Guideline 3: Preventing HCV transmission in hemodialysis units

INTRODUCTION
Dialysis units have responsibility for ensuring that blood-borne viruses are not transmitted among the patients in their care.

3.1 Hemodialysis units should ensure implementation of, and adherence to, strict infection-control procedures designed to prevent transmission of blood-borne pathogens, including HCV. (Strong)

• Isolation of HCV-infected patients is not recommended as an alternative to strict infection-control procedures for preventing transmission of blood-borne pathogens. (Weak)
• The use of dedicated dialysis machines for HCVinfected patients is not recommended. (Moderate)
• Where dialyzer reuse is unavoidable, it is suggested that the dialyzers of HCV-infected patients can be reused provided there is implementation of, and adherence to, strict infection-control procedures. (Weak)

3.2 Infection-control procedures should include hygienic precautions (Tables 18 and 19) that effectively prevent the transfer of blood-or fluids contaminated with blood-between patients, either directly or via contaminated equipment or surfaces. (Strong)

• It is suggested to integrate regular observational audits of infection-control procedures in performance reviews of hemodialysis units. (Weak)

BACKGROUND
Transmission of HCV is primarily via percutaneous exposure to infected blood. HCV can remain viable in the environment (on equipment, clothing, and so on) for at least 16 hours.²¹¹

The prevalence of HCV infection in hemodialysis patients is significantly higher than in the general population.^36,58 Transfusions before donor blood screening for HCV undoubtedly caused many cases of HCV in dialysis units. Still, correlation between HCV infection and time on dialysis, higher prevalence in hemodialysis than peritoneal dialysis or home hemodialysis, and the highly variable prevalence from unit to unit all suggest that nosocomial transmission has also contributed to the high prevalence.^36,212,213 The occurrence of nosocomial transmission was confirmed when phylogenetic analysis identified clusters of closely related isolates of HCV, both in studies of individual units with high seroconversion rates^214,215 and multicenter studies.^216,217 Parts of the HCV genome (especially hypervariable region 1) are highly variable and lend themselves to fingerprinting of each isolate or quasispecies using nucleic acid sequencing. This may be used to establish a firm basis for studies of spread and routes of infection by HCV.^218,219

Since the dramatic reduction in the 1990s of the risk of post-transfusional HCV, nosocomial transmission is the most likely source when hemodialysis patients develop HCV antibodies. Tables 20 and 21 show the results of a systematic review of studies of HCV infections in hemodialysis in which nosocomial transmission was confirmed by phylogenetic analysis, and the route of transmission was investigated by the authors.

RATIONALE

3.1 Hemodialysis units should ensure implementation of, and adherence to, strict infection-control procedures designed to prevent transmission of blood-borne pathogens, including HCV. (Strong)

Nosocomial transmission of HCV in hemodialysis units has been confirmed using epidemiology and/or molecular virology by many authors (see Background). The most likely cause of HCV transmission between patients treated in the same dialysis unit is cross-contamination from supplies and surfaces (including gloves) as a result of failure to follow infection-control procedures within the unit. Transmission via the internal pathways of the dialysis machine can be excluded for most machines (see below). Other possible transmission routes are direct contact between the patients, a common infected blood donor, and invasive procedures outside the unit with contaminated material used for both the source and the newly infected patient.²¹⁹ The two latter causes are currently very unusual^220-222 and can generally be excluded using the patient's medical records. The sharing of a contaminated medication vial was identified as the transmission route in one study involving the simultaneous infection of five patients.²²³

A systematic review of molecular virology papers that included both confirmation of the source patient(s) and an investigation of possible transmission routes was carried out. Twenty studies, involving between 1 and 22 newly diagnosed cases of HCV, were identified (see Tables 20 and 21). The authors of all 20 studies were unable to conclusively establish the specific transmission route(s), but all considered breaches in infection control, including failure to decontaminate pressure ports in one case,²²⁴ to be the probable cause of the outbreak.

