Diagnostic Microbiology and Infectious Disease xxx (2014) xxx–xxx

Contents lists available at ScienceDirect

Diagnostic Microbiology and Infectious Disease journal homepage: www.elsevier.com/locate/diagmicrobio

Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter prospective observational study☆,☆☆ B.A. Lipsky a,⁎, L.M. Napolitano b, G.J. Moran c, L. Vo d, S. Nicholson d, S. Chen e, L. Boulanger e, M. Kim d a

University of Oxford, Oxford, UK University of Michigan, Ann Arbor, MI, USA c UCLA Medical Center, Sylmar, CA, USA d Janssen Scientific Affairs, LLC, Raritan, NJ, USA e United BioSource Corporation, Lexington, MA, USA b

a r t i c l e

i n f o

Article history: Received 12 April 2013 Received in revised form 10 February 2014 Accepted 13 February 2014 Available online xxxx Keywords: Complicated skin and soft tissue infections Inappropriate antibiotics treatment Economic outcomes

a b s t r a c t This study examined economic outcomes associated with inappropriate initial antibiotic treatment (IIAT) in complicated skin and soft tissue infections using data from adults hospitalized and treated with intravenous antibiotic therapy. We specifically analyzed for the subsets of patients infected with methicillin-resistant Staphylococcus aureus (MRSA), with healthcare-associated (HCA) infections, or both. Data from 494 patients (HCA: 360; MRSA:175; MRSA + HCA: 129) showed the overall mean length of stay (LOS) was 7.4 days and 15.0% had the composite economic outcome of any subsequent hospital admissions, emergency department visits, or unscheduled visits related to the study infection. A total of 23.1% of patients had IIAT; after adjustments, these patients had longer LOS than patients without IIAT in the HCA cohort (marginal LOS = 1.39 days, P = 0.03) and the MRSA + HCA cohort (marginal LOS = 2.43 days, P = 0.01) and were significantly more likely to have the composite economic outcome in all study cohorts (odds ratio: overall = 1.79; HCA = 3.09; MRSA = 3.66; MRSA + HCA = 6.92; all P b 0.05). © 2014 Elsevier Inc. All rights reserved.

1. Introduction Skin and soft tissue infections (SSTIs) are a frequent clinical problem and are associated with substantial morbidity, often including the need for hospitalization (Stevens et al., 2005). The economic costs of patients hospitalized for SSTIs are substantial (Hatoum et al., 2009). In the past decade, the increased incidence of antimicrobial resistance among common pathogens causing complicated SSTIs (cSSTIs), such as methicillin-resistant Staphylococcus aureus (MRSA) (Klevens et al., 2007; Moran et al., 2006), has contributed to an increase in hospital admissions, emergency department (ED) visits, and outpatient department visits for these infections (Edelsberg et al., 2009; McCaig et al., 2006). Managing MRSA infections has been shown to be more costly than those caused by methicillin-sensitive strains, as they are associated with increased screening for colonization, more expensive antimicrobial regimens, need for personal protective equipment and occasionally isolation rooms, and likely longer length of stay (LOS) in the hospital (Nathwani, 2003). ☆ Funding: The work was supported by Janssen Scientific Affairs, LLC, Raritan, NJ, USA. ☆☆ Conflict of interests: B.A.L., G.M., and L.N. have served as consultants to Janssen Scientific Affairs, LLC; L.V., M.K., and S.N. are employees and shareholders of Janssen Scientific Affairs, LLC; S.C. and L.B. are employees of United BioSource Corporation that were contracted with Janssen Scientific Affairs, LLC to conduct this study. ⁎ Corresponding author. E-mail address: [email protected] (B.A. Lipsky).

Several published studies have reported that patients with various types of bacterial infections who receive inappropriate initial antibiotics therapy (IIAT) have worse clinical outcomes than those treated appropriately (Eagye et al., 2009; Ibrahim et al., 2000; Kollef et al., 2008; Kollef et al., 1999; Kumar et al., 2006; Micek et al., 2007; Schramm et al., 2006; Shorr et al., 2008; Zilberberg et al., 2010). There have been, however, very few studies examining such association in cSSTIs (Zilberberg et al., 2010). Although several studies have examined the effect of infection on clinical outcomes with specific causative pathogens in cSSTIs (Itani et al., 2011; Zilberberg et al., 2010), none have reported the possible effect of IIAT on economic outcomes by type of pathogen isolated. We designed this study to assess the association between IIAT and economic outcomes among a heterogeneous population of patients hospitalized for cSSTI in a variety of settings in the United States. We also attempted to assess the modifying effect of healthcare-associated (HCA) and MRSA infections on the association between IIAT and outcomes.

