WOUND HEALING AND WOUND INFECTION What Surgeons and Anesthesiologists Can Do Thomas K. Hunt, MD, and Harriet Williams Hopf, MD
The sixteenth-century French surgeon Ambroise Pare is famous for saying, “I dressed the wound. God healed it.’’ The implication was that wounds heal by a mysterious, incomprehensible force as long as local care is adequate. This attitude, unfortunately, has endured. In truth, it is only a quaint reminder of the ignorance that lasted well into the present century. Times change! This article is a brief discussion of how surgeons (with the help of anesthesiologists) can heal wounds. A BRIEF TOUR THROUGH INJURY AND REPAIR
Healing is a sequence. It starts with injury (e.g., thermal, knife, antigen, antibody) that perturbs the local environment, mainly by damaging its microcirculation. Platelets aggregate and release a variety of chemoattractants and growth factors from their granules. The resultant coagulation occludes vessels and prevents exsanguination but also widens the area of impaired circulation. Fibrin contributes growth factorlike sequences and chemoattractants. The microcirculation is then further perturbed by a flush of bradykinin and complement, and histamine release by mast cells. Inflammatory cells, including first leukocytes and then macroThis article was supported by NIH grant no. GM-27345.
From the Departments of Anesthesia (HWH) and Surgery (TKH,HWH), University of California, San Francisco, California
SURGICAL CLINICS OF NORTH AMERICA VOLUME 77 * NUMBER 3 JUNE 1997
phages, are led into the wound and activated by endothelial integrins, fibrin, lactate, hypoxia, foreign bodies, and some growth factors. In turn, macrophages (and lymphocytes) produce more growth factors, including, most importantly, insulin-like growth factor (IGF-I),leukocyte growth factor, interleukins (1L)-1 and 2, transforming growth factor (TGF), and vascular endothelial growth factor (VEGF).14Inflammatory cells also increase demand for oxygen. Owing to hypoxia and constitutive (aerobic) lactate production by inflammatory cells and fibroblasts, lactate accumulates to 5 to 15 mh4.53Lactate in turn stimulates collagen 38 Anti-inflammatory steroids suppress insecretion and angiogenesi~.~~, flammation and therefore suppress healing. Angiogenesis, essential to all but the most minor wounds, involves growth and movement of endothelial cells. Platelets release a preliminary wave of angiogenic factors, including platelet derived growth factor (PDGF), TGF, IGF-I, and others. A second wave comes from fibroblast growth factor (FGF), which is released from its normal binding sites on connective tissue molecules. VEGF, contributed largely by macrophages stimulated by fibrinopeptides, by hypoxia, and most importantly by the accumulated lactate, constitutes a third and dominant wave.61The capillary endothelial response to angiogenic agents (i.e., migration into the wound, tubule formation, and connecting to sources of blood flow) requires oxygen at a partial pressure (Po2)that is high enough to make this step critically dependent upon blood perfusion and arterial PO^^^ Fibroblasts replicate and migrate mainly in response to growth factors and chemoattractants.The extent of fibroplasia is proportional to oxygen tension. Once in the wound, fibroblasts adopt a collagen synthetic phenotype under the influence of growth factors and the high lactate concentration. Lactate and some growth factors regulate collagen mRNA synthesis and the activity of oxygenases that post-translationally hydroxylate proline and lysine residues. Hydroxylation allows collagen peptides to aggregate into triple helices. The triple helical structure is necessary to allow release from the cell and confers tissue strength when these units condense into collagen fibers. The activity of the hydroxylases depends critically upon vitamin C and tissue oxygen tension.56 Wound strength, which results from collagen deposition, is therefore highly vulnerable to respiratory and perfusion variables. Epithelial cells move and replicate in response to growth factors and oxygen tension, and epithelization occurs most rapidly in hydrated, well-oxygenated tissues.49 Reparative cells assemble in a characteristic order (Fig. 1).Figure 1 also illustrates the signaling processes of repair. This healing "module," led by macrophages (which produce chemoattractants and growth factors), progressively replaces the wound space. The oxygen and lactate gradients are shown as they have been measured in a particularly powerful model system, the rabbit ear chamber.64 Disruption of the normal skin barrier requires that wounds have the ability to clear foreign material and resist infection. Nonspecific phagocytosis and intracellular killing are the major pathways activated
WOUND HEALING AND WOUND INFECTION
Figure 1. The wound module. Cross-section of the wound module in a rabbit ear chamber (diagram of chamber is in left upper corner). Note that Po,, depicted graphically above the cross-section, is highest next to the vessels, with a gradient down to zero at the wound edge. Note also the lactate gradient, high in the dead space and lower (but still above plasma) toward the vasculature. (From Silver IA: The physiology of wound healing. In Hunt TK, Dunphy JE (eds): Fundamentals of Wound Management. New York, Appleton-CenturyCrofts, 1980, p 30; with permission.)
in wounds. Conversion of oxygen to superoxide in phagocytic vacuoles is responsible for the first step in nonspecific intracellular killing? so that resistance to infection depends critically on perfusion and arterial oxygen tension (Fig. 2). All these steps are influenced by nutrition. Repair and remodeling require a sufficient supply of precursors and cofactors, as well as sufficient energy supply to drive the processes. The functions of vitamin C are perhaps the best understood. Vitamin C is required for many processes critical to healing, including hydroxylation of collagen, IGF-I production, and leukocyte function. Other factors identified as particularly important for wound healing are vitamins A, B, D (in the case of bone healing), and possibly E; zinc; protein (particularly sulfated proteins and arginine); and calories.31In general, specific nutritional requirements are not well delineated. Wounds shrink in the late phases. As fibroblasts move, they pull
%2 H O W
IPhagosorneFigure 2. Superoxide and other oxidant production within the phagosome. SOD = superoxide dismutase; MP = myeloperoxidase.
