Decreasing mortality and changing patterns of causes of death in the Swiss HIV Cohort Study

Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich, Switzerland.
HIV Medicine (Impact Factor: 3.99). 04/2013; 14(4):195-207. DOI: 10.1111/j.1468-1293.2012.01051.x.
ABSTRACT
Background:
Mortality among HIV-infected persons is decreasing, and causes of death are changing. Classification of deaths is hampered because of low autopsy rates, frequent deaths outside of hospitals, and shortcomings of International Statistical Classification of Diseases and Related Health Problems (ICD-10) coding.

Methods:
We studied mortality among Swiss HIV Cohort Study (SHCS) participants (1988-2010) and causes of death using the Coding Causes of Death in HIV (CoDe) protocol (2005-2009). Furthermore, we linked the SHCS data to the Swiss National Cohort (SNC) cause of death registry.

Results:
AIDS-related mortality peaked in 1992 [11.0/100 person-years (PY)] and decreased to 0.144/100 PY (2006); non-AIDS-related mortality ranged between 1.74 (1993) and 0.776/100 PY (2006); mortality of unknown cause ranged between 2.33 and 0.206/100 PY. From 2005 to 2009, 459 of 9053 participants (5.1%) died. Underlying causes of deaths were: non-AIDS malignancies [total, 85 (19%) of 446 deceased persons with known hepatitis C virus (HCV) status; HCV-negative persons, 59 (24%); HCV-coinfected persons, 26 (13%)]; AIDS [73 (16%); 50 (21%); 23 (11%)]; liver failure [67 (15%); 12 (5%); 55 (27%)]; non-AIDS infections [42 (9%); 13 (5%); 29 (14%)]; substance use [31 (7%); 9 (4%); 22 (11%)]; suicide [28 (6%); 17 (7%), 11 (6%)]; myocardial infarction [28 (6%); 24 (10%), 4 (2%)]. Characteristics of deceased persons differed in 2005 vs. 2009: median age (45 vs. 49 years, respectively); median CD4 count (257 vs. 321 cells/μL, respectively); the percentage of individuals who were antiretroviral therapy-naïve (13 vs. 5%, respectively); the percentage of deaths that were AIDS-related (23 vs. 9%, respectively); and the percentage of deaths from non-AIDS-related malignancies (13 vs. 24%, respectively). Concordance in the classification of deaths was 72% between CoDe and ICD-10 coding in the SHCS; and 60% between the SHCS and the SNC registry.

Conclusions:
Mortality in HIV-positive persons decreased to 1.33/100 PY in 2010. Hepatitis B or C virus coinfections increased the risk of death. Between 2005 and 2009, 84% of deaths were non-AIDS-related. Causes of deaths varied according to data source and coding system.

Full-text

Available from: Justyna D Kowalska, Jul 14, 2015
Decreasing mortality and changing patterns of causes of
death in the Swiss HIV Cohort Study*
R Weber,
1†
M Ruppik,
1†
M Rickenbach,
2
A Spoerri,
3
H Furrer,
4
M Battegay,
5
M Cavassini,
6
A Calmy,
7
E Bernasconi,
8
P Schmid,
9
M Flepp,
10
J Kowalska,
11
B Ledergerber
1
and the Swiss HIV Cohort Study (SHCS)
1
Division of Infectious Diseases and Hospital Epidemiology, University Hospital Zurich, University of Zurich, Zurich,
Switzerland,
2
SHCS Data Center, University Hospital Lausanne, University of Lausanne, Lausanne, Switzerland,
3
Institute of Social and Preventive Medicine, University of Bern, Bern, Switzerland (for the Swiss National Cohort),
4
Division of Infectious Diseases, Bern University Hospital, University of Bern, Bern, Switzerland,
5
Division of Infectious
Diseases and Hospital Epidemiology, University Hospital Basel, Basel, Switzerland,
6
Division of Infectious Diseases,
University Hospital Lausanne, University of Lausanne, Lausanne, Switzerland,
7
Division of Infectious Diseases,
University Hospital Geneva, Geneva, Switzerland,
8
Division of Infectious Diseases, Regional Hospital, Lugano,
Switzerland,
9
Division of Infectious Diseases, Cantonal Hospital of St. Gallen, St. Gallen, Switzerland,
10
Center for
Infectious Diseases, Clinic “im Park”, Zurich, Switzerland and
11
Copenhagen HIV Programme, Copenhagen, Denmark
(for the CoDe Collaboration)
Background
Mortality among HIV-infected persons is decreasing, and causes of death are changing.
Classification of deaths is hampered because of low autopsy rates, frequent deaths outside of
hospitals, and shortcomings of International Statistical Classification of Diseases and Related
Health Problems (ICD-10) coding.
Methods
We studied mortality among Swiss HIV Cohort Study (SHCS) participants (1988–2010) and causes
of death using the Coding Causes of Death in HIV (CoDe) protocol (2005–2009). Furthermore, we
linked the SHCS data to the Swiss National Cohort (SNC) cause of death registry.
Results
AIDS-related mortality peaked in 1992 [11.0/100 person-years (PY)] and decreased to 0.144/100
PY (2006); non-AIDS-related mortality ranged between 1.74 (1993) and 0.776/100 PY (2006);
mortality of unknown cause ranged between 2.33 and 0.206/100 PY. From 2005 to 2009, 459 of
9053 participants (5.1%) died. Underlying causes of deaths were: non-AIDS malignancies [total,
85 (19%) of 446 deceased persons with known hepatitis C virus (HCV) status; HCV-negative
persons, 59 (24%); HCV-coinfected persons, 26 (13%)]; AIDS [73 (16%); 50 (21%); 23 (11%)];
liver failure [67 (15%); 12 (5%); 55 (27%)]; non-AIDS infections [42 (9%); 13 (5%); 29 (14%)];
substance use [31 (7%); 9 (4%); 22 (11%)]; suicide [28 (6%); 17 (7%), 11 (6%)]; myocardial
infarction [28 (6%); 24 (10%), 4 (2%)]. Characteristics of deceased persons differed in 2005 vs.
2009: median age (45 vs. 49 years, respectively); median CD4 count (257 vs. 321 cells/mL,
respectively); the percentage of individuals who were antiretroviral therapy-naïve (13 vs. 5%,
respectively); the percentage of deaths that were AIDS-related (23 vs. 9%, respectively); and the
percentage of deaths from non-AIDS-related malignancies (13 vs. 24%, respectively).
Correspondence: Dr Rainer Weber, Division of Infectious Diseases and Hospital Epidemiology, University Hospital, CH-8091 Zurich, Switzerland.
Tel: +41 44 255 38 26 (direct); +41 44 255 11 11 (hospital); fax: +41 44 255 32 91; e-mail: rainer.weber@usz.ch.
Alternate author for correspondence: Dr Bruno Ledergerber, Division of Infectious Diseases and Hospital Epidemiology, University Hospital, CH-8 091
Zurich, Switzerland. Tel: +41 44 255 33 57 (direct); +41 44 255 11 11 (hospital); fax: +41 44 255 32 91; e-mail: infled@usz.uzh.ch
*Part of this work was presented at the 18
th
Conference on Retroviruses and Opportunistic Infections, 27 February to 2 March 2011, Boston, MA
(Poster number O-194).
These authors contributed equally to this work.
See Acknowledgements.
DOI: 10.1111/j.1468-1293.2012.01051.x
© 2012 British HIV Association HIV Medicine (2012)
ORIGINAL RESEARCH
1
Page 1
Concordance in the classification of deaths was 72% between CoDe and ICD-10 coding in the
SHCS; and 60% between the SHCS and the SNC registry.
Conclusions
Mortality in HIV-positive persons decreased to 1.33/100 PY in 2010. Hepatitis B or C virus
coinfections increased the risk of death. Between 2005 and 2009, 84% of deaths were
non-AIDS-related. Causes of deaths varied according to data source and coding system.
Keywords: causes of death, hepatitis C virus coinfection, HIV infection, national death registry,
prospective observational database.
Accepted 30 July 2012
Introduction
Data on causes of death inform priorities in care and
prevention, are an element of quality control of care, and
may indicate adverse effects of medical interventions. Pat-
terns of morbidity and mortality of HIV-positive persons
with access to antiretroviral therapy (ART) are changing as
a result of immune reconstitution, prolonged survival
[1–10], and aging [11]. Despite the success of combination
antiretroviral therapy (ART) since the mid-1990s, mortality
and causes of death still differ between HIV-positive and
HIV-negative persons [12–14] for various reasons: presen-
tation at a late stage of infection and late start of ART;
treatment failure; prolonged immunodeficiency prior to
treatment; a state of lasting chronic inflammation despite
complete suppression of HIV replication; coinfections with
hepatitis B virus (HBV), hepatitis C virus (HCV), or onco-
genic viruses; medication-related toxicities; use of illicit or
recreational drugs; life-style-related risks for disease; or no
access to care or ART [5,11].
Clinical endpoints, including deaths, are nowadays
occurring rarely during short-term controlled treatment
trials, but rather are observed in long-term cohort
studies including large numbers of participants [15].
However, diagnostic accuracy of causes of death in HIV-
positive persons is hampered because of many factors,
including low autopsy rates, increasing numbers of deaths
outside of medical institutions, and logistic and legal
difficulties in receiving appropriate information on clini-
cal events occurring outside of study sites. Furthermore,
different data sources, coding algorithms or classification
systems may result in different patterns of causes of death
[16,17].
