Table 3.

GRADE evidence profile

Certainty assessment							№ of patients		Effect	Certainty
№ of studies	Study design	Risk of bias	Inconsistency	Indirectness	Imprecision	Other considerations	Nursing intervertions delivered totally or partially at home	usual care	Absolute (95% CI)
Self-care Maintenance
7	randomised trials	seriousa	not serious	not serious	seriousb	none	205	211	MD 7.26 higher (5.2 higher to 9.33 higher)	⨁⨁◯◯ Low
Self-care Management
6	randomised trials	seriousa	not serious	not serious	seriousc	none	194	198	MD 5.02 higher (1.34 higher to 8.69 higher)	⨁⨁◯◯ Low
Self-care behaviours
2	randomised trials	not serious	not serious	seriousd	not serious	none	193	201	MD 7.91 lower (9.29 lower to 6.54 lower)	⨁⨁⨁◯ Moderate

CI: confidence interval; MD: mean difference

Explanations

a. Most studies were at moderate risk of bias

b. Choosing a Δ of 3.2 (derived from a clinically relevant improvement of 8 standardized points suggested by the authors of the instrument) and using the standard deviations associated with the studies included in the analysis, the optimal information size (OIS) criterion is not met

c. Choosing a Δ of 1.6 (derived from a clinically relevant improvement of 8 standardized points suggested by the authors of the instrument) and using the standard deviations associated with the studies included in the analysis, the optimal information size (OIS) criterion is not met

d. The two studies included only patients with an LVEF lower than 45% and hospitalized for acute decompensation. These criteria do not fully represent the overall population with HF; therefore, considering the applicability of these results, we deemed to downgrade indirectness by one level