History of the Innovation of Damage Control for Management of Trauma Patients: 1902-2016.

OBJECTIVE: To review the history of the innovation of damage control (DC) for management of trauma patients. BACKGROUND: DC is an important development in trauma care that provides a valuable case study in surgical innovation. METHODS: We searched bibliographic databases (1950-2015), conference abstracts (2009-2013), Web sites, textbooks, and bibliographies for articles relating to trauma DC. The innovation of DC was then classified according to the Innovation, Development, Exploration, Assessment, and Long-term study model of surgical innovation. RESULTS: The "innovation" of DC originated from the use of therapeutic liver packing, a practice that had previously been abandoned after World War II because of adverse events. It then "developed" into abbreviated laparotomy using "rapid conservative operative techniques." Subsequent "exploration" resulted in the application of DC to increasingly complex abdominal injuries and thoracic, peripheral vascular, and orthopedic injuries. Increasing use of DC laparotomy was followed by growing reports of postinjury abdominal compartment syndrome and prophylactic use of the open abdomen to prevent intra-abdominal hypertension after DC laparotomy. By the year 2000, DC surgery had been widely adopted and was recommended for use in surgical journals, textbooks, and teaching courses ("assessment" stage of innovation). "Long-term study" of DC is raising questions about whether the procedure should be used more selectively in the context of improving resuscitation practices. CONCLUSIONS: The history of the innovation of DC illustrates how a previously abandoned surgical technique was adapted and readopted in response to an increased understanding of trauma patient physiology and changing injury patterns and trauma resuscitation practices.

Resuscitation for injured patients requiring massive transfusion: A personal perspective.

ABSTRACT: The past century has seen many advances in the field of resuscitation. This is particularly true in the subset of patients who sustain major injuries causing hemorrhagic shock and require massive transfusion over 10 units of blood within the first 24 hours. Controversies on how best to resuscitate these patients include the role of fresh whole blood (WB), stored WB, fresh frozen plasma (FFP), platelets (PLTS), colloid solutions, balanced electrolytes solution, vasopressors and diuretics. This review summarizes the often-contradictory recommendations observed and studied by a single trauma surgeon working in a busy urban acute care center for 65 years.

Directed cutaneous neurectomy for cutaneous neuralgia.

BACKGROUND: Cutaneous neuralgia (CN) is a common challenge for surgical consultation. This report describes directed cutaneous neurectomy (DCN) for persistent CN. METHODS: From 2010 through 2022, DCN was performed 112 times in 100 patients. All had complete temporary relief of CN by outpatient percutaneous proximal blockade. DCN involved a successful proximal blockade with blue dye added to the injectate, and all blue stained tissue was excised. The site of DCN included groin (49 patients), abdomen (38 patients), chest (7 patients), extremity (4 patients), or skull (2 patients). Relief was judged continuous (C), none (N), or temporary (T). RESULTS: Pain relief was C in 82 patients (27 â€‹± â€‹20 â€‹mo), N in 6 patients, and T in 12 patients (22 â€‹± â€‹2 â€‹mo). The presence of microscopic nerve fibers (46 patients) or mesh (42 patients) did not affect outcome. A second DCN was done in two N patients, followed by C relief. A second DCN was done in seven T patients, and a third DCN was done in three T patients after recurrent CN. CONCLUSIONS: Refractory CN can usually be successfully treated by DCN.

The significance of a serum creatinine in defining renal function in seriously injured and septic patients.

