Content uploaded by Keith Hawton
Author content
All content in this area was uploaded by Keith Hawton
Content may be subject to copyright.
10.1192/bjp.168.1.43Access the most recent version at doi:
1996 168: 43-48 The British Journal of Psychiatry
K Hawton, C Ware, H Mistry, J Hewitt, S Kingsbury, D Roberts and H Weitzel
reduction
Paracetamol self-poisoning. Characteristics, prevention and harm
References
http://bjp.rcpsych.org/cgi/content/abstract/168/1/43#otherarticles
Article cited in:
permissions
Reprints/
permissions@rcpsych.ac.ukto
To obtain reprints or permission to reproduce material from this paper, please write
to this article at
You can respond
http://bjp.rcpsych.org/cgi/eletter-submit/168/1/43
service
Email alerting
click heretop right corner of the article or
Receive free email alerts when new articles cite this article - sign up in the box at the
from
Downloaded
The Royal College of PsychiatristsPublished by
on July 13, 2011 bjp.rcpsych.org
http://bjp.rcpsych.org/subscriptions/
go to: The British Journal of PsychiatryTo subscribe to
British Journal of Psychiatry(1996), 168, 43-48
A steady increasein paracetamoloverdoses sincethe
mid-1970s has meant that paracetamol has now
become the substance used most frequently in
deliberate self-poisoning in the UK (Hawton &Fagg,
1992a). In Oxford, for example, 47.8°loof all over
doses in 1993involved paracetamol or paracetamol
containing drugs (unpublished data). It is especially
common in overdoses by adolescents (Hawton &
Fagg,1992b).Thisphenomenon isnotconfinedto
the UK, having also been reported from Denmark
(Ott et a!, 1990) and Germany (Augustin &
Schmoldt, 1991). Hepatic toxicity (Bray, 1993)is the
main reason for concern about paracetamol self
poisoning and why methods of prevention or harm
reduction need to be found. Paracetamol liver
damage accounts for over half of all cases referred
to liver units because of fulminant liver failure, is
the definite cause of at least 150 deaths per year in
the UK (Spooner & Harvey, 1993), and has now
become a common reason for liver transplantation
(O'Grady et a!, 1991).
While a study of paracetamol self-poisoners
admitted to a tertiary referral centre for patients with
liverproblemsinthe 1970srevealedthatmost patients
did not know that paracetamol was dangerous in
overdose (Gazzard eta!, 1976), we have found that
paracetamol self-poisoners today are mostly aware
of the dangers, and the main reason they take
paracetamol is its ready availability (Hawton et a!,
1995). We report on the characteristics of
the overdoses and their effects, and examine the
potential effectiveness of possible measures to reduce
the likelihood or dangers of paracetamol self
poisoning.
Method
The subjects were 80 consecutive self-poisoning
patients who were admitted to the John Radcliffe
Hospital in Oxford between September 1992 and
March 1993, having taken overdoses of paracetamol
or paracetamol-containingcompounds, or both, and
who consented to participate in the research
interview. The patients had first undergone a full
psychosocial assessment (Hawton & Catalan, 1987),
following which members of the research team
approached them to seek their permission for
inclusion in the study. Eighteen other patients refused
the research interview. A further 20 patients left
hospital before their consent could be obtained and
five were not invited to participate because they were
confused or too ill. There were no significant age or
sex differences between the included and excluded
patients.
43
Paracetamol Self-Poisoning
Characteristics, Prevention and Harm Reduction
KEITH HAWTON, CHRISTOPHER WARE, HAMANT MISTRY, JONATHAN HEWITT,
STEPHEN KINGSBURY, DAVE ROBERTSand HEATHER WEITZEL
Background.Paracetamolis now the most commondrugusedfor self-poisoningin the UK
andisassociatedwith potentiallyfatal liverdamage.Patientsadmittedto hospitalbecause
of paracetamoloverdoseswerestudiedinorderto determinetheircharacteristicsandfactors
which might have deterredthem from taking paracetamolor reducedthe dangersof the
overdose.
Method.Eightypatientswerestudiedin hospitalusinga structuredinterviewschedule,
measuresofdepressionandsuicidalintent,informationcollectedthroughtheOxfordMonitoring
Systemfor Attempted Suicide,and the resultsof liverfunctiontests.