2. Methods 2.1. Data source and sample selection We obtained the data for this analysis as part of a prospective, multicenter, observational study designed to investigate the clinical

0732-8893/$ – see front matter © 2014 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.diagmicrobio.2014.02.013

Please cite this article as: Lipsky BA, et al, Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter pro..., Diagn Microbiol Infect Dis (2014), http://dx.doi.org/10.1016/j.diagmicrobio.2014.02.013

2

B.A. Lipsky et al. / Diagnostic Microbiology and Infectious Disease xxx (2014) xxx–xxx

characteristics and treatment patterns of cSSTIs in hospitalized patients. This study enrolled patients hospitalized between June 2008 and December 2009 at 62 US hospitals with varying size, ownership status, academic affiliation, and geographic regions. We included patients with diabetic foot infection (DFI), surgical site infection (SSI), deep soft tissue abscess (DSTA), or cellulitis who were treated with intravenous (IV) antibiotics and who had a positive culture from the wound site. Investigators prospectively collected the required data at study enrollment, the end of intravenous antibiotic therapy, the end of hospitalization, during a telephone follow-up visit 28–35 days post-discharge from the hospital, and between these events, as required. We have previously described the study inclusion and exclusion criteria (Lipsky et al., 2012). In brief, we enrolled patients for whom IV antibiotics were the primary treatment regimen during their hospitalization, who had a culture from the cSSTI site taken within 24 hours of admission that yielded at least 1 pathogen, and had an expected inpatient hospital stay of at least 48 hours. We considered wound isolates to be contaminants, rather than pathogens, if they were coagulase-negative staphylococci isolated as one among other probable pathogen(s) or were collected by a swab of the wound or if they were Corynebacterium species. In order to evaluate outcomes after hospitalization, we excluded patients who died in the hospital. All treating physicians agreed to collect data prospectively on standardized forms that we provided, but they each independently made all decisions about patient management. 2.2. Definition We considered initial IV antibiotic treatment to be inappropriate if the selected agents were not active against the identified pathogens based on in vitro susceptibility testing or usual spectrum of coverage (a proxy measure we used for susceptibility when test results were unavailable) or when they were not given within 24 hours of hospital admission. For patients who lacked adequate data on pathogen susceptibility results or spectrum of coverage of selected agents, we classified the appropriateness of antibiotic therapy by achieving consensus among 3 physician authors (BAL, GJM, and LMN), who have expertise in treating cSSTIs. We classified infections as HCA if the patient met at least 1 of the following criteria: had a reported hospitalization within the 6 months immediately preceding admission for the study-related infection; was a resident of, or was admitted from, a nursing home; was in an immunosuppressed state; had been treated with antibiotics during the 30 days prior to hospital admission for the study infection; or, was receiving renal dialysis. 2.3. Data elements We obtained baseline patient characteristics, including age, gender, race/ethnicity, infection type, whether or not they were admitted from the ED, type of hospital ward to which they were admitted, and the presence of active comorbid conditions (i.e., chronic lung disease, diabetes mellitus, hepatic dysfunction, peripheral vascular disease, renal insufficiency, or systemic cancer). We collected microbiological information, including the specific pathogens isolated, the Gram-stain characteristics of the culture isolates, and their antibiotic susceptibility. We documented the initial IV antibiotic regimens used for the study infection and determined the proportion of patients who received 2 or more classes of antibiotics. Furthermore, we documented all surgical or non-surgical procedures that the patients underwent related to the study infection, as well as whether or not we considered the procedures a “source control” of the infection (i.e., incision and drainage, surgical debridement, excision, or amputation of the wound).

We recorded LOS, excluding the number of days that discharge was delayed due to any procedure not related to the study infection. We created a composite economic outcome composed of subsequent hospital admissions, ED visits, or unscheduled visits to a healthcare provider specifically related to the study infection.

2.4. Statistical analysis We developed descriptive results for both the entire study population and 3 non-mutually exclusive sub-populations: 1) those whose infections were HCA (HCA cohort), 2) those whose pathogen was MRSA (MRSA cohort), 3) and those whose infections were HCA and whose pathogen was MRSA (MRSA + HCA cohort). We reported continuous variables with mean and SD and summarized categorical variables using frequencies and percentages. We calculated the rates of IIAT for the 4 study cohorts (overall, HCA, MRSA, and MRSA + HCA), then compared outcomes, including mean LOS and the proportion of patients who met the composite economic outcome of subsequent hospital admissions, ED visits, or unscheduled visits between patients who received IIAT and patients who did not receive IIAT. We used a chi-square test to evaluate the unadjusted differences in the proportion of patients in each cohort with the composite economic outcome. We compared mean LOS based on whether or not patients received IIAT using an unadjusted, generalized linear model with the log link function assuming negative binomial distribution. We performed regression analyses to assess the association between IIAT and economic outcomes while controlling for patient characteristics including age, whether the infection was SSI (yes/no), hospitalization within the past 6 months (yes/no), and pathogen type (only gram-positive, only gram-negative, or mixed). We used logistic regressions to examine the composite economic outcome of subsequent hospital admissions, ED visits, or unscheduled visits for the study infections. We reported odds ratios (ORs) with 95% confidence intervals (CI). For LOS, we used generalized linear regression models with the log link function assuming negative binomial distribution. We estimated marginal effects of LOS of IIAT by computing the expected change in mean LOS as a function of a change in IIAT, while setting all the other covariates constant at their mean values (Greene, 2003).