We aimed, first, to study the characteristics of partici-
pants of the Swiss HIV Cohort Study (SHCS) who died from
2005 to 2009, and their causes of death: we used the
Coding Causes of Death in HIV (CoDe) protocol [16,18] to
better categorize causes of death, to reduce the rate of
missing data, to distinguish immediate and underlying
causes of death, and to adjudicate causes of deaths. Fur-
thermore, for quality control purposes, we linked the SHCS
data with the Swiss National Cohort (SNC) which includes
a national cause of death registry [19]. Secondly, we
describe the mortality, autopsy rates, and location of death
in SHCS participants from 1988 to 2010.
Methods
Study design and data collection
The SHCS is a prospective observational cohort study with
continued enrolment of HIV-infected persons, aged 18
years, who attend out-patient clinics of seven cohort
centres, affiliated regional hospitals, or private practition-
ers collaborating with the centres [20]. Standardized data
collection forms containing demographic, psychosocial,
clinical, laboratory and treatment information are com-
pleted every 6 months (http://www.shcs.ch).
HIV-associated diseases, non-AIDS-related malignan-
cies, and causes of death were documented since 1988, and
endpoints of the Data Collection on Adverse Events of
Anti-HIV Drugs (D:A:D) since 1999 [21].
Causes of death and definitions
From 1988 to 1998, the SHCS classified causes of death in
broad categories: AIDS-related, overdose of narcotics,
suicide, accident, homicide, other and unknown. AIDS-
related deaths included deaths according to the 1986 and
1993 Centers for Disease Control and Prevention defini-
tions of AIDS-defining infections and AIDS-defining
malignancies [22,23]. Non-AIDS-related deaths included
all other known causes of death (excluding deaths of
unknown cause which were analysed separately).
In 1999, coding of deaths according to the International
Statistical Classification of Diseases and Related Health
Problems (ICD-10) [24] was introduced.
In 2005 the SHCS started to use, in addition, the
CoDe protocol, which was designed by an international
2 R Weber et al.
© 2012 British HIV Association HIV Medicine (2012)
Page 2
collaboration of HIV care physicians to deliver a standard-
ized method for coding causes of death in HIV-positive
persons [16] (protocol and questionnaire: [18]). Using a
four-page structured questionnaire (CoDe form), the
patients clinical and autopsy data are reviewed and docu-
mented by local study personnel, and causes of death are
coded within 33 categories according to pathomechanisms
of death or the underlying organ system morbidity
(Appendix Tables A1 and A2). Immediate, contributing and
underlying causes of death are distinguished; and the ques-
tionnaire is used to adjudicate causes of death by a central
endpoint review committee. An immediate cause of death
is a disease or injury directly leading to death; a contrib-
uting cause contributes to the fatal outcome; and an under-
lying cause is the morbidity that initiated the sequence of
events leading directly or indirectly to death [18]. For this
analysis, we added codes for HCV- or HBV-related hepa-
tocellular carcinoma (HCC). Immediate and underlying
causes of death are reported.
At the time of data analysis, 151 CoDe forms of the
SHCS had been adjudicated by the D:A:D coordinating
centre at the Copenhagen HIV Program. The remaining
308 deaths were reviewed within the SHCS by two
assigned physicians according to the CoDe principles (MR
and RW). In case of a disagreement, a review of the source
data was carried out in collaboration with the correspond-
ing SHCS centre.
Active HBV infection was defined as positive HBV
surface or positive HBV e antigen or detectable HBV DNA.
Active HCV infection was defined as positive HCV antibody
and detectable HCV RNA.
Swiss National Cohort (SNC)
To investigate whether deaths and causes of death were
missed in the SHCS, we linked the SHCS and the SNC
databases. The SNC is a long-term, census-based, multi-
purpose cohort and research platform which is based on the
linkage of individual data from the 1990 to the 2000
census [19,25]. This basic database has been enhanced with
information on all-cause and cause-specific mortality by
linking it to the national cause of death registry based on
death certificates collected by the Swiss Federal Statistical
Office [26]. In the SNC, mortality data and causes of deaths
are available until 2008. Probabilistic record linkage
between the SHCS database (2005–2008) and the SNC
death registry (2005–2008) was carried out using the Gen-
eralized Record Linkage System package developed by Sta-
tistics Canada [27], and was based on date of birth, sex,
marital status, education, nationality, type of household,
region of residence, date last known to be alive, and, where
applicable, date of death.
Statistical analyses
We used c
2
or Fisher’s exact tests to compare categorical
variables. Continuous variables were analysed with Wil-
coxon rank-sum and Kruskal–Wallis tests. To measure
trends across ordered groups we used the nonparametric
test for trend developed by Cuzick [28].
Associations between death and demographic and clini-
cal variables (including HBV and HCV infection status) and
ART status were analysed in univariable and multivariable
Poisson regression models. Fixed covariables were sex,
mode of HIV acquisition and ethnicity. Time-updated cov-
ariables were age, arterial hypertension, diabetes mellitus,
prior cardiovascular event, active HBV infection and active
HCV infection; further time-updated variables were lagged
by 6 months, including CD4 cell count, smoking status,
body mass index (BMI), and ART status (never treated,
on ART, or ART interrupted for >1 month). The ‘latest’
measurement for the lagged variables had to have been
measured at least 6 months prior to death (to reduce
confounding with the ‘process of dying’), or at the latest
time-point for those remaining alive.
We used
STATA (Version 12.0; StataCorp, College Station,
TX) for analyses.
Results
Deaths in the SHCS, 1988–2010
A total of 5023 (31%) of 16 134 SHCS participants died
from 1988 to 2010. Potent ART became available in
Switzerland in 1996. Causes of death in the periods 1988–
1995, 1996–2004 and 2005–2010, respectively, were: AIDS,
78, 41 and 15%; suicide, 3, 3 and 6%; substance use, 2, 3
and 5%; accident/homicide, 0, 2 and 2%; and non-AIDS-
related diseases (excluding suicide, substance use and
accidental deaths), 17, 51 and 71%.
AIDS-associated mortality peaked in 1992, with 11.0
[95% confidence interval (CI) 9.94-12.1] deaths per 100
person-years (PY), and decreased to 0.144 (95% CI 0.077-
0.267)/100 PY in 2006; non-AIDS-associated mortality
decreased from 1.74 (95% CI 1.36-2.23) in 1993 to 0.776
(95% CI 0.594-1.01)/100 PY in 2008; and mortality of
unknown cause decreased from 2.33 (95% CI 1.88-2.89) in
1994 to 0.207 (95% CI 0.125-0.343)/100 PY in 2007
(Fig. 1a,b). In 2010, mortality rates because of AIDS, non-
AIDS, or unknown causes were 0.211 (95% CI 0.122-0.363),
0.860 (95% CI 0.657-1.13) and 0.260 (95% CI 0.159-0.424)
per 100 PY.
Autopsy rates
From 1988 to 1996 the autopsy rate fell from 40 to 7%.
Subsequently, when AIDS as a cause of death decreased,
Causes of death and HIV infection 3
© 2012 British HIV Association HIV Medicine (2012)
Page 3
autopsy rates increased to around 30%. From 2005 to 2009,
autopsies were performed in 89 (19%) of 459 deaths
(Fig. 1c).
Place of death
Places of death are documented since 1995 (Fig. 1d).
Overall, less than 50% of patients died in acute care hos-
pitals. From 2005 to 2009, 45% of 459 deaths occurred in
hospitals, 22% at home, 11% in nursing homes, 4% in
hospices and 18% at other or unknown places.
Characteristics of SHCS participants who died between
2005 and 2009
A total of 459 (5.1%) (340 male and 119 female) of 9053
SHCS participants under follow-up died between 2005 and
2009. Mortality was 1.25 (95% CI 1.14-1.37) per 100 PY.
Table 1 summarizes participants’ characteristics at the time
of death: their median age was 47 years; the median
duration since diagnosis of HIV infection was 14 years; ART
was ever taken by 93% of participants; and their median last
CD4 lymphocyte count was 251 [interquartile range (IQR)
114–428] cells/mL. Active HCV or HBV infection was present
in 45% and 11% of persons who died, respectively. Thirty-
six per cent had a history of injecting drug use (IDU).
Causes of death
The most frequent underlying causes of death (according to
the CoDe classification) were: 19% non-AIDS-related
malignancies (including HCC); 16% AIDS; 15% liver failure
excluding HCC (18% liver failure including HCC); 9% non-
AIDS-related infections; 7% substance use-related; 6%
suicide; and 6% myocardial infarction (Table 2). The pro-
portion of unknown causes of death was 3%.
