Journal of Hospital Infection (1990) 15, 95-102
Double-blind vancomycin catheter M. R. Ranson, Departments
placebo controlled study of prophylaxis for central venous insertion in cancer patients
B. A. Oppenheim *, A. Jackson, J. H. Scarffe
of Medical
A. G. Kamthan
Oncology and *Microbiology, Manchester M20 9BX, UK
Accepted for publication
31 August
Christie
and
Hospital,
1989
Summary: To assess whether vancomycin administration at the time of central venous catheter insertion would prevent catheter-related sepsis (CRS) in immunocompromised patients, 98 cancer patients were entered into a randomized placebo-controlled trial. Patients were stratified according to whether they were having therapy for acute leukaemia or undergoing bonemarrow transplantation (group A) or required chemotherapy for a solid tumour (group B). Eligible patients were randomized to receive either 500 mg vancomycin in 250 ml of 0.9% saline prior to catheter insertion followed by a further 500 mg infused via the established catheter, or the same regimen with 0.9% saline alone. In group A, there were 32 instances of CRS occurring in 20 of the 3.5 evaluable catheters(57%). Six catheters were removed because of CRS. No significant difference was found in the incidence of CRS between the two arms. Of the 37 evaluable catheters in group B, CRS occurred in 6 (16%), and none of the catheters required removal because of CRS. Again, no significant differences were found in the incidence of CRS between the patients given vancomycin or placebo. These findings indicate that the incidence of CRS is greater in patients who have more severe and prolonged immunosuppression and that vancomycin prophylaxis fails to reduce CRS in patients undergoing chemotherapy for malignant disease. Keywords:
Vancomycin
prophylaxis;
catheter-related
sepsis.
Introduction
In recent years there has been a steady increase in the use of central venous catheters in the management of patients with malignant disease. Whilst these have greatly facilitated and improved the management of such patients, catheter related sepsis (CRS) remains a significant problem. The major pathogens associated with central venous catheter use are the Gram-positive cocci, particularly coagulase-negative staphylococci (Press et al.‘; 1984). A number of centres, including our own have seen a rise in the Correspondence 01954701/90/010095+0s
to: Dr M. R. Ranson. so3.Mljo
0 1990 The Hospital
95
Infection
Society
96
M. R. Ranson
et al.
incidence of coagulase negative staphylococcal (CNS) bacteraemias attributable to the presence of an indwelling central venous catheter. The employment of specialist nursing staff to manage the catheters, and the provision of a subcutaneous tunnel, have previously been noted to reduce the incidence of CRS (Johnstone 1982, Garden & Sim, 1983). The question of whether prophylactic antibiotics at the time of catheter insertion are of benefit has yet to be resolved. In a randomised prospective trial in patients receiving parenteral nutrition, the administration of vancomycin prophylaxis was of no benefit in reducing CRS (McKee et al., 1985). In contrast, in a non-randomized study in patients undergoing bone marrow transplantation, the use of vancomycin resulted in an increase in lifespan of the catheter and a reduction in the number of catheters removed because of sepsis. The incidence of CRS in the first 30 days was also found to be reduced in vancomycin-treated patients but this was not statistically et al., 1987). significant (Mackinnon Randomized prospective trials of antibiotic prophylaxis are required to resolve these discrepancies. In addition, since the principal organisms associated with central venous catheters are of low pathogenicity, the immunocompetence of the patient group may have a significant bearing on the development of clinical CRS. The aim of the present study was to investigate in a randomized double-blind prospective study whether prophylactic intravenous vancomycin was effective in reducing the subsequent incidence of CRS in patients undergoing chemotherapy for malignant disease. Patients
and methods
Following informed verbal consent, 98 eligible patients were randomized to receive either vancomycin or placebo at the time of catheter insertion. Patients were excluded if they had a known hypersensitivity to vancomycin, or had evidence of renal impairment defined as a serum creatinine of > 0.2 mmol 1-l. Patients already receiving vancomycin were also excluded. Patients were randomized into two groups. Group A-Patients undergoing induction or consolidation therapy for acute leukaemia or undergoing bone marrow transplantation (BMT). Group B-All other patients requiring a central venous catheter for chemotherapy. The former group of patients experienced more severe and prolonged myelosuppression than the latter group. Group A required predominantly in-patient management in contrast to group B where out-patient management was the norm. Administration of either vancomycin or placebo was carried out as a double-blind procedure with randomisation and drug reconstitution being carried out by the hospital pharmacy. Patients randomized to receive vancomycin received 500 mg of vancomycin in 250 ml of 0.9% saline
