Acute bacterial skin and skin structure infections (ABSSSI) are among the most common infections observed in the emergency department.1,2 More than 15% of patients who present to the emergency department with an ABSSSI are admitted to the hospital, and the average hospitalization costs per patient range from $6300 to $13,000, with multiday room and board expenses comprising 50% of the total costs of care.3-5 Although patients with ABSSSI have historically received care in the hospital, many of these patients have few or no active comorbidities and can be effectively and safely managed in the outpatient setting.3
Data suggest that as many as 40% of patients with ABSSSI are admitted to the hospital solely for the administration of intravenous (IV) antibiotics.6 Because inpatient ABSSSI care is approximately 2 to 4 times more costly than outpatient care,7 it is imperative that clinicians identify appropriate patients for outpatient therapy at emergency department presentation and develop a treatment plan that can safely and effectively shift care from the inpatient to the outpatient setting among hospitalized patients.
Omadacycline is an oral and IV once-daily aminomethylcycline antibiotic with broad-spectrum activity, including for methicillin-resistant Staphylococcus aureus, that was recently approved by the US Food and Drug Administration for the treatment of patients with ABSSSI.8,9 To date, omadacycline has been shown to be noninferior to linezolid for patients with ABSSSI in 2 clinical trials.10,11 The first study, Omadacycline in Acute Skin and Skin Structure Infections Study (OASIS), was an IV-to-oral treatment-switch trial demonstrating that patients in both treatment arms could be safely and effectively transitioned from IV treatment in the inpatient setting to outpatient treatment with an oral formulation.10 The second study, OASIS-2, was an oral treatment–only trial, and therapy in both arms was initiated in the outpatient setting.11
Recognizing the financial burden that the treatment of patients with ABSSSI places on the healthcare system, we developed conceptual healthcare decision models to examine the potential cost-savings of treatment with antibiotics that have bioequivalent IV-oral formulations, such as omadacycline, because of avoidable hospitalizations and hospital stay reductions among patients who present to the hospital from the emergency department for the treatment of ABSSSI.
We first examined the potential cost-savings associated with treating adults with ABSSSI who had no or only limited comorbidities and no life-threatening conditions with omadacycline in the outpatient setting compared with managing patients in the hospital with IV vancomycin—the hospital avoidance, or the Stay Home model. We also developed models to determine the reduction in hospital stay required with omadacycline therapy to confer cost-savings compared with the current standard of inpatient care for patients with ABSSSI who had ≥2 comorbidities and no life-threatening conditions—the hospital discharge, or the Go Home model.
Although no clinical studies have compared omadacycline and vancomycin head-to-head, vancomycin was selected for this study as the comparator rather than linezolid, because vancomycin is the most widely prescribed antibiotic for hospitalized adults with ABSSSI.12 Recent hospital prescribing data indicate that vancomycin prescriptions comprised 26% of all first-course antibiotic prescriptions for ABSSSI, which is more than 20 times the volume of linezolid prescriptions.13 Furthermore, our model assumptions for omadacycline were based on several studies across various infection types, which demonstrated that antibiotics with IV and oral formulations can shorten a hospital stay compared with treatment with antibiotics with IV-only formulations, such as vancomycin.14,15
Methods
The assumptions used in the Stay Home and the Go Home models are described in the Supplemental Table (see Appendix). The Go Home model has the additional assumption that all patients receiving IV vancomycin or IV omadacycline were initially treated in the inpatient setting and were discharged to the outpatient setting using IV vancomycin or a generic oral antibiotic to complete treatment.
Because studies comparing omadacycline with vancomycin have not been performed, treatment nonresponse and rehospitalization rates were excluded from the models.10,11 Patients with life-threatening conditions were also excluded, because omadacycline has not yet been studied in that population. Here, the models solely focused on ABSSSI severity and site of care.
Hospital Avoidance: Stay Home Model
The first model compared the costs of inpatient IV vancomycin versus outpatient omadacycline for the treatment of patients with ABSSSI who had few or no comorbidities (Charlson Comorbidity Index [CCI] score of 0 or 1) and no life-threatening conditions among patients with ABSSSI presenting to the emergency department (Figure 1). All patients who received vancomycin were assumed to be admitted as inpatients.
