance can prevent patients who experience severe injury or illness from obtaining immediate transport to an acute care hospital. Further, some emergency patients do not need an acute inpatient level of care but simply require monitoring and observation for limited periods. In some instances, when patients are unable to be transported, remote rural clinics designed and used primarily for outpatient ambulatory care and staffed by physicians or other health care professionals may offer observational or stabilization services until the patient can be transferred or is no longer in need of transport.

These services require additional staffing, equipment, and quality assurance programs beyond those usually found in rural clinics—services that are similar to, but not as extensive as, those provided in acute care hospitals. However, these extended stay services in rural health centers are not paid under Medicare and Medicaid programs, or by other third-party payers. Lack of funding for these services raises concern about the quality of care at clinics in isolated rural areas when emergency transport is not immediately available or when emergency transport is available but observation and monitoring on an extended stay basis is all that is required. Frontier clinics have always provided extended stay services when necessary, but have had to rely on existing revenue streams to cover the cost of extended stay services, challenging their ability to fulfill their core mission, which is to meet the primary health care needs of their communities.

In 2010, the Centers for Medicare & Medicaid Services (CMS) implemented the Frontier Extended Stay Clinic (FESC) Demonstration, as mandated under Section 434 of the Medicare Prescription Drug Improvement and Modernization Act of 2003 [1]. The purpose of the demonstration was to test the feasibility of providing extended stay services to Medicare beneficiaries at clinics in isolated rural areas. The FESC Demonstration allowed certified clinics in isolated rural areas to treat Medicare beneficiaries for more extended periods, including overnight stays, than are provided in routine clinic visits and to bill for those services at an enhanced payment rate under the Medicare program.

As part of the demonstration, CMS developed a set of requirements for ensuring the safety of patients who received observation and emergency services (and, in some cases, overnight care) in ambulatory health care settings [2]. To qualify as an extended stay facility and bill Medicare for extended stay services under the demonstration, participating clinics had to meet the federal conditions of participation (in addition to those they already had to meet, to obtain their level of federal and state licensure as outpatient clinics) relating to staff type and coverage, facility services and physical structure, and administrative procedures. The most resource-intensive components of these requirements were expanding staff to ensure appropriately credentialed staff were either on site or on call and within 30 minutes of the clinic 24 hours a day, seven days a week; complying with ambulatory health care occupancy life safety codes (primarily related to fire safety); and obtaining specific laboratory and radiology equipment.

The purpose of this unique study was to assess the feasibility and advisability of providing enhanced reimbursement for providing extended stay services to Medicare beneficiaries living in remote and frontier communities of the United States. We provide a descriptive analysis of extended stay encounters at demonstration clinics using patient-level encounter forms. We also provide a multivariate regression analysis of the marginal impact of providing enhanced payment for extended stay services in remote and frontier regions of the country on the total cost of care using Medicare claims and a matched comparison group of nonparticipating frontier clinics.

 

Patients and Methods


Demonstration Clinics

The five original demonstration clinics provided a wide array of family-oriented primary care services, including prenatal, pediatric, and adult medical; diagnostic (laboratory and X-ray); pharmacy; dental; behavioral health; drug and alcohol counseling; preventive health; and wellness programs (Table 1). In addition, several clinics provided itinerant primary and preventive care services to residents in remote villages; others rented office space to visiting specialists for scheduled specialty clinics. Each participating clinic also offered either on-call or on-site after-hours care.

Table 1: Characteristics of clinics that participated in the FESC demonstration, by certification date. (Source: University of Alaska Anchorage, 2010)

Top ↑

The first clinic obtained certification and began receiving Medicare payments for extended stay services in April 2013, triggering the start of the FESC Demonstration. The demonstration lasted for 36 months (through April 2013), regardless of when a clinic received certification and began receiving enhanced payments. The five clinics that participated in the demonstration were located in remote areas or islands in two states: four in Alaska and one in Washington [foot note 1]. Four clinics were certified as FQHCs and received supplemental funding under Section 330 of the Public Health Service Act, two of which were also tribally affiliated clinics and received funding under the Indian Health Service (IHS). The nearest hospitals to these clinics were generally distant, so air (and, in some cases, boat) transport is the only option for emergency transport services, and severe weather could prevent transfer for days [3,4].

Finally, at the time of FESC certification, all participating clinics had a minimum of five providers able to independently treat patients and prescribe medications, at least one of whom was a physician. They also had a minimum of four registered nurses, plus a wide range of other medical personnel, including licensed practical nurses, medical assistants, pharmacists, lab technicians, and emergency medical technicians.

