Clinical features
A total number of 177 inpatients were enrolled, including PV (80.2%, 142/177), pemphigus erythematosus (PE) (14.1%, 25/177), PF (5.1%, 9/177), and pemphigus vegetans (Pveg) (0.6%, 1/177). The median age was 50.4 years (range: 14-80 years), and the female/male ratio was 0.90. The incidence of diabetes mellitus (DM) (18.6%, 33/177), hypertension (27.7%, 49 /177), and osteoporosis (11.3%, 20/184) were high. Twelve (6.8%), 31 (17.5%), 52 (29.4%) and 82 (46.3%) of 177 patients had mild, moderate, severe and extensive skin lesions, respectively. Most inpatients (70.6%, 125/177) received oral glucocorticoids, and 114 (91.2%) of 125 had detailed records of the dosage. Of them, 43.9% (50/114) patients took medium dose (0.5mg/kg·d -1.0mg/kg·d), and 24.6% (28/114) patients took high dose (≥1.0mg/kg·d) of prednisone or the equivalent. Eighty-seven of 177 (49.2%) inpatients developed BSIs. LOS was analyzed by a non-parametric test, and inpatients with BSIs had longer LOS compared with inpatients without BSIs (median: 18.9 days vs. 14.1 days, p = 0.008).
Bacterial spectrum and drug resistance
Eighty-seven of 177 (49.2%) patients developed BSIs, and 17 of 87 (19.5%) inpatients had multiple bacterial infections. If a patient, who was diagnosed with BSIs, had the isolations of both pathogenic bacteria and underlying contaminating bacteria, such as Staphylococcus epidermidis (S. epidermidis) or streptococcus viridans, it was hard to differentiate whether the S. Epidermidis or streptococcus viridans was contaminating bacteria or not. Therefore, we showed all the isolated bacteria of these inpatients. The Bacterial spectrum of all inpatients with BSIs was presented in Figure 2. Sixty-two (71.3%, 62/87) inpatients were infected with Staphylococcus aureus (S. aureus), followed by Escherichia coli (8.0%, 7/87), and Methicillin-resistant coagulase-negative staphylococcus (8.0%, 7/87). Eighty-one (93.1%, 81/87) inpatients presented with community-acquired infections. Patients with hospital-acquired infections had a higher incidence of S. aureus (83.3%, 5/6). The bacterial spectrum of community-acquired infections and hospital-acquired infection infections was illustrated in Figure 3.
It was a pity that many patients did not record the location of collected strains. Forty-three patients (43/87, 49.4%) detailed recorded the location of collected strains, including the scalp (13/43, 30.2%), the back (12/43, 27.9%), the axilla (12/43, 27.9%), and the other sites (33/43, 76.7%). The most common bacteria of the scalp, the back, and the axilla were S.aureus (7/13, 53.8%), S.aureus (8/12, 66.7%), and Enterobacter cloacae (3/12, 25.0%), respectively.
As for the drug resistance of S. aureus, penicillin G (91.9%, 57/62), erythromycin (75.8%, 47/62), and clindamycin (45.2%, 28/62) had the highest resistance rate. The resistance rate of rifampicin (1.6%, 1/62), linezolid (1.6%, 1/62), quinupristin (1.6%, 1/62), and tigecycline (1.6%, 1/62) was the lowest. All patients were sensitive to vancomycin (Figure 4). As the second most common bacteria, Methicillin-resistant coagulase-negative staphylococcus was also highly resistant to penicillin (100.0%, 7/7) and erythromycin (100.0%, 7/7). Methicillin-resistant coagulase-negative staphylococcus remained 100% sensitive to tigecycline and vancomycin.
Risk factors for BSIs
Sex (p = 0.49), age (p = 0.96), obesity (p = 0.24), and smoke (p = 0.13) did not significantly associate with BSIs. Cerebrovascular disease (CVD) (p = 0.96), chronic heart disease (CHD) (p = 0.34), diabetes mellitus (DM) (p = 0.28), hypertension (p = 0.72), and osteoporosis (p = 0.58) did not associate with higher incidence of BSIs. Similarly, severity of skin lesions (p = 0.08), subtype of pemphigus (p = 0.51), and hypoalbuminemia (p = 0.09) were not underlying risk factors for BSIs. Higher levels of anti-Dsg1 autoantibodies (> 124.2U/mL) (p < 0.001) and anti-Dsg3 autoantibodies (>169.5U/mL) (p = 0.01) were associated with BSIs (Table 1). The variables with p ≤ 0.10 were reanalyzed by binary regression analysis and the results showed in Table 2. Higher levels of anti-Dsg1 autoantibodies (>124.2U/mL) (p < 0.001, odds ratio [OR]=3.564, 95% CI: 1.784-7.123) and anti-Dsg3 autoantibodies (>169.5U/mL) (p = 0.03, OR=2.074, 95% CI: 1.084-3.969) were significantly associated with BSIs.
Associated factors for the type of bacteria
In inpatients with BSIs, 60 (69.0%, 60/87), 13 (14.9%, 13/87), and 14 (16.1%, 14/87) had infections of Gram-positive bacteria, Gram-negative bacteria and coinfections, respectively. Females had a higher rate of coinfection (54.3% vs. 35.7%) and a lower rate of Gram-negative infection (25.4% vs. 84.6%) compared to males (p = 0.03). Though the result was not significantly different, a higher rate of coinfections was observed in patients <53.5 years compared to ≥53.5 years’ (71.4% vs. 28.6%) (p = 0.37). Comorbidities, including CVD (p = 0.72), CHD (p = 0.72), DM (p = 0.39), hypertension (p = 0.86), and osteoporosis (p = 0.92), were not associated with Gram’s stain. Bedridden was not significantly associated with Gram’s stain (p = 0.09). Within 2 weeks before this admission, patients taking oral antibiotics had a higher rate of Gram-negative isolation (p = 0.05). Patients who were hospitalized in other hospitals within two weeks before the current admission had higher rates of Gram-negative infection and co-infections (p = 0.03) (Table 3).