Sex differences
Characteristics of men and women at time of death were
similar, with a few exceptions: age (median 48 years for
Mortality rate
per 100 patient-years
1988
1990
1992
1994
1
996
1998
2000
2002
2004
2006
2008
2010
0
5
10
15
AIDS-associated death
Not AIDS associated death
Causes of death unknown
(a) (b)
(c)
Number of deaths per year
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
0
100
200
300
400
AIDS-associated death
Not AIDS associated death
Causes of death unknown
(d)
Number of deaths per year
Deaths with autopsies
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
0
100
200
300
400
500
600
% with autopsies
Number of deaths
0%
10 %
20 %
30 %
40 %
50 %
Percentage of deaths
1995
1998
2001
2004
2007
2010
0%
20%
40%
60%
80%
100%
AIDS hospice
Hospital
Nursing home
At home
Other/unknown
Place of death
Fig. 1 Mortality, autopsy rates, and location of death in SHCS participants from 1988 to 2010. (a) Numbers of deaths per year in the Swiss HIV Cohort
Study (SHCS), 1988–2010, stratified into deaths because of clinical AIDS (i.e. deaths because of AIDS-defining infections or malignancies [22,23]),
non-AIDS-associated deaths, and deaths of unknown aetiology. (b) Mortality in the SHCS, 1988–2010. AIDS-associated mortality was highest in 1992
[11.0/100 patient-years (PY)] and lowest in 2006 (0.144/100 PY); non-AIDS-associated mortality was highest in 1993 (1.74/100 PY) and lowest in
2006 (0.776/100 PY); and mortality of unknown cause was highest in 1994 (2.33/100 PY) and lowest in 2007 (0.207/100 PY). From 1988 to 1998, the
SHCS classified causes of death into broad categories (HIV-related, overdose of narcotics, suicide, accident, homicide, other and unknown). In 1999,
coding of deaths according to the International Statistical Classification of Diseases and Related Health Problems (ICD-10) [24] system and in 2005
the CoDe protocol [18] were introduced. (c) Numbers of deaths per year in the SHCS, 1988–2010, and proportions of autopsies. (d) Proportion of
deaths at different locations (AIDS hospice, hospital, nursing home, at home, other or unknown places) in the SHCS, 1988–2010.
4 R Weber et al.
© 2012 British HIV Association HIV Medicine (2012)
Page 4
men and 45 years for women; P < 0.001); age at time of
first positive HIV test (median 34 and 28 years, respec-
tively; P < 0.001); duration of HIV infection (median 13.6
and 15.1 years, respectively; P = 0.083); HIV transmission
risk; and ethnicity (Caucasian 94 and 81%; African 3 and
12%; Asian 1 and 4%, respectively; P < 0.001).
The distribution of underlying causes of death was not
different except for the rate of suicides (8% in men and 3%
in women), and types of malignancies (Table 2, footnote).
Changes over calendar time
From 2005 to 2009, there were changes in patient char-
acteristics at the time of death: compared with 2005,
median age was higher in 2009 (45 and 49 years,
respectively; P < 0.001); duration of HIV infection was
longer (median 13 and 16 years, respectively; P = 0.002);
duration of ART was longer (median 8 and 12 years,
respectively; P < 0.001); CD4 lymphocyte counts were
higher (median 257 and 321 cells/mL, respectively;
P = 0.005); and the rate of ART-naïve patients was
lower (13 and 5%, respectively; P = 0.018); whereas the
age at first positive HIV test was similar, as was the
nadir CD4 cell count (median 119 and 107 cells/mL,
respectively).
The proportion of participants who died from AIDS
decreased (23 and 9%, respectively) and the proportion of
Table 1 Patient characteristics at time of death, stratified by CD4 lymphocyte count
Variable Total
CD4 lymphocyte count at time of death
P
*<50 cells/mL 50–199 cells/mL 200–499 cells/mL >500 cells/mL
No. of deceased participants (%) 459 (100) 63 (13.7) 119 (25.9) 187 (40.7) 90 (19.6)
Female 119 (25.9) 18 (28.6) 30 (25.2) 48 (25.7) 23 (25.6) 0.96
Age at time of death (years) [median (IQR)] 47 (42–56) 45 (38–50) 46 (41–56) 47 (43–58) 49 (44–57) 0.006
Duration of HIV infection (years) [median (IQR)] 13.5 (0.1–27) 13.4 (0.1–24) 14.1 (0.4–24) 13.9 (0.3–25) 13.5 (0.1–27) 0.80
HIV transmission [n (%)] 0.37
MSM 108 (23.5) 8 (12.7) 25 (21.0) 47 (25.1) 28 (31.1)
Heterosexual 130 (28.3) 18 (28.6) 31 (26.1) 55 (29.4) 26 (28.9)
IDU 201 (43.8) 33 (52.4) 57 (47.9) 76 (40.6) 35 (38.9)
Blood products 3 (0.7) 0 (0) 2 (1.7) 1 (0.5) 0 (0)
Perinatal 2 (0.4) 2 (3.2) 0 (0) 0 (0) 0 (0)
Unknown/other 15 (3.3) 2 (3.2) 4 (3.4) 8 (4.3) 1 (1.1)
Ethnicity [n (%)] 0.15
Caucasian 415 (90.4) 52 (82.5) 111 (93.3) 169 (90.4) 83 (92.2)
African 23 (5.0) 4 (6.4) 5 (4.2) 10 (5.4) 4 (4.4)
Asian 8 (1.7) 3 (4.8) 0 (0) 2 (1.1) 3 (3.3)
Other 13 (2.8) 4 (6.4) 3 (2.5) 6 (3.2) 0 (0)
ART at time of death [n (%)] 0.008
Naïve 33 (7.2) 5 (7.9) 9 (7.6) 12 (6.4) 7 (7.8)
Interrupted for longer than 1 month 113 (24.6) 23 (36.5) 34 (28.6) 42 (22.5) 14 (15.6)
On treatment 312 (68.0) 35 (55.6) 76 (63.9) 132 (70.6) 69 (68.0)
ART initiation with mono or dual regimen 112 (26.3) 20 (34.5) 35 (31.8) 38 (21.7) 19 (22.9) 0.78
ART duration (years) [median (IQR)] 9.5 (5.5–12.2) 9.9 (3.8–12.5) 9.1 (4.9–12.2) 9.1 (5.5–12) 10.8 (8.4–12) 0.055
Nadir CD4 lymphocyte count [median (IQR)] 106 (38–203) 19 (6–38) 66 (26–107) 140 (86–232) 194 (108–336) <0.001
Prior clinical AIDS [n (%)] 196 (42.7) 42 (66.7) 60 (50.4) 66 (35.3) 28 (31.1) <0.001
CVD risks
Smoking, ever [n (%)] 377 (82.1) 50 (79.4) 92 (77.3) 158 (84.5) 77 (85.6) 0.41
Hypertension [n (%)] 90 (19.6) 7 (11.1) 13 (10.9) 43 (23.0) 27 (30.0) <0.001
Diabetes mellitus [n (%)] 28 (6.1) 2 (3.2) 8 (6.7) 10 (5.4) 8 (8.9) 0.29
Dyslipidaemia [n (%)] 84 (18.3) 7 (11.1) 11 (9.2) 40 (21.4) 26 (28.9) 0.001
Body mass index (kg/m
2
) [median (IQR)] 21.3 (19–24) 20 (19–22.8) 20.8 (19–24) 21.6 (19–24) 22.4 (20–25) 0.003
Prior cardiovascular event [n (%)] 40 (8.7) 3 (4.8) 8 (6.7) 21 (11.2) 8 (8.9) 0.54
Hepatitis virus infection [n (%)]
Active HCV
202 (44.0) 28 (44.4) 59 (49.6) 78 (41.7) 37 (41.1) 0.65
Active HBV
51 (11.1) 12 (19.1) 14 (11.8) 16 (8.6) 9 (10.0) 0.30
Depression or psychosis [n (%)] 183 (39.9) 21 (33.3) 41 (34.5) 83 (44.4) 38 (42.2) 0.68
Substance use [n (%)]
Alcohol 130 (28.3) 17 (27.0) 28 (23.5) 55 (29.4) 30 (33.3) 0.88
IDU during last year of life 92 (20.0) 14 (22.2) 26 (21.9) 38 (20.3) 14 (15.6) 0.22
Non-injecting drug use 128 (27.9) 15 (23.8) 32 (26.9) 56 (30.0) 25 (27.8) 0.47
ART, antiretroviral therapy; CVD, cardiovascular disease; HBV, hepatitis B virus; HCV, hepatitis C virus; IDU, injecting drug use; IQR, interquartile range; MSM,
men who have sex with men.
*Nonparametric test for trend across latest CD4 count categories.
Active HBV infection: positive HBV surface or HBV e antigen or HBV DNA. Active HCV infection: positive HCV antibody and positive HCV RNA.
Causes of death and HIV infection 5
© 2012 British HIV Association HIV Medicine (2012)
Page 5
non-AIDS-related malignancies increased (13 and 24%,
respectively).
Deaths in different CD4 cell strata
Forty per cent of patients died with a last CD4 count <200
cells/mL, and 20% with a CD4 count >500 cells/mL (Table 1).
More cardiovascular risk factors were observed in patients
in higher CD4 cell strata at death. Among the 182 who died
with CD4 counts <200 cells/mL, 33% died from AIDS
(Table 2); among the 277 who died with CD4 counts >200
cells/mL, causes of death were: non-AIDS-related malig-
nancies, 23%; liver failure (excluding HCC), 12%; myocar-
dial infarction, 9%; substance use, 9%; non-AIDS-related
infections, 8%; suicide, 8%; and AIDS, 6%.
Deaths in different age groups
Sixteen per cent of individuals died before the age of 40
years (categorized as ‘younger’ persons below); 20% were
older than 60 years (categorized as ‘older’ persons). Younger
deceased persons were also younger at the time of first
positive HIV test than older deceased persons, whereas the
duration of HIV infection at time of death was not different.
Younger patients were more likely to be ART-naïve (younger
16%; older 2%); had shorter ART exposure (median 7.2 and
9.3 years, respectively); were more likely to have a history of
IDU (51 and 3%, respectively); and were more likely to be
HCV-coinfected (45 and 7%, respectively). More cardiovas-
cular risk factors were observed in older patients.