BACKGROUND: Most current analyses of multiple organ failure after injury use the serum creatinine (SCr) as a surrogate for defining renal insufficiency (RI) or renal failure (RF). This study correlates SCr with glomerular filtration rate, renal perfusion, and renal excretion in injured and septic patients. METHODS: The 289 injured patients were in shock for an average of 32 minutes and received an average of 13.9 transfusions by the end of the operation. The 34 septic patients were in shock for an average of 23 minutes and received 8.4 crystalloid during operation. The renal studies included (SCr), inulin clearance (CIn), and creatinine clearance (CCr), renal plasma flow (CPAH), renal blood flow, and the clearance of sodium (CNa++), osmolar clearance (COsm), and urine output. All clearance studies followed the classic methodologies described by Homer Smith, including weight-guided leading dose, steady-state serum levels, and urine collections made exactly 15 minutes after serum collections. RESULTS: The average SCr in 289 trauma and septic patients was 1.23 mg/dL and 1.3 mg/dL, respectively. The average CCr was 106 mL/min and 103 mL/min, whereas the average CIn was 96 mL/min and 95 mL/min, respectively. The CIn correlated (p<0.0005) with CCr in all patients, whereas the CIn was lower than CCr due to the tubular excretion of creatinine. For the group of patients with RI (CIn between 10 and 30 mL/min) and nonoliguric RF (CIn<10 mL/min), the average CCr was 3.1. Other values in this subgroup included an average CCr 23.6 mL/min, CIn 14.6 mL/min, CPAH 69.9 mL/min, renal blood flow 138 mL/min, CNa 0.7 mL/min, COsm 1.5 mL/min, and urine output 1.4 mL/min. Although nephrectomy in 15 of 36 patients with renal injury or death in 21 patients was associated with a higher SCr, the relationship between SCr and renal function studies remained the same as with survivors and patients without renal injury. The best SCr value for defining RI was 2.4 mg/dL and for RF was 3.1 mg/dL. CONCLUSION: Based on these findings, one can recommend that when SCr data are extracted from large trauma registries, the definition of RI should be inferred when the SCr exceeds 2.4 mg/dL, and RF should be diagnosed when the SCr exceeds 3.1 mg/dL.

Congenital pancreaticobiliary anomalies in an urban medical center in the United States.

BACKGROUND AND AIM: Pancreaticobiliary anomalies are rare and often present with cryptic signs and symptoms, thus delaying appropriate treatment. METHODS: Endoscopic retrograde cholangiopancreatography (ERCP) was used to define pancreaticobiliary anomalies. A retrospective review was performed of 5522 ERCPs conducted at a tertiary care center from 1972 to 2015. RESULTS: There were 249 (4.5%) patients with pancreaticobiliary anomalies, including 179 patients with pancreas divisum (PD), 44 patients with choledochal cyst (CC) (Todani's classification Type I: extrahepatic cyst 31, Type III; choledochocele 9, Type V: Caroli's disease 4), 20 patients with anomalous pancreaticobiliary ductal union (APDU), and 6 patients with other abnormalities. Of 179 patients with pancreas divisum, 8 (4.5%) required minor sphincterotomies for multiple unexplained acute pancreatitis. Of the 31, 15 (48%) Type I CC patients underwent an operation. In patients with Type III CC (choledochocele), seven of the nine were treated by endoscopic sphincterotomy, and two patients were treated by surgery. Four patients with Type V CC (Caroli's disease) were managed nonoperatively. Of the 20 patients with APDU, 8 (40%) required operative intervention. Six patients were found to have other anomalies: two with pancreas bifidum, one with a duplication of the gallbladder, one with a cystic duct diverticulum, one with an annular pancreas, and one with an abnormal cystic duct origin. These patients were treated based on their etiology. CONCLUSION: Pancreaticobiliary anomalies are rare and can be defined using ERCP. The appreciation of these abnormalities is important for the proper diagnosis and treatment of these rare biliary and pancreatic disorders.

Nonoperative Management (NOM) of most liver injuries impairs the mastery of intraoperative hemostasis.

BACKGROUND: Nonoperative management (NOM) of most liver injury (LI) compromises teaching of technical skills required for intraoperative LI hemostasis. This study assesses this void. METHODS: The records of patients (pts) admitted for LI during two years (1/1/16-12/31/17) were compared to pts treated during two-year intervals for the last six decades. Treatment included NOM, operation only (OR/only), suture (Sut), tractotomy (Tra), dearterialization (HAL), and resection (Res). RESULTS: During 2016/2017, 41 pts had penetrating (23) or blunt (18) LI. Treatment for penetrating LI was NOM (4), OR/only (12), and hemostasis (7) with Sut (3), HAL (1), Tra (1), and Res (2). Treatment for blunt LI was NOM (16) and OR/only (2). 14 residents performed an average of 0.5 procedures. During six decades, LI requiring hemostasis was 121, 114, 30, 48, 17, and 7 per decade. Concomitantly, the percent having NOM or OR/only was 46%, 47%, 62%, 59%, 72%, and 83%. CONCLUSION: NOM precludes adequate training for hemostasis of LI. Technical proficiency for LI hemostasis requires training in Advanced Trauma Operative Management (ATOM), Advanced Surgical Skills for Exposure in Trauma (ASSET), and rotation through a liver transplant or hepatobiliary service.

