Pseudomonas aeruginosa infection increases the readmission rate of COPD patients

Figures & data

Figure 1 Study design.

Abbreviations: AECOPD, acute exacerbations of COPD; PCR, polymerase chain reaction.

Table 1 Baseline characteristics in readmission and no-readmission groups before and after propensity score matching

	Before matching				After matching
	Readmission group (n=65)	No-readmission group (n=639)	P-value	d	Readmission group (n=52)	No-readmission group (n=52)	P-value	d
Age (years)a	73.9±9.6	71.8±9.3	0.078	0.2251	72.9±9.2	73.1±7.6	0.908	0.0237
Gender
Maleb	48 (73.8%)	444 (69.5%)	0.465	0.1186	40 (76.9%)	38 (73.1%)	0.651	0.1133
Femaleb	17 (26.2%)	195 (30.5%)			12 (23.1%)	14 (26.9%)
Smoking history
Current smokerb	4 (6.2%)	102 (16.0%)	0.499	0.1695	4 (7.7%)	4 (7.7%)	0.898	0.1072
Ex smokerb	43 (66.2%)	367 (57.4%)			37 (71.2%)	34 (65.4%)
Nonsmokerb	18 (27.7%)	170 (26.6%)			11 (21.2%)	14 (26.9%)
Comorbidities
Hypertensionb	31 (47.7%)	324 (50.7%)	0.644	0.0665	26 (50.0%)	29 (55.8%)	0.556	0.1278
Diabetesb	15 (23.1%)	149 (23.3%)	0.965	0.1175	11 (21.2%)	17 (32.7%)	0.185	0.3272
Coronary artery diseaseb	13 (20.0%)	99 (15.5%)	0.344	0.1710	11 (21.2%)	11 (21.2%)	1.000	0.0000
Cerebrovascular accidentb	2 (3.1%)	31 (4.9%)	0.760	0.2612	2 (3.8%)	3 (5.8%)	1.000	0.2347
Severity of COPD
GOLD Ib	5 (7.7%)	50 (7.8%)	0.434	0.1020	4 (7.7%)	3 (5.8%)	0.798	0.1159
GOLD IIb	25 (38.5%)	246 (38.5%)			18 (34.6%)	22 (42.3%)
GOLD IIIb	22 (33.8%)	266 (41.6%)			20 (38.5%)	20 (38.5%)
GOLD IVb	13 (20.0%)	77 (12.1%)			10 (19.2%)	7 (13.5%)
Length of hospital stay (days)a	14.0±10.4	9.6±8.1	0.001	0.5278	12.2±8.0	10.5±8.2	0.307	0.2099
ICU admission	10 (15.4%)	63 (9.9%)	0.164	0.2802	7 (13.5%)	8 (15.4%)	0.780	0.0860
Use of mechanical ventilation	8 (13.8%)	38 (5.9%)	0.031	0.4396	6 (11.5%)	8 (15.4%)	0.566	0.1831
Use of NIV	1 (1.5%)	10 (1.6%)	1.000	0.0096	1 (1.9%)	0 (0.0%)	1.000	Infinity

Notes:

a Data are presented as mean ± SD.

b Data are presented as number of patients (%).

Abbreviations: d, standardized mean difference; GOLD, Global Initiative for Chronic Obstructive Lung Disease; ICU, intensive care unit; NIV, noninvasive ventilation.

Table 2 Pulmonary medication or treatment in readmission and no-readmission groups before and after propensity score matching

