Financial burden of postoperative complications following colonic resection

Abstract Background: Colonic resection is a common surgical procedure that is associated with a high rate of postoperative complications. Postoperative complications are expected to be major contributors to hospital costs. Therefore, this systematic review aims to outline the health costs of postoperative complications following colon resection surgery. Methods: MEDLINE, Excerpta Medica database, Cochrane, and Economics literature medical databases were searched from 2010 to 2019 to identify English studies containing an economic evaluation of postoperative complications following colonic resection in adult patients. All surgical techniques and indications for colon resection were included. Eligible study designs included randomized trials, comparative observational studies, and conference abstracts. Results: Thirty-four articles met the eligibility criteria. We found a high overall complication incidence with associated increased costs ranging from $2290 to $43,146. Surgical site infections and anastomotic leak were shown to be associated with greater resource utilization relative to other postoperative complications. Postoperative complications were associated with greater incidence of hospital readmission, which in turn is highlighted as a significant financial burden. Weak evidence demonstrates increased complication incidence and costlier complications with open colon surgery as compared to laparoscopic surgery. Notably, we identified a vast degree of heterogeneity in study design, complication reporting and costing methodology preventing quantitative analysis of cost results. Conclusions: Postoperative complications in colonic resection appear to be associated with a significant financial burden. Therefore, large, prospective, cost-benefit clinical trials investigating preventative strategies, with detailed and consistent methodology and reporting standards, are required to improve patient outcomes and the cost-effectiveness of our health care systems.


Rationale
Cost-effective health care, particularly in the hospital setting, is crucial for the sustainability of our health care systems. On the international level, health care expenditure has increased at a faster annual rate than economy growth between the years 2000 and 2016. [1] Rising health care costs combined with the continual necessity for high quality care, has resulted in growing demand by policy makers for high quality health economic assessments.
Vonlanthen et al [2] report that postoperative complications are the strongest indicators of in-hospital costs. These findings were reinforced by a systematic review demonstrating increased hospital costs from surgical complications after major surgery. [3] Editor: Young-Kug Kim.
The authors have no funding and conflicts of interest to disclose.
Supplemental Digital Content is available for this article.
The datasets generated during and/or analyzed during the current study are not publicly available, but are available from the corresponding author on reasonable request. The datasets generated during and/or analyzed during the current study are available from the corresponding author on reasonable request.
Limitations of this review were that it failed to report colorectal complications. Given that colon resection surgery is a common procedure with a high rate of postoperative adverse events relative to other major surgeries, [4,5] it is expected to be a major contributor to hospital costs. The development of cost-effective management strategies targeting colon resection surgery is dependent on accurate financial data and a deep understanding of the relationship between postoperative complications and the drivers of increased hospital costs.

Objectives
The primary aim of this systematic literature review is to outline the health costs of postoperative complications in adult patients who undergo colon resection surgery. We highlight the importance of evaluating the components of healthcare cost profiles relevant to patients undergoing colon resections and consider the quality of the studies with reference to how they measure and report costing data.

Protocol and registration
We conducted a systematic review of the literature in accordance with Cochrane guidelines [6] and reported under the guidance of the preferred reporting items for systematic reviews and metaanalyses statement. [7] The protocol of this review was registered in PROSPERO, an international prospective register of systematic reviews, and is available from: http://www.crd.york.ac.uk/ PROSPERO/display_record.php?ID=CRD42019128618. The Austin Health Research Ethics Committee waivered the requirement for ethics approval as collection of data did not involve patient contact or patient data.

Eligibility criteria
We included studies containing a full or partial economic evaluation of postoperative complications in adult patients (≥18 years of age) undergoing any form of colonic resection. Colon resection was defined as complete excision of any part of the large bowel (excluding rectum). Studies that did not report the cost of colon resection specifically were excluded.
Eligible study designs included randomized controlled trials (RCTs), non-RCTs, comparative observational studies and conference abstracts. Letters, opinion papers and editorials were excluded. Only studies in the English language were considered and no restriction by country or currency was applied.
Enhanced recovery after surgery [8] was established globally and adopted by many hospitals as a uniform approach to minimizing variability of perioperative care. Taking this recent advancement into account, we included studies published from January 2010 until February 2019 in order to retrieve up to date cost data.

