Healthcare Resource Utilization After Surgical Treatment of Cancer: Value of Minimally Invasive Surgery

Background As the US healthcare system moves towards value-based care, hospitals have increased efforts to improve quality and reduce unnecessary resource use. Surgery is one of the most resource-intensive areas of healthcare and we aim to compare health resource utilization between open and minimally invasive cancer procedures. Methods We retrospectively analyzed cancer patients who underwent colon resection, rectal resection, lobectomy, or radical nephrectomy within the Premier hospital database between 2014 and 2019. Study outcomes included length of stay (LOS), discharge status, reoperation, and 30-day readmission. The open surgical approach was compared to minimally invasive approach (MIS), with subgroup analysis of laparoscopic/video-assisted thoracoscopic surgery (LAP/VATS) and robotic (RS) approaches, using inverse probability of treatment weighting. Results MIS patients had shorter LOS compared to open approach: − 1.87 days for lobectomy, − 1.34 days for colon resection, − 0.47 days for rectal resection, and − 1.21 days for radical nephrectomy (all p < .001). All MIS procedures except for rectal resection are associated with higher discharge to home rates and lower reoperation and readmission rates. Within MIS, robotic approach was further associated with shorter LOS than LAP/VATS: − 0.13 days for lobectomy, − 0.28 days for colon resection, − 0.67 days for rectal resection, and − 0.33 days for radical nephrectomy (all p < .05) and with equivalent readmission rates. Conclusion Our data demonstrate a significant shorter LOS, higher discharge to home rate, and lower rates of reoperation and readmission for MIS as compared to open procedures in patients with lung, kidney, and colorectal cancer. Patients who underwent robotic procedures had further reductions in LOS compare to laparoscopic/video-assisted thoracoscopic approach, while the reductions in LOS did not lead to increased rates of readmission. Supplementary Information The online version contains supplementary material available at 10.1007/s00464-022-09189-8.


3
robotic-assisted surgery to a wide range of specialty fields within surgery.
As surgery is one of the most resource-intensive areas of clinical medicine, there is a growing trend for quality improvement initiatives to improve the efficiency, quality, and safety of surgical care, to reduce unnecessary consumption of resources, and to increase patient satisfaction [6]. Some commonly used indicators to measure surgical care resource utilization include hospital length of stay (LOS), reoperation, and readmission. With the increasing adoption of MIS, especially robotic surgery, it is necessary to better understand their impact on healthcare resource utilization and quality of surgical care. This is of greater importance in the context of the COVID-19 pandemic given the need to free up hospital beds, staff shortages, and other competing resource needs.
Some prior studies have evaluated healthcare resource utilization after MIS; however, they tended to focus on specific procedures or were performed in single institutions [7,8]. This limits their generalizability to the broad range of surgical oncological procedures performed at the national level. The aim of this study was to leverage a national hospital discharge database in the US.

Data source
The Premier Healthcare Database (PHD) was used for this study. The database contains service-level information for hospital-based inpatient admissions and outpatient encounters for over 231 million patients in the United States. Clinical, billing, and financial information can be tracked for patients within the same hospital in the database [9]. Institutional Review Board approval was not necessary for this study because PHD is commercially available and de-identified.

Study population
Hospital encounters for adults 18 years of age and older were included in the study if the patient underwent one of the following primary, elective inpatient procedures between January 1, 2014 and December 31, 2019 using either an open, laparoscopic/video-assisted thoracoscopic surgery (LAP/VATS), or robotic-assisted (RAS) surgical approach: (1) colon resection for colon cancer, (2) rectal resection for rectal cancer, (3) lobectomy for primary lung cancer, or (4) radical nephrectomy for kidney cancer. Procedures and their corresponding surgical approaches were defined using International Classification of Diseases, Ninth Revision (ICD-9) codes; ICD-10 Codes; Current Procedural Terminology (CPT) codes; and hospital billing records (Supplemental Table 1). An encounter was excluded from the analysis if the corresponding procedure's operating room time or total cost was less than or equal to zero minutes or dollars, respectively.

