Article, Infectious Diseases

Incidence of bacteremia and antimicrobial resistance, and associated factors among patients transferred from long-term care hospital

Incidence of bacteremia and Antimicrobial resistance, and associated factors among patients transferred from long-term care hospital

Sangmin Kim, MD a, Sion Jo, MD, PhD a,?, Jae Baek Lee, MD, PhD a, Youngho Jin, MD, PhD a, Taeoh Jeong, MD, PhD a, Jaechol Yoon, MD, PhD a, So Eun Kim, MD a,

Boyoung Park, MD, PhD b, Hasan Bhally, MD, PhD c

a Department of Emergency Medicine, Research Institute of Clinical Medicine of Chonbuk National University and Biomedical Research Institute of Chonbuk National University Hospital, Jeonju-si,

Republic of Korea

b Department of Medicine, Hanyang University College of Medicine, Seoul, Republic of Korea

c Department of Medicine and Infectious Diseases, North Shore Hospital, Private Bag 93503, Auckland 0620, New Zealand

a r t i c l e i n f o

Article history:

Received 1 September 2018

Received in revised form 29 October 2018

Accepted 6 November 2018

Keywords:

Bacteremia

Long-term care hospitals Antimicrobial resistant bacteremia Procalcitonin

a b s t r a c t

Objective: To evaluate the prevalence of bacteremia and antimicrobial resistance, and associated factors among infectious patients transferred from long-term care hospitals (LTCHs).

Methods: Consecutive adult patients who were transferred for suspected infection from affiliated LTCH’s to study hospital emergency department (ED) over a 12 month period from January to December 2016 were included ret- rospectively. Patients with positive blood cultures (excluding contaminants as clinically determined) were de- fined as primary measure and subjected to further analysis according to antimicrobial resistance pattern. The latter was categorized into 4 subgroups based on groups of antimicrobial choices for empiric choices of suspected Bloodstream infections. R-Group 0: bacteria susceptible to penicillin and amoxicillin; R-Group 1: bacteria resis- tant to penicillin/amoxicillin, first, second, or third generation cephalosporins. R-Group 2: ESBL-producing bacte- ria or bacteria resistant methicillin, fourth generation cephalosporin, or fluoroquinolone. R-Group 3: highly resistant pathogens including vancomycin resistant enterococci, carbapenem or colistin resistant Gram nega- tives. Blood culture isolate could therefore be included in N1 group.

Results: Among 756 patients who were transferred from LTCHs, we excluded 278 patients who were not suspi- cious of infection and 65 patients who were not checked blood culture at ED. In total, 422 patients were enrolled. The incidence of bacteremia was 20.4% (n = 86). The most frequent pathogen was E. coli (n = 25) followed by

S. aureus (n = 10), S. epidermidis (n = 8), and K. pneumonia (n = 6). The incidences of the R-Group 1, 2, and 3 groups were 16.8% (n = 71), 14.4% (n = 61), and 1.4% (n = 6), respectively. Of the Gram-positive pathogens (n = 44), the R-Group 1, 2, and 3 groups were 84.1% (n = 37), 75.0% (n = 33), and 9.1% (n = 4), respectively. Of the Gram-negative pathogens (n = 46), the R-Group 1, 2, and 3 groups were 82.6% (n = 38), 69.6% (n = 32), and 4.3% (n = 2), respectively. Among tested variables, initial serum procalcitonin level was significantly associ- ated with the presence of bacteremia (AOR 1.03, 95% confidence interval 1.00-1.05), R-Group 1 (1.04, 1.01-1.07) and the R-Group 2 (1.04, 1.00-1.06). Conclusions: The prevalence of bloodstream infections in patients admitted from LTCH was high (20.4%) with ma- jority of these infections from resistant bacteria. Procalcitonin levels were significantly higher in bacteremic pa- tients with an increasing trend towards bacteria in the antimicrobial resistant groups.

(C) 2018

  1. Introduction

The increasing complexity in medical care in developed countries over the last few decades has improved longevity. However, an

* Corresponding author at: 567 Baekje-daero, Deokjin-gu, Jeonju-si, Jeollabuk-do 54907, Republic of Korea.

E-mail address: [email protected] (S. Jo).

unintended consequence is people with chronic illnesses that are resid- ing in long term care facilities (LTCF) or hospitals (LTCH). In Korea, LTCHs are distinguished from LTCF by increased staffing of both doctors and nurses up to 24 h a day. LTCH patients who are suspected of infec- tion usually receive treatment that includes empirical antimicrobials during their initial illness courses. Patients who respond poorly to the treatment are frequently transferred to the emergency department (ED) of tertiary care or referral hospitals. Once these patients show

https://doi.org/10.1016/j.ajem.2018.11.013

0735-6757/(C) 2018

improvement, they return to the LTCHs. Once this link between LTCHs and tertiary care hospital is established and repeated, it eventually be- comes a cycle.

Thus, notable features should be present among infectious LTCH pa- tients. First, considering the advanced age and high rate of comorbidi- ties of LTCH patients, the incidence of bacteremia should be significant. Second, considering the cycle between LTCHs and Tertiary care hospitals, the incidence of antimicrobial resistant microorganisms should be high. Since the tertiary care hospital is the primary location where antimicrobial resistance occurred [1,2], the antimicrobial resis- tance acquired at the tertiary care hospital can subsequently be spread into LTCHs. Previous studies have already supported these propositions [3-6]. Bacteremia itself and infection by antimicrobial resistant patho- gens can cause significant mortality and morbidity, which makes sub- stantial demands on Health care resources [7-10]. Additionally, selection of appropriate empiric antimicrobials in patients from LCTH is clinically challenging due to a higher risk of colonization with multi- resistant bacteria. Therefore, the incidence of bacteremia, antimicrobial resistance and other associated factors are important areas of concern. Korea, like many developed nations, is challenged with an increasing elderly population with a proportionate rise in people residing in LTCH.

In fact, the need for long-term care beds in Korea has increased the most among all OECD countries since 2000 [11].

To our knowledge, there are no local studies on the prevalence of in- fections including bacteremias, and multi-resistant pathogens in resi- dential care facilities. Furthermore, previous world-wide studies regarding this issue were reported decades ago [3-6]. Therefore, this issue is worthy of being revisited and would be helpful for determining treatment, especially in the choice of empirical antimicrobials. In the present study, the authors evaluated the incidence of bacteremia and antimicrobial resistance, and associated factors among infectious pa- tients transferred from LTCHs.

  1. Methods
    1. Study design and setting

This was a single center, retrospective observational study con- ducted at a local university hospital, a 1200-bed urban, academic, tertiary-care center with the highest ED referral service in the region.

All patients transferred from approximately 50 neighboring LTCHs be- tween January 1st and December 31st, 2016 for suspected infection were screened. This study was approved by the local Institutional Re- view Board (IRB) and informed consent was waived for all subjects in this study. We referred to the Standards for Reporting of Diagnostic Ac- curacy Studies (STARD) recommendations when analyzing the results [12].

Blood cultures were obtained before starting antimicrobials at the study hospital ED. At least two sets of 10 mL blood samples were col- lected in two sets of bottles – BacT/ALERT(R) FA Plus (for detection of aer- obic and facultative microorganisms) and BacT/ALERT(R) FN Plus (for detection of anaerobic microorganisms) – and were incubated under aerobic and anaerobic conditions in an automated blood culture ana- lyzer (BacT/ALERT(R) 3D, BIOMERIEUX, Durham, NC, USA). The bacterial growth was automatically monitored by the instrument.

In the study hospital, department of laboratory medicine take au- thority with the results of blood culture. VITEK MS (bioMerieux, Marcy l’Etoile, France) which is an automated microbial identification system based on matrix-assisted laser desorption ionization time-of- flight mass spectrometry (MALDI-TOF MS) is used. Isolated were inocu- lated into the cards and were run on the VITEK 2 (bioMerieux, Hazel- wood, MO, USA) system to assess the antimicrobial susceptibility test. After the system report, faculties of department of laboratory medicine who are sub-special for infectious disease confirm the result of blood culture. They have authorization to view clinical data of a patient to as- sess the actual pathogenicity after all.

There is no specifically distinct policy for antimicrobial prescription for infectious patients in LTCHs or Transferred patients from LTCHs to study hospital ED who are suspicious of infection in this area.

Selection of participants

Consecutive all adult (aged 18 or above) patients transferred from the LTCHs to the study hospital ED between January 1, 2016 and Decem- ber 31, 2016 were screened for this study. To determine whether a pa- tient was suspected of infection or not, the senior resident of emergency medicine reviewed the ICD 10 code diagnosis at discharge from ED of screened patients. The senior resident was blinded to the aim of the study and patients’ final outcome. Those who were compatible with infection-related diagnosis were only eligible (Fig. 1). Patients who

Image of Fig. 1

Fig. 1. STARD flow diagram.

Table 1

Characteristics of enrolled patients.