3. Results 3.1. Baseline characteristics A total of 494 patients met our criteria for inclusion in this analysis. Of these, 360 (72.9%) were in the HCA cohort, 175 (35.4%) were in the MRSA cohort, and 129 (26.1%) were in the MRSA + HCA cohort. Table 1 shows the baseline patient characteristics overall, as well as in the 3 sub-cohorts. Overall, the mean age for the study population was 52.8 years (SD = 15.4), 56.7% were male, and 73.3% were white. By type of infection, 36.2% had SSI, 27.5% had DFI, 20.5% had DSTA, and 15.8% had cellulitis. The majority (70.0%) of the patients were admitted from the ED, and most were hospitalized on a medical ward (70.9%) or a surgical ward (23.3%). Most patients (72.9%) had at least 1 HCA risk factor, the most common of which were antibiotic use within the past 30 days (51.8%) and hospitalization within the past 6 months (48.4%). Diabetes was the most frequent comorbidity in the study population (52.8%), followed by peripheral vascular disease (16.0%) and renal insufficiency (13.2%). The baseline characteristics for the HCA, MRSA, and MRSA + HCA cohorts were generally similar to those in the overall study population. The 2 HCA cohorts had a higher proportion of patients with SSI (HCA: 46.1%; MRSA + HCA: 41.9%), while the 2 MRSA cohorts had a higher proportion of patients with DSTA (MRSA: 30.3%; MRSA + HCA: 24.0%).

Please cite this article as: Lipsky BA, et al, Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter pro..., Diagn Microbiol Infect Dis (2014), http://dx.doi.org/10.1016/j.diagmicrobio.2014.02.013

B.A. Lipsky et al. / Diagnostic Microbiology and Infectious Disease xxx (2014) xxx–xxx

3

Table 1 Baseline characteristics by study cohort.

Age, mean (SD) Male Race White Black or African American Asian American Indian or Alaska Native Native Hawaiian or Pacific Islander Infection type Diabetic foot infection Surgical site infection Deep soft tissue abscess Cellulitis Admitted from ED Admitted to hospital ward Medical ward Surgical ward ICU Other HCA risk factors Hospitalization within past 6 months Resident of or admission from nursing home Immunosuppressed state Previous antibiotic use (past 30 days) Receiving dialysis Comorbidity Chronic lung disease Diabetes Hepatic dysfunction Peripheral vascular disease Renal insufficiency Systemic cancer

Overall (N = 494)

HCA (n = 360)

MRSA (n = 175)

MRSA + HCA (n = 129)

52.8 (15.4) 280 (56.7)

53.4 (15.7) 189 (52.5)

51.24 (15.4) 105 (60.0)

52.67 (15.5) 69 (53.5)

362 (73.3) 100 (20.2) 10 (2.0) 3 (0.6) 19 (3.9)

264 (73.3) 72 (20.0) 5 (1.4) 3 (0.8) 16 (4.4)

123 37 4 2 9

(70.3) (21.1) (2.3) (1.1) (5.1)

90 27 1 2 9

(69.8) (20.9) (0.8) (1.6) (7.0)

136 179 101 78 346

84 (23.3) 166 (46.1) 58 (16.1) 52 (14.4) 244 (67.8)

35 57 53 30 126

(20.0) (32.6) (30.3) (17.1) (72.0)

22 54 31 22 91

(17.1) (41.9) (24.0) (17.1) (70.5)

350 (70.9) 115 (23.3) 8 (1.6) 21 (4.3) 360 (72.9) 239 (48.4) 11 (2.2) 56 (11.3) 261 (52.8) 13 (2.6)

245 (68.1) 93 (25.8) 5 (1.4) 17 (4.7) 360 (100.0) 239 (66.4) 11 (3.1) 56 (15.6) 249 (69.2) 13 (3.6)

120 44 4 7 129 79 8 25 93 5

(68.6) (25.1) (2.3) (4.0) (73.7) (45.1) (4.6) (14.3) (53.1) (2.9)

88 32 2 7 129 79 8 25 89 5

(68.2) (24.8) (1.6) (5.4) (100.0) (61.2) (6.2) (19.4) (69.0) (3.9)

45 256 17 79 65 15

40 (11.1) 181 (50.3) 12 (3.3) 66 (18.3) 50 (13.9) 15 (4.2)

18 83 9 28 27 8

(10.3) (47.4) (5.1) (16.0) (15.4) (4.6)

17 59 8 24 22 8

(13.2) (45.7) (6.2) (18.6) (17.1) (6.2)

(27.5) (36.2) (20.5) (15.8) (70.0)

(9.1) (51.8) (3.4) (16.0) (13.2) (3.0)

“Overall” means all patients enrolled; number of patients in the sub-cohorts added up to more than the overall population because patients could be in more than 1 cohort. Note: Data are number (%) of patients, unless otherwise indicated.

