The application of laparoscopic techniques to colon cancer surgery has been a standard approach since the 1990s,1 with reported benefits of less intraoperative blood loss, fewer adhesions, shorter hospital stay, higher postoperative quality of life, and faster return to work;2,3,4,5,6,7,8 however, whether this surgical approach should be applied to rectal cancer surgery is unclear. Rectal resection often involves a more complicated and morbid procedure than for other gastrointestinal cancers and requires very precise dissection to reduce the chance of cancer recurrence.9 This has been shown in the Australasian Laparoscopic Cancer of the Rectum Trial (ALaCaRT)10,11 and the American College of Surgeons Oncology Group Z6051 trial,12 where the non-inferiority of the laparoscopic approach compared with conventional open resection for rectal cancer treatment for short-term pathological outcomes was not established. However, laparoscopic-assisted surgery did not show significantly worse oncologic outcomes at 24 months11 and may have facilitated more Australian patients to return to work at 12 months.13

Decisions about whether to undertake laparoscopic or open surgery for rectal cancer are also influenced by their costs and the overall impact of each type of surgery on hospital budgets. The difference in healthcare use and costs between laparoscopic-assisted surgery and open surgery is uncertain, however it is generally agreed that laparoscopic-assisted surgery has higher upfront costs as the procedure involves more disposable equipment and longer operative time.14,15 Whether the higher upfront cost can be offset by lower subsequent costs remains unknown from previous studies.14 In the Australian state of Queensland, a non-randomized study demonstrated laparoscopic-assisted surgery had a similar operating time and lower costs than open surgery for colorectal cancer treatment due to a shorter postoperative hospital stay and shorter time in the intensive care unit (ICU).16 A recent cost-comparison study of the COLOR II trial, a non-inferiority, open-label, multicenter randomized controlled trial (RCT) in eight European and Asian countries, showed that the average healthcare costs per patient at 28 days and 3 years were higher for laparoscopic-assisted surgery than open surgery, however length of hospital stay was shorter in the laparoscopic group (8 vs. 9 days; p = 0.036).15,17 However, in costing rectal cancer surgery within the COLOR II trial, unit costs of health resource use were only collected from Swedish sources. In addition, ICU admissions, length of intensive care during index and subsequent hospital stay, and other postoperative care, including colonoscopy, follow-up computed tomography (CT) scans and x-ray tests were not included. Furthermore, cost estimates within the COLOR II trial were originally reported in Swedish Krona (SEK) and converted to US dollars (US$) using the 2013 average exchange rate, rather than the accepted purchasing power parity (PPP). For these reasons, the applicability of the results for decision making in other healthcare systems (e.g. Australia) are limited as total costs are crucial for reimbursement and policy formulation.

This study aimed to evaluate the healthcare costs of laparoscopic-assisted surgery compared with open resection for rectal cancer patients participating in ALaCaRT, for surgery and up to 12 months, from an Australian health system perspective. It is hypothesized that the higher upfront costs of laparoscopic-assisted surgery for rectal cancer were offset by lower postoperative care during the first 12-month period

Materials and Methods

Study Participants

Rectal cancer patients who participated in ALaCaRT were included in this prospective analysis (ACTRN12609000663257). Healthcare use and cost effectiveness of the surgical approaches were listed a priori as secondary objectives at trial registration. ALaCaRT was a multicenter randomized, non-inferiority, phase III trial evaluating the safety and efficacy of laparoscopic resection versus open surgery for rectal cancer.10,11 All ALaCaRT participants were adults aged 18 years or older, with a histological diagnosis of adenocarcinoma of the rectum within 15 cm of the anal verge, and a life expectancy of at least 12 weeks.10 Overall, 475 patients from 24 hospitals in Australia and New Zealand were recruited between March 2010 and November 2014, with 473 eligible for analysis.10 Patients were randomized on a 1:1 basis to undergo laparoscopic or open surgery, and stratified by the site of the tumor, registering surgeon, planned operative procedure (low anterior resection or abdominoperineal resection), body mass index (BMI), preoperative radiotherapy, and distant metastases. Postoperative care was by the operative surgeon, except for pain control procedures, the initiation of oral intake and its monitoring, ingredients of intravenous fluids, and criteria for hospital discharge. Data on included participants were investigated on an intention-to-treat (ITT) basis.10 During the trial, electronic clinical record forms (eCRFs) were administered, two (Randomization and Baseline) for the preoperative stage, one (Surgery) for the intraoperative stage, and two (Pathology and Early Complications) for the postoperative stage. Postoperative follow-up was undertaken to collect information on patient status, subsequent admission, and elective procedure at days 3 and 14, 4–6 weeks, and 3, 6, 9, and 12 months after discharge. All trial participants gave written informed consent before randomization. Central ethics approval was obtained by the Sydney Local Health District (LHD) Human Research Ethics Committee.

