Skip to main content

Table 4 Poisson goodness-of-fit tests

From: The impact of daily temperature on renal disease incidence: an ecological study

Disease

Statistical test

Daily maximum temperature (1980 df)

  

ED

Inpatient

Total renal disease

Deviance goodness-of-fit statistic

2216.946

2271.521

 

P-value

<0.001

<0.001

 

Pearson goodness-of-fit statistic

2199.826

2244.133

 

P-value

<0.001

<0.001

Urolithiasis

Deviance goodness-of-fit statistic

2290.703

2325.260

 

P-value

<0.001

<0.001

 

Pearson goodness-of-fit statistic

2150.742

2071.219

 

P-value

0.004

0.075

Renal failure

Deviance goodness-of-fit statistic

2470.910

2387.224

 

P-value

<0.001

<0.001

 

Pearson goodness-of-fit statistic

2241.799

2133.986

 

P-value

<0.001

0.008

AKI

Deviance goodness-of-fit statistic

2438.874

2407.545

 

P-value

<0.001

<0.001

 

Pearson goodness-of-fit statistic

2294.101

2171.447

 

P-value

<0.001

0.002

CKD

Deviance goodness-of-fit statistic

1934.498

2279.640

 

P-value

0.764

<0.001

 

Pearson goodness-of-fit statistic

1906.882

2069.318

 

P-value

0.878

0.079

UTI

Deviance goodness-of-fit statistic

2121.826

2131.721

 

P-value

0.014

0.009

 

Pearson goodness-of-fit statistic

2102.839

2079.242

 

P-value

0.027

0.059

Lower UTI

Deviance goodness-of-fit statistic

2108.504

2168.087

 

P-value

0.022

0.002

 

Pearson goodness-of-fit statistic

2083.343

2103.063

 

P-value

0.052

0.027

Pyelonephritis

Deviance goodness-of-fit statistic

2456.306

2257.211

 

P-value

<0.001

<0.001

 

Pearson goodness-of-fit statistic

2159.492

1949.885

 

P-value

0.003

0.681

  1. Poisson regression was used to estimate incidence rate ratios (IRRs) for daily emergency department (ED) and inpatient admissions for renal diseases in relation to an increase in daily maximum temperature per 1°C during the warm season (October – March) in Adelaide from 1 July 2003 to 31 March 2014. Deviance and Pearson goodness-of-fit tests were then applied to assess whether a Poisson model fits the data
  2. The deviance and Pearson goodness-of-fit test show how well a Poisson model fits the data. Each test yields a test statistic, comparing the observed count to the expected count under a Poisson model. The test statistics and degrees of freedom (df) are used to generate P-values. A P-value of <0.05 indicates a poor fit of the Poisson model – most of the results indicated a poor fit