| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
*Departments of Anesthesiology and Critical Care Medicine, Surgery, and Pediatrics, The Johns Hopkins University School of Medicine; and
Department of Biostatistics, The Johns Hopkins University School of Public Health, Baltimore, Maryland
Address correspondence and reprint requests to Deborah A. Schwengel, MD, Blalock 904, The Johns Hopkins Hospital, 600 N. Wolfe St., Baltimore, MD 21287. Address e-mail to dschweng{at}jhmi.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
IMPLICATIONS: In this randomized, controlled trial, we found that peripherally inserted central catheters safely and effectively reduce needle punctures and improve patient satisfaction. The technique is cost-effective if anesthesiologists insert the catheters during surgical preparation time.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
We hypothesized that preemptive placement of PICCs during elective surgery would decrease or eliminate painful venipunctures and improve patient (parental) satisfaction when used for 4–7 days. Additionally, we hypothesized that patients with PICCs would have fewer catheter-related complications than patients with PIVs and that this technology would be cost-effective even with short-term use.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
After approval of the institution’s IRB and receipt of informed parental consent and, when applicable, patient assent, patients were randomized by using a sealed, opaque envelope to receive either PICC or PIV. Randomization was performed before the induction of anesthesia for the proposed surgical procedure. There was no patient stratification within groups. Anesthetic management was not standardized, but after the induction of general anesthesia, all patients had a PIV inserted. Patients randomized to the PICC group had a PICC (Cook Incorporated, Bloomington, IN) inserted by an attending anesthesiologist (Fig. 1). Veins for cannulation were identified by palpation or visualization. An ultrasound technique was not used. In the PICC group, the PIV placed after the induction of anesthesia was either converted to a heparin lock and not used or was removed.
|
Postoperative fluid management and laboratory studies were dictated by the patient care team. Bedside nurses kept a running log of problems with catheter care and the number of needle punctures a patient received. A study nurse collected data on complications associated with IV therapy, including catheter occlusion and malfunction, broken catheters, catheter dislocation, IV infiltration, bacteremia, fever, fluid leakage or blood at the insertion site, phlebitis, and missed or delayed fluid or drug administration. Parents and the patient (if old enough) were given a satisfaction survey to complete at the time of PICC removal, at discharge, or at 7 days.
For statistical analysis, Stata Version 7.0 (Stata Corp., College Station, TX) and Prism Version 3.0 (GraphPad Software, San Diego, CA) were used. Exploratory data analysis was conducted on all data collected and presented in appropriate graphical displays and tables. Statistical significance was defined as P < 0.05.
The sample size estimate was based on our satisfaction hypothesis. At the inception of the study, we had estimated a sample size of 97 patients in each group, assuming a 40% excellent satisfaction score in the PIV group versus a 60% excellent satisfaction score in the PICC group (
error = 0.05; and ß error = 0.8). We planned an interim analysis based on O’Brien and Fleming’s stopping boundary (11).
Our primary hypothesis was that patient/parent satisfaction levels would differ between the PIV and the PICC groups. We compared the proportion of patients in each group who reported the highest level of satisfaction on each of the eight questions of the patient satisfaction survey. We calculated the difference in proportions between groups and the 95% confidence intervals (95% CI). Statistical significance was assessed with Fisher’s exact tests.
Patients randomized to PICC who had failed PICC placement were analyzed in the PIV group rather than in an intention-to-treat group. A sensitivity analysis was performed to determine whether our results would change when the failed PICC patients were removed from analysis. Ninety-five percent CIs for group differences in responses to each of the eight satisfaction items were recomputed without including data on the failed PICC patients in the PIV groups; P values were also recomputed. The results of both analyses (with and without failed PICC patients) were compared to see whether the substantive interpretation and statistical results differed.
We also compared the complications in the two groups. Objective measures of patient complications, such as the mean number of insertion attempts on the first study device, and other continuous measures were compared via a two-sample Student’s t-test, and 95% CIs were computed for mean differences between the PIV and PICC groups.
