Evidence review

Clinical and technical evidence

Regulatory bodies

A search of the Medicines and Healthcare Products Regulatory Agency website revealed no manufacturer Field Safety Notices or Medical Device Alerts for this device. No reports of adverse events were identified from a search of the US Food and Drug Administration (FDA) database: Manufacturer and User Device Facility Experience (MAUDE).

Clinical evidence

A literature search found 124 citations (after removal of duplicates) and a relevant Cochrane review that reported on auto‑inflation using Otovent. Studies were included in this briefing if they reported randomised control trials of Otovent in English.

The Cochrane review (Perera et al. 2013) selected randomised controlled trials that compared any form of auto‑inflation (including Otovent and other treatments) with no auto‑inflation (observation without treatment) in people with otitis media with effusion. Using pooled data from 8 selected trials, the Cochrane review concluded that auto‑inflation, with the various methods used in the trials, seemed to have a beneficial effect on the resolution of otitis media with effusion, and could be considered as a treatment during the observation period.

Of the 8 trials identified in the Cochrane review, 4 used Otovent. One of these (De Nobili and Bellomo 2008) was excluded from this briefing because there was no English translation; the remaining 3 studies (Blanshard et al. 1993; Ercan et al. 2005; Stangerup et al. 1992) are included in this briefing, along with a more recent randomised controlled trial that was funded by the National Institute for Health Research Health Technology Assessment programme (Williamson et al. 2015).

The Williamson et al. (2015) study was set in 43 family practices in the UK (tables 1 and 2). Children aged 4 to 11 years with a recent history of ear symptoms and otitis media with effusion in 1 or both ears, confirmed by tympanometry, were allocated to have either auto‑inflation using Otovent 3 times daily for 1 to 3 months plus standard care, or standard care alone. Details of standard care were not further described in the study. Clearance of middle‑ear fluid at 1 and 3 months was assessed using tympanometry by experts masked to group allocation. Of the 320 children enrolled, those having auto‑inflation were more likely to have normal tympanograms at 1 month than those having standard care alone (47.3% [62/131] versus 35.6% [47/132]; relative risk [RR] 1.36) and at 3 months (49.6% [62/125] versus 38.3% [46/120]; RR 1.37). Auto‑inflation also produced greater improvements in ear‑related quality of life assessed by a validated tool (the OMQ‑14, a 14‑point questionnaire on the effect of otitis media with effusion). Compliance with Otovent treatment (use of the device 'most' or 'all of the time') was 89% at 1 month and 80% at 3 months. Adverse events were mild, infrequent and comparable between groups.

The Blanshard et al. (1993) trial enrolled 85 children aged between 3 and 10 years, who had chronic otitis media with effusion in both ears and were on the waiting list for grommet insertion surgery at a UK children's hospital (tables 3 and 4). After randomisation, 2 children were excluded from the analysis, leaving 42 children in the intervention group and 41 in the observation group. Those in the intervention group were given Otovent to use 3 times a day for the duration of the study, and children in both intervention and observation groups were then seen at monthly intervals for 3 months for pneumatic otoscopy and tympanometry. For the analysis, the treatment group was split into a high compliance group (greater than 70% compliance; n=19) and a low compliance group (less than 70% compliance; n=23). Compliance was measured as the number of times a child used the device as a percentage of the maximum possible. Statistically significant improvements, in both tympanometry and pneumatic otoscopy, were seen at 1, 2 and 3 months in those using Otovent in the subgroup with greater than 70% compliance when compared with the observation group. No statistically significant differences were seen between the low (less than 70%) compliance group and the observation group. No adverse treatment effects were seen.

The Ercan et al. (2005) study aimed to evaluate the longer‑term effects of auto‑inflation using the Otovent device on the need for surgical implantation of grommets (tables 5 and 6). The trial enrolled 60 children aged from 4 to 10 years with chronic otitis media with effusion in at least 1 ear (93 ears in total). The children were randomly divided into intervention and control groups using the total number of affected ears. The intervention group had auto‑inflation 3 times a day for 6 weeks with nasal saline irrigation, and the control group had nasal saline irrigation alone for 6 weeks. Compared with the control group, the auto‑inflation group showed a statistically significantly lower need for ventilation tubes (grommets) than the control group at 3 months (p=0.017), 6 months (p=0.040), and 9 months (p=0.015). Except for at the third month, the auto‑inflation group showed statistically significantly higher recovery rates from effusion than the control group.