Acute hemorrhagic rectal ulcer: Experience in 11 patients at an urban acute care center in the USA: A case series.

INTRODUCTION: Acute hemorrhagic rectal ulcer (AHRU) is a rare entity which has most frequently been described in Japan and Taiwan literature. This study characterizes 11 AHRUs identified and managed at an urban acute care hospital in the United States of America (USA). METHODS: A total of 2253 inpatients underwent colonoscopy. In 1172 patients (52%), colonoscopy was performed for evaluation of lower gastrointestinal (LGI) bleeding. Eleven (0.9%) of the 1172 patients with LGI bleeding had AHRU. RESULTS: AHRU is characterized by a sudden onset of painless and massive lower rectal bleeding in elderly, bedridden patients (pts) with major underlying diseases. The endoscopic findings were classified into 4 types. All 11 ulcers were located in the distal rectum within 10 cm of the dentate line. All 11 patients required blood transfusion (mean = 3.7 units; range 2-9 units). Seven patients responded to blood, plasma, and platelet transfusions. The other 4 patients required endoscopic hemostasis.Three patients died within a month of colonoscopy from comorbidities. None had bleeding as a cause of death. Eight surviving patients did not have recurrent bleeding. CONCLUSION: AHRU does exist in the USA and should be considered as an important cause of acute lower GI bleeding in elderly, critically ill, and bedridden patients. AHRU should be recognized and managed correctly.

The academic challenge of teaching psychomotor skills for hemostasis of solid organ injury.

BACKGROUND: This study highlights the inherent challenges of achieving psychomotor skills in an era of nonoperative therapy for solid organ injuries. Technical procedures on the liver, the most frequent intra-abdominal solid organ injured, were assessed in five decades. METHODS: Guided by prospective assessment and registry data, all patients with liver injury seen during 24 months in five consecutive decades were reviewed. Initially (1960s), all injuries were explored; currently (2000s), most injuries are observed. The number of patients was 235 (1960s), 228 (1970s), 79 (1980s), 116 (1990s), and 64 (2000s). The greater number in the 1990s reflects the diagnosis of minor, clinically insignificant, blunt injuries after abdominal CAT scan became available. Each injury was categorized by cause, severity (Abbreviated Injury Scale), associated shock, and primary therapy (observe [OBS], operation alone [OR], hepatorrhaphy [SUT], tractotomy [TRACT] with intraparenchymal hemostasis, hepatic dearterialization [HAL], and resection [RESECT]). Packing, used in each decade, was placed in one of the above primary treatment groups. RESULTS: The primary techniques for hemostasis are shown in the text table.Shock and Abbreviated Injury Scale correlated with mortality averaged 16%; 40 of 116 deaths (34%) exsanguinated from hepatic injury. During training, a resident performed an average of 12.0, 12.0, 2.4, 4.0, and 1.3 procedures for hemostasis. CONCLUSIONS: Reduced incidence and decreased therapeutic laparotomies for liver injury have created a training vacuum for future trauma surgeons. Surgical residents will need to supplement their clinical experience with solid organ hemostasis by practice on appropriate animal models of injury and cadaver dissections.

Systemic not just mesenteric lymph causes neutrophil priming after hemorrhagic shock.

BACKGROUND: Inflammatory mediators in postshock mesenteric lymph have been causally linked to systemic polymorphonuclear cells (PMNs) priming resulting in acute lung injury (ALI) and multiple organ failure. Earlier human and animal studies demonstrated ALI after lower limb ischemia/reperfusion (I/R) injury. As hemorrhagic shock (HS) is in essence a systemic I/R insult, we postulated that systemic lymph after HS would exhibit PMN priming and this was studied in vitro. METHODS: Lymph was collected at intervals from the hind limb of dogs subjected to sham or HS and crystalloid resuscitation. Human PMNs isolated from heparinized blood of normal volunteers were incubated with buffer, sham lymph, or lymph after 120 minutes of shock or resuscitation. PMN priming was indexed by CD11b expression (mean fluorescence intensity), superoxide anion (O2(-)) generation (nanomoles/mg protein), and elastase release (%) after the addition of fMLP (1 micromol). PMNs with buffer served as control. RESULTS: PMN priming after exposure to either shock or postshock resuscitation lymph was noted by increased expression of CD11b, superoxide generation, and elastase release after exposure to fMLP. No priming effect was noted with sham lymph. Maximal bioactivity of shock or postresuscitation shock lymph was noted at 2 hours postresuscitation. CONCLUSIONS: Exposure with systemic lymph after HS resulted in PMN priming. These results question the unique properties attributed to post-HS lymph from the splanchnic bed in causing PMN priming and ALI after shock. The causal agent(s) for these effects are unclear.

