No patient had AIDS. The rest of the data therefore refer to HIV
positive and HIV negative patients only. Of the 402 patients admitted
to the unit during the six months, 52 (13%) tested positive for HIV.
Though the male to female distribution in the two groups was similar,
they differed significantly in age (P<0.002; table 1).
| Table 1 - Age and sex distribution of HIV negative
and
HIV positive patients |
|
HIV negative (n=350) |
HIV
positive (n=52) |
| No (%)
male | 228 (65) | 39 (75)
|
| No (%) female | 122 (35) | 13 (25)
|
| Mean age in years (SD) | 33 (18) | 28
(9)* |
| *P=0.0018. |
Most patients in both groups were admitted after trauma (table 2).
| Table 2 - Interdisciplinary distribution of patients |
| Discipline |
No (%) HIV negative |
No (%)
HIV positive |
Overall % HIV positive in discipline
|
| Trauma | 188
(54) | 30 (57) | 14 |
| Obstetrics and
gynaecology | 40 (11) | 8
(15) | 17 |
| Paediatrics | 30
(9) | 2 (4) | 6 |
| Vascular
surgery | 27 (8) | 3 (6) | 10
|
| General surgery | 24 (7) | 4
(8) | 14 |
| Internal medicine | 17
(5) | 3 (6) | 15 |
| Ear, nose, and
throat/maxillofacial | 11
(3) | 0 | 0 |
| Orthopaedic
surgery | 9 (3) | 2 (4) | 18
|
| Urology | 4 (1) | 0 | 0
|
| Total | 350 (100) | 52
(100) | 13 |
HIV
infection was more common in patients referred from orthopaedic surgery
and obstetrics and gynaecology. There was no significant difference in
intensive care unit or hospital mortality or in the duration of
intensive care unit or hospital stay (tables 3 and 4).
| Table 3 - Comparison of mortality between HIV
negative
and HIV positive patients |
|
No (%) HIV negative |
No (%) HIV
positive |
P value |
Odds
ratio |
Age adjusted odds ratio (95% confidence interval)
|
| Intensive care unit | 84/350
(24) | 15/52
(29) | 0.558 | 1.28 | 1.45 (0.75 to
2.80) |
| Hospital | 16/247 (6) | 1/37
(3) | 0.308 | 0.16 | † |
| †Data on hospital mortality for HIV negative patients were
available for only 247 patients, therefore maximum likelihood ratios
could not be calculated. |
| Table 4 - Mean and median (range) number of days'
stay
in intensive care unit and hospital for HIV positive and HIV negative
patients |
| HIV
negative | HIV
positive |
|
|---|
| Mean | Median | Mean | Median | P
value |
| Intensive care
unit | 6 | 4
(1-44) | 7 | 5
(1-41) | 0.1
|
| Hospital | 10 | 7
(1-50) | 8 | 6
(2-42) | 0.08 |
| Data were available for only 247 HIV negative patients. |
Intensive care
unit mortality in HIV negative patients was 24% (84/350) compared with
29% (15/52) in HIV positive patients (odds ratio 1.45; 95% confidence
interval 0.75 to 2.80); hospital mortality was 6% (16/247) and 3%
(1/37) in the two groups respectively. There was no significant
difference in survival distribution between the groups (mean survival
time 7.3 (SE 2.6) days in the HIV positive group, 6.9 (0.78) days in
the HIV negative group; P=0.88) (fig 1). Information regarding
hospital mortality and duration of hospital stay could not be retrieved
for 19 HIV negative patients. There was no significant difference in
mean APACHE II score between HIV negative and HIV positive patients
(scores 9 and 8 respectively).
Fig 1 - Survival distribution function versus duration
of stay in intensive care unit.
Organ failure was more prevalent in HIV positive patients.