Healthcare Costing Methodology

This cost analysis compared healthcare resource use and the direct healthcare costs of laparoscopic-assisted and open surgery (randomization allocation as per the ITT principle) from enrolment into the trial, including index surgery, first 30 days, 56 days, and up to 365 days (12 months) after surgery from a health system perspective. Clinical effectiveness of treatment was not included in this analysis as no significant differences in 2-year disease-free survival and overall survival were previously found between the two surgery types in ALaCaRT.11 Potential bias attributed to censoring and missing data was considered small and immaterial as only 1.9% of participants had died and the trial had a high eCRF response rate of 95–99% during the 12-month follow-up period. It was therefore assumed that, in this analysis, patients with missing data would not differ from those with complete data.

Resource Use

For the base case, data on healthcare use were mainly collected through eCRFs in ALaCaRT. This included ‘open to close’ operative time, whether anastomosis was performed, whether ostomy was created, hospital length of stay, reoperations, subsequent hospital admissions, and postoperative care (i.e. colonoscopy, abdominal CT, chest CT and chest x-ray). Healthcare use that was not collected during the trial (including surgical staff attending, operating theatre, surgical equipment, sterilization of surgical equipment, and anesthesia time) was determined in collaboration with senior surgeons (ARLS and MS) and by external sources.15,18,19 Details of healthcare use in rectal cancer surgery that were not covered in the ALaCaRT database are listed in Table 1.

Table 1 Key assumptions of resource use in rectal cancer surgery not covered by the ALaCaRT database
Full size table

Unit Costs

Unit costs were obtained from numerous sources (Table 2). Hospital staffing costs were based on the staff award rates of pay at public hospitals in the state of New South Wales, Australia. The per minute rate for operating theatres and the unit costs for surgical instruments were made available from the Agency for Clinical Innovation, NSW Health. Costs of specific equipment, such as the linear stapler (if anastomosis was not performed) and circular stapler (if anastomosis was performed), were taken from the purchase price for public hospitals participating in ALaCART for these products in 2021. All non-specific equipment was assumed to be incorporated into the operating theatre costs. Unit costs for sterilization of surgical equipment were provided by the Sydney LHD. Reoperations within 12 months were mapped by CKL and RLM into six Australian Refined Diagnosis-Related Groups (AR-DRGs), based on their reasons recorded in text format in the ALaCART database (see Appendix 1). Unit costs of reoperation for each AR-DRG were estimated by multiplying their specific price weight with the national efficient price (NEP). Per diem rates for acute hospital and ICU stays, and unit costs of other postoperative care, were calculated from figures published by the Independent Hospital Pricing Authority (IHPA), Medicare Benefits Schedule (MBS), and online retailers.

Table 2 Unit cost of healthcare resources
Full size table

Sensitivity Analyses

Unit cost calculations for operating room resources are often surrounded by uncertainty as it can be difficult to disentangle the potential effect of different cost components.18 One-way sensitivity analyses were therefore carried out to challenge some of the assumptions adopted in this analysis. This included varying staffing costs, time in anesthesia, equipment costs, reoperation costs, ICU costs, and hospital length of stay. In each case, a ±30% deviation in value per variable for each procedure was adopted to determine the robustness of the cost estimates between the two types of surgery. A non-parametric bootstrapping exercise with 10,000 replications was included for both the base case and sensitivity analyses as a robustness check of the costing results.19,20,21,22 T-tests, given the normality assumption of the central limit theorem, were not used to test for differences in mean healthcare costs due to the skewed distribution of costs.23

Results

Of 473 ALaCaRT participants, 468 (236 laparoscopic, 232 open) were included in this analysis. Five were excluded as their operative time was missing at the time of analysis. Baseline characteristics of the study sample, by treatment arm (laparoscopic‐assisted surgery, n = 236; open surgery, n = 232), are shown in Table 3. Three participants were of Aboriginal or Torres Strait Islander origin (laparoscopic‐assisted surgery, n = 1; open surgery, n = 2). No significant difference in baseline characteristics by surgery type was identified.