	Before matching				After matching
	Readmission group (n=65)	No-readmission group (n=639)	P-value	d	Readmission group (n=52)	No-readmission group (n=52)	P-value	d
Inhaler use before admission
LABAs	3 (4.6%)	4 (0.6%)	0.277	0.3522	0 (0.0%)	2 (3.8%)	0.730	0.0348
LAMAs	13 (20.0%)	75 (11.7%)			10 (19.2%)	6 (11.5%)
LABAs + LAMAs	16 (24.6%)	97 (15.2%)			12 (23.1%)	15 (28.8%)
ICS/LABAs	3 (4.6%)	51 (8.0%)			3 (5.8%)	3 (5.8%)
Triple therapy (ICS/LABAs + LAMAs)	24 (36.9%)	245 (38.3%)			21 (40.4%)	20 (38.5%)
Oral medication use
β2 Adrenoreceptor agonist	11 (16.9%)	78 (12.2%)	0.276	0.2106	7 (13.5%)	6 (11.5%)	0.767	0.0971
N-acetylcysteine	11 (16.9%)	50 (7.8%)	0.013	0.4826	6 (11.5%)	12 (23.1%)	0.120	0.4592
Roflumilast	9 (13.8%)	43 (6.7%)	0.046	0.4416	6 (11.5%)	3 (5.8%)	0.488	0.4170
Mucolytic agent	44 (67.7%)	339 (53.1%)	0.024	0.3404	36 (69.2%)	32 (61.5%)	0.410	0.1880
Oral steroid	10 (15.4%)	32 (5.0%)	0.003	0.6826	6 (11.5%)	4 (7.7%)	0.506	0.2470
Oral antibiotics	8 (12.3%)	27 (4.2%)	0.011	0.6380	4 (7.7%)	2 (3.8%)	0.678	0.4047
Home oxygen therapy	25 (38.5%)	126 (19.7%)	<0.001	0.5149	16 (30.8%)	16 (30.8%)	1.000	0.0000

Note: Data are presented as number of patients (%).

Abbreviations: d, standardized mean difference; LABAs, long-acting β-agonists; LAMAs, long-acting antimuscarinic antagonists; ICS, inhaled corticosteroids.

Table 3 Identified pathogen in readmission events and no-readmission groups before and after propensity score matching

	Before matching			After matching
	Readmission group (n=65)	No-readmission group (n=639)	P-value	Readmission group (n=52)	No-readmission group (n=52)	P-value
Identification of bacteria or virus
Only bacterial pathogen identification	27 (41.5%)	158 (24.7%)	0.003	20 (38.5%)	14 (26.9%)	0.210
Only viral pathogen identification	9 (13.8%)	149 (23.3%)	0.081	8 (15.4%)	13 (25.0%)	0.222
Bacterial–viral coidentification	8 (12.3%)	91 (14.2%)	0.669	8 (15.4%)	2 (3.82%)	0.092
No pathogen identification	21 (32.3%)	241 (37.7%)	0.390	16 (30.8%)	23 (44.2%)	0.156
Analysis of major pathogen
Pseudomonas aeruginosa	15 (23.1%)	73 (11.4%)	0.007	12 (23.1%)	4 (7.7%)	0.030
Streptococcus pneumoniae	6 (9.2%)	76 (11.9%)	0.524	4 (7.7%)	3 (5.8%)	1.000
Haemophilus influenza	3 (4.6%)	34 (5.3%)	1.000	3 (5.8%)	2 (3.8%)	1.000
Influenza virus	5 (7.7%)	83 (13.0%)	0.219	5 (9.6%)	5 (9.6%)	1.000
Rhinovirus	3 (4.6%)	62 (9.7%)	0.177	2 (3.8%)	4 (7.7%)	0.678
Parainfluenza virus	4 (6.2%)	33 (5.2%)	0.768	4 (7.7%)	2 (3.8%)	0.678

Note: Data are shown as number of patients (%).

Table 4 Multivariate analysis of Pseudomonas aeruginosa identification rate after propensity score matching

Parameter	Odds ratio	95% CI	P-value
Adjusted for propensity and selected variablesa	3.600	1.077–12.035	0.038
Adjusted for propensity and all variablesb	4.749	1.296–17.401	0.019

Notes:

a Selected variables included variables associated with prognosis such as age, comorbidity (hypertension, diabetes, coronary artery disease, cerebrovascular accident), COPD severity, smoking history, length of hospital stay, intensive care unit admission, and use of mechanical ventilation.

b All variables in Tables 1 and 2.