Skip to main content

Table 1 Learning variables for training the ANN were selected from this group on the basis of potential HCV-specific indicators such as clues of extrahepatic disease manifestations [10]

From: Neural-network analysis of socio-medical data to identify predictors of undiagnosed hepatitis C virus infections in Germany (DETECT)

Variables

Age (learn only, no best match)

Gender (learn only, no best match)

Relative frequency of HCV-infections within 2-digit ZIP codes

Frequency of drug prescriptions by ATC for

 Levothyroxine (n = 5450)

 Ibuprofen (n = 3534)

 Metamizol (n = 2790)

 Diclofenac (n = 2324)

 Diclofenac topical (n = 2112)

 Zopiclon (n = 1854)

 Prednisolone (n = 1798)

 Metformin (n = 1517)

Frequency of four common drug prescriptions by PZN for

 Cholecalciferol (n = 473)

 Decristol 20.000 iE (n = 473)

 Lyrica 75 mg (n = 473)

 Pregabalin (n = 369)

Frequency of out-patient diagnoses

 E04.9 nontoxic goitre, unspecified (n = 1123)

 M81.9 osteoporosis, unspecified (n = 1122)

 F32.9 major depressive disorder, single episode (n = 1073)

 L30.9 dermatitis, unspecified (n = 1039)

 M54.5 low back pain (n = 1016)

 E14 diabetes mellitus, unspecified (n = 948)

 R53 malaise and fatigue (n = 743)

 E1190 diabetes mellitus, not primary insulin dependent (n = 728)

One HCV uncorrelated but in the HCV group statistically heaped diagnosis

 H52.2 astigmatism

Average length of hospital stay

  1. Of note, no variables directly linked to the diagnosis of HCV (e.g. diagnosis of hepatitis C, drug treatment with interferon or direct antiviral agents) were chosen