collagen fibers together. This process shortens scars and shrinks open wounds and is called contraction. It is influenced by growth factors and inhibited by anti-inflammatory steroids. PRINCIPLES OF WOUND CARE
Complications of wounds include failure to heal, infection, and excessive scarring or shrinkage (contracture). Rapid repair has the least potential for infection and excess scarring. Surgical goals, therefore, are to ensure rapid tissue synthesis, minimize contamination, maximize immune mechanisms, and provide protection and "guidance." These goals are best met by maximizing blood perfusion, oxygen supply, and nutrition and protecting against reinjury. Rapid perfusion of injured tissue is probably the most important goal of wound care. Ischemic tissue is easily infectable and heals poorly if at all. The converse is equally true. Healing and resistance to infection become better and better as local perfusion and oxygenation improve. Wounds in the anus, for instance, where blood flow is exceptionally great, rarely become infected and with few exceptions heal uneventfully and without excess scar despite massive contamination. Common surgical expectations are that wounds of the extremities and trunk heal more slowly than those of the face and anus. These differences seem due to variations in blood perfusion, and, as a first approximation, the degree of oxygenation, that is, the local tissue Po,. Wound tissue oxygenation depends upon vascular anatomy, vasomotor control, and arterial Po2. Peripheral vasoconstriction generally limits wound perfusion and Po, before blood pressure and cardiac output fall. Clearly, wound care
WOUND HEALING AND WOUND INFECTION
demands attention to local blood supply, respiratory variables, and cardiac output. The ultimate objective is to maintain a high PO, in the injured tissue. This is not necessarily easy to do and certainly requires more than just giving supplemental oxygen. When perfusion is poor, breathing oxygen has little or no effect on PO, in tissue.20 The emphasis on tissue oxygen tension has been difficult for many to understand because standard teaching emphasizes that oxygen delivery depends more on the amount of oxygen that can be carried by the hemoglobin, that is, its oxygen content, than on arterial Po,. Although this is true of working muscle, it is not true of wound healing. In muscle, where intercapillary distances are small and oxygen consumption is high, hemoglobin has the primary effect of ensuring support of ATP production. Thus, production continues and sustains muscle function even at extremely low local oxygen tensions. In wounds, on the other hand, intercapillary distances are large and oxygen consumption is relatively low. In uninjured subcutaneous tissue, intercapillary distance is greater than in muscle and other highly perfused organs. When the microvasculature is damaged, diffusion distances increase tenfold. Peripheral vasoconstriction also increases diffusion distance. The driving force of diffusion is partial pressure. Hence, a higher Po, is needed to force oxygen into injured and healing tissues, particularly in subcutaneous tissue, fascia, tendon, and bone-the slowly healing tissues. Although oxygen consumption is relatively low in wounds, it is consumed by processes that require a high oxygen tension. Inflammatory cells use little oxygen for respiration, producing energy largely via the hexose-monophosphate shunt. Most of the oxygen consumed in wounds is used for (1) oxidant radical production (bacterial killing), a mainstay of immunity to bacterial infections due to Staphylococcus, Escherichiu coli, Klebsiellu, and other frequent wound infectors; (2) collagen synthesis; and (3) epithelization. The rate constants for oxygen for these components of repair all fall within the physiologic range of 25 to 100 mm Hg. This means that the rate at which repair proceeds varies according to tissue Po, from zero to at least 250 mm Hg.l, 8, 15, 18, 34, 49, 52 Fibroblast replication is optimal at a Po1 of about 40 to 60 mm Hg, higher by some estimates. Any degree of hypoxia impairs immunity and repair. Animal and human measurements clearly show that high oxygen tensions can be reached in wounds but only if perfusion is rapid and arterial Po, is high.20,63 Wound Po, can be elevated even to levels over 100 mm Hg (full saturation of hemoglobin) because (1) subcutaneous tissue serves a reservoir function, so flow is normally in excess of nutritional needs, and (2) wound cells consume relatively little oxygen, about 0.7 mL/100 mL of blood flow at normal perfusion rate.l6rZ7At high levels of Pao, this small volume can be carried by plasma alone. Contrary to popular belief, therefore, oxygen-carrying capacity (i.e., hemoglobin concentration) is not particularly important to wound healing provided that perfusion is normal. Wound Po, and collagen synthesis are easily maintained in individuals who have hematocrit levels as low
as 15%to 187'0, provided that they have healthy hearts and blood vessels and are not vasoconstricted. Peripheral vasoconstriction, which results from central sympathetic control of subcutaneous vascular tone, is probably the most frequent and the most clinically important impediment to maintenance of oxygen supply (PO,) in wounds. Subcutaneous tissue is particularly vulnerable to vasoconstriction because blood flow has little local regulation other than local heating57,63 because subcutaneous tissue is both a reservoir to maintain central volume and a major site of thermoregulation. Sympathetically induced peripheral vasoconstriction is stimulated by cold, pain, fear, and blood volume deficit,l0,22 and by various medications including nicotine, beta-adrenergic antagonists, and alpha, agonists, all commonly present in surgical situations. Some regions-the extremities and gut, for instance-have many adrenergic receptors in their microvessels and are subject to extreme vasoconstriction. In most cases, vasomotor tone is under the surgeon's control. Vasodilation enhances the chance of uncomplicated healing. Decreased infection rates due to prevention or correction of hypothermiaa and blood volume deficitsz4are particularly well documented. These effects are clearly mediated, at least in major part, by raising the Po, in the injured tissue. Smoking and some medications also are clinically significant vasoconstrictors. Smoking one cigarette lowers wound and tissue Po, in normal volunteers by 30% for the better part of 1 The propensity of smokers to develop wound complications of all sorts is also well documented.5*Beta-blocking agents, by allowing unopposed alpha activity, are likely to have similar effects. When vasomotor tone is high, vasodilating drugs such as clonidine clearly raise wound PO, and promote repair.29 Some pathologic conditions, particularly the long-term effects of smoking, diabetes, and hypertension, leave a damaged microcirculation.I7These conditions increase the rate of both wound dehiscence and intestinal leaks. When these conditions exist, especially when revascularization is not feasible, they warrant special attention to enhancing perfusion and arterial Po,. Delivery of antibiotics also depends on perfusion. Parenteral antibiotics given so that high levels are present in the blood at the time of wounding clearly diminish, but do not eliminate, wound infections6 In one third of all wound infections, the bacteria cultured from the wound are sensitive to the prophylactic antibiotic given to the patient, even when the antibiotics were given according to standard procedure. The vulnerable one third of patients appear to be the hypoxic and vasoconstricted group. Antibiotics that are present in blood at the time of injury are trapped in the fibrin clot, where they exert their effect. Given later, they diffuse poorly into clotted fibrin and have little effect on wound infection. The anti-infective actions of antibiotics and oxygen are additive, but because oxygen enters fibrin clots more easily, its duration of action is longer.40 Nutrition is another factor central to repair. Protein, carbohydrates,
WOUND HEALING AND WOUND INFECTION
essential fats, and sufficient calories are all important. So are micronutrients such as zinc, iron, copper, magnesium, manganese, and vitamins A, B, C, D, E, and K. A functioning apparatus for efficient absorption and use of nutrients, including growth hormone, IGF, and insulin, is also essential. Nevertheless, astonishingly cachectic patients often manage to heal. The solution to this apparent contradiction is that recent intake is 26 No food for a week seems worse than limited the most imp~rtant.'~, regular diet for a longer period, and nutritional depletion is correlated with wound complications.", 25, 67 Similarly, rapid replenishment of depleted nutritional pools is surprisingly effective, and a few days of parenteral nutrition seems to return the patient to relatively normal reparative capacity.26 Although specific nutritional deficiencies probably exist, they are not usually sought. On the contrary, the most economical response is to give an excess of all important nutrients. Arginine and glutamine accelerate repair when given at very high doses of 15 to 30 g/day,3 an effect more pharmacologic than nutritional. The mechanisms are not well understood. NOW THAT THE PRINCIPLES ARE ESTABLISHED, WHAT CAN THE SURGEON ACTUALLY DO?
