Materials and Methods
An open, prospective, randomized, 7-day clinical trial was conducted to assess the complications associated with arterial catheter use for hourly blood sampling to monitor glycemia. The eligible population was all patients admitted to a 10-bed medicalsurgical ICU in a community hospital from April 2008 through December 2011. We consecutively included all patients requiring IIT for glycemia control and who had been fitted no more than 5 days previously with a radial arterial catheter (polyurethane, 20- gauge × 6.35 cm; Mayo Healthcare).
Patients were excluded when their arterial catheter was not permeable, the arterial catheter was overdamped or underdamped according to a square-wave test, a Flotrac system (Edwards Lifesciences) was used, they had effort limitations based on therapeutic orders, or they declined to participate in the study. All included patients, or their guardians, provided informed consent. The hospital's Committee on Ethics and Clinical Research approved the protocol.
Intervention Group
Clinical trial participants were randomized to 1 of 2 arterial catheter setups: Needleless, with Lueractivated needle-free valve (MicroCLAVE, Victus Inc), single-use syringe, and fast flush from the continuous flushing system or syringe, adding to the needleless setup a permanent syringe that requires manual flushing after each blood sample is obtained by aspirating and injecting 2 mL of the flush solution (Figure 1). In both configurations, heparinized isotonic sodium chloride solution was used for flushing, with a pressure bag at 300 mm Hg for a delivery rate of 3 mL/h. Nonwaste method was used for all samples: 2 mL of heparinized isotonic sodium chloride solution was withdrawn from the stopcock farthest from the insertion site, leaving the syringe attached. The sample was withdrawn from the stopcock nearest the insertion, and then blood in the attached syringe was reinfused.
(Enlarge Image)
Figure 1.
Design and assembly of arterial catheter setups.
Control Group
All remaining patients admitted to the ICU during the same period, fitted with the ICU's standard arterial catheter (Figure 1) and not included in the clinical trial, constituted the control group. In controls, the arterial catheter was used only for blood pressure measurement and blood extraction for daily routine laboratory analysis.
Study Protocol
Two consecutive recordings were made of the patient's pressure waveform, first with the ICU's standard arterial catheter setup and then with the randomly assigned arterial catheter setup. The following paired blood samples from the same arm were obtained once daily for 2 consecutive days:
One sample through the arterial catheter and another by fingerstick, which were analyzed by using the OptiumXceed glucometer (Abbott Diabetes Care, MediSense Products). A single device was assigned to each patient and calibrated according to the manufacturer's instructions. Strips appropriate to that calibration were documented and kept in each patient's room.
Another sample was drawn by direct venipuncture and sent to the central laboratory within 15 minutes.
In both arms of the study, arterial blood was extracted from the arterial catheter for hourly glycemia monitoring during the first 2 days of the study. After those 2 days, if the patient continued under IIT, hourly glycemia monitoring was performed by fingerstick. Blood samples for culture were collected from the hub and the tip of the arterial catheter was cultured once the catheter had been in place for 7 days. In addition, 2 to 4 days after the catheter was withdrawn, arterial flow of the puncture zone was assessed by Doppler echography (Acuson S2000, Siemens).
Objectives
Our hypothesis was that an increase in arterial catheter manipulation with closed circuits would not increase the incidence of catheter-related infection or of arterial ischemia and that blood glucose levels in samples obtained from the arterial catheter and analyzed at the POC would be as accurate as those obtained by venipuncture and analyzed by the hospital laboratory.
The study objectives were as follows:
To evaluate the effectiveness and complications of a blood sampling technique that uses an arterial catheter, configured either as a needleless or a syringe setup.
To compare the glycemia results by type of sampling used in 3 different clinical situations: administration of vasoactive drugs, presence of peripheral edema, and hematocrit range.