(Foot Note 1: Inter-Island Medical Center in Friday Harbor, Washington, converted to a CAH in November 2012 and thus became ineligible to receive FESC payments five months before the end of the demonstration. A sixth site located in Broadus, Montana, applied for and was accepted into the demonstration, but later withdrew, citing the high cost of becoming a Medicare-certified extended stay facility)

Data Sources

We relied on two sources of data for this study. For a descriptive analysis of the number and type of extended stays, we used self-reported patient encounter forms from March 2010 through April 2013. Clinical staff from participating sites completed the encounter forms and sent them to CMS or the fiscal intermediary to verify that the beneficiary met the medical requirements for receiving extended stay services. Encounter forms provided more clinical information about extended stays than could be obtained from Medicare claims. For example, encounter forms included information on the chief complaint at extended stay admission, the time and duration of the stay, the type of admission (monitoring and observation versus stabilization pending transfer), the clinical outcome of the extended stay, and the discharge diagnosis. The encounter forms also included an assessment by the clinician of the impact of an extended stay on subsequent care (for example, averting an emergency transfer and hospitalization). We obtained 166 encounter forms for Medicare extended stay beneficiaries from March 2010 through April 2013.

To estimate the marginal impact of the demonstration on the use and cost of hospital inpatient and emergency transfer services, we first extracted outpatient claims for all beneficiaries who received care (including those with a visit that did not result in an extended stay) at demonstration and comparison group clinics from January 2008 to December 2012, approximately three years before all of the participating clinics became certified as extended stay facilities (2008 –2010) and two years after (2011 and 2012). Using this reference file, we extracted all inpatient, outpatient, and professional claims for these beneficiaries during the same five-year period. The file contained 33,810 claims for outpatient visits at demonstration and comparison clinics, representing 3,919 unique beneficiaries across all clinics and years in the study (2,872 beneficiaries from demonstration clinics and 1,047 from comparison clinics).

Comparison Clinics

To identify the comparison clinics, we began with a list of 11 clinics in Alaska that met the distance or accessibility requirements of the demonstration. To further match on the participating clinics before the demonstration, we shortened the list to those clinics that met the following criteria: (1) designated as a RHC or FQHC; (2) provided comprehensive primary care services, including laboratory testing and radiology services; (3) staffed by physicians and advanced practice nurses or advanced practice nurses working with a visiting physician; (4) served a similar number of Medicare beneficiaries; and (5) did not operate an emergency or urgent care unit. Only two of the 11 clinics met all of these conditions. The other nine clinics were smaller, provided only limited services, and operated with fewer and less well-trained staff.

We then compared the demographic characteristics and chronic condition prevalence among beneficiaries with at least one visit at a demonstration clinic with those seeking care at either of the two comparison clinics in 2010, using information from the Master Beneficiary Summary File and the Medicare enrollment database. Beneficiaries who received care at the comparison clinics were younger (less than 75 years old) and less likely to be American Indian or Alaska Native than those at the demonstration clinics. However, the beneficiary populations at the two types of facilities had similar chronic condition prevalence, suggesting that although the beneficiaries differ demographically, they share similar clinical characteristics. Like all frontier clinics, the comparison clinics were likely required to treat patients needing emergency care. However, without the staffing, equipment, and infrastructure to appropriately treat patients on an extended stay basis, practice would have required them to transfer these patients as soon as weather and conditions permitted.

Defining Episodes of Care

We defined the first episode of care for a given beneficiary as the seven day period following the end of the first outpatient visit (including the extended stay) and included all outpatient visits, professional services, emergency transfers, and inpatient services occurring within that seven day period as part of the initial episode. We used seven day episodes to capture all follow-up services, including transfers and inpatient services, which might be associated with an ambulatory visit at a frontier clinic. It is important to consider emergency transfer and inpatient services rendered over several days because seriously ill patients can often be transferred twice, once from the frontier clinic to a community hospital and then from the community hospital to a tertiary hospital for specialized care.

A second episode was triggered by another outpatient visit at a demonstration or comparison group clinic after a “clean” period of 14 days with no outpatient claims following the end of the previous seven day episode. We required a 14-day period with no outpatient services before starting a new episode in an effort to identify discrete episodes of care and to increase the chances that the follow-up care received during the episode was related to the index visit. In the end, we identified a total of 20,379 seven-day post-outpatient visit episodes among 3,919 beneficiaries during the study period. Medicare beneficiaries had on average 5.2 episodes during the five-year study period, approximately one episode per year.