Vancomycin
prophylaxis
for line insertion
97
infused over 20-30 min via a peripheral vein just prior to catheter insertion. A further 500 mg of vancomycin in 250 ml 0.9% saline was infused via the established central venous catheter over 20-30 min, once the satisfactory position of the catheter had been confirmed by chest X-ray. Control patients received exactly the same regimen with 0.9% saline infusions alone. Catheter insertion was carried out on the ward using a standard technique. Chlorhexidine in spirit was used as skin preparation. The insertion of the catheter (Neutricath S 35 cm, Vigon, France) was via the infraclavicular approach using a Seldinger technique. A 4-10 cm subcutaneous tunnel was created with the supplied introducer and the line held in place by silk sutures and a sterile occlusive transparent dressing (Tegaderm, 3M, USA). Standardized redressing of the line was carried out at weekly intervals for in-patients and at each out-patient attendance for out-patients. When the line was not in use a heparin lock (Hepsal, CP Pharmaceuticals, UK) was used to prevent catheter blockage. Outpatients had their central line heparinized at each attendance at the clinic. Blood cultures taken both peripherally and via the central line were obtained at the onset of pyrexial episodes before the institution of antibiotic therapy. Cultures were obtained subsequently, as indicated by the clinical status of the patient. Swabs from the catheter site were taken if clinically indicated and routine surveillance swabs were obtained weekly from in-patients undergoing therapy for acute leukaemia or BMT. Catheter-related sepsis was defined as follows. (a) Coagulase negative staphylococcal bacteraemia. Pyrexial episode (>38”C) during which CNS were isolated from blood cultures with or without evidence of localized skin tunnel infection and which resolved on appropriate antibiotic therapy or catheter removal. (b) Vancomycin responsive fever. Pyrexial episode (> 38°C) with negative blood cultures with or without evidence of localized skin tunnel infection in which pyrexia resolved on institution of vancomycin intravenously or catheter removal. (c) Skin tunnel/exit site sepsis without bacteraemia or fever requiring antibiotic therapy and/or catheter removal. Localized exit site infection not requiring antibiotic therapy and/or catheter removal was not included in the definition of CRS. Results
Group A (leukaemic + BMT patients) A total of 35 catheters were evaluable in 48 patients randomized. There was a failure to site the catheter in 9 patients, 4 patients were excluded because of protocol violation (3 patients already receiving IV vancomycin, and one patient had their catheter repaired rather than new catheter inserted). Patient details are shown in Table I.
98 Table
M. R. Ranson
et al.
I. Patient characteristics according to prophylactic insertion Treatment Vancomycin
A
Treatment
group
B
Control
Vancomycin
Control
17
18
19
:: 8 37 16-68
7 E 17-66
8 55 26-73
1; z-71
0.8 (O-8.8)
1 .O (o-21.8)
5.0 (2.8-14.5)
4.4 (1.4-8.7)
No of catheters Male Female Mean age (years) Age range (years) Median neutrophil count at insertion X IO9 1-l (Range) Antibiotic therapy* at time of insertion
group
treatmentgroups at the time of catheter
7
6
13
13
5
2
0
2
0
1
4
8
G
4 7
Diagnosis Acute myeloid leukaemia Acute lymphoblastic leukaemia Chronic granulocytic leukaemia Myeloma Small cell carcinoma of bronchus Other malignancy *antibiotics
other than vancomycin
In total, there were 32 instances of CRS and these occurred in 20 of the 35 evaluable catheters (57%). A total of 18 episodes of CNS bacteraemia were documented in 11 catheters. Tunnel sepsis occurred in 4 catheters. Six catheters were removed because of CRS, 2 because of Staphylococcus aureus tunnel infection and 4 because of repeated episodes of CNS bacteraemia. Coagulase-negative staphylococci were by far the commonest isolate, accounting for 18 of the 25 (72%) positive blood cultures in the group. Both the patient characteristics and the duration of neutropenia, during the first 30 d following catheter insertion, were similar in the vancomycin-treated and control groups. No significant difference was found in the incidence of CRS between the two arms and this remained the case whether one considered CRS in the first 30 d or during the total duration of catheter placement (Table II). Group B (Non-leukaemic patients) A total of 37 catheters were evaluable in 50 patients randomized. Initial attempts at catheter placement failed in 10 patients and 3 patients were excluded from analysis (1 died before catheter placement, 1 patient opted 1 patient was randomized twice for for no therapy after randomization, same catheter). Patient details are shown in Table I.
