Predicting Length of Stay Following Radical Nephrectomy Using the National Surgical Quality Improvement Program Database

Hypogonadism, frailty, and postoperative outcomes among men undergoing radical nephrectomy

BACKGROUND AND OBJECTIVE

Hypogonadism and frailty may impact postoperative outcomes for men undergoing radical nephrectomy (RN). We aimed to determine the prevalence of hypogonadism in men undergoing RN and whether hypogonadism and frailty are associated with adverse postoperative outcomes.

METHODS

We identified men undergoing RN between 2012 and 2021 using the IBM Marketscan database. Frailty was determined using the Hospital Frailty Risk Score (HFRS). Patients were considered to have hypogonadism if diagnosed <5 years before RN. Length of stay (LOS), complications, emergency department (ED) visits, and readmissions were evaluated between men with and without hypogonadism at the time of surgery. Subgroup analysis of men with hypogonadism was performed to determine the effect of testosterone replacement therapy (TRT) on clinical outcomes.

RESULTS

Among 13 598 men who underwent RN, 972 (7.1%) had hypogonadism. Men with hypogonadism were more frail compared to men without hypogonadism (HFRS: median: 8.2, interquartile range [IQR]: 5.2-11.7 vs. median: 7.0, IQR: 4.3-10.7, p < 0.001) and had increased incidence of postoperative ileus (13.0% vs. 10.8%, p = 0.045), acute kidney injury (25.5% vs. 21.6% p = 0.005), and cardiac arrest (1.2% vs. 0.6%, p = 0.034). Hypogonadism was not associated with LOS, 90-day ED visit or readmission. However, high-risk frailty was associated with increased risk of 90-day ED visit (hazard ratio [HR]: 2.1, 95% confidence interval [95% CI]: 1.9-2.4, p < 0.001) and 90-day inpatient readmission (HR: 2.6, 95% CI: 2.2-3.1, p < 0.001), compared to low-risk frailty patients. Among men with hypogonadism, TRT was not associated with any postoperative outcomes.

CONCLUSIONS

Hypogonadism and frailty should be considered in the preoperative evaluation for men undergoing RN as risk factors for adverse postoperative outcomes.

Are markers of survival associated with perioperative outcomes for tumor thrombectomy patients?

INTRODUCTION

Despite modern advances in surgical and perioperative technologies, management of renal cell carcinoma (RCC) with tumor thrombus (TT) is a morbid procedure that necessitates careful patient selection. It is not known whether established prognostic models for metastatic RCC are suitable prognostic tools for more immediate perioperative outcomes in patients with RCC with TT. We evaluated if established risk models for cytoreductive nephrectomy, as a potential extension of their purpose-built use, are associated with immediate perioperative outcomes in patients undergoing nephrectomy and tumor thrombectomy.

METHODS

Perioperative outcomes of patients who underwent radical nephrectomy and tumor thrombectomy for RCC were compared to presences of established predictors of long-term outcomes from prior risk models individually and as stratified by risk grouping (International Metastatic Renal-Cell Carcinoma Database Consortium [IMDC], Memorial Sloan Kettering Cancer Center [MSKCC], M.D. Anderson Cancer Center [MDACC], and Moffitt Cancer Center [MCC]). Wilcoxon rank-sum test or the Kruskal-Wallis test compared continuous variables and the chi-square test or Fisher's exact test compared categorical variables.

RESULTS

Fifty-five patients were analyzed with 17 (30.9%) being cytoreductive. Eighteen (32.7%) patients had a level III or higher TT. Individually, preoperative variables were inconsistently associated with perioperative outcomes. Poorer risk patients per the IMDC model had more major postoperative complications (Clavien-Dindo grade≥3, P = 0.008). For the MSKCC model, poorer risk patients had increased intraoperative estimated blood loss (EBL), longer length of stay (LOS), more major postoperative complications, and more likely to discharge to a rehabilitation facility (P < 0.05). Less favorable risk patients per MDACC model had increased LOS (P = 0.038). Poorer risk patients per the MCC model had increased EBL, LOS, major postoperative complications, and 30-day hospital readmissions (P < 0.05).

CONCLUSION

Overall, cytoreductive risks models were heterogeneously associated with perioperative outcomes in patients undergoing nephrectomy and tumor thrombectomy. Of available models, the MCC model is associated with more perioperative outcomes including EBL, LOS, major postoperative complications, and readmissions within 30 days when compared to the IMDC, MSKCC, and MDACC models.

