Over the past five years, we were confronted with a couple of clinical cases, in which the grading of complications turned out to be difficult. Such cases were prospectively collected at our center.
A 63-year-old man underwent a right hemiliver resection for a hepatocellular carcinoma. The patient complained about right upper quadrant pain and developed a fever on postoperative day 4. A CT scan revealed an infected biloma in the right upper quadrant, which was drained percutaneously. After successful drainage of the biloma, the patient became acutely dyspnoeic due to a significant pneumothorax. A thorax drain was inserted to treat the pneumothorax.
Complication After a Surgical Procedure Not Caused by a Surgeon
This scenario describes a postoperative complication that is not caused by the surgical team responsible forthe patient; this may include complications after an intervention by the radiology or gastroenterology team. Although the “causative” physician is not a surgeon, the complication would not have occurred if the patient had not undergone surgery. Therefore, such postoperative events should be recorded irrespective of the team involved in caring for the patient.
The correct labeling of both complications presented in this scenario was biloma (grade IIIa) and pneumothorax (grade IIIa).
A right hemihepatectomy was performed elsewhere due to a metastatic disease in a 75-year-old woman. Then, the patient developed signs of sepsis on postoperative day 3. The same team performed a relaparotomy, revealing a perforation of the small intestine, which required a partial colonic resection. The patient showed persisting signs of sepsis and hemodynamic instability, and was later referred to our center. Due to the deterioration of the patient’s condition, including respiratory and renal failures, the patient was reintubated. A third laparotomy disclosed an insufficiency of the intestinal anastomosis, which was corrected with a new anastomosis. An intraperitoneal VAC system had to be applied because of the retraction of the abdominal wall. Subsequently, several laparotomies were necessary over the next few weeks because of recurrent leakages in the small intestine. Finally, the patient expired in our ICU after 25 days.
Complication Occurring in a Patient Transferred After a Surgical Procedure Performed in Another Center
This patient was referred to a surgical unit from another hospital due to a complicated postoperative course. Typically, such patients are in poor conditions, yielding a high risk of developing additional problems. The surgical unit may be tempted to omit recording such events because the initial surgery including a rough postoperative course, originates from another team. The consensus at the University of Zurich was that all complications must be properly recorded. However, for a quality assessment of the surgical unit, the patients should be identified as transferred from another unit. This team used the addendum “referred patient” in brackets after the complication grade to indicate that the patient was initially not treated in the center reporting the outcome.
The grading in this case was grade V (referred patient).
A 57-year-old patient underwent an elective anterior sigma resection for a cancer. The patient developed excruciating abdominal pain, a high-grade fever, and leukocytosis 5 days after surgery. The patient was taken back to the OR, where an anastomotic breakdown was documented. A Hartmann’s procedure was performed. Then, the patient was admitted to the ICU, and developed an ARDS and renal failure requiring dialysis over the next few days. A number of other problems occurred including a wound infection treated at bedside and deep vein thrombosis, requiring anticoagulation. The patient died 17 days later.
Patients Developing Complications of Increasing Severity Depending One From Each Other
This scenario illustrates cases with the sequential development of a complication gradually becoming more severe. Another example would be a reoperation of a small bowel obstruction, associated with a bronchoaspiration of gastric content and, eventually, death. The question is whether each grade of the same line of complications must be recorded separately, or only the most severe ones. The University of Zurich team records only the most severe grade, when a complication clearly occurs as consequence from a prior, less severe complication.
In this case, a grade I (wound infection), a grade II (deep vein thrombosis), and a grade V complication (anastomotic dehiscence leading to death) were recorded because those complications do not depend on each another.
The same scenario as in scenario 3, but the patient recovered, and was, finally, able to leave the hospital after 3 months of hospitalization. He was very frail and could hardly walk. He was transferred to a rehabilitation home.
Complication Still Present at the Time of Discharge or at Follow-Up Visits
In this case, the patient remains in a compromised health condition at the time of discharge, as a result of his shaky postoperative course and long hospital stay. The suffix “d” (disabling) was introduced in the classification to catch persistent complications at the time of discharge or at follow-up visits, which often raised the question which of the conditions may qualify for this suffix. A frail health status after a long course does not qualify for the label d. Conditions should only be labeled as such, if they describe a specific complication. For example, paresis of a hand after surgery, persisting wound dehiscence etc.
A 92-year-old patient presented with severe acute cholecystitis. Although the patient also suffered from severe cardiovascular disease and was considered as ASA IV, a laparoscopic cholecystectomy was performed. During surgery, the patient died of cardiac arrest.
Intraoperative Death of the Patient
This scenario describes an intraoperative death. The classification of 2004 was developed to record postoperative complications, recognizing that intraoperative complications, for example, a bleeding or a tear on the small bowel, is often difficult to define, and, therefore, is unlikely to be reported unless it leads to postoperative negative events. The team in Zurich had, however, considered that death of a patient during surgery must be an exception, and must be caught by the classification.
Hence, a grade V complication was recorded in such a scenario.
