Skip to main content

Table 3 Results for the independent variable ‘private healthcare provision’ in Cox proportional hazard model analyses of various time periods between symptom onset, healthcare presentation, investigations and treatment commencement among adult Victorian TB adult patients, 2002–2015

From: Comparing tuberculosis management under public and private healthcare providers: Victoria, Australia, 20022015

Time period outcome

Private sample

Private healthcare provision

Public healthcare provision

Results of Cox regression analysis with inclusion of private healthcare as binary exposure variable

Multivariate survival analysisa

Median (IQR) (days)

Median (IQR) (days)

Number of obs

HR

(95%CI)

P value

Other independent variables

Patients with pulmonary involvement

 Health system delay cx sp

59

40 (17–90)

22 (6–52)

1399

0.81

(0.58–1.12)

0.195

a sx pe p h o r

sp.

 Diagnostic delay - presentation to first chest x-ray/CT scan

35

29 (5–59)

9 (0–32)

1216

0.84

(0.60–1.18)

0.307

sx pe r p h o y

 Treatment initiation delay 1 sp

83

19 (7–64)

10 (3–34)

1409

1.35

(0.82–2.21)

0.235

a p s c h o y sp. h#p a#o a#sp. (this interaction term contained no smear positive in the youngest age-group)

Extrapulmonary patients

 Health system delay cx

143

61 (25–99)

46 (20–91)

1504

NA

sx e f‡ c‡ y‡ cx cx#m‡

 Diagnostic delay - presentation to first chest x-ray/CT scan

81

54 (19–93)

27 (6–65)

1080

0.79

(0.63–0.99)

0.043

s‡ sx p‡ y‡

All patients

 Patient delayb

39

1 (0–28)

18 (0–68)

871

1.36

(1.02–2.00)

0.065

sx pe ly e d p c o

 Laboratory delay - specimen collection to culture resultc

53

43 (40–53)

43 (35–52)

1380

0.90

(0.68–1.20)

0.474

sx pe ly‡ e‡ d‡ r s‡ p c o c#sx‡ sx#o‡

 Treatment initiation delay 2

60

10 (5–25)

11 (3–21)

577

NA

a sx y‡ sm‡

  1. Definition of abbreviations: TB tuberculosis, NA not applicable, private p > 0.25 in univariate analysis, obs observations, IQR interquartile range, HR hazard ratio, # interaction operator
  2. aVariables considered in univariate analysis – age (a) (five groups: 0–9 years; 10–17 years; 18–34 years; 35–64 years, ≥65 years); sex (sx); manifestation (m) (categories: pulmonary only, pulmonary plus other sites [pe], lymph node [ly], disseminated [d] or other extrapulmonary [e]); saw rural health provider (r); saw private health provider (p); has a history of substance abuse (s); has ever resided in an aged care facility (f); household member or close contact with TB (c); born in high burden country (h); overseas born (o); year of notification (y) (groups: 2002–2005; 2006–2011; 2012–2015);. First CXR results (cx) and sputum smear results (sp) and smear results (sm) were also considered where noted
  3. bPatient delay could only be considered from 2012 to 2015 due to small sample numbers in prior years.
  4. cLaboratory delay could only be considered from 2011 to 2015 due to year of notification variable not being proportional
  5. Independent variables included in multivariate analyses with p<0.05