Three omadacycline treatment scenarios were considered, including (1) patients were sent home with oral omadacycline treatment, (2) patients received an IV loading dose of omadacycline and were sent home from the emergency department with oral omadacycline, and (3) patients were treated under observation status with IV omadacycline and were sent home from the emergency department with oral omadacycline.
Data from a retrospective, observational study that used hospital discharge data from the Premier Perspective hospital database were used to determine the average costs associated with hospital inpatient treatment of patients with ABSSSI receiving IV vancomycin.16 Patients were included in this study if they had a medical encounter with a primary diagnosis of ABSSSI between January 1, 2012, and December 30, 2012, and IV vancomycin use on day 1 or 2 of hospital admission. The presence of comorbid conditions was based on the CCI score noted during the qualifying admission. Patients were included in this analysis if they had a CCI score of 0 or 1 without any life-threatening conditions. The severity of the condition based on the CCI score is equally applicable to the vancomycin and omadacycline groups.
In the omadacycline arm, the daily costs of omadacycline varied from $0 to $1000 (US dollars), and all patients who received omadacycline received the drug orally to complete a 7-day course of therapy. We considered the daily acquisition costs to be the same for the IV and oral formulations. The costs associated with observation care for omadacycline were derived from Medicare national limitation amounts,17 which were also used as a proxy for the direct hospital costs of omadacycline IV therapy administration.16
The key output was the per-patient cost differences between treatment with vancomycin and with omadacycline. As part of this analysis, we estimated the proportion of patients who initially receive oral omadacycline in the outpatient setting and can be subsequently admitted into the hospital while still conferring cost-savings compared with vancomycin therapy in the inpatient setting. For this, the acquisition cost of IV and oral omadacycline treatment varied from $0 to $1000 daily to estimate the upper end of daily omadacycline acquisition cost that still conferred cost-savings using outpatient omadacycline treatment with different hospital admission rates.
The admission rates were fixed at 0%, 10%, and 20%. We assumed that patients who received omadacycline had 7 days of oral omadacycline treatment, did not respond to this treatment, and were subsequently admitted to the hospital. The hospitalization costs for these patients were assumed to be identical to the inpatient treatment cost of patients with ABSSSI who received IV vancomycin in the Premier Perspective research database study.16
For these analyses, we did not consider any reimbursement penalties that may occur. However, because patients were never admitted with omadacycline therapy in this hypothetical modeling exercise, there was no risk for readmission.
Hospital Discharge: Go Home Models
The second model compared the costs associated with inpatient current standard of care (ie, the reference condition) versus inpatient omadacycline therapy for the treatment of patients with ABSSSI with ≥2 comorbidities (CCI score ≥2) and no life-threatening conditions from a clinical (ie, hospital) perspective (Figure 2). For hospitalized patients, 4 categories of disease severity were examined, including a CCI score of 2 with and without systemic symptoms, and a CCI score of ≥3 with and without systemic symptoms.
The third model was identical to model 2 and included the cost of care for subsequent outpatient care. This model presented the costs for all treatment arms from the hospital and payer perspectives. All patients in this model were assumed to have received outpatient treatment after initial hospitalization to complete a 10-day course of therapy. Cost-savings could be diminished from a reimbursement perspective because of readmission penalties. However, the model was developed to consider only costs, not reimbursement. In addition, because no comparator data are available for readmission, we did not consider reimbursement in this model.
In the standard-of-care arm (ie, the reference case), the patients were assumed to have completed therapy after hospital discharge either with IV vancomycin in a hospital-based infusion suite (50%) or with a generic oral antibiotic (50%) in the outpatient setting. In the omadacycline arm, all patients were assumed to have received oral omadacycline after discharge to complete the 10-day course of therapy. A longer course of treatment (10 days vs 7 days) was assumed in the Go Home models to allow for the differences in severity (ie, higher CCI levels) than in the Stay Home model.
Data from a retrospective, observational study that used hospital discharge information from the Premier Perspective hospital database were used to determine the median hospital length of stay associated with the inpatient treatment of patients with ABSSSI in the standard-of-care arm.3 Patients from this analysis were included in the hospital avoidance study if they were (1) admitted, (2) had a CCI score of ≥2, and (3) had no life-threatening conditions. Different hospital lengths of stay were associated with levels of the CCI and whether patients had systemic symptoms.
