Evaluation of the
Reliability of Electronic Medical Record
Data in Identifying Comorbid Conditions
among Patients with Advanced Non-Small Cell Lung Cancer
Table 2
Measured reliability of algorithm (combination of programmatic queries and key word searches).
Electronic chart review group
True negatives (via chart review)
False negatives (via chart review)
Negative predictive value
95% confidence interval
patients identified as negative via algorithm
414
36
0.92
0.89–0.94
Subgroup analyses of patients negative for comorbidities via algorithm
Group 1: Negative for all comorbidities ()
134
16
0.89
0.83–0.94
Group 2: Negative for diabetes ()
41
9
0.82
0.69–0.91
Group 3: Negative for cardiovascular disease ()
47
3
0.94
0.83–0.99
Group 4: Negative for cerebrovascular disease ()
50
0
1.00
0.93–1.00
Group 5: Negative for moderate/severe renal disease ()