Younger patients were more likely than older patients to
die from AIDS (25 and 14%, respectively), liver failure
(excluding HCC) (21 and 1%, respectively), non-AIDS-
related infections (14 and 9%, respectively), substance
use (13 and 0%, respectively) or suicide (4 and 3%,
respectively); and less likely to die from non-AIDS-related
malignancies (6 and 27%, respectively) or cardiovascular
diseases (3 and 14%, respectively).
Table 2 Underlying causes of death between 2005 and 2009 according to the Coding Causes of Death in HIV (CoDe) protocol
Total
CD4 lymphocyte count at time of death
P
§
<50 cells/mL 50–199 cells/mL 200–499 cells/mL >500 cells/mL
No. of deceased participants (%) 459 (100) 63 (13.7) 119 (25.9) 187 (40.7) 90 (19.6) -
AIDS, total 74 (16.1) 33 (52.4) 25 (21.0) 12 (6.4) 4 (4.4) <0.001
Opportunistic infections 37 (8.1) 22 (34.9) 9 (7.6) 4 (2.1) 2 (2.2) <0.001
AIDS-defining malignancies* 30 (6.5) 8 (12.7) 13 (10.9) 7 (3.7) 2 (2.2) 0.001
Other or not classified 7 (1.5) 3 (4.8) 3 (2.5) 1 (0.5) 0 (0) 0.007
Non-AIDS-defining malignancies, incl. HCC
87 (19.0) 4 (6.4) 20 (16.8) 46 (24.6) 17 (18.9) 0.020
Liver failure, excl. HCC 68 (14.8) 8 (12.7) 27 (22.7) 26 (13.9) 7 (7.8) 0.070
HCV infection 55 (12.0) 6 (9.5) 22 (18.5) 23 (12.3) 4 (4.4) 0.082
HBV infection 5 (1.1) 2 (3.2) 1 (0.8) 1 (0.5) 1 (1.1) 0.27
Other 8 (1.7) 0 (0) 4 (3.4) 2 (1.1) 2 (2.2) 0.79
Non-AIDS-related infections
42 (9.2) 5 (7.9) 16 (13.5) 17 (9.1) 4 (4.4) 0.18
Heart failure, total 30 (6.5) 3 (4.8) 1 (0.8) 17 (9.1) 9 (10.0) 0.015
Myocardial infarction 28 (6.1) 3 (4.8) 1 (0.8) 15 (8.0) 9 (10.0) 0.018
Other 2 (0.4) 0 (0) 0 (0) 2 (1.1) 0 (0) 0.61
Central nervous system, total 13 (2.8) 0 (0) 1 (0.8) 9 (4.8) 3 (3.3) 0.057
Stroke 5 (1.1) 0 (0) 1 (0.8) 2 (1.1) 2 (2.2) 0.20
Other 8 (1.7) 0 (0) 0 (0) 7 (3.7) 1 (1.1) 0.16
Renal failure 4 (0.9) 0 (0) 1 (0.8) 2 (1.1) 1 (1.1) 0.47
Gastrointestinal/pancreatic disease 7 (1.5) 1 (1.6) 2 (1.7) 2 (1.1) 2 (2.2) 0.88
Lung disease (excl. cancer and infection) 8 (1.7) 0 (0) 3 (2.5) 4 (2.1) 1 (1.1) 0.79
Substance use, total 33 (7.2) 1 (1.6) 6 (5.0) 14 (7.5) 12 (13.3)
Chronic alcohol use 2 (0.4) 0 (0) 1 (0.8) 0 (0) 1 (1.1) 0.61
Injecting drug use 25 (5.5) 1 (1.6) 5 (4.2) 11 (5.9) 8 (8.9) 0.040
Other 6 (1.3) 0 (0) 0 (0) 3 (1.6) 3 (3.3) 0.029
Suicide or psychiatric disease 29 (6.3) 1 (1.6) 6 (5.0) 12 (6.4) 10 (11.1) 0.017
Accident 9 (2.0) 1 (1.6) 0 (0) 5 (2.7) 3 (3.3) 0.16
Other 41 (8.9) 5 (7.9) 9 (7.6) 17 (9.1) 10 (11.1) 0.40
Unknown 14 (3.1) 1 (1.6) 2 (1.7) 4 (2.1) 7 (7.8) 0.026
Values are n (%).
HCC, hepatocellular carcinoma; HCV, hepatitis C virus; HBV, hepatitis B virus.
*AIDS-defining malignancies, n (%): total, 30 deaths (100%); non-Hodgkin lymphoma, 22 (73%); primary brain lymphoma, 2 (7%); Hodgkin lymphoma, 2
(7%); Kaposi’s sarcoma, 3 (10%); cervical carcinoma, 1 (3%).
Non-AIDS-defining malignancies, n (%): bronchial tree, 30 (34.5); oesophagus, 2 (2.3); biliary tract, 3 (3.4); HCC, 13 (14.9); pancreas, 5 (5.7); stomach, 1 (1.1);
colon, 1 (1.1); rectum, 1 (1.1); anus, 6 (6.9); breast, 5 (5.7); vulva/vagina, 2 (2.3); kidney, 2 (2.3); bladder, 2 (2.3); prostate, 2 (2.3); skin, 3 (3.4); oropharynx,
3 (3.4); haematological, 3 (3.4); glioblastoma, 1 (1.1); and sarcoma, 2 (2.2).
Non-AIDS-related infections, n (%): pneumonia, 18 (42.9); sepsis, 14 (33.3); endocarditis, 7 (16.7); other, 3 (7.1).
§
Test for trend.
6 R Weber et al.
© 2012 British HIV Association HIV Medicine (2012)
Page 6
Impact of HCV coinfection
In 446 (97%) of 459 deceased persons, HCV status was
available; of these patients, 45% had active HCV infection.
HCV prevalence in men and women was similar. HCV-
infected patients were significantly younger than HCV-
negative persons at the time of death (median 44 and 52
years, respectively); younger at the time of HIV infection
(median 28 and 39.5 years, respectively); had a longer
duration of HIV infection until death (median 16.7 and 11.7
years, respectively); were more likely to have HIV trans-
mission via IDU (86 and 9%, respectively); and were more
likely to be ART-naïve (10 and 4%, respectively); whereas
the median ART duration (10 years) was not different.
The most frequent underlying causes of death among
HCV-coinfected persons were: liver failure, including HCC
(32%), non-AIDS-related infections (14%), substance use
(11%) and non-AIDS-related malignancies (8%, excluding
HCC). Causes of death of HCV-negative persons differed
substantially (Fig. 2).
Immediate vs. underlying causes of death
Differences between immediate and underlying causes
of death were substantial in several disease categories:
AIDS-defining infection or malignancy (immediate cause,
10%; underlying cause, 16%), non-AIDS-related infections
(17 and 9%, respectively), chronic HCV or HBV infection
(9 and 16%, respectively), and non-AIDS-related malig-
nancies (10 and 16%, respectively) (Table 3).
CoDe vs. ICD-10 classification of causes of death
Comparing the CoDe vs. the ICD-10 coding in individual
SHCS participants, concordance in the classification of
underlying causes of death was 72%. Main discrepancies of
classification were: liver failure caused by hepatitis virus
infections (13% in CoDe; 5% in ICD-10), other liver failure
(2 and 6%, respectively), and cardiovascular diseases (8 and
11%, respectively) (Table 3).
Comparison of the SHCS and SNC databases
Linkage of the SHCS (2005–2008: 384 deaths) and the SNC
(2005–2008) resulted in probable matches for 308 deaths.
Seventy-six dead persons from the SHCS could not be
linked to the SNC either because they died abroad, or
because matching variables were too imprecise or errone-
ous to result in a sufficiently high linkage score. In addition
to the participants who were known to have died in the
SHCS, an additional 36 (9.4%) probable matches were
identified in the SNC, potentially representing SHCS par-
ticipants who were lost to follow-up (i.e. not seen in the
SHCS for >12 months).
Comparing the classification of underlying causes of
death in individual patients in the SHCS vs. the SNC,
concordance was 60% among the 308 linked persons. Main
discrepancies were: AIDS (SHCS, 16%; SNC, 34%), non-
AIDS-defining infections (9 and 4%, respectively), liver
failure caused by hepatitis virus infections (13 and 3%,
respectively) and HCC (3 and 8%, respectively), while data
for other diagnostic categories showed smaller differences
(Table 3).
Associations of clinical variables, HBV and HCV
infection status, and death
The results of the uni- and multivariable Poisson regression
analyses for the association between demographic and
clinical variables, HBV and HCV infection status and death
are displayed in Table 4. An increased risk of death was
found for persons with HIV transmission via IDU, increas-
ing age, low CD4 cell counts, current smoking, diabetes
mellitus, low BMI, a prior cardiovascular event, active HBV
infection [incidence rate ratio 1.60 (95% CI 1.12-2.27)],
active HCV infection [1.49 (95% CI 1.07-2.07)] and
interrupted ART.
Discussion
Mortality in HIV-positive persons with access to care is
continuously decreasing, and causes of death are changing
[2,4,8–10,29–32]. In 2010, the total mortality rate in the
SHCS was 1.33 per 100 PY, including AIDS-associated
Fig. 2 Proportion of underlying causes of death in participants of the
Swiss HIV Cohort Study (SHCS) with active hepatitis C virus (HCV)
coinfection vs. HCV RNA-negative participants. A total of 202 (45%) of
446 participants with available HCV status had active HCV infection. All
hepatocellular carcinomas (HCCs) in HCV-negative persons were a con-
sequence of hepatitis B virus (HBV) coinfection.