The 5-factor frailty index for radical nephrectomy: Simplifying personalized preoperative risk-stratification

Radical nephrectomy is the gold standard treatment for large renal cell carcinoma. Given the rising incidence of renal cell carcinoma and higher prevalence of geriatric patients in the population, readily identifying patients preoperatively that are at risk for a more complicated postoperative course is critical. The 5-factor modified frailty index (5-IFi) is a scoring system that assigns 1 point for each of the following comorbidities: dependent functional status, diabetes, chronic obstructive pulmonary disease, congestive heart failure, and hypertension. Patients with higher 5-IFi scores have been shown to be at significant risk for increased postoperative morbidity and mortality in many cohorts, including patients that undergo radical nephrectomy. This simplified comorbidity index with only 5 components is much more clinically pragmatic than its predecessors. As we encounter an increasing volume of patients with renal cell carcinoma and geriatric surgical candidates, readily risk stratifying patients on a personalized basis may be informative for shared clinical and surgical-decision making.

Patient characteristics predicting prolonged length of hospital stay following robotic-assisted radical prostatectomy

Objective:

The objective of this study is to determine the preoperative patient characteristics predicting prolonged length of hospital stay (pLOS) following robotic-assisted radical prostatectomy (RARP).

Methods:

The National Surgical Quality Improvement Program (NSQIP) database was used to select patients who underwent RARP without other concomitant surgeries between 2008 and 2016. Patients’ demographics, comorbidities, and laboratory markers were collected to evaluate their role in predicting pLOS. The pLOS was defined as length of stay (LOS) >2 days. A multinomial logistic regression was constructed adjusting for postoperative surgical complications to assess for the predictors of pLOS.

Results:

We obtained data for 31,253 patients of which 20,774 (66.5%) patients stayed ⩽1 day, 6993 (22.4%) patients stayed for 2 days, and 3486 (11.2%) patients stayed for >2 days. Demographic variables – including body mass index (BMI) <18.5: odds ratio (OR) = 2.8, 95% confidence interval (CI) = [1.7–4.8]; smoking: OR = 1.2, 95% CI = [1.1–1.4]; and dependent functional status: OR = 3.1, 95% CI = [1.6–6.0] – were predictors of pLOS. Comorbidities – such as heart failure: OR = 4.6, 95% CI = [2.0–10.8]; being dialysis dependent: OR = 2.7, 95% CI = [1.4–5.0]; and predisposition to bleeding: OR = 2.0, 95% CI =  [1.5–2.7] – were the strongest predictors of extended hospitalization. In addition, pLOS was more likely to be associated with postoperative bleeding, renal, or pulmonary complications.

Conclusion:

Preoperative patient characteristics and comorbidities can predict pLOS. These findings can be used preoperatively for risk assessment and patient counseling.

Predicting Prolonged Length of Hospital Stay for Peritoneal Dialysis-Treated Patients Using Stacked Generalization: Model Development and Validation Study

Background

The increasing number of patients treated with peritoneal dialysis (PD) and their consistently high rate of hospital admissions have placed a large burden on the health care system. Early clinical interventions and optimal management of patients at a high risk of prolonged length of stay (pLOS) may help improve the medical efficiency and prognosis of PD-treated patients. If timely clinical interventions are not provided, patients at a high risk of pLOS may face a poor prognosis and high medical expenses, which will also be a burden on hospitals. Therefore, physicians need an effective pLOS prediction model for PD-treated patients.

Objective

This study aimed to develop an optimal data-driven model for predicting the pLOS risk of PD-treated patients using basic admission data.

Methods

Patient data collected using the Hospital Quality Monitoring System (HQMS) in China were used to develop pLOS prediction models. A stacking model was constructed with support vector machine, random forest (RF), and K-nearest neighbor algorithms as its base models and traditional logistic regression (LR) as its meta-model. The meta-model used the outputs of all 3 base models as input and generated the output of the stacking model. Another LR-based pLOS prediction model was built as the benchmark model. The prediction performance of the stacking model was compared with that of its base models and the benchmark model. Five-fold cross-validation was employed to develop and validate the models. Performance measures included the Brier score, area under the receiver operating characteristic curve (AUROC), estimated calibration index (ECI), accuracy, sensitivity, specificity, and geometric mean (Gm). In addition, a calibration plot was employed to visually demonstrate the calibration power of each model.