All

Non-bacteremia group

Bacteremia group

p-Value

(N = 422)

(N = 336)

N (=86)

Age, yr

77.5 +- 10.7

77.6 +- 11.1

77.5 +- 9.1

0.926

Age category

0.608

b65 yr

45 (10.7)

36 (10.7)

9 (10.5)

65-74 yr

82 (19.4)

63 (18.8)

19 (22.1)

75-84 yr

189 (44.8)

148 (44.1)

41 (47.7)

>=85 yr

106 (25.1)

89 (26.5)

17 (19.8)

Male sex, %

200 (47.4)

160 (47.6)

40 (46.5)

0.854

Hemiplegia, %

53 (12.6)

43 (12.8)

10 (11.6)

0.770

Sore

158 (37.4)

122 (36.3)

36 (41.9)

0.343

L-tube, %

45 (10.7)

32 (9.5)

13 (15.1)

0.134

Foley catheter, %

91 (21.6)

68 (20.2)

23 (26.7)

0.191

PEG, %

21 (5.0)

15 (4.5)

5 (5.8)

0.780

Tracheostomy, %

18 (4.3)

15 (4.4)

3 (3.5)

0.689

PICC, %

27 (6.4)

15 (4.5)

12 (13.9)

b0.01

AVF, %

19 (4.5)

16 (4.7)

3 (3.8)

0.776

PTGBD, %

2 (0.5)

1 (0.3)

1 (1.3)

0.366

ECOG

0.066

0

58 (13.7)

51(15.2)

7(8.1)

1

110 (26.1)

92 (27.4)

18 (20.9)

2

53 (12.6)

43 (12.8)

10 (11.6)

3

79 (18.7)

63 (18.8)

16 (18.6)

4

122 (28.9)

87 (25.9)

35 (40.7)

Hypertension, %

219 (51.9)

172 (51.2)

47 (54.6)

0.567

Diabetes mellitus, %

114 (27.0)

94 (28.0)

20 (23.3)

0.379

Uncomplicated

48(11.4)

38(11.3)

10(11.6)

0.934

end-organ damage

66(15.6)

56(16.7)

10(11.6)

0.251

Chronic liver disease, %

12 (2.8)

9 (2.7)

3 (3.5)

0.716

Mild

1 (0.2)

1 (0.3)

0 (0.0)

0.612

Moderate to severe

11 (2.6)

8 (2.4)

3 (3.5)

0.565

Chronic kidney disease, %

45 (10.7)

40 (11.9)

5 (5.8)

0.119

Stroke, %

169 (40.1)

135 (40.2)

34 (39.5)

0.913

Dementia, %

167 (39.6)

127 (37.8)

40 (46.5)

0.140

Heart failure, %

45 (10.7)

33 (9.8)

12 (14.0)

0.268

Malignancy, %

78 (18.5)

61 (18.2)

17 (19.8)

0.731

Any leukemia or localized solid tumor

36 (8.5)

27 (8.0)

9 (10.5)

0.472

Metastatic solid tumor

42 (10.0)

34 (10.1)

8 (9.3)

0.821

Myocardial infarction, %

25 (5.9)

22 (6.6)

3 (3.5)

0.441

Chronic airway disease, %

49 (11.6)

41 (12.2)

8 (9.3)

0.454

Peptic ulcer disease, %

9 (2.1)

5 (1.5)

4 (4.7)

0.088

Charlson comorbidity

5.4 +- 1.9

5.4 +- 2.0

5.5 +- 1.9

0.97

Fever, %

225 (53.3)

171 (50.9)

54 (62.8)

0.048

Chill, %

71 (16.8)

48 (14.3)

23 (26.7)

b0.01

Respiratory symptoms, %

210 (49.8)

173 (51.5)

37 (43.0)

0.161

abdominal symptoms, %

131 (31.0)

110 (32.7)

21 (24.4)

0.137

urinary symptoms, %

40 (9.5)

26 (7.7)

14 (16.3)

0.016

Other symptoms, %

153 (36.3)

117 (34.8)

36 (41.9)

0.226

Previous antibiotics use, %

119 (28.2)

100 (29.2)

19 (23.8)

0.330

Sepsis

b0.01

None

179 (42.4)

157 (46.7)

22 (25.6)

Sepsis without septic shock

193 (45.7)

154 (45.8)

39 (43.4)

Septic shock

50 (11.9)

25 (7.4)

25 (29.1)

SBP, mmHg (missing at 6)

120.5 +- 33.7

122.7 +- 31.4

112.7 +- 40.3

b0.01

DBP, mmHg (missing at 6)

69.3 +- 18.1

70.5 +- 16.8

64.5 +- 22.2

b0.01

PR, bpm (missing at 3)

92.1 +- 24.2

91.1 +- 22.3

96.0 +- 30.2

0.096

RR, bpm (missing at 3)

19.8 +- 3.1

19.9 +- 3.1

19.6 +- 3.1

0.439

Body Temperature, ?C (missing at 2)

36.7 +- 2.7

36.7 +- 2.2

36.6 +- 4.2

0.752

SpO2, % (missing at 0)

95.6 +- 5.3

95.5 +- 5.7

96.3 +- 3.7

0.221

Mentality (missing at 0)

0.097

Alert. %

277 (65.6)

229 (68.2)

48 (55.8)

Verbal. %

72 (17.1)

56 (16.7)

16 (18.6)

Pain. %

62 (14.7)

44 (13.1)

18 (20.9)

Unresponsive, %

11 (2.6)

7 (2.1)

4 (4.7)

NEWS (missing at 0)

4.8 +- 3.8

4.6 +- 3.7

5.8 +- 4.1

b0.01

KTAS (missing at 0)

0.019

1

21 (5.0)

14 (4.2)

7 (8.1)

2

74 (17.5)

51 (15.2)

23 (26.7)

3

167 (39.6)

138 (41.1)

29 (33.7)

4

148 (35.1)

121 (36.0)

27 (31.4)

5

12 (2.8)

12 (3.6)

0 (0)

White blood cell, 103/uL (missing at 3)

11.9 +- 6.6

11.7 +- 6.4

12.3 +- 7.4

0.502

SEG NEU, % (missing at 4)

79.2 +- 14.3

78.3 +- 13.8

82.9 +- 15.7

b0.01

Hematocrit, % (missing at 3)

32.1 +- 6.2

32.2 +- 6.3

31.6 +- 5.9

0.448

Hemoglobin, g/dL (missing at 3)

10.7 +- 2.2

10.8 +- 2.2

10.6 +- 2.1

0.652

Platelet, 103/uL (missing at 3)

236.4 +- 114.5

249.4 +- 108.9

186.1 +- 122.2

b0.01

Table 1 (continued)

All

Non-bacteremia group

Bacteremia group

p-Value

(N = 422)

(N = 336)

N (=86)

Sodium, mmol/L (missing at 2)

136.5 +- 6.1

136.5 +- 6.2

136.3 +- 6.1

0.709

Albumin, g/dL (missing at 2)

3.1 +- 0.6

3.2 +- 0.6

2.9 +- 0.7

b0.01

BUN, mg/dL (missing at 2)

26.5 +- 20.5

25.9 +- 21.2

29.1 +- 17.3

0.188

Creatinine, mg/dL (missing at 2)

1.3 +- 1.7

1.3 +- 1.9

1.3 +- 1.2

0.916

eGFR, mL/min/1.7 (missing at 2)

72.9 +- 37.2

75.1 +- 37.4

64.3 +- 35.4

0.016

Glucose, mg/dL (missing at 6)

144.3 +- 63.7

145.6 +- 63.6

139 +- 64.3

0.410

HbA1c, % (missing at 40)

5.8 +- 1.0

5.8 +- 0.9

5.6 +- 1.1

0.092

pH (missing at 6)

7.4 +- 0.1

7.4 +- 0.1

7.4 +- 0.1

0.674

HCO3, mmol/L (missing at 6)

20.6 +- 4.5

21.0 +- 4.4

19.0 +- 4.7

b0.01

Lactate, mmol/L (missing at 7)

2.2 +- 2.6

2.0 +- 2.4

3.1 +- 3.0

b0.01

hs-CRP, mg/L (missing at 2)

98.0 +- 82.3

95.0 +- 78.9

109.7 +- 93.8

0.138

PCT, ng/mL (missing at 82)

6.7 +- 18.3

3.0 +- 9.6

19.4 +- 31.3

b0.01

UA category (missing at 81)

0.004

b5

127 (37.2)

106 (39.4)

21 (29.2)

6-30

95 (27.9)

81 (30.1)

14 (19.4)

N30

119 (34.9)

82 (30.5)

37 (51.4)

ED LOS, h (missing at 0)

36.9 +- 36.1

36.8 +- 36.0

37. +- 36.6

0.947

Hospital LOS, day (missing at 0)

11.6 +- 14.7

11.4 +- 13.7

12.4 +- 18.7

0.592

ICU admission, % (missing at 0)

55 (13.0)

45 (13.4)

10 (11.6)

0.664

Ward admission, % (missing at 0)

231 (54.7)

181 (53.9)

50 (58.2)

0.478

Hospital death, % (missing at 0)

87 (20.6)

66 (19.6)

21 (24.1)

0.329

Abbreviations L-tube, Levin tube; PEG, percutaneous endoscopic gastrostomy; PICC, peripherally inserted central catheter; AVF, arteriovenous fistula; PTGBD, percutaneous transhepatic gallbladder drainage; ECOG, Eastern Cooperative Oncology Group; SBP, systolic blood pressure; DBP, diastolic blood pressure; PR, Pulse pressure; RR, respiratory pressure; SpO2, Blood oxygen saturation; NEWS, national early warning score; KTAS, Korean triage and acuity scale; SEG NEU, segmented form neutrophil; BUN, blood urea nitrogen; eGFR, estimated glomerular filtration rate; HCO3, bicarbonate; hs-CRP, high sensitive C-reactive protein; PCT, procalcitonin; UA, Urine analysis; ED LOS, emergency department length of stay; ICU, intensive care unit.

did not have blood cultures taken in the ED were excluded. Before he reviewed the medical chart, he was trained to select patient who were compatible with infection based on ICD 10 code diagnosis and 20% of the randomly selected patients’ data were reviewed by another author. In addition, patients who did not have blood cultures taken in the ED were excluded because we could not identify bacteremia.