Results.Acuteliverdysfunction(25 patients)was associatedwith consumptionofmorethan
25 tablets(oddsratio4.46,95% Cli .31 to 17.41, P=0.Oi 4). Theproportionateuseoftablets
fromblisterpacks(60%) andloosepreparations(46%; 5 patientsusingbothtypes)reflected
theirgeneralavailability.Moreof thosewhotooktabletsfroma loosepreparationconsumed
25 or more tablets (69%) than those who used a blister-packpreparation(40%; odds
ratio = 3.0,95% CI 1.12 to 9.95, P= 0.028). Only 20 patients thought that any type of warning
label would have deterred them from taking a paracetamol overdose.
ConclusIons. Establishing a maximum number of tablets (e.g. 25) that can be available in
individualpreparationsis likelyto reducethe dangersof paracetamolself-poisoning.The
potentialeffects of other measuresare uncertain.
44
HAWTON ET AL
There were eight researchinterviewers, all psy
chiatristsor membersof the psychiatricconsultation
service in the general hospital who were familiar with
interviewing self-poisoning patients.
The research questionnaire contained questions
regarding details of previous overdoses (including
whether paracetamol was used and treatment
received), and of the current overdose (including
substances used, extent of premeditation and alcohol
use before and with the overdose). Specific questions
about paracetamol use (including paracetamol
compounds) in the current overdose concerned
where the paracetamol was originally obtained,
where it was kept in the house, and the form
of paracetamol (loose tablets or blister-pack
preparation). Enquiry was then made about factors
which might have affected the patient's decision
to take an overdose of paracetamol. These included
full knowledge of its dangers, specific warnings
on the paracetamol container, blister packaging,
availability of smaller amounts, and being available
only on prescription. The final 42 subjects in
the study were also asked how their decision to
take the paracetamol preparation might have been
influenced if they had known it to have included
an antidote which would reduce its dangers. Patients
were asked for their responses to statements
presented on typed cards about the anticipated
dangers of paracetamol self-poisoning and about
possible specific warnings on labels of containers
of paracetamol which might deter them from taking
it in overdose. The warnings included the following:
“¿Anoverdose of paracetamol can kill―,“¿Due
to the addition of an antidote, exceeding the stated
dose may cause vomiting―,“¿Thisdrug is dangerous
in overdosage―, “¿Donot exceed the stated dose―,
and “¿Itmay be harmful to exceed the stated
dose―. Findings from further questions about
why paracetamol was chosen, the effects expected
by the patient, and knowledge about the dangers of
paracetamol self-poisoning, including the anticipated
timing of any adverse effects, are reported elsewhere
(Hawton et a!, 1995).
In each case the research interviewer assessed the
patient's suicidal intent using the Suicidal Intent
Scale (SIS; Beck et a!, 1974), which includes items
about the circumstances of the overdose and the
patient's response to questions concerning pre
meditation, intent, and so on, and the level of
depression by means of the Montgomery—Asberg
Depression Rating Scale (MADRS; 1979). The
clinical assessor for each case was asked to indicate,
from a list of motives (Bancroft eta!, 1979), the likely
reasons for the overdose, and whether or not in their
opinion the patient had wanted to die when taking
the overdose. The inter-rater reliability of the
measures was not investigated.
The information collected in the study was
combined with data routinely collected on each
patient by the Oxford Monitoring System for
Attempted Suicide (Hawton & Fagg, 1992a). This
provided details concerning each patient's current
problems (e.g. alcohol abuse).
We reviewed the results of the physical investigations
that had been conducted in order to ascertain whether
the patients had any signs of acute liver dysfunction.
Evidence of dysfunction was defined as an INR
(International Normalised Ratio), based on a pro
thrombin coagulation test, of greater than 1.2 which
had risenfrom or fallento a baselinevalue(usually 1.0).
We also noted whether or not the liver function tests
were abnormal. In defming a cut-off point with regard
to the number of tablets taken, weused the mid-point
(25 tablets) of the range of 20—30500-mg tablets
(10-15 g) that has been estimated as the threshold dose
for liver toxicity (Mitchell eta!, 1974; Prescott, 1983).
Statistical methods
The differences in proportions (odds ratios and 95%
confidence intervals (CIs)) were calculated using the
EPI-INFO program (Dean et a!, 1990). The signifi
cance of differences between the means of con
tinuous variables were assessed by the t-test.
Results
The 80 paracetamol self-poisoners included more
females than males, and were predominantly relatively
young (Table 1). According to their MADRS scores,
Sex
females
males
Age: years
13-20
21-35
36 +
MADRSscores
<22
>22
SISscores
0-6 (low)
7—12(moderate)
13—20(high)
21+ (very high)
Previousself-harm
Previousparacetamoloverdose
Table 1
Subjects included in the study
No. (%) of
patients(n=80)
53 (66)
27 (34)
32 (40)
32 (40)
16 (20)
42 (53)
38 (47)
17 (21)
31(39)
27 (34)
5 (6)
38 (48)
23 (29)
MotivesNo.