Significant
differences were found when cardiac, respiratory, and haematological
system failures were compared (table 5).
| Table 5 - Comparison of total and individual organ
failures between HIV negative and HIV positive patients |
|
No (%) HIV negative (n=350) |
No
(%) HIV positive (n=52) |
P value |
Odds
ratio |
Age adjusted odds ratio (95% confidence
interval) |
| Total | 171
(49) | 37 (71) | <
0.003 | 2.58 | 2.87 (1.51 to 5.46)
|
| Cardiac | 84 (24) | 21
(40) | <0.014 | 2.11 | 2.38 (1.28
to 4.42) |
| Respiratory | 150 (43) | 33
(63) | <0.005 | 2.32 | 2.60 (1.41
to 4.78) |
| Haematological | 31 (9) | 11
(21) | <0.007 | 2.76 | 3.22 (1.47
to 7.09) |
| Renal | 47 (13) | 9
(17) | 0.45 | 1.35 | 1.75
(0.78 to 3.92) |
| Neurological | 19 (5) | 5
(10) | 0.23 | 1.85 | 1.72
(0.61 to 4.85) |
Though there was no difference
in the incidence of severe sepsis and nosocomial sepsis, septic shock
was significantly more common in HIV positive patients (table
6).
| Table 6 - Incidence of sepsis in HIV negative and HIV
positive patients |
|
No (%) HIV negative(n=350) |
No
(%) HIV positive(n=52) |
P
value |
Odds ratio |
Age adjusted odds ratio
(95% confidence interval) |
| Septic
shock | 54 (15) | 20
(38) | <0.001 | 3.43 | 3.64 (1.91
to 6.89) |
| Severe sepsis | 71 (20) | 9
(17) | 0.62 | 0.82 | 0.84
(0.39 to 1.81) |
| Nosocomial sepsis | 83
(24) | 13
(25) | 0.84 | 1.07 | 1.16
(0.59 to 2.31) |
It was not possible to perform flow cytometry on all HIV positive
patients. This was because either the patient died soon after admission
or the request for testing came after the patient was discharged from
the unit and could not be reached. Compared with normal values there
were significant differences in T4 count, T4:T8 ratio, and B4 count
(table 7). The T4:T8 ratio was reversed and B4 count reduced in HIV
positive patients.
| Table 7 - Flow cytometry results in HIV positive
patients. Cell counts are means (SE) |
| Cell count (x
106/l) |
|
|---|
| HIV
positive patients (n=24) | Normal
| P value |
| T3 | 949
(86) | 800-2800 | g 0.05
|
| T4 | 425
(41) | 550-1955 | 0.011
|
| T8 | 549 (74) | 250-1200 | g
0.05 |
| T4:T8 ratio | 1.2
(0.2) | 2.0 | <0.001
|
| B4 | 186
(18) | 245-850 | 0.001
|
| NKH-1 | 170 (24) | 25-360 | g
0.05 |
Accidental disclosure of HIV status occurred in one instance as a
result of a laboratory error. The researcher who became aware of the
result did not divulge it to other staff and did not participate in
management decisions regarding the patient. Data for this patient were
collated by another member of the team, who was unaware of the result.
As permitted by the protocol, the HIV status of five other patients
became known to relevant staff members before patient discharge. One
case involved a needlestick injury to a staff member, and the remaining
disclosures were in preparation for haemodialysis. In all instances
management decisions and collation of patient data were by other,
blinded researchers. Of all 402 patients tested, only three wished to
be informed of their HIV status. No patient objected to having been
included in the study without prior informed
consent.
Mortality is the best measure of outcome of patients treated in an
intensive care unit. Markers of morbidity may be subjective and are
therefore less reliable end points. In this study there was no
difference in intensive care unit or hospital mortality between HIV
positive and HIV negative patients when results were adjusted for age
(table 3).
HIV positive patients were more prone to septic shock and organ
failure, and we were therefore surprised that the duration of hospital
stay and mortality were not increased. This finding is unlikely to have
been the result of observer error because, except for one inadvertent
disclosure, HIV results were not available until all other data were
collated. Abnormalities in flow cytometry results may have been an
important factor. HIV positive patients had low T4 counts whereas T8
counts were comparatively high. As a consequence the T4:T8 ratio was
reduced but the total (T3) was not affected. The B4 count was also low.
All these features are consistent with the latent phase of HIV
infection.(9) Though the behaviour of these cell
populations
predicts clinical progression of HIV disease to AIDS,(10) to
our knowledge its impact on intensive care unit patients admitted for
non-HIV related disease has not been described. Conceivably the immune
response to major trauma and sepsis is altered. The observation by
Munoz et al that HIV negative patients with sepsis and
impaired macrophage responsiveness are more prone to subsequent
sepsis(11) lends credence.
