Predicting occurrence of spine surgery complications using big data modeling of an administrative claims database

John K. Ratliff, Raymond Balise, Anand Veeravagu, Tyler S. Cole, Ivan Cheng, Richard A. Olshen, Lu Tian

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Background: Postoperative metrics are increasingly important in determining standards of quality for physicians and hospitals. Although complications following spinal surgery have been described, procedural and patient variables have yet to be incorporated into a predictive model of adverse-event occurrence. We sought to develop a predictive model of complication occurrence after spine surgery. Methods: We used longitudinal prospective data from a national claims database and developed a predictive model incorporating complication type and frequency of occurrence following spine surgery procedures. We structured our model to assess the impact of features such as preoperative diagnosis, patient comorbidities, location in the spine, anterior versus posterior approach, whether fusion had been performed, whether instrumentation had been used, number of levels, and use of bone morphogenetic protein (BMP). We assessed a variety of adverse events. Prediction models were built using logistic regression with additive main effects and logistic regression with main effects as well as all 2 and 3-factor interactions. Least absolute shrinkage and selection operator (LASSO) regularization was used to select features. Competing approaches included boosted additive trees and the classification and regression trees (CART) algorithm. The final prediction performance was evaluated by estimating the area under a receiver operating characteristic curve (AUC) as predictions were applied to independent validation data and compared with the Charlson comorbidity score. Results: The model was developed from 279,135 records of patients with a minimum duration of follow-up of 30 days. Preliminary assessment showed an adverse-event rate of 13.95%, well within norms reported in the literature. We used the first 80% of the records for training (to predict adverse events) and the remaining 20% of the records for validation. There was remarkable similarity among methods, with an AUC of 0.70 for predicting the occurrence of adverse events. The AUC using the Charlson comorbidity score was 0.61. The described model was more accurate than Charlson scoring (p < 0.01). Conclusions: We present a modeling effort based on administrative claims data that predicts the occurrence of complications after spine surgery. Clinical Relevance: We believe that the development of a predictive modeling tool illustrating the risk of complication occurrence after spine surgery will aid in patient counseling and improve the accuracy of risk modeling strategies.

Original languageEnglish (US)
Pages (from-to)824-834
Number of pages11
JournalJournal of Bone and Joint Surgery - American Volume
Volume98
Issue number10
DOIs
StatePublished - May 18 2016
Externally publishedYes

Fingerprint

Spine
Databases
Area Under Curve
Comorbidity
Logistic Models
Bone Morphogenetic Proteins
ROC Curve
Counseling
Physicians

ASJC Scopus subject areas

  • Surgery
  • Orthopedics and Sports Medicine

Cite this

Predicting occurrence of spine surgery complications using big data modeling of an administrative claims database. / Ratliff, John K.; Balise, Raymond; Veeravagu, Anand; Cole, Tyler S.; Cheng, Ivan; Olshen, Richard A.; Tian, Lu.

In: Journal of Bone and Joint Surgery - American Volume, Vol. 98, No. 10, 18.05.2016, p. 824-834.

Research output: Contribution to journalArticle

Ratliff, John K. ; Balise, Raymond ; Veeravagu, Anand ; Cole, Tyler S. ; Cheng, Ivan ; Olshen, Richard A. ; Tian, Lu. / Predicting occurrence of spine surgery complications using big data modeling of an administrative claims database. In: Journal of Bone and Joint Surgery - American Volume. 2016 ; Vol. 98, No. 10. pp. 824-834.
@article{5144b64eee414e22be3b801a4a1e42cb,
title = "Predicting occurrence of spine surgery complications using big data modeling of an administrative claims database",
abstract = "Background: Postoperative metrics are increasingly important in determining standards of quality for physicians and hospitals. Although complications following spinal surgery have been described, procedural and patient variables have yet to be incorporated into a predictive model of adverse-event occurrence. We sought to develop a predictive model of complication occurrence after spine surgery. Methods: We used longitudinal prospective data from a national claims database and developed a predictive model incorporating complication type and frequency of occurrence following spine surgery procedures. We structured our model to assess the impact of features such as preoperative diagnosis, patient comorbidities, location in the spine, anterior versus posterior approach, whether fusion had been performed, whether instrumentation had been used, number of levels, and use of bone morphogenetic protein (BMP). We assessed a variety of adverse events. Prediction models were built using logistic regression with additive main effects and logistic regression with main effects as well as all 2 and 3-factor interactions. Least absolute shrinkage and selection operator (LASSO) regularization was used to select features. Competing approaches included boosted additive trees and the classification and regression trees (CART) algorithm. The final prediction performance was evaluated by estimating the area under a receiver operating characteristic curve (AUC) as predictions were applied to independent validation data and compared with the Charlson comorbidity score. Results: The model was developed from 279,135 records of patients with a minimum duration of follow-up of 30 days. Preliminary assessment showed an adverse-event rate of 13.95{\%}, well within norms reported in the literature. We used the first 80{\%} of the records for training (to predict adverse events) and the remaining 20{\%} of the records for validation. There was remarkable similarity among methods, with an AUC of 0.70 for predicting the occurrence of adverse events. The AUC using the Charlson comorbidity score was 0.61. The described model was more accurate than Charlson scoring (p < 0.01). Conclusions: We present a modeling effort based on administrative claims data that predicts the occurrence of complications after spine surgery. Clinical Relevance: We believe that the development of a predictive modeling tool illustrating the risk of complication occurrence after spine surgery will aid in patient counseling and improve the accuracy of risk modeling strategies.",
author = "Ratliff, {John K.} and Raymond Balise and Anand Veeravagu and Cole, {Tyler S.} and Ivan Cheng and Olshen, {Richard A.} and Lu Tian",
year = "2016",
month = "5",
day = "18",
doi = "10.2106/JBJS.15.00301",
language = "English (US)",
volume = "98",
pages = "824--834",
journal = "Journal of Bone and Joint Surgery - American Volume",
issn = "0021-9355",
publisher = "Journal of Bone and Joint Surgery Inc.",
number = "10",

