Use of a Standardized Dalbavancin Approach to Facilitate Earlier Hospital Discharge for Vulnerable Patients Receiving Prolonged Inpatient Antibiotic Therapy

Abstract

Twenty-seven patients receiving prolonged inpatient antibiotic therapy for a serious bacterial infection received a single dose of dalbavancin 7–10 days before the planned end date to facilitate earlier hospital discharge. Eighty-one percent met criteria for clinical success, 7% experienced a potential adverse event, and 182 hospital days were averted.

INTRODUCTION

Vulnerable populations including people who inject drugs (PWID), people experiencing homelessness, and those with alcohol use disorder are at increased risk of developing serious bacterial infections such as infective endocarditis (IE) and osteomyelitis (OM). Injection drug use–related IE hospitalizations have been on the rise since 2000, with a doubling in prevalence in the last decade [1–3]. Many patients with these infections are not considered candidates for outpatient parenteral antimicrobial therapy (OPAT) and are therefore kept in the hospital to complete prolonged parenteral antibiotic courses [4]. Extended hospitalizations for completion of antibiotic therapy in patients who are otherwise stable for discharge lead to negative consequences for patients, including feelings of isolation, financial stress, inability to maintain family responsibilities, and the potential for nosocomial infections while straining the finite resources of hospitals. Although recent randomized trials have demonstrated that oral antibiotics are effective for bone and joint infections and endocarditis, these studies have included few patients with substance use disorders, and concerns about nonadherence may limit use of oral therapy in this population [5, 6]. A recent retrospective study showed similar readmission rates among PWID with invasive infections who completed inpatient parenteral therapy and those who received partial oral therapy, but published evidence with oral therapy in this population remains scarce [7]. Novel care models to successfully treat these patients while shortening durations of hospitalization are urgently needed.

Parenteral lipoglycopeptides, such as dalbavancin, have a long half-life and are approved for single-dose treatment of acute bacterial skin and skin structure infections (ABSSSIs). Off-label use of dalbavancin for more complicated, deep-seated infections such as IE and OM has been supported by an increasing number of case series, especially among vulnerable populations who are not suitable candidates for OPAT [8–14]. However, to date, there has been only 1 published randomized trial in which a multidose dalbavancin regimen was as effective as standard of care for osteomyelitis [13]. In February 2018, we implemented a standardized approach in which a single dose of dalbavancin was given 7–10 days before the planned end date of therapy for vulnerable patients receiving prolonged inpatient parenteral antibiotic therapy for bloodstream infection, IE, or OM, thus allowing an earlier hospital discharge with no further therapy needed. In this report, we describe the clinical outcomes, adverse events, and cost avoidance associated with this standardized treatment approach.

METHODS

Setting and Population

This was a retrospective observational case series conducted at Denver Health Medical Center in Denver, Colorado. Denver Health Medical Center is a 525-bed, academic, tertiary care, level 1 trauma center that serves as the primary safety net institution for the city and county of Denver. Adult patients ≥18 years of age hospitalized from February 1, 2018, to November 30, 2019, with a gram-positive infection in which a single dose of dalbavancin was used to complete the planned antibiotic course were included. Patients who received dalbavancin in the outpatient setting were excluded. Refer to the Supplementary Data for our specific institutional approach for use of dalbavancin.

Data Collection

Data extracted from the electronic health record included patient demographics, medical comorbidities, microbiology, indication for antibiotics, antibiotics administered, hospital length of stay, time to follow-up, and dalbavancin-related adverse events. Hospital readmissions, emergency department or urgent care visits, and deaths occurring within 90 days of the dalbavancin dose were recorded.

Definitions

Clinical failure was defined as the need for additional antibiotic therapy related to the primary infection or clinical or microbiological evidence of active infection within 90 days of the dalbavancin dose. Cases not meeting criteria for clinical failure were classified as clinical success. For the primary analysis, patients lost to follow-up, defined as having no clinical encounters within 90 days of the dalbavancin dose, were classified as clinical failure. Potential dalbavancin-related adverse events were defined as hypersensitivity, dermatologic, gastrointestinal, hepatic, hematologic, or other reactions documented in the electronic health record within 7 days of the dose. Potential adverse events were adjudicated by 3 study authors (A.V.D., K.S., H.Y.) to determine relatedness to dalbavancin.

Statistical and Cost Analyses

The projected hospital length of stay (LOS) was the date of admission to the planned end date of antibiotic therapy as documented by the Infectious Diseases service. For each hospitalization, hospital days averted were determined by subtracting the actual LOS from the projected LOS. Estimated cost avoidance was determined by multiplying the total hospital days averted by the estimated cost of 1 hospital day ($1800) minus an estimated $3000 for each dalbavancin dose.

