Role of Roux-en-Y feeding jejunostomy for patients with acute head injury

The total care of a patient with severe head injury is challenging and may extend for weeks, months, or even years. A major challenge of this care includes nutritional support: swallowing is impaired, aspiration accompanies gastric tube feeding, parenteral nutrition is limited to short term in hospital care, and needle jejunostomy or transabdominal jejunostomy are prone to inadvertent removal. The role of Roux-en-Y feeding jejunostomy was evaluated in 13 patients with acute head injury. Procedure related complications include prolapse of the ostomy (1 patient) and stoma-ischemia requiring revision (1 patient). The effect of ostomy tube feedings on gastric acid secretions was studied in five patients, and no significant change was noted when saline feeding was compared to blenderized diet feeding. Blenderized diet feedings were advanced gradually, and antidiarrheals were added as needed once gastrointestinal function returned. In conclusion, Roux-en-Y feeding jejunostomy provides an attractive, safe method for long-term enteral nutrition in the head injury patient. Easy replacement of the feeding tube facilitates nursing care, and the threat of acid-induced stress gastric bleeding is not enhanced.

Significance of hypocalcemia following hypovolemic shock

Changes in calcium levels during and after resuscitation from severe shock were studied in 22 seriously injured patients who received an average of 21 blood transfusions and 26 mEq supplemental calcium. Total serum proteins (TSP), serum albumin (SA), total calcium (TC), and ionized calcium (CA++), were studied intraoperatively after the tenth transfusion and postoperatively at 5 hours, 15 hours, day 2, day 4, and during convalescence (day 25). The intraoperative TSP fell to 3.7 gm%; the TC and Ca++ fell to 7.2 mg% and 1.4 mEq/L. The TSP and SA remained low throughout day 4 (4.8 and 2.6 gm%); the TC was also low on day 4 (7.5 mg%), whereas the Ca++ rose to normal (2.1 mEq/L) by day 2. The severity of hypocalcemia paralleled the hypoproteinemia, the number of transfusions given during resuscitation, and the duration of shock; paradoxically, hypocalcemia correlated inversely with Ca++ supplementation of blood transfusions during resuscitation, suggesting increased extravascular Ca++ flux with more severe shock. Further studies in comparably injured patients are needed to identify the concomitant responses of the calcium homeostatic factors such as parathormone in order to help identify the optimal role of calcium manipulation during resuscitation from hypovolemic shock.

Impaired renal concentrating ability during resuscitation from shock

Renal concentrating ability was tested in 20 severely injured patients who received an average of 21 blood transfusions. Timed measurements were made of urine output (UO), and the clearance of creatinine glomerular filtration rate (GFR), sodium (CNa), osmols (COsm), and free water (CH2O) during operation and were repeated at 12 and 18 hours postoperatively, and on postoperative days 2 and 4. During operation the GFR was markedly reduced (36 mL/min), while UO, CNa, and COsm were all markedly increased (8.5, 5.9, and 8.1 mL/min, respectively). The CH2O was positive (0.4 mL/min). Following operation the rate of renal excretion of water and solutes was still high: UO, 4.0 mL/min; CNa, 4.3 mL/min; and COsm, 6.0 mL/min. Five hours postoperatively the CH2O had returned to normal. By day 2 the excretion rate of water had returned closer to normal: UO, 2.1 mL/min; CNa, 2.6 mL/min; COsm, 4.0 mL/min; the CH2O was normal. Subsequent study results were normal. These data demonstrate a renal concentrating impairment during and following operation. An osmotic diuresis, transient tubular ischemia, a washout of interstitial inner medullary osmoles, or some cryptic factor may be causative.

Mycotic aneurysm. New concepts in therapy

Fifty patients were treated for 52 mycotic aneurysms secondary to intravenous drug abuse. An initial misdiagnosis of cellulitis or abscess in 17% of the patients was corrected after arteriography or bleeding following operative drainage. There was no ischemia following ligation and excision of aneurysms of the radial, brachial, external iliac, deep femoral, and superficial femoral arteries. Excision of the common femoral artery in four patients and femoral bifurcation in 25 led to marked morbidity in 28 patients without simultaneous revascularization. Ischemia occurred in 53% of these patients; it was mild in 21% with claudication only. Severe, limb-threatening ischemia occurred in 32% and led to amputation in 21%. Six patients underwent artificial bypass, including one for absent back-bleeding at the time of ligation, four for immediate severe ischemia, and one for late ischemia. Two infected grafts were removed; another became thrombotic. Cultures were positive for 73% of aneurysms and blood of 46% of the patients.