Vancomycin Table
II. Details
of catheter
prophylaxis related
for line insertion
sepsis according Group
Vancomycin No of catheters No of catheters
Mean duration < 1 .O X lo9 l-’
1-231 81.3 14.4
neutropenia* in first 30 days
CNS bacteraemia in first 30 days
episodes
Vancomycin responsive total in first 30 days Tunnel sepsis episodes in first 30 days Total episodes sepsis
142
days total range mean
99
to prophylactic
treatment
A
Group
groups B
Control
Vancomycin
Control
17 1459 9-189 85.8
18 2278 88462 126
19 2042 21-329 107
18.0
8 3
10 3
2 0
1 0
4 4
6
1
4
1
2 1
3 1
1 1
0 0
0 0
15
17
3
3
total
fever
total
of catheter-related
Catheter-related sepsis in first 30 days per 100 days
8 1.02
8 1.16
A.13
IL5
No of lines removed
3
3
0
0
*Incidence
recorded
for CRS during
first
30 d following
catheter
insertion.
Catheter-related sepsis affected 6 of the 37 catheters (16%). No multiple episodes of CRS occurred. Coagulase-negative staphylococcal bacteraemia occurred in 3 catheters (8.1%). None of the catheters required removal because of CRS. Coagulase-negative staphylococci were the most frequent pathogens isolated from blood cultures, accounting for 3 out of the 5 isolates. No significant differences were found in the incidence of CRS between control patients and patients given vancomycin prophylaxis. This applied whether one considered CRS -in the first 30 d following catheter insertion or CRS during the total period of catheter placement (Table II). Discussion
The present study illustrates that CRS is an important complication in patients with malignancy who have central venous catheters. CRS occurred in 57% of the catheters in leukaemic patients and 16% of catheters in patients treated for a range of other malignancies. The incidence of CRS in
100
M. R. Ranson
et al.
non-leukaemic patients in the. present study (0.14 per 100 catheter days) is identical to that reported by Press et al. (1984) in a review of over 1000 Hickman catheters in patients with neoplastic disease. In comparing the incidence of CRS between centres, it is important to consider the severity of immunosuppression to which the patient is subjected since both the present study and the work of Hartman & Shochat (1987) illustrates that neutropenia is one of the major risk factors for CRS. Coagulase-negative staphylococci are important pathogens in patients with malignant disease and they produce significant morbidity (Wade et al., 1982; Bodey, 1986). In the present study CNS accounted for 18 out of 25 bacteraemias (72%) in the leukaemic group and 3 out of 5 in the non-leukaemic group. In most instances both localized and systemic infection due to these organisms was successfully treated with antibiotic therapy. Only in instances of recurrent bacteraemia or tunnel infection with S. aweus was catheter removal necessary, emphasizing again that catheter removal is not required in the majority of patients with catheter-related sepsis (Hartman & Shochat, 1987). Skin preparation with antiseptics does not remove all skin bacteria and it has been reported that significant catheter-tip contamination occurs in almost a quarter of patients at the time of insertion (McKee et al., 1985). Investigators have therefore sought to assess whether the administration of antibiotics at the time of catheter insertion can reduce the incidence of CRS. The activity of vancomycin against skin commensals makes this agent an obvious choice. Vancomycin prophylaxis failed to reduce the incidence of CRS in either of the two groups of patients in this study. The explanation probably lies in the mechanisms by which CRS occurs. Investigators who have examined phage types of CNS have documented that there is little correlation between phage types isolated from skin around the catheter site and those recovered from the blood or catheter tip (McKee et al., 1985). However, there does appear to be an association between isolates from the catheter hub and those from catheter segments or blood (Sitges-Serra et al., 1984; Weightman et al., 1988). It would appear therefore that bacteria are more likely to be introduced during and following hub manipulation than via spread from the skin exit-site or from tunnel infection. Certainly, many episodes of CRS arise in the absence of clinically evident skin exit-site infection. Our results conflict with those of Mackinnon et al. (1987) who administered prophylactic vancomycin to patients undergoing BMT, and reported that this significantly increased catheter life and resulted in fewer catheters being removed due to infection. CRS was however not significantly improved by vancomycin. The study by Mackinnon et al. was neither blind nor randomized and it is possible that differences in catheter management occurred in patients who were given vancomycin prophylaxis. The very act of entering a patient into a clinical study may affect the incidence of CRS. Wagman,, Kirkemo & Johnston (1984) showed that
