Diagnostic intervals before and after implementation of cancer patient pathways – a GP survey and registry based comparison of three cohorts of cancer patients

Table 3 Estimated differences in diagnostic interval (DI) (calendar days) during and after the implementation of CPPs compared to before the implementation (Model 1), and also according to referral route after the implementation: to a CPP (after-CPP) or not (after-no CPP) (Model 2) (N=11,640)

	Model 1¹		Model 2¹after-group split by referral to a CPP or not
	During vs. before	After vs. before	After-CPP vs. before	After-no CPP vs. before	After-CPP vs. during	After-no CPP vs. during
	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)	Estimate (95% CI)
Percentile
25^th	−7 (−8;-5)	−10 (−11;-8)	−12 (−13;-11)	−6 (−8;-5)	−8 (−9;-7)	0 (−2;1)
50^th	−14 (−16;-11)	−17 (−19;-15)	−23 (−25;-21)	−9 (−12;-7)	−15 (−17;-12)	4 (1;7)
75^th	−22 (−27;-16)	−27 (−34;-20)	−46 (−51;-41)	−11 (−21;-1)	−32 (−37;-28)	10 (1;19)
90^th	−53 (−76;-30)	−44 (−65;-23)	−110 (−153;-67)	−6 (−77;66)	−80 (−126;-34)	48 (−49;145)

Estimates with 95% confidence intervals (95%CI) are displayed for the 25^th, the 50^th, the 75^th percentile and the 90^th percentiles. Bold estimates indicate statistical significance at p = 0.05 level or less.
Model 1 reference: before implementation group, cohort, female, 45 years of age, cancer sites, no co-morbidity, high disposable income and high educational level.
Model 2 = model 1, but with ‘after group’ split by referral route (CPP).
¹Adjusted for gender, age, cancer site, co-morbidity, educational level and disposable income.

ISSN: 1471-2407