Study outcomes
The primary outcome for this analysis was length of stay (LOS), which is directly captured in the PHD and is calculated as the discharge date minus the admission date. Secondary outcomes included reoperation during hospital stay, discharge to home, and 30-day readmission rates. Reoperation was defined as any return to operating room billing record after index surgery.

Study covariates
Patient, surgeon, and hospital characteristics were used as covariates in the analysis. Patient characteristics included age, gender, race/ethnicity, insurance, Charlson Comorbidity Index (CCI; excluding cancer), presence of metastasis, obesity, smoking history, and year of surgery. Surgeon characteristics included surgeon specialty and surgeon procedure volume. Hospital covariates included hospital procedure volume, geographic region, teaching status, rural/urban, and hospital bed size.

Statistical analysis
All descriptive and statistical testing analyses were conducted by procedure comparing open surgical approach to MIS, and LAP/VATS to RAS. Unstratified descriptive statistics were also calculated across all procedures. For both the crude and adjusted analyses, the gtsummary v1.4.2 package in R was used to calculate frequencies and proportions for categorical outcomes and covariates, and means, medians, standard deviations, and interquartile ranges for continuous outcomes and covariates.
Adjusted analyses were achieved using Inverse Probability of Treatment Weighting (IPTW) through the WeightIt v0.12.0 package in R. Stabilized propensity score weights were used to estimate the average treatment effect and all patient, surgeon, and hospital covariates were used to create balance between the groups [10]. A covariate was considered balanced if the absolute value of the standardized mean difference after adjustment was less than 0.10. Using the IPTW-adjusted data, adjusted mean differences and odds ratios were calculated. A gamma regression with an identity link was used to calculate the mean difference and 95% confidence interval between comparison groups for LOS. A logistic regression model was used to calculate the odds ratio and 95% confidence interval between comparison groups for reoperation, discharge to home, and 30-day readmission rates. Mean differences and odds ratios were considered significant if the p-values were less than 0.05. For the lobectomy procedure comparing open surgical approach to MIS, surgeon procedure volume and hospital procedure volume were added as additional adjustment variables to the models because the absolute values of the standardized mean differences for both covariates after IPTW were not less than 0.10. In the sensitivity analysis, we assessed the conversion to open surgery, ICU admission for at least 1 day, ICU admission for at least two days, and mechanical ventilation usage. All analyses were conducted using R version 4.1.1.

