Two hundred and twenty-five stage II to stage III thoracic empyema patients, who underwent video-assisted thoracoscopic decortication and had a positive pleural fluid or pleural peel tissue culture, were included in this study. From these patients, one to four sets of sputum cultures were collected during the period of hospitalization. For 76 patients, at least one sputum culture set showed a positive finding, and 149 patients had only negative sputum culture results.
The clinical demographic data of the patients are shown in Table 1. The mean age was about 63 years old, and most of the patients were males (75.6%). Right-side empyema was predominant (56.0%), and only one patient had bilateral empyema. There was more stage II empyema (76.0%) than stage III empyema (24.0%). As for the comorbidities, congestive heart failure was the only one for which the sputum culture positive group had a statistically significantly higher ratio (15.8% versus 6.0%, p = 0.033). No differences in serum leukocyte count and pleural fluid analysis were noted between the two groups. Treatment outcome measures such as preoperative antibiotics use days, chest tube drainage duration, ICLOS, HLOS, and hospital mortality rate were statistically significantly lower in the sputum culture negative group. There was no 30-day mortality case in the positive sputum culture cohort.
Table 1
Clinical characteristics of the patients
Factors (Mean ± SD or count) | Total cohort (n = 225) | Sputum culture positive (n = 76) | Sputum culture negative (n = 149) | P-value |
Age (years) | 62.86 ± 16.52 | 64.59 ± 15.98 | 61.98 ± 16.77 | 0.263a |
Gender | | | | |
Male | 170(75.6%) | 54(71.1%) | 116(77.9%) | 0.566 |
Female | 55(24.4%) | 22(28.9%) | 33(22.1%) | 0.327 |
Location | | | | |
Right | 126(56.0%) | 40(52.6%) | 86(57.7%) | 0.635 |
Left | 98(43.6%) | 35(46.1%) | 63(42.3%) | 0.670 |
Bilateral | 1(0.4%) | 1(1.3%) | 0(0.0%) | 0.317 |
Phase | | | | |
II | 171(76.0%) | 60(78.9%) | 111(74.5%) | 0.747 |
III | 54(24.0%) | 16(21.1%) | 38(25.5%) | 0.466 |
Comorbidity | | | | |
Malignancy | 61(27.1%) | 24(31.6%) | 37(24.8%) | 0.354 |
Diabetes mellitus | 82(36.4%) | 34(44.7%) | 48(32.2%) | 0.138 |
Hypertension | 134(59.6%) | 51(67.1%) | 83(55.7%) | 0.321 |
Liver cirrhosis | 19(8.4%) | 7(9.2%) | 12(8.1%) | 0.808 |
End-stage renal disease | 25(11.1%) | 11(14.5%) | 14(9.4%) | 0.221 |
Congestive heart failure | 21(9.3%) | 12(15.8%) | 9(6.0%) | 0.033* |
COPD | 54(24.0%) | 21(27.6%) | 33(22.1%) | 0.396 |
Lab data | | | | |
WBCC (/µL) | 15,596.62 ± 6,379.68 | 14,544.73 ± 6,656.69 | 16,133.15 ± 6,187.66 | 0.077a |
ANC (/µL) | 13,328.48 ± 5,992.72 | 12,590.03 ± 6,223.27 | 13,707.69 ± 5,856.05 | 0.187a |
Pleural data | | | | |
Pleural pH ≦ 7.2 | 112(57.1%) | 39(59.1%) | 73(56.2%) | 0.780 |
Pleural glucose ≦ 40 mg/dL | 112(54.1%) | 34(48.6%) | 78(56.9%) | 0.437 |
Pleural LDH ≧ 1000 IU/L | 152(73.8%) | 51(72.9%) | 101(74.3%) | 0.934 |
Pre-OP ABX usage (days) | 7.28 ± 19.38 | 13.32 ± 31.78 | 4.21 ± 5.30 | 0.015a* |
Chest tube drainage duration (days) | 18.02 ± 24.72 | 25.43 ± 34.67 | 14.24 ± 16.53 | 0.009a* |
