Study designs of evaluations included in the review
Randomised controlled trials (RCTs). Trials were excluded if they had no control group or were not randomised.
Specific interventions included in the review
For the aromatherapy intervention (with no independent replication, i.e. only 1 trial for each intervention) participants received either:
1. Inhalation of vaporised aromatic mixture or water,
2. Oral administration of aromatic liquid or placebo.
3. Dummy cigarette with black pepper smell, or menthol smell or no treatment.
4. Bath water with natural or synthetic lavender oil or another synthetic oil with a smell.
5. Inhalation of steam plus aroma or steam alone.
6. Daily massage of carrier oil into scalp with or without a mixture of essential oils.
For the massage intervention (with independent replications, i.e. more than one trial for each intervention) participants received various forms of massage with or without the addition of oils or aromas to the foot, limbs, head, 'areas available to the therapist', or full body. Other interventions for the massage group were: routine care, no treatment, undisturbed rest, or a 20-minute chat. The frequency and duration of interventions varied between studies.
Trials were excluded if they pertained to studies of local effects such as the antiseptic effects of tea tree oil.
Participants included in the review
For the aromatherapy intervention (with no independent replication), healthy adults with common cold, chronic bronchitis patients, male smokers, post-partum women, healthy volunteers, or patients with alopecia areata. In the trials for anxiety (with independent replications), participants were cancer patients and post-operative patients (following cardiac surgery) being treated in a hospital setting and a mixed group of patients in an intensive therapy unit.
Trials were excluded if they were pre-clinical studies of healthy volunteers.
Outcomes assessed in the review
Anxiety, well-being, symptom scores, and stress measured using questionnaires such as the Speilberger State-Trait Anxiety Inventory or the Rotterdam Symptom Checklist.
How were decisions on the relevance of primary studies made?
Two authors independently performed the selection of papers for inclusion.