Risk Assessment Is Useful Wound infection can be predicted to a certain extent. Because wound defenses and wound repair are vulnerable to many of the same defects, predictors of infection are usually also predictors of dehiscence or other wound failures. The Centers for Disease Control, in their study of the effect of nosocomial infection control (SENIC), developed a remarkably useful and simple predictive indexz1based on a score of 0 or 1 for each of the following four patient factors: an abdominal operation, an operation that lasts 2 hours or more, an operation that is contaminated, and a patient who will have three or more diagnoses at discharge, exclusive of wound infection. The risk of infection with a score of 0 is 1%,with a score of 1 is 3.6%, with a score of 2 is 9%, with a score of 3 is 17%, and with a score of 4 is 27%. These percentages may seem high, but this index was constructed on 3% of the American surgical patients in 1975-1976 and 1983, and the overall results are consistent with numerous other studies. More recent risk analyses by the same group, based on simpler predictors (eg., American Society of Anesthesiologists Physical Status Classification), have yielded less sensitivity but about the same overall infection rate.7 Wound oxygen tension in surgical patients generally correlates with the SENIC score. In patients who develop surgical wound infections, however, wound Po, is almost invariably low, regardless of SENIC score. Moreover, wound Po, is a significantly better predictor of infection
HUNT & HOPF
Table 1. PREOPERATIVE CHECKLIST Assess and optimize cardiopulmonary function. Correct hypertension. Treat vasoconstriction: Attend to blood volume, thermoregulatory vasoconstriction, pain, and anxiety. Assess recent nutrition and treat appropriately. Treat existing infection. Among other actions, clean and treat skin infections. Assess wound risk by SENIC* score in order to decide upon the extent to which prophylactic measures should be taken. Prescribe vitamin A or anabolic steroid therapy in patients taking prednisone. Improve or maintain blood sugar control. 'Study of the effect of nosocomial infection control, Centers for Disease Control.
than SENIC score. Therefore, the SENIC indices must relate to the physiologic variables that influence blood and oxygen supply. For instance, long operations often lead to increased fluid losses, hypothermia, and greater postoperative pain. The number of diagnoses correlates with poorer cardiopulmonary function and more medications.
Preoperative Preparation Risk assessment guides preoperative preparation (Table 1).Cardiopulmonary function is obviously important. Hypertension is associated with infections and dehi~cence.'~ Therefore, vasodilating antihypertensives are helpful. We use clonidine patches (usually a No. 2) applied 2 days before operation because of their ease of use. Correct blood volume deficits preoperatively if possible. Once vasoconstriction is established, blood volume deficits are difficult to correct until the vasoconstrictive stimulus (e.g., hypothermia, pain) has been corrected, as the administered volume is lost as urine. When planning antibiotics, remember that not every infection need be prevented! The significance of wound complications varies. Antibiotic usage should be judicious. Surgeons have a duty to each other to keep antibiotic use to appropriate levels in order to perturb the environment minimally. Preoperative antibiotics lose effectiveness when given too long before or after incision. Administration within 30 minutes to 1 hour of incision is ideal. Assist nutrition if the patient has not eaten in the past few days or more or if the patient is expected to eat little in the days after surgery. Surgery should be delayed significantly only if malnutrition is severe.9, 26, 51 In less severely malnourished patients, much can be accomplished in 2 to 3 days. In all cases, if nutritional support is begun preoperatively, it should be carried into the postoperative period. Any appropriate method can be used. Control diabetes, including insulin, glucose, and blood volume abnormalities. Constant insulin infusions are generally superior to periodic injections in acute surgical circumstances. Hyperglycemia, among other detrimental effects, decreases vitamin C uptake into
WOUND HEALING AND WOUND INFECTION
cells.47This can be partially overcome by supplementing vitamin C at 500 to 2000 n;g/day. Anti-inflammatory steroids hinder wound healing. For reasons that are not understood, they seem to affect the colon less than other tissues.60 Vitamin A in oral doses of 25,000 U/day, or perhaps half that intravenously, tends to overcome the inhibitory effect temporarily.3O It may also be used topically in chronic wounds. The effect may be mediated via TGF-P, which is decreased by steroids and returned to normal by vitamin A. Patients need the inflammatory effect of vitamin A for about l week after injury. Once inflammation has developed, steroids affect healing less, but their negative effect on contraction persists. Vitamin A has also been shown to correct the adverse healing effects of diabeteP and radiation." Anabolic steroids are a useful alternative to vitamin A 50 and may also elevate IGF-I levels, probably a useful lntraoperative Management
New wounds are made and contaminated as operations proceed. Maintain antibiotic levels in long operations (Table 2). All anesthetics tend to cause hypothermia by (1)causing vasodilation, which redistributes heat from core to periphery in previously vasoconstricted patients, and increases heat loss, and (2) decreasing heat p r o d u ~ t i o n Vasoconstriction .~~ is uncommon intraoperatively, as the threshold for thermoregulatory vasoconstriction is decreased, but is often severe in the immediate postoperative period, when anesthesia is discontinued and the thermoregulatory threshold returns to normal in the face of a core temperature of 34 to 35.5"C. The onset of pain with emergence from anesthesia often adds to this vasoconstriction. Maintenance of normothermia intraoperatively has recently been shown to decrease the wound infection rate by two thirds in patients undergoing colon Rapid rewarming of hypothermic patients in the postanesthesia care unit appears also to be effective.68Maintenance of a high room temperature and forced air warming before, during, and after the
Table 2. INTRAOPERATIVE MANAGEMENT Appropriate prophylactic antibiotics should be given at the start of any procedure in which infection is highly probable and/or has potentially disastrous consequences. Keep antibiotic levels high during long operations. Keep patient warm. Observe gentle surgical technique with minimal use of ties and cautery. Keep wounds moist. Irrigate with antibiotics in contaminated cases. Elevate Pao?. Delay closure for heavily contaminated wounds. Use appropriate sutures (and skin tapes). Use appropriate dressings.