Transmission of the virus via internal fluid pathways of the dialysis machine was considered to be a possibility in only one study,²²⁵ whereas in 18 of 20 studies, the authors reported that some or all patients with new HCV infection had never shared the dialysis machine with the source patient (see Table 21), either because they dialyzed at the same time or because the unit policy was to assign HCV-positive patients to separate machines.

Overall, the evidence from this systematic review of molecular virology studies strongly suggests that the internal hemodialysis machine circuit is, at most, a minor contributor to the nosocomial transmission of HCV among hemodialysis patients. There is no reason to believe that a publication bias has suppressed the reporting of nosocomial transmission related to the dialysis equipment or favored reporting of transmission due to breaches in infection-control procedures.

• The isolation of HCV-infected patients is not recommended as an alternative to strict infection-control procedures for preventing transmission of blood-borne pathogens. (Weak).

In the absence of any good RCTs of the impact of isolation on the risk of transmission of HCV to hemodialysis patients, the available evidence is limited to observational studies.

Many authors have reported a reduction (but not full prevention) of HCV transmission in hemodialysis after the adoption of an isolation policy, either dedicated machines for HCV-infected patients or a separate ward. All these studies were of the after-before (or before-after) type and none included a control group.^226-228 Thus, it is unclear whether the reported improvement resulted from the adoption of an isolation policy or rather from the simultaneous raising of awareness and reinforcement of the application of hygienic precautions.

Currently, the best available evidence on the impact of isolation measures on HCV transmission to hemodialysis patients derives from two large prospective observational studies. The DOPPS⁴⁰ and an Italian study²²⁹ concur that, after multivariate adjustment for potential confounders- especially the prevalence of HCV infection within each hemodialysis unit-isolation does not protect against HCV transmission in hemodialysis patients.

Some prospective observational studies^230,231 have reported a reduction of HCV transmission after the reinforcement of basic hygienic precautions, without any isolation measures. In particular, one Belgian prospective multicenter study²³⁰ showed a reduction from 1.4 to 0% of the yearly incidence of seroconversion for HCV. This demonstrated that complete prevention of HCV transmission to hemodialysis patients was possible in the absence of any isolation policy.

Additional arguments against relying on the use of isolation to prevent transmission of HCV include the possibility of increased risk of HCV infection with more than one genotype and the time between infection and seroconversion. The seroconversion time ('window') can be over a year²³² and has a median length of 5 months in hemodialysis patients even with third-generation EIA tests.⁸⁶ This will result in inadequate selection of patients to be isolated, unless costly NAT is performed frequently.

If nosocomial transmission continues to occur, despite reinforcement and audit of the precautions listed in Tables 18 and 19, a local isolation policy may be deemed necessary. HCV-infected patients should be treated by dedicated staff in a separate room, area, or shift (morning, afternoon, or evening), as there is no rationale for using dedicated machines. It should be realized that accepting the 'need' for isolation equates to accepting the impossibility of full implementation of basic hygienic precautions, a regrettable situation that entails the risk of transmission of pathogens other than HCV.

• The use of dedicated dialysis machines for HCV-infected patients is not recommended. (Moderate)

In theory, it is almost impossible for a virus to pass through the dialyzer membrane of an infected patient and migrate from the drain tubing to the fresh dialysate circuit to pass through the dialyzer membrane to infect another patient. The impermeability of an intact dialyzer membrane and the requirement for backfiltration in the second session add to the improbability of this mode of transmission.

Of the studies summarized in Table 20, the majority were able to discount transmission via the internal pathways of the dialysis machine easily as the patients involved in the outbreak were dialyzed at the same time and/or on separate machines. Interestingly, several of these reports (Table 20) documented nosocomial HCV transmission despite the existence of a policy of dedicated monitors for HCV-infected patients. This underscores the substantial limitations of such a policy.

