RESULTS
A total of 312 patients were included, 150 from the control group and 162 from the intervention group (Fig 1). 13 patients presented protocol violation in the intervention group. Likewise, 13 patients were lost during follow-up in the control group and 3 in the intervention group, leaving a total of 137 patients in the control group and 146 in the intervention group, in the per protocol analysis. Table 1 shows the baseline characteristics of the participants.
Table 2 shows the main results in the control and intervention groups. The overall mortality at 90 days, 180 days and 1 year per protocol was 8 (2.57%), 10 (3.22%) and 14 (4.50%), respectively. There were no significant differences in mortality after one year between both groups, both unadjusted and when adjusted for the Charlson comorbidity index (ITT: adjusted OR 0.89, 95% CI 0.30-2.69, p=0.84; PP: OR adjusted 1.04, 95% CI 0.33-3.26, p=0.94). The rate of new admissions per year in the control group was 27.01% while it was 25.52% in the intervention group, p = 0.84. The same analysis was performed for each of the hospitals without showing any significant differences. Fig 2a and 2b show the analysis of 1-year survival by intention to treat (Hazard ratio (95% CI) =0.91 (0.32, 2.61), p=0.87), and per protocol (Hazard ratio (95% CI) =1.08 (0.36, 3.22), p =0.89), respectively, with no significant differences detected in either case. On the other hand, 16 (10.74%) cardiovascular events per year were observed in the control group and 23 (14.38%) in the intervention group (adjusted OR 1.40, 95% CI 0.71-2.77, p value=0.33) in the per protocol analysis (Table 2).
A sample was obtained for biomarker analysis from 65 patients in the control group and 81 in the intervention group, all from one of the hospitals. The biomarker levels at admission, day 5, and day 30 as well as the difference from day 5 to day 30 are shown in Table 3. In the per protocol analysis, after adjusting for the value at day 5, no significant differences were observed between the control and intervention groups with respect to the difference in proadrenomedullin levels from day 5 to day 30 (p=0.52). Neither were significant differences detected for the PCR (p=0.2910) or in the PCT (p=0.44).
Last, the effect of change over time in biomarker values on the rate of cardiovascular events at follow-up was assessed (Table 4). 22 cardiovascular events were observed in this patient sample. Changes in proADM values from day 5 to day 30 showed an OR of 1.11 (95% CI: 0.09, 13.65) () of having any cardiovascular event, adjusted for its value on day 5 and by the Charlson index (p=0.94). The same analysis was performed taking into account the control and intervention groups, and again no significant effect was detected for cardiovascular events at 1 year in either the control or intervention group (proADM difference: p=0.79 and p=0.86, respectively; PCR difference: p=0.38 and p=0.20, respectively; and PCT difference: p=0.94 and p=0.63, respectively).