The fluid problem in the critically ill

Fluid therapy in the critically ill patient must be adjusted to accommodate continuing changes in the plasma volume, interstitial space, and intracellular space. During and after hemorrhagic shock, replacement of crystalloid is needed to replenish the plasma and interstitial spaces during operation and then interstitial and intracellular spaces after operation. Severe sepsis leads to a more pronounced expansion of the interstitial space than that of hemorrhagic shock. Continuing therapy after both hemorrhage and sepsis should be directed toward maintaining effective plasma volume and levels of hemoglobin while the interstitial and intracellular spaces return to normal. Concomitantly, effective circulatory volume must be guided by continuing changes in cardiac, pulmonary, and renal function.

The value of prophylactic antibiotic therapy during endoscopic retrograde cholangiopancreatography

Prophylactic antibiotics have been recommended during endoscopic retrograde cholangiopancreatography in the hope of preventing septic complications. However, their effectiveness has not been proved clinically. In this study, the ability of the pancreas to secrete ampicillin, gentamicin and clindamycin after a single intravenous dosage, given prior to endoscopic retrograde cholangiopancreatography in 12 patients, was examined. Simultaneously obtained serum antibiotic levels were within the expected therapeutic range. In contrast, pancreatic ductal levels of ampicillin and gentamicin were too low to be measured in most patients. Most patients had measurable clindamycin levels which ranged from 12.0 to 3.1 micrograms per milliliter in seven patients and was 8.0 micrograms per milliliter in one patient. Parenteral prophylactic antibiotic coverage for endoscopic retrograde cholangiopancreatography is ineffective with gentamicin or ampicillin. Clindamycin may be of value, but it needs further study at higher dosages.

Anergy and altered lymphocyte function in the injured patient

The failure of the host-defense mechanism following trauma has been recognized, but the site of the deficit is unknown. The immunologic competence of 16 patients, including 15 who had minor injuries and required less than 4 transfusions, was prospectively studied with skin testing, leukocyte counts, and protein electrophoresis. The stimulation ratio (SR) and index of lymphocyte response to phytohemagglutanin (PHA) were measured as the ratio of thymidine uptake in stimulated cells to that of resting cells in both pooled normal serum as well as each patient's serum. Six patients, including the one patient with major injury, had no response to any of the four skin test antigens and were considered anergic (AN). Ten patients with minor trauma responded to at least one of the antigen skin tests and were considered nonanergic (NA). The anergic patients had significantly more shock and more blood transfusions and had significantly lower serum albumin levels. There was no statistical difference between the anergic and nonanergic groups in leukocyte count, absolute lymphocyte count, gamma-globulin fraction, or age. The average PHA Stimulation Ratio of patient lymphocytes in patient serum was significantly higher than the PHA Stimulation Ratio of patient lymphocytes in control serum. This suggests that the lymphocytes of injured patients respond in a greater magnitude when bathed in autologous-serum than when bathed in control pooled serum. Furthermore, the autologous serum did not inhibit the PHA response of control or normal lymphocytes. The presence of an enhancing factor in injured patients' sera or an absence of a supressor substance may be responsible for this phenomenon.

Altered coagulation protein content after albumin resuscitation

Previous studies showed that the random addition of supplemental albumin to a resuscitation regimen of blood, salt, and frozen plasma caused a significant (p = less than 0.05) fall in fibrinogen clotting activity (FC) and a rise in prothrombin times (PT) in seriously injured patients; the partial thromboplastin times (PTT) were insignificantly prolonged. Based upon these findings, frozen plasma samples, prospectively collected in 41 non-albumin patients and 35 albumin patients, were analyzed immunologically, in duplicate, for protein content of coagulation factor VIII (VIIIAg), prothrombin (IIAg), fibrinogen (FAg), antithrombin III (ATAg), and fibrin(ogen) split products (FSP). Supplemental albumin resuscitation was associated with a significant fall in FAg (83 +/- 9 versus 124 +/- 10 SE mg/dl), VIIIAg (97 +/- 13 versus 127 +/- 135 SE %), IIAg (54 +/- 3 versus 80 +/- 4 SE %), and ATAg (14 +/- 0.8 +/- 19 +/- 0.8 SE mg%) with no significant changes in FSP. FSP, however, were more than 20 micrograms/ml in 13 of 41 nonalbumin patients versus four of 35 albumin patients (X2 = 4.5, p less than 0.05). Reduced coagulation activity following albumin supplementation seems partly caused by a decrease of coagulation protein content; increased fibrinolysis in the albumin patients is not the cause. Decreased coagulation protein content parallels the fall in coagulation activity and the need for postresuscitation blood transfusions. The role of reduced coagulation synthesis in these changes needs further study.

