Impact of pre-and peri-operative risk factors on length of stay and hospital readmission following minimally-invasive partial nephrectomy

Vness A.Luks,Rhul Dutt,Ashok K.Heml,Mtvey Tsivin,Timothy E.Crven,Nihols A.Deeel,Dvid D.Thiel,Rm Anil Pthk,＊

a Brigham and Women’s Hospital,Division of Urology,Department of Surgery,75 Francis Street,Boston,MA,USA

b Atrium Health Wake Forest Baptist Medical Center,Department of Urology,1 Medical Center Blvd,Winston-Salem,NC,USA

c Mayo Clinic Florida,Department of Urology,4500 San Pablo Road,Jacksonville,FL,USA

KEYWORDS Minimally-invasive partial nephrectomy;The American College of Surgeons National Surgical Quality Improvement Program;Length of stay;Hospital readmission

Abstract Objective: We conducted an analysis of the American College of Surgeons National Surgical Quality Improvement Program database for minimally-invasive partial nephrectomy cases reported with the goal to identify pre-and peri-operative variables associated with length of stay (LOS) greater than 3 days and readmission within 30 days.Methods: Records from 2008 to 2018 for “laparoscopy,surgical;partial nephrectomy” for prolonged LOS and readmission cohorts were compiled.Univariate analysis with Chi-square,t-tests,and multivariable logistic regression analysis with odds ratios (ORs), p-values,and 95% confidence intervals assessed statistical associations.Results: Totally,20 306 records for LOS greater than 3 days and 15 854 for readmission within 30 days were available.Univariate and multivariable analysis exhibited similar results.For LOS greater than 3 days,undergoing non-elective surgery(OR=5.247),transfusion of greater than four units within 72 h prior to surgery (OR=5.072),pre-operative renal failure or dialysis(OR=2.941),and poor pre-operative functional status(OR=2.540)exhibited the strongest statistically significant associations.For hospital readmission within 30 days,loss in body weight greater than 10% in 6 months prior to surgery (OR=2.227) and bleeding disorders(OR=2.081)exhibited strongest statistically significant associations.Conclusion: Multiple pre-and peri-operative risk factors are independently associated with prolonged LOS and hospital readmission within 30 days of surgery using the American College of Surgeons National Surgical Quality Improvement Program data.Recognizing the risks factors that can potentially be improved prior to minimally-invasive partial nephrectomy is crucial to informing patient selection,optimization strategies,and patient education.

1.Introduction

Partial nephrectomy(PN)is the standard of care for T1 renal tumors less than 4 cm[1].Minimally invasive techniques are well suited for PN.While pure laparoscopic PN (LPN) has proved to be technically challenging with limited provider adoption,the growing ubiquity of robotic surgical systems and surgeon comfort with robotic techniques have advanced the utilization of PN versus radical nephrectomy(RN).

In addition to surgeon preference,robotic-assisted PN(RAPN) as compared to LPN has been reported to have shorter warm ischemia time,decreased rates of positive surgical margins,fewer conversions to open surgery,fewer major complications,and fewer complications overall [2].The benefits of RAPN are also upheld as compared to open PN with reduced blood loss,surgical complications,and length of stay (LOS) [3,4].All of the above support the trend demonstrating that RAPN accounted for nearly 66% of the increase in PN cases based on an analysis of renal mass management trends from 2000 to 2011 [5].Contemporary reviews state that RAPN is positioned to become the reference standard for PN [1,6,7].

With minimally-invasive PN (MIPN) supported by both guidelines and clinical trial data,recognizing the modifiable risk factors for poor surgical outcomes is crucial.Elderly and medically complex patients are at increased risk for prolonged LOS and readmission;both are indicators of poor surgical outcomes [8].Both general surgical complications,such as surgical site infection,hemorrhage,and thromboembolic events [9,10],as well as PN-specific complications,such as urinary leak and gross hematuria [10],are cited as causes of readmission within 30 days after surgery.