Length of ICU stay (days) | 10.77 ± 15.64 | 18.83 ± 22.20 | 6.66 ± 8.37 | < 0.001a* |
Length of hospital stay (days) | 51.04 ± 149.40 | 87.40 ± 222.98 | 32.50 ± 86.95 | 0.042a* |
Hospital mortality | 36(16.0%) | 20(26.3%) | 16(10.7%) | 0.014* |
Mortality in 30 postoperative days | 3(1.3%) | 0(0.0%) | 3(2.0%) | 0.157 |
COPD: chronic obstructive pulmonary disease; WBCC: white blood cell count; ANC: absolute neutrophil count; LDH: lactate dehydrogenase; OP: operative; ABX: antibiotics; ICU: intensive care unit |
a: Independent t test; *: significant at the 5% level |
We analyzed the patients based on the number of sets of sputum cultures (Fig. 1). Those with only one sputum culture set had a 24% positive rate, and similar positive culture rates were found in the patients with two sets (26%) and three sets (26%). In comparison, patients with four sets of sputum cultures had a significantly higher positive rate (56%).
The collection method of a sputum specimen was taken into consideration. In our center, we collected patients’ sputum in three ways: (1) patient expectorates by oneself, (2) aspiration through endotracheal tube or tracheostomy, (3) BAL under bronchoscopy. We analyzed the positive culture rates of these three methods separately (Fig. 2) and discovered significantly higher positive culture rates when the specimen was derived through bronchoscopy (40%) or aspiration via endotracheal tube or tracheostomy (32%).
Figure 3 shows the pathogen distributions for the sputum cultures and the intraoperative pleural fluid and tissue peel cultures. In our study, the most common species of pathogens in the sputum cultures was Pseudomonas aeruginosa (44%), which was followed by Klebsiella pneumoniae (16%), Staphylococcus aureus (10%), Acinetobacter baumannii (10%), and Streptococcus spp. (4%). The most common species of pathogens in the pleural fluid cultures and pleural tissue peel cultures was Streptococcus spp. (20%), Staphylococcus aureus (11%), and viridans (9%).
To see the concordance rate of pathogens between the sputum cultures and the intraoperative pleural fluid/tissue peel cultures, we divided those patients with positive sputum cultures into an identical group and a distinct group (Table 2). We defined the identical group as the patients for which a sputum culture and the pleural fluid/tissue culture had a common pathogen (n = 30). The distinct group was defined as the patients who had no pathogen in common between the sets of sputum culture and the pleural fluid/tissue culture (n = 46). In comparing these two groups, we noted that the age, gender, location and phase of empyema, laboratory data, and pleural analysis data were similar, but a longer use of antibiotics preoperatively was found in the identical group (19.03 ± 37.66 days versus 9.59 ± 27.09 days, p = 0.006). For the surgical outcomes, the duration of chest tube drainage, ICLOS, and HLOS were longer in the identical group, but the differences were not statistically significant. However, the mortality rate during hospitalization was significantly higher in the identical group (40.0% versus 17.4%, p = 0.002).