Vancomycin
prophylaxis
for line
insertion
101
patients who entered a randomized prospective trial had a significantly lower incidence of septic complications than a similar group of patients who were not randomized into the study. It is now evident that dramatic improvements in the incidence of CRS can result when catheter management is undertaken by both trained and interested staff (Keohane et al., 1983; Weightman et al., 1988). We have used only non-cuffed silastic catheters because of simplicity in insertion and generally good performance. Although studies have been few, infection rates with these catheters appear to be similar to those found with Hickman catheters (Decker & Edwards, 1988). Our policy of only ‘heparinizing’ lines after blood sampling at each out-patient attendance may have a positive effect in minimizing the incidence of CRS. In conclusion, this randomized double-blind prospective trial confirms that the incidence of CRS is higher in patients who have more severe and prolonged immunosuppression. Vancomycin prophylaxis fails to reduce the incidence of CRS in patients undergoing chemotherapy for a range of malignancies. Th e weight of present evidence suggests that meticulous adherence to catheter care by trained staff may be the single most important factor in minimizing the incidence and morbidity arising from CRS. Financial support for microbiological procedures was received from Eli Lilly and Co. We wish to thank the specialist nursing staff who undertook the catheter management, Professor D. Crowther, Dr N. Thatcher and Dr A. Howell for permission to enter patients into the study and Mrs L. Chamberlain for preparing the manuscript.
References G. P. (1986). Infection in cancer patients. A continuing association. The American Journal of Medicine 81, 1 l-26. Decker, M. D. & Edwards, K. M. (1988). Central venous catheter infections. Pediatric Clinics of North America 35, 579-612. Garden, 0. J. & Sim, A. J. W. (1983). A comparison of tunnelled and non-tunnelled subclavian vein catheters-a prospective study of complications during parenteral feeding. Clim’cal Nutrition 2, 51-54. Hartman, G. E. & Shochat, S. J. (1987). Management of septic complications associated with silastic catheters in childhood malignancy. Pediatric Infectious Disease 6, 1042-l 047. Johnstone, J. D. (1982). Infrequent infections associated with Hickman catheters. Cancer Bodey,
Nursing 5, 125-l 29. Keohane, P. P., Jones, B. M. J., Attrill, H., Cribb, A., Northover, T., Frost, P. & Silk, D. B. A. (1983). Effect of catheter tunnelling and a nutrition nurse on catheter sepsis during parenteral nutrition. Lancet ii, 1388-l 390. Mackinnon, S., Garden, 0. J., Gribben, J. G., Burns, H. J. G., Baird, D. & Burnett, A. K. (1987). The use of Hickman dwelling catheters in marrow transplant patients. Intensiwe Therapy and Clinical Monitoring 8, 122-126. McKee, R., Dunsmuir, R., Whithy, M. & Garden, 0. J. (1985). Does antibiotic prophylaxis at time of catheter insertion reduce the incidence of catheter related sepsis in intravenous nutrition. Journal of Hospital Infection 6, 419-425. Press, 0. W., Ramsey, P. G., Larson, E. B., Fefer, A. & Hickman, R. 0. (1984). Hickman catheter infections in patients with malignancies. Medicine 63, 189-200. Sitges-Serra, A., Puig, P., Linares, J., Perez, J. L., Farrero, N., Jaurieta, E. & Garau, J. (1984). Hub colonisation as the critical step in an outbreak of catheter related sepsis due
M. R. Ranson
102 to coagulase
negative
staphylococci
and Enteral Nutrition
8, 668-672.
during
et al.
parenteral
nutrition.
Journal of Parenteral
Wade, J. C., Stephen, C., Shimpff, M. D., Kathryn, A., Newman, R. N. & Wiernik, P. H. cause of infection in patients with (1982). Staphylococcus epidermidis: An increasing granulocytopenia. Annals of Internal Medicine 97, 503-508. Wagman, L. D., Kirkemo, A. & Johnston, M. R. (1984). Venous access: a prospective, randomised study of the Hickman catheter. Surgery 95, 303-308. Weightman, N. C., Simpson, E. M., Speller, D. C., Mott, M. G. & Oakhill, A. (1988). Bacteraemia related to indwelling central venous catheters: prevention, diagnosis and treatment. European Journal of Clinical Microbiology and Infectious Disease 7, 125-l 29.
