Negative results in long COVID clinical trials: choosing outcome measures for a heterogeneous disease