Catheter related blood stream infections – prevalence and interventions

Summary

Introduction:
Catheter related blood stream infections are a significant complication of intensive care with worldwide prevalence rate around 5 cases per 1000 catheter-days. Only scanty Czech data have been published. Our study monitored the occurrence of catheter-related blood stream infections in a high dependency unit of regional hospital.

Methods:
In 2008 we commenced to monitor the occurrence rate of catheter-related blood stream infections in short-term central venous catheters without antimicrobial coating. We organized a training session for medical staff and started to strictly adhere to published guidelines. After two years of keeping a register we analysed individual cases as proven, possible, or not proven blood stream infections.

Results:
From March 2008 to March 2010 we inserted 142 central venous catheters for total time of 1423 catheter days (median 9 days). Ten catheters were removed after median of 17 days due to unexplained pyrexia. Blood stream infection was proven in 4 cases and possible in 2 cases. We have noted total 2.81 proven cases, and 4.22 proven and/or possible cases of blood stream infection per 1000 catheter-days.

Conclusion:
The register of catheter related blood stream infections is an inexpensive and time-efficient tool that improves the quality of intensive care.

Key words:
catheter-related bloodstream infection, prevalence, hospital-acquired infection

Introduction

The most frequently encountered hospital acquired infections include surgical site infections, respiratory and urinary tract infections, and blood stream infections (BSI). Catheter related BSIs are significant cause of increased morbidity and mortality in intensive care. The published data of the Centers for Disease Control (CDC), World Health Organization (WHO) and meta-analyses of epidemiological studies show the prevalence rate of CR-BSI in the North America and Europe around 5 cases per 1000 catheter days (1-4). The highest risk of CR-BSI is present in short-term non-tunneled central venous catheters (CVC) without antimicrobial coating (3,5). Apart from the type of used material, CR-BSI occurrence is influenced by other factors – site and circumstances of insertion, the quality of care over the insertion sites and also patient-related factors (1, 4-6).

The type of CVC used and composition of patients may be influenced only to some degree in our circumstances, however, the circumstances of catheter insertion and subsequent care of the inserton site may be largely improved by influencing the medical and nursing staff. Education of doctors and nurses and adherence to simple measures may result in a significant reduction of CR-BSI occurrence (1, 7-10). Large datasets of prevalence rate and interventions may be found in English literature, however, original Czech data on prevalence rate or interventions are scanty (11).

The objective of our study was to monitor the effect of educational and technical intervention on the prevalence rate of CR-BSI in a dedicated high dependency unit (HDU) in the Department of Infectious Diseases of Ceske Budejovice Hospital, Inc., and to benchmark data for future interventions to further reduce the prevalence rate of these serious complications of hospital care.

Methods

HDU of the Department of Infectious Diseases in Ceske Budejovice Hospital, Inc., provides level 2 intensive care in six non-ventilated beds for patients with a broad spectrum of infectious diseases, including perioperative care for patients with septic neurosurgical and orthopedic conditions. In 2008, after several septic complications of central venous catheter, we proactively started to influence the occurrence rate of catheter-related BSI's in short-term CVC's without antimicrobial coating.

First, we organized a workshop for doctors and all nurses in the department to discuss the topic of infectious complications of CVCs, where we addressed the continuum of biofilm development from colonization though to sepsis (12). During the same workshop, we reviewed the principles included in the Guideline for Management of Central Venous Catheters (Guideline no. 21, Ceske Budejovice Hospital), which was based on published guidelines (13-15).

We also made some adjustments in the procedure of insertion and subsequent care for the CVC. We started to require strict adherence in using large drapes (16), complete personal protective gear, triple disinfection of the skin of the shoulder, neck, and upper half of the chest with 70% alcohol, or iodpovidone, cleaning the insertion site after the procedure with hydrogen peroxide to remove blood residuals, and subsequent disinfection with iodpovidone. To affix the CVC to the skin, we developed own method of fixation with suture causing mild tension, which prevents a piston-like movement of the CVC in a mobile patient, along with a small bending of the CVC before the second place of suture (Figure).

