Which goal for fluid therapy during colorectal surgery is followed by the best outcome: near-maximal stroke volume or zero fluid balance?

BACKGROUND: We aimed to investigate whether fluid therapy with a goal of near-maximal stroke volume (SV) guided by oesophageal Doppler (ED) monitoring result in a better outcome than that with a goal of maintaining bodyweight (BW) and zero fluid balance in patients undergoing colorectal surgery. METHODS: In a double-blinded clinical multicentre trial, 150 patients undergoing elective colorectal surgery were randomized to receive fluid therapy after either the goal of near-maximal SV guided by ED (Doppler, D group) or the goal of zero balance and normal BW (Zero balance, Z group). Stratification for laparoscopic and open surgery was performed. The postoperative fluid therapy was similar in the two groups. The primary endpoint was postoperative complications defined and divided into subgroups by protocol. Analysis was performed by intention-to-treat. The follow-up was 30 days. The trial had 85% power to show a difference between the groups. RESULTS: The number of patients undergoing laparoscopic or open surgery and the patient characteristics were similar between the groups. No significant differences between the groups were found for overall, major, minor, cardiopulmonary, or tissue-healing complications (P-values: 0.79; 0.62; 0.97; 0.48; and 0.48, respectively). One patient died in each group. No significant difference was found for the length of hospital stay [median (range) Z: 5.00 (1-61) vs D: 5.00 (2-41); P=0.206]. CONCLUSIONS: Goal-directed fluid therapy to near-maximal SV guided by ED adds no extra value to the fluid therapy using zero balance and normal BW in patients undergoing elective colorectal surgery.

Short hospital stay and low complication rate are possible with a fully implemented fast-track model after elective colonic surgery.

PURPOSE: The aim was to implement a fast-track model in a colorectal unit. We evaluated its effects on hospital stay and complication rate after elective open colonic surgery. METHODS: A fast-track programme was fully implemented, with a dedicated staff and a schedule for all perioperative procedures which included provision of information to patients, surgical guidelines, mobilisation and postoperative care. All previously existing procedures were modified according to previously published guidelines. Criteria for success with regard to the length of hospital stay and rates and types of complications were defined. Fast-tracked patients were compared with patients not included in the programmes that were operated during the same period. RESULTS: 131 fast-tracked patients were compared with 39 control patients. The fast track significantly reduced the median hospital stay from 7 to 3 days (p < 0.0001). There was no difference in complication rates between the two groups, and no major complications were observed after early discharge from the hospital. The readmission rate was 15% in the fast-track group and 16% with the control patients. CONCLUSION: Implementation of a fast track after open elective colonic surgery is safe and reduces the length of hospital stay.

Primary hyperparathyroidism: intraoperative PTH-measurements.

BACKGROUND: With the development of rapid assays and intraoperative measurement of intact parathyroid hormone (PTH), new strategies in the handling of patients with primary hyperparathyroidism (pHPT) have evolved. AIM: The aim of our study was to illustrate the performance of the intraoperative PTH measurement as a predictor of successful cure. MATERIAL AND METHODS: From September 1999 to April 2002 143 patients with pHPT underwent a parathyroid operation (bilateral neck exploration with identification of all parathyroid glands) with intraoperative measurements of plasma PTH (immediately prior to surgery (T0) and 5 minutes after gland excision (T5)). A positive test result was defined as plasma PTH values at T5 below 20% of T0 or a value in the normal range below 7.6 pmol/l. Hence T5 values above 20% of T0 and above 7.6 pmol/l were considered test negative. RESULTS: 122 patients (85%) were test positive and cured, 11 patients (8%) were test negative but cured, and 10 patients (7%) were test negative and not cured by the primary operation. Consequently, the sensitivity of the test was 0.92 and the specificity 1.00. CONCLUSIONS: The rapid PTH test used is a reliable predictor of a successful outcome in pHPT patients undergoing parathyroid surgery.

Epidural anaesthesia with bupivacaine does not cause increased oedema in small gut anatomoses in pigs.