Primary outcome
Total hospital costs associated with complications following colonic resection surgery

Study selection
The titles and abstracts of all retrieved studies were screened by 2 authors (ML and SJ) in an independent and blinded manner. The full texts of eligible studies were retrieved and independently evaluated for eligibility by 3 authors (ML, SJ, and LW). Disagreements were resolved by a fourth author (RM) and by consensus.

Data collection process and data items
Data from included studies was extracted in an independent manner by 2 authors (ML and SJ) into a predetermined data extraction

Risk of bias in individual studies
Risk of bias of the included studies were assessed by 2 authors (ML and SJ) using the Cochrane Collaboration's risk of bias tool to assess randomized controlled trials (RCTs) and the Scottish Intercollegiate Guidelines Network (SIGN) Checklist for Cohort Studies to assess cohort studies. Discrepancies were resolved by consensus following review by a third author (LW).

Summary measures and synthesis of results
Findings are reported in the form of a narrative synthesis. This is structured around the type of complications and their hospital costs. Cost of complications was derived from either the stated value within the study or by calculation of the cost difference between the group with and the group without complications. Costs were converted to United States dollar (USD) ($) based on the annual average conversion rate [9] of the specified base currency year or the year of publication if a currency year was not reported. Costs were then inflated to February 2019 from January of the specified or assumed cost year using the Bureau of Labor Statistics Consumer Price Index inflation calculator. [10] We referenced complication costs to the complication type, complication severity, surgical technique, indication for surgery, surgical urgency, readmissions, mortality, and LOS. A critical review of the data showed significant discrepancies in economic environment and hospital characteristics of the included studies resulting in significant heterogeneity of the studies, therefore a meta-analysis could not be performed.

Risk of bias across studies
Homogenous effect sizes across studies were unavailable. Therefore, formal assessment of publication bias using a funnel plot was not presented. Each of the reported study's outcomes and results were compared to assess for selective reporting bias.

Additional analyses
Subgroup analyses were undertaken for resource utilization measure, readmissions, surgical technique, number and severity of complications and type of complication.

Study selection
The search strategy resulted in a total of 2289 articles being sourced. Twelve additional studies were identified by manual searching of bibliographies. Thirty-four articles  met the eligibility criteria. The preferred reporting items for systematic reviews and meta-analyses flow diagram [7] representing the selection of studies is presented in Figure 1.

Population
Number of participants ranged from 46 patients [29] to 217,939 patients. [41] Reported mean and median patient age ranged from 49.9 years [24] to 78.5 years. [43] Most studies incorporated broad inclusion criteria, including all colon resections performed within the specified time frame. Details on patient demographics are presented in Table S2, Supplemental Digital Content, http://links. lww.com/MD2/A260.

Risk of bias within studies
Thirty-one out of the thirty-four included studies were retrospective cohort studies with their retrospective nature preventing them from being considered "high quality" evidence based on the SIGN cohort study checklist. All twenty-two retrospective cohort full paper publications [11,12,[16][17][18][19][20]22,24,25,27,28,31,32,35,37,38,[40][41][42][43][44] were of "acceptable quality" with clear focused research questions. The risk of bias in the nine conference abstracts [13,15,21,23,26,30,33,36,39] was not formally assessed due to the incomplete nature of their reporting. The cross-sectional cohort study [34] was deemed of "acceptable quality" as it is classified as a database study by the SIGN cohort study checklist. The overall risk of bias for the randomized controlled [14] and the randomized clinical trial [29] was judged to be low with satisfactory randomization and no evidence of selection bias.

Risk of bias across studies
Noninclusion of grey literature in this review may have resulted in publication bias. In addition, inclusion of studies only in the English language may have resulted in a language bias. The principally objective nature of the financial and clinical outcomes assessed in this review makes outcome measurement an unlikely bias. Conference abstracts are at a high risk of selective reporting bias and this was considered in our data synthesis.
Synthesis of results: all costs are presented as: study reported costs (adjusted USD cost).