Model and Estimation Strategy

Finally, we conducted a difference-in-differences regression analysis to estimate the impact of the demonstration on the probability of having a transfer or hospitalization within seven days of a clinic visit. Our analytic approach relies on the assumption that the change in the rate of emergency transfers and hospitalizations would, in the absence of the demonstration, have been the same for the demonstration clinics as it was in the comparison clinics in 2011 and 2012. To control for potential bias caused by differences in patient mix between demonstration and comparison clinics, we included a set of beneficiary-level demographic and health characteristics (age, gender, race/ethnicity, dual eligibility, disability, and mortality) likely to be associated with being transferred and hospitalized. Differences in the change in transfer and hospitalization rates among demonstration clinics after versus before the demonstration relative to the change among the comparison clinics over the same period are ascribed to the effect of the demonstration. To obtain the standard errors on the interaction terms, we estimated the model using a linear probability regression and corroborated the results with a logit model using the inteff command in Stata [5].

The credibility of the difference-in-difference results rests on the appropriateness of the comparison group. Although this cannot be definitively established, our investigation showed that comparison clinics are mostly well matched to demonstration clinics on both facility (including staffing and service mix) and patient characteristics. We also supplemented our claims-based regression analysis with a descriptive analysis of the information available from the extended stay encounter forms (corroborated by interviews with clinic staff) and found that the two data sources provided similar results.

 

Results

 

Types of Extended Stays

Based on our descriptive analysis of encounter data from FESC clinics, 65.1% of the 166 extended stays that were billed for enhanced reimbursement from Medicare under the demonstration (and for which we received an encounter form) were admitted for monitoring and observation (Table 2). The attending clinicians (specifically, medical doctors, advanced practice nurses, and registered nurses) reported that, with appropriate diagnostic information and clinical expertise, these beneficiaries could likely be sent home after several hours of monitoring and observation and possibly referred to nonemergency follow-up care in the community. The remaining 32.5% of all extended stay cases during the demonstration were clinically determined at admission to require an emergency level of care. The attending clinicians reported that these beneficiaries needed to be medically stabilized and transported by air to a hospital as soon as weather conditions permitted. (The encounter forms for the remaining 2.4% extended stays did not specify the type of admission).

Table 2: Characteristics of Extended Stays Paid under the FESC Demonstration. (Source: Extended stay patient encounter forms submitted to CMS by participating clinics for services rendered from March 2010 to April 2013. Number of admissions might not sum to total admissions due to missing information. Percentages are based on the number of encounter forms with reported information)

Top ↑

Timing and Duration of Extended Stays

As a condition of participation in the demonstration, clinics were required to maintain on-call physicians or non-physician clinicians 24 hours a day, seven days a week, who could be on site within 30 minutes of a patient’s after-hours arrival and that a clinician is on site throughout each extended stay episode. Information reported on the encounter form indicates that 59.0% of all extended stay cases billed for enhanced reimbursement from Medicare under the demonstration was admitted to the clinic outside of regular business hours and 41.0% were admitted during regular working hours (Table 2). Although 44.0% of all extended stays lasted fewer than 8 hours, 39.1% of all beneficiaries requiring extended stay care remained in the clinic for more than 12 hours and 12.1% stayed for 24 hours or longer.

Conditions Requiring Extended Stay Care

The most common type of illnesses (in descending order) requiring extended stay services were cardiovascular conditions (68%), metabolic conditions (49%), orthopedic problems (36%), respiratory conditions (29%), gastroenterological conditions (19%), and genitourinary conditions (19%). A review of patient encounter forms indicates that beneficiaries with cardiovascular conditions were more likely to be transferred to a hospital at the end of their extended stay; beneficiaries with respiratory and genitourinary conditions were more likely to be discharged home after their extended stay at the community clinic, with or without follow-up care.

Discharge Disposition of Extended Stay Beneficiaries

A review of patient encounter data also shows that 48.2% of all Medicare-eligible beneficiaries admitted for extended stay services at frontier clinics were discharged home at the end of their stays (Table 2). Five of the extended stay beneficiaries discharged home was referred to nonemergency follow-up care and another beneficiary recovered in the clinic while waiting for emergency transport, eliminating the need for transport. Another 48.2% of all extended stay beneficiaries, including 27 admitted for monitoring and observation, were transferred to a hospital following their extended stays. Clinicians reported that beneficiaries admitted for monitoring and observation but then transferred to an acute care hospital at the end of their stay either failed to improve as expected or their conditions deteriorated during their extended stays.