The urban injury severity score (UISS) better predicts mortality following penetrating gunshot wounds (GSW).

BACKGROUND: The Injury Severity Score (ISS) and the New ISS (NISS) underscore injury severity after GSW. This study assesses the Urban ISS (UISS), which incorporates all injuries. METHODS: Complete trauma program registry (TPR) data and chart analyses were performed on 585 patients (pts) over 28 months. Factors analyzed included age, gender, ISS, NISS, UISS, time of admission, intent of injury, race, number GSW, weapon, and outcome. RESULTS: The 585 patients could be categorized within three groups. The first group included 98 pts with low ISS (1-2), no organ injuries, and early discharge; the second group included 47 patients with severe shock who died during operation; the third group of 442 pts were admitted after operation. All injury scores correlated (p < 0.001) with assault, number GSW, death, and length-of-stay (LOS). Death and LOS correlated closely with assault and the resultant number of GSW, best seen with UISS compared to ISS or NISS. Race and admission time did not correlate with death or LOS. CONCLUSIONS: UISS correlates better than ISS and NISS in victims of inner-city firearm injuries.

Clinical evaluation and management of caustic injury in the upper gastrointestinal tract in 95 adult patients in an urban medical center.

BACKGROUND: Caustic ingestion causes a wide spectrum of injuries; appropriate treatment varies according to the severity and extent of the injury. This retrospective study of adult patients with caustic injury presents the endoscopic findings, treatment regimen, and clinical outcome. METHODS: Over a 28-year period, 95 consecutive adult patients admitted to an urban emergency hospital for ingestion of caustic materials were studied. Each patient underwent early endoscopy and the injury was graded for severity. There were 61 men and 34 women with an average age of 37.2 years (range 17 to 81). Ingestion was due to a suicide attempt in 49 patients and accidental in 46 patients. RESULTS: Ten patients showed no mucosal damage. The remaining 85 patients had grade I superficial injury in 47 patients, grade II moderate injury in 25 patients, and deep grade III injury in 13 patients. The ingestion of strong acid or strong alkali often produced deep grade III changes while bleach, detergent, ammonia or other substances usually caused grade I injury. Operative interventions were required for 11 patients with grade III injury and 6 patients with grade II injury. Endoscopic grading was predictive for the onset of complications including late esophageal stricture. There were no complications due to endoscopy; one patient with grade III and multiple comorbidities died from multiple organ failure. CONCLUSION: Upper gastrointestinal endoscopy after caustic ingestion should be performed early to define the extent of injury and guide appropriate therapy. Grade I injuries heal spontaneously. Grade II injuries may be treated conservatively but repeat endoscopy helps define when intervention is needed. Grade III injuries ultimately require surgical intervention.

Variability of the f-cell ratio after treatment of traumatic hemorrhagic shock.

PURPOSE: Measuring total blood volume (TBV) in critically ill patients, using isotope techniques to determine red cell volume (RBCV) and plasma volume (PV) is laborious. Recently, PV measurement using a single bolus dose of tracers has been validated, thus, allowing TBV calculation using large venous hematocrit (LVHCT). However, this technique relies on using a correlation factor, the f-cell ratio, to adjust for differences in LVHCT and total body hematocrit (TBHCT). The normal f-cell ratio is 0.9 but has never been studied in patients recovering from hemorrhagic shock (HS). This study assesses the f-cell ratio at different phases after HS to determine accuracy in calculating TBV. METHODS: 114 injured patients requiring immediate operation for HS were studied. All patients had measurements of PV and RBCV via isotope dilution enabling measurements of TBHCT. Correlation of LVHCT and TBHCT were used to calculate the f-cell ratio in the fluid sequestration (n = 81) and in the fluid mobilization period (n = 108). RESULTS: The f-cell ratio (mean ±â€¯SD) averaged 0.89 ±â€¯0.15 and 0.90 ±â€¯0.01 in the first and second halves of the fluid sequestration period versus 0.90 ±â€¯0.2 and 0.80 ±â€¯0.07 in the first and second 48 h of the fluid mobilization period. The f-cell ratio was significantly lower (p=<0.001) in the mobilization period. CONCLUSIONS: These data show for the first time that using PV and LVHCT to calculate TBV after HS is unreliable. The mechanisms causing this variation in f-cell ratio is unknown but likely related to capillary/interstitial dynamics and needs further scientific study.

