Skip to main content

Table 5 Unadjusted and adjusted multilevel logistic regression analysis of health centre and client level factors on composite indicator of recommended STI testing and counselling services (n = 137 health centres; 15,212 patient records)

From: Wide variation in sexually transmitted infection testing and counselling at Aboriginal primary health care centres in Australia: analysis of longitudinal continuous quality improvement data

Outcome is 100% service delivery

Unadjusted

Model A

Model B

Model C

Predictors

Fixed effects

OR

95% CI

OR

95% CI

OR

95% CI

OR

95% CI

Audit Year

2007

  

1.00

(reference)

1.00

(reference)

1.00

(reference)

2008

  

0.65

(0.46-0.91)

0.79

(0.55-1.12)

0.82

(0.57-1.19)

2009

  

0.65

(0.47-0.92)

0.73

(0.51-1.04)

0.74

(0.51-1.06)

2010

  

0.44

(0.31-0.63)b

0.50

(0.34-0.73)b

0.49

(0.33-0.73)b

2011

  

0.34

(0.24-0.47)b

0.43

(0.30-0.63)b

0.43

(0.29-0.63)b

2012

  

0.51

(0.36-0.72)b

0.56

(0.38-0.83)a

0.57

(0.38-0.85)

2013

  

0.61

(0.43-0.86)

0.71

(0.48-1.05)

0.71

(0.48-1.06)

2014

  

0.57

(0.40-0.82)a

0.66

(0.43-1.00)

0.62

(0.40-0.94)

Health Centre Characteristics

Location

Urban

1.00

(reference)

  

1.00

(reference)

1.00

(reference)

Regional

0.25

(0.12-0.56)a

  

0.41

(0.19-0.86)

0.38

(0.18-0.80)

Remote

1.71

(0.88-3.34)

  

2.09

(1.12-3.90)

1.98

(1.05-3.74)

Governance

Community-controlled

1.00

(reference)

      

Government operated

0.78

(0.49-1.23)

      

Service population

≥1000

1.00

(reference)

      

>500- < 1000

1.30

(0.78-2.17)

      

≤500

1.31

(0.88-1.96)

      

Length of accreditationa

Never accredited

1.00

(reference)

      

Accredited for some of the time

1.45

(0.94-2.24)

      

Accredited for all of the time

0.73

(0.49-1.06)

      

Duration of participation in ABCD CQI project

Baseline

1.00

(reference)

  

1.00

(reference)

1.00

(reference)

1-2 cycles

1.49

(1.35-1.64)b

  

1.30

(1.12-1.50)b

1.34

(1.16-1.55)b

≥3 cycles

1.11

(0.98-1.24)

  

0.97

(0.78-1.20)

1.01

(0.81-1.26)

Provided adult health checkb

No

1.00

(reference)

  

1.00

(reference)

1.00

(reference)

Yes

3.17

(2.88-3.50)b

  

3.25

(2.94-3.59)b

3.40

(3.07-3.77)b

Patient Characteristics

Sex

Male

1.00

(reference)

    

1.00

(reference)

Female

1.41

(1.30-1.52)b

    

1.45

(1.33-1.57)b

Age (years)

15-19

1.82

(1.48-2.22)b

    

2.10

(1.67-2.63)b

20-24

3.21

(2.63-3.92)b

    

3.84

(3.07-4.80)b

25-29

2.82

(2.30-3.47)b

    

3.01

(2.39-3.79)b

30-24

2.15

(1.74-2.65)b

    

2.35

(1.85-2.97)b

35-39

1.72

(1.38-2.13)b

    

1.77

(1.39-2.26)b

40-44

1.35

(1.08-1.67)

    

1.36

(1.07-1.73)

45-49

1.17

(0.93-1.47)

    

1.27

(0.99-1.63)

50-54

1.00

(reference)

    

1.00

(reference)

Aboriginal Status

Non-Aboriginal

1.00

(reference)

    

1.00

(reference)

Aboriginal

2.10

(1.70-2.61)b

    

1.46

(1.15-1.84)a

Not recorded

1.55

(1.13-2.12)

    

1.09

(0.77-1.55)

Random effects (intercepts)

  State (variance (SE))

  

1.05

(0.71)

0.54

(0.39)

0.55

(0.395)

  MORSTATE

  

2.66

 

2.02

 

2.03

 

   PCV (% explained variance)

    

48%

 

47%

 

  Health Centre (variance (SE))

  

0.84

(0.14)

0.54

(0.09)

0.55

(0.095)

  MORHC

  

2.39

 

2.01

 

2.03

 

   PCV (% explained variance)

    

36%

 

34%

 

  State & Health Centre (variance)

  

1.89

 

1.08

 

1.11

 

  MORSTATE-HC

  

3.70

 

2.69

 

2.73

 

   PCV (% explained variance)

    

43%

 

41%

 

  Patient (variance (SE))

  

0.28

(0.14)

0.12

(0.055)

0.03

(0.015)

   PCV (% explained variance)

    

59%

 

89%

 
  1. CQI continuous quality improvement, MOR Median Odds Ratio: odds of receiving above median service delivery if client was to change health centre or jurisdiction, OR Odds Ratio, PCV proportional change in variance: per cent variation explained in odds for better healthcare delivery by introduction of health centre or client level factors, SRH Sexual and Reproductive Health
  2. Due to multiple testing, the significance level is set at p < 0.005 (Bonferroni correction); a p < 0.005; b p < 0.0001
  3. a(n = 13,989 with accreditation status recorded); b(n = 13,235 with AHC recorded, introduced to audit tool in 2007)