To calculate the cost of inpatient treatment from the median length of stay, each hospital day was estimated to cost $1346 in 2016 US dollars (the amount in the Premier database analysis inflated to July 2016 using the US Medical Care Consumer Price Index).5 The costs for healthcare utilization inputs are summarized in the Supplemental Table. For the model from the payer’s perspective, the patients completed a 10-day course of therapy with either IV vancomycin in a hospital-based clinic (50%) or a generic oral antibiotic (50%) in an outpatient setting in the standard-of-care arm. We used the wholesale acquisition cost of linezolid 600 mg orally twice daily, double-strength trimethoprim plus sulfamethoxazole 1 tablet orally twice daily, and vancomycin 1000 mg IV twice daily for drug cost.18 For vancomycin IV administration, we used the reimbursement rate for Current Procedural Terminology code 96365 as the cost associated with 1-hour infusions of vancomycin.16
All the costs associated with standard-of-care inpatient treatment were also assumed to be the same for patients receiving omadacycline. All patients were assumed to receive a 10-day course of omadacycline, which was split between the inpatient (IV administration) and outpatient (oral administration) setting. The daily costs for omadacycline varied as described below.
The key output was the per-patient cost differences between treatment with standard-of-care inpatient treatment and with omadacycline. Oral antibiotic dosing formulation has been shown to shorten hospital stays across several disease states, including skin and soft-tissue infections.14
We determined the impact of 1- and 2-day hospital length of stay reductions with omadacycline compared with the current inpatient standard of care on the overall cost of care from the hospital and payer’s perspectives. Regardless of the hospital length of stay reduction put into the model, the minimum length of stay specified for omadacycline in the model was 3 days. The daily cost of omadacycline varied between $0 and $1000 to characterize the upper end of daily omadacycline acquisition cost that still conferred cost-savings with 1- to 2-day hospital stay reductions compared with inpatient standard-of-care treatment from the hospital and payer perspectives.
Results
Stay Home Model
The average cost associated with the hospital inpatient treatment of patients with ABSSSI who were receiving IV vancomycin was $6511.89 per course of treatment. Switching an individual patient from inpatient IV vancomycin to outpatient omadacycline at the beginning of treatment was estimated to result in cost-savings, the amount of which depended on receipt of a loading dose, the use of an observation unit, and an inputted daily cost of omadacycline. The most cost-minimizing regimen was discharge to home from the emergency department with oral omadacycline. The administration of a loading dose of IV omadacycline, followed by discharge to home, was slightly more expensive ($98 per infusion) as a result of the cost of infusion. Keeping the patient under observation for less than 48 hours (<2 midnights) was more expensive ($885 per patient) than the other omadacycline regimens because of the added cost of using the observation facility and multiple infusions.
When admission rates after nonresponse to oral treatment were set at 0, the upper bounds of the cost-minimizing daily omadacycline drug acquisition costs ranged from $705 to $891, depending on the receipt of an IV loading dose in the emergency department and the use of an observation unit. The upper ends of the cost-minimization daily omadacycline acquisition costs ranged from $625 to $798 when the admission rates were set at 10%. At 20% admission rates, the maximum cost-minimizing daily omadacycline acquisition costs subsequently decreased by nearly $100 for each omadacycline treatment scenario and ranged from $525 to $708. The actual admission rates in the oral medication–only trial were <1%.
Assuming a net acquisition cost of $375, near the midpoint of the sensitivity range, pegged to the capability of omadacycline treatment to save 1 day in the hospital on average in the Go Home scenario, the Stay Home model projected considerable cost-savings per patient. As shown in the Table, per-patient cost reductions of >$2000 were projected in this scenario.
Go Home Models
The upper bounds of cost-minimizing daily omadacycline drug acquisition costs with hospital length of stay reductions of 1 to 2 days from the hospital perspective are shown in Figure 3.
From the hospital perspective, omadacycline would be associated with cost-savings if the omadacycline cost was ≤$383 daily, with a 1-day hospital length of stay reduction, and ≤$936 daily with a 2-day hospital length of stay reduction. From a payer perspective, omadacycline would be associated with cost-savings if the omadacycline cost was ≤$173 daily with a 1-day length of stay reduction, and <$317 daily with a 2-day length of stay reduction.
