Abstract
OBJECTIVES: The aim of this study was to assess the reliability of patency rates calculated on basis of data from a standard vascular registry.
DESIGN AND SETTING: Since 1989, all patients undergoing infrainguinal bypass procedures have been offered a standard follow-up programme at 3 month intervals and all data have been recorded prospectively in a vascular registry. As part of a randomised trial on adjuvant medication in femorocrural bypass surgery, 102 patients, operated on between 1990 and 1992 were independently and simultaneously monitored. This subgroup was examined at 3 and 12 months postoperatively and 100% follow-up was obtained. After completion of the trial we calculated the patency and survival rates using life-table methods and compared the results based on the vascular registry with those achieved in the clinical trial.
RESULTS: Comparing the results from the two databases revealed a marked discrepancy between the calculated figures: primary (68% in the registry and 52% in the trial) and secondary patency rates (90% vs. 63%), limb survival (97% vs. 77% as well as patients survival rates (95% vs. 85%). The differences could be explained by a substantial number of patients being lost to follow-up according to the vascular registry database and the fact that these patients turned out to have a significantly increased rate of graft thrombosis, limb amputation and death, respectively.
CONCLUSIONS: Life-table statistics may inadvertently become unreliable if a large proportion of patients is lost to follow-up, since failure to examine the patient for any reason may be related to the patients health. In addition to the number of patients at risk, it is suggested, that life-table plots should be supplemented with information on the number of patients lost to follow-up.
Originalsprog | Engelsk |
---|---|
Tidsskrift | European journal of vascular and endovascular surgery : the official journal of the European Society for Vascular Surgery |
Vol/bind | 12 |
Udgave nummer | 3 |
Sider (fra-til) | 282-6 |
Antal sider | 5 |
ISSN | 1078-5884 |
Status | Udgivet - okt. 1996 |
Udgivet eksternt | Ja |