Publications (8)

  • J F Fang · R J Chen · Y C Wong · [...] · M F Chen
    [Show abstract] [Hide abstract] ABSTRACT: Pooling of contrast material on computed tomographic (CT) scan represents free extravasation of blood as a result of active bleeding. For patients with blunt hepatic injury, aggressive management such as angiography or celiotomy is usually indicated if this sign is detected. The purposes of this study were to further categorize this CT scan finding and to correlate its characteristics with clinical outcomes. This CT scan classification might be helpful for the selection of appropriate management. During a 42-month period, 276 patients with blunt hepatic injury were treated. Two hundred twelve of them were hemodynamically stable after initial resuscitation and underwent abdominal CT scan examination. Pooling of contrast material was detected on the CT scans of 15 patients. The CT scans and medical records were reviewed. Special attention was paid to the presence, location, and character of the extravasated contrast material. The finding of pooling of contrast material on CT scan was categorized into three types according to its location and character. Type I showed extravasation and pooling of contrast material in the peritoneal cavity (six patients). All patients with type I CT scan findings became hemodynamically unstable soon after CT scan examination and required emergent laparotomy. Type II findings showed simultaneous presence of hemoperitoneum and intraparenchymal contrast material pooling (six patients). Four patients with type II CT scan findings required laparotomy for hemostasis. Type III findings showed intraparenchymal contrast material pooling without hemoperitoneum (three patients). All patients with type III CT scan signs remained hemodynamically stable. With the use of a high-speed spiral CT scanner, it is possible to predict the necessity of operative management or angiography for patients with blunt hepatic injury before deterioration of hemodynamic status. The presence of pooling of contrast material within the peritoneal cavity indicates active and massive bleeding. Patients with this CT scan finding show rapid deterioration of hemodynamic status. Most of these patients might require emergent surgery. Pooling of contrast material in a ruptured hepatic parenchyma indicates active bleeding. Close monitoring and emergent angiography should be performed. Deterioration of hemodynamic status in these patients usually requires prompt surgical intervention. Intraparenchymal pooling of contrast material with unruptured liver capsule often indicates a self-limited hemorrhage. Patients with this CT scan finding have a high possibility of successful nonoperative treatment.
    Article · Jan 2001 · The Journal of trauma
  • J F Fang · R J Chen · B C Lin · [...] · M F Chen
    [Show abstract] [Hide abstract] ABSTRACT: Many publications recommend nonoperative treatment for stable blunt hepatic injury patients. Unstable hemodynamic status is the only indication for surgery. When operation is indicated, controversies exist regarding which operative procedure will be more beneficial to the patients. The purposes of this study are to compare the results of operative and nonoperative management of patients with blunt hepatic injuries and to identify the optimal surgical approach when surgery is indicated. Different prospective protocols of treating adult blunt hepatic injuries were conducted. From 1992 to 1993 (group I), urgent surgery would be performed in the presence of hemoperitoneum. The policy shifted to aggressive nonoperative approach between 1996 and 1997 (group II). The patients from each period were divided into three subgroups. Group A included the patients who received nonoperative treatment in either period. Group B consisted of the patients who received surgery in the first period and nonoperative management in the second period. Group C included the patients who were operated on in either group. Comparisons were made between matched groups. Groups IA and IIA patients had minor injuries and could be successfully treated nonoperatively. The results of groups IB and IIB were similar concerning hospital stay, morbidity, and mortality. Transfusion requirements of group IIB patients were significantly higher (2.2 vs. 1.1 units,p = 0.01) than those of group IB. However, 25 (58%) celiotomies of group IB patients were nontherapeutic. When surgery was indicated, group IC patients had significantly higher liver-related mortality (14 of 49 vs. 3 of 55, p = 0.002). Anatomic resection was performed more frequently in that period. Nonoperative treatment significantly decreased the rate of nontherapeutic laparotomy but carried the risks of higher transfusion requirements and delaying operation. When surgery was indicated, the policy of minimal intervention positively affected the patients' outcomes. The goal of surgery should be hemorrhage control rather than resection of the injured liver tissues.