Table 3 Baseline characteristics of ALaCaRT participants included in the cost analysis
Full size table

Healthcare Resource Use

The main items of healthcare use in rectal cancer surgery for ALaCaRT participants are shown in Table 4. Rectal cancer patients randomized to laparoscopic-assisted surgery incurred a longer average operative time (219.6 vs. 196.6 min), more anastomosis (88.6 vs. 87.1%), and creation of an ostomy (80.5 vs. 71.6%), but slightly shorter average stay in hospital (10.6 vs. 10.9 days), than patients randomized to open surgery. Fewer patients randomized to laparoscopic-assisted surgery required intensive care (10.6 vs. 13.8%), but their average length in ICU was longer (5.0 vs. 4.5 days) than those randomized to open surgery. For patients who underwent ostomy during surgery, 15.7% (n = 56) had colostomy and 84.3% (n = 300) had ileostomy. No obvious difference in the proportion of colostomies was observed between the two randomized groups (laparoscopic, 15.3%; open, 16.3%).

Table 4 Operation-related healthcare use for rectal cancer surgery by type of surgery
Full size table

For postoperative care, 27 (5.8%) ALaCaRT participants required a reoperation during the first 47 days after the index surgery (see Table 5). More rectal cancer patients randomized to laparoscopic surgery had a reoperation than those randomized to open surgery (7.3 vs. 4.3%). About two-thirds of reoperations (n = 18; laparoscopic, n = 12; open, n = 6) were related to anal and stomal procedures.

Table 5 Postoperative use of healthcare resource by type of surgery and time of analysis
Full size table

Over the 12-month postoperative period, patients randomized to laparoscopic surgery had 241 hospital admissions, with a total length of hospital stay of 1434 patient-days (see Table 5). Patients randomized to open surgery had 248 hospital admissions, with a total length of hospital stay of 1615 patient-days during the first 12 months. For hospital stays, 48 patient-days (3.3%) in the laparoscopic surgery group and 67 patient-days (4.1%) in the open surgery group required intensive care. Lastly, no obvious difference in other postoperative care during the first 12 months (ileostomy closure rate: laparoscopic 59.6% vs. open 61.9%; colonoscopy: laparoscopic 75 vs. open 74; abdominal CT: laparoscopic 252 vs. open 237; chest CT: laparoscopic 198 vs. open 180; and chest x-ray: laparoscopic 54 vs. open 60) was observed between the two randomized groups.

Per Patient Cost of Resource Use

The average costs of health care for randomized patients at different time periods are shown in Table 6. Due to a longer operative time and more surgical instruments involved, the per patient cost of laparoscopic-assisted surgery was AUD$4542 (standard deviation [SD] AUD$1050), which was about $520 higher (bootstrap 95% confidence interval [CI] AUD$354–AUD$692) than that in the open group at AUD$4021 (SD AUD$804). During the index hospitalization after surgery, patients randomized to open surgery had a longer ward and ICU stay than those randomized to laparoscopic-assisted surgery. This resulted in a lower average hospitalization cost for laparoscopic surgery at AUD$23,444 (SD AUD$25,338) than that in the open group at AUD$24,325 (SD AUD$25,188), although the cost difference was not statistically significant (bootstrap 95% CI −AUD$5483 to AUD$3628). At discharge from index hospitalization, no significant difference in overall healthcare costs (both surgery and index hospitalization) was detected between the two randomized groups (difference −AUD$360; bootstrap 95% CI −AUD$4955 to AUD$4165).

Table 6 Average cost of healthcare for randomized patients at different time periods
Full size table

Following discharge from the index hospitalization, patients randomized to laparoscopic surgery had a higher cost of reoperations (AUD$347 vs. AUD$128) and stoma care (AUD$609 vs. AUD$513) than those randomized to open surgery during the first 365 days; however, their costs of subsequent ward (AUD$11,981 vs. AUD$13,611) and ICU stays (AUD$1083 vs. AUD$1537) were lower than that of open surgery. As a result, the total average cost for rectal cancer patients randomized to laparoscopic surgery in the first 365 days, at AUD$43,288 (SD AUD$40,883), was slightly lower than open surgery, at AUD$45,384 (SD AUD$38,659), however the mean difference was not statistically significant (bootstrap 95% CI −AUD$9358 to AUD$5003).

