In cases of pneumothorax with a slow persistent air leak, do you consider discharging patients with a Heimlich valve?  

The question arises as there are data on outpatient management of primary spontaneous and secondary spontaneous pneumothoraces but data seems to be sparse for iatrogenic pneumothoraces.

In patients who do not have sufficient air leak for an offending lobe to be identified for bronchoscopic one way valve placement, but a slow air leak causing pneumothorax build-up after a chest tube clamping trial, do you consider the use of a Heimlich valve with outpatient follow up?