    Article · Nov 2000 · The Journal of trauma
  • J F Fang · R J Chen · B C Lin · [...] · M F Chen
    [Show abstract] [Hide abstract] ABSTRACT: The neurologic outcome of comatose patients has a wide variation from complete reawakening to death. Methods of predicting the outcome of coma caused by either head injury or cardiac arrest have been the subject of much discussion in the literature. However, prediction of neurologic prognosis in comatose trauma patients without head injury has rarely been discussed. We reviewed our experience in treating patients with presumptive hypoxic-ischemic coma after trauma and tried to identify factors relating to their neurologic outcomes. Thirty-six patients with normal brain computed tomographic scans, who remained comatose 10 minutes after stabilization of their hemodynamic status, were studied. Serial motor response, verbal response, pupillary light reflex, presence of spontaneous breathing and seizure, and blood glucose level were recorded to evaluate their roles in predicting neurologic outcomes. There were five deaths (mortality rate, 14%) and 11 patients (31%) with neurologic deficits. An absence of spontaneous breathing, a blood glucose level greater than 300 mg/dL during resuscitation, and a presence of seizure signified a poor prognosis. Initial neurologic evaluation at 10 minutes after stabilization of hemodynamic status was not accurate in predicting outcome. A motor response worse than withdrawal from painful stimuli at 24 hours after injury and an absence of pupillary light reflex at 48 hours after injury predicted a poor neurologic outcome, with a 100% accuracy rate. Hypoxic-ischemic coma in patients sustaining major trauma yielded a significantly better survival and neurologic outcome than that induced by cardiac arrest or head injury. Decision-making in the first 24 hours after injury should not be affected by the patient's neurologic status at that time. A motor response worse than withdrawal at 24 hours after injury and an absence of pupillary light reflex at 48 hours after injury predicted a poor neurologic outcome.
    Article · Dec 1999 · The Journal of trauma
  • J F Fang · R J Chen · B C Lin · [...] · M F Chen
    [Show abstract] [Hide abstract] ABSTRACT: Controversies regarding how urgent bowel perforation should be diagnosed and treated exist in recent reports. The approach for early diagnosis is also debatable. The purposes of this study were to evaluate the relationship between treatment delay and outcome of small bowel perforation after blunt abdominal trauma and to determine the best assessment plan for the diagnosis of this injury. One hundred eleven consecutive patients with small bowel perforations caused by blunt abdominal trauma were retrospectively reviewed. The patients were divided into four groups according to the time interval between injury and surgery. Hospital stay, time to resume oral intake, and mortality and morbidity rates were compared between groups. Physical signs, laboratory and computed tomographic findings, and the results of diagnostic peritoneal lavage were analyzed to find the most sensitive and specific test for early diagnosis of small bowel perforation. Delay in surgery for more than 24 hours did not significantly increase the mortality with modern method of treatment; however, complications increased dramatically. Hospital stay and time to resume oral intake increased significantly when surgery was delayed for more than 24 hours. Abdominal tenderness was a common finding, but it was not specific for bowel perforation. Only 40% of the computed tomographic scans were diagnostic for bowel perforations: 50% of them showed suggestive signs, and 10% were considered as negative. Persistence of abdominal signs indicated peritoneal lavage. By using cell count ratio in diagnostic peritoneal lavage and/or increased lavage amylase activity, presence of particulate matter and/or bacteria in the lavage fluid, all patients with intraperitoneal bowel perforation were diagnosed accurately before operation. Small bowel perforation has low mortality and complication rates if it is treated earlier than 24 hours after injury. The principle of "rushing to the operation suite" for a stable blunt abdominal trauma patients without detailed systemic examination is not justified. The priority of treatment for the small bowel perforation should be lower than the limb-threatening injuries. Diagnostic peritoneal lavage provides high sensitivity and specificity rates for the diagnosis of small bowel perforation if a specially designed positive criterion is applied.
    Article · Oct 1999 · The Journal of trauma
  • J F Fang · R J Chen · B C Lin · [...] · M F Chen
    [Show abstract] [Hide abstract] ABSTRACT: Delay in surgical treatment and duodenal wound dehiscence are two major causes of extensive retroperitoneal abscess formation after blunt duodenal injury. This complication is traditionally treated with primary repair of the duodenal wound and drainage of the abscess through anterior laparotomy. Pyloric exclusion is sometimes added as an adjunctive procedure. The anterior approach, however, may result in inadequate drainage, and repeat surgery is sometimes needed. We reviewed our experiences and evaluated the effectiveness of retroperitoneal laparostomy for the treatment of retroperitoneal abscess with continuous soiling. There were 52 blunt duodenal injuries during a 7-year period. Eleven patients developed extensive retroperitoneal abscesses. All 11 patients were treated with anterior laparotomy initially. Five patients recovered after this procedure. Six patients continued to have retroperitoneal abscesses and were under septic status. Two patients received another anterior drainage, and had recurrent abscesses later. Retroperitoneal laparostomy was performed for these six patients. After retroperitoneal laparostomy, daily wound care, and antibiotic treatment, all six patients recovered. Only two patients developed incisional hernia. Retroperitoneal laparostomy is effective in treating extensive intractable retroperitoneal abscess after blunt duodenal injury. Patients with the complications of duodenal leak and extensive retroperitoneal abscess should be treated with pyloric exclusion and drainage through anterior laparotomy first. If the duodenal wound does not heal after pyloric exclusion and retroperitoneal abscess persists, retroperitoneal laparostomy should be performed without further attempt to repair the wound.