Independent Variables
Primary: Random assignment to needleless or syringe arm of the clinical trial. Secondary:
Sociodemographic variables (age and sex)
Diagnosis at admission, diagnostic category (medical or surgical)
Length of ICU stay
ICU mortality
Length of hospital stay
Clinical characteristics including Acute Physiology and Chronic Health Evaluation (APACHE II) score
Level of vasopressor drug therapy: none, low doses (<0.25 μg/kg per minute of norepinephrine or <5 μg/kg per minute of dopamine), or higher doses
Presence of peripheral edema according to the edema rating scale
Hematocrit range (<25%, 25%-30%, >30%)
Number of arterial puncture attempts to insert the arterial catheter
Presence of hematoma at the puncture point
Administration of systemic anticoagulant or antiplatelet agents
Presence of coagulopathy, defined as prothrombin time >13 seconds or activated partial thromboplastin time (APTT) >40 seconds or platelet count <150 x 10/L
Dependent Variables
Primary outcome: episodes of catheter-related bloodstream infections (CRBSIs) per 1000 catheter-days.
Secondary outcome: arterial catheter colonization, distorted pulse waveform or catheter obstruction, concordance in glycemia values, presence of hand or finger ischemia, and radial artery pseudoaneurysm, thrombosis, or air embolism.
Definitions
According to the criteria of the Centers for Disease Control and Prevention (CDC) and the Infectious Diseases Society of America:
Colonized catheter was defined as a growth of at least 15 colony-forming units (CFUs) in a semiquantitative tip culture.
CRBSI was defined as the presence of the same microorganism in both peripheral blood and in cultures of the catheter tip obtained 48 hours before or after removal of the arterial catheter in a patient with clinical signs of sepsis and no other source of bacteremia except the arterial catheter, and 1 of the following 2 criteria:
Simultaneous quantitative blood cultures with a ratio of more than 3:1 CFU/mL between catheter and peripheral blood;
Differential time to positivity, as detected by an automated blood culture system (at least 2 hours difference between cultures of blood obtained through a catheter hub and of an equal volume of simultaneously sampled peripheral blood).
Vascular Complications
Doppler ultrasound for venous and arterial hemodynamic functioning was used to detect the following arterial complications.
Pseudoaneurysm: actual disruption of the intima and media layer, preserving the adventitia
Thrombosis: decreased or absent pulse distal to puncture, used to determine temporal occlusion of radial artery
Hand ischemia: depth of ischemia according to lower-extremity ischemia standards
Retrograde air embolism: macro air bubbles visible in the arterial circuit during reintroduction of the clearing volume
Permeability of the Arterial Catheter
An arterial catheter was considered unobstructed when the square-wave test and the patient's pressure waveform remained unchanged after interposing a 3-way stopcock and free backflow of blood from the catheter was observed. When the stopcock is turned off to the transducer and a syringe is attached, a blood flashback to the tubing occurs within 1 second.
Sample Size Calculation
Sample size was calculated by using the CRBSI variable because glycemia values were not expected to differ between the intervention groups. Of the 77 ICU patients who met the inclusion criteria during the 1 year before the recruitment period, 2 had catheter-related bacteremia. Therefore, at a 95% confidence interval (CI), 80% power, and 10% expected losses, the estimated sample size was 90 patients (45 per study arm).
Randomization
A computer-generated list in permuted blocks of 4 per stratum was used to randomize patients into intervention groups. The hospital epidemiologist prepared the patients' codes in closed envelopes, labeled with numbers indicating the syringe or needleless setup. The ICU nurses opened these envelopes to assign the patient to an intervention arm. Therefore, this was an open clinical trial because the assigned setup was visible to researchers, patients, and nurses. Diagnostic testing was done after the arterial catheter was removed. The vascular surgeon who performed the vascular Doppler echogram and the microbiologists who determined catheter-related infection and/or colonization were blinded to the study arm assignments.
Statistical Analysis
Categorical variables are expressed as frequency and percentage, using the Fisher or χ test to compare groups. Results for quantitative variables are shown as mean and standard deviation (SD) or median and interquartile range (P25-P75). Depending on distribution, comparisons were made by using the Student t test or Mann-Whitney U tests. Wilcoxon test was used to compare results of the 3 sample types. Statistical significance was defined as a P value less than .05, and SPSS v18 (IBM Statistics) was used for analysis.