Fig 1.
Study design graphic representation.
Six Italian hospitals from different geographical regions were enrolled in the study (North: Feltre, Tolmezzo, Vigevano; Centre: Rome; South: Foggia, Messina). Five hospitals were randomly allocated in two Intervention groups (I1, I2) and one further hospital represented an external control (extC): I1-group included Roma, Foggia and Feltre hospitals, entering the study on January 1st 2016; I2-group included Vigevano and Tolmezzo hospitals, entering 5-months later, on May 1st 2016; extC hospital was represented by Messina hospital, receiving no intervention and monitored from May 1st 2016. The phases of the study are indicated by colours: orange, 6-months pre-intervention period (pre-PCHS); light green, stabilization period, when PCHS was introduced; green, 6-months post-intervention period (PCHS), when PCHS was routinely applied. Sampling campaigns for microbiological analyses are indicated by circles: conventional microbiological analyses were performed monthly (black circles), and molecular analyses were performed quarterly (red circles) in all enrolled hospitals.
Table 1.
Population characteristics of study participants in pre-PCHS and PCHS phases, stratified by enrolled hospitals.
Fig 2.
HAI incidence rates in the I1-I2 intervention hospitals.
Results are expressed as bimonthly value of incidence rate per 1,000 patient-days, respectively in the pre-PCHS (red) and PCHS periods (blue). 95% CI intervals are also reported.
Table 2.
Patient characteristics of the I1-I2 hospitals in the pre-PCHS and PCHS periods (11,461 patients).
Table 3.
HAIs in pre-PCHS and PCHS phases, stratified by type.
Table 4.
Microorganisms isolated from HAIs during pre-PCHS and PCHS phases in I1-I2 hospitals.
Table 5.
Risk factors associated with HAI onset in patients of I1-I2 hospitals: Multivariable model*.
Fig 3.
Surface contamination in the surveyed hospitals.
(A) Pathogen load on hospital surfaces, expressed as CFU/m2. Six pathogens were measured by direct CFU counting on specific Rodac plates, as described in Methods (Staphylococcal spp., Enterobacteriaceae spp., Acinetobacter spp., Candida spp., Pseudomonas spp., Clostridium spp.). Graphed results represent the sum of the median values obtained for each measured pathogen. Median values (lower part of the box) and Q3 values (upper part of the box, representing the 75% percentile values) are shown for each hospital, and for pre-intervention (pre-PCHS) and intervention (PCHS) phases. Values reported for the external control hospital (Messina), correspond to those detected in the 1st and 2nd 6-month periods of the study. (B) Total bacterial load and PCHS-Bacilli count, respectively measured by a pan-bacterial qPCR (panB) and a specific qPCR for Bacillus genus (spo0A). Results are expressed as genome copy number per 100 ng of tested DNA. The median values ± SD of pre-PCHS and PCHS phases are shown. Values reported for the external control hospital (Messina), correspond to those detected in the 1st and 2nd 6-month periods of the study.
Fig 4.
Resistome analysis of PCHS-Bacillus strains.
Antibiotic resistance genes were analyzed by microarray both in the PCHS detergent prior to application, containing a blend of three Bacillus species (Original) and in the Bacillus isolates (Isolates) collected from hospital surfaces in the PCHS phase of I1 and I2 hospital groups. For original PCHS-Bacilli, results are expressed as mean values ± SD of six replicates. For Isolates, results are expressed as the mean value ± SD of 120 Bacillus isolated from hospital surfaces. Both Original and Isolates values were compared to negative control values (NTC). Each Bacillus isolate was identified by PCR and sequencing prior to microarray analysis, as previously described [29].