TAVR and Cancer: Machine Learning-Augmented Propensity Score Mortality and Cost Analysis in Over 30 Million Patients

2020 ◽  
Author(s):  
Dominique Monlezun ◽  
Logan Hostetter ◽  
Prakash Balan ◽  
Nicolas Palaskas ◽  
Juan Lopez-Mattei ◽  
...  
Keyword(s):  
Machine Learning ◽  
Aortic Valve ◽  
Propensity Score ◽  
Cancer Patients ◽  
Hospital Admissions ◽  
Treatment Options ◽  
The United States ◽  
United States Department ◽  
Primary Malignancy ◽  
Multivariable Regression

Abstract Introduction: Cardiovascular disease (CVD) and cancer are the top mortality causes globally, yet little is known about how the diagnosis of cancer affects treatment options in patients with hemodynamically compromising aortic stenosis (AS). Patients with cancer often are excluded from aortic valve replacement (AVR) trials including both trials with transcatheter AVR (TAVR) and surgical AVR (SAVR). This study looks at how cancer may influence treatment options, and assess the outcome of cancer patients who undergo surgical or TAVR intervention. Additionally, we sought to quantitate and compare both clinical and cost outcomes for cancer and non-cancer patients. Methods: This population-based case-control study uses the most recent year available National Inpatient Sample (NIS (2016) from the United States Department of Health and Human Services’ Agency for Healthcare Research and Quality (AHRQ). Machine learning augmented propensity score adjusted multivariable regression was conducted based on the likelihood of undergoing TAVR versus MM and TAVR versus SAVR with model optimization supported by backward propagation neural network machine learning.Results: Of the 30,195,722 total hospital admissions, 39,254 (0.13%) TAVRs were performed, with significantly fewer performed in cancer versus non-cancer patients even in those of comparable age and mortality risk (23.82% versus 76.18%, p<0.001) despite having similar mortality. Multivariable regression in cancer patients demonstrated that mortality was similar for TAVR, MM, and SAVR, though LOS and cost was significantly lower for TAVR versus MM and comparable for TAVR versus SAVR. Patients with prostate cancer constituted the largest primary malignancy among TAVR patients including those with metastatic disease. There were no significant race or geographic disparities for TAVR mortality.Discussion: Comparison of aortic valve intervention in cancer patients with those without co-existing malignancy suggests that intervention is underutilized in the cancer population. This study suggests that as cancer patients including those with metastasis have similar clinical outcomes, patients who are symptomatic and those with higher risk aortic valve lesions should be offered the benefit of intervention. Modern techniques have reduced intervention-related adverse events, provided improved quality of life, and appear to be cost effective; these advantages should not be denied to patients on the basis of co-existing malignancy.

scholarly journals TAVR and cancer: machine learning-augmented propensity score mortality and cost analysis in over 30 million patients

2021 ◽  
Vol 7 (1) ◽  
Author(s):  
Dominique J. Monlezun ◽  
Logan Hostetter ◽  
Prakash Balan ◽  
Nicolas Palaskas ◽  
Juan Lopez-Mattei ◽  
...  
Keyword(s):  
Machine Learning ◽  
Aortic Valve ◽  
Propensity Score ◽  
Hospital Admissions ◽  
Treatment Options ◽  
The United States ◽  
Healthcare Research ◽  
United States Department ◽  
Patients With Cancer ◽  
Multivariable Regression

