Abstract
Objectives:
A large portion of the body of orthopedic surgical research is conducted retrospectively, using the operative note as a main data source. Operative notes tend to infrequently report quantitative data, significantly limiting the precision with which research questions can be assessed (Scherer et al., 2003). Large scale surgical outcomes studies have identified specific surgical parameters as valid predictors of key surgical outcomes (Kaeding, Pedroza, Reinke, Huston, & Spindler, 2015). Thus, it is now feasible to develop a data capture system that collects more accurate, robust surgical data quickly and with lower cost than operative note dictation. The objectives of this study are as follows: 1. Develop a provider-friendly, evidence-based, data capture system for lower limb orthopedic surgeries.2. Assess the performance of the data capture system on the dimensions of agreement with operative note and implant log, consistency of data quality, and speed of provider input.
Methods:
A multidisciplinary team including administration, orthopedic surgery staff, and software development was assembled for the design of the surgical database. The database was developed in Redcap and user interfaces were created for Android and IOS operating systems. Branching logic was created to streamline provider data input. All data input was created with discrete fields. Cleveland Clinic owned I-Phones were used for surgeon data input. All surgeons at Cleveland Clinic Sports Health were included in the dataset. One hundred cases were randomly selected from the first four months of data collection (February- June 2015). Cases were limited to those undergoing ACL repair, meniscal repair, or both. Duplicate data on these 100 cases was collected via chart review. These two datasets were compared. Cohen’s Kappa statistic was used to assess agreement.
Results:
Table 1 shows the Kappa statistic and 95% confidence interval for agreement between the database and the chart review collected data. Agreement across all variables was greater than Kappa of 0.85. These variables include, but are not limited to: ACL status (Kappa = 0.904, 95% C.I.[0.717, 1.090]), medial meniscus treatment (Kappa = 0.992, 95% C.I0.860, 1.045]), and tibial screw type (Kappa =0.933, 95% C.I.[0.860, 1.007]). Moreover, the frequency of reporting key information was higher in I-phone based data capture system as compared to traditional operative note (graft strand number: 100% vs. 80%; femur tunnel size: 100% vs. 85%; tibial tunnel size: 100% vs. 81%). Median surgeon data input time was 151 seconds.
Conclusion:
The database and smartphone data capture tool shows almost perfect agreement (Kappa>0.81) for all data tested. Additionally, data are more comprehensive with near perfect provider completion (100% for all data tested). Additionally, provider data entry is efficient (median 151 s completion time). A well designed database and user friendly interface have greater potential for research utility, clinic efficiency, and thus cost effectiveness when compared to traditional operative notes. Widespread utilization of such tools can accelerate the pace and improve the quality of orthopedic clinical research.