Survival analysis of orthodontic retainers.

Abstract

Retainers play a critical role in the long-term success of orthodontic treatment. The aims of this study were to evaluate the survival time of different retainers and to investigate the potential predictors of the retainer failures. A total of 591 retainers from 309 patients (between 2003 and 2014) were included in the study, including Hawley retainers (n = 199), vacuum-formed retainers (n = 34), lingual fixed retainers (n = 278), and the COMBO (a combination of two different retainers in the same arch; n = 80). Patient's demographics, retention procedures, time to failure (survival time), and reasons for failures were extracted from patient files. Failure of retainer was defined as any events after which the retainers needed to be replaced or repaired; loss of retainer was treated as a failure because the retainer could not fulfil its role. Survival analyses were performed to compare the survival time between different retainer groups. The survival time was the longest for lingual fixed retainers (median 1604 days) and Hawley retainers (1529 days), followed by COMBO (258 days) and vacuum-formed retainers (105 days; overall P < 0.001). No statistical significance of survival time of lingual fixed retainers was found between maxilla (1497 days) and mandible (1604 days; P = 0.341), nor between different types of the COMBO (overall P = 0.078). These results were unchanged before and after adjusting for the age, gender, and socioeconomic status (SES) of the patients. The reasons for failures varied among different retainers: most failures of the Hawley retainers were 'lost' (52.0%), most failures of the lingual fixed retainers were debond (63.5%), and fracture was the most common cause of failure for both vacuum-formed retainers (43.5%) and the COMBO (41.9%). Lingual fixed retainers and Hawley retainers had the longest survival times, followed by the COMBO retainers and vacuum-formed retainers. The reasons of failure were mainly mechanical (debond and fracture) and patient-related (loss).