Causes of death and HIV infection 7
© 2012 British HIV Association HIV Medicine (2012)
Page 7
mortality (0.211/100 PY), non-AIDS-associated mortality
(0.860/100 PY), and mortality of unknown cause (0.260/
100 PY). In the period 2005–2009, the proportion of non-
AIDS-related death was 84%, and malignancies became the
most frequent underlying cause of death (26%; including
19% non-AIDS-defining plus 7% AIDS-defining malignan-
cies). Smoking and other modifiable cardiovascular risks,
substance use, and HCV coinfection substantially influ-
enced the distribution of causes of death. Moreover,
characteristics of HIV-positive persons who died were
changing: from 2005 to 2009, median age increased from
45 to 49 years, and the median CD4 cell count at time of
death increased from 257 to 321 cells/mL.
The distribution of causes of death in different studies is
not directly comparable because of different coding, or
dissimilarities in the selection of study participants such
as differences in the proportions of male vs. female patients
[33], those with IDU [34] and those with HCV coinfection
[35]; differences in age group distributions [11]; and dif-
ferences in terms of socioeconomic status [36], access to
health care [37] or type of health insurance [38]. Never-
theless, recent reports show that non-AIDS-related diseases
became major causes of death: proportions of non-AIDS-
related malignancies were increasing (5 to 17% in refer-
ences [5,8–10,31,32]; 19% in this report), as were liver
diseases (9–15% and 15%, respectively), non-AIDS-related
infections (4–8% and 9%, respectively), and cardiovascular
diseases (6–12% and 7%, respectively); while AIDS-related
causes (10–74% and 16%, respectively) were decreasing.
Confirming causes of death is challenging: the autopsy
rate was 19% in our study, more than half of patients died
outside a hospital, and many legal and logistical problems
hindered collection of data for patients who died outside of
a study site. The diagnostic specificity of the clinical
assessment of causes of death is lower compared with
autopsy [39–41]. Furthermore, it was questioned whether a
Table 3 Comparison of the Swiss HIV Cohort Study (SHCS) and Swiss National Cohort (SNC) databases, and comparison of different codings of
causes of death
Causes of death
SHCS database 2005–2009* SNC database 2005–2008
CoDe classification ICD-10 coding ICD-10 coding
Immediate cause Underlying cause Primary cause Definitive primary cause
n
%
n
%
n
%
n
%
Total 459 100.0 459 100.0 459 100.0 308 100.0
AIDS, total 45 9.8 74 16.1 62 13.5 105 34.1
AIDS, infections 21 4.6 37 8.1 18 3.9 24 7.8
AIDS, malignancies 19 4.1 30 6.5 26 5.7 21 6.8
AIDS, other 5 1.1 7 1.5 18 3.9 60 19.5
Malignancy, non-AIDS-defining (excl. hepatocellular
carcinoma)
47 10.2 74 16.1 60 13.1 46 14.9
Hepatocellular carcinoma 5 1.1 13 2.8 11 2.4 25 8.1
Liver failure caused by hepatitis virus infection (excl.
hepatocellular carcinoma)
35 7.6 60 13.1 22 4.8 8 2.6
Liver failure, other than hepatitis virus infection 7 1.5 8 1.7 28 6.1 9 2.9
Infection, non-AIDS-related 78 17.0 42 9.2 48 10.5 13 4.2
Myocardial infarction 26 5.7 28 6.1 14 3.1 8 2.6
Cardiovascular disease, other 10 2.2 2 0.4 22 4.8 16 5.2
Stroke 14 3.1 5 1.1 14 3.1 7 2.3
Central nervous system disease 8 1.7 8 1.7 3 0.7 1 0.3
Renal failure 3 0.7 4 0.9 3 0.7 1 0.3
Gastrointestinal bleeding 18 3.9 1 0.2 7 1.5 1 0.3
Gastrointestinal disease, other 2 0.4 5 1.1 5 1.1 3 1.0
Pancreatitis 1 0.2 1 0.2 1 0.2 0 0.0
Lactic acidosis 2 0.4 0 0.0 0 0.0 0 0.0
Pulmonary hypertension 1 0.2 2 0.4 0 0.0 0 0.0
Pulmonary embolism 4 0.9 3 0.7 3 0.7 0 0.0
Chronic obstructive pulmonary disease 1 0.2 3 0.7 3 0.7 2 0.6
Lung disease, other (excl. malignancies and infections) 6 1.3 0 0.0 5 1.1 1 0.3
Substance use 28 6.1 33 7.2 22 4.8 24 7.8
Suicide, psychiatric disease 26 5.7 29 6.3 34 7.4 16 5.2
Accident or other violent death (not suicide) 7 1.5 9 2.0 13 2.8 6 1.9
Other causes 11 2.4 41 8.9 4 0.9 2 0.6
Unknown causes 74 16.1 14 3.1 75 16.3 14 4.5
CoDe, Coding Causes of Death in HIV; ICD-10, International Statistical Classification of Diseases and Related Health Problems.
*In 2005–2009 there were 459 deaths; in 2005–2008 there were 384 deaths in the SHCS.
8 R Weber et al.
© 2012 British HIV Association HIV Medicine (2012)
Page 8
Table 4 Univariable and multivariable Poisson regression of demographic, clinical and anthropometric covariables potentially affecting the risk of
death*
IR, per 100 PYFU
(95% CI)
IRR, univariable models
(95% CI)
IRR, multivariable model
(95% CI)
Sex/mode of HIV infection
Male heterosexual 1.20 (0.943-1.52) 1.45 (1.07-1.97) 1.16 (0.850-1.59)
Female heterosexual 0.714 (0.547-0.932) 0.865 (0.624-1.20) 1.04 (0.732-1.47)
Men who have sex with men 0.824 (0.682-0.997) 1 (reference) 1 (reference)
Male injecting drug user 3.19 (2.69-3.78) 3.87 (3.00–4.99) 2.05 (1.39-3.03)
Female injecting drug user 2.36 (1.83-3.05) 2.86 (2.08-3.94) 1.56 (1.01–2.42)
Other 1.29 (0.812-2.05) 1.56 (0.949-2.58) 1.39 (0.834-2.30)
Age
16–39 years 0.675 (0.531-0.858) 1 (reference) 1 (reference)
40–49 years 1.23 (1.07-1.42) 1.82 (1.38-2.41) 1.38 (1.03-1.84)
50–59 years 1.58 (1.30-1.94) 2.35 (1.72-3.21) 2.26 (1.63-3.13)
60 years 2.90 (2.35-3.58) 4.30 (3.12-5.91) 5.92 (4.09-8.59)
Ethnicity
White 1.42 (1.28-1.56) 1 (reference) 1 (reference)
Black 0.470 (0.300-0.737) 0.332 (0.210-0.526) 0.852 (0.510-1.42)
Hispanic 0.716 (0.298-1.72) 0.506 (0.209-1.22) 1.23 (0.503-2.99)
Asian 0.684 (0.342-1.37) 0.483 (0.240-0.973) 0.993 (0.484-2.04)
Unknown 4.78 (2.28-10.0) 3.38 (1.60-7.13) 1.49 (0.698-3.18)
CD4 count
<200 cells/mL 5.18 (4.40-6.11) 1 (reference) 1 (reference)
200–499 cells/mL 1.19 (1.03-1.37) 0.230 (0.185-0.285) 0.323 (0.258-0.404)
500 cells/mL 0.68 (0.561-0.823) 0.131 (0.102-0.169) 0.215 (0.165-0.280)
Smoking status
Never 0.665 (0.529-0.835) 1 (reference) 1 (reference)
Previous 1.00 (0.789-1.27) 1.50 (1.08-2.09) 1.16 (0.829-1.63)
Current 1.81 (1.62-2.03) 2.73 (2.11-3.52) 1.86 (1.39-2.49)
Arterial hypertension
†§
No 1.24 (1.11-1.37) 1 (reference) 1 (reference)
Yes 1.48 (1.20-1.83) 1.20 (0.951-1.52) 1.19 (0.939-1.51)
Diabetes mellitus
No 1.22 (1.11-1.35) 1 (reference) 1 (reference)
Yes 2.63 (1.89-3.67) 2.15 (1.52-3.04) 1.77 (1.23-2.57)
BMI
<18.5 kg/m
2
3.35 (2.65-4.22) 2.62 (2.02–3.41) 1.68 (1.28-2.21)
18.5–24.9 kg/m
2
1.28 (1.13-1.43) 1 (reference) 1 (reference)
25–29.9 kg/m
2
0.77 (0.604-0.984) 0.605 (0.462-0.792) 0.677 0.512-0.894)
30 kg/m
2
1.26 (0.853-1.87) 0.990 (0.657-1.49) 1.02 (0.664-1.55)
Prior cardiovascular event
No 1.21 (1.10-1.33) 1 (reference) 1 (reference)
Yes 2.99 (2.19-4.07) 2.47 (1.79-3.42) 1.72 (1.22-2.41)
Active HBV infection
No 1.24 (1.12-1.37) 1 (reference) 1 (reference)
Yes 2.05 (1.47-2.85) 1.65 (1.17-2.33) 1.60 (1.12-2.27)
Active HCV infection
No 0.854 (0.749-0.974) 1 (reference) 1 (reference)
Yes 2.59 (2.27-2.96) 3.03 (2.52-3.66) 1.49 (1.07-2.07)
ART
Naïve 0.737 (0.571-0.951) 0.651 (0.491-0.864) 0.899 (0.637-1.20)
On treatment 1.13 (1.00–1.28) 1 (reference) 1 (reference)
Interrupted >1 month 3.63 (3.03–4.34) 3.21 (2.58-3.98) 3.16 (2.53-3.96)
ART, antiretroviral therapy; BMI, body mass index; CI, confidence interval; HBV, hepatitis B virus; HCV, hepatitis C virus; IR, incidence rate; IRR, incidence
rate ratio; PYFU, person-years of follow-up.