Ethical Approval

This programmatic evaluation was reviewed by the Quality Improvement Review Committee of Denver Health and Hospital Authority (authorized by the Colorado Multiple Institutional Review Board) and deemed not to constitute human subjects research.

RESULTS

Of 28 patients who received an inpatient dose of dalbavancin, 27 met criteria for inclusion. One patient was excluded because treatment was planned to be completed with a second outpatient dose of dalbavancin and therefore did not meet the institutional criteria for use being evaluated. The median age (interquartile range [IQR]) was 49 (40–54) years. A large proportion of patients were PWID (67%), were experiencing homelessness (56%), or had alcohol use disorder (22%). Twenty-five (93%) were bacteremic at hospital admission. Other baseline and clinical characteristics are shown in Table 1.

Table 1.

Baseline and Clinical Characteristics (n = 27)

Age, median (IQR), y	49 (40–54)
Male	17 (63)
White/Caucasian	25 (93)
Experiencing homelessness	15 (56)
Substance use disordera
People who inject drugs	18 (67)
Noninjection drug use	7 (26)
Alcohol use disorder	6 (22)
Concomitant diseases
Hepatitis C virus infection	13 (48)
HIV infection	4 (15)
Diabetes mellitus	2 (7)
Cirrhosis	1 (4)
Chronic kidney disease	0
Bacteremia	25 (93)
Infection
Documented bacteremia
Uncomplicated bacteremia	7 (26)
Complicated bacteremia	4 (15)
Right-sided infective endocarditis	7 (26)
Bone or joint infection with bacteremia	4 (15)
Pneumonia with bacteremia	2 (7)
Left-sided infective endocarditis	1 (4)
Bone or joint infection without bacteremia	2 (7)
Microorganisms
Methicillin-susceptible Staphylococcus aureus	12 (44)
Methicillin-resistant Staphylococcus aureus	11 (41)
Polymicrobial	4 (15)
MRSA and group A streptococci	1
MSSA and Streptococcus anginosus	1
Methicillin-resistant Staphylococcus epidermidis and group A streptococci	1
MRSA, Enterococcus durans or hirae, and alpha-hemolytic streptococci	1
Days of antibiotic therapy before dalbavancin administration, median (IQR)	21 (8.5–35)
Hospital length of stay, actual, median (IQR)	26 (10–34.5)
Hospital days averted per patient, median (IQR)	7 (6–7)
90-d inpatient re-admission, all-cause	4 (15)
90-d inpatient re-admission due to initial infection	1 (4)
90-d re-presentation to the emergency department/urgent care, all-cause	10 (37)
90-d re-presentation to the emergency department/urgent care due to initial infection	0
Follow-up appointment with either ID clinic, primary care provider, or other outpatient clinic to assess infection	13/25 (52)
Loss to follow-up	4 (15)
Potential dalbavancin-related adverse events	2 (7)
90-d mortality, all-cause	0
90-d mortality due to initial infection	0
Clinical success	22 (81)

Data reported as No. (%) unless otherwise noted.

Abbreviations: ID, infectious diseases; IQR, interquartile range; MRSA, methicillin-resistant Staphylococcus aureus; MSSA, methicillin-susceptible Staphylococcus aureus.

^aCategories are not mutually exclusive.

The most common indications for the prolonged inpatient antibiotic course included uncomplicated Staphylococcus aureus bacteremia (26%), right-sided IE (26%), and bone or joint infection with concurrent bacteremia (15%). S. aureus was isolated in all 27 patients; 48% were methicillin-resistant S. aureus (MRSA). Four cases (15%) involved a polymicrobial infection. Dalbavancin was administered a median (IQR) of 21 (8.5–35) days after clearance of blood cultures or source control was achieved. The median length of hospital stay (IQR) was 26 (10–34.5) days.

Of the 27 patients, 2 planned to follow up at a clinic outside of the Denver Health system. Of the remaining 25 patients, 52% followed up in the infectious diseases clinic, with a primary care provider, or in another clinic in which the initial infection was assessed. Within the 90-day follow-up period, 37% of patients had an emergency department or urgent care visit, and 15% were re-hospitalized. Only 1 of these encounters was likely due to recurrence of the primary infection. Four (15%) patients, including 2 with planned follow-up outside of Denver Health, had no documented clinical encounters and were classified as lost to follow-up. Overall, 22 (81%) patients met criteria for clinical success. When excluding the 4 patients lost to follow-up, 96% achieved clinical success. Two patients experienced adverse effects potentially related to dalbavancin within 1 day of dose administration. Both reported self-resolving dizziness, and 1 reported nausea requiring ondansetron. Additional details of the cases are presented in Table 2.