BACKGROUND AND OBJECTIVE: Epidural analgesia is widely used for abdominal surgery due to the properties of 'stress-free' anaesthesia and superior pain control. Nevertheless, sympathomimetics are known to antagonize inflammation. The present study was performed to investigate if epidural local anaesthetics caused increased local oedema formation. METHODS: Thirty Dansk Landrace pigs were randomized into three groups: epidural bupivacaine, epidural morphine or intravenous (i.v.) fentanyl. All animals were anaesthetized with isoflurane and i.v. midazolam and received an identical fluid regimen. Six small bowel resections were performed over a 3-h period and during the following 3 h the anastomoses were resected. Primary end-points were water content in small bowel and mesentery samples before and after gut anastomosis, lymph flow and urine production. RESULTS: The water content in the small bowel samples was not changed by surgery or by the different anaesthetic protocols. In the mesenteric tissue, there was a highly significant increase in water content of the postanastomotic samples compared to pre-anastomotic samples (P < 0.001) and a significant time treatment interaction was revealed (P < 0.05) suggesting an increase in oedema formation in the epidural local anaesthetic group. Lymph flow did not change during the experiments and there were no significant differences between the groups (P = 0.80). The mean total urine output was 44% higher in the epidural morphine group compared to the local anaesthetic group (P = 0.17). CONCLUSIONS: Surgery did not increase gut wall water content, but acute oedema formation resulted in the peri-resectional mesenterial tissue, more prominently so in the bupivacaine group.

Effects of preoperative oral carbohydrates and peptides on postoperative endocrine response, mobilization, nutrition and muscle function in abdominal surgery.

BACKGROUND: Surgery is succeeded by long-lasting state of relative peripheral insulin resistance, which is reduced by giving glucose infusion or oral carbohydrate-rich drinks immediate before operating instead of fasting. The aim of the present study was to investigate whether oral carbohydrate or carbohydrate with peptide drinks preoperatively instead of fasting would improve postoperative voluntary muscle strength, nutritional intake and ambulation, decrease postoperative fatigue, anxiety and discomfort, and reduce the endocrine response to surgery. METHODS: Forty-eight patients were included and randomized into three groups to receive 2 x 400 ml of carbohydrate-rich drinks or to fast overnight and allowed only water. Voluntary grip and quadriceps strength, body composition, pulmonary function, VAS-score of eight parameters of wellbeing, muscle biopsies and insulin, glucagon, IGF-1 and free fatty acids were measured before and after the operation. The basic postoperative regimen for all groups were immediate oral nutrition and early enforced mobilization. RESULTS: Significant postoperative decrease in glycogen synthase activity in the muscle biopsies was reduced in the intervention groups, and in combination, the intervention groups had a less reduced quadriceps strength after one week (-10% vs. -16%, NS) and one month (-5% vs. -13%, P < 0.05). Minor changes in the endocrine response to surgery were found without differences between the groups, and there were no differences between the groups in ambulation time, nutritional intake or subjective measures of wellbeing.

The incidence of transient neurologic symptoms (TNS) after spinal anaesthesia in patients undergoing surgery in the supine position. Hyperbaric lidocaine 5% versus hyperbaric bupivacaine 0.5%.

BACKGROUND: The incidence of TNS after spinal anaesthesia is a problem. Especially the use of hyperbaric lidocaine in patients placed in the lithotomy position during surgery has been associated with a high incidence of TNS. The present study was performed to investigate whether TNS is present more frequently in patients undergoing surgery in the supine position with use of hyperbaric lidocaine compared with hyperbaric bupivacaine. METHOD: Seventy patients were included and randomised to receive either hyperbaric lidocaine or hyperbaric bupivacaine. All patients were contacted on the first and third postoperative days by an anaesthesiologist blinded to the local anaesthetic used. The patients were asked about symptoms of TNS, pain not associated with the operation area, and asked to grade the complaints after a verbal analogue score from 0 to 10. RESULTS: We found a total of ten patients who showed signs of TNS. There were nine patients in the lidocaine group (26%) who showed signs of TNS compared to only one patient in the bupivacaine group (3%) (P<0.01). The average score of TNS complaints was 3.5. A total of 13 patients (19%) complained of back pain. There were no significant differences with regard to which local anaesthetic was used. The average score of back pain was 3.3. CONCLUSION: TNS is a significant problem in patients having spinal anaesthesia with hyperbaric lidocaine compared to hyperbaric bupivacaine, both in the supine position. For day-case surgery, TNS would start after dismissal from hospital. The use of hyperbaric lidocaine is therefore questionable, even though these problems are of an order that the majority of patients would still choose spinal anaesthesia for future operations.

[Fiber optic tracheoscopy in percutaneous dilation tracheostomy]. / Fiberoptisk trakeoskopi ved perkutan dilatationstrakeostomi.

