OMQ-14 replaces OM2-13 as OME outcome measure

More powerful OMQ-14 instrument for health status & outcomes in OME replaces OM2-13 as Patient Reported Outcome Measure (PROM) for ENT

by Mark Haggard 7 Dec 2010

Since March 2009, a downloadable questionnaire, OM2-13, has been located in the Clinical Outcomes page of the ENT UK website. It will shortly be superseded by the better OMQ-14. The formatted questionnaire is anticipated to be available for the end of January 2011, and the scoring spreadsheet by the end of February. Anyone planning audits with such an instrument in the near future is invited to contact Mark Haggard (mph38@cam.ac.uk).

Previous OM2-13 lacked hearing items (assumed objective audiometry available)

The previous OM2-13 respected the domain structure of impact in OME (hearing, physical health, developmental impact); however on the assumption that audiometry (HL, objective) would normally be available, it did not include reported hearing items. Thus the acceptable number of items for a short form was "spent" on supporting the other two domains, hence the 2 in the title. As is usual in generating such instruments, a consistency criterion was followed: there was a progressive retention of the best items for reflecting these known domains, via the facets within them, from the full TARGET measures (76 items), down through OM8-30 (32 items) to OM2-13 (13 items).

OMQ-14 has better correlation with generic Quality of Life measures

The principle was endorsed on Friday Dec 3rd 2010 by ENT-UK's Clinical Audit and Practice Advisory Group that a PROM should bear as close an overall relationship to a generic measure of quality of life (QoL) as is possible,  that is, even if many of its constituent elements are signs or symptoms from the specific organ system affected in the disease in question. It is not always easy to achieve this, which is why such an external prediction approach is less common than the internal consistency approach. There are several statistical reasons why it is difficult, but various new steps to overcome these obstacles, including the use of very large samples have now been taken. A strand of work using data from the GOME trial and the recruitment stages of the TARGET trial has now succeeded in doing this (see Dakin et al Qual Life Res, 2010, 19: 65-80, for the first of three relevant publications).

Consequences for clinical ENT

Two main consequences for clinical ENT follow:
 
(1) It has become possible to aggregate across the relevant outcome measures to make a single composite QoL one which aggregates for reliability and determines the relative weightings of the various outcome measures in a rationally and empirically justified way. This aggregation simplifies and strengthens the TARGET clinical trial results, showing in persistent OME moderate benefits from a policy of selectively allocated ventilation tubes and adjuvant adenoidectomy, as described to BAPO 17th September 2010.

(2) The work has substantiated, and corroborated in two separate clinical trial databases, a method for generating a PROM which, desirably, is a short form of the lengthier trial and OM8-30 measures for QoL. This is done by retaining the best items from the point of view of their collectively best predicting separate QoL measures, but not necessarily all the items marking a good (ie predictive) facet. The two data bases each demand an item appropriate to maximising discrimination in their type of sample but not demanded by the other (more variation related to hearing problems in secondary care, more to AOM in primary care). Apart from this, they select the same 12 remaining items as best, making 14 in all with only one disagreement over the best 13 (ie independent  replication shows 92% agreement on selection of items).

"Better value" OMQ-14 is 70 percent more powerful for only one extra question

If the major criterion is accepted that such a PROM should correspond to QoL, then this new OMQ-14 is 1.7 times as good (widely expressed as "70% better") than OM2-13 at doing that, whilst being only 1.08 times as long (ie including 1 item more). It is thus a better buy for the respondent, the auditing clinician and the Department of Health.