The American College of Surgeons National Surgical Quality Improvement Program (ACS-NSQIP) database is a peer-developed,validated,risk-adjusted,and outcomes-based program for the measurement and improvement of surgical care [11].Unlike other quality improvement efforts,ACS-NSQIP is compiled by clinically trained personnel.Patient outcomes are manually assessed within 30 days after surgery without reliance on billing codes.It has demonstrated superiority over single-[12]and multi-institution [13]datasets for capturing outcome measures within 30 days.We conducted an analysis of MIPN cases reported to the ACS-NSQIP database with the goal to identify pre-and peri-operative variables associated with LOS greater than 3 days and readmission within 30 days.

2.Materials (patients) and methods

The Current Procedural Code 50 543,“laparoscopy,surgical;partial nephrectomy”,was searched in the ACS-NSQIP participant use data files for years of surgery 2008-2018.The data for readmission within 30 days after surgery were only available beginning in 2012.Patient demographics and pre-operative characteristics were examined overall,by LOS greater than or less than 3 days,and by whether or not patients were readmitted to the hospital within 30 days after surgery.

Patient demographics,pre-operative characteristics,and post-surgical outcomes were extracted.Only patients with complete observations for LOS,readmission,and all selected covariates were included in the final analysis.Ages greater than 90 years old were truncated for patient confidentiality purposes by ACS-NSQIP.The pre-operative characteristics that were assessed included: year surgery conducted,Hispanic ethnicity,other race as compared to Caucasian,body mass index (BMI) greater than 30 kg/m2,diabetes requiring treatment,smoking within the last year,dyspnea,severe chronic obstructive pulmonary disease(COPD),congestive heart failure (CHF) within 30 days,hypertension (HTN) requiring treatment,chronic steroid use,bleeding disorders,open wound or wound infection,greater than 10% loss of body weight in last 6 months,emergent case,transfusion of greater than 4 units within 72 h prior to surgery,non-elective case,renal failure or dialysis,systemic sepsis,poor baseline functional status,American Society of Anesthesiologists (ASA) classification score greater than or equal to 3,and contaminated or dirty wound.

Descriptive statistics for continuous and categorical factors included means and standard deviations (SDs) or frequencies and percentages,respectively.Univariate comparisons between groups were performed usingt-test for continuous factors with Satterthwaite’s adjustment for unequal variances when appropriate,and Chi-square tests for categorical factors with Fisher’s exact test in situations when warranted by small expected cell counts.LOS and readmission cohorts were each analyzed by logistic regression models,and the relationships between pre-operative characteristics and those outcomes were examined.Multivariable logistic regression models were used to estimate odds ratios (ORs),95% confidence intervals (CIs),andp-values in separate models for each outcome.

3.Results

ACS-NSQIP data from 20 402 MIPN cases were identified.Full datasets for the following variables were available:20 306 patients for LOS and 15 854 patients for readmission.Within these groups,17.3% of patients (3506/20 306) reported a LOS of greater than 3 days and 6.0% of patients(958/15 854)were readmitted within 30 days after surgery.

3.1.Overall population

3.1.1.Demographics

Mean age was 59 (SD 12) years and BMI was 30.9 (SD 6.7)kg/m2.Distribution of surgery by year was skewed toward later year with 3%,19%,24%,and 55% for 2008-2010,2011-2013,2014-2015,and 2016-2018,respectively.Table 1 summarized pre-operative characteristics for the total population.

3.2.LOS

3.2.1.Demographics

For LOS greater than 3 days and less than or equal to 3 days,mean ages were 61 (SD: 13) years and 59 (SD: 12) years,BMIs were 31.2 (SD: 7.1) kg/m2and 30.8 (SD: 6.6) kg/m2,and total LOS was 5.9 (SD: 5.0) days and 1.9 (SD: 0.8) days(p＜0.01 for all).Distribution of LOS by surgery year favored LOS of less than or equal to 3 days with 72%,80%,81%,and 85% for 2008-2010,2011-2013,2014-2015,and 2016-2018,respectively (p＜0.0001 for all) (Table 1).