Discussion
Because hospital reimbursement and antimicrobial stewardship programs are increasingly tied to quality, efficiency, and cost of care,19,20 this study developed conceptual healthcare decision models to assess the potential cost-saving opportunities with antibiotics that have bioequivalent IV-to-oral formulations, such as omadacycline, compared with the current standard of care for the treatment of patients with ABSSSI. We opted to evaluate patients who presented to the hospital for the treatment of ABSSSI, because studies have demonstrated that more than 75% of these patients present to the emergency department for initial treatment.3,5
The hospital avoidance (Stay Home) model considered managing patients in the outpatient setting with oral omadacycline compared with inpatient treatment with IV vancomycin. Despite the availability of newer agents, vancomycin is used in the overwhelming majority of hospitalized adults with ABSSSI.5 More than 60% of admitted patients with ABSSSI who receive vancomycin have 0 or 1 comorbid condition and limited systemic symptoms.5 Data show that these patients have an extremely low mortality risk (0.08%) and can be effectively and safely managed in the outpatient setting.3
The results of the OASIS-2 study indicate that it may be possible to treat some patients who have ABSSSI and no life-threatening conditions completely in the outpatient setting.9 No patients were hospitalized for skin infections at the start of the OASIS-2 study; however, hospitalization for worsening of ABSSSI from the first dose through the follow-up visit occurred in 2 (0.6%) patients in the omadacycline group and 1 (0.3%) patient in the linezolid group. Therefore, 99.4% of patients were treated with omadacycline as outpatients by the end of the study.9
In this model, cost-savings for the payer for patients with ABSSSI may be realized with outpatient oral omadacycline compared with inpatient IV vancomycin. This conclusion holds when the costs of infusing an initial IV dose of omadacycline and an observation stay are added to the outpatient omadacycline drug acquisition cost.
For various omadacycline outpatient treatment scenarios considered, the models estimated the upper bounds of daily omadacycline costs that still conferred cost-savings to be $725 to $900. Although this hospital avoidance approach is intuitive, one of the major reservations among clinicians is the perception that patients will not respond to treatment in the outpatient setting and will return to the hospital for subsequent care, negating all the potential cost-savings associated with outpatient treatment.
Recognizing this, we considered a few different scenarios that allowed for a proportion of patients to be readmitted after a 7-day course of omadacycline in the outpatient setting and incur the full costs associated with inpatient management. Even with 20% subsequent admissions among omadacycline outpatients, the daily cost with omadacycline could still be upwards of $600 to $800 daily, and the total patient cost-savings could still be realized.
The second set of models (Go Home) considered 1- or 2-day hospital stay reductions with omadacycline compared with inpatient treatment with IV vancomycin among patients who had ≥2 comorbid conditions, with and without systemic signs of infections.
The rationale for hospital stay reductions with omadacycline in this conceptual model is based on the OASIS-1 trial and previous studies among patients with various infection types, including skin and soft-tissue infections, which demonstrated that antibiotics with IV and oral formulations, such as omadacycline, can effectively facilitate earlier hospital discharge compared with agents with only IV formulations, such as vancomycin.10
In fact, any antibiotic, regardless of the type of antibiotic, that has IV and oral formulations should have the capability to help patients transition to oral dosing, provided that the same antibiotic treatment that is effective in the inpatient setting is continued after the IV-to-oral switch.21 It should be noted, however, that patients should be able to meet certain clinical criteria22 and continue maintenance treatment with the same antibiotic that has demonstrated effectiveness during the inpatient stay.23
In this model, the maximum cost-minimizing daily omadacycline cost from the hospital perspective varied between approximately $400 to approximately $1000, depending on the number of active comorbidities or systemic symptoms, and the reduction in hospital length of stay. Adding outpatient treatment options, which are the liability of the payer, significantly reduced the maximum cost-minimizing daily omadacycline acquisition cost compared with the hospital perspective. However, the upper threshold of the cost-minimizing daily omadacycline acquisition cost from the payer perspective was still substantial, especially with a 2-day hospital length of stay reduction.
Limitations
Several limitations should be noted when evaluating these findings. This modeling exercise was conducted to understand the potential cost-savings opportunities with antibiotics that have bioequivalent IV-to-oral formulations, such as omadacycline, compared with current ABSSSI inpatient treatment practices among patients who present to the hospital for initial care. Several assumptions were made for this study, and the results should be interpreted with caution until these assumptions are validated by appropriate clinical studies; a comparative, prospective cost-effectiveness study with economic and healthcare utilization end points is being considered to validate these findings.
