Accurate placement of central venous catheters: a prospective, randomized, multicenter trial.

OBJECTIVES: a) To define the frequency of dangerous (intracardiac) central venous catheter placement in a multicenter study of large community hospital intensive care units (ICUs) and to evaluate physician responses to this finding. b) To validate right atrial electrocardiography as a technique to assure adherence with recent Food and Drug Administration (FDA) guidelines regarding the location of central venous catheter tips. c) To conduct a literature review of vascular cannulation and its associated potentially lethal complications. DESIGN: Prospective, randomized, blinded, multicenter study. SETTING: Multidisciplinary ICUs in five large community teaching hospitals. PATIENTS: Consecutive patients (n = 112) who required a central venous catheter by either internal jugular vein or subclavian vein at four separate hospitals were assessed using 30-cm catheters. Consecutive patients (n = 50) in a fifth hospital who subsequently required a central venous catheter via the internal jugular vein or subclavian vein route were prospectively randomized to receive a 20-cm central venous catheter with either conventional surface-landmark guidance, or with the right atrial electrocardiography-guided technique. MAIN OUTCOME MEASURES: a) Occurrence rate of malpositioned central venous catheters. b) Ability of right atrial electrocardiography to aid in the accurate placement of central venous catheters. RESULTS: a) Using conventional placement techniques with a 30-cm catheter, 53 (47%) of 112 initial central venous catheter placements resulted in location of the catheter tip within the heart. Catheter tips were not repositioned to locations outside the right atrium after this finding was identified on initial post-procedure films. b) Using the right atrial electrocardiography technique to place 20-cm central venous catheters resulted in no catheter tip locations within the heart (0/25) vs. 14 (56%) of 25 (p < .0001) intracardiac placements using conventional techniques. c) The literature suggests that serious mechanical complications of central venous catheterization, although uncommon, are associated with a high mortality rate. Deaths are associated with intracardiac placement. CONCLUSIONS: a) The FDA guidelines regarding catheter tip location (catheter tip should not be in the right atrium) have not been widely publicized. b) The average safe insertion depth for a central venous catheter from the left or right internal jugular vein or subclavian vein is 16.5 cm for the majority of adult patients; a central venous catheter should not be routinely inserted to a depth of > 20 cm. Catheters longer than this size are rarely needed, and potentially dangerous. Catheter tip location is important to document following central venous catheter insertion. Thirty-centimeter central venous catheters should not be used when accessing the central circulation via internal jugular or subclavian veins. c) Right atrial electrocardiography is a technique that assures initial tip position outside the heart in accordance with FDA guidelines. This technique would virtually eliminate the major risk of death (i.e., cardiac perforation) associated with this procedure. d) Recently available, 15- and 16-cm central venous catheters have significant potential to minimize intracardiac placement of central venous catheters by either the internal jugular or subclavian vein route and may become the standard of care.

Effects of pentoxifylline on tumor necrosis factor production and survival during lethal E. coli sepsis vs. disseminated candidiasis with fungal septic shock.

Lethal circulatory shock during microbial sepsis is thought to be initiated by early molecular events, including production of tumor necrosis factor (TNF) and cytokine-mediated upregulation of neutrophil (PMN) function, irrespective of the causative organism. The phosphodiesterase inhibitor pentoxifylline (PTX) inhibits TNF gene transcription and modulates PMN function, and has been shown to improve outcome in experimental sepsis. We hypothesized that PTX would attenuate gram-negative and fungal septic shock by different mechanisms: reduced TNF production in Escherichia coli (EC) sepsis vs. enhanced PMN-mediated defense during Candida albicans (CA) fungemia. Conscious chronically catheterized rats received PTX (25 mg/kg, i.v.) before i.v. challenge with 10(10) viable EC (serotype 055:B5), 10(9) viable serotype A yeast-phase CA (each the LD100 in < 24 hr in naive rats), or normal sterile saline (NSS), and then PTX posttreatment (6.5 mg/hr x 4.5 hr). Treatment controls received NSS before and after challenge. Serum TNF peaked 1.5 hr after EC infection in NSS-treated animals (1654 +/- 390 U/ml, mean +/- SE), and was significantly reduced by PTX (120 +/- 32 U/ml, P < 0.01), but PTX did not improve 24 hr survival. PTX also aggravated systemic hypotension after EC, and did not modify neutropenia, thrombocytopenia, or microvascular permeability assessed by organ wet/dry weight (W/D) ratios. Peak serum TNF in CA + NSS animals (130 +/- 45 U/ml) was delayed 8 hr compared to EC animals, and were not reduced by PTX (67 +/- 25 U/ml, P = NS). Moreover, PTX did not alter CA-induced mortality, hypothermia, hypotension, neutropenia, increased lung W/D, or interstitial and alveolar hemorrhage. We conclude that PTX-induced suppression of endogenous TNF production does not prevent gram-negative shock in this model, possibly due to impaired TNF-mediated antibacterial host defense. Since fungal septic shock with acute disseminated candidiasis evolves prior to significant increases in circulating TNF, PTX also appears ineffective in its treatment.

Pelvic fracture classification: correlation with hemorrhage.

Hemorrhage remains the leading cause of death in patients with pelvic fractures. To identify patients at greatest risk for massive hemorrhage, we retrospectively reviewed charts and initial emergency room anterior-posterior (AP) radiographs of 245 consecutive patients. Pelvic fractures were classified according to our modification of the Pennel and Sutherland classification scheme. A simple pelvic fracture classification scheme was developed. Using this classification, we can be 90% confident that 50 to 69% of patients with "unstable" pelvic fractures will require 4 or more units of blood, 30 to 49% will require greater than 10 units of blood, 36 to 55% will have an intra-abdominal injury, and 6 to 18% will have a pelvic arterial injury. Therefore we conclude that this pelvic fracture classification based on the initial emergency-room AP X-ray can predict a patient population at high risk for massive hemorrhage for which an aggressive treatment protocol is justified.