Abstract Introduction Cardiovascular disease (CVD) and cancer are the top mortality causes globally, yet little is known about how the diagnosis of cancer affects treatment options in patients with hemodynamically compromising aortic stenosis (AS). Patients with cancer often are excluded from aortic valve replacement (AVR) trials including trials with transcatheter AVR (TAVR) and surgical AVR (SAVR). This study looks at how cancer may influence treatment options and assesses the outcome of patients with cancer who undergo SAVR or TAVR intervention. Additionally, we sought to quantitate and compare both clinical and cost outcomes for patients with and without cancer. Methods This population-based case-control study uses the most recent year available National Inpatient Sample (NIS (2016) from the United States Department of Health and Human Services’ Agency for Healthcare Research and Quality (AHRQ). Machine learning augmented propensity score adjusted multivariable regression was conducted based on the likelihood of undergoing TAVR versus medical management (MM) and TAVR versus SAVR with model optimization supported by backward propagation neural network machine learning. Results Of the 30,195,722 total hospital admissions, 39,254 (0.13%) TAVRs were performed, with significantly fewer performed in patients with versus without cancer even in those of comparable age and mortality risk (23.82% versus 76.18%, p < 0.001) despite having similar hospital and procedural mortality. Multivariable regression in patients with cancer demonstrated that mortality was similar for TAVR, MM, and SAVR, though LOS and cost was significantly lower for TAVR versus MM and comparable for TAVR versus SAVR. Patients with prostate cancer constituted the largest primary cancer among TAVR patients including those with metastatic disease. There were no significant race or geographic disparities for TAVR mortality. Discussion Comparison of aortic valve intervention in patients with and without cancer suggests that interventions are underutilized in the cancer population. This study suggests that patients with cancer including those with metastasis have similar inpatient outcomes to patients without cancer. Further, patients who have symptomatic AS and those with higher risk aortic valve disease should be offered the benefit of intervention. Modern techniques have reduced intervention-related adverse events, provided improved quality of life, and appear to be cost effective; these advantages should not necessarily be denied to patients with co-existing cancer.

scholarly journals Effect Of Prior Cancer History on The Survival of Gallbladder Cancer Patients: A Propensity Score Matching Study Based on The SEER Database

2020 ◽  
Author(s):  
Jie Ren ◽  
Wei Liu ◽  
Qinglin Li ◽  
Ruixia Cui ◽  
Yingmu Tong ◽  
...  
Keyword(s):  
Propensity Score ◽  
Gallbladder Cancer ◽  
Propensity Score Matching ◽  
Cancer Patients ◽  
The United States ◽  
Primary Cancer ◽  
Seer Database ◽  
Primary Malignancy ◽  
Kaplan Meier ◽  
The Impact

Abstract Background: The effect of previous malignancy history on the survival of individuals with a second primary gallbladder cancer remains unclear. Therefore, this study was conducted to analyze the impact of previous malignancy history on the survival of individuals with gallbladder cancer and to compare the prognostic differences between gallbladder cancer patients with and without previous cancer.Methods: Extract the United States Surveillance, Epidemiology, and End Results (SEER) database from 2004 to 2015 for cases diagnosed with gallbladder cancer. The Kaplan-Meier curves and log-rank test were used to compare the survival difference between gallbladder cancer individuals with and without previous malignancy. Cox proportional hazards regression model was used to explore the risk factors of gallbladder cancer.Results: A total of 5861 patients with gallbladder cancer were enrolled, including 5622 (95.9%) patients without prior primary cancer and 239 (4.1%) patients with prior primary cancer. Patients with gallbladder cancer with prior primary malignancy were older, and the tumors were at localized and regional stages more frequently and more early stages. The Kaplan-Meier curves showed that gallbladder cancer patients with prior cancer had better overall survival (OS) (P=0.027) and gallbladder cancer-specific survival (GCSS) (P<0.001) before propensity score matching (PSM), and gallbladder cancer patients with prior cancer had better GCSS (P<0.001), and there was no difference in OS (P=0.113) between gallbladder cancer patients with and without prior cancer after PSM. Multivariable cox regression analysis revealed that prior malignancy history was not a risk factor for OS (HR=0.875, 95%CI: 0.752-1.018, P=0.084), but it was beneficial to GCSS (HR=0.404, 95%CI: 0.318-0.513, P<0.001).Conclusions: Gallbladder cancer individuals with previous primary malignancy have different clinical characteristics from those without previous primary malignancy. Gallbladder cancer patients with previous primary malignancy have better progress than those without previous malignancy.

Patterns of second-line chemotherapy for metastatic prostate cancer in the United States.

