Hepatitis C virus (HCV) is a blood borne pathogen which has been recognized to be one of the major causes of chronic liver disease worldwide [1].

In Italy the highest prevalence of HCV infection (usually higher than 10%) is reported among intravenous drug users, subjects who received blood coagulation factors before 1987, subjects who received blood transfusions or organ transplantation before 1992 and haemodialysis (HD) patients [2].

HCV is an important cause of morbidity and mortality in patients with end-stage renal disease [2, 3], in fact, despite the introduction of anti-HCV routine screening for blood donors and the introduction of erythropoietin in the early '90 s [4, 5], in Italy the prevalence of HCV among HD patients is much higher than in the general population [2, 6]. Nevertheless, recent lines of evidence indicate that HCV can hardly be transmitted during dialysis procedures when state-of-the-art machines are used [7]; therefore it has been suggested that the increased incidence of infections may result from poor application of standard precautions, such as hand hygiene, proper use of gloves, and safe injection practices[8, 9].

Between 27 July and 29 September 2005, 13 cases of newly acquired HCV infection occurred among the 50 end-stage renal disease patients who were cared for in a privately run HD unit in Lazio Region (central Italy). In response to this outbreak an epidemiological investigation was conducted by the National Institute for Infectious Disease "Lazzaro Spallanzani". The present paper reports the epidemiological investigation carried out to confirm the presence of the outbreak, to identify the source(s) of infection, to elucidate the transmission pathway(s) and to define the clinical outcome of infected patients. The report has been written according to "The ORION statement: guidelines for transparent reporting of outbreak reports and intervention studies of nosocomial infection" [10].

The outbreak was suspected on the ground of a peculiarly high incidence of anti-HCV sero-conversion between July and the September 2005 and eventually confirmed through the retrospective cohort study and molecular characterization of HCV molecular variant.

This investigation showed that the 13 incident cases occurred by the end of September 2005 and for 10 of them HCV genotype was available in the clinical records (all genotype 2c). The Poisson regression which included all the 45 susceptible subjects enrolled in the retrospective study found that undergoing HD on MWF-pm shift was the only independent factor associated with the infection. The subsequent analysis of patients' flow in this shift showed that all the 11 susceptible subjects who underwent HD on 20 April afternoon eventually became incident cases. Having analyzed the medical activities reported on the logbook for that day, we found that a prevalent case (index cases) underwent HD in the morning; she had several problems with her arterovenous fistula, bled and needed several additional doses of heparin. According to this evidence and the data obtained from the audit, we hypothesised that the 11 susceptible subjects who underwent HD in the subsequent shift were infected through contaminated heparin doses prepared during the morning by the HCWs who were simultaneously involved in caring after the index case and preparing the heparin doses for the next shift. It is possible that, during the care of the index case, shared 250 ml saline solution(s) and/or shared 20 ml heparin vial(s) were contaminated by repeated drawings and injections of materials and then re-used to prepare the heparin doses for the afternoon shift. It is also possible that during the bleeding emergency the area where the heparin doses were prepared was contaminated by HCWs who failed to change gloves after caring for the index case.

Of these 11 incident cases 6 were confirmed by the molecular analysis of both NS5B and HVR-1 viral region, 2 were proved to be infected with a genotype 2c HCV with an identical NS5B sequence to the index case and other confirmed cases while for 3 incident cases we know they were infected with genotype 2c HCV by data recovered from their clinical records.

Two additional incident cases did not attended the 20 April HD. We cannot clearly define how these patients contracted the infection but there is a possibility they might have been infected as a result of: lacks of standard precautions, the increased prevalence of infection occurred after the onset of the first cluster of infections and the decision of spreading the incident cases over different shifts instead of cohorting them. The analysis of NS5B and HVR-1 viral region confirmed that both these case were infected with the same HCV molecular variant of the index cases and all other confirmed cases.

Several mechanisms such as: reuse of dialysis filters, internal contamination of HD machines and contamination of environment, can potentially contribute to patient-to-patient HCV transmission in HD unit. However lack of standard precautions, including hand-washing, proper use of gloves and safe injection practices, have been recognized as leading causes of HCV outbreak in HD settings [7, 8, 15–19].

