The prevalence of anti-HCV antibodies in patients receiving HD (31.1%) was remarkably high and is approximately 25 times higher than in the general population
. It is also higher than the prevalence of 20.5% reported by Daw et al. in a sample of 200 HD patients in 2001 in Libya
. Globally the prevalence of HCV among patients receiving HD varies from as low as 6.1% in Germany
 to as high as 76% in Casablanca
. In general, North Africa and the Middle East are high prevalence areas both in the general population and in HD patients
. Previous studies from the region have reported a prevalence of anti-HCV antibodies in HD patients of 50% in Saudi Arabia
, 42% in Tunisia
, 20.2% in Turkey
 and 21% in Jordan
. In contrast, the observed prevalence of HBV infection (2.6%) is similar to the general population and similar to that reported in HD patients in other regions including Europe (4.1%), Japan (2.2%) and the USA (2.4%)
. A study sample from the Dialysis Outcome and Practice Patterns Study that included 8615 adult HD patients from 308 dialysis facilities in Western Europe and the United States, reported prevalence rates for HBV infection ranging from 0% to 6.6%
. Studies from less developed countries estimated that the proportion of HBsAg carriers in the HD population varies from 2% to 20%
[29–32]. According to the 2008 Saudi Centre for Organ Transplantation (SCOT) report, HBV sero-positivity was 4.6% in the Saudi HD population
 while among Jordanian HD patients it was 5.9%
. In general, the prevalence and incidence of HBV and HCV infections in HD patients reflects the prevalence of these infections in the general population, the quality of healthcare services in a community and the standards of infection control practices in HD units. The importance of HBV and HCV as a health risk in patients on HD is illustrated by our observation that 3% of deaths in Libyan HD patients during a 1 year observation period were due to liver failure and that 13 of the 14 patients who died of liver failure were sero-positive for HCV and/or HBV
Our data show that sero-positive patients were significantly younger on average than sero-negative patients. This observation is in agreement with a previous report from Libya showing that the highest prevalence of HCV antibodies was observed in HD patients aged 36–55 years
. Other studies
[28, 36, 37] have reported a higher prevalence of HBV or HCV sero-positivity in older patients and the reason for this difference in not clear. On the other hand, the prevalence and incidence of HBV or HCV sero-positivity was significantly related to the length of time on HD. This is consistent with nosocomial transmission related to dialysis since longer duration of dialysis represents a longer period at risk of acquiring an infection. Similar observations have been reported by other authors
[38–41]. Prevention of nosocomial transmission is of vital importance in Libya as HCV antiviral treatment is expensive and its availability is limited to only a few centres.
A positive history of blood transfusions as well as the number of blood transfusions was strongly associated with HBV or HCV infection at baseline, but not with new infections. Prior to the introduction of effective screening of blood donors, blood transfusions were recognised as the leading source of HCV infection and some of these infections may have been acquired before adequate screening was introduced
[21, 40]. In addition it is possible that some blood donors with HCV infection are being missed by current screening procedures and these may need to be reassessed
[42, 43]. On the other hand the lack of association between blood transfusions and new infections suggests that fewer infections are acquired by this route than previously. A large proportion of patients had previously received blood transfusions. The risk of infection could therefore be further reduced by more effective management of anaemia with iron supplementation and erythropoietin. In accordance with other studies
[41, 44, 45], HBV or HCV infection was more prevalent in patients with a history of previous renal transplant. Infection in these patients might have been transmitted from an infected donor kidney or blood transfused peri-operatively. This observation emphasizes the need for adequate screening of potential kidney donors, which is deficient in some countries. The shortage of donated kidneys in Libya induces many patients to seek a transplant abroad.
Another concern raised by the current study is that HBV or HCV infection was associated with a history of HD in another centre either in Libya or abroad. Many patients travel for social reasons but some also transfer to a maintenance HD centre after initiating dialysis as an emergency in a specialised centre providing acute services or may travel to another centre for surgery to create an arteriovenous fistula
. The association of hepatitis virus infection with travel suggests that the risk of nosocomial infection varies between dialysis centres within Libya and abroad. The former is confirmed by our data showing a marked variation in both prevalence and incidence of HBV and HCV infection among Libyan HD units (Figures
2). These observations emphasize the importance of isolating patients following their return and monitoring them for sero-conversion.
Prospective follow up of sero-negative HD patients enabled us to verify 89 sero-conversions for HBV or HCV during 1 year, giving an overall incidence of 7.7% for new infections. The incidence rate of 0.6% for HBsAg sero-conversion is similar to that reported in Europe, Japan and the USA (0.4-1.8 per 100 patient-years)
. Three new HBsAg positive patients were detected in a single centre that was treating 20 other HBsAg positive patients and 2 new cases were detected in another centre that was treating 14 HBsAg positive patients, suggesting that nosocomial transmission probably occurred. We observed a high incidence of new HCV infections during the 1-year observation period (7.2%). The reported incidence of new HCV infections varies considerably between countries. A rate as low as 0.4% was observed in France from 1997 to 2000
 but higher rates have been reported from the Mediterranean region. According to the 2008 SCOT report, the annual rate of HCV sero-conversion in Saudi HD patients was 7-9%
 while in Jordan it was 2.6%
. In our study most new cases were observed in centres treating other patients with HCV infection, suggesting nosocomial transmission. Interestingly 9 new HCV infections were observed in one unit and 2 in another that previously accepted only patients without HCV infection. This raises the possibility of transmission from a carrier that was not detected by current screening procedures.
A striking observation from this study is the wide variation in incidence and prevalence of HBV and HCV infections among different HD units (Figures
2). Interestingly none of the potential centre-related factors that we assessed formally explained this variation. On the other hand, we observed variations in other practices that may be relevant. Most facilities faced a problem of increasing number of patients and most of them responded by adding more HD stations at expense of space and staff. Infection control precautions also varied widely between centres. They were strictly enforced in some places but frequently breached in others. This seemed to depend on staff initiative rather than national guidelines. On the other hand, dialyser reuse was not permitted and all bloodlines as well as other consumables were disposed after a single use
. Some brands of HD machines were equipped with a sphygmomanometer. Otherwise, most non- disposable instruments used in HD environment were shared between sero-positive and sero-negative patients. The use of multi-dose vials of heparin was common and is likely to have been an important cause of nosocomial infections. Many patients started HD without being vaccinated against HBV. Even in vaccinated patients the antibody titre was not assessed. The wide variation in HBV and HCV prevalence and incidence between dialysis centres implies that there is potential to reduce blood-borne virus infection by transferring best practice from HD centres with low infection rates. In particular infection control procedures should be investigated in centres with high infection rates and the use of multidose heparin vials must be stopped urgently. Previous studies from the region show that with appropriate intervention HCV infection rates in HD centres may be substantially improved. For example in Iran, HCV prevalence reduced from 14.4% in 1999 to 4.5% in 2006
 and in Saudi Arabia from 2.4% in 2001 to 0.2% in 2005
Several limitations of this study should be conceded. Medical records were often incomplete and additional clinical information was frequently obtained by interviewing staff and patients. Serological testing was done in local laboratories and it is likely that there was some variation in the quality of testing. Data regarding hepatitis B core antibodies (HBcAb) or hepatitis B DNA were not available. In one recent study of haemodialysis patients in Egypt who were negative for HBsAg, hepatitis B DNA was detected in 4.1% and HBcAb in 20%
. It is therefore possible that we failed to detect cases of occult hepatitis B infection. Testing for HCV relied on a third generation ELISA to detect anti-HCV antibodies and confirmation or genotyping with PCR is currently not available in most centres.