operation are far superior to other methods of warming such as circulating water blankets placed under the patient and humidification of the breathing circuit." Replace fluid losses aggressively: 10 to 15 mL/kg/hour plus three to four times the blood loss (crystalloid) during major abdominal surgery. This takes into account third-space losses as well as the preoperative fluid deficit, which is especially great after a bowel preparation. Dark venous blood in the incision usually indicates vasoconstriction, not poor ventilation. Use gentle and clean surgical technique. Plan incisions with regard to blood supply, particularly when operating near or in old incisions. Release mechanical retractors from time to time to allow perfusion. Judicious antibiotic irrigation of contaminated areas is effective. Because dried wounds lose perfusion, keep wounds moist, especially during long operations. Not all wounds can be anatomically closed. Edema, obesity, the possibility of unacceptable respiratory compromise, and need to d6bride abdominal or chest wall tissues can interfere with closure. A variety of polymeric "fabrics" is available to interpose between the wound edges. Their use is preferable to evisceration on the one hand or respiratory compromise on the other. If the problem is temporary, the implant can be removed in a few days. Polypropylene mesh was designed to allow granulation tissue to penetrate the interstices and can be left permanently. When stiff mesh materials are left longer, fistulas sometimes develop. Unfortunately, soft Teflon meshes, although useful in primary healing, are poorly incorporated into granulation tissue and eventually require removal when involved in secondary healing. Use modern sutures. They are very good. Although silk is acceptable for some uses and surgeons traditionally prefer its handling characteristics, multifilament nylon handles well and is less inflammatory and more permanent. If skin sutures are necessary, use monofilament absorbable material, preferably in a subcutaneous closure. Modem monofilament absorbable suture materials are preferred for deep closures as well. A spectrum of absorption times is available. The time to half loss of suture strength is usually the index of absorbability. Mass is lost more slowly. Polylactide sutures are most suitable for subcutaneous closures and are preferable to catgut in essentially all circumstances. Use the finest gauge compatible with the strength needed. Most surgeons use sutures that are a size larger than necessary. The tendency to suture sinus formation is proportional to the size of the suture as well as the material. Deep tissues, abdominal and chest walls, deep fascia, and synovium can be closed with running or interrupted sutures as indicated by other considerations. However, closure should never be excessively tight and should allow for the inevitable swelling. In high-risk abdominal closures, large monofilament sutures should be placed at least 2 cm back from the wound edge and about the same distance apart. Suction drains are usually more successful than subcutaneous sutures for closing potential dead spaces. Irrigate out such small particles as fat detritus,
WOUND HEALING AND WOUND INFECTION
which must be phagocytosed and removed by leukocytes and thus slows the rate of bacterial clearance. Skin tapes with or without buried subcuticular closure are still the most infection-free closure! Staples are slick and quick but difficult to place without compressing local microvasculature and entering the wound space. If you cannot be truly expert in their use, do not use them. To tell the difference, examine staple-closed wounds at about 7 days. Erythema and scarring at the entrance points indicate that the staples were placed too tightly. Dressing technique is not critical on primarily closed operative wounds because resistance to external contamination builds rapidly, within a few hours. Skin tapes covered by a simple gauze dressing usually suffice. Open, incised wounds should be dressed with nonadherent material covered by an occlusive moisture-retaining material that can be changed daily if necessary. However, if delayed primary closure is intended, the dressing should not be disturbed until the fourth or fifth day or until troubles develop. Healing is most rapid and pain is least with oxygen-permeable, plastic adherent sheets over skin graft donor sites and extensive abrasions. Delayed primary closure should be considered when wound infection seems highly probable. Wisely used, delayed closure can prevent infection and unnecessary scarring. Closure can almost always be accomplished on the ward or in the office using skin tapes and no anesthesia. Consider leaving contaminated subcutaneous tissue open for delayed primary closure on the fourth or fifth postoperative day-not before or after.66 Postoperative Management
Wounds are most vulnerable in the early hours (Table 3). Although antibiotics lose their effectiveness after the first hours, natural wound immunity lasts longer. Even a short period of vasoconstriction during the first day is sufficient to reduce oxygen supply. Correction of vasoconstriction in the first 24 to 48 hours after surgery has significant beneficial effects. Third-space losses continue for about 24 hours and then abate. Using tissue oxygen probes, we find that the routine postoperative fluid order of 100 to 150 mL/hr is inadequate to maintain wound perfusion for the first 12 to 24 hours after most abdominal and chest operation^.^^ Routine measurement of wound Po, is not yet available, so it is useful to make routine assessment of the degree of vasoconstriction, an indirect indicator of wound Po,. Unfortunately, urine output is a poor, often misleading guide to peripheral perfusion. Markedly low output may indicate decreased renal perfusion, but normal or even high urine output has little correlation with wound and tissue Po,. Many factors commonly present in the postoperative period, including hyperglycemia, dye administration,
Table 3. POSTOPERATIVE MANAGEMENT Keep patient warm. Provide analgesia to keep patient comfortable, if not pain free. Patient report and the ability to move freely are the best signs of adequate pain relief. Administer one more dose of antibiotic unless an infection is present or contamination continues. Keep up with third-space losses. Remember that fever increases fluid losses. Assess petfusion and react to abnormalities. Avoid diuresis until pain is gone and patient is warm. Assess losses (including thermal losses) if wound is open. Assess need for ParenteraVenteral nutrition and respond. Continue to control hypertension and hyperglycemia.