Only two studies considered that the machine may have been involved in the transmission of HCV. One concluded that lack of internal disinfection between patients was a possible (but not proven) transmission route, although one of the three confirmed cases of nosocomial transmission that were followed up never dialyzed on the same machine as the source patient.²²⁵ Breaches in infection-control procedures, such as failure to change gloves in emergencies, were also cited and thus appear as an alternative (more likely) explanation.

The strongest case for nosocomial transmission via the dialysis machine was reported in one study.²²⁴ The infected patient was not dialyzed at the same time as the source patient in the period when infection took place, but was dialyzed on the same machine 21 times. However, transmission via the internal pathways could be excluded as the machine was disinfected between shifts. Environmental contamination could not be excluded, but the authors concluded that contamination of the venous pressure port was the likely transmission route. If the precautions in Table 19 are followed, transmission via this route will be prevented without the need to use dedicated machines for patients with HCV.

The high quality of the evidence against transmission of HCV via the internal pathways of the dialysis machine is the basis for the recommendations on disinfection in Table 19 and the recommendation that dedicated machines should not be used for patients with HCV infection.

A single study has claimed that the random assignment of hemodialysis units to dedicated machines for HCV-infected patients reduced the incidence of seroconversion for HCV.⁴³ However, the authors did not disclose details of the randomization procedure, the policy of the participating units before randomization (four to use of dedicated machines, eight to shared machines), or whether the patients who seroconverted had actually shared machines with infected patients. In addition, the authors stated that machines in all units were disinfected with bleach between sessions and that interviews with nurses revealed some deviation from the CDC guidelines on hygienic precautions. One unit in the group using shared machines was eliminated from the analysis due to nonadherence with CDC guidelines. The incidence of seroconversion for HCV was substantial, even in the dedicated machines group, for a relatively low prevalence at study start. Overall, this strongly suggests that the transmission was related to breaches in infection-control procedures and not to the sharing of machines.

The possibility that use of dedicated machines acts as a reminder to the staff to implement procedures cannot be discounted, but it should be possible to raise awareness using methods that do not (i) restrict the availability of dialysis or (ii) restrict the choice of dialysis location, shift, or treatment modality of HCV-positive patients compared to uninfected patients.

• Where dialyzer reuse is unavoidable, it is suggested that the dialyzers of HCV-infected patients can be reused provided there is implementation of, and adherence to, strict infection-control procedures. (Weak).

The main risk for HCV transmission associated with the reprocessing of dialyzers is to the staff involved (beyond the scope of this guideline). Theoretically, contaminated blood could be transferred if dialyzers or blood port caps²³³ that have not been sterilized effectively are switched between patients, but this should not occur if procedures are followed correctly. In addition, there are risks associated with the transport of contaminated equipment, but these risks should be eliminated by strict adherence to hygienic precautions (Table 19).

Dialyzer reuse was not identified as a risk factor for seroconversion for HCV in the CDC surveillance data³⁴ or in the Belgian prospective multicenter study.²³⁴ The weak association of dialyzer reuse with HCV infection in one study carried out in Portugal²³⁵ may reflect an association with unmeasured confounders, such as the degree of actual implementation of basic hygienic precautions.

3.2. Infection-control procedures should include hygienic precautions that effectively prevent the transfer of blood-or fluids contaminated with blood-between patients, either directly or via contaminated equipment or surfaces. (Strong)

As HCV is transmitted by percutaneous exposure to infected blood, effective implementation of the hygienic precautions detailed in Tables 18 and 19 should prevent nosocomial transmission. The precautions listed differ very little from the extensive recommendations of the CDC⁵³ and the rather brief guidelines provided by the European Renal Association.²³⁶

The recommendation of the CDC is that 'single-pass' machines do not require disinfection between shifts on the same day, even when a blood leak has occurred. There is a very low risk that a virus leaving the dialyzer could be trapped in the Hansen connector and transferred to the fresh dialysate side through accidental misconnection. Under normal conditions, the dialyzer membrane should provide an effective barrier and make the risk of transmission negligible, but when a blood leak occurs, the risk is slightly greater. Although the extra disinfection may be unnecessary, as blood leaks are now relatively rare, it is unlikely to affect the management of the unit.

