For classified variables hazard ratios for each class were estimated in these "univariate" models; two patients with missing gender were excluded from the models including male as independent variable.
In univariate models, seven patient related variables were significantly associated with the patient score (table 3).
In this population modeling, nine patients were excluded: two patients did not meet the inclusion criteria, five patients had an incomplete course of medication, one patient self-medicated with other drugs, and one patient withdrew consent.
Sixthly, despite the fact that most patients with the eventual diagnosis of pulmonary embolism were identified as having a high probability of pulmonary embolism by the validated diagnostic prediction models, three patients were missed by all of the rules.
To illustrate, the estimated CD4 slopes from the model in two patients with specific baseline characteristics are shown in Table 3 and in Figure 2. The bold cells are when the estimated CD4 count slope falls between -20 and +20 cell/μL per year, which we considered as indicative of borderline CD4 count decreases.
We constructed subject-specific models of two patients implanted with instrumented knee prostheses that measured knee forces in vivo.
Separate models for the two patient groups had a low predictive value because of the small numbers of IUA-OLs, especially in patients who received autocross-linked polysaccharide.
After feature selection of the GC-MS data, 152 components were used for discrimination modeling between these two patient groups, of which 59 were identified as metabolites.
Using this model, differences between the two patient groups were adjusted for modality performance differences exhibited by healthy controls (anticipated from previous work; Hailstone et al., 2010).
The application of propensity weights reduced differences between the two models for all but two patient characteristics (gender, and whether the patient was in a depressive state at the start of the study) and attained standardised differences of less than 0.1, indicating negligible differences [ 31], for five of the twelve characteristics.
