Statistical Analysis
Statistical analyses were performed using SAS version 9.4 (SAS
Institute, Inc., Cary, NC., USA). For the two surgical approaches,
baseline and postoperative outcomes were analyzed using frequency tables
and histograms. Continuous variables are reported as mean and standard
deviation if normally distributed, or as median and interquartile range
if not normally distributed. Continuous variables were compared between
the two groups using either two sample independent t-tests or the
Wilcoxon rank sum test, depending on normality. Categorical variables
were compared using the Chi square or Fischer’s exact test, depending on
expected values.
Unadjusted survival between groups was compared by Kaplan-Meier methods
and comparisons made between groups by the Log-Rank test. Death was
accounted for in models of cumulative incidence of mitral reoperation
and progression of MR as described in the Fine-Grey Model [15,16].
Multivariable logistic regression was performed to determine which
preoperative variables were associated with a minimally invasive
surgical approach. A minimally invasive surgical approach was modeled as
the dependent variable and the following independent variables were
included in the model based on the univariate analysis, the medical
literature, and biologic plausibility: age, race (white vs. non-white),
previous cardiac surgery (yes vs. no), need for concurrent cardiac
procedure (yes vs. no), location of MV leaflet involvement on
preoperative echo (anterior vs. posterior vs. bileaflet), MV lesion
etiology (degenerative vs. non-degenerative), and diabetes (yes vs. no).
The C-statistic and Hosmer and Lemeshow goodness-of-fit test were
reported for the model.
To estimate the impact of surgical approach on the outcomes of interest,
the multivariable logistic regression model was used to calculate a
subject’s propensity for having a minimally invasive surgical approach.
Propensity scores were then entered in to a separate Cox proportional
hazard regression model to calculate adjusted hazard ratios for each
outcome. Statistics with a p value less than 0.05 were considered
statistically significant.