We were conservative with model assumptions and limited data inputs to the critical healthcare resource components. For the standard-of-care arm, the model inputs were largely derived from previous studies of the Premier Perspective Hospital Database, which contains coding and billing information for approximately 50 million admissions from approximately 500 US acute-care hospitals.5,24 Despite its size, it may not be generalizable to all other US hospitals, and hospitals should rely on institution-specific costs when deriving the costs associated with inpatient treatment of patients with ABSSSI at their institutions.
Similarly, we relied on Medicare national limitation payment amounts to represent the observation and drug administration costs, which may not be reflective of the cost associated with these healthcare resource components at a given facility.17 We purposefully excluded parameters such as peripherally inserted central catheter (PICC) line placement and subsequent PICC-line infections, outpatient laboratory and vancomycin assay costs, and physician and nursing time in outpatient antibiotic infusion suites, to provide a conservative estimate of cost-savings opportunities with omadacycline.
We also did not include treatment response rates, adverse events, or rehospitalization occurrence, because phase 3 comparator ABSSSI studies have not been conducted between omadacycline and vancomycin.10
Finally, the proposed potential cost-savings opportunity with omadacycline compared with the current standard of care can be applied to any antibiotic that has IV-to-oral formulation capabilities. The use of omadacycline compared with these other antibiotics should be guided by the individual patient and the potential risks versus benefits.
Conclusion
We are living in a rapidly evolving US healthcare environment, in which hospital reimbursement is increasingly tied to quality and efficiency of care. In the recent value-based hospital payment model, it is important to develop new therapies and treatment approaches that maintain or improve outcomes at the lowest overall cost. The concept of high-quality care at the lowest cost is congruent with the goals of antibiotic stewardship programs, which have a dual responsibility to optimize clinical outcomes and minimize healthcare utilization and costs without adversely affecting quality of care. Given that IV-to-oral antibiotics have been shown to reduce hospitalization length of stay in various studies, the findings from these economic models are in keeping with previous research and suggest that targeted prescribing of antibiotics with bioequivalent IV-to-oral formulations, such as omadacycline for the treatment of select patient populations with ABSSSI, could result in cost-savings compared with the current standard of care.
Acknowledgments
Samantha Scott, PhD, of Innovative Strategic Communications, LLC, provided assistance in preparing and editing the manuscript.
Funding Source
This study was funded by Paratek Pharmaceuticals, Inc.
Author Disclosure Statement
Dr LaPensee is an employee of, and Dr Lodise is a consultant and scientific advisor to Paratek Pharmaceuticals.
Dr LaPensee is Director, Health Economics and Outcomes Research, Paratek Pharmaceuticals, King of Prussia, PA; Dr Lodise is Professor, Albany College of Pharmacy and the Health Sciences, NY.References
- Centers for Disease Control and Prevention. National Hospital Ambulatory Medical Care Survey: 2015 emergency department summary tables. www.cdc.gov/nchs/data/nhamcs/web_tables/2015_ed_web_tables.pdf. Accessed October 25, 2017.
- Moran GJ, Krishnadasan A, Gorwitz RJ, et al; for the EMERGEncy ID Net study group. Methicillin-resistant S. aureus infections among patients in the emergency department. N Engl J Med. 2006;355:666-674.
- Lodise TP, Fan W, Sulham KA. Hospital admission patterns in adult patients with skin and soft tissue infections: identification of potentially avoidable hospital admissions through a retrospective database analysis. Hosp Pract (1995). 2015;43:137-143.
- Sulham K, LaPensee K, Fan W, Lodise TP. Severity and costs of acute bacterial skin and skin structure infections by treatment setting: an application of the Eron classification to a real-world database. Value Health. 2014;17:A282.
- LaPensee K, Fan W. Economic burden of hospitalization with antibiotic treatment for ABSSSI in the US: an analysis of the Premier Hospital Database. Poster presented at the International Society for Pharmacoeconomics and Outcomes Research Annual International Meeting; June 2-6, 2012; Washington, DC.
- Talan DA, Salhi BA, Moran GJ, et al. Factors associated with decision to hospitalize emergency department patients with skin and soft tissue infection. West J Emerg Med. 2015;16:89-97.