We evaluated the potential effect of PICC placement on economic outcomes. We calculated the provider costs for each patient by collecting the labor (anesthesiologist and phlebotomist time), equipment (PICC trays and IV catheters), and operating room (OR) time costs. Anesthesiologist cost was calculated per unit of time on the basis of average salary and benefits at our institution. Phlebotomist cost was estimated on the basis of average salary plus benefits at our institution divided by the average time required for each venipuncture. We measured the number of minutes required to place the initial study catheter in the OR. To establish an average cost for either IV starts or phlebotomy, we had previously recorded time spent by pediatric phlebotomists for 358 patient encounters. On the basis of these data, the average amount of time per encounter was calculated for IV starts and phlebotomy. Therefore, PICC and PIV costs were calculated according to the following formula:
|
|
To evaluate, we calculated the satisfaction score based on the results of the satisfaction survey. An aggregate satisfaction score was obtained based on the scaled Likert score. Each question was given a possible maximum value of 1 and a minimum value of 0. The maximum possible score for each patient was 8, and the lowest possible score was 0. Each of the five possible responses on the Likert scale was weighted as follows: highest satisfaction = 1, some satisfaction = 0.75, neither satisfied nor dissatisfied = 0.5, some dissatisfaction = 0.25, very dissatisfied = 0. The scores were totaled for each patient. The cost-effectiveness was calculated as total costs/satisfaction score.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
There were no significant differences between groups in terms of patient age, sex, weight, or number of previous hospitalizations. All patients received routine maintenance and third-space IV fluid replacement, as well as surgical and patient-specific IV antibiotic prophylaxis.
We scored all of the patient surveys that responded with highest satisfaction to each of the eight questions of the survey. The results were all statistically significant except for Question 5 (Table 1). The aggregate score for all questions was also significantly higher among patients in the PICC group (mean, 6.76; SD, 1.57) compared with the PIV group (mean, 4.95; SD, 1.84) (P < 0.0001).
|
We observed a significantly different number of needle punctures between groups. The number of postoperative IV restarts (SD) was 0.27 (1.07) in the PICC group and 1.5 (2.68) in the PIV group. The number of venipunctures was 0.3 (1.0) in the PICC group and 1.4 (1.9) in the PIV group.
Major complications were not different between groups. There were two study patients with positive blood cultures, one in each group, and neither was clearly due to the IV catheter. There were no complications such as broken or occluded PICCs, and there were no differences in maximum daily temperatures between groups. The two most common complications of the PICC were old blood at the site (56%) and failed placement (17%). One (2%) catheter fell out prematurely, and 36 (92%) functioned well during the study period. Other relatively minor complications were more common in the PIV group (Table 2).
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Hospitalization and the anticipation of painful medical procedures and needles ("shots") are terrifying experiences for many children (12,13). PIV catheters inserted at the time of surgery cannot be expected to last the duration of a patient’s postoperative hospital course, nor can they reliably be used to sample blood. Consequently, venipuncture becomes a necessary, albeit unfortunate, consequence of hospitalization and illness. However, it does not have to be painful. Topical local anesthetics such as EMLA® cream can minimize venipuncture pain, but these topical drugs do not eliminate pain or prevent dread of the procedure (14).
PICCs eliminate the need for multiple venipunctures. These catheters have become commonplace and have revolutionized the care of chronically ill children and children in whom IV access is difficult or impossible. Janes et al. (15) concluded that PICCs used in very-low-birth-weight infants significantly reduced IV insertions and needle punctures without adding morbidity in that patient population. Our study extends this technology to postoperative patients in whom the duration of use would be expected to last four to seven days.
Regardless of where and how they are placed, serious complications such as pneumothorax, thrombosis, infection, and catheter fracture have been associated with the use of central catheters. PICCs are less likely to incur major complications compared with central catheters placed in the subclavian or internal jugular veins. However, minor complications are common (16,17). Long-duration (up to 77 days) series reports from neonatal intensive care units describe complications as infectious (bacteremia/sepsis) or mechanical (2–7). Mechanical complications, such as catheter occlusion, fluid extravasation, dislodgment, or thrombosis, were the reason for removal of the lines in 10%–43% of the cases (2,4–6). Complications in older patients are infrequent (8–10), but there are isolated case reports of difficulty removing catheters and broken catheters (18,19). There are no reports of short-term PICC use, but complication rates reported in one case series increased once catheters were in place for 
20 days (20). Others have reported that central placement results in less phlebitis, occlusion, and leaking than noncentral tip locations (21).