3.2. Microbiology data, treatment patterns, and outcomes As shown in Table 2, for all patients, only gram-positive pathogens were isolated in 66.8% while 12.8% were infected with only gramnegatives, and 20.5% were infected by both gram-positives and gram-

negatives (i.e., a mixed infection). Of the patients in the MRSA and MRSA + HCA cohorts, 18.9% and 19.4% had mixed infections, respectively. The most commonly used initial antibiotics were glycopeptides (always vancomycin, 65.2%) and some type of penicillin (40.1%). A higher proportion of patients in the MRSA cohort (71.4%) and in the

Table 2 Microbiology data, treatment patterns, and outcomes by study cohort.

Infecting pathogen type Gram (+) only Gram (−) only Mixed gram (+) and gram (−) Initial IV antibiotic(s) for study infectiona Glycopeptidesb Penicillinsc Cephalosporinsd Lincosamides Fluoroquinolones Proportion receiving ≥2 classes of initial IV antibiotics Surgical source control Incision and drainage Surgical debridement Excision of wound Amputation Rate of IIAT LOS, mean (SD) Additional care related to study infection post-discharge ED visit Hospital admission Unscheduled visit

Overall (N = 494)

HCA (n = 360)

MRSA (n = 175)

MRSA + HCA (n = 129)

330 (66.8) 63 (12.8) 101 (20.5)

230 (63.9) 55 (15.3) 75 (20.8)

142 (81.1) 0 33 (18.9)

104 (80.6) 0 25 (19.4)

322 (65.2) 198 (40.1) 91 (18.4) 69 (14.0) 36 (7.3) 262 (53.0) 225 (45.6) 150 (30.4) 82 (16.6) 8 (1.6) 31 (6.3) 114 (23.1) 7.4 (6.2) 74 (15.0) 42 (8.5) 29 (5.9) 42 (8.5)

233 (64.7) 140 (38.9) 68 (18.9) 42 (11.7) 24 (6.7) 193 (53.6) 153 (42.5) 99 (27.5) 59 (16.4) 5 (1.4) 14 (3.9) 87 (24.2) 7.8 (6.5) 55 (15.3) 29 (8.1) 20 (5.6) 30 (8.3)

125 (71.4) 66 (37.7) 29 (16.6) 31 (17.7) 8 (4.6) 93 (53.1) 86 (49.1) 66 (37.7) 27 (15.4) 1 (0.6) 7 (4.0) 42 (24.0) 7.0 (5.9) 25 (14.3) 15 (8.6) 7 (4.0) 15 (8.6)

91 (70.5) 50 (38.8) 22 (17.1) 21 (16.3) 7 (5.4) 69 (53.5) 60 (46.5) 45 (34.9) 20 (15.5) 1 (0.8) 2 (1.6) 33 (25.6) 7.3 (5.8) 18 (14.0) 10 (7.8) 4 (3.1) 11 (8.5)

Note: Data are number (%) of patients, unless otherwise indicated. a Not mutually exclusive as patients may receive more than 1 antibiotic. b Solely vancomycin. c Include beta-lactamase inhibitors, beta-lactamase resistant, non–beta-lactamase inhibitors. d Include first generation, third generation, and fourth generation.

Please cite this article as: Lipsky BA, et al, Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter pro..., Diagn Microbiol Infect Dis (2014), http://dx.doi.org/10.1016/j.diagmicrobio.2014.02.013

4

B.A. Lipsky et al. / Diagnostic Microbiology and Infectious Disease xxx (2014) xxx–xxx

MRSA + HCA cohort (70.5%) received glycopeptides as their initial treatment. At least 2 classes of IV antibiotics were initially prescribed in 53% of patients. A surgical source control procedure was performed in 45.6% of patients, most often incision and drainage (30.4% of overall patients) or surgical debridement (16.6% of overall patients). Overall, 23.1% of patients received IIAT. The rates of IIAT were similar among patients in the 3 sub-cohorts of interest: 24.2% for HCA, 24.0% for MRSA, and 25.6% for MRSA + HCA. The average LOS was 7.4 days (SD = 6.2), ranging from 7.0 days (SD = 5.9) in the MRSA cohort to 7.8 days (SD = 6.5) in the HCA cohort. A total of 15.0% of patients had the composite economic outcome of a hospital admission, an ED visit, or a unscheduled clinic visit due to the study infection after hospital discharge. As shown in Table 2, results were similar (ranging from 14.0% to 15.3%) for the 3 cohorts.

3.3. Outcomes between patients with and without IIAT Across all cohorts, patients who received IIAT had significantly longer LOS than patients who did not receive IIAT (all P b 0.05) (Fig. 1). The difference was largest in the MRSA + HCA cohort (difference = 3.56 days) compared with the MRSA cohort (difference

= 2.39 days), HCA cohort (difference = 2.05 days), or the overall population (difference = 1.58 days). Across all cohorts, the rate of having a subsequent unscheduled visit, ED visit, or hospital admission related to the study infection was higher among those who received IIAT (all P b 0.05). Similar to LOS, the difference was the greatest in the MRSA + HCA cohort (22%) compared with the MRSA cohort (15.7%), HCA cohort (14.7%), or all patients overall (9.0%).

