Twenty-one case studies were identified (32 patients, range 1 to 5). Eight studies met the suggestive quality level, 10 were at the preponderance level and three were at the conclusive level.
Diet manipulation (15 studies): The mean NAP across studies was 90.9%. The most common approach was increasing the amount of food consumed by providing additional food during meals and snacks during the day. Other approaches were encouraging slower eating by pacing food presentation, removing specific types of food or liquids from meals and changing the consistency or texture of food.
Sensory-base interventions (five studies): The mean NAP across four studies was 82.6%. Approaches included providing oral stimulation by chewing gum or a plastic chew ring (three studies), automatic reinforcement from spraying a liquid containing a preferred flavour in their mouths (one study) and visual screening for 30 seconds after vomiting (one study).
Socially mediated reinforcers (six studies): The mean NAP across studies was 86.3%. Approaches included attention during or following mealtimes (three studies), verbal praise for not vomiting (two studies) and escaping from the demands of staff (one study). Punishment only formed part of the intervention approach in two studies.
Functional analysis (five studies): Four out of five studies concluded that vomiting and/or rumination was maintained by automatic reinforcement. One study concluded that vomiting was maintained by escaping staff demands.
Details were provided of three studies that considered the mechanism of action of interventions.