Failure of gastric prophylaxis against stress bleeding reflects inadequate pH control.

BACKGROUND: Gastric prophylaxis against stress ulceration with histamine 2 blockers or protein pump inhibitors is a quality standard in septic Surgical Intensive Care Unit (SICU) patients to reduce gastric pH below 3.5. This study assesses the efficacy of gastric prophylaxis with pH monitoring. METHODS: A total of 100 patients were studied for 481 days. All received histamine 2 blockers or protein pump inhibitors. Daily pH control was judged as poor (1-3), moderate (4-5), or good (>5). Patients with poor pH received double-dose or an infusion gastric prophylaxis. Nasogastric tube bile or blood and transfusions for stress ulceration were recorded. RESULTS: Gastric prophylaxis was poor for 37 days, moderate for 83 days, and good for 279 days. NGT blood occurred on 15 days (40%) with poor, 17 days (20%) with moderate, and 17 days (6%) with good control. Transfusions for stress ulceration occurred on 5 days (14%) during poor, 3 days (4%) during moderate, and 1 day (0.3%) during good control (P < .05). Enhanced gastric prophylaxis improved pH control and reduced nasogastric tube blood. Transfusion for stress ulceration after enhanced therapy was required on 1 day (8%) with poor control and never for moderate or good control. CONCLUSION: Gastric prophylaxis against stress ulceration should be monitored by nasogastric tube pH.

Kindness kills: the negative impact of pain as the fifth vital sign.

BACKGROUND: The current emphasis on pain assessment as the fifth vital sign and the use of unscientific pain scales is causing serious injury and death from overmedication. STUDY DESIGN: This premise was tested by reviewing the case reports of all trauma center site surveys performed by the authors for the American College of Surgeons Committee on Trauma verification program during 2 separate time periods: 1994 through 1998 and 2000 through 2004. A total of 2,907 and 2,282 reports summarized by one of the authors, plus a total of 53 and 50 other reviewers, respectively, were analyzed from the records of 120 and 94 trauma centers. Most patients were men (71% and 66%) and had sustained blunt injury (83% and 79%). Average age was 35 years for both periods, with a range of 3 weeks to 97 years and 3 days to 98 years, respectively. The most common injuries involved head (33% and 34%), chest (13% and 13%), abdominal (22% and 21%), orthopaedic (18% and 18%), or multiple (9% and 14%). There were 1,459 and 867 deaths, respectively; all had a multidisciplinary peer review. RESULTS: Overmedication with sedatives/narcotics, during the two periods, clearly contributed to deaths in 13 and 32 patients and probably contributed to deaths in 5 and 14 patients, respectively. This occurred in 17 and 43 patients, respectively, after blunt injury and in 1 and 3 patients, respectively, after penetrating injury. Two clinical scenarios predominated, ie, overmedication in preparation for an imaging study and overmedication after discharge from ICU to the floor. The sequel of hypotension and compromised airway requiring intubation initiated a cascade of negative events that led to death. One patient in each period died as a result of prehospital overmedication. CONCLUSIONS: The current assessment of pain by computer-stored pain scales is in a state of imbalance, with excessive emphasis on undermedication at the same time ignoring overmedication. This imbalance reflects pain-service attempts to comply with external accrediting agencies. This preventable cause of death and disability in trauma patients is also occurring in noninjured patients. Surgeons must correct this problem by insisting on a balanced assessment of overmedication versus undermedication.

Life-threatening tertiary hyperparathyroidism in the critically ill.

BACKGROUND: Tertiary hyperparathyroidism typically occurs in patients who have recovered from renal failure after renal transplantation. This report describes a syndrome of tertiary hyperparathyroidism after recovery from multiple organ failure (MOF) with acute oliguric renal failure (AORF). METHODS: Six patients with MOF including AORF are presented. Increased parathyroid hormone (PTH) levels were documented as early as 3 weeks after injury or septic insult and remained increased in some patients for several weeks. RESULTS: The resultant increase in calcium levels led to recurrent bouts of bradycardia, often leading to asystole requiring cardiopulmonary resuscitation. Hypercalcemic-induced bradycardia was refractory to hydration, loop diuresis, atropine, and external pacing. Definitive treatment requires bisphosphonate therapy, which must be repeated until organ function has returned to normal. CONCLUSIONS: A new syndrome of life-threatening tertiary hyperparathyroidism is described in patients with critical illness. This syndrome probably is being overlooked frequently in critical care units. Early diagnosis and prophylactic treatment with bisphosphonate may preclude the life-threatening cardiac arrhythmias.