Percutaneous dilatational tracheostomy (PDT) was performed on a 64 year-old female multitraumatised patient. Because of difficulties during the procedure, fiberoptic tracheoscopy was carried out and revealed perforation and dilatation of the endotracheal tube. The error was corrected with little loss of time, and no complications were observed after finishing the procedure. We recommend the use of routine fiberoptic tracheoscopy during and after percutaneous dilatational tracheostomy as suggested by Ciaglia et al (2).

alpha-Trinositol prevents increased negativity of interstitial fluid pressure in rat skin and trachea induced by dextran anaphylaxis.

The new anti-inflammatory agent alpha-trinositol (D-myo-inositol-1,2,6-trisphosphate), is suggested to act on the cellular adhesion receptor towards extracellular matrix components, the beta1-integrins, and may therefore represent a novel principle for therapy of the phenomena associated with acute inflammation. Increased negativity of interstitial fluid pressure (p(if)) is a major driving force for the rapid edema formation in trachea and skin associated with dextran anaphylaxis in the rat. We therefore used this experimental model to study the effect of alpha-trinositol in skin and trachea of pentobarbital anesthetized rats. p(if) was measured with sharpened glass capillaries (3-7 microm) connected to a servocontrolled counterpressure system. alpha-Trinositol (10 mg) was given before or after dextran. Circulatory arrest was induced 2 min after i.v. dextran to limit the increased capillary fluid filtration associated with the anaphylactic reaction. This increased filtration will otherwise raise interstitial volume and thereby p(if) and cause an underestimation of a potential increased negativity of p(if). In the trachea, p(if) was 0.0 +/- 1.0 mmHg (S.D.) and -1.4 +/- 0.5 mmHg in controls given saline vehicle and alpha-trinositol (P > 0.05), respectively, and fell to -8.5 +/- 2.7 mmHg after dextran (P < 0.01). alpha-Trinositol given 2 min prior to or after dextran resulted in p(if) of -1.7 +/- 1.2 mmHg (P > 0.05 versus control, P < 0.01 versus dextran) and -4.7 +/- 3.0 mmHg (P < 0.01 versus control and dextran), respectively. In skin, i.v. dextran caused p(if) to fall from -0.6 +/- 0.5 to -4.6 +/- 1.9 mmHg (P < 0.001). When alpha-trinositol was given prior to dextran the corresponding figures were -0.4 +/- 0.8 and -0.9 +/- 1.1 mmHg, respectively (P > 0.05). Subdermal administration of alpha-trinositol after i.v. dextran and circulatory arrest normalized p(if) in concentration of 100, 10 and partly at 1 mg/ml. Thus, alpha-trinositol prevented the increased negativity of p(if) induced by dextran anaphylaxis when administered prior to as well as after dextran showing that the alpha-trinositol also could influence an already started inflammatory reaction.

A novel physiological function for platelet-derived growth factor-BB in rat dermis.

1. The present experiments describe a role for platelet-derived growth factor-BB and cellular adhesion receptors towards extracellular matrix molecules (beta 1-integrins) in control of interstitial fluid pressure (Pif). 2. Pif was measured in rat skin with sharpened glass capillaries (3-7 microns) connected to a servocontrolled counter-pressure system. 3. The collagen and laminin-binding alpha 2 beta 1-integrin is involved in the control of Pif since subdermal injection (5 microliters) of monoclonal hamster anti-rat alpha 2 beta 1-integrin IgG (anti-alpha 2 beta 1) resulted in increased negativity of Pif. Control Pif averaged -0.88 +/- 0.23 mmHg (+/- S.D.) and decreased to -2.50 +/- 0.35 mmHg (P < 0.05) and -3.88 +/- 1.45 mmHg (P < 0.05) at anti-alpha 2 beta 1 concentrations of 0.56 and 1.12 mg ml-1, respectively. 4. The effect of anti-alpha 2 beta 1 was abolished when platelet-derived growth factor-BB (PDGF-BB) (200 ng ml-1) was injected together with anti-alpha 2 beta 1. 5. The time- and dose-responses of PDGF-BB to counteract increased negativity of Pif were studied further using dextran anaphylaxis as an experimental model inducing increased negativity of Pif in skin. Control Pif averaged -0.33 +/- 0.43 mmHg and fell to -4.10 +/- 1.47 mmHg within 10 min after dextran (P < 0.01). Subsequent subdermal injection of PDGF-BB at 200 ng ml-1 normalized Pif in 10-20 min which became -1.37 +/- 1.23 mmHg (P < 0.01 versus dextran, P > 0.05 versus control). PDGF-BB had little or no effect at 50 ng ml-1. PDGF-AA and basic fibroblast growth factor had no effect on Pif. 6. The in vivo function reported for PDGF-BB has not been described previously and provides further evidence for active participation of connective tissue cells in control of Pif by altering tension on extracellular matrix structures.

