Appendix Table 2.

Estimated Logistic Regression Coefficients (b) for the Patient-Reported Symptom Data

Score ModelSymptombPoints95% CI
A priori symptomsCough2.8561.41, 4.80
Subjective_fever1.9941.21, 2.82
Acute_onset−0.27−1−1.02, 0.45
Chills_sweats1.2420.26, 2.29
Myalgia−0.69−1−1.85, 0.44
LASSOChills_sweats1.0720.13, 2.07
Cough2.7451.34, 4.66
Subjective_fever1.8141.09, 2.56
Ebell flu score symptomsAcute_onset−0.39−1−1.13, 0.31
Myalgia−0.70−1−1.85, 0.43
Chills_sweats1.1620.20, 2.21
Cough:subjective_fever2.1941.46, 2.99
CF (unweighted)Cough:subjective_fever2.1721.50, 2.88
CFA (unweighted)Cough:subjective_fever:acute_onset1.2310.55, 1.96
CFM (unweighted)Cough:subjective_fever:myalgia1.9321.28, 2.61
LASSO variables (unweighted)Chills_sweats:cough:subjective_fever2.0521.39, 2.75
CF (weighted)Cough2.6351.24, 4.55
Subjective_fever2.0441.35, 2.77
CFA (weighted)Cough2.6151.21, 4.53
Subjective_fever2.1041.39, 2.87
Acute_onset−0.27−1−1.00, 0.44
CFM (weighted)Cough2.6851.27, 4.60
Subjective_fever2.1141.38, 2.89
Myalgia−0.30−1−1.34, 0.75
  • Abbreviations: LASSO, Least Absolute Shrinkage and Selection Operator; CF, presence of cough and fever; CFM, presence of cough, fever, and myalgia; CFA, presence of cough and fever with acute onset of disease.

  • Notes: All models were fit only to the derivation set. Confidence intervals for the coefficients were calculated using the Wald Method.