Brodsky scale most reliable for tonsil size evaluation in kids

By Rob Goodier

NEW YORK (Reuters Health) - The Brodsky scale for evaluating the size of children's tonsils may give the most consistent results compared to two other tools - the Friedman scale and a new modified three-grade scale, a new study has found.

The Brodsky gave the most reliable measurements, both between observers and between different measurements by the same observer, according to the research.

"We would recommend discontinuing the usage of other grading scales and recommend routinely using the Brodsky scale for both clinical care and research," Dr. Neil K. Chadha of British Columbia Children's Hospital and the University of British Columbia in Vancouver, Canada told Reuters Health by email.

"It is important to reflect on the reliability and validity of any scale or scoring system used in medicine - just because something has been used for many years doesn't necessarily mean it is reliable, or cannot be improved. In our case this well-established scale (Brodsky) was found to be reliable, but it is notable that research has been published relying on other scales which may not have been as reliable as assumed."

As reported online October 9 in JAMA Otolaryngology-Head & Neck Surgery, Chadha and his team conducted a cross-sectional study at an outpatient clinic in the British Columbia Children's Hospital system in which four observers independently assessed the size of the left tonsil of 116 children ages 3 to 14 years old.

The observers graded each child twice at two different times using each of the three scales.

They assessed the reliability between raters using intraclass correlation coefficients and Cronbach's alpha (a coefficient of internal consistency). The Brodsky scale had a mean ICC of 0.721 and Cronbach's alpha of 0.911, while the Friedman scale had an ICC of 0.647 and Cronbach's alpha 0.879. The modified three-grade scale had an ICC of 0.599 and a Cronbach's alpha of 0.927.

None of the scales reached an acceptable level of reliability, which the researchers had defined as an ICC of 0.75. The Brodsky, however, came close.

The researchers also measured the reliability between the two successive evaluations of each observer, the intra-rater reliability, using the Pearson correlation coefficient. The Brodsky had a mean reliability of 0.954 compared to Friedman's 0.932 and the three-grade scale's 0.927.

The Brodsky four-point scale and the Friedman five-point scale are the two most widely used. Brodsky assesses the percentage of the airway that the tonsil takes up and Friedman grades the tonsil according to its position relative to surrounding structures.

The third scale included in the comparison is a modified three-grade scale that, like the Friedman, evaluates the relative position of the tonsils. This scale is unique, developed in house with the assumption that fewer grades might improve its reliability and the chance that different observers give similar grades.

The three-grade scale performed unexpectedly poorly, the researchers noted in their paper. That may be because the observers are less familiar with it compared to the other two scales or because it may be more difficult to assess size in thirds, the researchers write in their paper.

Grading the tonsil size to check for enlargement is important in part because enlarged tonsils can reduce airflow to and from the lungs and they are a risk factor in obstructive sleep apnea in children, the researchers write.

SOURCE: http://bit.ly/1veND3e

JAMA Otolaryngol Head Neck Surg 2014.

(c) Copyright Thomson Reuters 2014. Click For Restrictions - http://about.reuters.com/fulllegal.asp