We did not see more frequent complications in the PICC group in this study. In fact, there was more local injury in the PIV group (phlebitis and infiltrates). One concern about the use of PICCs is the possibility of thrombotic complications. We did not have any clotted catheters or symptomatic thrombotic events; we did not evaluate the presence of asymptomatic thromboses, nor do we know the significance of asymptomatic thromboses. Screening for asymptomatic thromboses should be the topic of future studies.
The cost of PICC is higher than PIV, both in terms of labor and equipment. Our lowest total cost of $173.58 was 1.59 times the cost of PIV care ($108.49). The highest cost increased to $440.70 (4 times the cost of the PIV). However, the total cost in relation to a hospital stay is a small fraction of the total. Patients may ultimately have no awareness of the individual monetary cost of the PICC, but they will be very aware of their level of satisfaction and memory of the number of needle punctures, number of times awakened from sleep, and other unpleasant hospital events. Parents also commented on the survey that the PICC was more comfortable, with no arm board and better limb mobility. Patients might score the value of the PICC differently than the provider. The increased costs of PICCs might be viewed negatively by administrators. Satisfaction scores can be used as a way to evaluate the value of treatment to the patient. Longer hospitalizations make the PICC more cost-effective, and if PIV placement becomes impossible, PICC or central line placement becomes necessary and might require an additional anesthetic.
Each physician and each institution will have to decide whether their costs of PICC placement will be more or less than those we have reported. To achieve the cost of $173.58, we placed PICCs at the same time that the patient was being prepared for the procedure or when the procedure was under way. If OR time is used exclusively for PICC placement, then costs for OR time and the anesthesiologist’s time should be included in the analysis. We also placed our catheters in the midclavicular position unless the catheter would be used for parenteral nutrition or medications requiring central location; consequently, chest radiographs were not routinely performed. In absolute dollars, the cost might still be very positive for the patient and the provider because of the improvements in patient satisfaction that result from the reduction of painful events. Each institution must determine whether their costs warrant placement.
Finally, insertion failure was more frequent than previously reported. We do not believe that this was due to inexperience, the size of the catheters (3F or 4F), or operator inexperience. Rather, in many of the children studied, the basilic, cephalic, and greater saphenous veins could not be readily seen or palpated. Perhaps, in retrospect, this technique should not have been attempted in these patients. Alternatively, we could use techniques and technologies that could increase our ability to access peripheral veins, such as ultrasound imaging or transillumination. Ultrasound imaging uses relatively inexpensive technology to image veins and arteries in real time. It is commonly used with great success in the catheterization laboratories and in some ORs and deserves future study.
The satisfaction survey that we used is not a validated tool, but the results were highly clinically and statistically significant when analyzed in two different ways. Objective results, including the number of IV restarts and the number of venipunctures for blood sampling, were significantly more frequent in the control group. Because the control group also had lower satisfaction scores, we conclude that the objective and subjective data corroborate each other. Because this was a nonblinded study, we cannot eliminate the possibility of attitude bias introduced by staff members caring for the patients. To minimize bias in completing the survey, parents were given the survey to complete in private for return to a nurse in a sealed envelope.
Anesthesiologists, surgeons, or other physicians should consider placing PICCs in patients who require more than four days of in-hospital postoperative care, especially if frequent blood sampling or IV access is anticipated or required.
|
Acknowledgments |
|---|
The authors wish to acknowledge the faculty, fellows, and pediatric pain nurse practitioners of the Department of Anesthesiology and Critical Care Medicine of the Johns Hopkins University School of Medicine. Additionally, the authors wish to acknowledge the support of the faculty of the Divisions of Pediatric General Surgery, Pediatric Neurosurgery, Pediatric Orthopedics, and Pediatric Urology of the Johns Hopkins University School of Medicine and the nurses of the Children’s Center of the Johns Hopkins Hospital.
|
Footnotes |
|---|
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
