Abstract

Objective To evaluate the effectiveness of Listen, a self-management support intervention, for people living with long covid who were not in hospital. Design Pragmatic, multicentre, parallel group, randomised controlled trial. Setting Twenty four sites in England and Wales. Participants Identified from long covid clinic waiting lists, word of mouth, and adverts/social media self-referred to the trial, 554 adults with long covid were randomised to receive either the Listen trial intervention or NHS usual care. Interventions The Listen intervention involved up to six one-to-one personalised sessions with trained healthcare practitioners and an accompanying handbook co-designed by people with lived experience and health professionals. Usual NHS care was variable, ranging from no access, access to mobile applications and resources, and to specialist long covid clinics. Main outcome measures The primary outcome was the Oxford participation and activities questionnaire (Ox-PAQ) routine activities scale score at three months assessed in the intention-to-treat population. Secondary outcomes included Ox-PAQ emotional wellbeing and social engagement scale scores, the Short Form-12 (SF-12) health survey, the fatigue impact scale, and the generalised self-efficacy scale at three months. The EuroQol five-dimension five-level (EQ-5D-5L) assessed health utility. Serious adverse events were recorded. Results Between 27 May 2022 and 15 September 2023, 554 people with long covid (mean age 50 (standard deviation 12.3) years; 394 (72.4%) women) were randomly assigned. At three months, participants assigned to the intervention group reported small non-significant improvements in the primary outcome of capacity for daily activities as assessed by Ox-PAQ routine activities scale score (adjusted mean difference −2.68 (95% confidence interval (CI) −5.38 to 0.02), P=0.052) compared with usual NHS care. For the secondary outcomes, people receiving the intervention also reported significant improvements in mental health (Ox-PAQ emotional wellbeing −5.29 (95% CI −8.37 to −2.20), P=0.001; SF-12 2.36 (95% CI 0.77 to 3.96), P=0.004), reductions in fatigue (fatigue impact score −7.93 (95% CI −11.97 to −3.88), P<0.001), and increases in self-efficacy (generalised self-efficacy scale 2.63 (95% CI 1.50 to 3.75), P<0.001). No differences were found in social engagement (−2.07 (95% CI −5.36 to 1.22), P=0.218) or SF-12 physical health (0.32 (95% CI −0.93 to 1.57), P=0.612). No intervention related serious adverse events were reported. Conclusions The personalised self-management support intervention of the Listen trial resulted in non-significant short term improvements in routine activities when compared with usual care. Improvements in emotional wellbeing, fatigue, quality of life, and self-efficacy for people living with long covid were also reported. Physical health and social engagement were not affected by the trial intervention. The limited understanding of how much change is clinically meaningful in this population along with the unblinded design, the use of self-referral as a recruitment method and variable usual care may have introduced unintended bias and thus limits robust conclusions about this intervention. Further research is required to fully establish the impact of the intervention.