2012 ◽  
Vol 30 (5_suppl) ◽  
pp. 169-169
Author(s):  
Hongwei Wang ◽  
Laura Liao ◽  
Eliot Obi-Tabot ◽  
Robert Sands ◽  
Mathieu Rose ◽  
...  
Keyword(s):  
Prostate Cancer ◽  
Healthcare Costs ◽  
Metastatic Prostate Cancer ◽  
Hospital Admissions ◽  
Treatment Options ◽  
Patient Characteristics ◽  
The United States ◽  
Geographic Region ◽  
Treatment Groups ◽  
Comorbidity Burden

169 Background: Patients with prostate cancer progressing from 1st line (1L) docetaxel had limited treatment options available. This study is to evaluate patterns of 2nd line (2L) chemotherapy in US managed care between 2004 and 2010. Methods: Patients with metastatic prostate cancer (mPC) and treated with docetaxel as 1L chemotherapy after July 1, 2004 were ascertained from the OptumInsight database. We evaluated type and timing of chemotherapy and relationships between patient characteristics, physician specialty, healthcare costs and geographic region 6 months prior to 1L docetaxel and choice of 2L chemotherapy. Results: Patients (n=1,173) were on average 71 years old at the onset of 1L docetaxel. During a mean follow-up period of 18 months, 38% patients received 2L treatment. Out of the patients received 2L therapy, 32% received mitoxantrone (MITO), 24% with docetaxel rechallenge (RECH), 14% carboplatin (CARB), and 12% paclitaxel (PAC), plus 11% on combo therapy. An examination of the 2L treatment groups showed that during the 6 months prior to 1L docetaxel, the RECH group (n=101) was older (73yrs), had fewer hospital admissions (12%), lower comorbidity burden (Charlson Comorbidity Index (CCI)=7.4), lower total healthcare costs ($10,083), and 78% patients seeing an oncologist; relative to MITO group (n=143) with 70 years of age, 16% hospital admissions, CCI of 7.5, total healthcare costs of $12,074, and 80% seeing an oncologist. The combo group (n=52) was 66 years old, with 19% hospital admissions, CCI of 8, total healthcare costs of $20, 505, and 92% seeing an oncologist. Median time to MITO from the start of 1L docetaxel was 184 days, 309 days to RECH and 223 days to combo therapy. Midwest (36%) and West (37%) were more frequently using MITO than Northeast (26%) and South (31%), while RECH was more frequently used in Northeast. Conclusions: Patients with mPC in US were most frequently treated with MITO or RECH as 2L chemotherapy after 1L docetaxel. MITO was also given sooner than RECH, hence a valid comparator for comparative effectiveness evaluation on new 2L therapy. Rechallenge with docetaxel increased with time and was given to patients with lower disease burden and healthcare costs than the MITO or Combo group.

A Novel Data Mining on Breast Cancer Survivability Using MLP Ensemble Learners

2019 ◽  
Vol 63 (3) ◽  
pp. 435-447
Author(s):  
Mohsen Salehi ◽  
Jafar Razmara ◽  
Shahriar Lotfi
Keyword(s):  
Breast Cancer ◽  
Machine Learning ◽  
Data Mining ◽  
Cancer Patients ◽  
Cancer Incidence ◽  
The United States ◽  
Machine Learning Techniques ◽  
Breast Cancer Patients ◽  
Stacked Generalization ◽  
Learning Techniques

Abstract Breast cancer survivability has always been an important and challenging issue for researchers. Different methods have been utilized mostly based on machine learning techniques for prediction of survivability among cancer patients. The most comprehensive available database of cancer incidence is SEER in the United States, which has been frequently used for different research purposes. In this paper, a new data mining has been performed on the SEER database in order to investigate the ability of machine learning techniques for survivability prediction of breast cancer patients. To this end, the data related to breast cancer incidence have been preprocessed to remove unusable records from the dataset. In sequel, two machine learning techniques were developed based on the Multi-Layer Perceptron (MLP) learner machine including MLP stacked generalization and mixture of MLP-experts to make predictions over the database. The machines have been evaluated using K-fold cross-validation technique. The evaluation of the predictors revealed an accuracy of 84.32% and 83.86% by the mixture of MLP-experts and MLP stacked generalization methods, respectively. This indicates that the predictors can be significantly used for survivability prediction suggesting time- and cost-effective treatment for breast cancer patients.