Measurement and data collection

Data were collected from the patients’ electronic charts using a structured form for data collection according to the guidelines recom- mended by Gilbert et al. [13]. The senior resident of emergency medi- cine reviewed and abstracted the data. Data included the following: age; sex; concurrent conditions (hemiplegia, sore, Levin tube in situ, Foley catheter in situ, percutaneous endoscopic gastrostomy (PEG) in situ; tracheostomy in situ, peripherally inserted central catheter (PICC) in situ, arteriovenous fistula (AVF) in situ, and percutaneous transhepatic gallbladder drainage (PTGBD) in situ; Eastern Cooperative Oncology Group (ECOG) scale; comorbidities [hypertension (HTN), dia- betes mellitus (DM) – uncomplicated vs. end-organ damage, chronic liver disease (CLD) – mild vs. moderate to severe, chronic kidney disease (CKD), stroke, dementia, heart failure , malignancy – any leukemia or localized solid tumor vs. metastatic solid tumor, myocardial infarc- tion, chronic airway disease, and peptic ulcer disease]; Charlson co- morbidity index; presentation symptoms (fever, chill, respiratory symptoms, abdominal symptoms, urinary symptoms, other symp- toms); previous antimicrobial use; sepsis (none vs. sepsis without sep- tic shock vs. septic shock); physiology at ED presentation [systolic blood pressure (SBP), diastolic blood pressure , heart rate (HR), respira- tory rate (RR), body temperature, blood oxygen saturation(SpO2)], mentality using Alert Verbal Pain Unresponsive (AVPU) scale, and na- tional early warning score (NEWS); Korean triage and acuity scale (KTAS); initial laboratory findings of white blood cell count, seg- mented Neutrophil count, hematocrit (Hct), hemoglobin (Hb), platelet, sodium, albumin, Blood urea nitrogen , creatinine, estimate glo- merular filtration rate (eGFR), glucose, hemoglobin A1c (HbA1c), pH, bicarbonate (HCO3), lactate, high sensitivity C-reactive protein (hs- CRP), procalcitonin ; pyuria (b5 vs. 6-30 vs. N30); ED length of

stay(LOS); hospital LOS; disposition [admission to the intensive care unit (ICU) or ward]; and, hospital death.

Comorbidity status was based on the physician’s medical record. Most information were based on the guardian’s statement. Sepsis and septic shock diagnosis is based on the recent international consensus (Sepsis-3) [14].

Study measures

The primary measure was patients with confirmed blood stream in- fection. Isolation of coagulase-negative Staphylococcus, Corynebacte- rium, Propionbacterium sp., Bacillus non-anthracis, or Micrococcus in a single blood culture was considered contaminant and therefore ex- cluded from the case definition. However, single isolate of

S. epidermidis among patients with prosthesis or pacemaker was not regarded as contamination. Additionally, we determined the proportion of bacteremias from antimicrobial-resistant bacteria according to a pragmatic antimicrobials classification. R-Group 0: bacteria susceptible to penicillin and amoxicillin; R-Group 1: bacteria resistant to penicil- lin/amoxicillin, first, second, or third generation cephalosporins (e.g. cefazolin, ceftriaxone, cefotaxime). R-Group 2: ESBL-producing bacteria or bacteria resistant methicillin, fourth generation cephalosporin (e.g. cefepime), or fluoroquinolone (e.g. ciprofloxacin). R-Group 3: highly re- sistant pathogens including vancomycin resistant enterococci, carba- penem (e.g. meropenem) or colistin resistant Gram negatives. A single patient could therefore be included in N1 group for analysis. For exam- ple, a patient with an ESBL producing Klebsiella pneumoniae bacteremia from an isolate which is also resistant to ciprofloxacin was included in both groups, R-Groups 1 and 2. A patient with pan-resistant Pseudomo- nas bacteremia was included in R-Groups 1, 2 and 3.

Analysis

All continuous data are presented as the mean and standard devia- tion (SD), and discrete data are presented as both the count and per- centage. The results of logistic regression analyses are presented as the odds ratio with a 95% confidence interval. Statistical significance was defined as a two-sided p b 0.05.

Comparison of normally distributed data was performed using an in- dependent sample t-test. For non-normally distributed data, compari- sons were performed using the Mann-Whitney U test or the Kruskal- Wallis test. For categorical data, the Chi-square test was used. If neces- sary, the Chi-square test with a Fischer exact test for 2 x 2 tables was use. The results were considered significant at a threshold of p b 0.05 (two-tailed).

Associations between the presence of each measure (except R-group 0) and each potential variable were first quantified using univariate lo- gistic regression analysis. Next, multivariable logistic regression analysis was performed for trend factors (p b 0.05) in the univariate analysis. Re- gression results are expressed as ORs with 95% confidence intervals (CI). For the variable of interest, the area under the receiver operating characteristic (AUROC), sensitivity (SN), Specificity , positive likeli- hood ratio (+LR), negative likelihood ratio (-LR), positive predictive value (+PV), and negative predictive value (-PV) at various clinical

cut-off points were examined.

For the missing information, discrete variables were treated as dummy variable and continuous variable were imputed as the mean value of each variable. Thus all of the subjects were included in the final multivariate analysis. A sensitivity analysis was conducted with the complete data to identify whether the creating dummy variable and imputation would affect the result.

All analyses were performed using STATA 11.1 (StataCorp LP, TX, USA) and SAS 9.1 (SAS Institute Inc., Cary, NC, USA).

  1. Results

Four hundred and eighty-seven (63.7%) of 765 patients transferred from LTCH during the study period were suspected of having an infec- tion. A total of 422 patients were enrolled in the study after exclusion of 65 patients who did not have 2 or more sets of blood cultures per- formed at ED (Fig. 1).

Table 1 showed the characteristics of the enrolled patients. Mean age was 77.5 +- 10.7 years, and males ware 200 of the 422 enrolled (47.4%). There were 53 (12.6%) patients with hemiplegia and 158 (37.4%) with sores. Regarding catheters or any tube in situ, a Foley catheter was the most common (n = 91, 21.6%) followed by L-tube (n = 45, 10.7%) and PICC (n = 27, 6.4%). HTN was the most frequent chronic disease (n = 219, 51.9%), followed by stroke (n = 169, 40.1%) and dementia (n = 167, 39.6%). Respiratory symptoms were the most frequent symp- tom (n = 210, 49.8%) followed by abdominal symptoms (n = 131, 31.0%) and urinary symptoms (n = 40, 9.5%). Fifty patients (11.9%) had septic shock and 193 patients (45.7%) had sepsis without septic shock. Fifty-five patients (13.0%) were admitted to the ICU and 231 pa- tients (54.7%) were admitted to the ward. Hospital death occurred in 87 patients (20.6%). Most of the baseline characteristics and comorbidities were not significantly different between the bacteremia group and the non-bacteremia group. PICC in situ, fever/chill, urinary symptoms, and septic shock were more frequent in the bacteremia group. SBP and DBP were lower and NEWS and acuity as measured by KTAS was higher in the bacteremia group. segmented neutrophil count, procalcitonin, and lactate level were higher, and platelet count, albumin, and bicar- bonate level were lower, in the bacteremia group. Pyuria was more fre- quent in the bacteremia group. There were no significant difference in ED LOS, hospital LOS, ICU or ward admission, and hospital death be- tween the two groups.

Table 2 shows the microBiological characteristics of 86 patients with bacteremia. Among Gram (+) bacteria, S. aureus was the most fre- quently isolated (n = 10) followed by S. epidermidis (n = 8),

S. capitalis (n = 5) and S. hominis (N = 5). Forty-four Gram (+) bacteria were identified. Among Gram (-) bacteria, E.coli was the most fre- quently isolated (n = 25), and followed by K. pneumonia (n = 6). Forty-six Gram (-) bacteria were identified. Thus, when considering duplicate cases (n = 4), the actual number of bacteremic patients was

86. Majority were classified as R-Group 1 (37/44 (84.1%) in Gram (+), 38/46 (82.6%) in Gram (-), and 75/90 (83.3%)). There was a significant

Table 2

Microbiology of 90 bacterial isolates from 86 patients.