3.4. Effect of IIAT on outcomes After adjusting for patient characteristics, we found that patients in the HCA cohort and the MRSA + HCA cohort receiving IIAT had longer LOS than the other cohorts. Compared to patients who received appropriate initial antibiotic therapy, those who received IIAT stayed in the hospital 1.39 days (P = 0.03) longer in the HCA cohort and 2.43 days (P = 0.01) longer in the MRSA + HCA cohort (Fig. 2). In the overall and HCA cohorts, we also found those who were infected by only gram-negative pathogens had longer LOS compared with those who were infected by only gram-positive pathogens. The marginal LOS associated with infection with gram-negative pathogens was 1.57 days longer (P = 0.02) in the overall cohort and 1.92 days longer in

Mean length of stay (days)

12

P=0.0002 10

P=0.0016 P=0.0029

P=0.0036

8 6 4

9.33

8.66 7.28

7.08

9.91

8.83 6.44

6.35

2 0 Overall

HCA

Appropriate

MRSA

MRSA+HCA

Inappropriate

Percentage of Patients with at Least One Composite Economic Outcome 35%

Percentage of patients

P=0.0017 30% 25%

P=0.0009

P=0.0114

P=0.0177

20% 15%

30.3% 26.4% 21.9%

10% 5%

26.2%

12.9%

11.7%

10.5%

8.3%

0% Overall

HCA

Appropriate

MRSA

MRSA + HCA

Inappropriate

HCA= health care-associated MRSA= methicillin-resistant Staphylococcus aureus p-values are comparisons of patients who received appropriate to inappropriate initial antibiotic therapy *Composite Economic Outcome=hospital admission, emergency department visits, or unscheduled visits to a healthcare provider due to study infection after hospital discharge Fig. 1. LOS and percentage of patients with a composite economic outcome by appropriateness of initial antibiotic treatment among study cohorts.

Please cite this article as: Lipsky BA, et al, Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter pro..., Diagn Microbiol Infect Dis (2014), http://dx.doi.org/10.1016/j.diagmicrobio.2014.02.013

Marginal Length of Stay (days)

Marginal Length of Stay (days)

B.A. Lipsky et al. / Diagnostic Microbiology and Infectious Disease xxx (2014) xxx–xxx

3

unscheduled visit related to the study infection post-discharge (OR = 1.79; 95% CI: 1.01–3.16) (Table 3). In addition to finding a similar association between IIAT and the composite economic outcome in the sub-cohorts, we also observed a larger effect based on a higher OR in the HCA cohort (OR = 3.09; 95% CI: 1.52–6.30), MRSA cohort (OR = 3.66; 95% CI: 1.25–10.70), and MRSA + HCA cohort (OR = 6.92; 95% CI: 1.77–27.08). In the MRSA cohort, patients who were also infected with gram-negatives were more likely to have those subsequent events compared with patients infected with just grampositives (OR = 2.87; 95% CI:1.02–8.02). Other variables in the models were not statistically significant predictors of these economic outcomes.

Inappropriate Initial Antibiotic Treatment P=0.01

2 P=0.03

P=0.10

P=0.06

2.43

1 1.39

1.34

1.00

0

3

Overall

HCA

MRSA

MRSA + HCA

4. Discussion

Patients with Only Gram Negative Pathogens P=0.02

P=0.02

2

1

1.92 1.57

N/A

0

Overall

HCA

N/A

MRSA

5

MRSA + HCA

HCA= health care-associated MRSA= methicillin-resistant Staphylococcus aureus Note: Adjusted difference of length of stay for inappropriate initial antibiotic treatment estimated using generalized linear regression model with log function and negative binomial distribution, controlling for age, surgical site infection, and hospitalization during 6 months prior. N/A= not applicable

Fig. 2. Marginal LOS associated with IIAT and for gram-negative infection.

the HCA cohort (P = 0.02) than with infections associated with only gram-positive or mixed pathogens. Other characteristics associated with longer LOS included older age and having an SSI (Table 3). IIAT was also associated with a higher likelihood of having the composite economic outcome of a re-admission, an ED visit, or an

IIAT in patients with various infections has been found to be associated with adverse outcomes in most (Cheong et al., 2008; De Rosa et al., 2011; Erbay et al., 2009; Kollef, 2008; Kumar et al., 2009; Kuti et al., 2008; Micek et al., 2011; Micek et al., 2010) but not all (Kang et al., 2011; Teng et al., 2009) studies. Most studies have examined this issue in patients with some form of bloodstream infection or pneumonia; few have looked at patients with SSTIs (Teng et al., 2009). In addition, most have examined clinical, but not economic, outcomes associated with IIAT. We found that patients hospitalized with cSSTI who received IIAT had longer LOS (by 1.00 day) and a greater likelihood of subsequent healthcare utilization, especially if they had MRSA, an HCA infection, or both. Similarly, patients who received IIAT were significantly more likely to have a subsequent hospital admission, ED visit, or unscheduled visit related to the study infection. Consistent with our findings were those reported by Zilberberg et al. (2010), who found that inappropriate empiric therapy given to patients with HCA cSSTIs at a single large urban academic tertiary care medical center was associated with an increased LOS of 1.8 days. Our study confirmed this association from a larger sample of more heterogenous patients, thus affording greater generalizability. Our findings are also similar to those in a study by Eagye et al. (2009) that

