Twelve studies (9,963 patients overall) were included in the review: 5 RCTs (962 patients), 3 non-randomised controlled trials (1,912 patients) and 4 before-and-after studies (7,089 patients).
Clinical outcomes (2 RCTs): no level I or II studies showed that SOUs improve or worsen clinical outcomes. Both RCTs showed no significant difference between SOUs and usual care.
Length of stay (1 RCT and 1 non-randomised controlled trial): neither study showed that SOUs increased the length of stay. The RCT showed that SOUs plus a 'rapid' protocol reduced the average length of stay compared with routine care. The non-randomised controlled trial showed no difference between SOUs and direct hospital admission.
Efficiency of the emergency department (1 before-and-after study): no level I or II studies showed that SOUs hindered the organisation of the emergency department. The study showed that SOUs reduced the average time in the emergency department for some patient subgroups (chest pain and asthma), but had no effect on the time spent in the department for other subgroups (participants with sickle cell crisis or seizures).
Re-presentation to the emergency department (1 RCT and 1 before-and-after study): the RCT showed no difference in re-presentation rates between treatments for children with asthma, while the before-and-after study showed an increase (3 to 5%) in re-presentations for patients with asthma.
Medical admissions (2 RCTs and 1 before-and-after study): both RCTs showed that SOUs decreased the number of medical admissions. However, the before-and-after study found no difference in initial discharge rates, or the number of patients eventually hospitalised, for patients treated within an observation unit.
Quality of life (1 RCT): the RCT showed that the SOU was associated with an improvement in quality of life measures compared with hospital in-patients.