Species Bacteria classification Total R-Group 0 R-Group1 R-Group 2 R-Group 3

Gram (+) Staphylococcus aureus 10a 10 7a Coagulase negative Staph. epidermidis 8b,c 1 7c 6b

capitis 5c 5c 5

hominis

5

5

5

auricularis

2

1

1

caprae

1

1

1

wnarneri

1

1

1

Coagulase negative Staphylococcus

2

1

1

Enterococcus

faecalis

3

3

3

3

faeceum

2

2

2

1

Streptococcus

constellatus

1

1

gordonii

1

1

mitis

1

1

pneumoniae

1

1

1

viridans group

1

1

Total

44

5

37

33

4

Gram (-)

Escherichia coli

25a,b,d

2

23

22a,b,d

Klebsiella pneumoniae

6d

2

2d

Proteus mirabilis

2

2

2

1

Serratia marcescens

2

2

Stenotrophomonas maltophilia

2

Acinetobacter baumannii complex

1

1

1

Alcaligeness faecalis spp

1

1

1

Burkholderia cepacia

1

1

Elizabethkingia meningoseptica

1

1

1

1

Enterobacter agrogenes

1

1

Hemophilus influenza

1

1

1

Morganella morganni

1

1

Providencia stuartii

1

1

1

Psedomonas aeruginosa

1

1

1

Total

46

2

38

32

a, b, c, and d denote that bacteria of same alphabet was isolated from the same patient.

overlap in bacterial species belonging to both R-Groups 1 and 2 with

E. coli (22 of 23 isolates) due to quinolone resistance, and Staphylococci (22 of 26 isolates) due to both quinolone and methicillin resistance. Seven out of 10 S. aureus isolates (70%) were methicillin resistant (MRSA). There were only 6 bacteria in R-Group 3 including 3 vancomy- cin resistant Enterococci.

Results of multivariable logistic regression analysis are shown in Table 3. Although various potential factors showing a trend in univariable logistic analysis were tested, only procalcitonin level showed a significant relationship with bacteremia (AOR 1.03, 95% CI 1.00-1.05, p = 0.010), while the PICC and septic shock showed a trend (AOR 2.63, 95% CI 0.98-7.04, p = 0.055, and AOR 2.93, 95% CI

0.85-10.14, p = 0.089, respectively). Procalcitonin level and abdominal symptoms were significantly associated with R-Group1 (AOR 1.03, 95% CI 1.01-1.05, p = 0.005, and AOR 0.40, 95% CI 0.18-0.90, p = 0.026, re-

spectively). Procalcitonin level and chill were significantly associated with R-Group 2 (AOR 1.02, 95% CI 1.00-1.04, p = 0.030, and AOR 2.33,

95% CI 1.01-5.35, p = 0.047, respectively). Septic shock showed a trend for R-Group 2 (AOR 3.64, 95% CI 0.93-14.29, p = 0.064). Urinary symptoms showed a significant association with R-Group (AOR 8.37,

95% CI 1.53-45.74, p = 0.014). We confirmed that the results were not much different after a sensitivity analysis which was conducted with the complete data to identify whether the creating dummy vari- able and imputation would affect the result (Appendix Table 1).

Next, we evaluated the characteristics of procalcitonin level further. Fig. 2 shows the distribution of each measure according to the tertile of procalcitonin level. Incidence of bacteremia was 11.4% at 1st tertile, 14.2% at 2nd tertile, and 41.6% at 3rd tertile. Incidence of R1 group was 8.8% at 1st tertile, 9.7% at 2nd tertile, and 32.7% at 3rd tertile. Inci- dence of R2 group was 7.9% at 1st tertile, 9.7% at 2nd tertile, and 29.2% at 3rd tertile. Incidence of R3 group was 0.9% at 1st tertile, 0.9% at 2nd tertile, and 2.7% at 3rd tertile. The AUROC was determined to be 0.729 (95% CI 0.657-0.800) for the presence of bacteremia, 0.733 (95% CI

0.653-0.812) for R-Group 1, 0.728 (95% CI 0.646-0.809) for R-Group

2, and 0.691 (95% CI 0.351-1.000) for R-Group 3. Table 4 showed the di- agnostic performance of procalcitonin at the various clinical cut-off values for each outcome. +LRs for each outcome at procalcitonin levels of 5.0 and 10.0 were approximately 3.0 and 5.0, respectively. Multivar- iable logistic regression analyses using procalcitonin as tertile were shown in Appendix Tables 2 and 3.

Table 3

Logistic regression analysis for antibiotic resistance (422 patients were entered into analysis).

Variable

Bacteremia (n = 86) AOR

p-Value

R-Group 1

(n = 71) AOR

p-Value

R-Group 2 (n = 61) AOR

p-Value

R-Group 3

(n = 6) AOR

p-Value

L-tube, %

2.26 (0.79-6.47)

0.128

1.91 (0.66-5.57)

0.233

PICC, %

2.63 (0.98-7.04)

0.055

1.48 (0.45-4.83)

0.520

1.48 (0.44-4.94)

0.526

ECOG

0

Reference

Reference

Reference

1

1.11 (0.38-3.23)

0.844

0.86 (0.27-2.72)

0.792

0.61 (0.18-2.06)

0.429

2

1.39 (0.42-4.51)

0.585

1.14 (0.31-4.16)

0.842

0.90 (0.23-3.45)

0.872

3

1.52 (0.51-4.54)

0.452

0.97 (0.29-3.28)

0.966

0.78 (0.23-2.71)

0.701

4

1.97 (0.70-5.89)

0.201

1.68 (0.54-5.21)

0.370

1.43 (0.46-4.47)

0.537

Fever, %

1.21 (0.57-2.54)

0.623

Chill, %

1.66 (0.75-3.71)

0.213

2.02 (0.92-4.46)

0.081

2.33 (1.01-5.35)

0.047

Abdominal symptoms, %

0.40 (0.18-0.90)

0.026

Urinary symptoms, %

1.42 (0.55-3.66)

0.470

1.34 (0.51-3.55)

0.553

2.16 (0.82-5.75)

0.121

8.37 (1.53-45.74)

0.014

Other symptoms, %

0.99 (0.47-2.07)

0.983

Sepsis

None

Reference

Reference

Reference

Sepsis without septic shock

1.18 (0.57-2.48)

0.655

0.68 (0.30-1.55)

0.361

0.94 (0.39-2.24)

0.889

Septic shock

2.93 (0.85-10.14)

0.089

2.38 (0.65-8.68)

0.188

3.64 (0.93-14.29)

0.064

Systolic blood pressure, mm Hg

1.00 (0.99-1.01)

0.701

1.00 (0.98-1.01)

0.647

1.00 (0.98-1.01)

0.639

Body temperature, ?C

1.18 (0.78-1.79)

0.438

Mentality

Alert, %

Reference

Reference

Reference

Verbal, %

1.12 (0.45-2.81)

0.804

1.79 (0.66-4.83)

0.253

2.39 (0.85-6.72)

0.098

Pain, %

1.04 (0.34-3.18)

0.944

1.28 (0.36-4.52)

0.704

1.50 (0.40-5.56)

0.547

Unresponsive, %

2.18 (0.31-15.61)

0.437

3.05 (0.26-36.23)

0.377

3.43 (0.29-40.13)

0.326

NEWS

0.98 (0.87-1.09)

0.694

0.95 (0.83-1.08)

0.429

0.93 (0.81-1.07)

0.294

KTAS

1

2.12 (0.28-15.85)

0.464

1.27 (0.17-9.32)

0.817

2

1.15 (0.17-8.00)

0.887

0.80 (0.12-5.47)

0.818

3

2.01 (0.31-12.81)

0.461

1.69 (0.27-10.28)

0.571

4 & 5

Reference

Reference

SEG NEU %

1.01 (0.99-1.03)

0.317

Platelet, 103/uL

1.00 (0.99-1.00)

0.021

1.00 (1.00-1.00)

0.291

1.00 (0.99-1.00)

0.142

Albumin, g/dl

0.61 (0.37-1.03)

0.065

0.67 (0.38-1.18)

0.169

0.69 (0.38-1.24)

0.213

eGFR, mL/min/1.7

1.00 (0.99-1.01)

0.825

0.99 (0.98-1.00)

0.186

1.00 (0.99-1.01)

0.642

HCO3, mmol/L

0.94 (0.87-1.03)

0.179

0.95 (0.86-1.04)

0.237

0.94 (0.86-1.03)

0.216

Lactate, mmol/L

0.99 (0.85-1.16)

0.921

0.92 (0.78-1.09)

0.336

0.91 (0.76-1.08)

0.279

PCT, ng/mL

1.03 (1.01-1.05)

0.010

1.03 (1.01-1.05)

0.005

1.02 (1.00-1.04)

0.030

1.02 (0.99-1.04)

0.193

UA category

b5

Reference

Reference

Reference

6-30

0.66 (0.29-1.54)

0.341

0.77 (0.31-1.96)

0.590

0.78 (0.28-2.19)

0.639

N30

1.27 (0.58-2.73)

0.548

1.25 (0.54-2.88)

0.604

1.67 (0.68-4.10)

0.264

99

1.12 (0.47-2.65)

0.804

0.69 (0.24-1.97)

0.487

0.89 (0.30-2.64)

0.833

Abbreviations AOR, adjusted odd ratio; L-tube, Levin tube; PICC, peripherally inserted central catheter; ECOG, Eastern Cooperative Oncology Group; SBP, systolic blood pressure; DBP, di- astolic blood pressure; PR, pulse pressure; RR, respiratory pressure; SpO2, blood oxygen saturation; NEWS, national early warning score; KTAS, Korean triage and acuity scale; SEG NEU, segmented form neutrophil; eGFR, estimated glomerular filtration rate; HCO3, bicarbonate; PCT, procalcitonin; UA, urine analysis; ED LOS, emergency department length of stay; ICU, in- tensive care unit.