    Article · May 1999 · The Journal of trauma
  • J F Fang · R J Chen · YC Wong · [...] · YC Kao
    [Show abstract] [Hide abstract] ABSTRACT: Nonoperative management of blunt hepatic injury is currently a widely accepted treatment modality. Computed tomography (CT) is an important imaging study both for diagnosis and follow-up of these patients. There is, however, no reliable predictor of failure of nonoperative treatment other than the ultimate development of hemodynamic instability. Previous reports mostly were based on the data obtained from low-speed dynamic incremental scanners. The purpose of this study is to evaluate the value of a high-speed helical scanner in predicting the outcome of patients managed nonoperatively. During a 30-month period, 194 patients with blunt hepatic injury were treated, 150 of them were hemodynamically stable after initial resuscitation and underwent abdominal CT examination. All CT scans were performed with the High Speed Advantage Scanner. The CT scans and medical records were reviewed. Nonoperative management was successfully applied to all patients with grade I and II, 93% of grade III, 87% of grade IV, and 67% of grade V liver injuries. Twelve patients required liver-related celiotomy. Pooling of contrast material was detected on the CT scans of 8 patients. Six (75%) of these patients developed hemodynamic instability and required liver-related celiotomy later. Pooling of contrast material can be detected in 50% of the patients receiving liver-related celiotomy. The presence of pooling of contrast material within the hepatic parenchyma indicates free extravasation of blood as a result of active bleeding. In patients with blunt hepatic injury, if this sign is detected, nonoperative treatment should be terminated and angiography or celiotomy undertaken promptly. With the increasing use of high-speed spiral CT scanner and improvement in scanning technique, pooling of contrast material may become a sensitive sign for active bleeding and may be used as a guide for the selection of treatment modality.
    Article · Nov 1998 · The American Journal of Surgery
  • R J Chen · J F Fang · B C Lin · [...] · M F Chen
    [Show abstract] [Hide abstract] ABSTRACT: Most blunt hepatic trauma patients can be managed nonoperatively. The current failure rate in adult blunt hepatic trauma is reportedly 0 to 19%. We wished to evaluate the applicability of laparoscopy and fibrin glue as a minimally invasive alternative to laparotomy in these unsuccessfully nonoperative cases. All adult patients with blunt hepatic trauma managed nonoperatively at Linkou, Chang Gung Memorial Hospital Medical Center, Taipei, Taiwan, over a 2-year period from July 1, 1994, to June 30, 1996, were eligible for the study. A laparoscopic examination was performed on those who failed conservative care before undertaking an exploratory laparotomy. Fibrin glue was sprayed over the wound surface if ongoing hemorrhage was evident from any liver laceration. The clinical data, operative and laparoscopic findings, operative methods, and outcomes of these patients were studied. Of the 61 patients, 55 patients were successfully treated without operation. Of the six failures (10%) all were liver related. After the introduction of laparoscopy, the nontherapeutic laparotomy rate would have decreased from 100% (6 of 6) to 50% (3 of 6), and with the adjunctive use of fibrin glue, the laparotomy rate went down to 0% (0 of 6). There were no deaths among the six patients receiving laparoscopy and fibrin glues; and only one developed a liver abscess, for a morbidity rate of 17% (1 of 6). The selective use of laparoscopy and fibrin glue can effectively reduce the nontherapeutic laparotomy rate among blunt hepatic trauma patients who fail nonoperative management.
    Article · May 1998 · The Journal of trauma
  • [Show abstract] [Hide abstract] ABSTRACT: To evaluate the features and outcomes of delayed splenic rupture in the hope of developing a strategy to avoid future delayed diagnoses, all splenic injury patients treated over a 5-year period (1988-1992) in Chang Gung Memorial Hospital were reviewed, and 10 patients were found to meet the criteria for delayed splenic rupture. The average period of delay was 11.2 days. Eight had histories of trivial trauma, while the other two had histories of alcohol abuse and seizures. The diagnosis was made by abdominal ultra-sound in five, abdominal computed tomography in three, by diagnostic peritoneal lavage in one and on clinical grounds alone in the other. The average period of preoperative evaluation was 9.3 hours (4-17 hours). All underwent surgery and all required splenectomy. The average intraperitoneal blood volume at the time of laparotomy was 1870 ml. The average intraoperative blood loss was 110 ml. Differential diagnosis of left upper quadrant abdominal pain is limited. A history of trivial or questionable trauma, such as inebriation or seizure within the previous month, should always be elicited, followed by abdominal sonography or computed tomography whenever in doubt. The early diagnosis of delayed splenic rupture, allows for early surgery to reduce the amount of blood transfusion required.
    Article · Jan 1996