Abstract
Semiquantitative cultures of skin surrounding intravascular catheter entry sites and catheter hubs have high negative predictive values for catheter tip colonization. However, culturing samples from the inner side of the hub requires the catheter to be manipulated, thus increasing the risk of migration of microorganisms into the bloodstream. Today, hubs are closed using needleless connectors (NCs). Cultures of NCs could predict catheter colonization. Our objective was to compare the yield of NC sonicate cultures for prediction of catheter colonization with that of hub cultures. For 6 months, we prospectively collected all short-term central lines and systems removed from patients admitted to the cardiac surgery postoperative care unit, irrespective of the reason for withdrawal. Hub cultures were obtained immediately before withdrawal and were cultured using a semiquantitative method. Catheter tips were cultured using the roll-plate technique and sonication, and NCs were cultured using a semiquantitative technique after sonication. We considered NCs to be colonized when ≥1 culture was positive. We collected a total of 75 central systems. The catheter colonization rate was 10.7%. The rates for hub and NC colonization were 6.7% and 12.0%, respectively. The validity values for hubs and NCs for prediction of catheter colonization were as follows: sensitivity, 25.0% and 87.5%; specificity, 95.5% and 97.0%; positive predictive value, 40.0% and 77.8%; negative predictive value, 91.4% and 98.5%; validity index, 88.0% and 96.0%, respectively. Cultures of closed NCs can be used to rule out catheter tip colonization and are superior to hub cultures in ruling out short-term central venous catheter colonization.
INTRODUCTION
Catheter-related bloodstream infection (C-RBSI) is one of the most important nosocomial infections. It has high morbidity and mortality rates and occurs only in patients with colonized catheters (1, 2). The negative predictive value of superficial culture (skin and hub) results for prediction of catheter tip colonization is very high (3–7). However, taking cultures from the hub requires a swab to be rubbed inside the catheter lumen, thus potentially leading to migration of microorganisms into the bloodstream (8–10).
Currently, short-term central venous catheter (CVC) hubs are closed with needleless connectors (NCs), which are easier to culture. We compared the negative predictive values for catheter tip colonization of cultures taken from inside the hubs and from the NCs.
MATERIALS AND METHODS
Setting.
This prospective study was carried out in the laboratory of the clinical microbiology and infectious diseases department and the cardiac surgery postoperative care unit (PCU) of our institution.
Laboratory procedures.
Immediately before removing the CVC, we took culture samples from the inside of all catheter hubs and from the NC (Clave connector). After the CVC was removed, the catheter tip was sent to the microbiology laboratory for culture. Always before culture samples were taken, the investigator wore clean gloves and disinfected the hubs with alcoholic chlorhexidine.
The laboratory procedures were as follows. The catheter tip was cultured using the roll-plate technique and sonication on a blood agar plate, samples from all hubs were cultured on blood agar plates using a semiquantitative method, and NCs were sonicated for 1 min, vortex-mixed, and centrifuged in brain heart infusion (BHI) medium. One hundred microliters of sonicate was cultured on a blood agar plate (Fig. 1).
FIG 1.
Laboratory procedures.
The microorganisms recovered were fully identified using standard microbiological methods (11). We also used a preestablished protocol to record patient characteristics, underlying diseases, comorbid conditions, severity-of-illness scores (such as acute physiology and chronic health evaluation II [APACHE II] scores), the maximum severity reached before catheter withdrawal, and microbiological data on blood cultures.
Definitions.
The following definitions were used: catheter tip colonization, isolation of either ≥15 CFU/plate with the semiquantitative Maki technique or ≥100 CFU/segment with the sonication method; hub colonization, isolation of ≥15 CFU/plate in semiquantitative culture; closed NC colonization, isolation of ≥20 CFU/connector from at least 1 NC from the central venous system in semiquantitative culture. We considered an episode of C-RBSI to be confirmed when the same microorganism was isolated both from peripheral blood cultures and from the catheter tip. To calculate the validity values for NCs for predicting catheter colonization, we considered positive NCs to be those with the same microorganism(s) as isolated from the catheter tip.
Statistical analyses.
Continuous variables were expressed as mean ± standard deviation (SD) or median and interquartile range (IQR); categorical variables were expressed as percentages, with 95% confidence intervals (CIs), when applicable. Categorical variables were evaluated using the chi-square test or 2-tailed Fisher exact test. Statistical significance was calculated using a 2-tailed test, and P values of <0.05 were considered significant.