According to the encounter forms, none of the extended stay beneficiaries died while receiving extended stay services. However, an analysis of Medicare claims and enrollment data found that five of the beneficiaries who received extended stay care in a participating clinic died within 14 days of their extended stay, four of whom had been transferred and admitted to a hospital at the end of their stay.

Number and Frequency of Extended Stays

Based on an analysis of Medicare claims data, only 1% of all outpatient visits among Medicare beneficiaries at FESC clinics from January 2010 through December 2012 were for extended stay services. As shown in table 3, there were only 159 claims for extended stay services paid by Medicare across the five participating clinics during the demonstration, out of a total of 16,575 outpatient visits (slightly less than the number of extended stay encounter forms). A longer claims run-out period might reveal several more extended stays eligible for Medicare payment under the demonstration.

Table 3: Proportion of total visits for extended stay services among Medicare beneficiaries. (Source: Medicare claims for beneficiaries treated at demonstration clinics from January 2010 through December 2012. The data were extracted in April 2013. Because encounter forms did not always result in a final action claim and were available sooner than claims, the number of extended stay encounter forms exceeded the number of paid claims for extended stay services in this study)

Top ↑

Impact of Extended Stays on Use and Cost of Services

The results of the multivariate claims analysis indicate that the provision of extended stay services under the demonstration resulted in a 21.4% (p = 0.15) reduction in the probability of being transferred within seven days of a clinic visit, a decline in the seven-day transfer rate from 4.0 to 3.2 percent of all clinic visits (Table 4). The probability of being admitted to a hospital for inpatient care within seven days of a clinic visit also declined under the demonstration, by 23.9% (p = 0.12), from 3.7 to 2.7%. The estimated decline in the seven-day emergency transfer and hospitalization rates is equivalent to 26 avoided emergency transfers and hospitalizations per year associated with the provision of extended stay services. The results of the multivariate model are consistent with the absolute number of avoided transfers and hospitalizations per year as reported by the discharging clinicians on the encounter forms in 2011 and 2012.

Table 4: Impact of FESC Demonstration on 7-day emergency transfer and hospitalization rates. (Source: Medicare claims and enrollment data for beneficiaries treated at demonstration and comparison clinics from January 2008 through December 2012. The data were extracted in April 2013. The estimated percent change in 7-day emergency transfer rate was significant at the p = 0.15 level. The estimated percent change in the 7-day hospitalization rate was significant at the p = 0.12 level)

 

By reducing the number of beneficiaries who required transfer for inpatient care, the demonstration resulted in a net reduction in Medicare spending for medical services. To calculate federal savings, we first used claims data from demonstration and non-demonstration clinics to estimate average Medicare payments for emergency transfer and inpatient services, in 2012 dollars, over the seven-day period following a clinic visit. Actual average payment for emergency transfer services received by Medicare beneficiaries over the seven-day episode following a clinic visit was $10,983 and, for inpatient services, $16,304. We then multiplied the average payment amounts by the estimated number of transfers and hospitalizations avoided annually to calculate total Medicare savings per year. Total Medicare savings per year from averted transfers were an estimated $285,558 and, from averted hospitalizations, $423,904 (Table 5).

Table 5: Estimated annual Medicare savings from the FESC Demonstration. (Source: Medicare claims and enrollment data, calendar years 2008 – 2012, extracted in April 2013. We estimated the impact of the demonstration on emergency transfer and hospitalization rates using five years of Medicare claims [three before the demonstration and two after]. By taking the product of the estimated averted transfer and hospitalization rates and the number of extended stay counters per year, we express Medicare expenditures and savings on an annual basis)

 

To calculate net Medicare savings, we added total payments for extended stays ($101,680) and subtracted the estimated amounts that would have been paid in the absence of the demonstration for clinic visits ($8,800), emergency transfers ($285,558), and hospital stays ($423,904). This resulted in total estimated savings to Medicare of $616,582, or $7,707 per extended stay. Not included in our estimates, beneficiaries also benefited from approximately $90,000 in foregone copayments and deductibles for averted services per year, or slightly more than $1,100 per extended stay. Other averted out-of-pocket costs to patients and their families associated with a transfer and hospital stay in a distant city are not included in these estimates.