Results

The final cohort extracted from the HQMS database consisted of 23,992 eligible PD-treated patients, among whom 30.3% had a pLOS (ie, longer than the average LOS, which was 16 days in our study). Among the models, the stacking model achieved the best calibration (ECI 8.691), balanced accuracy (Gm 0.690), accuracy (0.695), and specificity (0.701). Meanwhile, the stacking and RF models had the best overall performance (Brier score 0.174 for both) and discrimination (AUROC 0.757 for the stacking model and 0.756 for the RF model). Compared with the benchmark LR model, the stacking model was superior in all performance measures except sensitivity, but there was no significant difference in sensitivity between the 2 models. The 2-sided t tests revealed significant performance differences between the stacking and LR models in overall performance, discrimination, calibration, balanced accuracy, and accuracy.

Conclusions

This study is the first to develop data-driven pLOS prediction models for PD-treated patients using basic admission data from a national database. The results indicate the feasibility of utilizing a stacking-based pLOS prediction model for PD-treated patients. The pLOS prediction tools developed in this study have the potential to assist clinicians in identifying patients at a high risk of pLOS and to allocate resources optimally for PD-treated patients.

Length of Stay Beyond Medical Readiness in a Neurosurgical Patient Population and Associated Healthcare Costs

BACKGROUND

Length of stay beyond medical readiness (LOS-BMR) leads to increased expenses and higher morbidity related to hospital-acquired conditions.

OBJECTIVE

To determine the proportion of admitted neurosurgical patients who have LOS-BMR and associated risk factors and costs.

METHODS

We performed a prospective, cohort analysis of all neurosurgical patients admitted to our institution over 5 mo. LOS-BMR was assessed daily by the attending neurosurgeon and neuro-intensivist with a standardized criterion. Univariate and multivariate logistic regressions were performed.

RESULTS

Of the 884 patients admitted, 229 (25.9%) had a LOS-BMR. The average LOS-BMR was 2.7 ± 3.1 d at an average daily cost of $9 148.28 ± $12 983.10, which resulted in a total cost of $2 076 659.32 over the 5-mo period. Patients with LOS-BMR were significantly more likely to be older and to have hemiplegia, dementia, liver disease, renal disease, and diabetes mellitus. Patients with a LOS-BMR were significantly more likely to be discharged to a subacute rehabilitation/skilled nursing facility (40.2% vs 4.1%) or an acute/inpatient rehabilitation facility (22.7% vs 1.7%, P < .0001). Patients with Medicare insurance were more likely to have a LOS-BMR, whereas patients with private insurance were less likely (P = .048).

CONCLUSION

The most common reason for LOS-BMR was inefficient discharge of patients to rehabilitation and nursing facilities secondary to unavailability of beds at discharge locations, insurance clearance delays, and family-related issues.

Risk factors affecting hospital stay among patients undergoing colon cancer surgery: a prospective cohort study

PURPOSE

To identify and validate risk factors that contribute to prolonged length of hospital stay (LOS) in patients undergoing resection for colorectal cancer.

METHODS

This prospective cohort study included 1955 patients admitted to 22 hospitals for primary resection of colorectal cancer. Multivariate analyses were used to identify and validate risk factors, randomizing patients into a derivation and a validation cohort. Multiple correspondence and cluster analysis were performed to identify clinical subtypes based on LOS.

RESULTS

The strongest independent predictors of prolonged LOS were postoperative reintervention, surgical site infection, open surgery, and distant metastasis. The multiple correspondence and cluster analysis provided three groups of patients in relation to prolonged LOS: patients with the longest LOS included the highest percentage of patients with open surgery, distant metastasis, deep surgical site infections, emergency admissions, additional diagnostic factors, and highly contaminated surgical sites. Patients with prolonged LOS (> 14 days) were more likely to develop adverse outcomes within 30 days after discharge.

CONCLUSIONS

Patients undergoing resection of colorectal cancer cluster into different groups based on LOS of the index admission. Those with prolonged LOS were more likely to develop adverse outcomes within 30 days after discharge. Some of the strongest independent predictors of prolonged LOS, such as surgical infections or open surgery, could be modified to reduce LOS and, in turn, other adverse outcomes.

TRIAL REGISTRATION

NCT02488161.

Pre- and intra-operative predictors of postoperative hospital length of stay in patients undergoing radical prostatectomy for prostate cancer in China: a retrospective observational study

Background

Hospital length of stay (LOS) has recently been receiving increasing attention as a marker of medical resource consumption. Identifying predictors of longer LOS can better equip doctors to counsel patients and facilitate more efficient patient flow and utilization of medical resources. The objective of this study was to identify pre- and intra-operative risk factors for postoperative hospital LOS in patients who had undergone radical prostatectomy in China.

