Skip to main content
. 2012 Apr;129(4):695–700. doi: 10.1542/peds.2011-2037

TABLE 3.

Interrater Reliability for Physical Examination Findings

Criterion No. of Patients Evaluated Characteristic Present per First Assessor, n/N (%) Patient for Whom at Least 1 Assessor Marked “Unsure”, n Missing Data, n Overall Agreement, % κ (95% CI)b κ (95% CI) “Unsure” Responses Recodedc
Abdominal tendernessa (none, mild, moderate, severe) 598 N/A 1 0 60.9 .33 (.26–.40) .33 (.26–.40)
Abdominal tendernessa (presence/absence) 598 593 (99.0) 1 0 98.7 .49 (.19–.80) .49 (.19–.80)
Right sided abdominal tenderness 583 562 (93.8) 15 1 91.1 .30 (.16–.44) .31 (.17–.44)
Maximal abdominal tenderness in RLQ 554d 447 (80.7) 29 0 80.0 .45 (.36–.54) .42 (.33–.51)
Rebound 548 224 (37.4) 51 0 68.4 .32 (.24–.40) .30 (.22 -.38)
Guarding 584 384 (64.1) 14 1 67.5 .30 (.23–.38) .30 (.22–.37)
Abdominal pain with walking, jumping, or coughing 494 418 (69.8) 103 2 83.6 .54 (.45–.63) .44 (.35–.53)
Right-sided abdominal pain with walking, jumping, or coughing 416e 311 (74.8) 76 2 77.6 .52 (.43–.60) .45 (.37–.53)
a

Abdominal tenderness variable analyzed as ordinal data (none, mild, moderate, or severe tenderness) as originally collected and as a dichotomous variable (presence or absence of tenderness).

b

“Unsure” responses excluded from analysis.

c

“Unsure” responses recoded as the presence of particular finding.

d

Excludes patients in whom clinician marked “unsure” for right-sided abdominal pain.

e

Excludes those patients in whom clinician marked “unsure” for abdominal pain with walking, jumping, or coughing.