Skip to main content

Table 3 TB transmission control measures and tuberculin skin test (TST) conversion: Risk of bias assessment and results (n = 10)

From: Effectiveness of control measures to prevent occupational tuberculosis infection in health care workers: a systematic review

Studies Risk of bias Results Notes
First author, year. Similar baseline outcome measures Similar baseline characteristics Incomplete outcome data (attrition bias) Blinding or objective assessment of outcome Protected against contamination Selective outcome reporting (reporting bias) Selection and confounding bias Conversion rate per 100 person years or proportion converting (%) (B = before, A = after) Relative rate or proportion (A/B) B1, B2, A1, A2, etc., if more than two cycles.
Studies reporting rates of conversion
Bangsberg, 1997 [31] N/A N/A Unclear Low N/A Low High B1: 5.8 B1:1.00  
B2: 5.1 B2: 0.88
A1: 0.0 A1: 0.00
A2: 2.3 A2: 0.40
A3: 0.0 A3: 0.00
Louther, 1997 [32] N/A N/A Unclear Low N/A Low High B: 7.2 (65/898) B: 1.00 Fella et al. [20] describing same study, mention possible overestimate of conversions in before period.
A: 3.3 (32/971) A: 0.46
Blumberg, 1998 [34] N/A N/A Unclear Low N/A Low High B: 5.98 B: 1.00  
A: 1.09 A: 0.18
Yanai, 2003 [35] N/A N/A Unclear Low N/A Low Low   Univariate: Multivariate adjustment for sex, age, area TB risk, BCG scar, duration of work (≤ 12 vs > 12 months) and frequency of patient contact.
aDiscrepancy between rates and rate ratios in Table 2 of article – figures here are recalculations.
B: 9.3 B: 1.00
A1: 6.4 A1: 0.69a
A2: 2.2 A2: 0.24a
Multivariate:
A1: 0.40
A2: 0.01
da Costa, 2009 [15] N/A N/A High Low N/A Low High B: 4.8c B: 1.00 cConverted from person-months to person-years.
High loss to follow up reported.
A: 3.1 A: 0.65
Studies reporting only counts, i.e. proportions converting in each period
Blumberg,
1995 [27]
N/A N/A Unclear Low N/A Low High B: 118/3579 (3.3%) 1.00 Each period 6 months.
A1: 51/2975 (1.7%) 0.51
A2: 67/4715 (1.4% 0.42
A3: 30/4775 (0.6%) 0.18
A4: 23/5153 (0.4%) 0.12
Jarvis, 1995 [28] High Unclear Highd Low Low Low High Intervention groups:
Hospital “A”
  dAuthors reported insufficient TST data from control hospital.
After period unspecified.
B: 7/29 (24.1%) 1.00
A: 0/23 (0%) 0.00
Hospital “D”  
B: 15/123 (12.2%) 1.00
A: 5/150 (3.3%) 0.27
Control:
Hospital “B”
 
B: 2/22 (9.1%) 1.00
A: 6/33 (18.2%) 2.00
Maloney, 1995 [29] High Unclear Unclear Low Unclear Low High Intervention group:
Wards with TB patients
ePeriods differed in duration: 18 vs 14 months. Before proportion standardised to 14 months.
B: 15/90 (16.7%)
12e/90 (13.3%)
-
1.00
A: 4/78 (5.1%) 0.38
Controls:
Other wards
 
B: 7/254 (2.8%) 1.00
A: 9/228 (3.9%) 1.39
Wenger, 1995 [30] N/A N/A Unclear Low N/A Low High B: 7/25 (28%) 1.00 Periods differed in duration: 5 vs 9 and 16 months. Given zero count in A2, no period adjustment needed.
A1: 3/17 (17.6%) 0.63
A2: 0/23 (0%) 0.00
Behrman, 1998 [33] High Low Unclear Low Unclear Low Low Intervention group:
Emergency Department
  Each period 12 months.
Possible co-intervention via increased staff vigilance regarding early diagnosis and use of respirators.
B: 6/50 (12%) 1.00
A: 0/64 (0%) 0.00
Controls: Other hospital employees  
B: 51/2514 (2.0%) 1.00
A: 36/3000 (1.2%) 0.60
  1. N/A not applicable