Image of Fig. 2

Fig. 2. Distribution of each outcome according to the procalcitonin level tertile.

The incidence of bacteremia of the top five LTCHs was compared. There were no significant differences in the incidence of bacteremia be- tween the top five LTCHs. (Table 5).

  1. Discussion

Our study provides insight about the prevalence of bacteremia, antimicrobial resistance and characteristics of bacteremic patients transferred from LTCHs with suspected infection. Key findings

from our cohort can be summarized as follows. Firstly, 1 in 5 pa- tients with suspected infection were bacteremic. Secondly, multi- resistant pathogens causing blood stream infections were very common, mostly ESBL producing Enterobacteriaceae, MRSA and flu- oroquinolone resistant Gram negatives. Thirdly, an elevated procalcitonin level was a significant factor for the presence of bac- teremia, as well as R-Group 1 and 2. These findings potentially have important implications in deciding empiric antibiotic treat- ment in this selected High risk patient group.

Table 4

Diagnostic performance of procalcitonin for each outcome.

Procalcitonin level

Outcome

SN

(95% CI)

SP

(95% CI)

+LR (95% CI)

-LR (95% CI)

+PV (95% CI)

-PV (95% CI)

N0.507 ng/mL

G1

72.4

59.5

1.8

0.5

34.0

88.2

(60.9-82.0)

(53.3-65.4)

(1.5-2.2)

(0.3-0.7)

(26.7-41.8)

(82.5-92.5)

G2

74.1

57.8

1.8

0.5

26.5

91.6

(61.0-84.7)

(51.8-63.6)

(1.4-2.2)

(0.3-0.7)

(19.9-34.0)

(86.5-95.2)

G3

73.6

57.1

1.7

0.5

24.1

92.1

(59.7-84.7)

(51.2-62.9)

(1.4-2.1)

(0.3-0.7)

(17.7-31.4)

(87.2-95.6)

G4

80.0

52.8

1.7

0.4

2.5

99.4

(28.4-99.5)

(47.3-58.3)

(1.1-2.7)

(0.1-2.2)

(0.7-6.2)

(96.9-100.0)

N1 ng/mL

G1

64.5

72.4

2.3

0.5

40.2

87.6

(52.7-75.1)

(66.5-77.7)

(1.8-3.0)

(0.4-0.7)

(31.4-49.4)

(82.5-91.7)

G2

65.5

70.2

2.2

0.5

31.1

90.8

(51.9-77.5)

(64.5-75.5)

(1.7-2.8)

(0.3-0.7)

(23.1-40.2)

(86.2-94.3)

G3

64.2

69.3

2.1

0.5

27.9

91.3

(49.8-76.9)

(63.6-74.6)

(1.6-2.7)

(0.4-0.7)

(20.1-36.7)

(86.7-94.7)

G4

60.0

64.5

1.7

0.6

2.5

99.1

(14.7-94.7)

(59.1-69.6)

(0.8-3.5)

(0.2-1.8)

(0.5-7.0)

(96.7-99.9)

N4.62 ng/mL

G1

46.1

87.5

3.7

0.6

51.5

84.9

(34.5-57.9)

(82.9-91.2)

(2.5-5.5)

(0.5-0.8)

(39.0-63.8)

(80.1-89.0)

G2

46.6

85.5

3.2

0.6

39.7

88.6

(33.3-60.1)

(80.8-89.4)

(2.2-4.8)

(0.5-0.8)

(28.0-52.3)

(84.2-92.1)

G3

45.3

84.7

3.0

0.7

35.3

89.3

(31.6-59.6)

(80.0-88.6)

(2.0-4.4)

(0.5-0.8)

(24.1-47.8)

(85.0-92.7)

G4

60.0

80.6

3.1

0.5

4.4

99.3

(14.7-94.7)

(75.9-84.7)

(1.5-6.5)

(0.2-1.5)

(0.9-12.4)

(97.4-99.9)

N10.15 ng/mL

G1

35.5

93.2

5.2

0.7

60.0

83.4

(24.9-47.3)

(89.4-95.9)

(3.0-8.9)

(0.6-0.8)

(44.3-74.3)

(78.6-87.5)

G2

39.7

92.2

5.1

0.7

51.1

88.1

(27.0-53.4)

(88.4-95.0)

(3.0-8.5)

(0.5-0.8)

(35.8-66.3)

(83.9-91.6)

G3

39.6

91.6

4.7

0.7

46.7

89.2

(26.5-54.0)

(87.8-94.6)

(2.9-7.9)

(0.5-0.8)

(31.7-62.1)

(85.0-92.5)

G4

60.0

87.5

4.8

0.5

6.7

99.3

(14.7-94.7)

(83.4-90.8)

(2.2-10.3)

(0.2-1.3)

(1.4-18.3)

(97.6-99.9)

Table 5

Comparison of bacteremia and antibiotic resistance between the top 5 long term care hospitals.

Hospital

N

G1

G2

G3

G4

A

44

10 (22.7%)

8 (18.2%)

8 (18.2%)

B

26

7 (26.9%)

6 (23.1%)

5 (19.2%)

C

24

5 (20.8%)

5 (20.8%)

4 (16.7%)

1 (4.2%)

D

21

2 (9.5%)

1 (4.8%)

1 (4.8%)

1 (4.8%)

E

18

3 (16.7%)

2 (11.1%)

2 (11.1%)

p-Value

0.657

0.444

0.614

0.313

Long-term care (LTC) is defined as a variety of medical and non- medical services for people with a chronic illness or disability who can- not care for themselves for long periods of time [15]. Therefore, LTC ser- vices encompass non-skilled care, such as daily activity assistance including dressing, feeding, or bathing, as well as medical care up to prolonged mechanical ventilator support. Provision of LTC can happen at home, in the community, in assisted living facilities, in nursing homes, in long-term care facilities, or in long-term care hospitals. In Korea, LTCH are distinguished from others because they are staffed with doctors and nurses 24 h a day.

In the United States, chronic, critically ill patients accounted for a substantial number of ICU resources, up to 40% [16]. Since 1983, this burden has been relieved substantially by the rapid growth of LTCHs [17]. Many patients requiring prolonged mechanical ventilation after acute illness have been adequately managed in LTCHs. However, in Korea, illnesses do not appear to be as severe as they are in the US. In the present study, approximately 40% of patients were ECOG grade 0-2, meaning that they were ambulatory and capable of all self-care at a minimum [18,19]. Co-morbidity status appears to be the reason for the admission to the LTHC in this cohort, considering that their Charlson comorbidity point was relatively high (5 point) [20,21].

Acute exacerbation of a previous disease or a newly occurring illness can be managed initially in the LTCHs in Korea because they are staffed by doctors and nurses, and empirical antimicrobials could be started for the infectious patients. However, if the patients were insufficiently re- sponsive, they were transferred to the referral center. After the patient’s condition is stabilized, almost of all return to the LTCHs where they were admitted previously. Since the pre-conditions were not changed, though they were discharged from the referral center, they were still in- trinsically vulnerable to acute exacerbation of a previous disease or a newly occurred illness. Therefore, the cycle between LTCHs and the re- ferral center is initiated. The transfer of infected or colonized patients with antimicrobial resistant pathogens from acute-care facilities to LTCHs is already a well-known phenomenon [22].

Therefore, notable and unique features can be found in LTCH pa- tients related to infection: 1) elderly with many comorbidities, 2) exces- sive exposure to broad-spectrum antimicrobials, and 3) likely to be exposed to patients infected with antimicrobial resistant pathogens or to surroundings colonized with antimicrobial resistant pathogens. As a consequence, it can be expected that the rate of infective complications would be high and in particular, bacteremia caused by antibiotic- resistant bacteria.

Studies that evaluated the incidence of bacteremia among LTC facil- ities were reported decades ago. In 1994, Nicolle et al. reported a rate of bacteremia of 6.3% (29/459) in long-term care facilities over an 8-year periods [4]. In 1992, Murder et al. reported a bacteremia rate of 0.20 to 0.36 cases/1000 patient-days from 1985 to 1989 [5]. In 1984, Setia et al. reported a bacteremia rate of 0.3 cases/1000 patient-days [6]. In the present study, we revisit the incidence of bacteremia among LTCH Declaration of interest“>patients and a higher rate of bacteremia, approximately 20%, was found. This finding suggests that the severity of infectious patients in LTCHs has been worsening.