Discussion
As the US healthcare system moves towards value-based healthcare, hospitals and surgeons have increased efforts to improve quality of care and reduce unnecessary resource utilization while achieving the goal of the procedure [11,12]. Hospital LOS is a common indicator for episode resource use, and readmission after surgery is often viewed as a quality measure by Medicare and other insurers. Our data demonstrates a significant outcomes advantage for MIS procedures compared to open procedures in patients with lung, kidney, and colorectal cancer. MIS is associated with shorter LOS, higher discharge to home rate, and lower rates of reoperation and readmission. Patients who underwent robotic procedures had further reductions in LOS compared to laparoscopic approach, while simultaneously not increasing readmission rates. These data demonstrate substantial outcomes gains for patients who undergo robotic procedures across cancer diagnoses.
As previously described, there has been substantial growth in robotic procedures throughout the world. In a review of data from the OptumLabs Data Warehouse in the United States and the Hospital Episodes Statistics in England, investigators demonstrated that robotic surgery has become the standard approach for radical prostatectomy in the United States and England [13]. Similarly, utilization of robotic proctectomy for rectal cancer has also steadily increased [14]. Confirming this practice change, our generalizable data reveal rapid gains in adoption of robotic procedures across cancer types by study end. With this rapid acceptance, we identified substantial advantages in LOS and  [16][17][18][19]. In contrast to these data, we can confirm a clear and consistent length of stay advantage for cancers of the colon, rectum, lung, and kidney approached in a robotic fashion.
Some of the LOS benefit for robotically approached procedures may be related to fewer conversions from minimally invasive to open surgery. In an analysis of administrative data including patients who underwent right colectomy, investigators found that patients who underwent robotic as compared to laparoscopic surgery were significantly less likely to undergo conversion [4]. Similarly, data from the Norwegian Registry for Gastrointestinal Surgery and from the Norwegian Colorectal Cancer Registry also revealed lower conversion rates with robotic-assisted rectal resections compared with conventional laparoscopic resections [20]. In the same manner, meta-analyses of patients with lung cancer have similarly identified lower conversion to open surgery for patients who underwent robotic surgery as compared to video-assisted surgery [15,21]. Although not all studies have demonstrated fewer conversions with robotic surgery [7], our data convincingly demonstrate significant reductions in conversions in all studied procedures except for nephrectomy. In fact, for rectal, colon, and lung cancer, our data reveal substantial reductions in conversions across the board. Given that minimally invasive conversions are reportedly associated with higher rates of postoperative complications [20] and increased length of stay, we propose reductions in conversion as a potential mechanism for robotic length of stay improvement.
Another variable that may be contributing to the significant reduction in LOS for MIS and especially robotic procedures relate to less pain and decreased dependency on opioids in post-operative care. Several studies have reported that better pain management reduces hospital length of stay [22,23]. MIS, especially robotic-assisted surgery, has been observed to have lower post-operative opioids use across multiple clinical specialties. In an analysis of thoracic lobectomy procedures from the Premier database, robotic patients received opioids less frequently, and with lower total and average daily doses, compared to those undergoing VATS and open procedures [24]. In a similar analysis of sigmoidectomies, robotic patients were administered lower doses of parenteral opioids in comparison to open or laparoscopic patients [25]. These findings are consistent with the results of an analysis within our own institution, where we found that minimally invasive techniques were associated with a reduced risk of prolonged opioid use [26].
Our study identified lower readmission rates when patients underwent minimally invasive procedures for colorectal, lung, and kidney cancer, with additional improvements for those patients who underwent robotic procedures. A 2017 study of robotic prostate surgery revealed a decreased LOS and 30-day readmissions for robotic surgery as compared to open surgery [13]. Similarly, reductions in readmission were noted for obese patients with robotic colorectal cancer procedures in a meta-analysis of laparoscopic versus robotic surgery [27]. Historically, shorter length of stay is often linked to higher risk of readmission [28,29], yet we did not identify an increased risk of readmission in our patients with minimally invasive procedures. Considering the importance of 30-day readmission for payers and policy makers, robotic procedures like other minimally invasive procedures do not seem to lead to a higher risk of readmission.
This study has several limitations. First, this represents a retrospective study of in hospital data without long-term follow-up. However, most acute postoperative complications and deaths often occurred during the initial postoperative period and should largely be captured in these data. Additionally, the policies and protocols regarding postoperative ICU admission may differ significantly across hospital systems with some prophylactically admitting major abdominal or thoracic surgery patients regardless of clinical status. While we could not truly assess hemodynamic status or vasopressor requirement of the patients within this study, the billing code of ICU admission was standardized across all groups and thus may serve as a standard estimate of this variable. Surgeon preference and decision-making for operative approach cannot be completely controlled for and may introduce selection bias in the open surgery though we included several hospital and surgeon characteristics in the IPTW model. Finally, the data and measured outcomes  within this study are dependent on appropriate ICD-9-CM, ICD-10, CPT, and billing coding and may be limited by misclassification or data entry error.
In conclusion, our study reveals substantial benefits in robotic surgery for patients with colorectal, lung, and kidney cancer. Many of the outcomes benefits for robotic procedures are shared by patients who undergo minimally invasive procedures, but the additional length of stay benefits are considerable. These additional outcomes benefits are without detriments in readmission, which is of particular importance when understanding downstream treatment effects. It is for these reasons that we can advise that there are both short term and sustained benefits to robotic procedures in the surgical treatment of cancer.

Declarations
Disclosure All the authors have not received payment or service from a third party for the submitted work. Rocco Ricciardi, Robert Neil Goldstone and Matthew Wszolek have no conflict of interest or financial ties to disclosure. Francone reported consultancy from Intuitive surgical outside the submitted works. Hugh Auchincloss is a proctor for Intuitive outside the submitted works. Alexander de Groot, I-Fan Shih and Yanli Li reported full-time employment from Intuitive Surgical outside the submitted works.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.