Table 2.

Case Descriptions

		Substance Use Disorders
Patient	Experiencing Home-lessness	Noninjection Drug Use	Injection Drug Use	Alcohol Use Disorder	Indication	Pathogen	Culture Source	Days of Antibiotic Therapy Before Dalbavancina	Inpatient Re-admission, Reason	Emergency Department or Urgent Care Representation, Reason	Clinical Failure or Loss to Follow-up
1	Yes	No	No	Yes	Osteomyelitis, foot	MRSA	Bone	35	No	Yes, foot pain with no evidence of infection	No
2	Yes	No	No	Yes	Osteomyelitis, hand	MRSE, Group A Streptococci	Bone	34	Yes, hand pain, no evidence of infection and alcohol intoxication	Yes, alcohol intoxication with delirium	No
3	Yes	Yes	No	Yes	Complicated bacteremia with septic superficial thrombophlebitis	MSSA	Blood	21	No	No	No
4	Yes	Yes	No	No	Complicated bacteremia with pneumonia	MRSA, Group A Streptococci	Blood, sputum	21	No	No	No
5	Yes	No	Yes	Yes	Bacteremia with septic arthritis, skin abscesses, pyomyositis, and osteomyelitis of the humerus complicated by septic pulmonary emboli concerning for tricuspid valve endocarditis	MRSA	Blood, joint synovial fluid, soft tissue, bone, wound abscess	35	Yes,b MRSA bacteremia with shoulder osteo-myelitis	Yes, fatigue and pain due to assault	Clinical failure
6	No	No	Yes	No	Complicated bacteremia with psoas abscess	MSSA	Blood, wound abscess	27	No	No	Lost to follow-up
7	Yes	No	Yes	No	Bacteremia with left upper extremity abscess	MSSA, S. anginosus	Blood, wound abscess	14	No	Yes, upper respiratory infection	No
8	No	No	Yes	No	Bacteremia with osteomyelitis, left shoulder	MSSA	Blood	35	No	No	No
9	Yes	No	Yes	No	Bacteremia with neck abscess, septic thrombophlebitis, and concern for L4-L5 septic joint facet and epidural phlegmon	MRSA	Blood, wound abscess	28	No	Yes, labial abscess	No
10	Yes	No	Yes	No	Tricuspid valve endocarditis	MSSA	Blood	8c	No	No	Lost to follow-up
11	Yes	No	Yes	No	Uncomplicated bacteremia with gluteal abscess	MSSA	Blood	9	No	No	Lost to follow-up
12	Yes	No	Yes	No	Uncomplicated bacteremia secondary to bilateral upper extremity abscesses	MRSA	Blood, wound abscess	7	Yes, hypoxia	Yes, phlebitis	No
13	Yes	No	Yes	No	Complicated bacteremia with probable septic pulmonary embolism and right-sided endocarditis	MRSA	Blood	21	No	Yes, chest pain	No
14	Yes	No	Yes	No	Uncomplicated bacteremia secondary to cutaneous abscess	MRSA	Blood	7	No	No	No
15	No	No	Yes	No	Tricuspid valve endocarditis with septic thrombophlebitis of right external jugular	MRSA	Blood	35	No	No	Lost to follow-up
16	Yes	No	Yes	No	Uncomplicated bacteremia secondary to purulent cellulitis	MSSA	Blood	7	No	Yes, wound check	No
17	No	Yes	Yes	No	Bacteremia with tricuspid valve endocarditis complicated by acute pacer lead infection status postextraction	MRSA	Blood	35	No	No	No
18	No	Yes	No	No	Bacteremia with sternomanubrial septic arthritis and osteomyelitis	MSSA	Blood	35	No	No	No
19	No	No	Yes	No	Uncomplicated bacteremia with cellulitis, foot	MSSA	Blood	7	No	No	No
20	No	No	No	Yes	Complicated bacteremia of unknown source	MSSA	Blood	21	No	No	No
21	No	Yes	No	No	Bacteremia with necrotizing pneumonia with pleural effusion and septic emboli suggestive of right-sided endocarditis	MRSA	Blood, sputum	21	No	Yes, left-sided Bell’s palsy	No
22	Yes	Yes	Yes	No	Tricuspid-valve endocarditis status post–valve replacement	MRSA, E. durans or hirae, alpha-hemolytic Streptococci	Blood	35	No	No	No
23	No	No	No	No	Uncomplicated bacteremia secondary to abscess, left buttock	MSSA	Blood	6	No	No	No
24	Yes	No	Yes	No	Bacteremia with aortic valve endocarditis status post–valve replacement complicated by multiple septic emboli	MRSA	Blood	35	Yes, presyncope and dizziness, urinary tract infection	No	No
25	No	No	No	Yes	Bacteremia of unknown source	MSSA	Blood	14	No	No	No
26	No	Yes	Yes	No	Uncomplicated bacteremia secondary to cellulitis, left lower extremity	MSSA	Blood	7	No	Yes, upper respiratory tract infection	No
27	No	No	Yes	No	Bacteremia with tricuspid valve endocarditis complicated by septic emboli likely secondary to abscess, right upper extremity	MRSA	Blood	35	No	No	No