Incidence and cost of complications
Postoperative complication incidence varied greatly between the studies ranging from 6.0% [32] to 66.0% [35] (Fig. 2). This variance can be attributed to the different definitions of complications  [21] this is in part due to the heterogenous definitions of hospital costs adopted by the different studies as well as the different complication types reported. Asgeirsson et al [11] and Knechtle et al [25] further demonstrate a positive correlation between the count of complications and the additional cost incurred by the hospital. All studies reporting hospital charges and hospital reimbursements demonstrated a positive increase in hospital resource utilization with postoperative complications (  [39] There was an exponential rise in reimbursements to $17,124 [$18,270] for grade III+ complications. [39] Despite this, grade I and grade II complications remain a significant health care burden due to their high prevalence rates (29% and 49%, respectively). [39] Table 1. SSI and anastomotic leak were associated with the greatest financial burden amongst postoperative complications in colon resection surgery. The additional hospital cost of SSI varied greatly across studies (Table 1). This significant variation in costs for SSI can be attributed to the geographic differences in health care systems. Asgeirsson et al [11] was the only study directly comparing SSI with postoperative ileus demonstrating significantly higher additional costs in the presence of SSI than with postoperative ileus (Table 1).

Cost of surgical technique
Nine studies [11,13,14,18,19,22,30,37,40] reported the cost of complications in open and laparoscopic surgery, however only one study [37] compared the two groups as its primary outcome.
Postoperative complications in open surgery were shown to be associated with higher hospital costs than postoperative complications in laparoscopic surgery across all included studies except Kashimura et al [22] ( The risk of selection bias due to the retrospective nature of the studies and existence of uncertainty around the statistical significance of data reported in some studies means that these findings are supported by weak evidence.

Length of stay
Postoperative complications resulted in an increased hospital LOS across all studies with additional LOS ranging from 1.5 days [38] to 19 days. [26] In addition, the greater the cumulative number of complications a patient experienced, the greater their hospital LOS. [25] Increasing SSI severity was also associated with increased hospital LOS. [18] No study assessed the direct cost impact of LOS.

Mortality
Three studies [28,36,40] reported increased mortality rates associated with incidence of postoperative complications. No study reported the cost impact of mortality.

Indication for surgery
Impact of indication for surgery on incidence of complications was inconclusive across all studies. [18,38,40] Zogg et al [40] was the only study assessing the cost impact of indication for surgery demonstrating increased costs associated with patients with colon cancer experiencing postoperative complications.

Surgical urgency
Impact of surgical urgency on incidence and cost of complications is inconclusive across all studies. [11,17] Fukuda et al [18] reported no statistical association between surgical urgency and risk of SSI. However, Asgeirsson et al [11] reported higher complication incidence and hospital costs with urgent/emergent admissions as compared to elective admissions.

Cost breakdown
Only two studies [32,38] reported the financial burden associated with complications broken down into different hospital cost centers (Table S2, Supplemental Digital Content, http://links. lww.com/MD2/A260). Both studies explored the costs associated with SSI, demonstrating the greatest cost difference in inpatient costs [38] and "room," "operative," and "other" hospital costs, which included the combined cost of nursing, medication, laboratory, and radiology services. [32] These increased costs can  Table 1 Summary of studies evaluating the financial burden of surgical site infection, postoperative ileus and anastomotic leak following colonic resection surgery.

Discussion
In an updated systematic review of thirty-four studies, we demonstrate strong evidence of high overall complication incidence arising from colonic resection surgery with associated increased costs and resource utilization. We found a considerable degree of heterogeneity among studies in factors such as study design, defining and reporting on complications, and methodology used to calculate "cost" and associated outcomes. Despite these limitations, our findings confirm that hospital readmissions are associated with significant financial burden, and postoperative complications are associated with greater incidence of hospital readmissions. We found weaker evidence that postoperative SSI and anastomotic leak are associated with greater costs and resource utilization relative to other postoperative complications.
Our review highlights significant shortcomings in defining and reporting of hospital resource utilization in economic studies of postoperative complications in colon resection surgery. First, the measure of hospital resource utilization adopted by the studies varied and was poorly defined in many. Second, the currency year was not reported in most of the studies thus had to be assumed to be the publication year. Thirdly, reporting of costs using means and medians varied, impeding on direct comparison between studies.
Hospital costs, hospital charges and hospital reimbursements are 3 resource use measures that represent different financial aspects of health economics. [45,46] Hospital charges for a given service may differ greatly between hospitals and health care systems and are considered a poor representation of hospital costs. [47] Similarly, hospital reimbursement systems demonstrate significant geographical variation in their coding classification and payment value. [48] In the USA, and many European countries, hospital reimbursements are predetermined and based on DRG codes [48,49] with cost variation within DRG codes acting as a source of uncertainty. As such, the most reliable measure of hospital costs involves recording actual resource consumption for each admission. [47] Secondary to this, studies should clearly define and report the utilized hospital resource use measure to enable accurate analysis of a study's results.
Poorly defined and inconsistent reporting of hospital resources acts as a barrier to accurate comparison of cost and clinical outcomes between studies. Hospital costs consists of fixed direct, variable direct and indirect costs. [46] Inclusion or exclusion of specific hospital cost components resulted in variation in total financial burden of complications amongst the included studies as presented in Figure 3. Furthermore, many studies in our review failed to report the cost currency and currency year which is essential in allowing comparison of cost data. The reporting of health cost studies should adhere to a minimal standard of reporting including the definition of hospital cost components analyzed including any adjustments for inflation that the authors performed. In addition, skewed distributions are expected in medical costing data, [50] therefore it is recommended that both mean and median costs are reported to avoid misinterpretation of results. [50] Most of the included studies reported the cost of a specific complication type improving the clinical relevance of these studies. However, complication definitions were inconsistent ; P-value