 

Limitations

 

The main challenge of this study was obtaining robust estimates of the impact of the demonstration on emergency transfer and hospitalization rates and, hence, the level of Medicare savings, if any, associated with the program. First, the number of Medicare beneficiaries living in frontier communities in Alaska is small and the proportion of elderly residents requiring extended stay care is even smaller, so observing statistically significant effects is difficult. Second, the lack of a billing code for extended stay services in ambulatory care settings under the traditional Medicare fee-for-service system makes it difficult to determine the number of beneficiaries receiving extended stay care who would have been transferred without the additional capacity and enhanced payments provided by the demonstration. Third, the clinics that chose to participate in the demonstration generally had more staffing and provided a higher level of services than other clinics in isolated rural areas in Alaska. The limited number of frontier clinics in Alaska and the unique characteristics of those that participated in the demonstration created challenges for identifying a valid comparison group of outpatient facilities for this study. Additional research in other remote and frontier regions of the United States and other countries is needed to corroborate (or challenge) these findings.

Top ↑

Discussion

 

The program administrators and frontline staff we interviewed for this study emphasized that the ability to provide high-level extended stay services improves the experience of care for patients seeking emergency treatment in frontier communities, particularly those needing medical attention during non regular hours of operation and those requiring monitoring and observation services only. In the absence of certified extended stay services, many beneficiaries with potentially serious illness or injury delay seeking care and, after they seek care, are transferred as soon as possible by single-or twin-engine, propeller-driven commuter plane to a hospital for inpatient care hundreds of miles away, without the benefit of their families and other social and cultural supports.

Many elderly residents in frontier regions are unaccustomed to traveling from their villages for medical care and often view a hospital transfer as being sent from their community to die. With the availability of extended stay services in the community, beneficiaries can be observed or stabilized closer to home, while local health care professionals determine whether monitoring and observation services are sufficient or emergency transport and hospitalization are required. Many beneficiaries receiving extended stay services for monitoring and observation purposes, and even a small proportion of those receiving stabilization services until conditions allow transfer, are able to go home after several hours of observation, sometimes with a referral for nonemergency follow-up care in the community.

Our analysis shows that, even with the enhanced payment rates offered under the demonstration, the provision of extended stay services in remote geographic areas lowers the per capita cost of care to Medicare for beneficiaries seeking treatment for potentially serious illness or injury. However, it is important to note that the FESC Demonstration was not budget neutral to the federal government. The savings reported in this study do not take into account the $1.5 million in annual cooperative agreement funding from HRSA to help participating clinics meet and maintain the federal extended stay certification requirements. For relatively small clinics with limited resources, complying with the certification requirements to ensure the safety of beneficiaries receiving extended stay care in ambulatory health care facilities and to ensure that medically appropriate and quality care was provided would be difficult without additional sources of external funding.

Funding Source

 

This study was conducted under funding from the Centers for Medicare & Medicaid Services, Contract No. GS-10F-0050L/HHSM-500-2011-00119G.

Disclosures

 

The authors have no conflicts of interest to disclose.

Acknowledgements

 

The authors are grateful to Negussie Tilahun and William Clark, program officers at the Centers for Medicare & Medicaid services, for providing guidance and review on this study. The views expressed in this study are solely those of the authors and do not necessarily reflect the views or opinions of CMS or its staff.

Top ↑

References

 

  1. Section 434 of the Medicare Prescription Drug, Improvement, and Modernization Act of 2003, United States Congress. United States Congress; 2003; Public Law 108–173.
  2. Frontier Extended Stay Clinic Demonstration. Department of Health and Human Services. Centers for Medicare & Medicaid Services (CMS), HHS. Agency: Centers for Medicare & Medicaid Services. Federal Register. 2006;71(165):50428–29.
  3. University of Alaska Anchorage. Frontier Extended Stay Clinic Project: Report on 48 Months of Data, Alaska Center for Rural Health and Institute of Social and Economic Research, College of Business and Public Policy. Anchorage, AK: Alaska Frontier Extended Stay Consortium. 2010.
  4. University of Alaska Anchorage. Frontier Extended Stay Clinic Demonstration: Volume 1, Institute of Social and Economic Research, College of Business and Public Policy. Anchorage, AK: Alaska Frontier Extended Stay Consortium. 2011.
  5. Norton EC, Wang H, Chunrong A. Computing Interaction Effects and Standard Errors in Logit and Probit Models. The Stata Journal. 2004;4(2):154–67.

Top ↑

Copyright: © 2017 Gilman B, et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.