The Stangerup et al. (1992) study was done to evaluate the effect of Otovent as a treatment for secretory otitis media (tables 7 and 8). One hundred children recruited from an ear, nose and throat hospital in Denmark were consecutively randomised to treatment and control groups and results were reported by ears (124 in total). The children were aged between 3 and 10 years and were entered into the study after having had secretory otitis media in 1 or both ears for at least 3 months, as verified by tympanometry. Tympanometry was repeated at 2 weeks and at 1, 2, and 3 months. After 2 weeks of auto‑inflation, 33 ears (65%) improved, 17 (33%) remained unchanged and 1 (2%) deteriorated. In the control group, 15% of ears improved, 71% remained unchanged, and 14% deteriorated (p<0.001). The children in the auto‑inflation group who still had type C2 tympanograms after 2 weeks were asked to continue using Otovent for at least a further 2 weeks. After further treatment of 27 ears, 7 (26%) improved and 20 (74%) remained unchanged. In the control group (55 ears), 14 ears (26%) improved, 33 (60%) remained unchanged and 8 (14%) deteriorated. No statistically significant differences were shown after 2 or 3 months.

Recent and ongoing studies

No ongoing or in‑development trials on Otovent nasal balloon for otitis media with effusion were identified.

Costs and resource consequences

According to NHS reference costs 2013–2014, NHS trusts and NHS foundation trusts in England did an annual total of 22,809 operations for the insertion of grommets, at an average cost of £802 per operation (Department of Health 2015). Introducing treatment with Otovent during the initial period of observation does not involve increased costs of monitoring above the cost of the device. Otovent might reduce costs by reducing the need for surgical intervention.

In England in 2014, 10,088 Otovent devices were dispensed at a net ingredient cost of £49,431 (Health and Social Care Information Centre 2015).

No changes would be needed in the organisation or delivery of current services and no other facilities or technologies would be needed.

Strengths and limitations of the evidence

The evidence included in this briefing is drawn from randomised controlled trials, the preferred study design when comparing treatments. The Cochrane review on auto‑inflation concluded that although pooled estimates of its effect favoured the treatment, there was no statistically significant effect on middle ear function as measured by tympanometry. The review deemed none of the included studies to be of high quality; all of the studies were relatively small, had limited treatment duration and short follow‑up periods. Of the studies included in the Cochrane review that used Otovent as the intervention, only Ercan (2005) assessed the effect on the need for surgery in the longer term. Although Otovent is intended for use in adults as well as children, no studies were found that assessed its effectiveness in adults.

The Williamson et al. study (2015) enrolled more children and was of a higher quality than the studies included in the Cochrane review, with a reduced risk of bias. The study was an economic assessment of treatment options during the 3‑month observation period, and did not follow up participants to consider the long‑term cost‑effectiveness implications associated with any eventual grommet insertion surgery.

Using the Cochrane Collaboration 'risk of bias' tool, the risk from generation of the randomised sequence was rated as low in the Blanshard et al. (1993) and Williamson et al. (2015) studies, and unclear in the Stangerup et al. (1992) and Ercan et al. (2005) studies. The risk from failing to conceal the treatment allocation was unclear in all studies except Williamson et al. (2015) where it was rated as low. The risk from comparability of groups pre‑test was rated as unclear in all the studies except Stangerup et al. (1992) where it was rated as low. The risk from failing to blind the outcome assessment was rated as unclear in all studies. The risk from incomplete outcome data was rated as low in all studies except Stangerup et al. (1992), for which it was rated as high due to 42% attrition in the intervention arm and 67% attrition in the control arm. The risk from selective reporting was rated as low in all of the studies. For the studies included in the Cochrane review (Blanshard et al.1993; Ercan et al. 2005; Stangerup et al. 1992), the review authors assessed the risk of bias, and the same methodology was applied by the External Assessment Centre to assess the Williamson et al. (2015) study.