Double-blind vancomycin catheter M. R. Ranson, Departments
placebo controlled study of prophylaxis for central venous insertion in cancer patients
B. A. Oppenheim *, A. Jackson, J. H. Scarffe
of Medical
A. G. Kamthan
Oncology and *Microbiology, Manchester M20 9BX, UK
Accepted for publication
31 August
Christie
and
Hospital,
1989
Summary: To assess whether vancomycin administration at the time of central venous catheter insertion would prevent catheter-related sepsis (CRS) in immunocompromised patients, 98 cancer patients were entered into a randomized placebo-controlled trial. Patients were stratified according to whether they were having therapy for acute leukaemia or undergoing bonemarrow transplantation (group A) or required chemotherapy for a solid tumour (group B). Eligible patients were randomized to receive either 500 mg vancomycin in 250 ml of 0.9% saline prior to catheter insertion followed by a further 500 mg infused via the established catheter, or the same regimen with 0.9% saline alone. In group A, there were 32 instances of CRS occurring in 20 of the 3.5 evaluable catheters(57%). Six catheters were removed because of CRS. No significant difference was found in the incidence of CRS between the two arms. Of the 37 evaluable catheters in group B, CRS occurred in 6 (16%), and none of the catheters required removal because of CRS. Again, no significant differences were found in the incidence of CRS between the patients given vancomycin or placebo. These findings indicate that the incidence of CRS is greater in patients who have more severe and prolonged immunosuppression and that vancomycin prophylaxis fails to reduce CRS in patients undergoing chemotherapy for malignant disease. Keywords:
Vancomycin
prophylaxis;
catheter-related
sepsis.
Introduction
In recent years there has been a steady increase in the use of central venous catheters in the management of patients with malignant disease. Whilst these have greatly facilitated and improved the management of such patients, catheter related sepsis (CRS) remains a significant problem. The major pathogens associated with central venous catheter use are the Gram-positive cocci, particularly coagulase-negative staphylococci (Press et al.‘; 1984). A number of centres, including our own have seen a rise in the Correspondence 01954701/90/010095+0s
to: Dr M. R. Ranson. so3.Mljo
0 1990 The Hospital
95
Infection
Society
96
M. R. Ranson
et al.
incidence of coagulase negative staphylococcal (CNS) bacteraemias attributable to the presence of an indwelling central venous catheter. The employment of specialist nursing staff to manage the catheters, and the provision of a subcutaneous tunnel, have previously been noted to reduce the incidence of CRS (Johnstone 1982, Garden & Sim, 1983). The question of whether prophylactic antibiotics at the time of catheter insertion are of benefit has yet to be resolved. In a randomised prospective trial in patients receiving parenteral nutrition, the administration of vancomycin prophylaxis was of no benefit in reducing CRS (McKee et al., 1985). In contrast, in a non-randomized study in patients undergoing bone marrow transplantation, the use of vancomycin resulted in an increase in lifespan of the catheter and a reduction in the number of catheters removed because of sepsis. The incidence of CRS in the first 30 days was also found to be reduced in vancomycin-treated patients but this was not statistically et al., 1987). significant (Mackinnon Randomized prospective trials of antibiotic prophylaxis are required to resolve these discrepancies. In addition, since the principal organisms associated with central venous catheters are of low pathogenicity, the immunocompetence of the patient group may have a significant bearing on the development of clinical CRS. The aim of the present study was to investigate in a randomized double-blind prospective study whether prophylactic intravenous vancomycin was effective in reducing the subsequent incidence of CRS in patients undergoing chemotherapy for malignant disease. Patients
and methods
Following informed verbal consent, 98 eligible patients were randomized to receive either vancomycin or placebo at the time of catheter insertion. Patients were excluded if they had a known hypersensitivity to vancomycin, or had evidence of renal impairment defined as a serum creatinine of > 0.2 mmol 1-l. Patients already receiving vancomycin were also excluded. Patients were randomized into two groups. Group A-Patients undergoing induction or consolidation therapy for acute leukaemia or undergoing bone marrow transplantation (BMT). Group B-All other patients requiring a central venous catheter for chemotherapy. The former group of patients experienced more severe and prolonged myelosuppression than the latter group. Group A required predominantly in-patient management in contrast to group B where out-patient management was the norm. Administration of either vancomycin or placebo was carried out as a double-blind procedure with randomisation and drug reconstitution being carried out by the hospital pharmacy. Patients randomized to receive vancomycin received 500 mg of vancomycin in 250 ml of 0.9% saline