We calculated the validity values for closed NC culture by comparing this method with the gold standard for colonization and also with culture of superficial samples. The sensitivity, specificity, and positive and negative predictive values, with their 95% CIs, were calculated using EPIDAT 3.1. Accuracy was defined as the sum of true-positive and true-negative results. Statistical analyses were performed using IBM SPSS Statistics for Windows, version 21.0 (IBM Corp., Armonk, NY).
Ethics.
The study was approved by the local ethics committee.
RESULTS
We included a total of 75 CVCs from 49 patients, with ages (mean ± SD) of 64.47 ± 14.2 years. The most common underlying conditions were congestive heart failure (36.7%) and diabetes mellitus (36.7%), followed by myocardial infarction (20.4%). The overall comorbidity index values (mean ± SD) and APACHE II scores at admission (mean ± SD) were 2.18 ± 1.7 and 8.67 ± 3.4, respectively. The main reason for catheter withdrawal was end of use (66.7%), followed by suspicion of infection (16.0%). Other patient and catheter data are detailed in Table 1.
TABLE 1.
Main patient and catheter characteristics
| Characteristica | Result |
|---|---|
| Patients (n = 49) | |
| Age (mean ± SD) (yr) | 64.47 ± 14.2 |
| Male/female (no./no.) | 28/21 |
| Underlying condition (no. [%]) | |
| Myocardial infarction | 10 (20.4) |
| Congestive heart failure | 18 (36.7) |
| Central nervous system (ACVA) | 2 (4.1) |
| Chronic obstructive pulmonary disease | 11 (22.4) |
| Diabetes mellitus | 18 (36.7) |
| Peptic ulcer disease | 7 (14.3) |
| Peripheral vascular disease | 5 (10.2) |
| Renal dysfunction | 8 (16.3) |
| EuroSCORE (mean ± SD) | 6.68 ± 3.1 |
| Comorbidity index (Charlson criteria) (mean ± SD) | 2.18 ± 1.7 |
| Nonfatal underlying disease (no. [%]) | 45 (91.8) |
| APACHE II score at admission (mean ± SD) | 8.67 ± 3.4 |
| Length of ICU stay (median [IQR]) (days) | 8.5 (5.2–13.0) |
| C-RBSI episodes (no. [%]) | 3 (6.1) |
| Crude mortality rate (no. [%]) | 8 (16.3) |
| Attributable mortality rate (no. [%]) | 0 (0.0) |
| Central venous systems (n = 75) | |
| Type of catheter (no. [%]) | |
| Nontunneled central venous catheter | 57 (60.0) |
| Swan-Ganz catheter | 1 (1.3) |
| Arterial catheter | 2 (2.7) |
| Sheldon catheter | 4 (5.3) |
| Introducer sheath | 11 (14.7) |
| Location (no. [%]) | |
| Jugular | 60 (80.0) |
| Subclavian | 8 (10.7) |
| Femoral | 7 (9.3) |
| Total parenteral nutrition/propofol (no. [%]) | 13 (17.3) |
| Reason for catheter withdrawal (no. [%]) | |
| End of use | 50 (66.7) |
| Suspicion of infection | 12 (16.0) |
| Other | 13 (17.3) |
| Catheter colonization (no. [%]) | 8 (10.7) |
| Catheter colonization density (cases per 1,000 catheter-days) | 9.91 |
ACVA, acute cerebrovascular accident; SD, standard deviation; IQR, interquartile range; ICU, intensive care unit; C-RBSI, catheter-related bloodstream infection; EuroSCORE, European System for Cardiac Operative Risk Evaluation.
Overall, we confirmed 3 episodes of C-RBSI (3.71 episodes/1,000 catheter days), and 16 patients (32.7%) had infections at other sites (Table 1). The overall catheter tip colonization rate was 10.7% (8/75 systems). The colonization rates for hubs and NCs were 6.7% and 12.0%, respectively.
Table 2 shows the validity values for hub and NC cultures for predicting catheter colonization. NC cultures showed a sensitivity value of 87.5% and a negative predictive value for catheter colonization of 98.5%, compared with 25.0% and 91.4% (P < 0.001), respectively, for hub cultures. Table 3 shows the different microorganisms isolated from the colonized central venous systems.