• It is suggested to integrate regular observational audits of infection-control procedures in performance reviews of hemodialysis units. (Weak).

There are few published studies of observational audits of infection control in hemodialysis. One study in Spain²³⁷ audited hand hygiene in nine hemodialysis units and showed gloves were used on 93% of occasions indicated by unit policy. Hands were washed only 36% of the time after patient contact and only 14% of the time before patient contact.

Observational audits of hygienic precautions were carried out in two of the outbreak investigations in Table 20.^85,238 Both identified a range of problems, including

• lack of basic hand hygiene;
• failure to change gloves when touching the machine interface to obtain biologic parameters, or when urgently required to deal with bleeding from a fistula;
• carrying contaminated blood circuits through the ward unbagged;
• lack of routine decontamination of the exterior of machines and other surfaces even when blood spillage had occurred;
• failure to change the internal transducer protector when potentially contaminated.

Where hygienic practice was reviewed through interviewing staff after an outbreak²³⁹ rather than by observation, no obvious breaches in procedure could be identified.

If HCV-negative patients are routinely screened for seroconversion, the absence of new infections can provide evidence of adherence to the procedures in units with high prevalence. Screening results are not a substitute for regular assessment of the implementation of hygienic precautions, especially in units with few or no infected patients. The frequency at which routine audits of infection-control procedures should be carried out will depend on staff turnover and training, and on the results of previous audits. When setting up a new program, audits should be at intervals of no more than 6 months to enable staff to gain experience with the process and ensure that any remedial actions taken have been effective.

IMPLEMENTATION ISSUES

• It is important for the designers of dialysis units to create an environment that makes infection-control procedures easy to implement. Adequate hand-washing facilities must be provided, and the machines and shared space should make it easy for staff to visualize individual treatment stations.
• The unit should ensure that there is sufficient time between shifts for effective decontamination of the exterior of the machine and other shared surfaces.
• The unit should locate supplies of gloves at enough strategic points to ensure that staff have no difficulty obtaining gloves in an emergency.
• When selecting new equipment, ease of disinfection should be taken into account.
• There are indications from the literature that the rate of failure to implement hygienic precautions increases with understaffing.^72,229 One study⁸⁵ describes how a large HCV outbreak occurred when an expansion in patient capacity that led to understaffing was aggravated by rapid staff turnover. Dialysis units that are changing staff-to-patient ratios, or introducing a cohort of new staff, should review the implications on infection-control procedures and educational requirements.
• Resource problems should be handled by carrying out a risk assessment and developing local procedures. For example, if blood is suspected to have penetrated the pressure monitoring system of a machine but the unit has no on-site technical support and no spare machines, an extra transducer protector can be inserted between the blood line and the contaminated system so that the dialysis can continue until a technician can attend to the problem.

RESEARCH RECOMMENDATIONS

• Large national or international prospective observational studies (such as the DOPPS) could be used to capture important epidemiologic data (HCV infection in new patients and seroconversions in prevalent patients). This information is expected to show a significant trend toward infection occurring before starting dialysis and lower rates of subsequent seroconversion.
• If studies such as DOPPS can also capture the hemodialysis unit's isolation policies, and the routine methods through which the use of strict infectioncontrol procedures is reiterated and monitored, the correlation between practice and nosocomial transmission could be strengthened.
• The systematic review carried out during the development of this guideline was limited to papers where nosocomial transmission was confirmed through molecular virology and some attempt had been made to identify the route of transmission. It is reasonable to assume that, unless the patient has other significant risk factors, all new infections in hemodialysis patients are the result of nosocomial transmission and that an investigation should be undertaken. If units are prepared to share their findings, the publications would provide useful educational material and help inform auditors.
• In addition to epidemiologic studies and reporting of outbreak investigations, generic educational materials, operating procedures, and audit tools that can be adapted for local implementation should be developed and made available in a range of languages.