Vancomycin
prophylaxis
for line insertion
97
infused over 20-30 min via a peripheral vein just prior to catheter insertion. A further 500 mg of vancomycin in 250 ml 0.9% saline was infused via the established central venous catheter over 20-30 min, once the satisfactory position of the catheter had been confirmed by chest X-ray. Control patients received exactly the same regimen with 0.9% saline infusions alone. Catheter insertion was carried out on the ward using a standard technique. Chlorhexidine in spirit was used as skin preparation. The insertion of the catheter (Neutricath S 35 cm, Vigon, France) was via the infraclavicular approach using a Seldinger technique. A 4-10 cm subcutaneous tunnel was created with the supplied introducer and the line held in place by silk sutures and a sterile occlusive transparent dressing (Tegaderm, 3M, USA). Standardized redressing of the line was carried out at weekly intervals for in-patients and at each out-patient attendance for out-patients. When the line was not in use a heparin lock (Hepsal, CP Pharmaceuticals, UK) was used to prevent catheter blockage. Outpatients had their central line heparinized at each attendance at the clinic. Blood cultures taken both peripherally and via the central line were obtained at the onset of pyrexial episodes before the institution of antibiotic therapy. Cultures were obtained subsequently, as indicated by the clinical status of the patient. Swabs from the catheter site were taken if clinically indicated and routine surveillance swabs were obtained weekly from in-patients undergoing therapy for acute leukaemia or BMT. Catheter-related sepsis was defined as follows. (a) Coagulase negative staphylococcal bacteraemia. Pyrexial episode (>38”C) during which CNS were isolated from blood cultures with or without evidence of localized skin tunnel infection and which resolved on appropriate antibiotic therapy or catheter removal. (b) Vancomycin responsive fever. Pyrexial episode (> 38°C) with negative blood cultures with or without evidence of localized skin tunnel infection in which pyrexia resolved on institution of vancomycin intravenously or catheter removal. (c) Skin tunnel/exit site sepsis without bacteraemia or fever requiring antibiotic therapy and/or catheter removal. Localized exit site infection not requiring antibiotic therapy and/or catheter removal was not included in the definition of CRS. Results
Group A (leukaemic + BMT patients) A total of 35 catheters were evaluable in 48 patients randomized. There was a failure to site the catheter in 9 patients, 4 patients were excluded because of protocol violation (3 patients already receiving IV vancomycin, and one patient had their catheter repaired rather than new catheter inserted). Patient details are shown in Table I.
98 Table
M. R. Ranson
et al.
I. Patient characteristics according to prophylactic insertion Treatment Vancomycin
A
Treatment
group
B
Control
Vancomycin
Control
17
18
19
:: 8 37 16-68
7 E 17-66
8 55 26-73
1; z-71
0.8 (O-8.8)
1 .O (o-21.8)
5.0 (2.8-14.5)
4.4 (1.4-8.7)
No of catheters Male Female Mean age (years) Age range (years) Median neutrophil count at insertion X IO9 1-l (Range) Antibiotic therapy* at time of insertion
group
treatmentgroups at the time of catheter
7
6
13
13
5
2
0
2
0
1
4
8
G
4 7
Diagnosis Acute myeloid leukaemia Acute lymphoblastic leukaemia Chronic granulocytic leukaemia Myeloma Small cell carcinoma of bronchus Other malignancy *antibiotics
other than vancomycin
In total, there were 32 instances of CRS and these occurred in 20 of the 35 evaluable catheters (57%). A total of 18 episodes of CNS bacteraemia were documented in 11 catheters. Tunnel sepsis occurred in 4 catheters. Six catheters were removed because of CRS, 2 because of Staphylococcus aureus tunnel infection and 4 because of repeated episodes of CNS bacteraemia. Coagulase-negative staphylococci were by far the commonest isolate, accounting for 18 of the 25 (72%) positive blood cultures in the group. Both the patient characteristics and the duration of neutropenia, during the first 30 d following catheter insertion, were similar in the vancomycin-treated and control groups. No significant difference was found in the incidence of CRS between the two arms and this remained the case whether one considered CRS in the first 30 d or during the total duration of catheter placement (Table II). Group B (Non-leukaemic patients) A total of 37 catheters were evaluable in 50 patients randomized. Initial attempts at catheter placement failed in 10 patients and 3 patients were excluded from analysis (1 died before catheter placement, 1 patient opted 1 patient was randomized twice for for no therapy after randomization, same catheter). Patient details are shown in Table I.