The issue of infection by antimicrobial resistant bacteria among LTC facilities came to light more recently. Regers et al. reported the annual

incidence of antimicrobial-resistant infection was 12.7 cases per 1000 long-term residents in California, Florida, Michigan, New York and Texas [23]. Recent studies focus on the rate of colonization of various an- timicrobial resistant bacteria, such as VRE [24], CRE [25], or multi-drug resistant strains [26]. In the present study, approximately 70% of isolates were either ESBL-producing or fluoroquinolone resistant Gram nega- tives, or MRSA (R-Group 2). Prevalence of VRE, and carbapenem and co- listin resistant Gram negatives was relatively low (6.7%). This finding has important implications for an emergency or acute physician in terms of appropriate empiric antibiotic choices, especially considering a significantly higher rate of septic shock (29.1%) and trend towards high in-hospital mortality (24.1%) in the bacteremic group.

The findings of the present study have a notable clinical implication. Approximately 70% of the bacteria isolated from blood cultures were from the R-Group 2, meaning they were ESBL-producing or resistant to methicillin or fluoroquinolone. Therefore, if the result of the Gram stain was reported, physicians should carefully consider changing the empirical antimicrobials to vancomycin or carbapenem. Of course, anti- microbial change should be based on multifaceted and meticulous con- sideration, not a single factor. Additionally, misuse of vancomycin or carbapenem can lead to the emergence of resistant bacteria, such as VRE or CRE. However, from Gram stain to bacterial identification and antimicrobial susceptibility testing generally takes 24-48 h, and this would be the “golden time” for patients in critical condition [27]. Physi- cians should consider these two conflicting possibilities carefully in their practice. The other point is that the higher the initial procalcitonin level is, the higher the incidence of bacteremia is, as well as that of R- Group 2.

Additionally, among potential factors including well-used in- flammatory markers, only procalcitonin was associated exclusively with bacteremia and R-Group 1 and 2. Several studies have exam- ined the performance of procalcitonin as a biomarker for predicting bacteremia [28,29]. In addition with those studies, we noted a trend towards relatively higher procalcitonin levels in R-Groups 1 and 2. Physicians should carefully choose the empirical antimicro- bials for patients transferred from LTCHs whose procalcitonin level is high.

Our study had several notable limitations. First, this is a retro- spective chart review single center study. Incidence of bacteremia and antimicrobial resistant bacteria would differ between regions and nations according to baseline characteristics, co-morbidities, and national systems. Local epidemiological surveillance systems would also differ. Therefore, multicenter and multi-national studies are warranted. Second, there would be unknown confounding fac- tors, although we collected extensive variables. We did not collect data on the type of empirical antimicrobials started before transfer or the patients’ colonization history. Factors related to baseline characteristics or co-morbidities were found not to be associated with outcomes in the present study, raising concerns with other po- tential pre-morbid conditions of LTCH patients. Notably, the rate of antimicrobial use in LTCHs before transfer was 28.2% in the present study, seemingly a low rate, considering that LTCHs were staffed with doctors. Missed documentation or mal- or under-medical practice was possible causes. Third, the number of R-Group 3 was very small, thus the statistical results regarding the R-Group 3 should be interpreted carefully.

In conclusion, the prevalence of bloodstream infections in patients admitted from LTCH was high (20.4%) with majority of these infections from resistant bacteria. Procalcitonin levels were significantly higher in bacteremic patients with an increasing trend towards bacteria in the an- timicrobial resistant groups.

Declaration of interest

The authors report no conflicts of interest. The authors are responsi- ble for the content and writing of the paper.

Acknowledgement

We specially thank to Er. Eun Ho Choe, who helped the discussion.

Conflict of interests

None.

Funding and support

None.

Authors’ contributions

S.K. and S.J. designed this study. S.K. collected the data. S.J. super- vised the overall data collection process, had full access to all data in the study, and takes responsibility for the integrity of the data. B.P. con- ducted the data analysis. S.J. and H.B. wrote the initial draft of the article. All authors provided substantial review and feedback on the final ver- sion of the article. S.J. takes responsibility for the paper as a whole.

All authors have read and approved the submitted manuscript. This manuscript has not been submitted or published elsewhere in whole or in part, except as an abstract (if relevant).

Appendix Table 1

Logistic regression analysis for antibiotic resistance with complete data.

Variable

Bacteremia (n = 61) AOR

(available at 285)

p-value

R-Group 1

(n = 49) AOR

(available at 276)

p-value

R-Group 2 (n = 45) AOR

(available at 276)

p-value

R-Group 3

(n = 6) AOR

(available at 422)

p-value

L-tube, %

4.22 (1.24-14.42)

0.022

3.97 (1.11-14.22)

0.034

PICC, %

3.54 (0.95-13.15)

0.060

1.50 (0.34-6.62)

0.595

1.23 (0.25-5.93)

0.799

ECOG

0

Reference

Reference

Reference

1

1.13 (0.43-4.05)

0.869

1.39 (0.28-6.79)

0.687

0.82 (0.16-4.32)

0.816

2

2.46 (0.56-10.88)

0.236

2.80 (0.52-15.20)

0.232

1.96 (0.33-11.74)

0.461

3

2.48 (0.60-10.22)

0.208

2.21 (0.42-11.49)

0.346

1.85 (0.34-10.07)

0.477

4

2.12 (0.54-8.35)

0.281

2.43 (0.50-11.89)

0.272

1.95 (0.40-9.49)

0.407

Fever, %

1.39 (0.58-3.34)

0.465

Chill, %

1.89 (0.77-4.65)

0.163

1.77 (0.68-4.57)

0.236

2.17 (0.78-6.09)

0.140

Abdominal symptoms, %

0.32 (0.11-0.90)

0.031

Urinary symptoms, %

1.04 (0.33-3. 27)

0.953

0.96 (0.29-3.19)

0.951

1.71 (0.49-6.01)

0.404

10.24 (2.00-52.57)

0.005

Other symptoms, %

1.02 (0.38-2.74)

0.972

Sepsis

None

Reference

Reference

Reference

Sepsis without septic shock

1.25 (0.49-3.22)

0.636

0.84 (0.30-2.30)

0.731

1.44 (0.46-4.47)

0.531

Septic shock

2.08 (0.46-9.47)

0.344

2.06 (0.43-9.86)

0.366

3.87 (0.67-22.20)

0.129

Systolic blood pressure, mm Hg

1.00 (0.99-1.01)

0.955

0.99 (0.98-1.01)

0.421

0.99 (0.97-1.01)

0.210

Pulse rate, bpm

1.01 (0.99-1.03)

0.439

1.01 (0.99-1.03)

0.339

Mentality

Alert, %

Reference

Reference

Reference

Verbal, %

1.14 (0.37-3.50)

0.824

1.59 (0.48-5.25)

0.446

2.53 (0.69-9.26)

0.160

Pain, %

1.83 (0.44-7.54)

0.404

1.36 (0.28-6.59)

0.702

2.02 (0.36-11.26)

0.422

Unresponsive, %

0.99 (0.08-11.58)

0.992

0.85 (0.05-14.88)

0.914

0.73 (0.04-12.63)

0.828

NEWS

0.92 (0.79-1.08)

0.305

0.87 (0.73-1.04)

0.126

0.81 (0.66-1.00)

0.047

KTAS

1

1.05 (0.09-12.49)

0.968

0.53 (0.41-6.79)

0.625

2

0.87 (0.08-9.26)

0.907

0.55 (0.05-6.05)

0.625

3

1.32 (0.14-12.25)

0.809

1.05 (0.12-9.45)

0.968

4 & 5

Reference

Reference

SEG NEU %

1.02 (0.99-1.05)

0.228

Platelet, 103/uL

1.00 (0.99-1.00)

0.045

1.00 (0.99-1.00)

0.062

0.99 (0.99-1.00)

0.009

Albumin, g/dL

0.55 (0.28-1.07)

0.079

0.79 (0.39-1.62)

0.527

0.87 (0.40-1.85)

0.711

eGFR, mL/min/1.7

1.00 (0.99-1.01)

0.746

0.99 (0.98-1.01)

0.384

1.00 (0.99-1.01)

0.992

HCO3, mmol/L

0.91 (0.82-1.02)

0.093

0.93 (0.83-1.04)

0.206

0.91 (0.80-1.03)

0.116

Lactate, mmol/L

0.97 (0.80-1.17)

0.742

0.95 (0.77-1.15)

0.593

0.92 (0.74-1.14)

0.456

PCT, ng/mL

1.03 (1.01-1.05)

0.014

1.03 (1.01-1.06)

0.003

1.02 (1.00-1.04)

0.035

UA category

b5

Reference

Reference

Reference

6-30

0.43 (0.16-1.18)

0.101

0.72 (0.24-2.16)

0.556

0.63 (0.18-2.25)

0.476

N30

1.24 (0.53-2.90)

0.625

1.28 (0.49-3.33)

0.612

1.71 (0.60-4.93)

0.317

Abbreviations AOR, Adjusted odd ratio; L-tube, Levin tube; PICC, peripherally inserted central catheter; ECOG, Eastern Cooperative Oncology Group; SBP, systolic blood pressure; DBP, diastolic blood pressure; PR, pulse pressure; RR, respiratory pressure; SpO2, blood oxygen saturation; NEWS, national early warning score; KTAS, Korean triage and acuity scale; SEG NEU, segmented form neutrophil; eGFR, estimated glomerular filtration rate; HCO3, bicarbonate; PCT, procalcitonin; UA, urine analysis; ED LOS, emergency department length of stay; ICU, intensive care unit.

