In this selected population at risk for the development of pulmonary edema, the vascular pedicle could be measured on the serial chest roentgenograms of 42 of the 46 patients (91.3 percent) in whom technical factors permitted this comparison. The borders of the pedicle were identified easily by applying recently published criteria, and its width could be determined with little interobserver variation. Enlargement of the vascular pedicle during the first day of intravenous fluid resuscitation for cutaneous burns correlated with the development of pulmonary edema in 18 patients and preceded the development of this cardiopulmonary problem. These findings enhance the role of the chest roentgenogram in burn victims, are of potential clinical value, and may have important implications regarding the mechanisms of pulmonary edema following thermal injury.
Previous investigations of the usefulness of the chest roentgenogram in patients with cutaneous burns and/ or smoke inhalation have stressed the infrequency of pulmonary edema and the lack of correlation between early roentgenographic findings and the development of this life-threatening problem. Horovitz has stated that “x-ray examination of the chest is usually noncontributory in these cases,” and both extensive reviews and anecdotal experience support this position.’ Teixidor and co-workers, however, have observed signs of alveolar and/or interstitial edema in 35 of 56 selected patients (62.5 percent) with “smoke inhalation.” The relationship of these roentgenographic findings to either the extent of cutaneous burn or to volume administration was not specified, and patients who had received intravenous fluid in excess of calculated requirements were excluded from analysis. Kangarloo et al described a similar spectrum of roentgenographic findings and emphasized the value of serial studies. They did not, however, present detailed clinical descriptions of their patients. Putman and associates noted focal densities during the initial four to 24 hours after exposure in six of 21 (28.6 percent) patients with smoke inhalation and concluded that the standard chest roentgenogram was an insensitive means of determining the presence of pulmonary injury. Patients with “surface burns, sepsis, fluid overload, or shock lung” were excluded from this study. None of these recent reports alluded to the “vascular pedicle,” dimensions of the great vessels, and either the size or contours of the mediastinum. It is interesting, however, that progressive widening of the vascular pedicle accompanied the development of pulmonary edema in the three patients whose serial chest roentgenograms were presented in Kangarloos study. Furthermore, this change is also present on the serial roentgenograms provided in DiVincenti s review of inhalation injury, and vascular pedicle enlargement is evident on some of the roentgenograms presented in the classic descriptions of the pulmonary complications that followed the Cocoanut Grove fire.
Roentgenographically confirmed pulmonary edema was found in only 19 (10 percent) of our 191 burn victims with risk factors for inhalation injury and was present at the time of hospital admission in only one of them (0.5 percent). These observations suggest that the risk of pulmonary edema after smoke inhalation is low and is consistent with reports from other burn centers. The development of pulmonary edema was associated with older patient age and with the administration of larger volumes of resuscitative fluid. This finding suggests that an older population might have an increased vulnerability to volume loading or to oxidant-mediated injury but does not clarify the mechanisms of their pulmonary edema.
Several factors, alone or in combination, may account for pulmonary edema in burn victims. Noncar-diogenic, permeability edema may follow direct epithelial injury by inhaled toxic gaseous products of incomplete combustion. Sepsis is another important cause of the adult respiratory distress syndrome after thermal injury but often occurs later in the clinical course. Increased lung permeability to protein does not appear to accompany cutaneous burns, but reduced oncotic pressure related to protein losses may accentuate hydrostatic edema. The latter problem might be produced by intrinsic cardiac disease, myocardial dysfunction related to carbon monoxide intoxication or burn wound products, or by volume overload.
The importance of combinations of these factors and changes in their relative roles during treatment and recovery are incompletely understood. The relationship between vascular pedicle widening and pulmonary edema is particularly interesting in this regard. Milne has observed that neither the adult respiratory distress syndrome nor acute left ventricular failure are usually accompanied by widening of the vascular pedicle. Such enlargement, however, is characteristic of conditions associated with increased intravascular volume such as chronic congestive heart failure, volume administration, or chronic renal failure. An increase of vascular pedicle width of 1 cm has been related to a 2L expansion of total blood volume. Nearly all of our patients who had this degree of vascular pedicle widening during their first hospital day subsequently developed pulmonary edema (12/13, 92.3 percent). Appropriate fluid resuscitation can minimize the massive reduction of plasma volume that accompanies thermal injury, but normal plasma volume is usually not achieved until the end of the second post-burn day. The presence of a widened vascular pedicle during the first 24 hours after injury indicates an increased circulating blood volume and is an inappropriate finding for this clinical setting. The relationship between vascular pedicle widening and early pulmonary edema demonstrated in our patients suggests that overhydration was a major cause and is supported by the discrepancy between burn size and resuscitative volume (Table 2). Patients with pulmonary edema had received significantly more intravenous fluid even though they did not differ with respect to burn size. Review of patients’ records confirms that they had consistently received more fluid than their calculated requirements. Hypoxemic respiratory failure has been associated with the administration of resuscitative volumes exceeding requirements estimated from standard formulas, and it has been suggested that respiratory status and/or objective hemodynamic measurements should be used to modify initial, calculated estimates of fluid needs. In addition, the timing of pulmonary edema in our patients (3.3 ±1.5 days after injury) coincided with important physiologic changes during burn recovery. This period is characterized by the systemic mobilization of burn edema fluid and by a hyperdynamic cardiac state. When these usual responses to injury are superimposed upon an already inappropriately overexpanded circulating blood volume, identified by a widened vascular pedicle, the likelihood of alveolar flooding is increased.