Is the pencil point spinal needle a better choice in younger patients? A comparison of 24G Sprotte with 27G Quincke needles in an unselected group of general surgical patients below 46 years of age.

Reports have indicated that there are less postoperative complaints after the use of pencil pointed spinal needles. We compared a 24G Sprotte needle with a 27G Quincke needle in a randomised study of 200 healthy patients (49% females), aged 15-46 years. Four patients (2%) reported postdural puncture headache, three with the 24G Sprotte needle and one with the 27G Quincke needle. Thirteen patients (7%) suffered with nonspecific headache, with no significant difference between the two groups. Of the 57 (29%) who reported backpain, a significantly higher proportion had received spinal anaesthesia with the Sprotte needle (OR = 2.06). There was a significantly higher incidence of insufficient blocks after dural puncture with the Sprotte needle. Ease of needle insertion and number of puncture attempts was the same for both needle types.

The anti-inflammatory agent alpha-trinositol exerts its edema-preventing effects through modulation of beta 1 integrin function.

Edema formation in acute inflammation can be induced through lowering of interstitial fluid pressure (Pif) and seems to involve dynamic beta 1 integrin-mediated interactions between dermal cells and extracellular matrix fibers. The present experiments investigate the role of beta 1 integrins in the control of Pif. The anti-inflammatory drug alpha-trinositol (1,2,6-D-myo-inositol trisphosphate) stabilizes Pif in acute inflammation. Pretreatment with 5 mg IV alpha-trinositol in pentobarbital-anesthetized rats inhibited the lowering in Pif and the edema formation induced by subdermal injection of anti-beta 1 integrin IgG. This stabilization of the beta 1 integrin function in vivo was paralleled by effects of alpha-trinositol on contraction of fibroblast-populated three-dimensional collagen lattices in vitro. alpha-Trinositol was additive to the known stimulatory effect of platelet-derived growth factor-BB on the final gel size in the collagen gel contraction assay. Furthermore, alpha-trinositol counteracted the inhibitory effect of anti-beta 1 integrin Fab fragments on collagen gel contraction. Finally, subdermal injection of dibutyryl-cAMP (db-cAMP) induced increased negativity of Pif to the same extent as did anti-beta 1 integrin antibodies, and in vitro db-cAMP reduced the ability of fibroblasts to contract collagen gels. The latter effect was opposed by alpha-trinositol. The data demonstrate that alpha-trinositol modulates beta 1 integrin function and may do so via intracellular pathways in turn affecting the function and/or cell surface expression of beta 1 integrins and suggest that alpha-trinositol can serve as a tool to study integrin function. Furthermore, the data indicate that the collagen contraction assays may provide important information of the control of Pif in vivo.

Increased hyaluronan flux in canine paw lymph is induced by histamine and the histamine-releasing agent compound 48/80.

The present experiments investigated hyaluronan (HYA) flux from skin of pentobarbital anesthetized mongrel dogs when transcapillary fluid flux was increased by local intraarterial injection of histamine (50 micrograms) or Compound 48/80 (C48/80) (100 micrograms) inducing mast cell degranulation. A prenodal lymphatic draining the hindpaw was cannulated and the paw flexed passively at 50 times/min. Grand mean (n = 18) of control lymph flow and HYA concentration was 16 +/- (SD) 14 microliters/min and 8.8 +/- 2.3 micrograms/ml, respectively. Lymph flow increased 11- and 15-fold within 10 min after histamine and C48/80 injection, respectively, and returned to control values after 3 h for histamine while it did not return fully in the C48/80 group. HYA concentration decreased by 30 and 40% during the first hour after histamine and C48/80, respectively, while HYA flux increased 11-15 times control. Control experiments (saline vehicle) showed an unexpected and gradual increase in HYA concentration during the 8-hour experimental period, regardless of unchanged lymph flow. This increase became statistically significant at the end of the experimental period, suggesting either an increased synthesis or increased rate of release of bound HYA from the paw. The present data show that HYA is loosely bound and easily mobilized from the interstitial matrix and that histamine and C48/80 cause a release of bound HYA from the interstitium. An increase in HYA concentration towards the end of the 8-hour experimental period most likely represents an increased synthesis of HYA.

