Abstract
Background
As the number of total knee arthroplasties (TKA) being performed continues to increase, the number of potential failures requiring revision surgery would also be expected to increase. This study analyzed the trends in revision TKA (rTKA) from 2009 to 2016.
Methods
The Nationwide Inpatient Sample (NIS) database was used to identify all rTKA by International Classification of Diseases, Ninth Revision (ICD-9) and Tenth Revision (ICD-10) codes. The diagnoses leading to revision, revision costs, patient and hospital characteristics, and major inpatient complications were compared between 2009 and 2016. Multivariate logistic regression analyses were used to calculate odds ratios (OR) for complications.
Results
A total of 453,770 rTKA were performed between 2009 and 2016 (Table 1). The incidence of rTKA between 2009 and 2016 decreased by 4.3% among all age groups and decreased 5.9% in those over 65 years of age. The mean age at revision remained unchanged at 65 years, the mean Length of Stay (LOS) decreased 19.5% from 4.1 to 3.3 days (p<0.001), and the mean hospital costs increased 18.7% from $23,103 to $27,427 (p<0.001). Male to Female ratio remained unchanged with Females making up 58% of rTKA. The rTKA patients in 2016 versus 2009 had significantly lower odds of requiring a blood transfusion (OR 0.26, p<0.001). The rate of myocardial infarction remained unchanged in 2016 versus 2009 (4.1% vs. 4.0%, p=0.52), as did mortality (0.1% vs. 0.2%, p=0.089). Patients in 2016 were significantly less likely to be discharged to a skilled nursing facility (SNF) (26.5% in 2016 vs. 31.6% in 2009, p<0.001). A significantly larger share of rTKA were performed in urban teaching hospitals in 2016 (66.6% in 2016 vs. 41.5% in 2009, p<0.001). Mechanical loosening remained the most common reason for rTKA and increased from 19.7% in 2009 to 28.4% in 2016. Infection remained the second most common reason for rTKA (14.7% in 2009 vs. 12.2% in 2016). Instability was the third most common reason for revision in 2016 (10.7%) but direct comparison with 2009 could not be made since ICD-9 did not include this diagnosis. When comparing 2009 to 2014, an increase in revision rates was noted. Upon inclusion of 2015 and 2016 data, a decrease is noted. The primary limitation of this study is the challenge faced when comparing results from before and after the October 2015 transition of ICD-9 to ICD-10 codes.
Conclusions
The incidence of rTKA had decreased slightly between 2009 and 2016 even while the volume of primary TKA continued to rise. Implant loosening became increasingly prevalent as a reason for revision and infection continued to plague patients after TKA. Positive trends were noted, such as decreased rates of blood transfusion, LOS and rates of discharge to SNF. Still, the revision burden remains large, costing approximately $1.4 billion annually and we must continue to attempt to improve outcomes in this patient population.
For figures, tables, or references, please contact authors directly.