Abbreviations: MRSA, methicillin-resistant Staphylococcus aureus; MRSE, methicillin-resistant Staphylococcus epidermidis; MSSA, methicillin-susceptible Staphylococcus aureus.

^aAfter clearance of blood cultures or appropriate surgical source control.

^bPatient was readmitted 60 days after dalbavancin administration with MRSA bacteremia and shoulder osteomyelitis in the setting of continued injection drug use. During the initial encounter, the patient had received 50 days of parenteral antibiotics (35 days with regards to when source control was achieved). The patient did not show up to the infectious diseases clinic appointment but did show up to a primary care provider appointment, and there was documentation of no concern of infection recurrence. The patient also had multiple emergency department encounters between the discharge date and the subsequent readmission, with normal laboratory tests and imaging not suggestive of infection.

^cThe patient had received 2 weeks of parenteral therapy at an outside hospital before being transferred to Denver Health Medical Center.

The total number of hospital days averted was 182 (median [IQR], 7 [6–7]). The projected LOS cost was $1 467 000, and the actual LOS plus dalbavancin drug cost was $1 220 400. Thus, the overall estimated cost avoidance was $246 600 (average of $9600 per patient).

DISCUSSION

To our knowledge, this is the first description of outcomes associated with a standardized approach for use of dalbavancin to facilitate earlier hospital discharge for patients with a serious bacterial infection. Administration of dalbavancin 7–10 days before the planned end date of therapy led to clinical success in 81% of cases, appeared to be safe, and averted about 7 hospital days per patient. This was associated with an estimated cost avoidance of $9600 per patient. Only 1 case was deemed to be a true clinical failure due to recurrent bacteremia and osteomyelitis, but this occurred 2 months after dalbavancin administration in the setting of continued injection drug use, making it difficult to determine if this was a recurrence of the initial infection or reinfection.

There have been an increasing number of reports of off-label dalbavancin use for the treatment of serious infections caused by gram-positive organisms [8–14]. These reports have been heterogeneous with respect to the types of infections treated, timing of dalbavancin dosing, and number of doses given. Our standardized approach represented a relatively conservative off-label use of dalbavancin in that patients were medically stable, nearing the end of their planned antibiotic course, and otherwise ready for discharge when the dose was given. This case series is also unique in that all patients had bloodstream or deep-seated infection (93% were bacteremic); several previous studies included a large proportion of patients of less severe infections such as ABSSSIs or bone or joint infection without bacteremia [10, 13]. The high rate of clinical success and low rate of potential dalbavancin-related adverse events were notable and support the ongoing use of this approach to shorten hospitalizations in these vulnerable patients.

There are several limitations to this report. First, given that this is a description from a single center, the results are not generalizable. Second, the sample size was small, which limits the evaluation of both safety and effectiveness. Third, the retrospective assessment of clinical success was limited to follow-up care in our system and relied on documentation in the medical record. Fourth, we did not perform a formal cost analysis of the program; the results presented are only general estimates. Finally, it is unknown whether any clinical failures were related to the use of dalbavancin or would have also occurred if treatment was completed with parenteral therapy.

In summary, our experience shows that a standardized approach to administer dalbavancin 7–10 days before the planned end date of therapy for vulnerable patients with serious gram-positive infections receiving prolonged inpatient therapy appears to be safe and effective, facilitates earlier hospital discharge, and results in cost avoidance as compared with completion of the full parenteral antibiotic course in the hospital. Until additional clinical data are available demonstrating the safety and efficacy of multiple-dose long-acting lipoglycopeptides for bloodstream or deep-seated infections, this relatively conservative off-label treatment approach may be considered to shorten hospitalizations in vulnerable patients who may not be candidates for OPAT.

Supplementary Data

Supplementary materials are available at Open Forum Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.