Skip to main content

Table 1 Univariate logistic regression of the association between clinician caseload and treatment outcome

From: Towards better guidance on caseload thresholds to promote positive tuberculosis treatment outcomes: a cohort study

Main exposure/potential confounders/a priori confounders

Overall cohort

Missing 12-month outcome

Unfavourable 12-month outcome

Univariate regressiona

No.

Col. %

No.

Row %

No.

Row %

OR (95 % CI)

Overall

48,838

100.0

1,468

3.0

6,777

13.9

Clinician caseloadb

       
 

10+

24,092

49.3

401

1.7

3,057

12.7

P <0.001

 

<10

16,803

34.4

767

4.6

2,518

15.0

1.20 (1.10–1.30)

 

Missing

7,943

16.3

300

3.8

1,202

15.1

Notification date

       
 

Post-toolkit

39,264

80.4

920

2.3

5,283

13.5

P <0.001

 

Pre-toolkit

9,574

19.6

548

5.7

1,494

15.6

1.28 (1.19–1.38)

 

Missing

0

0.0

0

0

Location

       
 

Outside London

28,212

57.8

1,338

4.7

4,069

14.4

P <0.001

 

Inside London

20,618

42.2

130

0.6

2,704

13.1

0.78 (0.70–0.86)

 

Missing

8

0.0

0

0.0

4

50.0

Sex

       
 

Male

26,910

55.1

829

3.1

4,051

15.1

P <0.001

 

Female

21,777

44.6

630

2.9

2,700

12.4

0.80 (0.76–0.85)

 

Missing

151

0.3

9

6.0

26

17.2

Age, years

       
 

<20

4,987

10.2

128

2.6

477

9.6

0.66 (0.58-0.75)

 

20–39

23,987

49.1

677

2.8

3,334

13.9

P <0.001

 

40–64

13,534

27.7

400

3.0

1,749

12.9

0.91 (0.85–0.97)

 

65+

6,325

13.0

263

4.2

1,215

19.2

1.42 (1.31–1.54)

 

Missing

5

0.0

0

0.0

2

40.0

Ethnic group

       
 

White

9,313

19.1

401

4.3

1,580

17.0

P <0.001

 

Black African

9,666

19.8

243

2.5

1,176

12.2

0.67 (0.61–0.74)

 

Black other

1,473

3.0

20

1.4

213

14.5

0.85 (0.71–1.01)

 

Indian subcontinent

21,082

43.2

486

2.3

2,753

13.1

0.74 (0.68–0.80)

 

Other

5,662

11.6

131

2.3

746

13.2

0.74 (0.67–0.83)

 

Missing

1,642

3.4

187

11.4

309

18.8

UK born

       
 

No

33,946

69.5

777

2.3

4,517

13.3

P <0.001

 

Yes

11,946

24.5

397

3.3

1,742

14.6

1.12 (1.05–1.20)

 

Missing

2,946

6.0

294

10.0

518

17.6

Site of diseasec

       
 

Extrapulmonary only

21,681

44.4

543

2.5

2,662

12.3

P <0.001

 

Pulmonary smear negative/unknown

17,483

35.8

578

3.3

2,441

14.0

1.18 (1.10–1.26)

 

Pulmonary smear positive

9,399

19.2

313

3.3

1,623

17.3

1.53 (1.41–1.65)

 

Missing

275

0.6

34

12.4

51

18.5

Drug sensitivity

       
 

Sensitive

26,955

55.2

782

2.9

3,607

13.4

P <0.001

 

Resistant to INH, EMB, or PZA

1,744

3.6

53

3.0

756

43.3

5.63 (5.01–6.32)

 

Not linked to a culture

19,865

40.7

627

3.2

2,364

11.9

0.86 (0.80–0.91)

 

Missing

274

0.6

6

2.2

50

18.2

Previous diagnosis

       
 

No

39,767

81.4

1,006

2.5

5,227

13.1

P <0.001

 

Yes

2,941

6.0

96

3.3

494

16.8

1.35 (1.20–1.51)

 

Missing

6,130

12.6

366

6.0

1,056

17.2

Social risk factorsd

       
 

No or unknown

47,048

96.3

1,394

3.0

6,352

13.5

P <0.001

 

One or more previous

829

1.7

19

2.3

144

17.4

1.42 (1.17–1.73)

 

One or more current

961

2.0

55

5.7

281

29.2

2.75 (2.34–3.23)

Shared clinical management

       
 

No or unknown

48,562

99.4

1,457

3.0

6,721

13.8

P = 0.02

 

Yes

276

0.6

11

4.0

56

20.3

1.45 (1.06–1.98)

  1. Descriptive analysis and univariate random effects logistic regression of the association between clinician caseload and treatment outcome, England, 2006–2012. All regression models adjusted for clustering by clinician
  2. aOdds of an unfavourable versus a good or neutral treatment outcome; records missing an outcome excluded
  3. bMean caseload per clinician over the preceding 3 years
  4. cPulmonary site could be with or without extrapulmonary disease
  5. dSocial risk factors a composite variable of homelessness, imprisonment, drug misuse and alcohol abuse; current risks override previous risks
  6. CI Confidence interval, Col. Column, EMB Ethambutol, INH Isoniazid, OR Cluster-specific odds ratio, P, P value, PZA Pyrazinamide