Background: Owing to its hierarchical sampling structure, the National Inpatient Sample (NIS) often involves clusters of study data. We aimed to (1) understand the statistical methods of currently published orthopaedic studies using NIS and (2) examine the role of hierarchical modeling versus traditional regression analysis in a retrospective cohort of patients from NIS.

Methods: We conducted a systematic review to examine statistical methods of orthopaedic studies (published between 2005 and 2014) using NIS. Ultimately, 132 studies were identified. We noted percentages of studies that used hierarchical modeling versus traditional regression analysis. Using NIS, we identified a retrospective cohort of patients aged 70 years and older who underwent operative fixation of intertrochanteric hip fractures between 2008 and 2012. Patient comorbidities were tested for association with in-hospital mortality and length of stay in hierarchical linear models and traditional regression analysis. Statistical outcomes were compared between the two models.

Results: Seven of 28 (25%) measured comorbidities overestimated the significance of in-hospital mortality in traditional regression compared to hierarchical modeling. Similarly, traditional regression analysis overestimated significance in four of 28 (14%) comorbidities in increasing length of stay. Of the 132 studies, most (74%) used traditional regression analysis, few (7%) used univariate statistics, and even less (2%) used hierarchical modeling. According to these findings, between 11% to 20% of all orthopaedic studies published between 2005 and 2014 using NIS data were at risk of overestimating their clinical results.

Conclusions: Traditional regression analysis may overestimate significance in linear and logistic models owing to the inability to address clustered data. Because healthcare data is often clustered, hierarchical linear modeling should be employed to increase the specificity of outcomes-based research using NIS.

Included in

Orthopedics Commons



To view the content in your browser, please download Adobe Reader or, alternately,
you may Download the file to your hard drive.

NOTE: The latest versions of Adobe Reader do not support viewing PDF files within Firefox on Mac OS and if you are using a modern (Intel) Mac, there is no official plugin for viewing PDF files within the browser window.