Patterns of second-line chemotherapy for metastatic prostate cancer in the United States.

2012 ◽  
Vol 30 (15_suppl) ◽  
pp. e15193-e15193
Author(s):  
Hongwei Wang ◽  
Laura Liao ◽  
Eliot Obi-Tabot ◽  
Robert Sands ◽  
Mathieu Rose ◽  
...  
Keyword(s):  
Prostate Cancer ◽  
Healthcare Costs ◽  
Metastatic Prostate Cancer ◽  
Hospital Admissions ◽  
Treatment Options ◽  
Patient Characteristics ◽  
The United States ◽  
Geographic Region ◽  
Second Line ◽  
Comorbidity Burden

e15193 Background: Patients with prostate cancer progressing from first-line (1L) docetaxel have limited approved treatment options available. Study is to evaluate patterns of second-line (2L) chemotherapy in a US managed care between 2004 and 2010. Methods: Patients with metastatic prostate cancer (mPC) and treated with docetaxel as 1L chemotherapy after July 1, 2004 were ascertained from the OptumInsight database. We evaluated type and timing of chemotherapy and relationships between patient characteristics, physician specialty, healthcare costs and geographic region, 6 months prior to 1L docetaxel and choice of 2L chemotherapy. Results: Patients (N=1,173) were on average 71 yrs old at onset of 1L docetaxel. During a mean follow-up period of 18 months, 38% of patients received 2L treatment. Out of the patients received 2L therapy, 32% received mitoxantrone (MITO), 24% with docetaxel rechallenge (RECH), 14% carboplatin (CARB), and 12% paclitaxel (PAC), plus 11% on Combo therapy. Examination of the 2L treatment groups showed that during the 6 months prior to 1L docetaxel, the RECH group (n=101) was older (73yrs), had fewer hospital admissions (12%), lower comorbidity burden [Charlson Comorbidity Index (CCI)=7.4], lower total healthcare costs ($10,083), and 78% seeing an oncologist; relative to MITO group (n=143) with 70 yrs, 16% hospital admissions, CCI of 7.5, total healthcare costs of $12,074, and 80% seeing an oncologist. The combo group (n=52) was 66 years old, with 19% hospital admissions, CCI of 8, total healthcare costs of $20,505, and 92% seeing an oncologist. Median time to MITO from the start of 1L docetaxel was 184 days, 309 days to RECH and 223 days to Combo therapy. Midwest (36%) and West (37%) were more frequently using MITO than Northeast (26%) and South (31%), while RECH was more frequently used in Northeast. Conclusions: Patients with mPC in US were most frequently treated with MITO or RECH as 2L chemotherapy after 1L docetaxel. MITO was also given sooner than RECH, hence a valid comparator for comparative effectiveness evaluation on new 2L therapy. Rechallenge with docetaxel increased with time and was given to patients with lower disease burden and healthcare costs than the MITO or Combo group.

Survival associated with cachexia among SEER-Medicare beneficiaries with lung cancer.

2017 ◽  
Vol 35 (15_suppl) ◽  
pp. e18117-e18117
Author(s):  
Shweta Shah ◽  
Joshua Noone ◽  
Christopher Michael Blanchette ◽  
Susan T Arthur
Keyword(s):  
Lung Cancer ◽  
Propensity Score ◽  
Elderly Patients ◽  
Cancer Patients ◽  
Median Survival ◽  
The United States ◽  
Population Characteristics ◽  
Treatment Modalities ◽  
Lung Cancer Patients ◽  
The Impact