3.2.2.Univariate analysis

Univariate analysis of factors for LOS greater than 3 days was performed and summarized in Table 2.Overall,factors displaying statistical significance (p＜0.01) included Hispanic ethnicity,Caucasian,other or unknown race,diabetes mellitus requiring treatment,dyspnea,severe COPD,CHF within 30 days,HTN requiring treatment,bleeding disorders,more than 10% loss of body weight in last 6 months,emergent case,transfusion of greater than four units within 72 h prior to surgery,non-elective case,renal failure or dialysis,systemic sepsis,partially or totally dependent functional status,and ASA classification score.

3.2.3.Multivariable analysis

Table 3 summarizes the OR and 95% CI for multivariate analysis.Compared to Caucasian patients,African-American,and other or unknown race patients exhibited statistically significant association for prolonged LOS with OR of 1.179 (95% CI: 1.039-1.337) and 2.021 (95% CI:1.825-2.239),respectively (p＜0.0001).OR for Hispanic ethnicity was 0.763(95% CI:0.646-0.901,p=0.0014).When compared to surgeries conducted between 2016 and 2018,earlier duration cohorts had an increased association with prolonged LOS for 2008-2010,2011-2013,and 2014-2015 exhibiting OR of 2.338(95% CI:1.908-2.866),1.427(95% CI:1.295-1.572),and 1.277(95% CI:1.165-1.400),respectively(p＜0.0001 for all).

Factors with strong associations,OR greater than 2.50,included undergoing non-elective surgery (OR=5.247,95% CI:4.123-6.677),transfusion of greater than four units within 72 h prior to surgery (OR=5.072,95% CI:1.760-14.617),pre-operative renal failure or dialysis(OR=2.941,95% CI: 1.763-4.906),and poor pre-operative functional status (OR=2.540,95% CI: 1.713-3.767)(p＜0.0001 for all,but for transfusionp=0.003).

3.3.Readmission

3.3.1.Demographics

Mean ages were 61(SD:13)years and 59(SD:12)years,BMIs 31.3 (SD: 6.9) kg/m2and 30.8 (SD: 6.7) kg/m2,and total LOSs 3.4 (SD: 3.2) days and 2.6 (SD: 2.8) days for readmission ＜30 days after surgery and no readmission ＜30 days after surgery,respectively (p＜0.0001 for age and LOS,p＜0.05 for BMI).Distribution of readmission by surgery year favored no readmission with 95%,95%,and 93% for 2011-2013,2014-2015,and 2016-2018,respectively(p＜0.0001 for all).

3.3.2.Univariate analysis

Univariate analysis of factors for readmission within 30 days was performed and summarized in Table 2.Overall,factors displaying statistical significance (p＜0.01) included BMI greater than or equal to 30 kg/m2,diabetes mellitus requiring treatment,dyspnea,severe COPD,HTN requiring treatment,chronic steroid use,bleeding disorders,greater than 10% loss of body weight in last 6 months,non-elective case of surgery,partially or totally dependent functional status poor baseline functional status,and ASA classification score.

3.3.3.Multivariable analysis

The multivariate analysis is summarized in Table 3.Race as compared to Caucasians was not statistically significant(p=0.250 for all).As compared to surgery performed from 2016 to 2018,surgeries performed in earlier years had a reduced and statistically significant (p＜0.0001) odds of readmission within 30 days for 2011-2013 (OR=0.601,95% CI: 0.504-0.718) and 2014-2015 (OR=0.678,95% CI:0.579-0.793).

Factors with strong associations,OR greater than 2,included bleeding disorders (OR=2.081,95% CI:1.475-2.935,p＜0.0001) and greater than 10% loss of body weight in last 6 months (OR=2.227,95% CI: 1.290-3.845,p=0.004).Although exhibiting strong associations,emergent cases (OR=2.823,95% CI: 0.986-8.086,p=0.053),or history of CHF within 30 days prior to surgery (OR=2.004,95% CI: 0.924-4.347,p=0.079) were not statistically significant.