Primary hemostasis after massive transfusion for injury

Primary hemostasis, the formation of a platelet plug, was studied in 22 injured patients receiving an average of 21 transfusions during the operation for control of bleeding. The storage age of the blood averaged 15 days; no platelet transfusions were given. Platelet counts (PLT) and bleeding time (BT) were studied intraoperatively; postoperatively at 6 hours, 25 hours, day 2, day 4; and during convalescence (8 days to 3 months). Serial PLT and BT levels were correlated with the number of transfusions and age of blood. During operation, the PLT fell to 109,000/mm3 and the BT was greater than 15 minutes. Thrombocytopenia did not correlate with the number of transfusions or age of blood. The PLT averaged 106,000/mm3 at 6 hours and then fell significantly to 73,000/mm3 at 15 hours and to 76,000/mm3 on day 2. The PLT rose significantly to 110,000/mm3 by day four and increased to supernormal levels by convalescence. The BT remained elevated at 6 hours, 15 hours, day 2, and day 4, and declined to normal by convalescence. Thrombocytopenia and prolonged BT after massive transfusion for injury indicate platelet dysfunction which may protect against disseminated intravascular coagulation. Correction of the thrombocytopenia should be reserved for patients with bleeding.

Reduced oncotic pressure after shock: a proposed mechanism

Hemorrhagic shock plus resuscitation leads to expansion of both interstitial fluid space (IFS) and intracellular fluid (ICF). THe IFS expansion is thought to reflect reduced serum colloid oncotic pressure (COP) from capillary leak of serum albumin (SA). This hypothesis was analyzed in 138 injured patients who received an average of 13.6 units of blood and 10.7 L of saline solution for correction of shock. Plasma volume (PV), extracellular fluid (ECF) by inulin space technique, SA level, percent albumin leak (PAL), total intravascular albumin content (TIAC), COP, IFS, weight gain, and time of injury were analyzed. Patients studied during the postoperative fluid sequestration phase had a low to normal PV (2.9 L) with a high IFS (15.2 L) and a low PV-ECF ratio (0.15). Although the SA level (2.8 g/dL), TIAC (81 g), and COP (11.9 mOsm/L) were low, the PAL was also low (6.0%/hr; normal, 7%/hr). Patients studied during the fluid mobilization phase had a high PV (3.2 L) and normal PV-ECF ratio (0.21), whereas the SA level (3.0 g/dL), TIAC (103 g), and COP (14.1 mOsm/L) remained low. Despite these findings, the PAL was high (8.0%/hr). The fall in the PV-ECF ratio during the postresuscitation fluid sequestration phas is associated with but not due to reduced TIAC and COP, both of which remained low after the PV-ECF ratio has normalized during the mobilization phase. Contrary to prior reports, albumin leak is not responsible for reduced COP, TIAC, and postresuscitation weight gain. Presumably, the low COP results from decreased reentry of albumin into the plasma volume due to entrapment within the IFS matrix.

Immunoglobulin changes after varied resuscitation regimens

Prior studies showed that albumin supplementation of the resuscitation for hypovolemic shock caused an increase in serum albumin but a fall in serum globulins; immunoglobulins were not measured. Using frozen sera, immunoglobulins (IgG, IgM, IgA, IgD) were measured in 184 severely injured patients including 40 patients prospectively randomized for supplemental steroid therapy and 46 patients prospectively randomized for supplemental albumin therapy. The remaining patients served as the control patients. Compared to normal, the control patients had a significant reduction in IgG, IgM, and IgA. This reduction was associated with a fall in total serum proteins and serum albumin concentrations. Supplemental albumin resulted in an increase in the serum albumin concentration but a reciprocal fall in the serum globulin fraction and in IgG, IgM, and IgA. The serum proteins and immunoglobulins in the steroid patients were significantly below normal but similar to that seen in the control patients. The reduction in serum proteins and immunoglobulins in the control and steroid patients paralleled the shock time (systolic pressure below 80 mm Hg) and the amount of plasma given during resuscitation. These correlations were not significant in the albumin- supplemented patients. These data indicate that albumin supplementation alters the normal immunoglobulin response to shock. These changes in the postinjury recovery period need further study.

Pulmonary response of massive steroids in seriously injured patients

The effects of massive steroids on pulmonary function after hypovolemic shock were tested in 114 injured patients who received an average of 13 transfusions, 760 ml plasma, and 11.7 L crystalloid solution; by random selection, 54 patients received methylprednisolone (1 g in operating room plus 3,578 mg average during the next three days). The patients who received steroids had a significant increase in central venous pressure and a decrease in arterial oxygen tension (PaO2) compared with control patients. The inspired oxygen concentration was similar for both groups; the FiO2/PO2, therefore, was significantly deranged (P = less than 0.05) in steroid patients (0.45 +/- 0.05 SE vs 0.37 +/- 0.02 SE). The patients who received steroids has an insignificantly increased pulmonary shunt (25 vs 22%), number of days on a volume ventilator (5.1 vs 3.0 days), and number of deaths (seven vs two), Massive steroids neither prevent nor ameliorate pulmonary failure after shock; indeed, steroids may aggravate pulmonary failure after shock.