The central lines were preferably inserted to the subclavian vein, only when not accessible, a jugular vein was used for insertion. We did not use femoral vein access (16).

To cover the CVC, we used sterile gauze pads, since in the past we did not have good experience with transparent covering – which required frequent changes in perspirating, pyrexial patients due to accumulated fluid, and it easily disattached in mobile patients (1,17).

We checked the catheter daily, and changed sterile gauze pads with a permeable adhesive tape, we also used in-line antibacterial filters and strictly disinfected the stop-cocks and connection site when changing infusion bags. Blood samples via the line were not drawn, which assured only one-way flow. Any handling of a CVC was done in gloves.

In case of local skin inflammation we considered removing the CVC, and we removed the CVC always when the patient was pyrexail over 38 degrees centrigrade with no apparent cause, or if any other clinical signs of potential blood stream infection developed. Any CVC had to be removed as soon as it was no more needed (10,18).

Along with educative and technical measures we prospectively monitored in an administratively undemanding way all CVCs inserted in our unit – a special ledger was kept with patient identification, day of insertion, date of removal and the reason for removal – the end of therapy or suspected infectious complication – pyrexial yes/no (Table 1). We did not evaluate CVCs inserted in other units in patients transferred to us, nor did we follow up on our CVCs in patients transferred to other departments. However, we monitored CVCs in patients transferred from our HDU to the wards of our department, as well as CVCs inserted in patients located in the wards as we wanted and were able to address the quality of CVC management there.

Suspected CR-BSI was clinically defined as fever over 38 degrees centrigrade in a patient with CVC and no other cause of fever, or a suspected infectious complication in afebrile patients based on deterioration of clinical condition or elevation of inflammatory markers. When this suspicion was voiced, CVC was removed after drawing blood culture via the catheter lumen, the catheter tip was sent for culture and another blood culture was drawn from a peripheral vein (5). Removed catheter tips were cultured using semiquantitative rolling technique (26).

After two years of keeping the register we retrospectively reviewed the patient notes and culture results to determine, which cases of suspected CR-BSI were not confirmed or excluded by another disease, which cases were catheter colonization without bacteriemia, since we considered those as possible BSI (positive catheter tip culture, negative peripheral blood culture), and which cases were confirmed CR-BSI – matching catheter tip and peripheral blood cultures (19). We also evaluated the cultured pathogens and sought potential risk factors for developing CR-BSI in the given patient.

Results

During 2 years after starting the register (from March 2008 to March 2010) we inserted total of 142 central venous catheters for total duration of 1423 days (median 9 days, range 1-28 days, 25^th percentil 4 days, 75^th percentil 15 days). During this time, 6 doctors participated in CVC insertion, there were 3 different nurse managers in HDU and total turnover 50% of nurses.

Total 10 catheters have been removed after median of 17 (11-28) days due to unexplained fever. In 4 cases CR-BSI was confirmed (after 13, 18, 19, and 21 days), in 2 other cases the CVC was colonized without bacteriemia in pyrexial patients with no other explanation for fever (after 11 and 28 days). (Graph)

Looking at microbiology results, with exception of Staphylococcus epidermidis in one case of CR-BSI and one CVC colonization, we detected several enteric bacteria (usually colonizing urinary or respiratory tract in the setting of intensive care) (table 2). In the remaining 4 cases with negative CVC culture and negative blood culture we had retrospectively diagnosed urinary tract infection in 2 cases, pneumonia in one case and a soft tissue abscess of the forearm in one case.

We have recorded total 2.81 proven cases of catheter related BSI per 1000 catheter days. Confirmed bloodstream infection or catheter colonization along with unexplained pyrexia, as confirmed and possible BSI, was noted in 4.22 cases per 1000 catheter days.

After excluding all catheters, in which we could not fully control subsequent care and usually not even follow their further course due to transfer to other units (total 30 CVCs, 157 catheter days, median 3 days, range 1-18 days), we came to a “net” cohort of CVCs fully under our control, total 112 CVCs inserted for total of 1266 days (median time of insertion 12 days, range 1-28 days, 25^th percentile 6 days, 75^th percentile 16 days). In this “net” cohort we noted 3.57 cases of CR-BSI per 1000 catheter days. After adding CVC colonisations with otherwise unexplained pyrexia we noted 4.73 possible CR-BSI per 1000 catheter days. All patients with confirmed or possible CR-BSI lived to discharge from the hospital.

