Konijnenberg, M. (Marleen) (2013) Effect van korte training op inter- en intraobserver overeenkomst bij de beoordeling van benauwde kinderen. thesis, Medicine.
![[img]](https://umcg.studenttheses.ub.rug.nl/style/images/fileicons/text.png) |
Text
KonijnenbergM.pdf Restricted to Registered users only Download (867kB) |
Abstract
Background: In children dyspnoea is a major cause for visits to the emergency department or hospitalization. Lung function examination often is not feasible in acute settings in young children, therefore we rely on clinical findings in clinical decision making concerning severity of disease and therapeutic effect. Earlier we found large observer variation in the clinical assessment of dyspnoeic children, obscuring the detection of possible clinically relevant changes in dyspnoea. We hypothesize that training of observers might improve the rater agreement. Goal: To evaluate the effect of short training of the observers on inter- and intrarater agreement. Methods: We used video recordings of 30 dyspnoeic children. Five observers (pediatricians) participated and received 30 minute training, where the observers assessed and discussed on a number of video recordings. The video recordings were assessed by five observers (pediatricians) twice before the training and twice afterwards, with intervals of at least one week. The following criteria were assessed: respiratory rate, heart rate, oxygen saturation, wheezing, prolonged expiratory phase, thorax movements, subcostal retractions, intercostal retractions, jugular retractions, supraclavicular retractions, nasal flaring, ability to speak, mental status and the degree of dyspnoea using a Likert-scale (0-10). For categorical variables (all but the degree of dyspnoea) we used percentage agreement and multirater kappa (>0,70) as measures of rater variation. For continuous variables ICC (>0,70) as well as the mean difference of the degree of dyspnoea was calculated for this purpose. The differences in inter- and intrarater agreement before and after training were considered the primary outcome. Results: Because of too many missing values, only the assessments of four observers were suitable for analysis (total of n=505 assessments). Intrarater multirater kappa was moderate to good (0,40-0,85) and percentage agreement 75-95%. Training did not significantly influence intrarater agreement (p >0,05). The interrater agreement was less, with kappa 0,10-0,40 and percentage agreement varying between 60-80%. Training did not have effect on interrater agreement either (p >0,05). Conclusion: Intrarater agreement is fair to good, but interrater agreement is moderate for the assessment of dyspnoeic children. A short training did not improve this observer variation. This does not exclude the possibility that other types of training or training of longer duration or higher intensity may improve this variation. Still at this moment clinical practitioners and researchers should account for the observer variation when using clinical assessment in dyspnoeic children and the use of more objective parameters is warranted.
| Item Type: | Thesis (Thesis) |
|---|---|
| Supervisor name: | Brand, prof. dr. P.L.P. |
| Supervisor name: | Bekhof, drs. J. and Isala Klinieken, Zwolle |
| Faculty: | Medical Sciences |
| Date Deposited: | 25 Jun 2020 10:44 |
| Last Modified: | 25 Jun 2020 10:44 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/546 |
Actions (login required)
 |
View Item |