• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

Table 1Model parameters

EPIDEMIOLOGYBase caseLower limitUpper limitSourceComments
Prevalence of active TB in cohort (at time of screening)pTB0.5%0.1%5%Bell et al (in preparation)Systematic review of new entrant screening programmes found reported rates of active disease from 0.24–4.5 % (17 studies, n=343,519), with most under 1% (13/17 studies). For comparison, Newham service identified 8 cases from 1223 screened (0.6%) in 2003. The Dover Induction Centre (AH) and Heathrow Port of Entry (Hardie and Wilson1993, Callister et al 2002) schemes, found a lower rate of active TB (0.13% and 0.1%–0.2% respectively), hence the lower limit for sensitivity analysis.
Prevalence of TB infection in cohort (at time of screening)pInfect20%5%55%Ormerod 1990 and Bell et al (in preparation)Systematic review found four studies reporting TST+ results: 38%–55%. However, some of these will be due to prior BCG or environmental mycobacteria. Ormerod 1990 found Heaf grade 2– 4 in BCG unvaccinated new entrants aged under 30: 20–25% from Pakistan, 35% from India.
Incidence of TB (15 year) in currently uninfected peoplerNI0.4%0.2%0.6%Challenor and Ormerod 2002, and Colditz et al 1994Challenor and Ormerod found a rate of 14.5/100,000 in a cohort of TST-new entrants given BCG. If we assume a relative risk of 0.5 (0.3–0.7) for BCG (Colditz 1994), then this gives an estimated rate of TB for unprotected TST-new entrants of 30 (21–43) per 100,000, or 0.4% (0.3%–0.6%) for 15 year risk. As a more conservative lower limit, the general population incidence of 13/100,000 is used (this comprises 4/100,000 for UK born and 90/100,000 for foreign born).
Incidence of TB (15 year) in infected people (without active TB)rLI2%1%5%Marks et al 2000, Sutherland 1968 and Hart and Sutherland 1977Over 15 years follow-up of school BCG, 121 out of 2550 (4.7%) unvaccinated participants with TST conversion developed TB. Incidence is expected to be lower in new entrants, as they are less likely to have been recently infected. Marks et al estimated the risk of TB for a new entrant with TST reaction over 15mm and normal CXR of 0.9% over 5 years, and 6.7% for life (for 35 year old new entrant).
Proportion of new entrants aged under 35pYoung65%50%88%ONS 2002 (Population Trends, table 7.1)Estimated proportion of new entrants aged under 35 from migration data for England and Wales. Upper limit from Bothamley et al study, in which only 28 out of 235 new entrants screened were aged over 35.
Proportion of new entrants previously vaccinatedpPrior70%60%80%Dover audit, Ormerod 1990, Ormerod 1998Prevalence of prior BCG in audit at Dover Induction Centre was 80%.However, two studies (Ormerod 1990, and Ormerod 1998) reported giving BCG to 24% and 31% of cohorts, which must imply a maximum of 69–76% with prior BCG, though some of this group will have been TST+ without prior BCG.
Mean number of secondary cases per primary casenSec0.20.10.75Estimate from Birmingham data333 secondary cases detected from 2866 index cases (11.6%).Assuming 50% transmission rate. Upper limit from Saeed et al 2004 estimates, as in schools BCG model.
Proportion of secondary cases prevented by early detectionpEarly50%0%80%Assumption
Mean delay in incidence with latent infection (years)lagLI315Assumption
Mean delay in incidence without current latent infection (years)lagNI538Assumption
Mean delay in transmission from primary to secondary (years)lagSec315Assumption
EFFECTIVENESSBase caseLower limitUpper limitSourceComments
Relative risk of TB with chemoprophylaxis (in patients with latent infection)rrPro0.400.310.52Smieja et al 2004Relative risk with isoniazid treatment of 6 months or more (95% CI).
Relative risk of TB with BCG (in patients without latent infection)rrBCG0.490.340.70Colditz et al 1994Meta-analysis of 6 CCTs and 7 RCTs (relatively more conservative than results from just RCTs).
CONCORDANCE
Proportion of new entrants attending for symptom screenpScreen30%20%80%Newham 2003, Bell et al 2005, Hogan et al 2005Basecase taken from Newham PCT New Entrant Service (1223 out of 4123 seen). Very wide range of uptakes in published studies (Bell et al 2005): 16–98% in 17 new entrant screening programmes. UK studies (Bothamley 2002, Grenville-Mathers 1979, Ormerod 1990 and Callister 2002) reported uptake rates of: 16%, 36%, 83%, 75%. Survey by Hogan et al also showed wide variation in follow-up of cases notified by Heathrow Port of Entry scheme.
Proportion of referred patients attending hospital clinicpAssess80%50%90%Levesque 2004Proportion attending clinic for investigation following suspicious X-ray or symptom screen. This is assumed higher than the proportion attending screening in the first place. In a Canadian new entrant screening programme (Levesque et al 2004) 39/49 patients with a positive skin test attended for clinic assessment.
Proportion of TST results available (at first or second attempt)pTST95%70%100%Newham 2003Readings available for 1118 out of 1173 heaf tests performed (95%). Lower limit from Bothamley study: 54/181 TST (30%) not read.
Proportion of repeat TSTs requiredpTST210%0%20%Assumption
Proportion of those offered prophylaxis who start treatmentpPro94%50%100%Levesque 2004In a Canadian study, of 35 offered prophylaxis, 33/35 commenced treatment, 24 of whom completed the full 6 months therapy.
Proportion of previously unvaccinated TST- given BCGpBCG98%50%100%Assumption18 out of 18 Heaf grade 0 or 1 with no BCG scar in Bothamley study.
TEST ACCURACY
Sensitivity of screening questionnaire for detecting TBseXR95%50%100%Schwartzman & Menzies 2000, Kelly et al 2002Kelly et al 2002 report 96% of those diagnosed with TB had abnormal chest X-ray. Similar rates from other studies reported in Schwartzman & Menzies.
