Smartphone video clip analysis yields ‘high’ clinician agreement on OSA scores in children

Key takeaways:

• Monash video score agreement between clinicians varied based on whether the child was healthy or had Down syndrome, autism or obesity.
• Future studies will evaluate video scores vs. polysomnography.

BOSTON — Clinicians reviewing smartphone video clips of children sleeping often agreed on Monash Obstructive Sleep Apnea Video Scores, according to data presented at the CHEST Annual Meeting.

“This work represents the first step to evaluate the use of parent-recorded video clips for evaluation of pediatric OSA,” Sherri L. Katz, MD, FCCP, division chief of pediatric respirology at Children’s Hospital of Eastern Ontario (CHEO) and senior scientist at the CHEO Research Institute, told Healio. “Now that we know that scoring can be reliably done by clinicians, the next step will be to determine whether video clips predict the presence of OSA on diagnostic tests including polysomnography. If, as we hope, video clips can identify children with moderate-severe OSA, they can be used to prioritize children for definitive diagnostic testing, shortening wait times for diagnosis and treatment, particularly in resource-limited settings.

“The use of video clips may ultimately change the paradigm for evaluation of OSA in children and may avoid morbidities associated with its delayed diagnosis,” Katz continued.

Katz and colleagues assessed 66 healthy children, 21 children with Down syndrome, 13 children with autism and nine children with obesity — all aged 3 to 18 years with suspected OSA — to find out if three smartphone video clips taken by parents while their child slept can be used to help detect OSA in each of these groups.

As Healio previously reported, video clips of children could assist in screening for OSA, as this method was nearly as effective in detecting OSA compared with other tools in two studies.

Two clinicians each reviewed the short videos for signs of OSA through the Monash Obstructive Sleep Apnea Video Score, in which moderate-severe OSA was signaled with scores of 3 or higher. Researchers than estimated inter-rater agreement via weighted Cohen’s kappa.

“The [Monash Obstructive Sleep Apnea] score is out of eight and based on the presence of inspiratory noises scored from one to four, four being the most severe, the presence of obstructive events, evidence of increased work of breathing, the presence of mouth breathing and neck extension,” Katz said during her presentation.

When considering the total study population, agreement between clinicians for the Monash score was 0.94.

“As expected, we found that video clips were of good quality for interpretation and were scored similarly by different raters,” Katz told Healio.

According to the presentation, healthy children had the highest inter-rater reliability of the four groups (0.96), followed by children with autism (0.85), Down syndrome (0.84) and obesity (0.76).

“Agreement between video clip scorers was greatest when scoring video clips of healthy children, compared to those with medical comorbidity, which was not surprising given the underlying pathophysiology of sleep disordered breathing in these populations,” Katz said. “Nonetheless, the video clips performed very well in children with Down syndrome, obesity and autism as well.”

Researchers further found that inter-rater reliability for the presence of moderate-severe OSA in the total cohort was 0.82.

“The Monash scoring system was excellent in discriminating presence or absence of OSA according to the clinicians’ gestalt impression,” Katz told Healio. “These findings are encouraging, as they mean that the scoring will be consistent between providers and is reliable.”

Unlike before, children with obesity had the highest inter-rater reliability (1), and healthy children had the lowest (0.79). Inter-rater reliability in children with autism (0.84) and children with Down syndrome (0.8) fell in between the above groups.

Future studies will evaluate video scores vs. polysomnography, Katz said.

“The work from this study has laid the foundation for larger scale studies comparing the diagnostic characteristics of video clips to those of polysomnography, the gold standard,” Katz told Healio. “A pilot study funded by the CHEST Foundation which compared video clips to polysomnography in healthy children has just been completed and results are being analyzed. Additional pilot studies in populations of children with Down syndrome, obesity and autism are underway.

“Our research team has also just been awarded over $1 million from the Canadian Institutes of Health Research to conduct a 5-year, multi-center study comparing video clips to polysomnography in children, which will more definitively determine the role of video clips in the diagnostic tool kit for OSA in childhood,” Katz added.

For more information:

Sherri L. Katz, MD, FCCP, can be reached at skatz@cheo.on.ca.