Trend in risk of delay in diagnosis of new pulmonary tuberculosis in Northwest China from 2008 to 2017

BMC Infectious Diseases

Table 5 Median delay times stratified by age and adjusted >45-to- ≤ 45 HRs over ten years

Year	Age>45 years		Age ≤ 45 years		Adjusted >45: ≤45 (HR, 95%CI)	P
Year	N (%)	Median delay time (IQR)	N (%)	Median delay time (IQR)	Adjusted >45: ≤45 (HR, 95%CI)	P
2008	526	61 (35–104)	769	51 (31–89)	0.84 (0.75–0.94) ^*	0.002
2009	644	42 (20–98)	929	32 (14–66)	0.82 (0.74–0.90) ^*	<0.001
2010	583	35 (18–72)	865	34 (16–66)	0.92 (0.83–1.02) ^*	0.128
2011	570	30 (15–63)	911	28 (14–56)	0.76 (0.68–0.85) ^*	<0.001
2012	511	33 (15–65)	783	28 (12–54)	0.74 (0.66–0.83) ^*	<0.001
2013	511	30 (16–61)	748	29 (13–55)	0.82 (0.73–0.93) ^*	0.001
2014	524	31 (14–63)	726	32 (15–57)	0.93 (0.82–1.04) ^*	0.195
2015	390	32 (16–63)	556	31 (13–57)	0.92 (0.80–1.05) ^*	0.198
2016	485	31 (14–62)	510	27 (12–57)	0.84 (0.73–0.95) ^*	0.007
2017	617	33 (15–70)	534	32 (17–72)	0.99 (0.88–1.11) ^*	0.857
2008–2017	5361	35 (17–70)	7331	32 (15–62)	0.87 (0.82–0.93)†	<0.001

Note: “*” Adjusted for gender, sputum smear results by using two-level mixed-effects survival model
“†” Adjusted for gender, sputum smear results and years. The estimate of variance in level 2 was 0.32 with the standard error of 0.45. An LR test comparing the model with the one-level survival model didn’t favor the random-intercept model with P > 0.05

ISSN: 1471-2334