Immunological mechanisms have been postulated to play a major part in
the pathogenesis of septic shock and multiple organ
failure.(12-13) The immune response is complex
and
paradoxical, pro-inflammatory and anti-inflammatory responses occurring
simultaneously and both being mediated by cytokines.(14)
This has prompted the use of new drugs which alter the immune response
in sepsis.(15-16) We therefore postulate that,
though HIV
positive patients have disturbances in immune function which make them
more susceptible to septic shock and multiple organ failure, the
inflammatory response is also altered such that there is no increase in
mortality.
The patients in this study were young, predominantly male, and
admitted
primarily after trauma or surgery. That no patient had AIDS concurs
with Gilks's observation that the pattern of HIV infection in Africa
differs from that in the developed world.(3) Non-
HIV disease
is far more prevalent in Africa, with rapid progression from
seroconversion to HIV to death from an AIDS defining
condition.(3) Data relating to outcome in patients
with AIDS
cannot therefore be extrapolated to our patients. This emphasises the
importance of describing the outcome in patients admitted to intensive
care with non-HIV related disease.
Issue of informed consent
Decisions on initiating and terminating care for critically ill
patients are difficult.(17) The unique nature of
the AIDS
epidemic in Africa,(3) the tremendous costs
associated with
advanced life support,(1) as well as the
particular ethical
considerations in patients with HIV infection(18)
are
compelling reasons for these decisions to be based on sound ethical
principles and objective evidence of disease outcome. In view of the
lack of clinical information in our patient population the acquisition
of objective data was imperative. A major ethical dilemma arose when
the decision was made not to seek informed consent. This was thought to
be essential, as patients who were likely to be at risk for HIV
infection would also be inclined to refuse the study, which would
seriously limit its value.
There were two consequences of the study. Firstly, patients were
denied
the option of being excluded and, secondly, they were at risk of having
their HIV status disclosed. The first consideration was evaluated in
terms of the potential benefit of the study to society as a whole. The
consensus of the research team and the ethics committee was that the
clinical implications of the study were enough to warrant denying
patients the right of refusal. With respect to the second consequence,
every effort was made in the design and execution of the study to
ensure that indiscriminate disclosure of HIV results did not occur. To
our knowledge HIV results were not disclosed except for study purposes
and, furthermore, patient care was not influenced by HIV status.
There was no reason to suspect that the racial background of our
patients would have any bearing on their outcome. Race as a demographic
variable is considered only rarely in South Africa.(19) Our
main criterion for denying patients admission to intensive care is
futility. This study showed no significant difference in mortality
between HIV positive and HIV negative patients. Though the incidence
rates of septic shock and organ failure were higher, this did not
influence mortality or duration of stay. We therefore conclude that in
our patient population HIV status cannot be used as a criterion for
denying patients admission to the intensive care unit. Our observations
regarding septic shock and organ failure require further
evaluation.
| Key Messages |
| HIV positive patients admitted to intensive care for diseases
unrelated to their HIV status have a similar mortality and duration of
stay when compared with HIV seronegative patients
The incidence of septic shock and multiple organ dysfunction
is higher in HIV seropositive patients and needs further
investigation
HIV status cannot be used to deny critically ill patients
admission to intensive care
The HIV and AIDS epidemic raises unique ethical considerations
that must be carefully addressed during clinical studies |
Acknowledgments: Part of this study was presented at the 12th annual
critical care congress of the South African Critical Care Society
(1995) and at the eighth European congress of intensive care
medicine. We thank Mrs Q A Karim, Dr S S A Karim, Professor H M
Coovadia, Dr E M Barker, Professor D J Pudifin, Professor A N Smith,
Professor J Lipman, and the ethics committee for advice and Miss E
Gouws for statistical analysis. We also thank Mr T Doorasamy,
technicians in the department of virology, Mr H Benimadho, Mr N
Bhimsan, and Mr R Loykisoonlal for technical help and Mrs A Pillay for
secretarial work.
References
1 Cheng E Y. Aids patients in the intensive care
unit.
Curr Opin Anesthesiol 1993;6:309-14.
2 Rosen M J, De Palo V A. Outcome of intensive
care for patients
with AIDS. Crit Care Clin 1993;9:107-14.
