Pious D Patel1, Sanjana Salwi2, Campbell Liles3, Akshitkumar M Mistry4, Eva A Mistry5, Matthew R Fusco6, Rohan V Chitale7, Chevis N Shannon8. 1. Vanderbilt University School of Medicine, Nashville, TN, USA; Surgical Outcomes Center for Kids, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, USA. Electronic address: pious.d.patel@vanderbilt.edu. 2. Vanderbilt University School of Medicine, Nashville, TN, USA; Surgical Outcomes Center for Kids, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, USA. Electronic address: sanjana.salwi@vanderbilt.edu. 3. Department of Neurosurgery, Vanderbilt University Medical Center, Nashville, TN, USA; Surgical Outcomes Center for Kids, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, USA. Electronic address: david.c.liles.1@vumc.org. 4. Department of Neurosurgery, Vanderbilt University Medical Center, Nashville, TN, USA. Electronic address: akshitkumar.m.mistry@vumc.org. 5. Department of Neurology, Vanderbilt University Medical Center, Nashville, TN, USA. Electronic address: eva.a.mistry@vumc.org. 6. Department of Neurosurgery, Vanderbilt University Medical Center, Nashville, TN, USA. Electronic address: matthew.r.fusco@vumc.org. 7. Department of Neurosurgery, Vanderbilt University Medical Center, Nashville, TN, USA. Electronic address: rohan.chitale@vumc.org. 8. Department of Neurosurgery, Vanderbilt University Medical Center, Nashville, TN, USA; Surgical Outcomes Center for Kids, Monroe Carell Jr. Children's Hospital at Vanderbilt, Nashville, TN, USA. Electronic address: cshannon@asrm.org.
Abstract
INTRODUCTION: The National Inpatient Sample (NIS) has led to several breakthroughs via large sample size. However, utility of NIS is limited by the lack of admission NIHSS and 90-day modified Rankin score (mRS). This study creates estimates for stroke severity at admission and 90-day mRS using NIS data for acute ischemic stroke (AIS) patients treated with mechanical thrombectomy (MT). METHODS: Three patient cohorts undergoing MT for AIS were utilized: Cohort 1 (N = 3729) and Cohort 3 (N = 1642) were derived from NIS data. Cohort 2 (N=293) was derived from a prospectively-maintained clinical registry. Using Cohort 1, Administrative Stroke Outcome Variable (ASOV) was created using disposition and mortality. Factors reflective of stroke severity were entered into a stepwise logistic regression predicting poor ASOV. Odds ratios were used to create the Administrative Data Stroke Scale (ADSS). Performances of ADSS and ASOV were tested using Cohort 2 and compared with admission NIHSS and 90-day mRS, respectively. ADSS performance was compared with All Patient Refined-Diagnosis Related Group (APR-DRG) severity score using Cohort 3. RESULTS: Agreement of ASOV with 90-day mRS > 2 was fair (κ = 0.473). Agreement with 90-day mRS > 3 was substantial (κ = 0.687). ADSS significantly correlated (p < 0.001) with clinically-significant admission NIHSS > 15. ADSS performed comparably (AUC = 0.749) to admission NIHSS (AUC = 0.697) in predicting 90-day mRS > 2 and mRS > 3 (AUC = 0.767, 0.685, respectively). ADSS outperformed APR-DRG severity score in predicting poor ASOV (AUC = 0.698, 0.682, respectively). CONCLUSION: We developed and validated measures of stroke severity at admission (ADSS) and outcome (ASOV, estimate for 90-day mRS > 3) to increase utility of NIS data in stroke research.
INTRODUCTION: The National Inpatient Sample (NIS) has led to several breakthroughs via large sample size. However, utility of NIS is limited by the lack of admission NIHSS and 90-day modified Rankin score (mRS). This study creates estimates for stroke severity at admission and 90-day mRS using NIS data for acute ischemic stroke (AIS) patients treated with mechanical thrombectomy (MT). METHODS: Three patient cohorts undergoing MT for AIS were utilized: Cohort 1 (N = 3729) and Cohort 3 (N = 1642) were derived from NIS data. Cohort 2 (N=293) was derived from a prospectively-maintained clinical registry. Using Cohort 1, Administrative Stroke Outcome Variable (ASOV) was created using disposition and mortality. Factors reflective of stroke severity were entered into a stepwise logistic regression predicting poor ASOV. Odds ratios were used to create the Administrative Data Stroke Scale (ADSS). Performances of ADSS and ASOV were tested using Cohort 2 and compared with admission NIHSS and 90-day mRS, respectively. ADSS performance was compared with All Patient Refined-Diagnosis Related Group (APR-DRG) severity score using Cohort 3. RESULTS: Agreement of ASOV with 90-day mRS > 2 was fair (κ = 0.473). Agreement with 90-day mRS > 3 was substantial (κ = 0.687). ADSS significantly correlated (p < 0.001) with clinically-significant admission NIHSS > 15. ADSS performed comparably (AUC = 0.749) to admission NIHSS (AUC = 0.697) in predicting 90-day mRS > 2 and mRS > 3 (AUC = 0.767, 0.685, respectively). ADSS outperformed APR-DRG severity score in predicting poor ASOV (AUC = 0.698, 0.682, respectively). CONCLUSION: We developed and validated measures of stroke severity at admission (ADSS) and outcome (ASOV, estimate for 90-day mRS > 3) to increase utility of NIS data in stroke research.
Authors: W Wahood; A A Rizvi; Y Alexander; M A Alvi; K R Rajjoub; H Cloft; A A Rabinstein; W Brinjikji Journal: AJNR Am J Neuroradiol Date: 2021-11-04 Impact factor: 3.825
Authors: Jason J Wang; Casey E Pelzl; Artem Boltyenkov; Jeffrey M Katz; Jennifer Hemingway; Eric W Christensen; Elizabeth Rula; Pina C Sanelli Journal: J Am Coll Radiol Date: 2022-04-25 Impact factor: 6.240
Authors: María Carmen Lea-Pereira; Laura Amaya-Pascasio; Patricia Martínez-Sánchez; María Del Mar Rodríguez Salvador; José Galván-Espinosa; Luis Téllez-Ramírez; Fernando Reche-Lorite; María-José Sánchez; Juan Manuel García-Torrecillas Journal: Int J Environ Res Public Health Date: 2022-03-08 Impact factor: 3.390