- Khachatryan A, Ektare V, Xue M, et al. Reducing total health care costs by shifting to outpatient (OP) settings of care for the management of gram+ acute bacterial skin and skin structure infections (ABSSSI). Value Health. 2013;16:A203.
- Villano S, Steenbergen J, Loh E. Omadacycline: development of a novel aminomethylcycline antibiotic for treating drug-resistant bacterial infections. Future Microbiol. 2016;11:1421-1434.
- Paratek Pharmaceuticals. FDA approves Nuzyra. Press release. October 2, 2018. www.drugs.com/newdrugs/fda-approves-nuzyra-omadacycline-community-acquired-bacterial-pneumonia-acute-skin-skin-structure-4836.html. Accessed November 16, 2018.
- O’Riordan WA, Green S, Overcash JS, et al. A phase 3 randomized, doubleblind, multi-centre study to compare the safety and efficacy of oral and IV omadacycline to linezolid for treating adult subjects with ABSSSI (the OASIS study). Presented at the European Congress of Clinical Microbiology and Infectious Diseases; April 22-25, 2017; Vienna, Austria.
- O’Riordan WD, Cardenas C, Sirbu A, et al. A phase-3 randomized, double-blind, multi-centre study to compare the safety and efficacy of oral omadacycline to oral linezolid for treating adult subjects with ABSSSI (OASIS-2 study). Presented at the European Congress of Clinical Microbiology and Infectious Diseases; April 21-24, 2018; Madrid, Spain.
- Sulham K, Fan W, Werner R. Real-world prescribing patterns for the treatment of acute bacterial skin and skin structure infections in the United States: a retrospective database analysis. Value Health. 2015;18:A247.
- Data on file based on Paratek Pharmaceuticals’ 2017 analysis of the Vizient (formerly MedAssets) clinical database. www.vizientinc.com/Our-solutions/Clinical-Solutions/Clinical-Data-Base. Accessed October 25, 2017.
- Nathwani D, Eckmann C, Lawson W, et al. Pan-European early switch/early discharge opportunities exist for hospitalized patients with methicillin-resistant Staphylococcus aureus complicated skin and soft tissue infections. Clin Microbiol Infect. 2014;20:993-1000.
- Dryden M, Saeed K, Townsend R, et al. Antibiotic stewardship and early discharge from hospital: impact of a structured approach to antimicrobial management. J Antimicrob Chemother. 2012;67:2289-2296.
- Lodise TP, Fan W, Sulham KA. Economic impact of oritavancin for the treatment of acute bacterial skin and skin structure infections in the emergency department or observation setting: cost savings associated with avoidable hospitalizations. Clin Ther. 2016;38:136-148.
- American Medical Association. CodeManager. https://ocm.ama-assn.org/OCM/mainMenu.do. Accessed May 20, 2014. (Requires subscription to access.)
- Wolters Kluwer. Medi-Span Price Rx. https://pricerx.medispan.com. Accessed October 25, 2017. (Requires subscription to access.)
- Burwell SM. Setting value-based payment goals—HHS efforts to improve U.S. health care. N Engl J Med. 2015;372:897-899.
- Barlam TF, Cosgrove SE, Abbo LM, et al. Implementing an antibiotic stewardship program: guidelines by the Infectious Diseases Society of America and the Society for Healthcare Epidemiology of America. Clin Infect Dis. 2016;62:e51-e77.
- Rodriguez-Pardo D, Pigrau C, Campany D, et al. Effectiveness of sequential intravenous-to-oral antibiotic switch therapy in hospitalized patients with gram-positive infection: the SEQUENCE cohort study. Eur J Clin Microbiol Infect Dis. 2016;35:1269-1276.
- Akhloufi H, Hulscher M, Melles DC, et al. Development of operationalized intravenous to oral antibiotic switch criteria. J Antimicrob Chemother. 2017;72:543-546.
- Davis SL, Delgado G Jr, McKinnon PS. Pharmacoeconomic considerations associated with the use of intravenous-to-oral moxifloxacin for community-acquired pneumonia. Clin Infect Dis. 2005;41(suppl 2):S136-S143.
- Makadia R, Ryan PB. Transforming the Premier Perspective hospital database into the Observational Medical Outcomes Partnership (OMOP) common data model. EGEMS (Wash DC). 2014;2:1110.