The diagnostic and surgical challenges of massive localized lymphedema.

BACKGROUND: Massive localized lymphedema (MLL) is a rare entity first described in 1998 in patients with morbid obesity; the incidence is rising with the increased prevalence of morbid obesity. This report defines the clinical presentation and surgical challenges in 6 patients with MLL. METHODS: The MLL in 6 patients with morbid obesity (weight range 270 to 585 lbs) involved the thigh in 3 patients, the calf in 1 patient, and the abdomen in 2 patients. The time from onset to presentation averaged 3 years (range 1 to 8 years). Two thigh lesions precluded ambulation because both legs could not be on the ground simultaneously; the 2 abdominal lesions were too heavy to permit ambulation. RESULTS: The surgical excision required the use of pulleys to elevate the MLL tissues, which, on excision, weighed between 24 and 78 lbs. A long oval horizontal incision and a long transverse incision were used for the 2 abdominal lesions. Long horizontal oval limb incisions with multiple perpendicular cross incisions had to be used to excise MLL in the 4 limb lesions. In 2 cases, the vessel-sealing device was employed successfully for dissecting subcutaneous edematous tissue. Loose wound closure permitted postoperative lymph leakage, which continued for 3 to 8 weeks. The histology demonstrated fibrotic lymphatic tissue with vascular and lymphatic proliferation and edema; all patients did well. CONCLUSIONS: MLL is rare and is best treated by surgical excision facilitated by pulleys and imaginative incisions to obtain primary closure. Long-term follow-up is necessary to assess for subsequent liposarcoma or angiosarcoma.

Etiology, treatment, and outcome of esophageal ulcers: a 10-year experience in an urban emergency hospital.

Esophageal ulcers are a rare cause of upper gastrointestinal bleeding. This report describes the etiology, treatment, complications, and outcome of esophageal ulcers. An esophageal ulcer is defined as a discrete break in the esophageal mucosa with a clearly circumscribed margin; esophageal ulcers were seen in 88 patients from a total of 7,564 esophagogastroduodenoscopies done by one surgeon at an urban hospital from 1991 to 2001. All hospital reports were reviewed. The etiology of esophageal ulcers included the following: gastrointestinal reflux disease (GERD) (n=58, 65.9%), drug induced (n=20, 22.7%), candidal (n=3, 3.4%), caustic injury (n=2, 2.3%), and herpes simplex virus (HSV), human immunodeficiency virus (HIV), marginal ulcer, foreign body, and unknown etiology (n=1 of each, 1.1%). The mean size of GERD-induced esophageal ulcers and drug-induced esophageal ulcers was 2.78 and 2.92 cm, respectively; 80.3% of GERD-induced esophageal ulcers and 13.8% of drug-induced esophageal ulcers were located in the lower thoracic esophagus. Morbidity (n=44, 50%) included hemorrhage (n=30, 34%), esophageal stricture (n=11, 12.5%), and esophageal perforation (n=3, 3.4%). Nonoperative therapy sufficed in 81 patients (92%). Three patients (3.4%) had a recurrence of esophageal ulcers. Fifteen patients (17.0%) required endoscopic intervention including esophageal dilatation for stricture in 11 patients and endoscopic hemostasis for esophageal bleeding in four patients. Surgery (n=7, 8.0%) was reserved for esophageal stricture and perforation. Two patients (2.3%) died from complications of esophageal ulcers: hemorrhage in one and perforation in one. Three patients died of their primary disease. GERD and drug ingestion are common causes of esophageal ulcers. Midesophageal ulcers have a greater tendency to hemorrhage compared with ulcers at the gastroesophageal junction; this may reflect the etiology. Strictures complicate GERD-induced esophageal ulcers but not drug-induced esophageal ulcers. Esophageal dilatation is an effective treatment for most strictures associated with esophageal ulcers. Esophageal ulcers rarely cause death.