Misclassification of incident conditions using claims data: impact of varying the period used to exclude pre-existing disease

Table 3 Incidence and prevalence proportions of conditions in breast cancer and non-cancer cohorts

	Incidence ^a proportion		Prevalence ^a proportion
Condition	Breast cancer	Non-cancer	Breast cancer	Non-cancer
Cerebrovascular disease	13.1	3.1	55.1	33.0
Chronic obstructive pulmonary disease	28.6	2.7	66.1	34.0
Congestive heart failure	19.8	2.8	76.9	32.1
Depression	14.2	1.5	35.1	21.1
Diabetes	33.5	3.4	91.1	46.8
Hip fracture	3.7	1.1	17.3	10.6
Hypertension	196.3	13.1	319.8	161.2
Liver disease	1.0	0.2	2.9	1.5
Myocardial infarction	12.7	2.0	34.8	23.1
Osteoarthritis	27.4	4.6	83.8	58.0
Osteoporosis	22.4	2.2	40.7	23.1
Renal disease	2.2	0.2	8.0	2.9

^a Both incidence and prevalence proportions reported as /1,000 patients at risk. Incidence period was 3 months after the date of cancer diagnosis (cancer patients) or sham date (non-cancer patients). Prevalence period was 36 months before the date of cancer diagnosis (cancer patients) or sham date (non-cancer patients).

ISSN: 1471-2288