thermoregulatory vasoconstriction, adrenal insufficiency, and various drugs, may cause inappropriate diuresis in the face of mild hypovolemia. Physical examination of the patient is a better guide to dehydration and vasoconstriction. Assess vasoconstriction by a capillary return time of more than 1.5 seconds at the forehead and more than 5 seconds over the patella. Dehydration is best assessed by eye turgor (use your own eye as a control if you have not missed lunch and do not have glaucoma), and patients can usually distinguish thirst from dry mouth. Skin should be warm and dry. Vasoconstriction due to any cause increases the urinary output. A high urine output when the patient is cold turns into a low one when he or she is warmed. If the patient has good peripheral perfusion on physical examination, a moderately low (but adequate) urine output is usually best ignored because aggressively administering fluids in this circumstance is one of the most common settings for pulmonary edema. When excessive tissue fluids have accumulated, diuresis should be undertaken gently so that transcapillary refill can maintain blood volume. This applies to patients who need renal dialysis as well. Fluid losses from the vascular system are not necessarily replaced from the tissues as rapidly as they are sustained. The average dialysis patient vasoconstricts sufficiently to lower tissue Po, by 30% or more during dialysis and needs about 24 hours to return vasomotor tone and wound and tissue Po, to Local perfusion is not ensured until patients have a normal blood volume, are warm and pain free, and are receiving no vasoconstrictive drugs, that is, until the sympathetic nervous system is inactivated. Warming should continue until patients are thoroughly awake and active and can maintain their own thermal balance. After major operations, warming may be useful for many hours or even days. The goal is skin warmth, and wound vasoconstriction due to cold surroundings often co-exists with core hyperthermia. Moderate hyperthermia is not itself a problem. When extensive wounds are left open, warmth should be continued, and heat losses due to evaporation should be prevented to avoid vasoconstriction and to minimize caloric losses.
WOUND HEALING AND WOUND INFECTION
Vasoconstrictive drugs should be avoided. The most common and most avoidable are smoking and beta blockers. Both are known to reduce wound and tissue Po,. Alpha-adrenergic agonists are clearly harmful to tissue Po,, but in a limited experience we have found that the so-called renal dose (3 to 5 pg/kg/min) of dopamine has little or no effect on wound and tissue Po,. Maintenance of tissue PO, requires attention to pulmonary function postoperatively. Pain control appears important because it favorably influences both pulmonary function and vascular tone. This is particularly true in patients at high risk for pulmonary complications postoperatively, such as morbidly obese patients and those with pulmonary disease.69Epidural analgesia may be the route of choice in these patients. It has several advantages over parenterally administered opioids in that it generally achieves lower pain scores with less sedation. Nonetheless, opioid-induced pruritus is more common with epidural administration and, in some patients, may be severe enough to counteract the benefits of pain control. Patient-controlled analgesia (PCA) is also quite effective at achieving low pain scores. It also has the benefit of giving control to the patient, leading to patient satisfaction as high as with epidural analgesia in many cases.54Nurse-administered, as-needed dosing of intravenous or intramuscular opioids should be avoided, as inadequate pain control is common (often exceeding 50%) with this approach.’, The key to pain control is recognition of the need for analgesia and attention to the patient’s complaints of pain. Opioid requirements vary enormously and are not always predictable, but even tolerant patients (intravenous drug abusers and those with cancer pain) can be given adequate pain relief with sufficient attention. IN THE LONG TERM
Some wounds heal to excess. Keloids and hypertrophic scars, however, seem to have different natures, although the distinction between them is often hazy. Hypertrophic scars are common after burns and correlate with the length of time required to close the wound. Rapid healing is less likely to become excessive. This is one of the strong arguments for early closure and the use of skin substitutes. Hypertrophic scars also tend to occur in wounds that cross lines of force in the skin. One suspects that repeated microinjuries prolong the course of repair. In any case, “hypertrophy” represents deposition of collagen in excess of collagen lysis. It is often best treated surgically with revisions designed to relieve tension but may also respond to pressure garments, repeated injections of steroids, or prolonged dressing with thick Silastic sheets. One might speculate that the effect of the pressure garments and silicone sheets is to raise tissue temperature, which favors collagen lysis. Keloids usually occur below the hair line and generally not on hands or below the groin, although exceptions are seen. Keloids appear
to have a "neoplastic" origin. The only treatment currently used is steroid injection with or without excision. Contractures occur when shrinking scars constrict mobility. All scars shrink unless the force of contraction is balanced by stretching forces. In general, prophylactic efforts, braces, physical therapy, stretching, and pressure dressings are the most effective prevention and therapy. Special compression garments can be made. Established contractures that resist stretching or recur are often best treated surgically by flaps or by so-called 2- or Y-plasties, which relax contractures and interpose normal tissues along the line of contracture. Suture sinuses commonly occur after use of nonabsorbable or slowly absorbable sutures near the skin. Their frequency is in proportion to the size of suture material used and the presence of prior wound infection. The cure is surgical removal of the offending stitch(es). CHRONIC WOUNDS
The considerations discussed above for the care of acute wounds apply to chronic wounds, with a few modifications (Table 4). Chronic wounds heal slowly and scar more. As the angiogenic stimuli become more and more remote from the initial wound margin, vessels coalesce or drop out. The net result is a relatively ischemic layer of scar. At a site of repeated injury and repair, such as venous ulcers, this ischemic tissue itself becomes a problem because it becomes a barrier to angiogenesis. Additionally, if work done on chronic periodontal disease can be generalized, the surface "clones" of fibroblasts gradually lose their "vigor"their ability to replicate and m0ve.3~The net practical conclusion to be drawn is that there comes a time for chronic wounds to be excisednot just dkbrided. This can be a painful decision to make, but if the arterial blood supply is adequate (this can be determined with transcutaneous oximetry) and if infection is carefully treated preoperatively, a skin graft can often be placed at the time of excision with expectation of "take." Many methods are available for measuring the degree to which arterial and venous disease contribute to chronic wounds. Arteriography, venography, plethysmography, and others are all useful. We have found transcutaneous oximetry (Ptco,) particularly useful for screening Table 4. CHRONIC WOUNDS CHECKLIST Assess and correct arterial perfusion abnormalities, including those that modify vasomotor tone. Assess and correct venous outflow and regurgitation abnormalities. Assess and treat infections. Assess and correct nutrition. Assess and lower cholesterol. Keep wounds moist and warm.
