Dysphonia in very preterm children: prevalence, presentation, influencing factors and response to treatment

Introduction

Dysphonia is a potential long-term outcome of extreme prematurity and has been linked with female gender, multiple intubations, extremely low birth weight, birth at <27 weeks gestation, complicated intubation procedure and surgical ligation of patent ductus arteriosus (Chapter 2). Dysphonia in extremely preterm children may be persistent (Chapter 3). The aforementioned risk factors may also be experienced following very preterm birth, yet systematic investigations of voice outcomes in very preterm children are lacking.

Aims

This thesis presents the prevalence of dysphonia in very preterm children aged between 6 and 12 years old in Western Australia, with reference to a term-born comparison group recruited from the same community (Chapters 6 and 7). Demographic and medical data were abstracted from medical charts to identify factors correlated with adverse voice outcomes at school age. Perceptual and acoustic data from the clinical voice assessments were used to determine the external validity of an index score of dysphonia severity, the Acoustic Voice Quality Index, in childhood voice disorders. Laryngeal examinations were conducted on consenting children with voice problems of at least moderate severity to document the nature and extent of any laryngeal injury underlying disturbance to the vocal signal. Finally, a trial of behavioural voice therapy was conducted, to determine whether non-invasive vocal exercises would have any effect on voice quality in very preterm children.

Results

The prevalence of dysphonia in this very preterm cohort was 61.2%, a higher prevalence than in the term-born reference group, at 30.5%. Perceptual judgements of voice quality were supported by the use of an acoustic evaluation of disturbance to the vocal signal (Chapter 11). Laryngeal examinations demonstrated that very preterm children present with laryngeal damage affecting the structure and function of the larynx during phonation, of varying degrees of severity (Chapter 8). The most common pathologies were incomplete glottic closure resulting from posterior chink and vocal fold atrophy and immobility. Each child presented with tightening of the supraglottic musculature during phonation, resulting in a strained vocal quality. Some preterm children experienced acceptable improvements in voice quality following behavioural intervention; however, most did not (Chapters 9 and 10).

Conclusions

Mild voice disorders, resulting from inefficient use of the vocal mechanism, are common in childhood, as demonstrated by the prevalence identified in this term-born cohort. However, the incidence of voice problems in very preterm children is higher than would be expected from voice overuse alone. The element of strain present in the voices of very preterm children, is hypothesised to be a maladaptive attempting to compensate for glottic incompetence. It is further hypothesised that those children who experienced improvements in voice quality had succeeded in reducing the supraglottic hyperfunctional component to their phonation. However, resolution of voice to a perceptually normal quality was not achieved, in part because of the likely persistence of structural laryngeal pathology. Implications of these findings in the context of the wider literature are discussed. Clinical recommendations arising from these findings are also presented (Chapter 12).