Altered interstitial fluid space dynamics and postresuscitation hypertension

Hypertensin occurred 24 to 48 hours after resuscitation in 35 of 86 injured patients, who had combined systolic and diastolic hypertensin (150/100 mmHg) for six or more consecutive hours. Plasma volume (PV), RBC volume, extracellular fluid (ECF) volume by the inulin dilution technique, renal plasma flow, glomerular filtration rate, and peripheral renin levels were measured in hypertensive and nonhypertensive patients an average of 40 hours after injury. The hypertensive patients had an average mean arterial pressure (MAP) of 114 mmHg, compared with 95 mmHg in the nonhypertensive patients. The RBC volume and ECF were comparable for both groups, whereas PV was increased in the hypertensive patients (3.6 L vs 3.3 L). Calculated interstitial fluid space (IFS) volume was greater in the nonhypertensive patients, as was the ratio PV/IFS. The MAP in both groups correlated directly with PV/IFS and serum albumin concentrations, and inversely with peripheral renin concentrations. This suggests that postresuscitative hypertension is not due to fluid overload but rather to the fluid maldistribution related to altered IFS compliance as reflected by the increased PV/IFS.

The cardiac effect of altered calcium homeostasis after albumin resuscitation

Supplemental albumin added to a standard non-albumin resuscitation regimen has been shown to significantly impair heartwork in seriously injured patients. The role of calcium dynamics in this myocardial depression was analyzed in 94 injured patients who were in shock for an average of 32 minutes, received an average of 14.5 transfusions, 9.2 L crystalloid, 0.9 L plasma, and 20.9 mEq calcium prior to the end of operation. By random selection, 44 patients received an average of 31 gms of albumin during operation, 207 gms during the early postoperative period (mean = 30 hrs) of extravascular fluid sequestration, and 402 gm during the mobilization period. The albumin resuscitated patients had normal total protein and serum albumin levels and higher total calcium (TC) levels, however, they had a significantly lower Ca++ and Ca++/TC. The accumulative slope for heartwork/filling pressure was significantly depressed in albumin patients as was the mean work unit/filling pressure index. The level of Ca++ and the Ca++/TC ratio correlated directly with the calculated work unit index in both the albumin and non-albumin patients. This suggests that a supplemental albumin binds serum Ca++ causing an increase in TC but a reduction in Ca++ and Ca++/TC. The fall in Ca++ and Ca++/TC seems responsible, in part, for heart failure and pulmonary edema in albumin resuscitated patients.

Postresuscitative hypertension: a reappraisal

New concepts of cause and therapy for postresuscitative hypertension (PRH) were evaluated in four patients with PRH. Each patient had severe injury and shock, and received an average of 28.3 transfusions, 15.4 L of electrolyte solution, and 1.4 L of plasma by the end of surgery for control of bleeding. Near the end of the sequestration phase, PRH developed. In two patients, PRH (190/100 mm Hg and 180/90 mm Hg) responded to previously recommended therapy; blood pressure fell to about 135/90 mm Hg. Shortly thereafter, bradycardia developed and both patients died. In the latter two patients, PRH (205/115 mm Hg and 150/120 mmHg) was treated less aggressively, maintaining intravenous fluids to keep urine output at a minimum of 50 mL/hr. Postresuscitative hypertension persisted for five and six days as both patients improved, continued to mobilize sequestered fluid, and maintained good organ perfusion. Postresuscitative hypertension may be a cell-mediated protective response to a need for increased capillary hydrostatic pressure to facilitate mitochondrial oxygenation. Fluid replacement should be guided by careful monitoring of cardiac, pulmonary, and renal function.

The natural history following venous ligation for civilian injuries

Contrary to prevailing views, documentation of major morbidity after venous ligation for trauma is sparse. Forty-six patients had ligation for injury to the major veins of the lower extremities, namely, vena cava (six), external iliac (five), common iliac (three), common femoral (eight), profunda femoris (six), superficial femoral (thirteen), and popliteal (five). The post-ligation management included: 1) early and extensive fasciotomy when indicated, 2) initial strict bed rest with elevation until edema free, 3) trial ambulation for 2 hours, 4) added elevation if trial ambulation leads to recurrent edema. Forty patients were discharged edema free, and six patients had mild edema. Followup examination in 39 patients revealed no edema in 30 patients, mild edema requiring no treatment in one patient, and moderate edema requiring support hose in eight patients. No severe or massive edema causing ulceration or ischemia occurred. Based on these findings, primary venous interruption is a safe, quick and effective means of managing venous injury whenever the extent of injury or the severity of associated injuries make primary repair hazardous.