<.001
Langelotz et al [26] Single  across studies limiting the ability to compare complication types. Many studies utilized local institutional definitions or the definitions of the national databases they analyzed. Only seven studies [19,24,[36][37][38]40,41] specified the use of the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes to classify complications. Reporting on postsurgical complications should be aligned with established international standards for definitions and use of outcome measures, designed for clinical effectiveness research in perioperative medicine. [51] Additionally, only three studies [11,18,39] assessed the cost impact of complication severity with only one [39] of these utilizing a prevalidated complication severity grading system (the Clavien-Dindo classification system). Our results demonstrated greater hospital resource use with greater complication severity, and as such management and outcomes of complications are heavily reliant on complication severity. Therefore, complication severity is an important factor to analyze and should be reported using internationally validated grading systems such as the Clavien-Dindo classification system. [52] Our review also highlights hospital readmissions as a significant source of costs. This has been recognized internationally. Specifically, in the USA, the hospital readmissions reduction program was introduced which penalizes hospitals who demonstrate high 30-day readmission rates for specific conditions/ procedures. [53] This initiative creates a significant financial incentive for hospitals to introduce measures that reduce readmissions. Our review demonstrated increased readmission rates in patients who experience postoperative complications, highlighting the financial benefit of reducing postoperative complication rates. We also recommend that the reporting of readmissions should be standardized to at least 30 days post discharge.
Prevention strategies that aim to mitigate risk factors for complications following colorectal surgery are well described. [54] These include preoperative weight loss, [54] nutritional optimization through immunonutrition [55] and early postoperative enteric nutrition, [56] intraoperative blood loss and blood transfusion minimization, [57,58] and use of laparoscopic surgery if feasible. [59,60,61] Incorporation of these strategies and others into enhanced recovery after surgery management programs has been shown to be associated with reduced postoperative complications, [60] reduced hospital LOS and costs. [62] Despite this, there is a paucity of studies analyzing the cost-benefit outcomes of these strategies. Therefore, large, prospective cost-benefit clinical trials investigating interventions aimed at reducing postoperative complications are still required to improve patient outcomes and the cost-effectiveness of our health care systems.
Our review has several limitations. Notably our review includes a high number of retrospective studies with few highquality detailed cost outcome studies. Despite an exhaustive search strategy, noninclusion of grey literature may subject our review to a risk of publication bias. In addition, the significant heterogeneity of the included studies prevented a quantitative analysis of the cost results, however the large number of articles identified allowed direct comparison of studies of similar characteristics. High quality prospective economic studies are needed to evaluate the cost of complications arising from colonic resection surgery.
Our systematic review has important clinical implications. We have demonstrated a high prevalence of complications following colon resection surgery and a substantial financial burden associated with complications. Therefore, large, prospective costbenefit analysis trials investigating strategies aimed at reducing surgical complications and their associated costs are required. Given the significant heterogeneity identified in our review, we propose a standardized approach for future costing studies of postoperative complications. Studies should follow a detailed and consistent methodology with the use of validated economic study guidelines [63] and must report, as a minimum, the following variables: complication definition, complication severity (reported using internationally validated grading systems such as the Clavien-Dindo classification system), follow-up duration for each outcome of interest, mortality reported at 90 days postoperatively and ideally at 1 year in clinical effectiveness research, readmissions standardized to at least 30 days post discharge, hospital cost definition (clearly define and report the utilized hospital resource use measure and the hospital cost components analyzed), cost currency, cost year adjusted for inflation, mean and median cost, confidence intervals and interquartile range as skewed distributions are expected in medical costing data.