J Am Acad Orthop Surg Glob Res Rev. 2019 Oct; 3(10): e19.00108.

The Use of Older Versus Newer Data in the National Surgical Quality Improvement Program Database Influences the Results of Total Hip Arthroplasty Outcomes Studies

Blake N. Shultz, BA, Anoop R. Galivanche, BS, Taylor D. Ottesen, BS, Patawut Bovonratwet, BS, and Jonathan N. Grauer, MDcorresponding author

Hip

Introduction:

National databases, such as the National Surgical Quality Improvement Program (NSQIP) database, are frequently used for total hip arthroplasty (THA) studies. NSQIP variables and the population included in the database have evolved over time. These changes may influence the results of studies using different periods of data.

Methods:

THA patients were aggregated from the 2005 to 2010 and 2011 to 2015 NSQIP data sets to define two era groups. Demographic data and 30-day perioperative outcomes were compared between the groups. As an example analysis, multivariate Poisson regression was used to determine the correlation between age and perioperative outcomes for each group.

Results:

Of 102,411 THA patients identified, 8098 cases were from 2005 to 2010 and 94,313 were from 2011 to 2015. A number of preoperative characteristics and perioperative outcomes were significantly different between the era groups. Multivariate analysis of the 2005 to 2010 group showed that increasing age was significantly associated with urinary tract infection and length of stay (LOS), and multivariate analysis of the 2011 to 2015 group showed that age was significantly associated with urinary tract infection, LOS, 30-day mortality, unplanned reintubation, extended LOS, pneumonia, deep vein thrombosis/thrombophlebitis, blood transfusion, and return to the operating room.

Conclusion:

Significantly more THA patients were enrolled in the NSQIP in the years between 2005 and 2015. Populations in 2005 to 2010 versus 2011 to 2015 were associated with differences in preoperative characteristics and perioperative outcomes. In an example analysis, it was shown that these differences together lead to different study results and conclusions. This needs to be considered when interpreting and conducting studies using earlier NSQIP data.

Link to article