e18117 Background: Lung cancer is the leading cause of cancer death in the United States. It is estimated that 60% of lung cancer patients are afflicted with cancer-associated cachexia syndrome (CACS) and approximately 10% of these patients will die due to CACS. We examined the impact of CACS on survival among lung cancer elderly patients. Methods: We conducted a retrospective study using SEER-Medicare data. Patients were included if diagnosed with first primary lung cancer between January 1, 2005 and December 31, 2010, at least 66 years of age, and continuously enrolled in Medicare Parts A and B in the 12 months prior to diagnosis. We identified cachexia in lung cancer patients using ICD-9 codes. Descriptive statistics were used to identify population characteristics. Propensity score (1:1 nearest neighbor) matching was performed between cachectic and non-cachectic lung cancer patients to compare survival. Results: We identified 84,518 lung cancer patients. Of these, 2,536 (3%) developed CACS after lung cancer diagnosis. The most common comorbid conditions among cachectic and non-cachectic groups were chronic obstructive pulmonary disease (50% versus 45.62%), congestive heart failure (8.56% versus 13.38%), diabetes (7.41% versus 14.75%), peripheral vascular disease (3.82% versus 6.85%), and renal disease (3.63% versus 6.14%). Propensity score 1:1 matching for confounding bias and adjustment for immortal time bias resulted in a cohort of 3734 matched patients. Eighty-eight percent of patients in the cachectic group died during the follow-up period compared to 78% in the non-cachectic group. Median survival time among non-cachectic lung cancer patients was significantly longer than cachectic lung cancer patients (log-rank p < 0.0001). Specifically, median survival in non-cachectic patients was 201 days compared to 92 days among cachectic patients. Conclusions: The occurrence of CACS is independently associated with a significant decrease in survival among lung cancer elderly patients. The results of this study may be useful for identifying healthcare burden and planning treatment modalities for this population.

scholarly journals 372. Comparing the Outcome of COVID-19 in Cancer and Non-Cancer Patients: an International Multicenter Study

2020 ◽  
Vol 7 (Supplement_1) ◽  
pp. S256-S256
Author(s):  
Ray Y Hachem ◽  
Tarcila Datoguia ◽  
Bilal Siddiqui ◽  
Ana Fernandez Cruz ◽  
Nobuyoshi Mori ◽  
...  
Keyword(s):  
Mortality Rate ◽  
Cancer Patients ◽  
Hospital Admissions ◽  
Clinical Course ◽  
Far East ◽  
The United States ◽  
Material Support ◽  
International Multicenter Study ◽  
International Multicenter ◽  
Research Grant

Abstract Background Our objective was to describe the clinical course, risk factors and outcomes of patients infected with COVID-19 around the globe comparing cancer to non-cancer patients. Methods We conducted a retrospective cohort study of COVID-19 confirmed cases through an international multicenter collaboration including 17 centers around the world including the United States of America, Brazil, Europe, Far East, Middle East and Australia from January to date. We evaluated the patients’ clinical characteristics, clinical course of the disease, hospitalization and outcome. Death was considered to be COVID-associated if it occurred within 30 days from the time of diagnosis. Results Preliminary data on 571 patients included 186 cancer patients and 385 non-cancer patients. Cancer patients were more likely to have COPD and received steroids but were less likely to have COVID-related symptoms compared to non-cancer patients (84% vs 97%, p&lt; 0.0001). The rate of pneumonia with hypoxia, non-invasive ventilation and mechanical ventilation were similar in both groups. Despite the fact that hospital admissions were significantly higher in non-cancer patients (70% vs 56%, p&lt; 0.001), promising antiviral and immune-related therapy including remdesivir, convalescent plasma and immunomodulators were more commonly used in cancer patients compared to non-cancer patients (P=0.04). Cancer patients had a higher COVID-associated mortality rate compared to non-cancer patients (20% vs 11%, p=0.006). Conclusion Despite the fact that cancer patients received more frequent antiviral and immune-related therapy, the mortality rate among cancer patients was significantly higher than non-cancer patients. Disclosures Roy F. Chemaly, MD, MPH, FACP, FIDSA, Chimerix (Consultant, Research Grant or Support)Clinigen (Consultant)Merck (Consultant, Research Grant or Support)Novartis (Research Grant or Support)Oxford Immunotec (Consultant, Research Grant or Support)Shire/Takeda (Research Grant or Support)Viracor (Research Grant or Support) Issam I. Raad, MD, Citius (Other Financial or Material Support, Ownership interest)Cook Medical (Grant/Research Support)Inventive Protocol (Other Financial or Material Support, Ownership interest)Novel Anti-Infective Technologies (Shareholder, Other Financial or Material Support, Ownership interest)

Sign in / Sign up

Export Citation Format

Share Document