4.Discussion

Through multivariable analysis of the ACS-NSQIP database,we found that LOS greater than 3 days is associated with undergoing non-elective surgery,transfusion of greater than four units within 72 h prior to surgery,pre-operative renal failure and/or dialysis,and poor pre-operative functional status.We also found that readmission within 30 days is associated with bleeding disorders and greater than 10% loss of body weight within 6 months of surgery.

The difference in total LOS between prolonged LOS cohorts was statistically significant-approximately 6 days in total for greater than 3 days and 2 days in total for less than or equal to 3 days.In addition to recent transfusion,renal failure and/or dialysis,and poor functional status,African-American race,and earlier year of surgery were also associated.These findings are not surprising,as non-Caucasian race and reduced functional status have been shown to be predictors of prolonged LOS [14].Increased surgeon experience over time may also decrease PN-specific adverse events,such as renal hemorrhage and urine leak;this may explain the finding of increased LOS with earlier surgical date [15,16].In ACS-NSQIP,the non-elective variable is potentially confounding as it is inclusive of patients that require pre-admission for any cause in addition to emergent or urgent cases.

Between the readmission within 30-day cohorts,there was a statistically significant difference in total LOS.While statistically significant,the clinical significance is difficult to appreciate as both cohorts exhibited approximately 3 days of hospitalization.Prolonged LOS may serve as a correlate for patient medical complexity as well as risk for readmission [14].Both of the variables associated with 30-day readmission are non-specific to PN.Frailty inclusive of pre-operative weight loss has been shown to be an independent risk factor for post-operative complication[17].Similarly,coagulopathy leads to increased risk of post-operative bleeding,which has been shown to be the most common urologic complication requiring readmission for RAPN [10].

The risk factors identified in this study are not absolute contraindications to MIPN,but may initiate discussion regarding pre-operative optimization or surgical delay given the slow rate growth and low rate of renal cell carcinoma stage progression or metastasis[18].To our knowledge,this is the first significant analysis of pre-and peri-operative factors affecting objective quality of care indicators for MIPN utilizing the ACS-NSQIP database.Overall,the majority of studies of prolonged LOS and hospital readmission after MIPN focus on comparing outcomes to RN [19],open approach [20],or general,and PN-specific perioperative factors,such as estimated blood loss,operative time,and ischemia time [21,22].Unlike the findings of this work,Shumate et al.[22]conducted an analysis of variables associated with prolonged LOS following RAPN completed by a single surgeon and found no predictive pre-operative variables.

Predicting LOS following MIPN is of interest to both patients and surgeons.The ACS-NSQIP Surgical Risk Calculator(http://riskcalculator.facs.org)[23]serves as a patient-and case-specific tool for guiding surgical decision-making and patient understanding of risk attributed to surgery.Its algorithm for estimated LOS is based on the ACS-NSQIP database.Unfortunately,prior efforts assessing the accuracy of the calculator in regards to LOS and complications for RAPN have demonstrated overestimations by the calculator,predicated LOS of 2.77 days versus actual LOS of 2.01 days(p＜0.0001)[24].Bazzi et al.[25]conducted a single center retrospective review to determine viable candidates for RAPN as an outpatient procedure with LOS cohorts of greater than or less than 1 day.Pre-operative factors such as higher ASA score and estimated glomerular filtration rate(GFR) were both statistically significant between groups,but their predictive model was unable to discriminate suitability for outpatient RAPN.A multi-institution study by Sentell et al.[26]provided strong evidence that outpatient RAPN is feasible without increased complications.

