Path: utzoo!censor!geac!torsqnt!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!usc!ucla-cs!news
From: ST602662%BROWNVM.BITNET@oac.ucla.edu (Bill Jesdale)
Newsgroups: sci.med.aids
Subject: Re: (2775) a question
Message-ID: <1990Dec2.232631.3753@cs.ucla.edu>
Date: 2 Dec 90 18:52:48 GMT
Sender: news@cs.ucla.edu (Mr. News)
Organization: UCLA, Computer Science Department
Lines: 61
Approved: phil@wubios.wustl.edu
Note: non-commercial reproduction.
Nntp-Posting-Host: squid.cs.ucla.edu
Archive-Number: 2777

In answer to the question of how dangerous are various exposures
to bodily fluids of HIV infected persons:

First, we do not know.
Some estimates are given by two recent articles in the Annals
of Internal Medecine (15 Nov 90).

Article 1) Infection with HIV-1 among Recipients of Antibody
           Positive Blood Donations.
  This study found 133 recipients of blood which was later found to
test positive for antibodies to HIV. After excluding 9 with other
high risks, they found that 111 out of 124 blood product recipients
without other high risks themselves tested positive to antibodies.
  So, for a direct innoculation of a large amount of HIV+
blood directly into the blood stream, they estimate that the risk
of subsequent infection is around 90%.
  With the advent of blood screening on top of self-deferral of people
deemed high-risk (men who have sex with men, Central Africans and
Hatians, along with some people who are in fact at high risk),
the chances of a blood product carrying HIV is much diminished.
  For those HIV- people who recieve one of these blood products,
the chances of infection are still as high as ever.

Article 2) Risk for Occupational Transmission of HIV-1 Associated
           with Clinical Exposures.
  This study was done on NIH employees, who reported a fairly large
number of accidental exposures to blood and other body fluids from
people known to be infected with HIV.
  When they pooled their data with 13 similar studies, they found
that for needlesticks with blood (HIV+ blood in and/or on a needle
which was accidentally stuck into the worker) 6 out of 2042 such
incidents probably led to the worker becoming infected.
  So, for an innoculation of HIV+ blood directly into the body, but
not necessarily into the bloodstream, they estimate the odds of
transmission to be 0.29%, per needlestick (95%CI 0.13 to 0.70).
  When they pooled their data with data from 11 similar studies
which looked at exposures of known HIV+ blood onto mucous
membranes, such as into the mouth or eyes, they found 0 out of 1051
exposures led to infection. (95%CI 0 to 0.28%)
  This study also looked at exposure body fluids from known HIV+
patients onto unbroken skin, and found 0 seroconversions out of
5,568 exposures. 2,712 of these exposures were to blood, 912 to
urine, 804 to sputum, 300 to feces, and 840 to other fluids.

  This evidence corroborates epidemiological evidence that the
risk of transmission through needlesticks is worrisomely large,
although still small when compared to other more efficient
transmission routes, and that the risk of transmission without
a significant exposure directly into the bloodstream is extremely
low. Please correct me if I'm wrong, but I don't think there
has yet been confirmed a case of transmission through mucous
membranes or unbroken skin.

I hope this helps you to understand the risks associated with
"casual" exposure to body fluids from an HIV infected person.
It is important to remember that these estimates are just that,
estimates, and that saying "The risk is X" is an unproductive
exercise in uncertainty.

Bill Jesdale, ST602662@BROWNVM             I need a job.
Box 5342, Brown U, Prov RI 02912. Ph: 401/863.4402