Appendix Table 2

Logistic regression analysis for antibiotic resistance using procalcitonin as tertile (422 patients were entered into analysis).

Variable

Bacteremia

p-value

R-Group 1

p-value

R-Group 2 (n = 61)

p-value

R-Group 3 p-value

(n = 86) AOR

(n = 71) AOR

AOR

(n = 6) AOR

L-tube, %

2.34 (0.80-6.86)

0.120

1.98 (0.67-5.88)

0.217

PICC, %

ECOG

2.99 (1.06-8.43)

0.038

1.58 (0.48-5.17)

0.452

1.65 (0.50-5.50)

0.415

Appendix Table 2 (continued)

Variable

Bacteremia (n = 86) AOR

p-value

R-Group 1

(n = 71) AOR

p-value

R-Group 2 (n = 61) AOR

p-value

R-Group 3

(n = 6) AOR

p-value

0

Reference

Reference

Reference

1

0.95 (0.33-2.74)

0.929

0.73 (0.23-2.28)

0.587

0.57 (0.17-1.91)

0.366

2

1.12 (0.34-3.65)

0.850

0.89 (0.24-3.26)

0.864

0.77 (0.20-2.97)

0.702

3

1.24 (0.41-3.74)

0.700

0.79 (0.23-2.65)

0.697

0.71 (0.20-2.46)

0.586

4

1.64 (0.64-4.19)

0.291

1.50 (0.49-4.58)

0.474

1.37 (0.44-4.26)

0.582

Fever, %

1.12 (0.52-2.41)

0.767

Chill, %

1.65 (0.72-3.78)

0.235

1.89 (0.85-4.21)

0.121

2.21 (0.95-5.14)

0.067

Abdominal symptoms, %

0.45 (0.21-0.98)

0.044

Urinary symptoms, %

1.64 (0.64-4.19)

0.305

1.69 (0.65-4.37)

0.279

2.48 (0.94-6.49)

0.065

10.24

0.005

(2.00-52.67)

Other symptoms, %

1.03 (0.49-2.17)

0.945

Sepsis

None

Reference

Reference

Reference

Sepsis without septic shock

1.07 (0.51-2.25)

0.860

0.62 (0.27-1.41)

0.255

0.85 (0.36-2.03)

0.719

Septic shock

2.18 (0.63-7.58)

0.220

1.74 (0.48-6.33)

0.402

2.78 (0.71-10.79)

0.140

Systolic blood pressure, mm Hg

1.00 (0.99-1.01)

0.666

1.00 (0.98-1.01)

0.590

1.00 (0.98-1.01)

0.511

Body temperature, ?C

1.18 (0.77-1.82)

0.444

Mentality

Alert, %

Reference

Reference

Reference

Verbal, %

1.04 (0.40-2.67)

0.937

1.74 (0.64-4.77)

0.279

2.29 (0.82-6.44)

0.116

Pain, %

1.05 (0.33-3.27)

0.937

1.34 (0.38-4.69)

0.649

1.48 (0.40-5.45)

0.557

Unresponsive, %

2.59 (0.27-24.67)

0.407

3.86 (0.34-44.19)

0.277

4.02 (0.36-45.60)

0.261

NEWS

0.97 (0.86-1.09)

0.626

0.95 (0.83-1.08)

0.437

0.93 (0.81-1.07)

0.308

KTAS

1

1.03 (0.16-6.61)

0.977

1.68 (0.26-10.87)

0.588

1.07 (0.16-7.07)

0.946

2

0.79 (0.13-4.68)

0.795

0.89 (0.15-5.42)

0.903

0.66 (0.11-4.06)

0.650

3

1.54 (0.28-8.51)

0.621

1.79 (0.32-10.08)

0.507

1.55 (0.28-8.56)

0.616

4 & 5

Reference

Reference

Reference

SEG NEU %

1.01 (0.99-1.03)

0.351

Platelet, 103/uL

1.00 (0.99-1.00)

0.020

1.00 (0.99-1.00)

0.197

1.00 (0.99-1.00)

0.087

Albumin, g/dL

0.58 (0.34-0.99)

0.044

0.68 (0.38-1.19)

0.178

0.70 (0.39-1.26)

0.232

eGFR, mL/min/1.7

1.00 (0.99-1.01)

0.996

0.99 (0.98-1.00)

0.274

1.00 (0.99-1.01)

0.750

HCO3, mmol/L

0.95 (0.87-1.04)

0.250

0.95 (0.86-1.04)

0.259

0.94 (0.86-1.04)

0.211

Lactate, mmol/L

1.03 (0.88-1.20)

0.718

0.96 (0.82-1.13)

0.620

0.93 (0.79-1.11)

0.424

PCT, ng/mL

1st tertile

Reference

Reference

Reference

2nd tertile

1.17 (0.56-2.43)

0.672

1.75 (0.76-4.01)

0.187

1.75 (0.73-4.15)

0.207

3rd tertile

3.45 (1.36-8.76)

0.009

4.48 (1.63-12.29)

0.004

2.97 (1.03-8.56)

0.043

UA category

b5

Reference

Reference

Reference

6-30

0.65 (0.27-1.54)

0.325

0.70 (0.27-1.79)

0.455

0.73 (0.26-2.05)

0.552

N30

1.27 (0.58-2.79)

0.546

1.25 (0.54-2.90)

0.610

1.67 (0.68-4.12)

0.264

99

1.25 (0.52-2.99)

0.622

0.79 (0.28-2.23)

0.651

0.97 (0.33-2.86)

0.951

Abbreviations AOR, Adjusted odd ratio; L-tube, Levin tube; PICC, peripherally inserted central catheter; ECOG, Eastern Cooperative Oncology Group; SBP, systolic blood pressure; DBP, di- astolic blood pressure; PR, pulse pressure; RR, respiratory pressure; SpO2, blood oxygen saturation; NEWS, national early warning score; KTAS, Korean triage and acuity scale; SEG NEU, segmented form neutrophil; eGFR, estimated glomerular filtration rate; HCO3, bicarbonate; PCT, procalcitonin; UA, urine analysis; ED LOS, emergency department length of stay; ICU, in- tensive care unit.

Appendix Table 3

Logistic regression analysis for antibiotic resistance using procalcitonin as tertile with complete data.

Variable Bacteremia

p-Value

R-Group 1

p-Value

R-Group 2 (n = 45)

p-Value

R-Group 3

p-Value

(n = 61) AOR

(available at 285)

(n = 49) AOR

(available at 276)

AOR

(available at 276)

(n = 6) AOR

(available at 422)

L-tube, %

4.08 (1.17-14.15)

0.027

3.78 (1.05-13.65)

0.042

PICC, % 4.49 (1.15-17.63)

ECOG

0 Reference

0.031

1.88 (0.42-8.42)

Reference

0.409

1.60 (0.34-7.61)

Reference

0.553

1 1.28 (0.33-5.03)

0.724

1.55 (0.34-7.02)

0.571

0.91 (0.18-4.59)

0.907

2 2.53 (0.58-11.09)

0.218

2.70 (0.53-13.91)

0.234

1.88 (0.32-10.98)

0.483

3 2.53 (0.62-10.29)

0.196

2.06 (0.42-10.07)

0.374

1.69 (0.32-8.94)

0.539

4 2.24 (0.59-8.57)

0.237

2.50 (0.56-11.25)

0.231

2.03 (0.43-9.52)

0.367

Fever, % 1.39 (0.58-3.34)

0.462

Chill, % 1.79 (0.73-4.37)

0.203

1.64 (0.63-4.25)

0.309

2.05 (0.73-5.74)

0.171

Abdominal symptoms, %

0.40 (0.15-1.03)

0.057

Urinary symptoms, % 1.19 (0.39-3.64)

0.761

1.21 (0.38-3.90)

0.747

1.86 (0.54-6.38)

0.327

10.24 (2.00-52.57)

0.005

Other symptoms, % Sepsis

None Reference

Reference

1.04 (0.38-2.82)

Reference

0.944

(continued on next page)

Appendix Table 3 (continued)

Variable

Bacteremia (n = 61) AOR

(available at 285)

p-Value

R-Group 1

(n = 49) AOR

(available at 276)

p-Value

R-Group 2 (n = 45) AOR

(available at 276)

p-Value

R-Group 3

(n = 6) AOR

(available at 422)

p-Value

Sepsis without septic shock

1.35 (0.51-3.53)

0.544

0.87 (0.31-2.41)

0.784

1.42 (0.46-4.42)

0.546

Septic shock

1.84 (0.41-8.19)

0.422

1.42 (0.31-6.57)

0.653

2.71 (0.49-14.94)

0.251

Systolic blood pressure, mm Hg

1.00 (0.99-1.01)

0.955

0.99 (0.98-1.01)

0.475

0.99 (0.97-1.01)

0.212

Pulse rate, bpm

1.01 (0.99-1.03)

0.352

1.01 (0.99-1.03)