TABLE 2.
Validity values for cultures of hubs and closed connectors from central venous systems to predict catheter colonization
| Culture type | Sensitivity (95% CI)a (%) | Specificity (95% CI) (%) | PPV (95% CI) (%) | NPV (95% CI) (%) | Validity index (95% CI) (%) | Prevalence (95% CI) (%) | Positive LR (95% CI) | Negative LR (95% CI) |
|---|---|---|---|---|---|---|---|---|
| Hub | 25.0 (0.0–61.3) | 95.5 (89.8–100) | 40.0 (0.0–92.9) | 91.4 (84.2–98.7) | 88.0 (79.9–96.0) | 10.7 (3.0–18.3) | 5.58 (1.09–28.56) | 0.79 (0.52–1.18) |
| NC | 87.5 (58.3–100) | 97.0 (92.2–100) | 77.8 (45.0–100) | 98.5 (94.8–100) | 96.0 (90.9–100) | 10.7 (3.0–18.3) | 29.31 (7.30–117.68) | 0.13 (0.02–0.81) |
CI, confidence interval; PPV, positive predictive value; NPV, negative predictive value; LR, likelihood ratio; NC, needleless connector.
TABLE 3.
Microorganisms isolated from colonized central venous systems
| Sample no. | Microorganism (CFU) |
||
|---|---|---|---|
| Catheter tip | Hub | Connector | |
| 40 | Staphylococcus capitis (100) | Staphylococcus epidermidis (50) | Staphylococcus capitis (1,000) |
| Staphylococcus hominis (100) | Staphylococcus hominis (100) | Staphylococcus hominis (500) | |
| Klebsiella pneumoniae (30) | |||
| 67 | Staphylococcus epidermidis (>1,000) | Staphylococcus epidermidis (>1,000) | Staphylococcus epidermidis (1,500) |
| 97 | Rhodotorula mucilaginosa (>1,000) | ||
| 53 | Staphylococcus epidermidis (1,000) | Staphylococcus epidermidis (600) | |
| Corynebacterium tuberculostearicum (1,000) | |||
| 36 | Staphylococcus epidermidis (100) | Staphylococcus epidermidis (1,300) | |
| Enterococcus faecalis (200) | |||
| Klebsiella pneumoniae (70) | |||
| 31 | Staphylococcus epidermidis (500) | Staphylococcus epidermidis (1,000) | |
| 109 | Staphylococcus aureus (5,000) | Staphylococcus aureus (300) | |
| 110 | Staphylococcus aureus (900) | Staphylococcus aureus (100) | |
DISCUSSION
Our study shows that sonication of closed NCs is a better diagnostic procedure than hub culture for predicting or ruling out short-term catheter colonization. It is well known that colonization of hubs or skin surrounding the catheter insertion site is the most common source of catheter tip colonization and C-RBSI (4, 5). Therefore, superficial cultures (cultures of skin surrounding the catheter insertion site and hubs) are recommended as a conservative technique for diagnosing catheter colonization and C-RBSI, as data suggest that patients with negative superficial culture results are unlikely to have C-RBSI (3, 4, 6, 12). Obtaining hub culture samples involves manipulating the catheter and rubbing a swab inside the hub, a process that could potentially introduce microorganisms into the catheter lumen (8–10).
Currently, NCs are routinely used to lock CVC hubs. They are easy to remove without manipulation and are easy to culture in the microbiology laboratory. Our findings demonstrated that NC cultures had better sensitivity values and negative predictive values for catheter colonization than did traditional hub cultures (87.5% and 98.5% versus 25.0% and 91.4%, respectively; P < 0.001).
The main limitation of the study was that we included a small sample of patients from a single institution. Our results prove, however, that negative NC culture results are almost always indicative of negative tip colonization.
Future studies must assess whether systematic cultures of withdrawn NCs can distinguish between two populations of patients, namely, those at risk of developing C-RBSI and those not at risk. Several such studies are ongoing in our institution. The use of an alternative and noninvasive diagnostic procedure, such as sonication of closed NC cultures, can replace hub cultures for the identification of patients at low risk of catheter colonization.
ACKNOWLEDGMENTS
We thank Thomas O'Boyle for his help with the preparation of the manuscript.
M.G. was supported by the Fondo de Investigación Sanitaria del Instituto de Salud Carlos III (grant CP13/00268).
