Despite the advances in dialysis technology, infections that develop in hemodialysis patients remain important, and prevention of these infections is mainly based on surveillance. Data from the National Nephrology, Dialysis and Transplantation Registry Report of Turkey showed that with a rate of 9.98%, infections are one of the most common death causes in HD patients. Type of vascular access is important for the risk of infection. Now, it is a well-known fact that, AV fistula has the lowest risk of infection. Therefore, it has been a core aim to increase its usage as a prevention measure for dialysis events. “Fistula First” initiative of the Centers for Medicare and Medicaid Services campaigned for use of AV fistula and set objectives to increase AV fistula usage above 66% meanwhile decreasing CVC below 10% among prevalent hemodialysis patients. Although nationwide data in Turkey show that the >66% goal has been achieved with a rate of 77.41%, the facilities included in this surveillance program had lower rates. In our study, rate of the patients receiving HD via CVC was 42.2% and for these patients, the risk of bacteremia was 17.6 times higher than the patients with AV fistula. NHSN Dialysis Safety Network reported that the bloodstream infection (BSI) rate ratio between CVC and AV fistula was 8.2. In our study population, proportion of patients receiving HD via catheter was higher than the rate reported in the national registry report. Since centers involved in the study are all tertiary hospitals’ HD units, patients’ characteristics can be different from the general population. Proportion of patients with tunneled catheter was high due to the centers included in the study being referral centers that predominantly give healthcare to complicated patients and existence of transient dialysis patients. Since the main goal of the study is not to compare infection rates across catheter types, baseline patient characteristics are different across these groups. These can explain lower infection rates obtained in non-tunneled CVC patients.
In 1999, the Centers for Disease Control and Prevention (CDC) developed the Dialysis Surveillance Network for monitoring hemodialysis related infections. Definitions updated over the years and a few reports regarding the data have been published. In the recent years, several countries started to implement this surveillance system. Surveillance plays critical role for improving health care quality and safety of HD patients and can guide infection control programs, find the gaps where improvements might be needed. Moreover, this approach can raise the awareness of the healthcare workers regarding infection prevention. Gork et al reported a significant decrease in dialysis related infections in a 9 year lasting study period 2. Their study consisted of both surveillance and intervention which includes checklists, ready kit for the care of vascular access, education, and an infection prevention team. They achieved a significant trend of decrease in access-related infection rates. Integration a HD specific surveillance system in HD units can cause considerable decrease in HD related infections as well as antimicrobial consumption 3. NHSN reported a significant decrease in BSI and access related infection rates in 2014 4. However, rates of intravenous antimicrobial start were similar with previous reports. This result highlights the need for different efforts for achieving similar lower rates for antimicrobial consumption and lowering rates of colonization and infection of HD patients with multidrug resistant organisms. Tracking antimicrobial use and antimicrobial resistance of organisms in HD patients is essential for programs to prevent antimicrobial resistance.
As in surveillance of hospital-acquired infections, standard definitions and calculations are important in terms of monitoring and comparing units' own rates. NHSN did not report DE rates since 2014. Our DE rates for all types are lower than the rates reported by the NHSN in 2014 and 20114,5. There are several reasons for lower DE rates. First, all centers participating in the study are at a tertiary care university hospital with high workload and they are all strictly controlled by the government. Benchmarking of data from a dialysis unit at a tertiary care university hospital with data from US outpatient dialysis units is not optimal, since there can be many differences in patient population, staff education and facility of the unit. On the other hand, NHSN reports significant differences in DE rates among facilities in US. Second, since this is the first study for surveillance of DE infections in our country, healthcare personnel are not familiar with this type of a data gathering system and this might have caused some underreporting. Even NHSN which implemented the system many years ago, discuss the quality of the data and problems understanding the system. However, in our opinion implementation of an official nationwide surveillance system would have a significantly positive effect on data quality. Third, in NHSN report 2014, 6005 outpatient HD facilities reported data. Both the number of patients and the number of centers are very high when compared to our study. Therefore, it is not appropriate to compare these DE rates. There have been many differences among DE rates reported from different countries. Quebec Public Health Expertise and Reference Centre reported that their vascular access related BSI rate was 0.22 cases per 100 patient-periods 6. They report that in 2016–2017, incidence rates for tunneled and non-tunneled catheters have significantly decreased compared to rates for 2012-2016 while rates for AV fistulas and grafts have remained stable. In a surveillance study from Kuwait reported rates of hospitalization, IV antimicrobial start, and positive blood culture were 4.3, 9.0 and 1.1 per 100 patient-months respectively 7. In an Irish study from two HD units for a period of 6 weeks, rate of hospitalizations, IV antimicrobial starts, and positive blood cultures were 13, 8.52 and 3.14 per 100 patient months, respectively 8. China reported 33 outpatient HD centers’ surveillance of dialysis events data in 2017 9. Overall DE rate was 1.47 per 100 patients-months.