}

TY - JOUR

T1 - Predicting occurrence of spine surgery complications using big data modeling of an administrative claims database

AU - Ratliff, John K.

AU - Balise, Raymond

AU - Veeravagu, Anand

AU - Cole, Tyler S.

AU - Cheng, Ivan

AU - Olshen, Richard A.

AU - Tian, Lu

PY - 2016/5/18

Y1 - 2016/5/18

N2 - Background: Postoperative metrics are increasingly important in determining standards of quality for physicians and hospitals. Although complications following spinal surgery have been described, procedural and patient variables have yet to be incorporated into a predictive model of adverse-event occurrence. We sought to develop a predictive model of complication occurrence after spine surgery. Methods: We used longitudinal prospective data from a national claims database and developed a predictive model incorporating complication type and frequency of occurrence following spine surgery procedures. We structured our model to assess the impact of features such as preoperative diagnosis, patient comorbidities, location in the spine, anterior versus posterior approach, whether fusion had been performed, whether instrumentation had been used, number of levels, and use of bone morphogenetic protein (BMP). We assessed a variety of adverse events. Prediction models were built using logistic regression with additive main effects and logistic regression with main effects as well as all 2 and 3-factor interactions. Least absolute shrinkage and selection operator (LASSO) regularization was used to select features. Competing approaches included boosted additive trees and the classification and regression trees (CART) algorithm. The final prediction performance was evaluated by estimating the area under a receiver operating characteristic curve (AUC) as predictions were applied to independent validation data and compared with the Charlson comorbidity score. Results: The model was developed from 279,135 records of patients with a minimum duration of follow-up of 30 days. Preliminary assessment showed an adverse-event rate of 13.95%, well within norms reported in the literature. We used the first 80% of the records for training (to predict adverse events) and the remaining 20% of the records for validation. There was remarkable similarity among methods, with an AUC of 0.70 for predicting the occurrence of adverse events. The AUC using the Charlson comorbidity score was 0.61. The described model was more accurate than Charlson scoring (p < 0.01). Conclusions: We present a modeling effort based on administrative claims data that predicts the occurrence of complications after spine surgery. Clinical Relevance: We believe that the development of a predictive modeling tool illustrating the risk of complication occurrence after spine surgery will aid in patient counseling and improve the accuracy of risk modeling strategies.

AB - Background: Postoperative metrics are increasingly important in determining standards of quality for physicians and hospitals. Although complications following spinal surgery have been described, procedural and patient variables have yet to be incorporated into a predictive model of adverse-event occurrence. We sought to develop a predictive model of complication occurrence after spine surgery. Methods: We used longitudinal prospective data from a national claims database and developed a predictive model incorporating complication type and frequency of occurrence following spine surgery procedures. We structured our model to assess the impact of features such as preoperative diagnosis, patient comorbidities, location in the spine, anterior versus posterior approach, whether fusion had been performed, whether instrumentation had been used, number of levels, and use of bone morphogenetic protein (BMP). We assessed a variety of adverse events. Prediction models were built using logistic regression with additive main effects and logistic regression with main effects as well as all 2 and 3-factor interactions. Least absolute shrinkage and selection operator (LASSO) regularization was used to select features. Competing approaches included boosted additive trees and the classification and regression trees (CART) algorithm. The final prediction performance was evaluated by estimating the area under a receiver operating characteristic curve (AUC) as predictions were applied to independent validation data and compared with the Charlson comorbidity score. Results: The model was developed from 279,135 records of patients with a minimum duration of follow-up of 30 days. Preliminary assessment showed an adverse-event rate of 13.95%, well within norms reported in the literature. We used the first 80% of the records for training (to predict adverse events) and the remaining 20% of the records for validation. There was remarkable similarity among methods, with an AUC of 0.70 for predicting the occurrence of adverse events. The AUC using the Charlson comorbidity score was 0.61. The described model was more accurate than Charlson scoring (p < 0.01). Conclusions: We present a modeling effort based on administrative claims data that predicts the occurrence of complications after spine surgery. Clinical Relevance: We believe that the development of a predictive modeling tool illustrating the risk of complication occurrence after spine surgery will aid in patient counseling and improve the accuracy of risk modeling strategies.

UR - http://www.scopus.com/inward/record.url?scp=84978785873&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84978785873&partnerID=8YFLogxK

U2 - 10.2106/JBJS.15.00301

DO - 10.2106/JBJS.15.00301

M3 - Article

C2 - 27194492

AN - SCOPUS:84978785873

VL - 98

SP - 824

EP - 834

JO - Journal of Bone and Joint Surgery - American Volume

JF - Journal of Bone and Joint Surgery - American Volume

SN - 0021-9355

IS - 10

ER -