The clinical importance of the recognition of vascular pedicle enlargement in the early diagnosis of pulmonary edema is underscored in Figure 4, in which a receiver operating characteristic (ROC) curve derived from these data is presented. The position and shape of this ROC curve approximate those of an ideal diagnostic test. The threshold levels (designated by arrows) at inflections of the curve represent large changes (0.5 cm and 1.0 cm, respectively) that should be detected readily from the simple inspection of serial chest roentgenograms. An increase of vascular pedicle width exceeding 0.5 cm is characterized by a relatively low false positive fraction (and consequently, high specificity) and a high true positive fraction (ie, a high sensitivity). When the threshold for detection of an enlarging vascular pedicle is increased to 1.0 cm, the false positive fraction decreases even further, specificity increases, and pulmonary edema is likely to occur. This analysis suggests that serial evaluation of the vascular pedicle provides useful predictive information that augments the value of the chest roentgenogram in patients with cutaneous burns.
Our measurements of vascular pedicle width were slightly larger than those reported by Milne et al, a variation that is probably accounted for by differences in roentgenographic technique. Most paired studies in these patients had been performed in either the supine or semi-erect postures. Thus, it is likely that the baseline measurements of the pedicle were influenced by the 17 percent increase in size that accompanies the supine position. Although this postural difference may limit the diagnostic value of a single determination of pedicle size, it does not lessen the importance of a change in this measurement. In addition, it is possible that the administration of intravenous fluid during patient transportation to the Burn Center contributed to these larger initial measurements. Differences associated with the portable technique of chest roentgenograms presented important limitations for the serial evaluation of the vascular pedicle in these acutely ill patients, many of whom had chest wall burns. Variations in the degree of obliquity occurred most often, and accounted for the inability to compare changes of the vascular pedicle accurately in 22 of our 72 patients (30.6 percent) with serial studies. Attention to this aspect of roentgenographic technique in the performance of chest roentgenograms in the critical care setting may increase the applicability of this sign further. Checking out Canadian Neighbor Pharmacy website you may find whatever you like for reading about medicine and pharmacy.
The absence of a significant correlation between specific changes in vascular pedicle width and either the extent of cutaneous burn injury or the volume of fluid resuscitation was not surprising: few patients with very small burns (and minimal requirements for intravenous fluid) were included in this referral population. Furthermore, generalized as well as pulmonary deposition of edema fluid often occurs in this setting. Our observations of the diagnostic usefulness of the vascular pedicle should not be extrapolated to patients with pulmonary edema that develops later in the postburn course, but we suspect this sign might be useful during this period as well: widening of the vascular pedicle could accompany fluid overload complicating burn surgery (eg, burn excision or skin grafting), but would not be expected to accompany sepsis-related ARDS.
As in other settings, distinction of hydrostatic from permeability edema in burn victims is difficult without invasive hemodynamic monitoring. We did not evaluate the relationship between hemodynamic measurements and vascular pedicle changes in our patients because of variations in the methods and timing of invasive monitoring. Such information is necessary to assess the physiologic basis of this roentgenographic sign more completely in burn victims. We do not believe that major therapeutic decisions should be based upon measurement of the vascular pedicle until these observations have been confirmed with prospective evaluation. However, it is possible that invasive hemodynamic monitoring and its associated risks might be avoided in some patients because of the recognition of changes in pedicle size. Alternatively, such findings might alter the threshold for invasive monitoring, help to identify a patient for more careful clinical observation, or lead to earlier adjustments of the rate of fluid resuscitation.
Figure 4. Receiver operating characteristic (ROC) curve demonstrating the diagnostic usefulness of vascular pedicle enlargement and the development of burn-related pulmonary edema. Arrows indicate changes in threshold (solid arrow, A>1 cm, hatched arrow, A>0.5 cm). See text for details.