*Analyses based on 8681 of 8782 persons seen between 2005 and 2009 with information on CD4 cell counts, followed for 34 249 person-years, of whom
438 died. The overall incidence rate of death was 1.28 (95% CI 1.16-1.40) per 100 PYFU. Characteristics that differ significantly from the reference category
(P < 0.05) are shown in bold.
Time-updated variable.
Time-updated variable lagged by 6 months.
§
Defined as systolic blood pressure > 160 mm Hg or diastolic blood pressure > 90 mm Hg.
Causes of death and HIV infection 9
© 2012 British HIV Association HIV Medicine (2012)
Page 9
monocausal explanation of causes of death is appropriate
in HIV-positive persons because several cofactors may sub-
stantially contribute to death, including comorbidities,
polypharmacological treatment, substance use [42] and
socioeconomic factors [36–38]. Also, it is an immanent
difficulty of observational cohort studies to obtain infor-
mation on participants who were lost to follow-up. The
latter problem was not the focus of this study, but linkage
of the SHCS with the SNC revealed an additional 9.4%
probable deaths among SHCS participants in the drop-out
category.
To improve the quality of classification of causes of
death, we used the CoDe protocol, which includes system-
atic collection of information specifically to classify causes
of death in HIV-positive persons; a classification system
adapted to morbidity in HIV-positive persons; and adjudi-
cation of diagnoses [18]. While, for HIV-positive persons,
the ICD-10 coding system often tends to relate deaths to
HIV also in non-HIV-related causes, the CoDe system uses
coding according to organ system or aetiology, and differ-
entiates among immediate, contributing and underlying
causes of death [16].
In our study, discordance between the routinely collected
ICD-10 coded diagnoses in the SHCS and the adjudicated
data according to the CoDe protocol was 28%. Reasons for
discordance were not systematically studied, but may
include incorrect use of the ICD-10 coding, neglect of cause
and effect of events leading to death, incorrect association
of causes of death with HIV infection, less careful collection
of necessary information used for coding, and lack of
adjudication. The discordance of 40% between the SHCS
and the SNC data may partially be explained by ICD-10
coding which overemphasizes the contribution of HIV to
death, and by the possibility that some death certificates
may have been completed by physicians not specialized in
HIV medicine. Also, a limitation of the SNC is its reliance on
routine death certificates [19,43,44], which has been found
to be satisfactorily reliable for clearly defined diagnoses
(e.g. malignancies and accidents) but less accurate for less
well-defined conditions (e.g. chronic obstructive pulmo-
nary disease and diseases of the nervous system) [45–47].
Strengths of our study include the prospective collection
of incident events using structured reporting forms, classi-
fication of causes of death according to the systematic CoDe
protocol, distinction between immediate and underlying
causes of death, and central adjudication of endpoints.
However, several limitations should be noted. First, the
autopsy rate was low, and coding of deaths that occurred
outside a medical institution was often not possible. Sec-
ondly, permission for record linkage to the national death
registry in Switzerland is limited to make use of anonymized
data only. Thirdly, some patients with end-stage disease may
have been lost to follow-up in the cohort because terminally
ill patients may prefer other institutions than study centres,
or may have died outside of Switzerland.
In conclusion, our data have important methodological
and clinical implications. First, the distribution of causes of
deaths varied significantly according to data source and
coding system [17]. Therefore, future efforts should rein-
force the awareness among HIV care physicians that inves-
tigation and accurate documentation of deaths are
important. Methods developed to harmonize coding algo-
rithms and classification systems must be made widely
available [16], and experience gained from their applica-
tion disseminated [17,32]. Future development should
ensure that methods remain sensitive enough to detect rare
events (e.g. ART-related portal hypertension without cir-
rhosis [15]). Secondly, overall mortality and AIDS-related
causes of death were decreasing, while malignancies, liver-
related morbidities, non-AIDS-related infections and car-
diovascular diseases became the major causes of death.
Many of these causes of death were associated with modi-
fiable risk factors which require increased attention in
primary and specialized care.
Acknowledgements
This study was financed in the framework of the Swiss HIV
Cohort Study, supported by the Swiss National Science
Foundation. The members of the Swiss HIV Cohort Study
are: J. Barth, M. Battegay, E. Bernasconi, J. Böni, H. C.
Bucher, C. Burton-Jeangros, A. Calmy, M. Cavassini, C.
Cellerai, M. Egger, L. Elzi, J. Fehr, J. Fellay, M. Flepp, P.
Francioli (President of the SHCS), H. Furrer (Chairman of the
Clinical and Laboratory Committee), C. A. Fux, M. Gor-
gievski, H. Günthard (Chairman of the Scientific Board), D.
Haerry (deputy of ‘Positive Council’), B. Hasse, H. H. Hirsch,
B. Hirschel, I. Hösli, C. Kahlert, L. Kaiser, O. Keiser, C. Kind,
T. Klimkait, H. Kovari, B. Ledergerber, G. Martinetti, B.
Martinez de Tejada, K. Metzner, N. Müller, D. Nadal, G.
Pantaleo, A. Rauch, S. Regenass, M. Rickenbach (Head of
Data Center), C. Rudin (Chairman of the Mother & Child
Substudy), P. Schmid, D. Schultze, F. Schöni-Affolter, J.
Schüpbach, R. Speck, P. Taffé, P. Tarr, A. Telenti, A. Trkola,
P. Vernazza, R. Weber and S. Yerly (Version, July 2011).
The Swiss National Cohort (SNC) is funded by the Swiss
National Science Foundation. The SNC was made possible
by the support of the Swiss Federal Statistical Office. We
thank Fabio Valeri for the probabilistic record linkage
between the SHCS and the SNC.
Conflicts of interest: RW has received travel grants or
unrestricted educational grants from Abbott, Boehringer
Ingelheim, Bristol-Myers Squibb, Gilead Sciences, Glaxo-
SmithKline, Merck Sharp & Dome, Pfizer, LaRoche, TRB
10 R Weber et al.
© 2012 British HIV Association HIV Medicine (2012)
Page 10
Chemedica and Tibotec; and was a member of an endpoint
adjudication panel of phase II and III antiretroviral treat-
ment studies of Tibotec. HF has participated in advisory
boards of ViiVHealthcare, Bristol-Myers Squibb, Gilead,
Merck Sharp & Dome, Boehringer-Ingelheim and Janssen.
HF’s institution has received unrestricted educational
grants from Abbott, ViiV Healthcare, BMS, Roche, Gilead,
Merck Sharp & Dome, Boehringer-Ingelheim and Janssen-
Cilag. MB has received travel grants, speaker’s honoraria
and/or research grants from Abbott, Boehringer Ingelheim,
Gilead Sciences, Merck Sharp & Dome, Tibotec and ViiV.
MC received travel grants from Abbott, Boehringer-
Ingelheim, Gilead and Merck Sharp & Dome. EB has
participated in advisory boards of Abbott, Boehringer-
Ingelheim, Gilead, Merck Sharp & Dome, Janssen, and ViiV
Healthcare. EB’s institution has received unrestricted edu-
cation grants from Abbott, Gilead and ViiV Healthcare. MF
has participated in advisory boards and received unre-
stricted educational or travel grants from Abbott, Boe-
hringer Ingelheim, Bristol-Myers Squibb, Gilead Sciences,
GlaxoSmithKline, Janssen, Merck Sharp & Dome, Pfizer,
Roche, TRB Chemedica and ViiV Healthcare. BL has
received travel grants, grants or honoraria from Abbott,
Aventis, Bristol-Myers Squibb, Gilead, GlaxoSmithKline,
Merck Sharp & Dohme, Roche and Tibotec. MRu, MRi, AS,
PS, and AS have no conflicts of interest.
Authors’ contributions: RW and BL had full access to all
the data of the study and take responsibility for the integ-
rity of the data and the accuracy of the data analyses. RW,
MRu and BL designed the study. MRu completed the data
set. MRu and BL analysed the SHCS data. AS and BL linked
data of the SHCS and SNC cohorts, and analysed the data.
RW, MRu and BL wrote the first draft of the manuscript. All
investigators contributed to data collection and interpreta-
tion of the data, reviewed drafts of the manuscript, and
approved the final manuscript.
Financial disclosure: This study has been financed in the
framework of the Swiss HIV Cohort Study, supported by the
Swiss National Science Foundation. The funding source
had no influence on the design or conduct of the study.
References
1 Mocroft A, Vella S, Benfield TL et al. Changing patterns of
mortality across Europe in patients infected with HIV-1.
EuroSIDA Study Group. Lancet 1998; 352: 1725–1730.
2 Mocroft A, Brettle R, Kirk O et al. Changes in the cause
of death among HIV positive subjects across Europe: results
from the EuroSIDA study. AIDS 2002; 16: 1663–1671.
3 Krentz HB, Kliewer G, Gill MJ. Changing mortality rates and
causes of death for HIV-infected individuals living in
Southern Alberta, Canada from 1984 to 2003. HIV Med
2005; 6: 99–106.
4 Lewden C, Salmon D, Morlat P et al. Causes of death among
human immunodeficiency virus (HIV)-infected adults in the
era of potent antiretroviral therapy: emerging role of
hepatitis and cancers, persistent role of AIDS. Int J
Epidemiol 2005; 34: 121–130.
