Patient Characteristics

Patient characteristics were poorly reported. Stage of breast cancer was rarely reported explicitly, but all studies included only operable cancer (or, in a few

TABLE 1 Characteristics of 61 Studies Presenting and/or Validating Prognostic Models for Breast Cancer Patients

Study characteristics n %

Dates of start and end of recruitment and end of follow-up all given 13 21 Study design

Prospective (including four randomized trials) 10 16

Retrospective 48 79

Unclear 3 5

Justification of sample size 0 0

Inclusion criteria stated explicitly 23 38 Menopausal status

Pre- and postmenopausal 39 64

Postmenopausal only 1 2

Not stated 21 34 Adjuvant therapy (at least some patients)

Chemotherapy 30 49

Hormone therapy 23 38

Radiotherapy 26 43

Immunotherapy 1 2 Patient endpoints analyzed

Explicitly any death 13 21

Not explicitly any death (includes three reporting "overall survival") 22 35

Specifically cancer death 16 26

Relapse (including recurrence, progression or disease free survival) 29 48 (includes 15 studies where it was unclear if death was taken as an event or censored)

Any event (death or relapse) 2 3

Time origin specified 17 29 Summary of length of follow-up

Median (or mean) and range 15 25

Median (or mean) only 19 31

Other 11 18

Not stated 16 26 Median (or mean) follow-up

Not stated 23 38

Clear statement on number of patients lost to follow-upa 14 23

Clear statement on how loss to follow-up treated was in analysisa 11 18

Discussion in text of missing data? (includes six studies with no 19 32 missing data)

Number of excluded patients due to missing data was 29 48 reported (includes six studies with no missing data)

including two studies that stated that no patients were lost.

cases, did not specify). In 33 studies, all patients were reported as having surgery but this information was missing for 23 studies. Some studies gave no information at all about patient characteristics (30,31).

0 0

Post a comment