Background
The scientific evidence regarding rehabilitation following pulmonary embolism (PE) is limited.

Aim
Our aim was to determine the effect of exercise-based rehabilitation compared to usual care on exercise capacity in PE survivors with persistent dyspnoea.

Methods
We conducted an RCT at two hospitals in Norway. Patients with persistent dyspnoea 6?72 months after acute PE, were randomized 1:1 to rehabilitation or usual care. The rehabilitation consisted of two weekly sessions of physical exercise for 8 weeks, and one educational session. The primary endpoint was the difference in Incremental Shuttle Walk Test (ISWT) between groups at follow-up. Secondary endpoints were differences in Endurance Shuttle Walk Test (ESWT), Quality of life (QoL; EQ-5D and Pulmonary Embolism-QoL questionnaires) and dyspnoea (Shortness of Breath questionnaire).

Results
Of 211 subjects included, 108 (51%) were randomized to rehabilitation and 103 (49%) to usual care. At follow-up, participants in the rehabilitation group performed better on the ISWT than controls (mean difference 53.0 m, 95% CI: 17.7?88.3, P=0.0035). The rehabilitation group reported better score on the Pulmonary Embolism-QoL (0?100% scale); mean difference -4% (95% CI: -9%?0%, P=0.041), but there were no differences in generic QoL, dyspnoea scores, or the ESWT. No adverse events occurred during the intervention.

Conclusions
In patients with persistent dyspnoea following PE, those who underwent rehabilitation had better exercise capacity at follow-up than those receiving usual care. Rehabilitation following PE is promising, but further research is needed to assess the optimal patient selection, timing, mode and duration of rehabilitation.