WOUND HEALING AND WOUND INFECTION
and prediction of therapeutic success. Such oximeters are available in newborn nurseries and should be available in noninvasive vascular laboratories. The transcutaneous electrode is a warmed device (44°C) that facilitates measurement of Po, as far as a few millimeters into the skin. Ptco, is the best-known predictive index of subsequent repair of amputation sites.5,13, 35 In general, if Ptco, is less than 20 mm Hg, healing does not occur. If it exceeds 40 mm Hg, eventual healing is virtually certain. We have developed a method that we find quite useful for evaluating nonhealing lower-extremity wounds.29 Ptco, is measured at the wound edge and on normal skin on both the affected and unaffected extremities. A reference site on the chest wall is also measured. The measurement is made with the patient breathing first room air and then supplemental oxygen with the extremity in three positions: neutral, elevated, and dependent. The method differentiates wound hypoxia caused by arterial insufficiency, venous insufficiency, or purely local factors (lack of angiogenesis or severe vasoconstriction within the wound bed) and therefore can be used to direct treatment. Repeated measurements are used to assess the efficacy of treatment. In ulcers that are due to arterial insufficiency, the fundamental problem is ischemia, of which hypoxia is the foremost element. Diabetes is a common cause. With the leg at heart level, Ptco, measurements show low values that decrease when the leg is raised and increase little or not at all in response to supplemental oxygen administration. The best solution is surgical revascularization, either a vascular procedure or a vascularized flap. Provision of oxygen alone is often followed by healing when revascularization is not possible. If the wound PO, is low and rises significantly with supplemental oxygen, successful therapy with hyperbaric oxygen is likely.65 Ptco, measurements have led to the recent finding that many patients with nonhealing lower-extremity ulcers have normal Ptco, elsewhere on the extremity, but marked tissue hypoxia within the wound area itself (Po, usually less than 20 mm Hg). These patients display a characteristic cluster of findings: hypertension, palpable distal pulses, severe pain in the ulcer, and a cold, often cyanotic foot. These findings appear to be due to local vasospasm induced by excess catecholamines in the circulation, excess reactivity to catecholamines (possibly due to upregulation of alpha, receptors in the wound), or both. Impaired angiogenesis likely also contributes. Tissue Po, in these patients is nearly always increased by percutaneous lumbar sympathetic blockz8or transThe combination of the two may dermal clonidine (No. 1 or 2 be additive. Other vasodilators may well be effective, but they have not yet been evaluated. Hypercholesterolemia and hypertension are associated with advancing arteriosclerosis and also with excessive vasomotor activity and should be controlled. Warming the extremity is also effective at increasing Ptco, in these patients. In fact, in some patients who display these characteristics, Ptco, is higher than expected. This appears to be the result of local warming
by the transcutaneous oximeter that overcomes the local vasoconstriction. These patients show marked improvement in perfusion when the extremity is submerged in a basin of warm water. Contrary to older beliefs, this calls for continuous warming as therapy. Venous insufjiciency is the most common cause of chronic ulcers in the legs. These ulcers are related largely to hypoxia, which in this case is due to blood congestion and subsequent local tissue hypoxia when the victim stands. Ptcq tends to rise and then fall shortly after the patient sits or stands. This kind of ischemia promotes local adherence of leukocytes, with further blockage of the microcirculation. Edema worsens the hypoxia. Therapy is intended to prevent the venous congestion and edema, and this is relatively easily accomplished with compression dressings. Several methods are used. All are effective in proportion to patient compliance. We prefer a tailored heavy compression stocking. Elderly arthritic patients sometimes cannot don them, and, if home nursing care is not available, compression bandages, changed about weekly, may be necessary. In the presence of exudative ulcers, compression dressings are less than ideal, as frequent washing of the ulcer is critical to minimize bacterial load. Hyperbaric oxygen is of no use for venous ulcers unless arterial disease is present (about 15% of cases). Failure of venous ulcers to heal or early recurrence often suggests that penetrating, valveless veins are present. Sonography and venography are useful to detect them. Such veins must be surgically divided. The modified Linton procedure is very effective, but skin loss is common and recovery is protracted. Newer videoscopic procedures have been described, but it is difficult to access the curving surface of the fascia, and incomplete division is a problem. Ulcers due to lymphatic insufjiciency are rare and are distinguished by the almost clear discharge that dries on the thickened skin, which is not as severely pigmented as in venous disease. Bed rest, compression, and sometimes local oxygen are useful. Neuropathic ulcers, found largely in diabetics and paraplegics, are often the most difficult to treat because insensate skin allows repeated injury. These patients, even the diabetics, usually have adequate circulation, and the problem is not so much healing as it is prevention of further injury. Both patient and surgeon have a coping job to do that involves questions of activity, special protective footwear, wheelchairs, nursing care, and the like. Pressure sores and neuropathic ulcers are often the same problem. Although they are commonly called signs of inadequate nursing care, they are also signs of a surgeon’s failure to cope with the reality of a crippled or unconscious patient. Prevention is the soundest policy because large pressure ulcers heal slowly, if at all. Chronic recurrent ulcers over protuberant bone can be excised together with the bony protuberance, and the wound can usually be safely closed at the same time. Larger ulcers can be covered with skin or muscle flaps. However, this approach is often an ineffective and expensive gesture because the flaps are also insensate and pressure ulcers recur quickly. A comprehen-
WOUND HEALING AND WOUND INFECTION
sive guide to cause and treatment can be found in the Clinical Practice Guideline published by the Agency for Health Care Policy and Dressings are extremely important in chronic wound therapy. Regardless of the cause of the wound, dressings that keep the wound moist are essential. Placing a wet dressing, letting it dry, and then tearing out the adherent tissue dkbride the wound but also delay repair. Many dressings that keep wounds moist are available. The savings in healing time and pain are considerable. Bactericidal agents such as povidone-iodine and hydrogen peroxide 46 These agents should be avoided unless used in very dilute s~lution.~, are cytotoxic as well and retard healing. In most wounds, daily washing or soaking with mild soap and water decreases the bacterial load so that antibacterial agents are not necessary. When contamination is a concern, 1%silver sulfadiazine cream or polysporin ointment keeps the wound moist and decreases bacterial load without impeding healing. What may the surgeon be able to do in the future? Although growth factors have failed to produce significant clinical benefit in chronic human wounds, human growth hormone given topically has proved beneficial in the healing of burns and recovery from surgery. Objective evidence of benefit from small doses has been obtained in elderly men.23 This effect is mediated through IGF-I, which may itself be useful when it is given systemically. The anabolism required for local protein synthesis depends heavily on growth hormone and local production of IGF-I. These factors serve almost as ”master switches” in repair.44Recent data show that synthetic hGH-like peptides are powerful stimulants of bone repair in humans. This is apparently near clinical status. Clearly, however, these do not obviate the need for oxygenation and perfusion.
