Does ‘healing’ benefit patients with chronic symptoms? A quasi-randomized trial in general practice