0.275

Mentality

Alert, %

Reference

Reference

Reference

Verbal, %

1.02 (0.33-3.20)

0.967

1.46 (0.43-4.94)

0.545

2.40 (0.66-8.82)

0.185

Pain, %

1.87 (0.46-7.59)

0.383

1.35 (0.29-6.41)

0.703

1.97 (0.37-10.64)

0.430

Unresponsive, %

1.51 (0.14-15.79)

0.731

0.98 (0.06-15.54)

0.990

0.79 (0.05-13.10)

0.868

NEWS

0.93 (0.80-1.07)

0.300

0.87 (0.73-1.03)

0.100

0.81 (0.67-0.99)

0.041

KTAS

1

0.49 (0.05-4.68)

0.537

0.33 (0.03-3.78)

0.370

2

0.34 (0.04-2.95)

0.329

0.29 (0.03-2.91)

0.293

3

0.78 (0.11-5.85)

0.813

0.74 (0.09-5.97)

0.777

4 & 5

Reference

Reference

SEG NEU %

1.02 (0.99-1.05)

0.322

Platelet, 103/uL

1.00 (0.99-1.00)

0.023

0.99 (0.99-1.00)

0.015

0.99 (0.99-1.00)

0.003

Albumin, g/dL

0.56 (0.28-1.10)

0.090

0.81 (0.40-1.67)

0.572

0.87 (0.41-1.88)

0.728

eGFR, mL/min/1.7

1.00 (0.99-1.01)

0.992

1.00 (0.98-1.01)

0.574

1.00 (0.99-1.02)

0.905

HCO3, mmol/L

0.92 (0.83-1.03)

0.154

0.94 (0.84-1.05)

0.291

0.12 (0.81-1.03)

0.138

Lactate, mmol/L

1.02 (0.85-1.23)

0.828

1.01 (0.83-1.22)

0.931

0.97 (0.79-1.19)

0.755

PCT, ng/mL

1st tertile

Reference

Reference

Reference

2nd tertile

0.84 (0.28-2.55)

0.762

0.64 (0.20-2.08)

0.461

0.70 (0.19-2.44)

0.553

3rd tertile

2.36 (0.78-7.12)

0.127

2.49 (0.83-7.47)

0.103

1.58 (0.48-5.26)

0.457

UA category

b5

Reference

Reference

Reference

6-30

0.45 (0.17-1.23)

0.121

0.81 (0.26-2.46)

0.705

0.71 (0.20-2.54)

0.598

N30

1.39 (0.60-3.23)

0.446

1.47 (0.56-3.84)

0.429

1.95 (0.68-5.57)

0.214

Abbreviations AOR, adjusted odd ratio; L-tube, Levin tube; PICC, peripherally inserted central catheter; ECOG, Eastern Cooperative Oncology Group; SBP, systolic blood pressure; DBP, di- astolic blood pressure; PR, pulse pressure; RR, respiratory pressure; SpO2, blood oxygen saturation; NEWS, national early warning score; KTAS, Korean triage and acuity scale; SEG NEU, segmented form neutrophil; eGFR, estimated glomerular filtration rate; HCO3, bicarbonate; PCT, procalcitonin; UA, urine analysis; ED LOS, emergency department length of stay; ICU, in- tensive care unit.

References

  1. Ofner-Agostini M, Varia M, Johnston L, et al. Infection control and antimicrobial re- striction practices for antimicrobial resistant organisms in Canadian tertiary care hospitals. Am J Infect Control 2007;35:563-8.
  2. Tiwari HK, Sen MR. Emergence of vancomycin resistant Staphylococcus aureus (VRSA) from a tertiary care hospital from northern part of India. BMC Infect Dis 2006;26(6):156.
  3. Chazan B, Raz R, Teitler N, et al. Epidemiology and susceptibility to antimicrobials in community, hospital and long-term care facility bacteremia in northern Israel: a 6 year surveillance. Isr Med Assoc J 2009;11:592-7.
  4. Nicolle L, McIntyre M, Hoban D, et al. Bacteremia in a long term care facility. Can J Infect Dis 1994;5:130-2.
  5. Muder RR, Brennen C, Wagener MM, et al. Bacteremia in a long-term-care facility: a five-year prospective study of 163 consecutive episodes. Clin Infect Dis 1992;14: 647-54.
  6. Setia U, Serventi I, Lorenz P. Bacteremia in a long-term care facility. Spectrum and mortality. Arch Intern Med 1984;144:1633-5.
  7. Paterson DL. Resistance in Gram-negative bacteria: Enterobacteriaceae. Am J Med 2006;119:S20-8 discussion S62-70.
  8. Cornejo-Juarez P, Velasquez-Acosta C, Diaz-Gonzalez A, et al. Trend of antimicrobial drug-susceptibility of blood isolates at an oncological center (1998-2003). Salud Publica Mex 2005;47:288-93.
  9. Turner PJ. Trends in antimicrobial susceptibilities among bacterial pathogens iso- lated from patients hospitalized in European medical centers: 6-year report of the MYSTIC Surveillance Study (1997-2002). Diagn Microbiol Infect Dis 2005;51:281-9.
  10. Oteo J, Lazaro E, de Abajo FJ, et al. Spanish members of EARSS. Antimicrobial- resistant invasive Escherichia coli, Spain. Emerg Infect Dis 2005;11:546-53.
  11. OECD. Long-term care beds in institutions and hospitals. Health at a glance 2015: OECD indicators. Paris: OECD Publishing; 2015. p. 206-7 Available at https://doi. org/10.1787/health_glance-2015-78-en.
  12. Bossuyt PM, Reitsma JB, Bruns DE, et al. Towards complete and accurate reporting of studies of diagnostic accuracy: the STARD initiative. BMJ 2003;326:41-4.
  13. Gilbert EH, Lowenstein SR, Koziol-McLain J, et al. Chart reviews in emergency med- icine research: where are the methods? Ann Emerg Med 1996;27:305-8.
  14. Singer M, Deutschman CS, Seymour CW, et al. The third international consensus def- initions for sepsis and septic shock (Sepsis-3). JAMA 2016;315:801-10.
  15. Shi Leiyu, Singh Douglas. Essentials of the U.S. Health Care System. Long-term care services. , 2nd ed.Jones & Bartlett Learning; 2010.
  16. Wagner DP. Economics of prolonged mechanical ventilation. Am Rev Respir Dis 1989;140:S14-8.
  17. Carson SS. Know your long-term care hospital. Chest 2007;131:2-5.
  18. Oken MM, Creech RH, Tormey DC, et al. Toxicity and response criteria of the Eastern Cooperative Oncology Group. Am J Clin Oncol 1982;5:649-55.
  19. Buccheri G, Ferrigno D, Tamburini M. Karnofsky and ECOG performance status scor- ing in Lung cancer: a prospective, longitudinal study of 536 patients from a single in- stitution. Eur J Cancer 1996;32A:1135-41.
  20. Charlson ME, Pompei P, Ales KL, et al. A new method of classifying prognostic co- morbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40:373-83.
  21. Quan H, Li B, Couris CM, et al. Updating and validating the Charlson comorbidity index and score for risk adjustment in hospital discharge abstracts using data from 6 countries. Am J Epidemiol 2011;173:676-82.
  22. Mylotte JM, Karuza J, Bentley DW. Methicillin-resistant Staphylococcus aureus: a questionnaire survey of 75 long-term care facilities in western New York. Infect Control Hosp Epidemiol 1992;13:711-8.
  23. Rogers MA, Mody L, Chenoweth C, et al. Incidence of antimicrobial-resistant infec- tion in long-term residents of skilled nursing facilities. Am J Infect Control 2008; 36:472-5.
  24. Ludden C, Cormican M, Vellinga A, et al. Colonisation with ESBL-producing and carbapenemase-producing Enterobacteriaceae, vancomycin-resistant enterococci, and meticillin-resistant Staphylococcus aureus in a long-term care facility over one year. BMC Infect Dis 2015;15:168.
  25. Prasad N, Labaze G, Kopacz J, et al. Asymptomatic rectal colonization with carbapenem-resistant Enterobacteriaceae and Clostridium Difficile among residents of a long-term care facility in New York City. Am J Infect Control 2016;44:525-32.
  26. Giufre M, Accogli M, Ricchizzi E, et al. Multidrug-resistant infections in long-term care facilities: extended-spectrum ?-lactamase-producing Enterobacteriaceae and hypervirulent antimicrobial resistant Clostridium difficile. Diagn Microbiol Infect Dis 2018;91:275-81.
  27. Kirn TJ, Weinstein MP. Update on blood cultures: how to obtain, process, report, and interpret. Clin Microbiol Infect 2013;19:513-20.
  28. Hattori T, Nishiyama H, Kato H, et al. clinical value of procalcitonin for patients with suspected bloodstream infection. Am J Clin Pathol 2014;141:43-51.
  29. Meier MA, Branche A, Neeser OL, et al. Procalcitonin-guided antibiotic treatment in patients with positive blood cultures: a patient-level meta-analysis of randomized trials. Clin Infect Dis 2018 Oct 25. https://doi.org/10.1093/cid/ciy917 [Epub ahead of print].