Vancomycin Table
II. Details
of catheter
prophylaxis related
for line insertion
sepsis according Group
Vancomycin No of catheters No of catheters
Mean duration < 1 .O X lo9 l-’
1-231 81.3 14.4
neutropenia* in first 30 days
CNS bacteraemia in first 30 days
episodes
Vancomycin responsive total in first 30 days Tunnel sepsis episodes in first 30 days Total episodes sepsis
142
days total range mean
99
to prophylactic
treatment
A
Group
groups B
Control
Vancomycin
Control
17 1459 9-189 85.8
18 2278 88462 126
19 2042 21-329 107
18.0
8 3
10 3
2 0
1 0
4 4
6
1
4
1
2 1
3 1
1 1
0 0
0 0
15
17
3
3
total
fever
total
of catheter-related
Catheter-related sepsis in first 30 days per 100 days
8 1.02
8 1.16
A.13
IL5
No of lines removed
3
3
0
0
*Incidence
recorded
for CRS during
first
30 d following
catheter
insertion.
Catheter-related sepsis affected 6 of the 37 catheters (16%). No multiple episodes of CRS occurred. Coagulase-negative staphylococcal bacteraemia occurred in 3 catheters (8.1%). None of the catheters required removal because of CRS. Coagulase-negative staphylococci were the most frequent pathogens isolated from blood cultures, accounting for 3 out of the 5 isolates. No significant differences were found in the incidence of CRS between control patients and patients given vancomycin prophylaxis. This applied whether one considered CRS -in the first 30 d following catheter insertion or CRS during the total period of catheter placement (Table II). Discussion
The present study illustrates that CRS is an important complication in patients with malignancy who have central venous catheters. CRS occurred in 57% of the catheters in leukaemic patients and 16% of catheters in patients treated for a range of other malignancies. The incidence of CRS in
100
M. R. Ranson
et al.
non-leukaemic patients in the. present study (0.14 per 100 catheter days) is identical to that reported by Press et al. (1984) in a review of over 1000 Hickman catheters in patients with neoplastic disease. In comparing the incidence of CRS between centres, it is important to consider the severity of immunosuppression to which the patient is subjected since both the present study and the work of Hartman & Shochat (1987) illustrates that neutropenia is one of the major risk factors for CRS. Coagulase-negative staphylococci are important pathogens in patients with malignant disease and they produce significant morbidity (Wade et al., 1982; Bodey, 1986). In the present study CNS accounted for 18 out of 25 bacteraemias (72%) in the leukaemic group and 3 out of 5 in the non-leukaemic group. In most instances both localized and systemic infection due to these organisms was successfully treated with antibiotic therapy. Only in instances of recurrent bacteraemia or tunnel infection with S. aweus was catheter removal necessary, emphasizing again that catheter removal is not required in the majority of patients with catheter-related sepsis (Hartman & Shochat, 1987). Skin preparation with antiseptics does not remove all skin bacteria and it has been reported that significant catheter-tip contamination occurs in almost a quarter of patients at the time of insertion (McKee et al., 1985). Investigators have therefore sought to assess whether the administration of antibiotics at the time of catheter insertion can reduce the incidence of CRS. The activity of vancomycin against skin commensals makes this agent an obvious choice. Vancomycin prophylaxis failed to reduce the incidence of CRS in either of the two groups of patients in this study. The explanation probably lies in the mechanisms by which CRS occurs. Investigators who have examined phage types of CNS have documented that there is little correlation between phage types isolated from skin around the catheter site and those recovered from the blood or catheter tip (McKee et al., 1985). However, there does appear to be an association between isolates from the catheter hub and those from catheter segments or blood (Sitges-Serra et al., 1984; Weightman et al., 1988). It would appear therefore that bacteria are more likely to be introduced during and following hub manipulation than via spread from the skin exit-site or from tunnel infection. Certainly, many episodes of CRS arise in the absence of clinically evident skin exit-site infection. Our results conflict with those of Mackinnon et al. (1987) who administered prophylactic vancomycin to patients undergoing BMT, and reported that this significantly increased catheter life and resulted in fewer catheters being removed due to infection. CRS was however not significantly improved by vancomycin. The study by Mackinnon et al. was neither blind nor randomized and it is possible that differences in catheter management occurred in patients who were given vancomycin prophylaxis. The very act of entering a patient into a clinical study may affect the incidence of CRS. Wagman,, Kirkemo & Johnston (1984) showed that