HCV transmission is a major concern in HD units in Italy [2] and other developed countries [8, 15, 16, 18, 19]. In fact, the close proximity of patients who require frequent vascular access and high prevalence of chronic HCV carriers in this population may enhance the chance of cross-infections when standard precautions are not strictly applied. Savey et al described a large outbreak of 61 patients in a French HD centre mainly supported by patient-to-patient transmission via HCWs' hands [18]. In addition a recent systematic review on transmission of HCV in healthcare settings in USA between 1998 and 2008 reported that inappropriate use on multiple patients of single use vials and failure to store and prepare medications under aseptic conditions were responsible for HCV transmission in 15 outbreaks occurring in nonhospital healthcare settings (6 of which were HD units)[16]. As a consequence, to prevent transmission of blood borne viruses in HD settings, Centers for Diseases Control and Prevention (CDC) recommend that all single-use injectable medications and solutions should be dedicated for use on a single patient and should be entered one at a time only [20].

In a recent paper CDC's investigators outline one possible way in which inappropriate use of medication vials intended for single-person may result in large HCV outbreaks. Through direct observation they found that HCWs used to draw medication from a single-use vial of propofol with a sterile needle. The medication was injected directly through an intravenous catheter into the patient's arm. If a patient required more sedation, the needle was removed from the syringe and replaced with a new needle; the new needle with the old syringe was used to draw more medication. Backflow from the patient's intravenous catheter or from needle removal contaminated the syringe with HCV and subsequently contaminated the vial [21]. A similar transmission pathway has been reported by Germain et al [17] and could have occurred in the present outbreak. An additional factor that may have contributed to the spreading of the infection was that medication were not handled, as recommended [22], in a clean separated area, in fact, heparin doses were prepared by HCWs, while attending a HD shift and in the same room where the patients were dialyzed.

Another important flaw we found in the organization was that the unit was not provided with any protocol for risk management nor was a single HCW appointed to act as risk manager so that, the alert to local health authority was given only by September 2005, while the first cases had already occurred by the end of July 2005. In addition at the onset of the outbreak, incident cases were allocated in different shifts, a decision likely to have favoured the occurrence of the 2 additional cases. With regard to the importance of this latter issue, a topical Canadian experience in dealing with the recent lethal Clostridium difficile outbreaks emphasizes that a comprehensive system of risk management including dedicated staff and prompt notification of cases was the single most important factor in containment of ongoing epidemics and in preventing new ones[23].

Although data is limited, this investigation suggests that, among end-stage renal disease subjects, HCV acute infection is not much more severe than acute infection in the general population, as no patient developed fulminant hepatitis and the 38.4% (5 out of 13) spontaneous recovery rate we found was not lower than the rate of spontaneous viral clearance reported in the general population[1].

The limitations of this study are that a specific mode of transmission could not be identified for 2 of the 13 patients. Moreover since HCV-RNA testing was performed only after the onset of the outbreak and the transaminases were tested only every 3 months, we could not define the exact time between the infection and the onset of viremia and/or the liver injury which would have been precious to further describing the natural history of HCV infection in subjects with end stage renal disease. In addition we could not perform a comprehensive investigation of the unit, so that the potential role of the contamination of HD machines and of the environment was not completely evaluated. Finally potential misclassification of the 5 possible cases may have occurred since it cannot be completely ruled out that these subjects were part of other cluster(s) trigged by one or more of the three prevalent cases without molecular characterization. Nevertheless we consider this latter hypothesis quite unlikely since all possible cases underwent HD on 20 April afternoon and 2 of them shared an identical genotype and N5SB with the other confirmed cases.

Apart from illicit drug addiction, medical procedures are the most important cause of HCV infection in developed countries. It has been estimated that in western countries more than 30% of all HCV infections results from poor application of standard precaution during percutaneous procedures [24]. The outbreak described here should serve as a reminder to HD providers that patients undergoing dialysis are at risk for HCV infection and that HCV may be easily transmitted whenever standard precaution, including safe injection practices and proper use of gloves are not strictly applied [25]. We believe that a comprehensive strategy including: educational programmes, periodical auditing on standard precaution, use of single-dose vials whenever possible, prospective surveillance for blood-borne infections (including a system of prompt notification) and risk assessment/management dedicated staff are the cornerstone to contain and prevent outbreaks in HD.