Table 3 Results of regressions analyses assessing effect of IIAT on economic outcomes. Overall

HCA

MRSA

MRSA + HCA

Outcome: hospital admission, ED visit, or unscheduled visit due to study infection OR Age b40 Age 40–64 Age 65+ Non-SSI SSI Hospitalization 6 months prior Gram-positive only Gram-negative only Mixed pathogens IIAT

Reference 0.99 0.67 Reference 0.98 1.50 Reference 0.86 1.31 1.79

95% CI (0.51–1.92) (0.29–1.55) (0.54–1.80) (0.83–2.72) (0.38–1.94) (0.70–2.45) (1.01–3.16)

OR Reference 0.84 0.50 Reference 1.05 1.72 Reference 0.92 1.06 3.09

95% CI (0.39–1.82) (0.19–1.33) (0.55–2.01) (0.81–3.65) (0.39–2.16) (0.50–2.27) (1.52–6.30)

OR Reference 0.76 0.27 Reference 0.64 1.86 Reference 2.87 3.66

95% CI (0.23–2.46) (0.05–1.47) (0.21–1.93) (0.66–5.29)

(1.02–8.02) (1.25–10.70)

OR Reference 1.44 0.31 Reference 0.68 2.19 Reference 2.43 6.92

95% CI (0.26–7.87) (0.03–2.93) (0.21–2.19) (0.57–8.44)

(0.70–8.43) (1.77–27.08)

Outcome: LOS Coefficient Age b40 Age 40–64 Age 65+ Non-SSI SSI Hospitalization 6 months prior Gram-positive only Gram-negative only Mixed pathogens IIAT

Reference 0.20 0.33 Reference 0.14 0.07 Reference 0.20 0.09 0.13

P-value 0.01 b0.01 0.03 0.27 0.02 0.24 0.06

Coefficient Reference 0.17 0.34 Reference 0.12 0.06 Reference 0.23 0.04 0.17

P-value 0.07 b0.01 0.11 0.49 0.02 0.61 0.03

Coefficient Reference 0.13 0.16 Reference 0.11 0.11 Reference 0.12 0.11

P-value 0.23 0.06 0.01 0.85 0.13 0.10

Coefficient Reference 0.15 0.34 Reference 0.11 0.03 Reference 0.21 0.31

P-value 0.31 0.06 0.37 0.83 0.12 0.01

Bold type indicates statistically significant values. Mixed = both gram-positive and gram-negative pathogens from wound culture.

Please cite this article as: Lipsky BA, et al, Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter pro..., Diagn Microbiol Infect Dis (2014), http://dx.doi.org/10.1016/j.diagmicrobio.2014.02.013

6

B.A. Lipsky et al. / Diagnostic Microbiology and Infectious Disease xxx (2014) xxx–xxx

found that inadequate antibiotic therapy in patients with SSI was associated with an additional LOS of 1.4 days. This reported marginal increase of LOS associated with IIAT is, however, much lower than the 5.4 days reported in a study of patients hospitalized with cSSTI by Edelsberg et al. (2008). That study, however, used treatment failure as a surrogate marker for inappropriate treatment since microbiological data were not reported. Our finding on the effect of IIAT on LOS underscores the substantial potential economic benefits to hospitals, providers, and payers, of avoiding IIAT. We also evaluated the importance of IIAT in specific populations of hospitalized patients, i.e., those with infections that were HCA, involved MRSA, or both. The 1-day greater duration of LOS associated with IIAT for our overall study population was not quite statistically significant (P = 0.06). The marginal LOS associated with IIAT was more substantial in patients with HCA infections (LOS = 1.39 days, P = 0.03) and patients with MRSA + HCA infections (LOS = 2.43 days, P = 0.01). This finding highlights the particular importance of choosing appropriate antibiotic therapy in hospitalized patients with HCA or risk factors for MRSA infections as IIAT was especially consequential in these more vulnerable populations. Our results also demonstrated that compared with patients who were infected only with gram-positive pathogens, those infected with only gram-negatives had a longer LOS (marginal LOS = 1.57 days in the overall population and 1.92 days in the HCA populations; both P = 0.02). As with the findings regarding IIAT, the impact of infection with gram-negatives was also greater in the HCA than the overall population. We were able to find only 1 study that reported on the association between pathogen type and LOS in cSSTIs (Itani et al., 2011). Unlike our results, Itani et al. (2011) found that infection by mixed pathogens, compared with just gram-positives, was associated with a longer LOS (5.8 days), but not when compared with gramnegatives. Although cSSTIs are most commonly caused by grampositive pathogens, infections with gram-negative pathogens were associated with longer LOS. This finding emphasizes the importance of providing appropriate initial antibiotic therapy, particularly in patients with gram-negative infections. Our study was designed to look not just at management during the hospitalization period but also evaluated subsequent economic outcomes after patient discharge. We found that patients who received IIAT were more likely to have subsequent re-admissions, ED visits, or unscheduled visits for the study infection. Similar to what was found with LOS, the likelihood of these events post-discharge was higher in the more vulnerable populations (i.e., those with HCA, MRSA, and MRSA + HCA infections). Interestingly, the type of infecting pathogen did not appear to influence economic outcomes after patients were discharged. To our knowledge, ours is the first study to assess and report post-discharge medical resource utilization in patients with cSSTIs. This study, of course, has a number of limitations. Although we made efforts to minimize selection bias by establishing a priori definitions and standardized data collection procedures, we could not completely eliminate these biases. The accuracy and completeness of the data collected varied with the investigators at the individual study sites. Enrollment in the study did not require that patients have a wound culture taken within 24 hours, but we analyzed only those who did, thus limiting the sample size. Consequently, some of the subanalyses had a small sample size and might not have sufficient power to detect a true difference. To minimize the number of patients who had wound culture results that represented colonization and not infection, we excluded patients with pathogens that we pre-defined as suspected contaminants. With a few patients for whom the preestablished algorithm was not useable, we assessed the appropriateness of initial antibiotic therapy by chart review and based our final determination on the consensus of the 3 physician authors. In addition, although we adjusted our regression models, they might not account for all unobservable confounders. Finally, our study