Vancomycin
prophylaxis
for line
insertion
101
patients who entered a randomized prospective trial had a significantly lower incidence of septic complications than a similar group of patients who were not randomized into the study. It is now evident that dramatic improvements in the incidence of CRS can result when catheter management is undertaken by both trained and interested staff (Keohane et al., 1983; Weightman et al., 1988). We have used only non-cuffed silastic catheters because of simplicity in insertion and generally good performance. Although studies have been few, infection rates with these catheters appear to be similar to those found with Hickman catheters (Decker & Edwards, 1988). Our policy of only ‘heparinizing’ lines after blood sampling at each out-patient attendance may have a positive effect in minimizing the incidence of CRS. In conclusion, this randomized double-blind prospective trial confirms that the incidence of CRS is higher in patients who have more severe and prolonged immunosuppression. Vancomycin prophylaxis fails to reduce the incidence of CRS in patients undergoing chemotherapy for a range of malignancies. Th e weight of present evidence suggests that meticulous adherence to catheter care by trained staff may be the single most important factor in minimizing the incidence and morbidity arising from CRS. Financial support for microbiological procedures was received from Eli Lilly and Co. We wish to thank the specialist nursing staff who undertook the catheter management, Professor D. Crowther, Dr N. Thatcher and Dr A. Howell for permission to enter patients into the study and Mrs L. Chamberlain for preparing the manuscript.
References G. P. (1986). Infection in cancer patients. A continuing association. The American Journal of Medicine 81, 1 l-26. Decker, M. D. & Edwards, K. M. (1988). Central venous catheter infections. Pediatric Clinics of North America 35, 579-612. Garden, 0. J. & Sim, A. J. W. (1983). A comparison of tunnelled and non-tunnelled subclavian vein catheters-a prospective study of complications during parenteral feeding. Clim’cal Nutrition 2, 51-54. Hartman, G. E. & Shochat, S. J. (1987). Management of septic complications associated with silastic catheters in childhood malignancy. Pediatric Infectious Disease 6, 1042-l 047. Johnstone, J. D. (1982). Infrequent infections associated with Hickman catheters. Cancer Bodey,
Nursing 5, 125-l 29. Keohane, P. P., Jones, B. M. J., Attrill, H., Cribb, A., Northover, T., Frost, P. & Silk, D. B. A. (1983). Effect of catheter tunnelling and a nutrition nurse on catheter sepsis during parenteral nutrition. Lancet ii, 1388-l 390. Mackinnon, S., Garden, 0. J., Gribben, J. G., Burns, H. J. G., Baird, D. & Burnett, A. K. (1987). The use of Hickman dwelling catheters in marrow transplant patients. Intensiwe Therapy and Clinical Monitoring 8, 122-126. McKee, R., Dunsmuir, R., Whithy, M. & Garden, 0. J. (1985). Does antibiotic prophylaxis at time of catheter insertion reduce the incidence of catheter related sepsis in intravenous nutrition. Journal of Hospital Infection 6, 419-425. Press, 0. W., Ramsey, P. G., Larson, E. B., Fefer, A. & Hickman, R. 0. (1984). Hickman catheter infections in patients with malignancies. Medicine 63, 189-200. Sitges-Serra, A., Puig, P., Linares, J., Perez, J. L., Farrero, N., Jaurieta, E. & Garau, J. (1984). Hub colonisation as the critical step in an outbreak of catheter related sepsis due
M. R. Ranson
102 to coagulase
negative
staphylococci
and Enteral Nutrition
8, 668-672.
during
et al.
parenteral
nutrition.
Journal of Parenteral
Wade, J. C., Stephen, C., Shimpff, M. D., Kathryn, A., Newman, R. N. & Wiernik, P. H. cause of infection in patients with (1982). Staphylococcus epidermidis: An increasing granulocytopenia. Annals of Internal Medicine 97, 503-508. Wagman, L. D., Kirkemo, A. & Johnston, M. R. (1984). Venous access: a prospective, randomised study of the Hickman catheter. Surgery 95, 303-308. Weightman, N. C., Simpson, E. M., Speller, D. C., Mott, M. G. & Oakhill, A. (1988). Bacteraemia related to indwelling central venous catheters: prevention, diagnosis and treatment. European Journal of Clinical Microbiology and Infectious Disease 7, 125-l 29.