5 Weber R, Sabin CA, Friis-Moller N et al. Liver-related deaths
in persons infected with the human immunodeficiency virus:
the D:A:D study. Arch Intern Med 2006; 166: 1632–1641.
6 Palella FJ, Jr, Baker RK, Moorman AC et al. Mortality in the
highly active antiretroviral therapy era: changing causes of
death and disease in the HIV outpatient study. J Acquir
Immune Defic Syndr 2006; 43: 27–34.
7 Monforte A, Abrams D, Pradier C et al. HIV-induced
immunodeficiency and mortality from AIDS-defining and
non-AIDS-defining malignancies. AIDS 2008; 22:
2143–2153.
8 Lewden C, May T, Rosenthal E et al. Changes in causes of
death among adults infected by HIV between 2000 and
2005: the ‘Mortalite 2000 and 2005’ surveys (ANRS EN19
and Mortavic). J Acquir Immune Defic Syndr 2008; 48:
590–598.
9 Smith C, Sabin CA, Lundgren JD et al. Factors associated
with specific causes of death amongst HIV-positive
individuals in the D:A:D Study. AIDS 2010; 24: 1537–1548.
10 Antiretroviral Therapy Cohort Collaboration. Causes of death
in HIV-1-infected patients treated with antiretroviral therapy,
1996-2006: collaborative analysis of 13 HIV cohort studies.
Clin Infect Dis 2010; 50: 1387–1396.
11 Hasse B, Ledergerber B, Furrer H et al. Morbidity and Aging
in HIV-Infected Persons: the Swiss HIV Cohort Study. Clin
Infect Dis 2011; 53: 1130–1139.
12 Jaggy C, von Overbeck J, Ledergerber B et al. Mortality in
the Swiss HIV Cohort Study (SHCS) and the Swiss general
population. Lancet 2003; 362: 877–878.
13 Keiser O, Taffe P, Zwahlen M et al. All cause mortality in
the Swiss HIV Cohort Study from 1990 to 2001 in
comparison with the Swiss population. AIDS 2004; 18:
1835–1843.
14 van Sighem A, Danner S, Ghani AC, Gras L. Anderson RM
and de Wolf F. Mortality in patients with successful initial
response to highly active antiretroviral therapy is still higher
than in non-HIV-infected individuals. J Acquir Immune
Defic Syndr 2005; 40: 212–218.
15 Kovari H, Ledergerber B, Peter U et al. Association of
noncirrhotic portal hypertension in HIV-infected persons and
antiretroviral therapy with didanosine: a nested case-control
study. Clin Infect Dis 2009; 49: 626–635.
16 Kowalska JD, Friis-Moller N, Kirk O et al. The Coding Causes
of Death in HIV (CoDe) Project: initial results and evaluation
of methodology. Epidemiology 2011; 22: 516–523.
Causes of death and HIV infection 11
© 2012 British HIV Association HIV Medicine (2012)
Page 11
17 Hernando V, Sobrino-Vegas P, Burriel M et al. Differences in
the cause of death in HIV-infected patients in the Spanish
AIDS Research Cohort (CoRIS) according to data sources and
coding algorithms. 13th European AIDS Conference.
Belgrade, Serbia, 2011 [Abstract PS11/13].
18 CoDe Working Group. Coding Causes of Death in HIV
Protocol Version 1.0. CoDe Website. Available at
www.cphiv.dk/CoDe/Documents/tabid/101/Default.aspx
(accessed 20 October 2011).
19 Bopp M, Spoerri A, Zwahlen M et al. Cohort Profile: the
Swiss National Cohort–a longitudinal study of 6.8 million
people. Int J Epidemiol 2009; 38: 379–384.
20 Schoeni-Affolter F, Ledergerber B, Rickenbach M et al.
Cohort profile: the Swiss HIV Cohort study. Int J Epidemiol
2010; 39: 1179–1189.
21 D:A:D Study Group. Data Collection on Adverse Events of
Anti-HIV Drugs (D:A:D): manual of Operations, Version 1.3
(February 2005). Available at www.cphiv.dk/DAD/tabid/57/
Default.aspx (accessed 20 October 2011).
22 Centers for Disease Control (CDC). 1993 revised classification
system for HIV infection and expanded surveillance case
definition for AIDS among adolescents and adults. MMWR
Recomm Rep 1992; 41: 1–19.
23 Centers for Disease Control (CDC). Classification system for
human T-lymphotropic virus type III/lymphadenopathy-
associated virus infections. MMWR Morb Mortal Wkly Rep
1986; 35: 334–339.
24 World Health Organization. International Statistical
Classification of Diseases and Related Health Problems,
10th Revision, Version 2010. Available at
http://apps.who.int/classifications/icd10/browse/2010/en
(accessed 20 October 2011).
25 Spoerri A, Zwahlen M, Egger M, Bopp M. The Swiss
National Cohort: a unique database for national and
international researchers. Int J Public Health 2010; 55:
239–242.
26 The Swiss Federal Statistical Office. Mortality, causes of
death: Data, indicators. Available at www.bfs.admin.ch/bfs/
portal/en/index/themen/14/02/04/key/01.html (accessed 30
August 2012).
27 Fair M. Generalized Record Linkage System Statistics
Canada’s Record Linkage Software. Aust J Stat 2004; 33:
37–53.
28 Cuzick J. A Wilcoxon-type test for trend. Stat Med 1985;
4: 87–90.
29 Bonnet F, Morlat P, Chene G et al. Causes of death among
HIV-infected patients in the era of highly active
antiretroviral therapy, Bordeaux, France, 1998-1999. HIV
Med 2002; 3: 195–199.
30 Mocroft A, Gatell J, Reiss P et al. Causes of death in HIV
infection: the key determinant to define the clinical response
to anti-HIV therapy. AIDS 2004; 18: 2333–2337.
31 Sackoff JE, Hanna DB, Pfeiffer MR, Torian LV. Causes of
death among persons with AIDS in the era of highly active
antiretroviral therapy: New York City. Ann Intern Med 2006;
145: 397–406.
32 Lifson AR, Belloso WH, Carey C et al. Determination of the
underlying cause of death in three multicenter international
HIV clinical trials. HIV Clin Trials 2008; 9: 177–185.
33 Hessamfar-Bonarek M, Morlat P, Salmon D et al. Causes of
death in HIV-infected women: persistent role of AIDS. The
‘Mortalite 2000 & 2005’ Surveys (ANRS EN19). Int J
Epidemiol 2010;
39: 135–146.
34 Murray M, Hogg R, Lima V et al. The effect of injecting drug
use history on disease progression and death among
HIV-positive individuals initiating combination antiretroviral
therapy: collaborative cohort analysis. HIV Med 2012; 13:
89–97.
35 Salmon-Ceron D, Rosenthal E, Lewden C et al. Emerging role
of hepatocellular carcinoma among liver-related causes of
deaths in HIV-infected patients: the French national
Mortalite 2005 study. J Hepatol 2009; 50: 736–745.
36 Rubin MS, Colen CG, Link BG. Examination of inequalities
in HIV/AIDS mortality in the United States from a
fundamental cause perspective. Am J Public Health 2010;
100: 1053–1059.
37 Blair JM, McNaghten AD, Frazier EL, Skarbinski J, Huang
P, Heffelfinger JD. Clinical and behavioral characteristics of
adults receiving medical care for HIV infection Medical
Monitoring Project, United States, 2007. MMWR Surveill
Summ 2011; 60: 1–20.
38 Palella FJ Jr, Baker RK, Buchacz K et al. Increased mortality
among publicly insured participants in the HIV Outpatient
Study despite HAART treatment. AIDS 2011; 25: 1865–
1876.
39 Roulson J, Benbow EW, Hasleton PS. Discrepancies between
clinical and autopsy diagnosis and the value of post mortem
histology; a meta-analysis and review. Histopathology 2005;
47: 551–559.
40 Pastores SM, Dulu A, Voigt L, Raoof N, Alicea M, Halpern
NA. Premortem clinical diagnoses and postmortem autopsy
findings: discrepancies in critically ill cancer patients. Crit
Care 2007; 11 : R48.
41 Cox JA, Lukande RL, Lucas S, Nelson AM, Van Marck E,
Colebunders R. Autopsy causes of death in HIV-positive
individuals in sub-Saharan Africa and correlation with
clinical diagnoses. AIDS Rev 2010; 12: 183–194.
42 Justice AC. Commentary: treated HIV infection is a chronic
disease: the case against cause of death analyses. Int J
Epidemiol 2010; 39: 146–148.
43 Johansson LA, Westerling R. Comparing hospital discharge
records with death certificates: can the differences be
explained? J Epidemiol Community Health 2002; 56:
301–308.
12 R Weber et al.
© 2012 British HIV Association HIV Medicine (2012)
Page 12
44 Villar J, Perez-Mendez L. Evaluating an educational
intervention to improve the accuracy of death certification
among trainees from various specialties. BMC Health Serv
Res 2007; 7: 183.
45 Minder C, Zingg W. Die Sterblichkeitsstatistik in der Schweiz.
Amliche Statistik der Schweiz No. 155. Bern: Bundesamt für
Statistik; 1989.
46 Lahti RA, Penttila A. Cause-of-death query in validation of
death certification by expert panel; effects on mortality
statistics in Finland, 1995. Forensic Sci Int 2003; 131:
113–124.
47 Jensen HH, Godtfredsen NS, Lange P, Vestbo J. Potential
misclassification of causes of death from COPD. Eur Respir
J 2006; 28: 781–785.