Sensitivity Analysis

In sensitivity analyses, variations in staffing costs, time in anesthesia, equipment costs, reoperation costs, ICU costs, and the cost of hospital stays slightly changed the mean difference in costs across the study period (Table 7); however, none of these variations had a significant impact on the base-case results. Across all individual participating sites, the average total costs per participant ranged between AUD$25,300 and AUD$64,600, but no significant pattern was found using one-way analysis of variance (ANOVA; F = 0.74, p = 0.7896).

Table 7 Sensitivity analyses
Full size table

Discussion

This study examined whether laparoscopic-assisted surgery in the first 12 months of follow-up was more expensive for the healthcare system than open resection. This data is important for decision making as healthcare budgets should consider not only the intraoperative costs but also the follow-up costs, particularly reoperations and readmissions in the short and medium term. From an Australian health system perspective, the results of this analysis showed that patients undergoing laparoscopic-assisted surgery for rectal cancer incurred higher upfront surgery costs than those randomized to open surgery, however no significant cost differences were identified between the two randomized groups when subsequent healthcare costs were included for index hospitalization, at 30, 56, or 365 days postoperation. The one-way sensitivity analyses showed that these results were robust to variations in major cost components.

In line with previous RCTs,15,18,19 the higher upfront cost for laparoscopic-assisted surgery was mainly due to longer operative time and involved more costly equipment. Like COLOR II15 and another non-randomized observational study in Queensland, Australia,16 higher upfront costs for laparoscopic surgery were offset by a shorter length of hospital stay and fewer ICU admissions, compared with open resection, during the index hospitalization. Furthermore, when compared with the COLOR II trial,15 fewer ALaCaRT participants required a reoperation shortly after the index procedure in both randomized groups (ALaCaRT, 5.8% vs. COLOR II, 14.9%). This may be attributed to Australia’s early adoption of laparoscopic resection for colorectal cancer and professional training of Australian colorectal surgeons.16

Under the Australian activity-based funding (ABF) framework, the cost of hospital services can be calculated with their assigned AR-DRG. Each AR-DRG indicates the workload and complexity of an inpatient episode by integrating the number and type of patients treated in a hospital to the resources required, in a clinically meaningful, fair, and equitable way.24 In the present analysis, the costs of colonoscopy and reoperation were estimated based on the ABF costing approach; however, this approach was not able to be used to value the rectal cancer surgery, as current AR-DRGs have not specified price weights by type of surgery. The use of ABF costing would inevitably underestimate resource use (i.e. longer operative time and involvement of more expensive surgical equipment) and genuine cost for laparoscopic-assisted surgery in Australia.

The findings of the present analysis extend our understanding of healthcare resource use and total hospital costs for each procedure in rectal cancer surgery, which provides useful information for hospital decision makers in Australia. In circumstances where laparoscopic-assisted surgery for rectal cancer is required, clinicians can be assured that laparoscopic-assisted surgery is not appreciably different than open surgery, and therefore the decision about the type of surgery can be made on clinical indications alone.

The present analysis was one of the few studies on the costs of rectal cancer surgery based on a large RCT with a multicenter design involving 24 hospitals in Australia and New Zealand.10 It also used a bottom-up micro-costing approach, with local unit costs for staffing, consumables, and surgical equipment, to estimate the cost of surgery and follow-up healthcare use in the Australian context. There are however some limitations to acknowledge. First, this analysis examined the cost of surgery for rectal cancer and follow-up healthcare costs from the healthcare system perspective; however, primary care services and other non-health-related costs (dietitian consultation, intake of healthy food and supplements, psychological counseling, support groups for cancer survivor) were not included due to a lack of data. In addition, for many patients, being unable to return to work as soon as feasible is a very important issue due to the financial toxicity associated with cancer care and treatment, especially in countries without a well-developed social welfare system, such as the US. A broader coverage of costs should be included in future research to ensure a more complete and relevant cost estimation for rectal cancer surgery. Second, as an RCT, ALaCaRT participants may be healthier than the average rectal cancer survivors in the community and with no concurrent or previous invasive pelvic malignant tumors within 5 years before study enrollment.10,13 The results of this study may therefore not be generalizable to all adults treated for rectal cancer.13

Conclusion

Despite higher upfront surgical costs, laparoscopic-assisted surgery for rectal cancer was similar in costs to open surgery for rectal cancer over a 12-month period. Clinicians may choose a surgical approach on the basis of clinical need and quality audits at the unit level examining variations in care (i.e. outliers) in achieving optimal outcomes.