Brandao et al.[10]conducted a single center retrospective review of 627 patients undergoing RAPN with a focus on 30-day readmissions.They found a 30-day readmission rate of 4.46%,which was less than that of the ACS-NSQIP database at 6.04%.Unlike in the ACS-NSQIP dataset,the mean LOSs between the readmission cohorts were not statistically significant,3.4 (SD 1.7) days for non-readmitted and 4.0 (SD 2.4) days for readmitted.Overall,readmitted patients had significantly higher Charlson comorbidity index (CCI).CCI equal to 5 or more resulted in a 2.77-time higher odds of readmission and was the only statistically significant factor studied.Though our study did not evaluate CCI directly,our findings conform with these results that multiple comorbidities (i.e.,CHF and renal failure)in addition to a solid tumor increase the likelihood of readmission after surgery [27,28].

Our study is not without limitations,many of which are derived from characteristics intrinsic to the ACS-NSQIP database.Principally,the data are pooled from nearly 700 institutions,and therefore,represent a non-uniform dataset that cannot control for variations in case complexity,institutional protocols,differences in training of individual surgeons[29],volume of PN cases[30],and potential skew due to the opt-in nature of participant hospitals and their selected case sampling schema based on volume.Efforts to systematically sample and to monitor collection ensure unbiased selection and equal opportunity to capture all surgeons at each institution.The “case-mixed-adjusted” nature of the ACS-NSQIP database also accounts for case complexity,which varies greatly for MIPN.

Lastly,the Current Procedural Code used (50 543“laparoscopy,surgical;partial nephrectomy”) is inclusive of both LPN and RAPN,therefore,creating a heterogeneous dataset,which limits the conclusions for either approach.To date,the largest meta-analysis comparing LPN to RAPN by Leow et al.[2]contests prior assertions that there were no significant differences in peri-operative outcomes between approaches.Due to changes in surgical preference over time [31],the results may not accurately reflect the current state of post-operative outcomes [19].Given the ACS-NSQIP goal for generalizability of data collection across surgical specialties,no specific details pertaining to renal mass (i.e.,R.E.N.A.L.score or pre-surgical GFR),urologic history,or peri-operative factors specific to PN (i.e.,warm ischemia time,collection system injury,or post-surgical GFR) are collected.Addition of these variables would improve the utility of the ACS-NSQIP for assessing urologic outcomes.Lastly,ACS-NSQIP reporting is non-specific;therefore,the reason for readmission or LOS cannot be attributed to cause.

Future approaches to this work may include completing univariate and multivariate analysis for the post-operative complications reported to ACS-NSQIP to determine associations to LOS and readmission.Given the limitations for urologic specific complications,advocating for supplemental report for ACS-NSQIP reporting may be beneficial to improve the database’s utility for the urologic community.

5.Conclusion

The ACS-NSQIP database provides insight into mitigating risk factors associated with poor surgical outcomes after MIPN.Of the independent risk factors identified,some are intrinsic (i.e.,non-elective surgery) and others,such as functional status,renal function,nutrition,and coagulopathy,may be improved through pre-operative optimization strategies.

Author contributions

Study concept and design: Vanessa A.Lukas,Rahul Dutta,

Ashok K.Hemal,Matvey Tsivian,Timothy E.Craven,Nicholas A.Deebel,David D.Thiel,Ram Anil Pathak.

Data acquisition:Vanessa A.Lukas,Rahul Dutta,Timothy E.

Craven,Nicholas A.Deebel,Ram Anil Pathak.

Data analysis: Vanessa A.Lukas,Rahul Dutta,Ashok K.

Hemal,Matvey Tsivian,Timothy E.Craven,Nicholas A.

Deebel,David D.Thiel,Ram Anil Pathak.

Drafting of manuscript: Vanessa A.Lukas,Rahul Dutta,Nicholas A.Deebel,Ram Anil Pathak.

Critical revision of the manuscript:Vanessa A.Lukas,Rahul

Dutta,Ashok K.Hemal,Matvey Tsivian,Timothy E.Craven,Nicholas A.Deebel,David D.Thiel,Ram Anil Pathak.

Conflicts of interest

The authors declare no conflict of interest.