Skip to main content
. 2015 Mar 12;17(3):e63. doi: 10.2196/jmir.4174

Table 8.

Summary of findings from the meta-analyses with the most direct evidence in each group for the outcome of all-cause hospitalizations.a-h

Outcome: Comparison Number of participants (studies)
Follow-up
Quality of evidence
GRADEa-d
Relative effect (95% CI) Anticipated absolute effects
Risk with comparator Risk difference with HT (95% CI)
All-cause hospitalization: Comparison 1 [34]

Population: Stable and recently discharged patients 2343 (8 studies)
6-12 months
LOW due to risk of bias, inconsistency, and imprecisione-g RR 0.91 (0.84 to 0.99) 521 per 1000 47 fewer per 1000 (from 5 fewer to 83 fewer)
Intervention: Home telemonitoring with clinical support provided during office hours or 24/7, without home visits for clinical assessment or educational purposes
Comparator group: usual care
All-cause hospitalization: Comparison 2 [37]

Population: Recently discharged patients (≤28 days) 831 (5 studies)
6-12 months
LOW due to risk of bias, inconsistency, imprecisione-g HR 0.67 (0.42 to 0.97)h 569 per 1000 138 fewer per 1000 (from 11 fewer to 271 fewer)
Intervention: Home telemonitoring with clinical support provided during office hours, without home visits for clinical assessment or educational purposes
Comparator group: usual care
All-cause hospitalization: Comparison 3 [37]

Population: Patients with stable heart failure 1267 (5 studies)
6-12 months
LOW due to risk of bias, imprecisione,f HR 1.17 (0.89 to 1.59)h 357 per 1000 47 more per 1000 (from 32 fewer to 148 more)
Intervention: Home telemonitoring with clinical support provided during office hours, without home visits for clinical assessment or educational purposes
Comparator group: usual care
All-cause hospitalization: Comparison 4 [37]

Population: Patients with stable heart failure 1258 (3 studies)
12-24 months
LOW due to risk of bias, inconsistency, imprecisione-g HR 0.84 (0.54 to 1.15)h 474 per 1000 57 fewer per 1000 (from 181 fewer to 48 more)
Intervention: Home telemonitoring with clinical support provided 24/7, without home visits for clinical assessment or educational purposes
Comparator group: usual care
All-cause hospitalization: Comparison 5 [35]

Population: Stable and recently discharged patients 787 (3 studies)
3-12 months
LOW due to risk of bias, imprecisione,f RR 0.79 (0.66 to 0.94) 438 per 1000 92 fewer per 1000 (from 26 fewer to 149 fewer)
Intervention: Home telemonitoring with clinical support provided during office hours or 24/7, with or without home visits for clinical assessment or educational purposes
Comparator group: usual care

aHigh quality: Further research is very unlikely to change our confidence in the estimate of effect.

bModerate quality: Further research is likely to have an important impact on our confidence in the estimate of effect and may change the estimate.

cLow quality: Further research is very likely to have an important impact on our confidence in the estimate of effect and is likely to change the estimate.

dVery low quality: We are very uncertain about the estimate.

eMost trials did not provide details of random sequence generation, allocation concealment, and blinding of data analysts or assessors (see Multimedia Appendix 5).

fThe optimal information size criterion was not met by the meta-analysis (power <80%).

gSerious unexplained inconsistency/heterogeneity (I2>70%). Point estimates and confidence intervals between RCTs varied considerably in magnitude and direction.

h95% credible intervals (Bayesian meta-analysis).