In contrast to published surveillance studies, we also searched vaccination rates for seasonal influenza and pneumococcal vaccine which are recommended for HD patients. Unfortunately, 32.2% for seasonal influenza vaccination and 5.8% for pneumococcal vaccination rates are both very low. Bond et al reported that vaccination against influenza and pneumococcal disease is associated with improved survival in dialysis patients 10. Infection prevention strategies must include topics regarding raising vaccination rates for HD patients.
Most reported microorganisms responsible for BSI were coagulase-negative staphylococcus and S. aureus. This data is similar with NHSN’s and other studies’ data 4,6,7,9. Rate of methicillin resistance among S. aureus was 46% in NHSN’s 2014 report and 1% in our report. In Qebec report S. aureus accounts most of the cases resulting in death (44%), however oxacillin resistance was 10.7% among S. aureus 6. China reported 17.86% methicillin resistance among S. aureus isolates 9.
The CDC published key interventions for prevention of BSI in HD patients. According to these recommendations, education for vascular access site care for all HD patients is essential. One of the most crucial results of this study is the low infection rate among patients who received education about the care of their vascular access site. This result revealed the importance of engaging patients for prevention strategies. Data on the HD patients’ perspective on infection prevention is limited in the literature. In a survey study, CDC staff evaluated the HD patients’ view and role on infection prevention strategies 13. Participating patients concluded that patients should take the responsibility for their vascular access site care and should be observant for infection prevention steps. HD patients spend much of their time in healthcare facilities, and they can have a positive effect for both their and other patients’ safety regarding infections 14. In our study population, 80% of the patients were educated. Next step for lowering the BSI rates among HD patients must be to increase this rate to 100%.
Making Dialysis Safer for Patients Coalition was founded by CDC and CDC Foundation in 2016 16. The coalition consists of a wide range of healthcare organizations and stakeholders and aims to prevent bloodstream infections in HD patients and raise awareness on recommended infection prevention practices. “Core Interventions for Bloodstream Infection Prevention” compiled by the coalition includes evidence-based practices for CVC care as well as benchmarking data collected through NHSN related with infection rate measures, education of staff and patients, audit, and competency assessments. Several centers reported significant, rapid and sustained reductions in DE rates after participating the Collaborative 17–19. After the early success of the Collaborative was shown, CDC compiled checklists regarding the Core interventions used by Collaborative participants 20. These checklists focused on hand hygiene and glove use, catheter exit site care, catheter connection and disconnection, arteriovenous fistula and graft cannulation and decannulation, and routine dialysis station disinfection. The most important outcome of our study was preparation of similar checklists for the Dialysis Services Unit of the Turkish Ministry of Health. After evaluating our project, Ministry of Health General Directorate of Health Services officially distributed these checklists to all dialysis centers in Turkey. Moreover, they made it mandatory to complete checklists for each HD patient. This was an unexpected result of utmost impact that we neither have intended nor foreseen at the beginning of the study.
This study has some limitations. First, like NHSN, we included data from all participating HD units regardless of the number of months reported. This can certainly lower the quality of the data. Second, all centers participating in the study are tertiary hospitals. Therefore, rates may not reflect the national data. Third, we conducted the study with a limited number of clinicians and research staff. High workload at these centers might have increased the probability of underreporting.
In conclusion, hemodialysis units are not covered in the National Nosocomial Infection Network run by the Turkish Ministry of Health General Directorate of Public Health. This first surveillance study revealed the baseline status of HD related infections in Turkey and showed that NHSN DE surveillance system can be easily implemented even in a high workload dialysis unit and be adopted as a nationwide DE surveillance program. Results have highlighted the importance of optimizing vascular access, appropriate care of catheters and the patient education for vascular access site care. Awareness of healthcare workers regarding infections in HD patients is one of the most important points of preventing, and this study provided a great contribution for raising awareness of healthcare workers in dialysis units. Revealing that DE rates are lower in patients who are educated about the care of their vascular access site will hopefully make healthcare workers more attentive. While there are caveats with international comparisons as discussed above, we have established a baseline that will facilitate us to demonstrate the effect of future infection prevention and control and antimicrobial stewardship strategies.