Increasing antimicrobial resistance in gram-negative bacilli isolated from patients in intensive care units.

BACKGROUND: We investigated gram-negative bacilli from patients in intensive care units to determine whether antimicrobial resistance was increasing. METHODS: Minimal inhibitory concentrations were determined by broth microdilution on 334 gram-negative bacilli collected in 1990, 1995, and 1998. RESULTS: During the 3 study years, the types of gram-negative bacilli encountered in our intensive care units changed with proportional increases of Pseudomonas sp and decreases of inducible enterics. Dramatic increases in resistance for ceftazidime, cefotaxime, and piperacillin were paralleled between respiratory-tract isolates and inducible enterics. By 1998, ticarcillin was more active than piperacillin against most isolates except Escherichia coli and Klebsiella sp, and most isolates became more resistant to gentamicin and tobramycin. CONCLUSIONS: Continuous changes in the types of gram-negative bacilli and antimicrobial resistance complicate empirical selection of antimicrobials in the intensive care units. These complications will place more emphasis on communication and strategy formations among health care workers (nurses, physicians, laboratorians, and pharmacists) in an effort to treat infections in a timely and effective manner.

Hemodynamics of revascularization for iliofemoral venous occlusion: a short-term canine model.

A short-term canine model of lower extremity venous hypertension was created to study the hemodynamics of crossfemoral venous bypass grafts (CFB). Specifically, the hemodynamic effects of bypass conduit diameter and adjunctive arteriovenous fistulas (AVFs) were investigated. Unilateral hind limb venous hypertension was produced by iliofemoral venous ligation in six groups of five greyhounds each. Group I had venous ligation alone. CFBs were constructed in the remaining five groups: group II, 3 mm bypass conduit alone; group III, 3 mm bypass plus sequential AVF; group IV, 3 mm bypass plus caudad AVF; group V, 3 mm bypass plus cephalad AVF; group VI, 6 mm bypass conduit alone. Venous hypertension was significantly reduced by CFB (group II, p less than 0.025; group VI, p less than 0.001); increasing the diameter of the bypass conduit from 3 to 6 mm produced significantly greater graft flow (p less than 0.05), while completely relieving venous hypertension. Addition of adjunctive AVFs significantly augmented graft flow (p less than 0.001) but tended to aggravate ipsilateral venous hypertension (group III, p less than 0.01; group IV increase, NSS; group V, p less than 0.001). During the 4 hours of pressure monitoring, venous hypertension diminished significantly (p less than 0.05) with the sequential AVF but not with the other AVF. We conclude that (1) AVFs may be required for adequate graft flow if a small-diameter (3 mm) bypass conduit is used to relieve venous hypertension; (2) adjunctive AVFs aggravate venous hypertension; (3) sequential AVFs seem to be the most hemodynamically efficacious; (4) AVFs may not be necessary if a large, isodiametric (6 mm) conduit is used.

Canine model for surgical correction of chronic venous hypertension.

A chronic model of venous hypertension was created by iliofemoral venous ligation in the left hindlimb of eight greyhounds (right limb control). Prior to ligation, immediately postligation (T0), and at 2, 4, 6, and 15 weeks postligation (T2-T15), bilateral hindlimb venous pressures were measured at rest. At T2-T15, exercise pressures were measured for 5 min after hindlimb muscle contraction induced under sedation by tetanic (20 sec, 50 Hz, 8 mA) stimulus (S30-S300 sec). Resting and exercise pressures were significantly higher in the ligated hindlimb at all time intervals (resting, P less than 0.001 at T0, T2, T4, T6, and P less than 0.025 at T15; exercise, P less than 0.001 for S30-S300 at T2-T15). In five of the animals at 15 weeks, a cross-femoral venous bypass (CFB) with adjacent adjuvant arteriovenous fistula (AVF) was constructed using autologous external jugular vein. Venous pressures were measured at 2 and 6 weeks post-CFB both at rest and after tetanic muscle stimulation. At 2 weeks, venous pressure in the ligated limb was significantly higher than that in control only after exercise (P less than 0.025 at S30-S120; P less than 0.05 at S150-S180; NS at S240-S300). At 6 weeks, venous pressures were not significantly different at rest or with exercise. All five grafts were patent at 6 weeks. Three of five dogs then successfully underwent ligation of the AVF. At 2 weeks post-AVF ligation there was no difference in resting or exercise venous pressure in the ligated limb compared to control. At 6 weeks post-AVF ligation only two dogs remained for monitoring and in these there was no trend toward venous hypertension in the ligated hindlimb. Graft patency was maintained despite AVF ligation.

Detergent-extracted small-diameter vascular prostheses.

The exact mechanism that leads to thrombosis of small-diameter vascular prostheses (less than 4 mm X greater than 6 cm) is unknown. This report presents preliminary patency and healing data on a sequential detergent-extraction technique for the preparation of autogenous small-diameter microvascular grafts. Fifteen carotid interpositional allografts (3 to 4 mm X 4 to 6 cm) were implanted in 15 mixed species adult greyhounds. Ninety days after implantation grafts were perfusion-fixed in situ, harvested, and evaluated by light microscopy and scanning and transmission electron microscopy. Two categories of acellular vascular matrix grafts were evaluated: non-cross-linked and cross-linked (1% carbodiimide). By objective morphologic analysis with blind random view, histologic sections were rated from 0 to 4 in five categories believed to be important for graft healing and patency. Overall graft patency was 80% (12 of 15), and there was no significant difference between cross-linked and non-cross-linked grafts. Non-cross-linked grafts were superior to cross-linked grafts in all areas of histologic evaluation except immunogenicity (p less than 0.01). Most important, non-cross-linked grafts demonstrated complete endothelial coverage (p less than 0.001). There was no significant difference (that is, normal) between control autografts and non-cross-linked grafts; however, there was a significant difference between control autografts and cross-linked grafts in all parameters except immunogenic reaction (p less than 0.01).

Delayed thrombosis of synthetic microvascular bypass grafts.

A 3-mm-diameter synthetic polytetrafluoroethylene (PTFE) cervical carotid bypass graft 20 cm in length was implanted in 30 dogs for the evaluation of blood flow, tissue response, and patency at intervals of 1 to 120 days. Although 4 of 5 grafts removed after 5 to 8 days were patent (80%), long term patency was observed in only 1 graft (10%). Aspirin treatment did not influence patency. Scanning electron microscopy demonstrated the lack of a neoendothelial layer upon the luminal surfaces of patent grafts, which were covered with a fibrin-blood cell lining. Subintimal fibrosis resulted in stenosis at sites of anastomosis in thrombosed grafts. The graft length, its small caliber, and a 40% decrease in blood flow after implantation may have contributed to thrombosis of the bypass graft in this model. Synthetic PTFE microvascular grafts may not be suitable for clinical use in extracranial-intracranial arterial bypass surgery.