SUMMARY Wound healing can be enhanced and wound infections prevented, often by simple, inexpensive, readily available means. Preoperative evaluation for impediments to healing, such as malnutrition, vasoconstriction, hyperglycemia, and steroid use, allows correction prior to operation. Intraoperatively, the surgeon should concentrate on surgical technique, appropriate antibiotic use, and prevention of vasoconstriction (volume, warming). Postoperatively, the focus should be on prevention of vasoconstriction through pain relief, warming, and adequate volume resuscitation and on maintaining nutrition and normoglycemia. These approaches apply as well to chronic wounds. Additionally, maintenance of a moist environment, correction of local vasospasm with sympathetic blockade or warming, and stimulation of angiogenesis through aggressive debridement or hyperbaric oxygen therapy enhance healing of chronic wounds.
References 1. Allen D, Marocci L, Scheuenstuhl H, et al: The respiratory burst of intact human neutrophils is impaired at the low oxygen tensions found in wounds. Wound Repair and Regeneration 272.1994 2. Babior -BM: Oxygen-dependent microbial killing by phagocytes. N Engl J Med 198:659. - , 1978 3. Barbul A, Lazarou S, Efron D, et al: Arginine enhances wound healing and lymphocyte immune response in humans. Surgery 108:331-337, 1990 4. Branemark P, Albrektsson B, Lindstrom J, et a1 Local tissue effects of wound disinfectants. Acta Chir Scand 357(Suppl):166-176, 1966 5. Burgess EM, Matsen F Determining amputation levels in peripheral vascular disease. J Bone Joint Surg Am 631493-1497,1981 6. Classen D, Evans R, Pestotnik S, et a1 The timing of prophylactic administration of antibiotics and the risk of surgical wound infection. N Engl J Med 326281-286,1992 7. Culver D, Horan T, Gaynes R, et a1 Surgical wound infection rates by wound class, operative procedure, and patient risk index. Am J Med 91:152%157S, 1991 8. DeJong L, Kemp A: Stoichiometry and kinetics of the prolyl 4-hydroxylase partial reaction. Biochim Biophys Acta 787105-111, 1984 9. Delany H, Demetriou A, Teh E, et al: Effect of early postoperative nutritional support on skin wound and colon anastomosis healing. JPEN J Parenter Enteral Nutr 14357361,1990 10. Derbyshire D, Smith G Sympathoadrenal responses to anaesthesia and surgery. Br J Anaesth 56:725-739, 1984 11. Dickhaut S, DeLee J, Page C: Nutritional status: Importance in predicting woundhealing after amputation. J Bone Joint Surg Am 66:71-75,1984 12. Donovan M, Dillon P, McGuire L Incidence and characteristics of pain in a sample of medical-surgical inpatients. Pain 3069-78, 1987 13. Dowd G S Predicting stump healing following amputation for peripheral vascular disease using the transcutaneous oxygen monitor. Ann R Coll Surg Engl69:31-35,1987 14. Dvonch V, Murphey R, Matsuoka J, et al: Changes in growth factor levels in human wound fluid. Surgery 11218-23, 1992 15. Edwards S, Hallett M, Campbell A Oxygen-radical production during inflammation may be limited by oxygen concentration. Biochem J 217851-854,1984 16. Evans NTS, Naylor PFD Steady states of oxygen tension in human dermis. Respir Physiol246-60,1966 17. Fawcett A, Shembekar M, Church J, et al: Smoking, hypertension, and colonic anastomotic healing: A combined clinical and histopathological study. Gut 38:714-718, 1996 18. Gabig TG, Bearman SI, Babior BM Effects of oxygen tension and pH on the respiratory burst of human neutrophils. Blood 531133-1139, 1979 19. Goodson WH, Lopez SA, Jensen JA, et al: The influence of a brief preoperative illness on postoperative healing. Ann Surg 20525G255, 1987 20. Gottrup F, Firmin R, Rabkin J, et al: Directly measured tissue oxygen tension and arterial oxygen tension assess tissue perfusion. Crit Care Med 15:1030-1036, 1987 21. Haley RW, Culver DH, Morgan WM, et a1 Identifying patients at high risk of surgical wound infection: A simple multivariate index of patient susceptibility and wound contamination. Am J Epidemiol 121:20&215,1985 22. Halter J, Pflug A, Porte D Jr: Mechanism of plasma catecholamine increases during surgical stress in man. J Clin Endocrinol Metab 45936-944,1977 23. Hamon G, Stotts N, Tiemey M, et a1 Effect of recombinant human growth hormone (rhGH) on wound healing in normal elderly men. Wound Repair and Regeneration 4:421425, 1996 24. Hartmann M, Jonsson K, Zederfeldt B: Effect of tissue perfusion and oxygenation on accumulation of collagen in healing wounds: Randomized study in patients after major abdominal operations. Eur J Surg 158:521-526,1992 25. Haydock D, Hill G: Impaired wound healing in surgical patients with varying degrees of malnutrition. JPEN J Parenter Enteral Nutr 10550-554, 1986
WOUND HEALING AND WOUND INFECTION
26. Haydock D, Hill G: Improved wound healing response in surgical patients receiving intravenous nutrition. Br J Surg 74320-323,1987 27. Houf H. Hunt T.. ,Tensen 1: Calculation of subcutaneous tissue blood flow. Sure " Forum 39:53-36, 1988 28. Hopf H, McKay W, West J, et al: Percutaneous lumbar sympathetic block increases tissue oxygen in patients with local tissue hypoxia in non-healing wounds. Anesth Analg, 84:5305, 1997 29. Hopf H, West J, Hunt T Clonidine increases tissue oxygen in patients with local tissue hypoxia in non-healing wounds. Wound Repair and Regeneration 4A129,1996 30. Hunt T, Ehrlich H, Garcia J, et a1 Effect of vitamin A on reversing the inhibitor effect of cortisone on healing of open wounds in animals and man. Ann Surg 170:633-641,1969 31, Hunt T, Hopf H: Nutrition in wound healing: In Fischer J (ed): Nutrition and Metabolism in the Surgical Patient, ed 2. Boston, Little, Brown & Co, 1996, pp 423-441 32. Hunt TK, Knighton DR, Thakral KK, et al: Studies on inflammation and wound healing: Angiogenesis and collagen synthesis stimulated in vivo by resident and