Infecciones asociadas a la red de agua hospitalaria en las unidades de cuidados intensivos

Resumen

Risk factors associated with acquisition of waterborne healthcare-associated infection and/or colonization in ICU Background: Water systems act as an important reservoir of microorganism causing high-risk unit-acquired colonization and/or infections. Aim: To conduct a systematic review to identify risk factors associated with acquisition of waterborne healthcare-associated infection (HAI) and/or colonization. Methods: A systematic review and meta-analysis was conducted. PubMed, Web science and Embase were searched to identify cohort and case-control studies that had reported on the risk factors associated with acquisition of waterborne-HAU and/or colonization from 1966 to 10 July 2018. Three investigators independently screened titles and abstracts for relevant articles, and two screened full-text articles. Data were extracted by two investigators. Summary estimates were obtained using random effect model. Findings: Ten relevant studies met the inclusion criteria for the systematic review (5 cohort studies and 5 case control studies). We identified absence of point-of-use filter installed on the sink faucet (OR: 5.79, 95%CI=1.29-30.62) and exposure to invasive ventilation (OR, 11.52; 95% CI: 0.55-240.94) as risk factors in neonatal population. Being admitted to a room with a contaminated tap (HR=1.76, 95% CI=1.09-2.84) or contaminated sink (OR: 11.2, 95% CI= 1.92-65.68), exposure to invasive ventilation >10 days (HR: 2.56, 95% CI= 1.46-4.50) and requiring hemofiltration (OR: 22.9, 95% CI= 1.63-203.91) were identified as risk factors in adult population. The case-control studies contributed to a meta-analysis. Waterborne healthcare-associated infection and/or colonization was positively associated in neonatal population with central venous catheter exposure (pooled OR: 8.02, 95% CI= 2.19-29.31, p=0.002). Umbilical catheter exposure (pooled OR: 2.41, 95% CI= 0.20-28.33, p=0.49) and total parenteral nutrition exposure (pooled OR: 11.52, 95% CI= 0.55-240.94, p=0.12) were not associated with acquisition. The overall quality of evidence as defined by GRADE was rated as low. Length of stay before colonization was positive associated with an increased risk of waterborne healthcare-associated infections and/or colonization acquisition in adult population. The overall quality of evidence as defined by GRADE was low. Duration of antibiotic treatment in ICU was identified as a protective factor in adult population. Conclusions: In this meta-analysis, central venous catheter exposure in neonates and length of stay in adult population were associated with an increased risk of waterborne healthcare-associated infection/colonization in high-risk units. The low quality of evidence of case-control studies makes it necessary to promote future studies in order to evaluate other risk factors. PHASE 2 A cluster of Chryseobacterium indologenes cases related to drainage water in intensive care units Objective: To describe a cluster of Chryseobacterium indologenes infections and colonizations related to sink-drain water in intensive care unit patients and the measures implemented to control it. Design: Outbreak investigation. Setting: Surgical and medical ICU in a tertiary care teaching hospital. Methods: We undertook a spatiotemporal analysis of cases to identify possible routes of cross-transmission or point sources. This analysis was reviewed against the sequential effect of the control measures implemented. Results: A total of 12 cases among 300 new patients admitted to ICU during the epidemic period (incidence rate 4.3%) were identified. Four of the case patients (33%) were classified as infected (3 pneumonia and 1 bacteremic pneumonia). Three of the infected patients died (75%). C. indologenes was recovered from environmental samples (hand-washing sinkholes and air close to the sinks while flushing tap water). All 12 patients from whom C. indologenes was isolated had a unique PFGE pattern indistinguishable from the isolate recovered from the air close to the reprocessing room sink. After closing down of patient hand washing stations, sealing hemodialysis drain connections, establishing waterless patient hygiene, redesigning the reprocess and storage of material, and training in reprocessing, no new cases appeared. Conclusions: Sinkholes and drains are potentially contaminated and can act as a reservoir for C. indologenes and other microorganisms. Tap water flushing may spread them to surrounding areas. Susceptible patients and their equipment should be protected from these splashes since contamination can occur. Water supplies number, location and design in ICUs should be rethought. PHASE 3 Impact of sink removal from intensive care units Objective: To identify the global rate of new cases of acquisition of GNPN and MDR-GNB bacteria in ICU after removal of drains. Design: A Seventy-three-month quasi-experimental study. Setting: Surgical and medical ICU in a tertiary care teaching hospital. Methods: The intervention consisted of removal of handwashing sinks per bed and bay in September 2016. It was implemented in two ICUs, leaving only central handwashing sinks (2) per unit. The pre-intervention period ran from April 2013 to September 2016 (34 months) and post-intervention period, from October 2016 to January 2020 (40 months). For each period, we collected epidemiological data from patients admitted to the ICUs and stay in ICU before acquisition. We identified the global rate of new cases of acquisition in ICU of NFGNB and MDR-GNB bacteria. Crude relative risk (cRR) and 95% confidence interval (CI) were calculated by dividing the incidence rates for the two periods. We evaluated factors associated with mortality. Multivariate logistic regression was undertaken with Stata 13.0 software (Stata StataCorp, College Station, US). We adjusted for gender, age, type of ICU and pre-acquisition hospital stay. Lina M. Parra Ramírez The Kaplan-Meier and log-rank test were used for time-to-event analysis with the following endpoint: survival after acquisition in the pre- and post-intervention period. Results: 736 new cases of NFGNB and MDR-GNB were identified among ICU-admitted patients over a period of 81.116 patient-days. The overall incidences of new NFGNB cases in the pre- and post-intervention periods were 8.70 and 4.44 per 1000 patient-days (RR= 0.51, CI95%=0.42-0.61) while the overall incidences of MDR-GNB were 3.19 and 2.91 per 1000 patient-days (RR= 0.91, CI 95% 0.70-1.19) respectively. CONCLUSIONS 1. In Intensive Care Units, the presence of contaminated taps or drains near the patient, the absence of point-of-use-Filters, prolonged stay and exposure to invasive procedures (central venous catheter, mechanical ventilation, haemofiltration) were associated with an increased risk of acquiring a waterborne microorganism. 2. Wastewater system and distal points (sinks, drains, taps, tap water) are the main reservoirs of non-fermenting gram-negative bacilli and Enterobacteriaceae. 3. Splashing and aerosolization are the indirect transmission mechanisms involved in the dissemination of micro-organisms from the water supply to the patient, surrounding areas and nearby sanitary equipment. 4. Prevention and control of outbreaks of waterborne microorganisms requires a multimodal strategy incorporating a. education, b. structural, c. barrier, d. disinfection, e. intensified surveillance, and f. feedback. 5. Splash barrier measures are necessary so that susceptible patients and their environment are protected. 6. In our study, intervention measures significantly reduced new cases of P. aeruginosa, S. maltophilia and A. baumanni in both units. 7. The reduction of multidrug-resistant Enterobacteriaceae isolates was only significant first year post-intervention and mainly for K. pneumoniae. 8. The proportion of patients with colonizations and/or HAI in both periods was significantly higher in the open unit. 9. Patients with a nearby water source (<1 meter) were eight times higher to die. 10. The number, location and design of water supplies in intensive care units need to be reconsidered.