The Association of Postoperative Complications and Hospital Costs Following Distal Pancreatectomy

Background Understanding the financial implications associated with the complications post-distal pancreatectomy (DP) may be beneficial for the future optimisation of postoperative care pathways and improved cost-efficiency. The primary outcome of this retrospective study was the characterisation of the additional cost associated with postoperative complications following DP. The secondary outcome was the estimation of the prevalence, type and severity of complications post-DP and the determination of which complications were associated with higher costs. Methods Postoperative complications were retrospectively examined for 62 adult patients undergoing distal pancreatectomy at an Australian university hospital between January 2012 and July 2021. Complications were defined and graded using the Clavien–Dindo (CVD) classification system. In-hospital cost of index admission was calculated using an activity-based costing methodology and was reported in US dollars at 2021 rates. Regression modelling was used to investigate the relationships among selected perioperative variables, complications and costs. Results 45 patients (72.6%) experienced one or more postoperative complications. The median (IQR) hospital cost in US dollars was 31.6% greater in patients who experienced complications compared to those who experienced no complications ($40,717.8 [27,358.0–59,834.3] vs. $30,946.9 [23,910.8–46,828.1]). Costs for patients with four or more complications were 43.5% higher than for those with three or fewer complications (p = 0.015). Compared to patients with no complications, the median hospital costs increased by 17.1% in patients with minor complications (CVD grade I/II) and by 252% in patients who developed major complication (i.e., CVD grade III/IV) complications. Conclusion Postoperative complications are a key target for cost-containment strategies. Our findings demonstrate a high prevalence of postoperative complications following distal pancreatectomy with number and severity of postoperative complications being associated with increased hospital costs. (Registered in the Australian New Zealand Clinical Trials Registry [No. ACTRN12622000202763]).

and Figure 1 below, summarises the association of costs severity of POPF. Interestingly, and not unexpectedly, there were no clinically or statistically significant increases in cost with a Grade 1 POPF compared to patients with no complications (p=0.99), given that the definition of a Grade is "no clinical symptoms but higher drain amylase levels". However, patients with Grade B POPF and Grace C had a clinically important increase in costs when compared to patients without any complication.
We have now included this addition Figure in the manuscript. Thank you once again for this excellent suggestion.   We agree that laparoscopic distal pancreatectomy should theoretically provide the same postoperative recovery advantages reputed to minimal access surgery, however our unit has had concerns as to the safety of laparoscopic distal pancreatectomy in terms of life-threatening intraoperative events, adequate oncological outcomes as compared to the traditional "open" distal pancreatectomy, as well as the safety challenges with a "minimal access approach" is the very obese, the elderly or the frail.
In our centre, we do consider that the indications for laparoscopic distal pancreatectomy are indeed very similar for open. We tend to consider laparoscopic distal pancreatectomy for benign, borderline, or malignant tumours of the pancreatic body and tail, however we advocate for open procedures when there are concerns from imaging that there may be invasion of surrounding organs or critical vasculature, distant metastasis in cancer, or acute pancreatitis and for radical cancer operations. Whilst obesity, the elderly and the very frail patient is not a formal contra-indication, in our experience laparoscopic distal pancreatectomy is more challenging, and an open approach is often preferred. We also agree that spleen preserving techniques may be easier to perform laparoscopically, however the preservation of the short gastric vessels, which is required in the Warshaw technique, might be more complex to perform laparoscopically. We certainly acknowledge that more centers are adopting the laparoscopic technique for technique distal pancreatectomy (as is our centre), however at present we think there is a lack of compelling level 1 evidence supporting laparoscopic distal pancreatectomies for all cases. Therefore, we embrace this technique selectively.
Current supporting evidence for this laparoscopic distal pancreatectomy exists can be found in retrospective case series and a few case-control studies. [1][2][3][4][5] Notable is the recent work of Kooby et al., 4 a large multiinstitutional case-control study in which 142 laparoscopic distal pancreatectomies were compared with 200 open distal pancreatectomies. Patients were matched based on age, pathologic findings, ASA criteria, and pancreatic specimen length. Similar to our findings, there was no increase in major morbidity or pancreatic leak rate with the laparoscopic approach. Their reported mortality was also 0%. We also acknowledge that many of the retrospective studies are limited by their multi-institutional, retrospective nature and the between centre variability.
In 2012, a systematic review and meta-analysis by Venkat at al. 6 reported that laparoscopic distal pancreatectomy (compared to open, all other factors being equal) was associated with less postoperative pancreatic fistula. Similarly, Khaled et al. 7 reported that the laparoscopic approach to distal pancreatectomy offers advantages over open surgery in terms of reductions in operative trauma and duration of postoperative recovery without compromising the oncologic resection. This was a single centre retrospective case-matched observational study.
More recently however, Røsok et al. 8 evaluated 582 studies, 52 (40 observational and 12 case-matched) were included in the assessment for outcome for laparoscopic distal pancreatectomy (n = 5023) vs. open (n = 16,306) whereas 16 observational comparative studies were identified for cancer outcome. No randomized trials were identified. The authors concluded that there was a tendency for lower blood loss and shorter hospital stay in the laparoscopic group, however they stated that available evidence for comparison of laparoscopic to open is "weak", although the number of studies is high and the observed outcomes of laparoscopic surgery are "promising", in the absence of randomized control trials, an international registry should be established. We agree with this statement.
In the 2016 Cochrane review by Riviera et al. 9 who examined laparoscopic versus open distal pancreatectomy for pancreatic cancer, it was reported that there is a dearth of randomised controlled trials that have compared laparoscopic distal pancreatectomy versus open distal pancreatectomy for patients with pancreatic cancers. This review did report that laparoscopic distal pancreatectomy has been associated with shorter hospital stay when compared with open distal pancreatectomy, however they also stated that currently, no information is available to determine a causal association in the differences between laparoscopic versus open distal pancreatectomy.
We also appreciate that the observed differences may be a result of confounding due to laparoscopic operation on less extensive cancer and open surgery on more extensive cancer. In addition, differences in length of hospital stay are relevant only if laparoscopic and open surgery procedures are equivalent from a positive oncological outcome perspective. This information is not available currently. Thus, randomised controlled trials are needed to compare laparoscopic distal pancreatectomy versus open distal pancreatectomy with at least two to three years of follow-up. Such studies should not only include postoperative complications (as we have reported), but patient-oriented outcomes, long-term mortality (at least two to three years), health-related quality of life, resection margins, and of course recurrence of cancer.

Question 2.
The strength of the study is to be able to demonstrate with a good analysis the considerable increase in costs in relation to postoperative complications. One of the weaknesses is not commenting on actions to try to reduce costs. For example: increase mini-invasive surgeries with fewer days of hospitalization, actions to reduce incidences of pancreatic fistulas, etc. Authors response: Thank you for this excellent and insightful comment, which we have now addressed in the discussion section.
We state "Finally, less than a quarter of patients in our study underwent laparoscopy surgery. Laparoscopic DP (compared to open), maybe associated with less postoperative pancreatic fistula, 26 and a tendency for lower blood loss, reductions in operative trauma and duration of postoperative recovery without compromising the oncologic resection, 27 and shorter hospital stay, 27-29 hence lower costs. The observed outcomes of laparoscopic surgery are promising, and further studies are required to comprehensively assess its cost effectiveness.
Four additional refences have been added to support this statement.
Question 3. Tables 1 and 4 are very long and difficult to follow. Could be summarized