Appendix
Table A1 Coding Causes of Death in HIV (CoDe) classification of causes
of death in HIV-positive persons [18]
01 AIDS (ongoing active disease)
01.1 Infection
01.2 Malignancy
02 Infection (other than 01.1)
02.1 Bacterial
02.1.1 Bacterial with sepsis
02.2 Others
02.2.1 Other with sepsis
02.3 Unknown aetiology
02.3.1 Unknown with sepsis
03 Chronic viral hepatitis (progression of/complication to)
03.1 HCV
03.1.1 HCV with cirrhosis
03.1.2 HCV with liver failure
03.1.3 HCV with hepatocellular carcinoma
03.2 HBV
03.2.1 HBV with cirrhosis
03.2.2 HBV with liver failure
03.2.3 HBV with hepatocellular carcinoma
04 Malignancy (other than 01.2 and 03, 03.1, 03.2)
05 Diabetes mellitus (complication to)
06 Pancreatitis
07 Lactic acidosis
08 Myocardial infarction or other ischaemic heart disease
09 Stroke
10 Gastrointestinal haemorrhage (if chosen, specify underlying cause)
11 Primary pulmonary hypertension
12 Lung embolus
13 Chronic obstructive lung disease
14 Liver failure (other than 03, 03.1, 03.2)
15 Renal failure
16 Accident or other violent death (not suicide)
17 Suicide
18 Euthanasia
19 Substance abuse (active)
19.1 Chronic alcohol abuse
19.2 Chronic intravenous drug use
19.3 Acute intoxication (indicate agent)
20 Haematological disease (other causes)
21 Endocrine disease (other causes)
22 Psychiatric disease (other causes)
23 CNS disease (other causes)
24 Heart or vascular (other causes)
25 Respiratory disease (other causes)
26 Digestive system disease (other causes)
27 Skin and motor system disease (other causes)
28 Urogenital disease (other causes)
29 Obstetric complications
30 Congenital disorders
90 Other causes (provide details)
91 Unclassifiable causes
92 Unknown
Table A2 Comparison of Coding Causes of Death in HIV (CoDe) and
International Statistical Classification of Diseases and Related Health
Problems (ICD-10) codes
CoDe code ICD-10 code
01 (AIDS) A021, A072-073, A15-A19, A31, A812,
B027, B20-B24, B25, B371,
B383-B389, B393-B399,
B451-B459, B582, C46, C53, C82,
C83, C85, C859
02 (Infection other than
AIDS-related)
A00-A020, A022-A071, A078-A09,
A20-A309, A32-A812, A818-A99,
B0-B09, B26-B370, B372-B382,
B39-B392, B40-B450, B46-B581,
B583-B941, B948-B99, G00-G02,
J01-J22
03 (Chronic viral hepatitis) B15-B19, B942
14 (Liver failure) K70-K77
04 (Malignacy other than 01.2) C00-C45, C47-C52, C54-C81, C84,
C88-D09
05 (Diabetes mellitus) E10-E149
21 (Endocrine disease, other) E01-E079, E20-E35
06 (Pancreatitis) K85-K861
07 (Lactic acidosis) E872
08 (Ischaemic heart disease) I21-I24
09 (Stroke) I61, I63-i64
24 (Heart or vascular disease) other I
10 (Gastrointestinal haemorrhage) K920-K922
26 (Digestive system, other) other K
11 (Primary pulmonary hypertension) I270
12 (Lung embolous) I26
13 (Chronic obstructive lung disease) J44
25 (Respiratory disease) J40-J99
15 (Renal failure) N17-N19
28 (Urogenital, other) other N
16 (Accident or violent death) V01-X59
17 (Suicide) X60-X84
18 (Euthanasia)
19 (Substance abuse) F10-F19
20 (Haematological disease) D50-D77
22 (Psychiatric, other) F other than F10-19
23 (CNS disease) G048-G99
90 (Other causes, provide details)
91 (Unclassifiable cause) R092, R96-R99
92 (Unknown)
Causes of death and HIV infection 13
© 2012 British HIV Association HIV Medicine (2012)
Page 13
  • Source
    • "Antiretroviral therapy (ART) has resulted in a remarkable decline in acquired immunodeficiency syndrome (AIDS)-related death among HIV-infected individuals worldwide123456. As prognosis has improved, reports from resource-rich countries have shown that the causes of death in HIV-infected individuals have changed, with cancers or cardiovascular diseases or liver-related diseases becoming the leading causes of mortality.78910. Although a detailed understanding of causes of death and associated risk factors is crucial to the appropriate management of HIV-related diseases and co-morbidities, the specific causes of death have not been well described in resource-limited settings. "
    [Show abstract] [Hide abstract] ABSTRACT: Although the prognosis for HIV-infected individuals has improved after antiretroviral therapy (ART) scale-up, limited data exist on the incidence of AIDS-defining opportunistic infections (ADIs) and mortality during ART in resource-limited settings.
    Full-text · Article · Mar 2016 · PLoS ONE
  • Source
    • "Subjects with HIV infection were well matched with controls for age, sex, and comorbidities (Table 1); although in keeping with previous reports, 19 there were more smokers in subjects with HIV infection (40% versus 12% of controls; P<0.005). Subjects with HIV infection had higher median serum levels of high-sensitivity C-reactive protein (323456 versus 101 mg/L in controls; P<0.005). Of subjects with HIV infection receiving ART, 68% were on non-nucleoside reversetranscription inhibitor–based regimen and 32% were on a protease inhibitor–based regimen.Table 2). "
    [Show abstract] [Hide abstract] ABSTRACT: Background-Patients with treated HIV infection have clear survival benefits although with increased cardiac morbidity and mortality. Mechanisms of heart disease may be partly related to untreated chronic inflammation. Cardiovascular magnetic resonance imaging allows a comprehensive assessment of myocardial structure, function, and tissue characterization. We investigated, using cardiovascular magnetic resonance, subclinical inflammation and myocardial disease in asymptomatic HIV-infected individuals. Methods and Results-Myocardial structure and function were assessed using cardiovascular magnetic resonance at 1.5-T in treated HIV-infected individuals without known cardiovascular disease (n=103; mean age, 45±10 years) compared with healthy controls (n=92; mean age, 44±10 years). Assessments included left ventricular volumes, ejection fraction, strain, regional systolic, diastolic function, native T1 mapping, edema, and gadolinium enhancement. Compared with controls, subjects with HIV infection had 6% lower left ventricular ejection fraction (P<0.001), 7% higher myocardial mass (P=0.02), 29% lower peak diastolic strain rate (P<0.001), 4% higher short-tau inversion recovery values (P=0.02), and higher native T1 values (969 versus 956 ms in controls; P=0.01). Pericardial effusions and myocardial fibrosis were 3 and 4× more common, respectively, in subjects with HIV infection (both P<0.001). Conclusions-Treated HIV infection is associated with changes in myocardial structure and function in addition to higher rates of subclinical myocardial edema and fibrosis and frequent pericardial effusions. Chronic systemic inflammation in HIV, which involves the myocardium and pericardium, may explain the high rate of myocardial fibrosis and increased cardiac dysfunction in people living with HIV.
    Full-text · Article · Mar 2016 · Circulation Cardiovascular Imaging
  • Source
    • "Adverse events related to agents such as clopidogrel have also been well documented [36]. In summary, hypothetical reasons for worse short-term outcomes in HIV+ patients include less frequently use required cathterization procedures [10], ongoing HIV replication and inflammation [5], drug-drug interactions [29], and asymptomatic coronary disease at younger ages [31] leading to a late diagnosis of CAD. Our study has several limitations. "
    [Show abstract] [Hide abstract] ABSTRACT: HIV infection may be associated with an increased recurrence rate of myocardial infarction. Our aim was to determine whether HIV infection is a risk factor for worse outcomes in patients with coronaray artery disease. We compared data aggregated from two ongoing cohorts: (i) the Acute Myocardial Infarction in Switzerland (AMIS) registry, which includes patients with acute myocardial infarction (AMI), and (ii) the Swiss HIV Cohort Study (SHCS), a prospective registry of HIV-positive (HIV+) patients. We included all patients who survived an incident AMI occurring on or after 1st January 2005. Our primary outcome measure was all-cause mortality at one year; secondary outcomes included AMI recurrence and cardiovascular-related hospitalisations. Comparisons used Cox and logistic regression analyses, respectively. There were 133 HIV+, (SHCS) and 5,328 HIV-negative [HIV-] (AMIS) individuals with incident AMI. In the SHCS and AMIS registries, patients were predominantly male (72% and 85% male, respectively), with a median age of 51 years (interquartile range [IQR] 46-57) and 64 years (IQR 55-74), respectively. Nearly all (90%) of HIV+ individuals were on successful antiretroviral therapy. During the first year of follow-up, 5 (3.6%) HIV+ and 135 (2.5%) HIV- individuals died. At one year, HIV+ status after adjustment for age, sex, calendar year of AMI, smoking status, hypertension and diabetes was associated with a higher risk of death (HR 4.42, 95% CI 1.73-11.27). There were no significant differences in recurrent AMIs (4 [3.0%] HIV+ and 146 [3.0%] HIV- individuals, OR 1.16, 95% CI 0.41-3.27) or in hospitalization rates (OR 0.68 [95% CI 0.42-1.11]). HIV infection was associated with a significantly increased risk of all-cause mortality one year after incident AMI.
    Full-text · Article · Dec 2015 · AIDS Research and Therapy
Show more