Methods

We retrospectively analyzed data of 793 eligible patients with prostate cancer who had undergone radical prostatectomy in our institution between January 2011 and March 2016. Relevant preoperative variables, including patient characteristics, medical comorbidities, prostate cancer disease-specific variables, urinary tract symptoms, preoperative laboratory values, and intraoperative variables including operation type, operation duration, and blood loss, were analyzed. The outcome was postoperative length of stay which was calculated as the time from the date of operation to the date of discharge. Multiple linear regression analysis was used to identify predictors of this outcome.

Results

The mean postoperative LOS was 11.7 days (±4.6 days) and the median 10 days (range, 5–46 days). According to univariate and multivariate analysis, operation type (open or laparoscopic), blood loss, Gleason score (≥8) and preoperative laboratory values of white blood count (WBC) were found to be the main explanatory predictors of postoperative LOS of patients with prostate cancer in our institution. Additionally, open surgery was the strongest significant predictor of longer LOS according to the standardized coefficients in this model.

Conclusions

Our findings indicate that significant predictors of longer postoperative LOS in patients who have undergone radical prostatectomy in China include both preoperative variables of Gleason score, WBC and intraoperative variables of operation type (open or laparoscopic), blood loss. To shorten hospital LOS in patients with prostate cancer and optimize utilization of Chinese medical resources, efforts should be made to improve the intraoperative process and reduce the prevalence of preoperative risk factors.

Length of hospital stay after craniotomy for tumor: a National Surgical Quality Improvement Program analysis

OBJECT

Although the length of hospital stay is often used as a measure of quality of care, data evaluating the predictors of extended hospital stay after craniotomy for tumor are limited. The goals of this study were to use multivariate regression to examine which preoperative characteristics and postoperative complications predict a prolonged hospital stay and to assess the impact of length of stay on unplanned hospital readmission.

METHODS

Data were extracted from the National Surgical Quality Improvement Program (NSQIP) database from 2007 to 2013. Patients who underwent craniotomy for resection of a brain tumor were included. Stratification was based on length of hospital stay, which was dichotomized by the upper quartile of the interquartile range (IQR) for the entire population. Covariates included patient age, sex, race, tumor histology, comorbidities, American Society of Anesthesiologists (ASA) class, functional status, preoperative laboratory values, preoperative neurological deficits, operative time, and postoperative complications. Multivariate logistic regression with forward prediction was used to evaluate independent predictors of extended hospitalization. Thereafter, hierarchical multivariate logistic regression assessed the impact of length of stay on unplanned readmission.

RESULTS

The study included 11,510 patients. The median hospital stay was 4 days (IQR 3-8 days), and 27.7% (n = 3185) had a hospital stay of at least 8 days. Independent predictors of extended hospital stay included age greater than 70 years (OR 1.53, 95% CI 1.28%-1.83%, p < 0.001); African American (OR 1.75, 95% CI 1.44%-2.14%, p < 0.001) and Hispanic (OR 1.68, 95% CI 1.36%-2.08%) race or ethnicity; ASA class 3 (OR 1.52, 95% CI 1.34%-1.73%) or 4-5 (OR 2.18, 95% CI 1.82%-2.62%) designation; partially (OR 1.94, 95% CI 1.61%-2.35%) or totally dependent (OR 3.30, 95% CI 1.95%-5.55%) functional status; insulin-dependent diabetes mellitus (OR 1.46, 95% CI 1.16%-1.84%); hematological comorbidities (OR 1.68, 95% CI 1.25%-2.24%); and preoperative hypoalbuminemia (OR 1.78, 95% CI 1.51%-2.09%, all p ≤ 0.009). Several postoperative complications were additional independent predictors of prolonged hospitalization including pulmonary emboli (OR 13.75, 95% CI 4.73%-39.99%), pneumonia (OR 5.40, 95% CI 2.89%-10.07%), and urinary tract infections (OR 11.87, 95% CI 7.09%-19.87%, all p < 0.001). The C-statistic of the model based on preoperative characteristics was 0.79, which increased to 0.83 after the addition of postoperative complications. A length of stay after craniotomy for tumor score was created based on preoperative factors significant in regression models, with a moderate correlation with length of stay (p = 0.43, p < 0.001). Extended hospital stay was not associated with differential odds of an unplanned hospital readmission (OR 0.97, 95% CI 0.89%-1.06%, p = 0.55).

CONCLUSIONS

In this NSQIP analysis that evaluated patients who underwent craniotomy for tumor, much of the variance in hospital stay was attributable to baseline patient characteristics, suggesting length of stay may be an imperfect proxy for quality. Additionally, longer hospitalizations were not found to be associated with differential rates of unplanned readmission.