We declare no conflicts of interest.
REFERENCES
- 1.Maki DG, Kluger DM, Crnich CJ. 2006. The risk of bloodstream infection in adults with different intravascular devices: a systematic review of 200 published prospective studies. Mayo Clin Proc 81:1159–1171. doi: 10.4065/81.9.1159. [DOI] [PubMed] [Google Scholar]
- 2.Palomar M, Alvarez-Lerma F, Riera A, Diaz MT, Torres F, Agra Y, Larizgoitia I, Goeschel CA, Pronovost PJ. 2013. Impact of a national multimodal intervention to prevent catheter-related bloodstream infection in the ICU: the Spanish experience. Crit Care Med 41:2364–2372. doi: 10.1097/CCM.0b013e3182923622. [DOI] [PubMed] [Google Scholar]
- 3.Mermel LA, Allon M, Bouza E, Craven DE, Flynn P, O'Grady NP, Raad II, Rijnders BJ, Sherertz RJ, Warren DK. 2009. Clinical practice guidelines for the diagnosis and management of intravascular catheter-related infection: 2009 update by the Infectious Diseases Society of America. Clin Infect Dis 49:1–45. doi: 10.1086/599376. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Bouza E, Alvarado N, Alcala L, Perez MJ, Rincon C, Munoz P. 2007. A randomized and prospective study of 3 procedures for the diagnosis of catheter-related bloodstream infection without catheter withdrawal. Clin Infect Dis 44:820–826. doi: 10.1086/511865. [DOI] [PubMed] [Google Scholar]
- 5.Cercenado E, Ena J, Rodriguez-Creixems M, Romero I, Bouza E. 1990. A conservative procedure for the diagnosis of catheter-related infections. Arch Intern Med 150:1417–1420. doi: 10.1001/archinte.1990.00390190077011. [DOI] [PubMed] [Google Scholar]
- 6.Guembe M, Martin-Rabadan P, Echenagusia A, Camunez F, Rodriguez-Rosales G, Simo G, Echenagusia M, Bouza E. 2013. Value of superficial cultures for prediction of catheter-related bloodstream infection in long-term catheters: a prospective study. J Clin Microbiol 51:3025–3030. doi: 10.1128/JCM.01351-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Bouza E, Munoz P, Burillo A, Lopez-Rodriguez J, Fernandez-Perez C, Perez MJ, Rincón C. 2005. The challenge of anticipating catheter tip colonization in major heart surgery patients in the intensive care unit: are surface cultures useful? Crit Care Med 33:1953–1960. doi: 10.1097/01.CCM.0000171842.63887.C1. [DOI] [PubMed] [Google Scholar]
- 8.Raad II. 1994. The pathogenesis and prevention of central venous catheter-related infections. Middle East J Anesthesiol 12:381–403. [PubMed] [Google Scholar]
- 9.Sitges-Serra A, Pi-Suner T, Garces JM, Segura M. 1995. Pathogenesis and prevention of catheter-related septicemia. Am J Infect Control 23:310–316. doi: 10.1016/0196-6553(95)90062-4. [DOI] [PubMed] [Google Scholar]
- 10.Bouza E, Munoz P, Lopez-Rodriguez J, Jesus Perez M, Rincon C, Martin Rabadan P, Sánchez C, Bastida E. 2003. A needleless closed system device (CLAVE) protects from intravascular catheter tip and hub colonization: a prospective randomized study. J Hosp Infect 54:279–287. doi: 10.1016/S0195-6701(03)00136-1. [DOI] [PubMed] [Google Scholar]
- 11.O'Hara CM, Weinstein MP, Miller JM. 2003. Manual and automated systems for detection and identification of microorganisms, p 185–207. In Murray PR, Baron EJ, Pfaller MA, Jorgensen JH, Yolken RH (ed), Manual of clinical microbiology, 8th ed American Society for Microbiology, Washington, DC. [Google Scholar]
- 12.Bouza E, Rojas L, Guembe M, Marin M, Anaya F, Luno J, López JM, Muñoz P. 2014. Predictive value of superficial cultures to anticipate tunneled hemodialysis catheter-related bloodstream infection. Diagn Microbiol Infect Dis 78:316–319. doi: 10.1016/j.diagmicrobio.2013.12.008. [DOI] [PubMed] [Google Scholar]