Skip to main content
. 2010 Apr 23;10:23. doi: 10.1186/1472-6947-10-23

Table 4.

Multilevel logistic regression analysis of frequency of ICD coded visits

Model A Model B
Fixed effects OR (95% CI)
Patient age group
1 (-28) - REF
2 (29-49) - 0.86 (0.84-0.89)
3 (50-67) - 0.84 (0.82-0.87)
4 (68-) - 0.75 (0.73-0.77)
Patient sex
Female - REF
Male - 0.98 (0.96-1.00)
Type of visit
Planned - REF
Not planned - 1.44 (1.41-1.47)
Random effects
HCC Variance (95% CI) 0.76 (0.40-1.54) 0.76 (0.41-1.50)
MOR (95% CI) 2.30 (1.82-3.26) 2.29 (1.84-3.22)
Physician Variance (95% CI) 2.28 (2.05-2.55) 2.25 (2.02-2.53)
MOR(95% CI) 4.22 (3.92-4.58) 4.19 (3.88-4.56)
HCC+Physician Variance 3.04 3.01
MOR 5.23 5.28
PCV
HCC - 0.3%
Physician - 1.3%
HCC+Physician - 0.9%
303170.55 301079.69
DIC (MCMC)

Outcome variable at the visit level, ICD coding (yes/no). Extracted from SPCD during 2002-2003 for the 24 Health Care Centres in Skaraborg Primary Care.

OR = odds ratio, CI = credible interval, MOR = median odds ratio, DIC = Deviance Information Criterion

HHS Vulnerability Disclosure