Intrathecal anaesthesia in patients under 45 years: incidence of postdural puncture symptoms after spinal anaesthesia with 27G needles.

Postoperative headache and backpain has limited the use of intrathecal anaesthesia in younger patients (15-45 years). We studied postoperative complaints among 133 healthy young patients (mean age 30.0 years, 47% females) who received spinal anaesthesia with a 27G needle. Postoperatively, 5 patients (4%) complained of postdural puncture headache (PDPH), 18 (14%) reported nonspecific headache, while 27 (20%) suffered from backpain. PDPH was not related to sex, age, day-care surgery, number of puncture attempts, or obstetric procedures. Backpain was significantly more common among females, and among in-patients. One hundred and sixteen patients (87%) would accept spinal anaesthesia if they were to undergo the same surgical procedure again. Compared to other studies, we find the incidence of postanaesthetic complaints to be acceptable, also among day-care patients. The PDPH seemed to be lightly incapacitating, and only one patient required blood patching.

Interstitial fluid pressure in rat skin becomes more negative in the initial phase of carrageenan-induced edema.

Increased negativity of the interstitial fluid pressure (P(i)) in the skin is a major driving force to create the initial edema in burn injury and, to a smaller degree, in the edemas formed in the initial phase of the chemical irritation induced by xylene and the anaphylactic reaction induced by dextran. The phenomenon of increased negativity of P(i) is the basis for stating that the connective tissues become 'active' in fluid exchange since the increased negativity of P(i) will be a major driving pressure for transcapillary fluid flux under these circumstances. The present study was performed to investigate if increased negativity of P(i) contributes to formation of the inflammatory edema induced by carrageenan in the rat paw skin, since carrageenan inflammation involves mechanisms principally different from burn injury, xylene and dextran anaphylaxis. Control P(i) averaged -0.42 mmHg +/- 0.12 (SE). Injection of normal saline increased P(i) by 1.4 mmHg in 6-10 min. In animals with intact circulation P(i) decreased to -4.75 +/- 0.60 mmHg in the same time period following subdermal injection of carrageenan and later returned to control values by 21-30 min. To prevent the edema formation which will increase P(i) and potentially lead to underestimation of an increased negativity in P(i), measurements were also performed after circulatory arrest when there will be no transcapillary fluid flux in response to the inflammatory agent. In this group P(i) fell to -4.10 +/- 0.71 mmHg (11-20 min) and remained at -2.75 to -4.10 mmHg throughout the observation period. Premedication with aprotinin (protease inhibitor), indomethacin (prostaglandin inhibitor) and 1,2 benzopyrone did not change the increased negativity in P(i) induced by carrageenan. In conclusion, the increased negativity of P(i) contributes to the formation of edema also in carrageenan-induced inflammation and is an effect of carrageenan which has previously not been described.

Blockade of beta 1-integrins in skin causes edema through lowering of interstitial fluid pressure.

The increased capillary fluid filtration required to rapidly create edema in acute inflammation can be generated by increased negativity of the interstitial fluid pressure (Pif). This observation suggests that connective tissues can "actively" enhance capillary fluid filtration. We now show that in vivo blockade of beta 1-integrin adhesion receptors in rat skin causes local edema concomitant with increased negativity of Pif. Experiments were performed on the dorsal side of the hind paw, and Pif was measured with sharpened glass capillaries (tip diameter, 3-7 microns) connected to a servo-controlled counterpressure system. Measurements were made after circulatory arrest had been induced with intracardiac potassium chloride in pentobarbital anesthesia. This procedure prevents the vascular phenomena of increased fluid and protein flux leading to edema formation, which in turn can increase Pif and therefore potentially mask an increased negativity of Pif. Control Pif averaged -0.58 +/- 0.81 (mean +/- SD) mm Hg (n = 37). Subdermal injection of 5 microliters monospecific rabbit anti-rat integrin beta 1-subunit immunoglobulin G caused increased negativity of Pif to average values between -4 and -6 mm Hg within 10 minutes after injection. Subdermal injection of 0.9% NaCl, preimmune immunoglobulin G, rat anti-fibronectin, and peptides with Arg-Gly-Asp and Arg-Gly-Glu sequences did not change Pif significantly. In another series of experiments, 5 microliters anti-beta 1 integrin immunoglobulin G was injected subdermally in rats with intact circulation and resulted in an increase in total tissue water corresponding to a doubling of the interstitial fluid volume in 10 minutes (p less than 0.05).(ABSTRACT TRUNCATED AT 250 WORDS)

Increased occurrence of exercise-induced silent ischemia after treatment with aspirin in patients admitted for suspected acute myocardial infarction.