included the 4 most common cSSTI infection types; the findings may not be generalizable to other types of cSSTIs, non-complicated infections, or excluded populations. Notwithstanding these few limitations, this was a large, multicenter, prospective study of a heterogeneous population and provides important information on a key issue, i.e., the economic outcomes of IIAT. Our results suggest that patients hospitalized with cSSTIs who received IIAT had a longer LOS and a greater likelihood of subsequent unscheduled clinic visits, ED visits, or hospital readmissions than patients who received appropriate initial antibiotic therapy. Thus, avoiding IIAT is not only clinically important, but also has substantial economic implications. Acknowledgment United BioSource Corporation performed the statistical analyses and provided assistance in preparing and editing the manuscript. Quality Review Associates, Inc., provided assistance in data review and query. References Cheong HS, Kang CI, Wi YM, Ko KS, Chung DR, Lee NY, et al. Inappropriate initial antimicrobial therapy as a risk factor for mortality in patients with communityonset Pseudomonas aeruginosa bacteraemia. Eur J Clin Microbiol Infect Dis 2008;27: 1219–25. De Rosa FG, Pagani N, Fossati L, Raviolo S, Cometto C, Cavallerio P, et al. The effect of inappropriate therapy on bacteremia by ESBL-producing bacteria. Infection 2011;39:555–61. Eagye KJ, Kim A, Laohavaleeson S, Kuti JL, Nicolau DP. Surgical site infections: does inadequate antibiotic therapy affect patient outcomes? Surg Infect (Larchmt) 2009;10:323–31. Edelsberg J, Berger A, Weber DJ, Mallick R, Kuznik A, Oster G. Clinical and economic consequences of failure of initial antibiotic therapy for hospitalized patients with complicated skin and skin-structure infections. Infect Control Hosp Epidemiol 2008;29:160–9. Edelsberg J, Taneja C, Zervos M, Haque N, Moore C, Reyes K, et al. Trends in US hospital admissions for skin and soft tissue infections. Emerg Infect Dis 2009;15:1516–8. Erbay A, Idil A, Gozel MG, Mumcuoglu I, Balaban N. Impact of early appropriate antimicrobial therapy on survival in Acinetobacter baumannii bloodstream infections. Int J Antimicrob Agents 2009;34:575–9. Greene WH. Econometric analysis. 5th ed. Upper Saddle River, NJ: Pearson Education, Inc.; 2003. Hatoum HT, Akhras KS, Lin SJ. The attributable clinical and economic burden of skin and skin structure infections in hospitalized patients: a matched cohort study. Diagn Microbiol Infect Dis 2009;64:305–10. Ibrahim EH, Sherman G, Ward S, Fraser VJ, Kollef MH. The influence of inadequate antimicrobial treatment of bloodstream infections on patient outcomes in the ICU setting. Chest 2000;118:146–55. Itani KM, Merchant S, Lin SJ, Akhras K, Alandete JC, Hatoum HT. Outcomes and management costs in patients hospitalized for skin and skin-structure infections. Am J Infect Control 2011;39:42–9. Kang CI, Chung DR, Ko KS, Peck KR, Song JH. Risk factors for mortality and impact of broad-spectrum cephalosporin resistance on outcome in bacteraemic intraabdominal infections caused by Gram-negative bacilli. Scand J Infect Dis 2011;43:202–8. Klevens RM, Morrison MA, Nadle J, Petit S, Gershman K, Ray S, et al. Invasive methicillin-resistant Staphylococcus aureus infections in the United States. JAMA 2007;298:1763–71. Kollef MH. Broad-spectrum antimicrobials and the treatment of serious bacterial infections: getting it right up front. Clin Infect Dis 2008;47(Suppl 1):S3–S13. Kollef MH, Sherman G, Ward S, Fraser VJ. Inadequate antimicrobial treatment of infections: a risk factor for hospital mortality among critically ill patients. Chest 1999;115:462–74. Kollef KE, Schramm GE, Wills AR, Reichley RM, Micek ST, Kollef MH. Predictors of 30day mortality and hospital costs in patients with ventilator-associated pneumonia attributed to potentially antibiotic-resistant gram-negative bacteria. Chest 2008;134:281–7. Kumar A, Roberts D, Wood KE, Light B, Parrillo JE, Sharma S, et al. Duration of hypotension before initiation of effective antimicrobial therapy is the critical determinant of survival in human septic shock. Crit Care Med 2006;34:1589–96. Kumar A, Ellis P, Arabi Y, Roberts D, Light B, Parrillo JE, et al. Initiation of inappropriate antimicrobial therapy results in a fivefold reduction of survival in human septic shock. Chest 2009;136:1237–48. Kuti EL, Patel AA, Coleman CI. Impact of inappropriate antibiotic therapy on mortality in patients with ventilator-associated pneumonia and blood stream infection: a metaanalysis. J Crit Care 2008;23:91–100. Lipsky BA, Moran GJ, Napolitano LM, Vo L, Nicholson S, Kim M. A prospective, multicenter, observational study of complicated skin and soft tissue infections in