Table 2
Concordance rate between species from sputum culture and pleural effusion or tissue culture (at least one pathogen is the same)
Factors (Mean ± SD or count) | Identical (n = 30) | Distinct (n = 195) | P-value |
Age (years) | 63.47 ± 15.91 | 65.33 ± 16.16 | 0.527a |
Male | 22(73.3%) | 32(69.6%) | 0.802 |
Located on right side | 19(63.3%) | 21(45.7%) | 0.103 |
Phase II | 22(73.3%) | 38(82.6%) | 0.423 |
Pre-OP ABX usage (days) | 19.03 ± 37.66 | 9.59 ± 27.09 | 0.006a* |
Lab data | | | |
WBCC (/µL) | 14,683.33 ± 7,082.18 | 14,454.35 ± 6,442.66 | 0.811a |
ANC (/µL) | 12,773.43 ± 6,457.72 | 12,470.42 ± 6,134.88 | 0.983a |
Pleural data | | | |
Pleural pH ≦ 7.2 | 16(66.7%) | 23(54.8%) | 0.277 |
Pleural glucose ≦ 40 mg/dL | 13(48.1%) | 21(48.8%) | 0.919 |
Pleural LDH ≧ 1000 IU/L | 22(81.5%) | 29(67.4%) | 0.219 |
Chest tube drainage duration (days) | 33.87 ± 51.33 | 19.94 ± 15.08 | 0.602a |
ICU duration (days) | 21.77 ± 28.85 | 16.91 ± 16.61 | 0.957a |
Hospital duration (days) | 129.60 ± 316.75 | 59.87 ± 127.60 | 0.983a |
Hospital mortality | 12(40.0%) | 8(17.4%) | 0.002* |
OP: operative; ABX: antibiotics; WBCC: white blood cell count; ANC: absolute neutrophil count; LDH: lactate dehydrogenase; ICU: intensive care unit |
a: Mann-Whitney U test; *: significant at 5% level |
In addition, we analyzed the factors associated with in-hospital mortality in Table 3, including the comparisons of the clinical characteristics, positive culture rates, bacteriology, ICLOS, and HLOS. Between survivors (n = 189) and non-survivors (n = 36), our analysis showed that end-stage renal disease, positive sputum cultures, and Pseudomonas aeruginosa as the causing pathogen were associated with higher in-hospital mortality rates. The stays in the intensive care unit (ICU) and the hospital were also clearly longer for the in-hospital mortality cases.
Table 3
Clinical characteristics and bacteriology between in-hospital survivors and non-survivors
Variables (Mean ± SD or count) | Survivors (n = 189) | Non-survivors (n = 36) | P-value |
Age (years) | 61.96 ± 16.94 | 67.61 ± 13.35 | 0.080a |
Male | 143(75.7%) | 27(75.0%) | 0.935 |
Location | | | |
Right | 105(55.6%) | 21(58.3%) | 0.851 |
Left | 84(44.4%) | 14(38.9%) | 0.583 |
Bilateral | 0(0.0%) | 1(2.8%) | 0.083 |
Phase | | | |
II | 147(77.8%) | 24(66.7%) | 0.361 |
III | 42(22.2%) | 12(33.3%) | 0.138 |
Comorbidity | | | |
Malignancy | 47(24.9%) | 14(38.9%) | 0.080 |
Diabetes mellitus | 64(33.9%) | 18(50.0%) | 0.081 |
Hypertension | 110(58.2%) | 24(66.7%) | 0.421 |
Liver cirrhosis | 14(7.4%) | 5(13.9%) | 0.180 |
End-stage renal disease | 13(6.9%) | 12(33.3%) | < 0.001* |
Congestive heart failure | 17(9.0%) | 4(11.1%) | 0.655 |
COPD | 47(24.9%) | 7(19.4%) | 0.366 |
Positive culture | | | |
Pleural effusion | 143(75.7%) | 29(80.6%) | 0.690 |
Tissue | 106(56.1%) | 23(63.9%) | 0.465 |
Sputum | 56(29.6%) | 20(55.6%) | 0.005* |
Causing pathogen | | | |
Pseudomonas aeruginosa | 31(16.4%) | 13(36.1%) | 0.006* |
Klebsiella pneumoniae | 40(21.2%) | 5(13.9%) | 0.237 |
Length of ICU stay | 7.77 ± 10.41 | 26.53 ± 26.06 | < 0.001a* |
Length of hospital stay | 37.82 ± 101.43 | 120.50 ± 285.82 | < 0.001a* |
COPD: chronic obstructive pulmonary disease; ICU: intensive care unit |
a: Mann-Whitney U test; *: significant at the 5% level |