Patients admitted for suspected acute myocardial infarction within 6 hours (mean 3 hours 42 minutes) after onset of symptoms were randomised to double-blind treatment with low-dose oral aspirin or placebo. Early exercise ischemic responses, exercise capacity and resting left ventricular ejection fraction (radionuclide ventriculography) were estimated in 77 survivors 2-4 weeks later. Exercise performance and ejection fraction in patients with confirmed acute myocardial infarction were equal in the two groups. During exercise, patients treated with aspirin had significantly more silent ischemia (ST depression without chest pain) compared to placebo (28% versus 6%; P = 0.015). The occurrence of positive exercise tests (chest pain or ST-segment depression), however, was similar in the two groups. The results indicate that the administration of aspirin early after acute myocardial infarction increases the occurrence of silent ischemia but has no effect on left ventricular function.

Increased negativity of interstitial fluid pressure during the onset stage of inflammatory edema in rat skin.

Interstitial fluid pressure (Pif) was measured in skin of pentobarbital-anesthetized rats during anaphylaxis toward dextran and after subdermal injection of histamine by using sharpened glass capillaries (tip diam 5-7 microns) connected to a servo-controlled counterpressure system. Control Pif averaged -1.5 mmHg (SD = 1.0). After intravenous dextran Pif in the rat paw fell transiently to -3 mmHg up to 20 min and thereafter increased to +1-2 mmHg when edema had developed. To study the full magnitude of the increased negativity of Pif, circulatory arrest was induced 1 min after dextran injection. This procedure prevents accumulation of edema that will cause underestimation of Pif. In this group Pif fell to about -10 mmHg in 20 min and remained at this level throughout the observation period of 90 min. Subdermal injection of 1-10 micrograms histamine in 10 microliters saline reduced Pif to about -6 mmHg within 5 min after injection. Injection of 10 microliters saline increased Pif by +2 mmHg. Indomethacin or cyproheptadine did not alter the response in the above situations. The increased negativity in Pif of about 6-8 mmHg will add directly to normal transcapillary net filtration pressure of 0.5 mmHg and increase the latter pressure 10-20 times.

[The laryngeal mask: a new tool for creating a free airway]. / Larynxmasken: et nytt hjelpemiddel for å holde frie luftveier.

The laryngeal mask airway represents a new concept in airway management of anesthetised patients. It is inserted without any instruments and is very easy to use, even for unskilled personnel. It reduces the problem of waste gas pollution in the operating theater, and it leaves the hands of the anaesthetist free for record keeping and monitoring. It is well tolerated by the patients. In this article the technique is presented together with indications and limitations. Based on our experience we find the laryngeal mask airway an excellent alternative to the face mask, even in procedures of short duration.

Increased negativity of interstitial fluid pressure contributes to development of oedema in rat skin following application of xylene.

Intradermal interstitial fluid pressure (Pi) has been studied in rat skin during formation of inflammatory oedema caused by application of xylene. Pi was measured with sharpened micropipettes connected to a servocontrolled counter-pressure system. Control Pi averaged -1.3 +/- 0.6 (SD) mmHg. Following xylene application Pi decreased to -5.0 mmHg after 5 min and then increased to stabilize at about 0 mmHg at 45-60 min and later. When the transvascular fluid shifts accompanying the inflammatory reaction were prevented by inducing circulatory arrest prior to xylene application, Pi fell to -7.5 mmHg within 5 min and remained at this level throughout the observation period of 90 min. Aprotinin in large doses (80,000 KIE kg-1) before xylene application reduced the fall in Pi, whereas indomethacin had no effect. The increased negativity in Pi will add directly to a normal transcapillary net filtration pressure of about 0.5 mmHg, resulting in a 10- to 20-fold increase in this pressure. The present experiments therefore suggest that the interstitium plays an active role in oedema formation in the initial phase of xylene-induced inflammation in rat skin through the development of an increased negativity of Pi.