3 Gilks C F. The clinical challenge of the HIV
epidemic in the
developing world. Lancet 1993;342:1037-9.
4 Luce J M. Conflict over ethical principles in
the intensive
care unit. Crit Care Med 1992;20:313-5.
5 1993 revised classification system for HIV
infection and
expanded surveillance case definition for AIDS among adolescents and
adults. MMWR 1992;41:1-19.
6 Knaus W A, Draper E A, Wagner D P, Zimmerman J E.
APACHE II: a
severity of disease classification system. Crit Care Med
1985;13:818-29.
7 Knaus W A, Draper E A, Wagner D P, Zimmerman J E.
Prognosis in
acute organ-system failure. Ann Surg 1985;202:685-93.
8 Members of the American College of Chest
Physicians/Society of
Critical Care Medicine Consensus Conference Committee. American College
of Chest Physicians/Society of Critical Care Medicine consensus
conference. Definitions for sepsis and organ failure and guidelines for
the use of innovative therapies in sepsis. Crit Care Med
1992;20:864-74.
9 Pantaleo G, Graziosi C, Fauci A S. The
immunopathogenesis of
human immunodeficiency virus infection. N Engl J Med
1993;328:327-34.
10 Stein D S, Korvick J A, Vermund S H. CD4+
lymphocyte
cell enumeration for prediction of clinical course of human
immunodeficiency virus disease: a review. J Infect Dis
1992;165:352-63.
11 Munoz C, Carlet J, Fitting C, Misset B,
Blériot J-P,
Cavaillon J-M. Dysregulation of in vitro cytokine production by
monocytes during sepsis. J Clin Invest 1991;88:1747-54.
12 Baue A E. The horror autotoxicus and
multiple-organ failure.
Arch Surg 1992;127:1451-62.
13 Goris R J, te Boekhorst T P, Nuytinck J K,
Gimbrère JS.
Multiple organ failure: generalised autodestructive inflammation?
Arch Surg 1985;120:1109-15.
14 Lin R Y, Astiz M E, Saxon J C, Saha D C, Rackow
E C. Relationships
between plasma cytokine concentrations and leukocyte functional antigen
expression in patients with sepsis. Crit Care Med
1994;22:1595-602.
15 Knaus W A, Harrell F E, Fisher C J Jr, Wagner
D P, Opal S M, Sadoff
J C, et al . The clinical evaluation of new drugs for
sepsis. A prospective study design based on survival analysis.
JAMA 1993;270:1233-41.
16 Dhainaut J-F, Tenaillon A, Le Tulzo Y,
Schlemmer B, Solet J P,
Wolff M, et al . Platelet-activating factor receptor
antagonist BN 52021 in the treatment of severe sepsis: a randomized,
double-blind, placebo-controlled, multicenter clinical trial.
Crit Care Med 1994;22:1720-8.
17 Ruark J E, Raffin T A. Initiating and
withdrawing life support.
Principles and practice in adult medicine. N Engl J Med
1988;318:25-30.
18 Brown J, Sprung C L. Ethical considerations
in the treatment of
AIDS patients in the intensive care unit. Crit Care Clin
1993;9:115-23.
19 Ellison G T H, De Wet T, Ijsselmuiden C B,
Richter L M.
Desegregating health statistics and health research in South Africa.
S Afr Med J 1996;86:1257-62.
Funding: The University of Natal's research and travel
committee and the Medical Research Council of South Africa.
Conflict of interest: None.
(Accepted 12 December 1996)
Faculty of Medicine,
University of Natal,
Private Bag 7,
Congella 4013,
South Africa
Satish
Bhagwanjee, lecturer in
anaesthetics
David J J Muckart, senior
lecturer in surgery
Prakash M
Jeena, lecturer in paediatrics
Prushini
Moodley, honorary lecturer in
haematology
Correspondence to: Dr
Bhagwanjee.
Commentary: Failing to seek patients' consent to research is
always wrong Rajendra Kale
Commentary: Why we did not seek informed consent before testing
patients for HIV
Satish Bhagwanjee, David J J Muckart, Prakash
Commentary: No simple and absolute ethical rule exists for every
conceivable situation
Y K Seedat
