Reply: Assessing Risk Factors of Respiratory Complications following Abdominal Wall Reconstruction.

Abstract

Sir: We appreciate the thoughtful response by Dr. Xue et al. to our article.1 In their letter, three main issues were raised regarding the study’s methodology and patient assessment: (1) intraoperative neuromuscular blockade and perioperative opioid analgesia; (2) preoperative assessment of cardiac function; and (3) ventilator-induced lung injury and blood transfusion strategies. We agree with the comments and insightful critiques presented by Dr. Xue et al. Further clarification on these issues is warranted. Several important overarching issues are worth mentioning first. Ventral hernia repair represents an extremely common disease process that creates a tremendous burden and adverse functional impact for patients. Complex abdominal wall reconstruction for large hernias represents a unique procedure in which high rates of both surgical and medical morbidity can occur, despite the best efforts at optimizing patient selection, refined surgical techniques, and multidisciplinary perioperative collaboration. Our study aimed to better delineate the incidence, risk factors, and downstream healthcare impact of respiratory morbidity in these procedures. We acknowledge that retrospective study design is fraught with intrinsic limitations that impact the applicability (and potentially the validity) of our data. Our work does, however, offer a unique assessment of postoperative respiratory morbidity in an area of surgery in need of clinical outcomes research. We approach this analysis with complete transparency of our outcomes and with an interest in improving processes of care for our patients. Intraoperative neuromuscular blockade and opioid administration are performed at the discretion of the attending anesthesiologist of record for the case. Short-acting neuromuscular blockage agents may be utilized for induction (i.e., succinylcholine); however, intermediate-acting neuromuscular blockade agents (i.e., vecuronium) are utilized most commonly during the remainder of the case. Long-acting agents are almost never utilized, and all patients are fully reversed at the conclusion of the case. The senior author (J.P.F.) favors complete neuromuscular blockade during repair, such that any undue tension generated by abdominal wall contractile forces is mitigated. The scope of this article did not focus directly on the use or type of neuromuscular blockade, since this is typical of institutional practice pattern and as such was not individually assessed for each patient. Perioperative pain management, however, is a slightly more challenging issue. Studies have demonstrated that epidural use can significantly improve perioperative metrics following open intra-abdominal operations.2 We have recently examined this in our ventral hernia population,3 and have demonstrated that epidurals may be associated with reduced perioperative morbidity and cost following abdominal wall reconstruction specifically. As such, we advocate for their use in all patients deemed appropriate by our anesthesiology team. When epidurals are placed, typically a combination of fentanyl and bupivacaine is utilized for analgesia. No opioids are administered, and this pattern is continued postoperatively with patient-controlled epidural analgesia devices for at least 48 hours postoperatively, assuming a working epidural. However, in cases where epidurals are not placed or if the patient has an underlying opioid requirement, intraoperative and postoperative analgesia is achieved initially by intravenous opioids. Intraoperatively, opioids are administered as morphine equivalents based on changes in vital signs as monitored by the anesthesiologist. Postoperatively, patients utilize patient-controlled analgesia devices, most commonly utilizing morphine or Dilaudid, with an initial maximum dose of 1 mg or morphine or its equivalent every 10 minutes. If pain is uncontrolled, this dosage is titrated accordingly. Of note, epidural patients with an underlying preoperative opioid requirement receive only bupivacaine in their epidural, and do receive opioid patient-controlled analgesia. Our study did not directly assess pain or total opioid use following ventral hernia repair, but anecdotally we do note improved pain control with epidurals. This is an area necessitating further research that exceeded the scope of this current analysis. Regarding the preoperative assessment of cardiac function, in our preoperative history and physical, if a cardiac condition is noted or if there is significant concern, medical clearance from a cardiologist is obtained. It is therefore at the discretion of the cardiologist to obtain cardiac imaging to assess function (i.e., echocardiography). Certainly patients with known coronary artery disease underwent appropriate assessment and imaging. We do not routinely order cardiac imaging, but preoperative chest radiographs are obtained according to published recommendations,4 especially in older patients or patients with significant comorbidities.5 Comorbid patients almost uniformly require medical clearance. Lastly, ventilator-induced lung injury is certainly a concern in any patient undergoing general anesthesia. The ventilation strategies for this study were not standardized, but at our institution they were most commonly volume or pressure controlled. Ventilation rates and volumes were adjusted based on continuous end-title carbon dioxide monitoring. Furthermore, recruitment procedures were often performed, especially if a slight decrease in the oxygen saturation trend was noted. During the cases, although we maintain nearly continuous communication with our anesthesiology colleagues, we do not on a minute-to-minute basis control their methods of perioperative patient care. Finally, as with many procedures, the decision regarding intraoperative blood product transfusion is based on a constellation of factors: estimated blood loss, patient vital signs and laboratory data, and anticipation of further blood loss. We typically transfuse if a hemoglobin value approaches 7g/dl (or <10 g/dl if the patient has significant coronary artery disease) or if significant blood loss is encountered with patient instability. This, however, rarely occurs. In our study, we attempted to further delineate risk factors of respiratory complications, but we did not retrospectively collect data on every variable that relates to pulmonary function and perioperative risk. We believe our work represents an important step forward in optimizing the perioperative care for complex abdominal wall reconstruction patients. The importance of risk and outcomes transparency, along with a desire to constantly strive to improve patient care coupled with an open-mindedness about interdisciplinary collaboration with other surgical colleagues and anesthesiologists, will serve to better our outcomes. We appreciate Dr. Xue et al. for bringing further attention to several of these additional factors. Further research will better delineate perioperative risk following abdominal wall reconstruction, ideally in a prospective fashion. DISCLOSURE The authors have no financial interest to declare in relation to the content of this communication. John Patrick Fischer, M.D. Jonas A. Nelson, M.D. Stephen J. Kovach, M.D. Division of Plastic Surgery Hospital of the University of Pennsylvania Philadelphia, Pa.