The same patient as in scenario 5, but the patient died before surgery, during intubation.
Death of a Patient Before Surgical Intervention
In this scenario, the patient dies during the intubation attempt, that is, before skin incision. Here, the question arises of whether a complication that occurs before a surgical intervention should be recorded as a surgical complication. In our opinion, each complication that occurs in a surgical patient has to be recorded in an “intention-to-treat” manner. Complication in the preparation for surgery, once the indication is established, should, therefore, be recorded irrespective of whether surgery has been performed or not.
This scenario therefore describes a grade V complication.
A 60-year-old diabetic patient underwent a rectal resection for adenocarcinoma in another facility. The patient had a long-term corticosteroid therapy due to a rheumatic disease. The patient developed abdominal pain, fever and leukocytosis 6 days later. A relaparotomy was performed, revealing an anastomotic insufficiency. The small pelvis was drained, and a protective ileostomy was performed. The patient was subsequently referred to us because of hemodynamic instability, but the patient died 3 days later in our ICU due to a fulminant sepsis.
Death of a Patient After an Operation Performed in Another Hospital
Similarly to case 2, the focus, here, is on a lethal outcome after the transfer of a patient operated elsewhere.
In such a situation, the fatal outcome must be recorded in both centers, and the patient should again be properly labeled as “initially operated elsewhere” (grade V, referred patient).
A 75-year-old patient underwent an aortic aneurysm repair and developed severe abdominal pain a few days after surgery. We were called to evaluate the patient for suspicion of an ischemic bowel. Despite a negative clinical and radiologic evaluation, we performed an exploratory laparotomy, but failed to document any intraabdominal pathology.
Patient Undergoing an Explorative “Blank” (Negative Laparotomy)
This is a relatively common scenario, in which the suspicion of an abdominal complication justified a laparotomy; but, the laparotomy is unrevealing and the patient subsequently recovers uneventfully. Should such a laparotomy be graded as IIIb complication or not be considered as a complication at all? Although such an intervention was a consequence of the former surgical intervention (aortic aneurysm repair), it was purely diagnostic. As the basis of the complication classification is the therapy required to treat a complication, blank laparotomy should not be considered a complication. This is also important to prevent proper management of the patient, as it might become tempting to avoid an invasive diagnostic procedure to minimize the reported morbidity.
After an uneventful hernia repair, a 28-year-old man was discharged 2 days after the procedure. Four days later, he came to the emergency room complaining of epigastric pain. He had no fever and laboratory results were normal. We performed an upper GI endoscopy, which revealed an uncomplicated duodenal ulcer. We prescribed a proton pump inhibitor and the patient was discharged on the same day.
Patient Developing a Medical Problem Not Obviously Related to Surgery
This scenario illustrates negative events occurring in the course of a surgical intervention that are not or only remotely related, to surgery. The correlation between such an event and surgery is, however, sometimes difficult to assess. In this case, the duodenal ulcer might or might not be due to the former surgical intervention. To prevent subjective interpretation, the Zurich team considered that any negative event occurring in a patient during hospital stay or within 30 days after surgery (whichever is longer), should be regarded as a surgical complication and should, therefore, be recorded; the labeling was therefore "grade II."
Our colleagues from gynecology called us to consult a patient in the operating room. They were involved in the relaparotomy of a 57-year-old woman, undergoing a Wertheim’s operation and found a severe peritonitis, with a perforation in the small intestine. We performed a segmental resection of the small intestine with end-toend anastomosis. Two days later, a relaparotomy was indicated again, due to the suspicion of an anastomotic leakage. We were called again and could resew a small leak at the anastomotic site. The patient eventually recovered fully. The patient remained in the gynecologic ward for the entire stay.
Complications After Procedures
This case presents a patient that was initially operated in the gynecology service, and then transferred to the surgical ward for a complicated course. An iatrogenic perforation of the colon occurred during the initial operation, which required further management by the surgical team. This is another scenario sharing mechanisms similar to cases 2 and 7. Such complications must be recorded irrespective of the origin or location of the patient; hence, the Zurich team recorded it as a grade IIIb (referred patient) in the displayed scenario.
A 37-year-old woman was operated on the thyroid due to cancer. On day 3, the patient developed severe pain in the left groin. A clinical evaluation revealed an incarcerated femoral hernia, for which the patient received an emergency operation. The following course was uneventful.
Patient With a Medical Problem Not Related to Surgery
In this scenario, the patient develops a medical problem unrelated to the previous surgery (incarcerated hernia after thyroid surgery). Despite the doubtful correlation, this scenario was evaluated with the same principle as in case 7 and considered it a grade IIIb complication. The relationship between surgery and an adverse medical event is often speculative and relies on subjective interpretation. Therefore, we reached a consensus to record all the negative events, which have occurred in a patient after surgery, irrespective of whether there was or was not a clear correlation with a former surgical intervention.
Clavien PA, Barkun J, de Oliveira ML, et al. The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg 2009; 250(2):187-196.