Pretest expectations strongly influence interpretation of abnormal laboratory results and further management

BMC Primary Care

Table 4 Influence of pretest expectations on interpretation, posttest probability estimates and management after abnormal results.

	Test interpretation	Posttest probability	Management
	= normal^†	= no disease^‡	= no action^§
	odds ratio (95% CI)	odds ratio (95% CI)	odds ratio (95% CI)
Reason for ordering lab tests	p = 0.067	p = 0.52	p = 0.06
exclude disease	1	1	1
reassure patient	1.1 (0.63-1.83)	1.46 (0.67-3.17)	2.25 (1.08-4.66)
confirm diagnosis	0.59 (0.37-0.93)	0.82 (0.42-1.59)	0.80 (0.46-1.40)
Pretest probability estimate	p < 0.001	p < 0.001	p = 0.19
definitely no disease	1	1	1
probably no disease	1.13 (0.56-2.26)	0.35 (0.11-1.08)	1.67 (0.64-4.32)
maybe	0.49 (0.24-1.03)	0.20 (0.06-0.63)	1.11 (0.41-3.00)
probably disease	0.33 (0.15-0.73)	0.07 (0.02-0.23)	1.75 (0.59-5.17)
definitely disease	0.18 (0.07-0.52)	0.04 (0.01-0.23)	2.88 (0.79-10.59)
Test interpretation	N/A*	p < 0.001	p < 0.001
normal		1	1
possibly abnormal		0.09 (0.06-0.15)	0.20 (0.10-0.40)
abnormal		0.01 (0.005-0.02)	0.43 (0.25-0.72)
Posttest probability estimate	N/A*	N/A*	p < 0.001
definitely no disease			1
probably no disease			0.48 (0.26-0.89)
maybe			0.14 (0.07-0.32)
probably disease			0.07 (0.03-0.16)
definitely disease			0.09 (0.04-0.23)

N = 742
*N/A not applicable
^†Goodness-of-fit: Chi-square = 5.29, p = 0.73, ^‡ Chi-square = 7.12, p = 0.52, ^§ Chi-square = 5.70, p = 0.68

ISSN: 2731-4553