Table 2.

Method and completeness of death ascertainment and method of following up patients who miss appointments among 17 HIV cohort studies from Europe and North America

	Death ascertainment			Lost to follow-up
Cohort	Linked to death registry (frequency in years)	Sources of information	Self-reported completeness of ascertainment (%)	% lost to clinical follow-up for AIDS	% lost to follow-up for death	Missed appointment follow-up
ATHENA, The Netherlands	No	Hospital, family/friends	90–95	4	4	Phone call to patient
Aquitaine, France	Yes (3)	Hospital, physician	90–95	6	6	Physician
Koln/Bonn cohort, Germany	No	Physician, local registry	90–95	16	16	Phone call to patient, physician
CoRIS, Spain	No		90–95	9	9	No
EuroSIDA, Europe	Some		90–95	5	5
FHDH, France	No	Local registry, family/friends	75–89	17	18	Unknown, phone call or letter to patient
ICONA, Italy	No	Local registry/hospital	90–95	12	12	Phone call to patient
PISCIS, Spain	Yes (2)	Hospital, physician	75–89	5	5	Unknown
Royal Free, UK	Yes (1)a	Physician, registry	75–89	7	8	Phone call or letter to patient
SHCS, Switzerland	No	Local registry	90–95	5	4	Letter to patient
S. Alberta, Canada	Yes (1)	Hospital, coroner	90–95	8	6	No
HOMER, Canada	Yes (1/12)	Provincial registry, physician	96–100	21	6	Physician (questionnaire)
HAVACS, USA	Yes (0.25)	Registry	96–100	3	3	Phone call to patient
UAB, USA	Yes (0.5)	Hospital, registry, family/friends	75–89	19	17	Phone call to patient
VACS, USA	Yes (1)	Hospital, physician	96–100	2	2
Vanderbilt, USA	Yes (0.5)	Hospital, registry, family/friends	90–95	12	13	Phone call to patient
WA, USA	Yes (0.25)	Hospital, registry	96–100	12	12	Phone call and letter to patient

^aLinkage to death registry was restricted to a subgroup of younger patients (aged <65 years).