Please cite this article as: Lipsky BA, et al, Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter pro..., Diagn Microbiol Infect Dis (2014), http://dx.doi.org/10.1016/j.diagmicrobio.2014.02.013

B.A. Lipsky et al. / Diagnostic Microbiology and Infectious Disease xxx (2014) xxx–xxx hospitalized patients: clinical characteristics, medical treatment, and outcomes. BMC Infect Dis 2012;12:227. McCaig LF, McDonald LC, Mandal S, Jernigan DB. Staphylococcus aureus-associated skin and soft tissue infections in ambulatory care. Emerg Infect Dis 2006;12: 1715–23. Micek ST, Kollef KE, Reichley RM, Roubinian N, Kollef MH. Health care-associated pneumonia and community-acquired pneumonia: a single-center experience. Antimicrob Agents Chemother 2007;51:3568–73. Micek ST, Welch EC, Khan J, Pervez M, Doherty JA, Reichley RM, et al. Empiric combination antibiotic therapy is associated with improved outcome against sepsis due to Gram-negative bacteria: a retrospective analysis. Antimicrob Agents Chemother 2010;54:1742–8. Micek ST, Reichley RM, Kollef MH. Health care-associated pneumonia (HCAP): empiric antibiotics targeting methicillin-resistant Staphylococcus aureus (MRSA) and Pseudomonas aeruginosa predict optimal outcome. Medicine (Baltimore) 2011;90:390–5. Moran GJ, Krishnadasan A, Gorwitz RJ, Fosheim GE, McDougal LK, Carey RB, et al. Methicillin-resistant S. aureus infections among patients in the emergency department. N Engl J Med 2006;355:666–74.

7

Nathwani D. Impact of methicillin-resistant Staphylococcus aureus infections on key health economic outcomes: does reducing the length of hospital stay matter? J Antimicrob Chemother 2003;51:ii37–44. Schramm GE, Johnson JA, Doherty JA, Micek ST, Kollef MH. Methicillin-resistant Staphylococcus aureus sterile-site infection: the importance of appropriate initial antimicrobial treatment. Crit Care Med 2006;34:2069–74. Shorr AF, Micek ST, Kollef MH. Inappropriate therapy for methicillin-resistant Staphylococcus aureus: resource utilization and cost implications. Crit Care Med 2008;36:2335–40. Stevens DL, Bisno AL, Chambers HF, Everett ED, Dellinger P, Goldstein EJ, et al. Practice guidelines for the diagnosis and management of skin and soft-tissue infections. Clin Infect Dis 2005;41:1373–406. Teng CS, Lo WT, Wang SR, Tseng MH, Chu ML, Wang CC. The role of antimicrobial therapy for treatment of uncomplicated skin and soft tissue infections from community-acquired methicillin-resistant Staphylococcus aureus in children. J Microbiol Immunol Infect 2009;42:324–8. Zilberberg MD, Shorr AF, Micek ST, Chen J, Ramsey AM, Hoban AP, et al. Hospitalizations with healthcare-associated complicated skin and skin structure infections: impact of inappropriate empiric therapy on outcomes. J Hosp Med 2010;5:535–40.

Please cite this article as: Lipsky BA, et al, Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter pro..., Diagn Microbiol Infect Dis (2014), http://dx.doi.org/10.1016/j.diagmicrobio.2014.02.013

Economic outcomes of inappropriate initial antibiotic treatment for complicated skin and soft tissue infections: a multicenter prospective observational study.

This study examined economic outcomes associated with inappropriate initial antibiotic treatment (IIAT) in complicated skin and soft tissue infections...
402KB Sizes 0 Downloads 3 Views