Overall willingness to pay was higher in HIC settings [ 16, 37, 39, 42, 44– 46] compared to MIC settings [ 35, 41, 47] or LIC settings [ 31], and lower in supervised HIVST [ 16, 37, 42, 44] than for unsupervised HIVST [ 41, 47].
Even though the numbers of dietician and LSA meetings were lower in the supervised program, our study showed that the participants of the supervised program were more satisfied with the intervention than the participants of the start-up program.
Despite a significantly lower number of supervised TEE examinations at M6 (17 [14–28] versus 30.5 [21.5–39.5], p = 0.0004, Table 2), trainees in the interventional group reached the same average score as in the control group (37.5 [33 39] versus 36 [33.5–37.5], respectively, p = 0.24, Table 2).
LBE is the ensemble-based supervised lower bound.
Values marked with bold font represent performances of the semi-supervised experiments that outperform the supervised lower bound.
The number of participants that were referred to a dietician was significantly lower in the supervised group (82.9% versus 67.5%).
Moreover, the proportion of participants which was referred to the dietician was approximately 15% lower in the supervised program.
The second column, LBE, shows the results of the supervised lower bound, which is also an ensemble, consisting of supervised classifiers.
Although PT guidance should be the only guidance that differs between the two programs, numbers of individual meetings with LSA and dietician were significantly lower in the supervised program.
Following baseline testing, participants were randomized to either 10 weeks of: 1) physiotherapist-supervised lower limb muscle strengthening exercises, or 2) no exercises (control).
