Skip to main content

Table 5 Multivariate Cox regression models for the prediction of poor prognosis

From: Role of crescents for lupus nephritis in clinical, pathological and prognosis: a single-center retrospective cohort study

Factors

95% confidence interval

P value

Proteinuria (g/24-h)

0.95–1.25

0.233

Positive anti-dsDNA

1.36–4.37

0.003

CI

1.07–1.33

0.001

Crescent

 Non-crescent

refer

 

 Crescent

1.72–10.68

0.002

Interaction item

 Non-crescent: proteinuria

refer

 

 Crescent: proteinuria

0.63–0.95

0.015

  1. Whole variables were included for constructing a full multivariate Cox regression model, and an optimal multivariate Cox regression model was selected using AIC with forward–backward steps. Because the model had interaction terms, the actual effect was not directly calculated. Only the coefficients and standard errors are displayed
  2. MBP mean blood pressure, dsDNA double-stranded DNA, SLEDAI systemic lupus erythematosus disease activity index, CI chronicity NIH index
Back to article page

European Journal of Medical Research

ISSN: 2047-783X

Contact us