activated wound macrophages. Surgery 96:4&54, 1984 33. Hussain M, Ghani Q, Zhang J, et al: Alterations of fibroblast metabolism in early ligature-induced periodontitis in the cynomolgus monkey. J Periodontol 65:771-775, 1994 34. Hutton JJ, Tappel AL, Udenfriend S: Cofactor and substrate requirements of collagen proline hydroxylase. Arch Biochem 118:231, 1967 35. Ito K, Ohgi S, Mori T, et al: Determination of amputation level in ischemic legs by means of transcutaneous oxygen pressure measurement. Int Surg 69:5941, 1984 36. Jensen JA, Goodson WH, Hopf HW, et al: Cigarette smoking decreases tissue oxygen. Arch Surg 1261131-1134,1991 37. Jensen JA, Goodson WH, Omachi RS, et al: Subcutaneous tissue oxygen tension falls during hemodialysis. Surgery 101:416-421, 1987 38. Jensen JA, Hunt TK, Scheuenstuhl H, et al: Effect of lactate, pyruvate and pH on secretion of angiogenesis and mitogenesis factors by macrophages. Lab Invest 543574578, 1986 39, Jonsson K, Jensen JA, Goodson WH, et a1 Assessment of perfusion in postoperative patients using tissue oxygen measurements. Br J Surg 74:263-267,1987 40. Knighton DR, Halliday B, Hunt T K Oxygen as an antibiotic: A comparison of the effects of inspired oxygen concentration and antibiotic administration on in vivo bacterial clearance. Arch Surg 121:191-195, 1986 41. Knighton DR, Silver IA, Hunt TK Regulation of wound-healing angiogenesis-ffect of oxygen gradients and inspired oxygen concentration. Surgery 90:262-270, 1981 42. Kurz A, Kurz M, Poeschl G, et al: Forced-air warming maintains intraoperative normothermia better than circulating water mattresses. Anesth Analg 7789-95, 1993 43. Kurz A, Sessler D, Lenhardt R, et a1 Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization. N Engl J Med 3341209-1215, 1996 44. Lee K, Stotts N Support of the growth hormone-somatomedin system to facilitate healing. Heart Lung 19:157-162, 1990 45. Levenson S, Gruber C, Rettura G, et al: Supplemental vitamin A prevents the acute radiation-induced defect in wound healing. Ann Surg 200:494, 1984 46. Lineaweaver W, Howard R, Soucy D, et al: Topical antimicrobial toxicity. Arch Surg 120267-270, 1985 47. Marhoffer W, Stein M, Maeser D, et al: Impairment of polymorphonuclear leukocyte function and metabolic control of diabetes. Diabetes Care 15256-260, 1992 48. Matsukawa T, Sessler D, Sessler A, et al: Heat flow and distribution during induction of general anesthesia. Anesthesiology 826624573, 1995 49. Medawar PS: The behavior of mammalian skin epithelium under strictly anaerobic conditions. Q J Microsc Sci 88:27, 1947 50. Mueller R, Hunt T, Tokunaga A, et a1 The effect of insulin-like growth factor I on wound healing variables and macrophages in rats. Arch Surg 129:262-263, 1994 51. Mullen J, Buzby G, Matthews D, et al: Reduction of operative morbidity and mortality
by combined preoperative and postoperative nutritional support. Ann Surg 192604613,1980 52. Myllyla R, Tuderman L, Kivirikko Kk Mechanism of the prolyl hydroxylase reaction: 2. Kinetic analysis of the reaction sequence. Eur J Biochem 80:349-357, 1977 53. Niinikoski J, Jussila P, Vihersaari T Radical mastectomy wound as a model for studies of human wound metabolism. Am J Surg 12653-58,1973 54. Owen H, McMillan V, Rogowski D Postoperative pain therapy: A survey of patients' expectations and their experiences. Pain 41:303-307, 1990 55. Panel on the Treatment of Pressure Ulcers: Treatment of Pressure Ulcers. Clinical Practice Guideline. US Department of Health & Human Services, Public Health Service, Agency for Health Care Policy and Research, 1994 56. Prockop DJ, Kivirikko KI, Tuderman L, et al: The biosynthesis of collagen and its disorders (first of two parts). N Engl J Med 301:13-23, 1979 57. Rabkin JM, Hunt TK Local heat increases blood flow and oxygen tension in wounds. Arch Surg 122221-225,1987 58. Rees TD, Liverett DM, Guy C L The effect of cigarette smoking on skin-flap survival in the face lift patient. Plast Reconstr Surg 73:911-915, 1984 59. Rudolph R, Vande Berg J, Ehrlich H: Wound contraction and scar contracture. In Cohen I, Diegelmann R, Lindblad W (eds): Wound Healing: Biochemical and Clinical Aspects. Philadelphia, WB Saunders, 1992, pp 96-114 60. Schrock T, Deveney C, Dunphy J: Factors contributing to leakage of colonic anastomoses. Ann Surg 177513-518, 1973 61. Schultz G, Grant M Neovascular growth factors. Eye 5370-180, 1991 62. Seifter E, Rettura G, Padawer J, et al: Impaired wound healing in streptozotocin diabetes: Prevention by supplemental vitamin A. Ann Surg 19442,1981 63. Sheffield C, Sessler D, Hopf H, et a1 Centrally and locally mediated thermoregulatory responses alter subcutanebus oxygen tension: Wound Repair and Regeneration 4:339345, 1996 ' 64. Silver I A Cellular microenvironment in healing and non-healing wounds. In Hunt TK, Heppenstall RB, Pines E (eds): Soft and Hard Tissue Repair. New York, Praeger, 1984, pp 50-66 65. Smith B, Desvigne L, Slade J, et a1 Transcutaneous oxygen measurements predict healing of leg wounds with hyperbaric therapy. Wound Repair and Regeneration 4:224229, 1996 66. Verrier E, Bossart K, Heer F Reduction of infection rates in abdominal incisions by delayed wound closure techniques. Am J Surg 13822-28,1979 67. Wamold I, Lundholm K Clinical significance of preoperative nutritional status in 215 noncancer patients. Ann Surg 199:299-305,1984 68. West J, Hopf H, Sessler D, et al: The effect of rapid postoperative rewarming on tissue oxygen. Wound Repair and Regeneration 1:93,1993 69. Wisner D A stepwise logistic regression analysis of factors affecting morbidity and mortality after thoracic trauma: Effect of epidural analgesia. J Trauma 30799-804,1990
Address reprint requests to Thomas K. Hunt, MD UCSF Department of Surgery Wound Healing Laboratory HSW 1652 513 Pamassus Avenue San Francisco, CA 94143-0522