Table 4. Median mesothelioma latency and adjusted time ratios for mesothelioma latency among British asbestos workers (1978–2005), by main industry sector.

Main industry sector during follow-up	No. of deaths	Person- years at risk	Median latency (years)	(95% CI)	Adj. time ratio	(95% CI)	LR test
Manufacturing							P=0.114
Textiles	16	246	34.0	(15.5–35.7)	1.07	(0.91–1.25)
Asbestos cement mixture, board and pipe	27	405	15.9	(6.6–34.6)	1.01	(0.95–1.08)
Asbestos/rubber/resin/bitumen mixtures	20	260	17.7	(4.8–32.2)	0.95	(0.88–1.02)
Asbestos board and paper	<5	NR	NR	NR	NR	NR
Garments	<5	NR	NR	NR	NR	NR
Insulation and plastering mixes	<5	NR	NR	NR	NR	NR
Maintenance	31	473	17.5	(10.5–36.4)	0.96	(0.91–1.02)
Missing	103	1672	26.1	(1.2–34.4)	0.98	(0.94–1.02)
Removal	200	2899	19.6	(13.9–26.8)	0.97	(0.94–1.06)
Other
Ship building, repair and breaking	23	329	8.2	(8.2–28.6)	1.05	(0.98–1.13)
Building and construction	6	71	25.5	(19.7–NE)	0.81	(0.72–0.90)
Miscellaneous	9	169	29.9	(15.4–35.8)	1.08	(0.97–1.20)
Missing	44	736	27.2	(6.2–35.4)	0.98	(0.93–1.03)
Insulation	129	1915	20.3	(2.2–31.3)	1.00	(ref)

Abbreviations: Adj. time ratio=time ratios adjusted for sex, main smoking status, year since first exposure, age at first exposure, duration of exposure, mesothelioma type, and presence of asbestosis estimated using a generalised gamma accelerated failure-time model; CI=confidence interval; LR test=likelihood-ratio test; NE=not estimable; NR=not reported since number of deaths was fewer than five; ref=reference category.