The Reliability of Photograph-Based Digital Measurements for Assessing the Pediatric Elbow Range of Motion—A Pilot Study

TL;DR

This study shows that measuring elbow motion via photos is reliable and could replace in-person visits for some pediatric patients.

Contribution

The study introduces and validates a photograph-based digital method for measuring elbow motion in children as a telemedicine alternative.

Findings

Photograph-based measurements overestimated extension by 1.2° and flexion by 5.5°, but underestimated the carrying angle by 3.5°.

Intra-rater and inter-rater reliability for digital measurements were in the good to excellent range (ICC ≥ 0.87).

Digital measurements are a viable alternative to in-office goniometer assessments for pediatric elbow motion evaluation.

Abstract

What are the main findings?
When comparing photograph-based digital measurements to in-office goniometer measurements, accuracy bias (the average difference between observer measurements and the goniometer) showed an overestimation of photographs for extension by 1.2°, an overestimation of photographs for flexion by 5.5°, and an underestimation of photographs for the carrying angle of 3.5°.Intra- and inter-rater reliability correlations were all within the good (0.75–0.9) to excellent (>0.9) ranges. The intra-rater intraclass correlation coefficient (ICC) was 0.91 for extension, 0.87 for flexion, and 0.87 for the carrying angle. The inter-rater ICC was 0.98 for extension, 0.93 for flexion, and 0.96 for the carrying angle.

When comparing photograph-based digital measurements to in-office goniometer measurements, accuracy bias (the average difference between observer measurements and the goniometer) showed an overestimation of photographs for extension by 1.2°, an overestimation of photographs for flexion by 5.5°, and an underestimation of photographs for the carrying angle of 3.5°.

Intra- and inter-rater reliability correlations were all within the good (0.75–0.9) to excellent (>0.9) ranges. The intra-rater intraclass correlation coefficient (ICC) was 0.91 for extension, 0.87 for flexion, and 0.87 for the carrying angle. The inter-rater ICC was 0.98 for extension, 0.93 for flexion, and 0.96 for the carrying angle.

What are the implications of the main findings?
Photograph-based digital measurements appear to be an acceptable alternative compared to in-office goniometer measurements in select pediatric patients.The incorporation of this technique in telemedicine evaluations could potentially enhance access to care, reduce the patient and caregiver burden of in-person visits and offer a new avenue to safely streamline upper extremity fracture care in pediatric patient populations.

Photograph-based digital measurements appear to be an acceptable alternative compared to in-office goniometer measurements in select pediatric patients.

The incorporation of this technique in telemedicine evaluations could potentially enhance access to care, reduce the patient and caregiver burden of in-person visits and offer a new avenue to safely streamline upper extremity fracture care in pediatric patient populations.

Background/Objectives: The follow-up for pediatric elbow injuries often involves in-office visits to assess the range of motion (ROM). There is evidence for the use of telemedicine and digital assessments of elbow motion in adults but none in pediatrics. The aim of this study was to compare measurements of the joint range of motion and elbow alignment in children, taken in a doctor’s office, with those obtained from photographs and to evaluate the intra- and inter-observer reliability of these measurements. Methods: Fourteen children were prospectively included in this study during follow-ups after an arm fracture. Only the uninjured arm was evaluated: goniometric measurements performed in the office were compared with digital measurements obtained from photographs. For the uninjured arm, in-office goniometer measurements were compared to picture-based digital measurements. Mean digital values were compared to in-office goniometer measurements using a paired t-test, and the intra- and inter-rater reliability of the digital measurements was calculated. Results: When comparing photograph-based digital measurements to in-office goniometer measurements, the mean absolute error (average absolute difference between observer measurements and goniometer) showed differences of 4.5° for extension, 7.4° for flexion, and 4.8° for the carrying angle. The accuracy bias (average difference between observer measurements and goniometer) showed an overestimation of photographs for extension by 1.2°, an overestimation of photographs for flexion by 5.5°, and an underestimation of photographs for a carrying angle of 3.5°. Intra- and inter-rater reliability correlations were all within the good (0.75–0.9) to excellent (>0.9) ranges. The intra-rater intraclass correlation coefficient (ICC) was 0.91 for extension, 0.87 for flexion, and 0.87 for the carrying angle. The inter-rater ICC was 0.98 for extension, 0.93 for flexion, and 0.96 for the carrying angle. Conclusions: Photograph-based digital measurements appear to be an acceptable alternative compared to in-office goniometer measurements in pediatric patients.

Figures