Discussion

The register was kept in a level 2 HDU that provides care to non-ventilated patients with infectious diseases such as meningitis, encephalitis, severe hepatitis, respiratory tract infections, kidney and urinary tract infections, severe gastroenteritis and medical septic conditions, however, there is a significant proportion of our patients with septic bone and joint infections that are frequently treated in cooperation with orthopedic surgeon or neurosurgeon (vertebral osteomyelitis, brain abscess, spinal abscess, psoatic abscess, septic complications of hip and knee arthroplasty). Many patients are transferred to our HDU from level 3 ICU in the Department of Anesthesiology and Resuscitation, who may have colonized urinary and respiratory tract with multidrug resistant bacteria, which may be highly prone to CVC infection, especially if they have tracheostomy.

Microbiological results of our cohort are slightly different from large meta-analyses (1,3), as we found a higher proportion of enteric bacteria compared to skin organisms. Apart from a small size of our cohort, we suggest two factors for this difference – first, successful aseptic conditions during CVC insertion and further management, and second, a higher risk in agitated patients - three cases of CR-BSI caused by enteric bacteria were non-relaxed, spontaneously ventilating patients with qualitative loss of consciousness (GCS 13-14), whose hands were the most likely vector of transfer of bacteria from their perigenital and perineal area to the external surface of the CVC. The fourth case was an immunosuppressed patient with bulbar syndrome, tracheostomy and colonized respiratory tract with massive sputum production (Table 2).

The low number of CVC colonization and CR-BSI caused by skin organisms (staphylococci) may be due to underdiagnosing (coagulase-negative staphylococci in debilitated patients may not cause apparent clinical symptoms), but also by measures taken to clean blood residuals around the site of insertion with hydrogen peroxide, and also thanks to the above mentioned limited piston-like movement of CVC shaft by mild tension of sutures (Figure). In a small number of patients with CVC in wards outside the HDU we noted no episode of CR-CVC in contrast from the literature (20).

As to elective replacement of CVC, some authors do not consider a routine replacement to be effective to limit the rate of BSI (21), other authors suggest elective CVC replacement after 14 days after insertion (22), in our experience we perform a thorough evaluation of CVC after 20 days and if there is ongoing need of CVC and any doubt of its condition, we proceed to elective replacement.

Conclusion

Thanks to the ongoing CVC register we were able to quantify that the implemented technical measures and persisting awareness of doctors and nurses concerning the risk of CR-BSI result in reassuring outcomes even when compared with published international data (1-5) and result in an improved quality of care and higher patient safety.

Regardless the type of catheter used and composition of patients, good results may be achieved with minimum demand on labor and costs. However, it is crucial to set and maintain the workplace culture by ongoing educative activities, especially in the environment of naturally high turnover of the staff (9,10). The register of CR-CVC is a time and labor undemanding tool that may be useful in reaching this target. The data obtained from such register should be, however, used as an educational feedback, not as a repressive tool, since its purpose is to set supportive environment in the workplace, where proactive detection of problems enables their solution, and where prevention of complications is esteemed as desirable (23).

Another area of intervention, to which this analysis pointed, will be to limit the effect of manipulation with the vascular access device by high-risk patients – again by educative effort towards the doctors, nurses and other healthcare workers in the HDU (6,10,18,24,25). The mere identification of this risk factor and its awareness in all members of the healthcare team is the first step to further reduce the incidence of CR-BSIs (9,10).

Abbreviations

• CVC – central venous catheter,
• BSI - blood stream infection,
• CR-BSI – catheter realated blood stream infection,
• ICU - intensive care unit,
• HDU – high dependency unit

Corresponding author:

MUDr. Aleš Chrdle,

3 Saxon Road, Crosby, L23 3DJ,

Liverpool, UK

Email: chrdle@email.cz

Tel: +44 7851 657932