Specificity of screening questionnaire for detecting TBspXR98%50%100%Assumption
Sensitivity of screening questionnaire for detecting TBseSQ80%50%100%AssumptionKelly et al 2002 report low sensitivity and specificity for a symptom questionnaire compared with chest Xray (no numerical data given).
Specificity of screening questionnaire for detecting TBspSQ70%50%100%Assumption
Sensitivity of TST for latent TBseTST90%50%100%AssumptionAs in tests for latent infection model. Given these assumptions of prevalence, sensitivity and specificity, about 27% will be TST+, 33% of whom actually have latent infection
Specificity of TST for latent TBspTST80%50%100%Assumption
Sensitivity of IGT for latent TBseIGT90%50%100%AssumptionAs in tests for latent infection model. Given these assumptions of prevalence, sensitivity and specificity, about 18% will test positive, 50% of whom actually have latent infection
Specificity of IGT for latent TBspIGT90%50%100%Assumption
COSTSBase caseLower limitUpper limitSourceComments
Cost of administering questionnaire (£ per patient)cSQ£1£0.5£2PSSRU 2004Assumes 2 (1–4) min consult with primary care nurse (£27 per hour in clinic). Time to administer questionnaire in general practice reported as 45 seconds (Bothamley 2005)
Cost of Chest X-ray (£ per patient)cXR£17£11£18DH tariff 2004/5Range: interquartile range from 2003 reference costs
Cost for TST administration (£ per patient)cTST£7£3£15VariousAssumptions as in latent infection model
Cost for TST reading (£ per patient)cTSTread£7£3£29VariousAssumptions as in latent infection model
Cost for interferon gamma test (£ per patient)cIGT£22£14£33VariousAssumptions as in latent infection model
Cost for assessment after positive test (£ per patient)cAssess£250£200£300VariousAssumptions as in latent infection model
Cost of BCG (£ per patient)cBCG£4£1£11VariousAssumptions as in school model
Cost of chemoprophylaxis (£ per patient)cPro£500£250£850VariousAssumptions as in school model
Cost of treating an active case of TB (£ per patient)cTB£5,100£1,900£12,000VariousAssumptions as in school model (estimated cost per active case all ages weighted by age of incidence)
Discount rate for costsdrM0.035
HEALTH OUTCOMES
QALY loss from reactions to TSTqTST0.000030.000000.0005AssumptionAssumes mean quality of life loss of 0.1 for 9 weeks for 0.3% of people tested.
QALY loss from reactions to BCGqBCG0.000030.000000.0005AssumptionAssumed similar to TST.
Incidence of isoniazid-induced hepatitispHep0.26%0.1%0.5%IUAT trial, Thompson 1982Increased rate of hepatitis with isoniazid (vs placebo) from IUAT trial was 0.26% for six months treatment (0.52% for one year).
Mortality rate for hepatitismHep1%0.5%3%IUAT Thompson 1982, Shelley et al 1997Hepatitis fatality rate of 3.1% in the IUAT trial (3/95), but this sample included older patients and the trial did not include active monitoring of hepatitis, hence we might expect a lower mortality rate in our cohort. Shelley et al estimate incidence of fatal hepatitis at about 0.001% for people aged 35 based on a review of trials with hepatitis monitoring.
Incidence of other non-fatal adverse reactions to prophylaxispAR2%Snider et al 1986Snider et al report incidence of GI and other side effects from two previous studies (Thompson 1982 and Kopanoff et al 1978).
Mean QALY loss for non-fatal hepatitisqHep0.025AssumptionAssumes mean quality of life loss of 0.1 for 3 months per case.
Mean QALY loss for other adverse reactionsqAR0.008AssumptionAssumes mean quality of life loss of 0.1 for 1 month per case.
QALY loss due to prophylaxisqPro0.00080.00030.0033Calculated from above assumptionsEstimated QALY loss due to side effects of prophylaxis (fatal and non- fatal hepatitis and other side effects) discounted to current age for
QALY loss due to active TB, under 35qTByoung0.2210.1100.441VariousSame assumptions as in other models. Estimated QALY loss due to TB morbidity and mortality for under 35s discounted and weighted by
QALY loss due to active TB, over 35qTBold0.5810.2911.162VariousAs above, but estimated for over 35 year olds weighted by age distribution of UK migrants over 35
QALY loss due to active TB, any ageqTBsec0.6760.3381.352VariousAs above, but estimated for all ages weighted by UK incidence of TB.
Gain in QALYs due to early diagnosis of active diseaseqEarly0.020.000.05AssumptionSame assumptions as in other models.
Discount rate for health outcomesdrH0.035

From: Appendix K, Health Economic Models

Cover of Tuberculosis
Tuberculosis: Clinical Diagnosis and Management of Tuberculosis, and Measures for Its Prevention and Control.
NICE Clinical Guidelines, No. 117.
National Collaborating Centre for Chronic Conditions (UK); Centre for Clinical Practice at NICE (UK).
Copyright © 2006, Royal College of Physicians of London.

For 2006 original guideline text, no part of the content may be reproduced in any form (including photocopying or storing it in any medium by electronic means and whether or not transiently or incidentally to some other issue of this publication) without the written permission of the Royal College of Physicians of London. Applications for the Royal College of Physicians of London's written permission to reproduce any part of this publication should be addressed to the publisher.

For 2011 updated text, the material may be freely reproduced for educational and not-for-profit purposes. No reproduction by or for commercial organisations, or for commercial purposes, is allowed without the express written permission of NICE.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.