(%) of
patients'
(n=80)To
escapefromanintolerablesituation60
(75)To
gainrelieffroman unbearablestateof mind53
(66)To
makeotherpeopleunderstandhowdesperatethe
personwasfeeling53
(66)To
seekhelp29
(36)To
influenceotherpeople25
(31)To
makeotherpeoplesorry21(26)To
findif someonecaredforthe person13
(16)To
showhow muchthepersonlovedsomeone9
(11)1.
In mostcasesmorethanonemotivewasattributed.
PARACETAMOL SELF-POISONING
45
38 were in the moderately or more depressedrange
(22 or more). Fourteen patients were in the severely
depressed range (35 or more). Nearly half had had
previous episodes of deliberate self-harm, and more
than a quarter had taken at least one previous
paracetamol overdose.
The index overdose
Of the overdoses, 56 (70%) included paracetamol
only, 16(20°lo)paracetamol-containing compounds
only, and eight (10°lo)both paracetamol and
paracetamol-containing compounds. Of those who
took paracetamol-containing compounds, 11(69%)
said they had not known they contained paracetamol,
but all but one of them said that if they had known
they would still have used the preparation for self
poisoning. Twenty patients took other substances in
addition to paracetamol, including antidepressants
(six), major tranquillisers (three), other prescribed
drugs (six), and salicylates (two). In addition, 18
patients consumed alcohol as part of the overdose
and 29 had been drinking during the six hours
beforehand.The numbersof paracetamol/compound
tablets taken were 1—25(35 patients), 26—50(27
patients), more than 50 (15 patients), or were not
known (three patients). Twenty-seven patients said
they took all the tablets (or capsules) available, 28
more than half, and 24 half or less (1 did not know).
The tablets had been contained loose in a tub or
bottle in 37 (46%) cases and in a blister pack in 48
(60°lo)cases (five patients having taken both types
of preparation). Significantly more of those who
used only tablets from a loose preparation pack
(22/32, 69%) took more than 25 tablets compared
with those who used tablets from a blister-pack
preparationalone (16/40,40%; odds ratio 3.0,95%
CI 1.12 to 9.95, P=0.028). On the other hand,
somewhat more of those who used a blister-pack
preparationalone said that they had taken more than
half of the tablets available (33/42, 79°lo)compared
with those who used a loose preparation (19/32,
59%; odds ratio=2.51, 95% CI 0.81 to 7.91,
P= 0.125). While patients who used a blister-pack
preparation did not differ in their premeditation or
suicidal intent from those who used loose tablets,
those who took more than 25 tablets (of either
type of preparation) had higher suicidal intent
(mean 13.3, s.d. 6.2) than those who took 25 or
less (mean 9.1, s.d. 4.4; t=3.49, P<0.01). In
the patients who took 25 tablets or less, there
was no difference between the mean suicidal
intent scores of those who used a blister-pack
preparation and those who took tablets from a
loose preparation.
Source of the tablets
The most frequentsourceof the tabletswas a chemist
(32, 40%), followed by some other type of shop (19,
24%), a family member (10, 13%), and the general
practitioner (9, 11%). Thirty-three patients (41%)
had obtained the tablets less than an hour before
taking the overdose, 11 (14%) between 1 and 24
hours beforehand, seven(9%)! to 7 days beforehand,
and 25 (31%) more than a week beforehand (four
did not know when they had obtained the tablets).
In42 cases, the tabletshad been obtained specifically
for the purpose of taking an overdose, whereas the
remaining 38 patients had initially obtained them for
some other reason, usually pain relief. In 43 cases
the tablets had a usual location in the house - for
18this was the kitchen, nine the bedroom, seven the
bathroom, and nine some other location.
Premeditation and motivation
The overdoses were often taken impulsively, 33
patients (41%) having seriously contemplated taking
the overdose for less than one hour beforehand, 26
(33%) for between 1and less than 3 hours, five (6%)
for between 3 and 24 hours, and 16(200lo)for more
than a day. The tablets were usually taken rapidly,
36 (45%) having taken the overdose in less than 5
minutes, 27(34%) in between 5 minutes and an hour,
and 17 (21%) in more than an hour.
The clinical assessors regarded escape from a
difficult situation or state of mind as the most
frequent explanation, followed by communication
of distress,help-seekingand interpersonalcommuni
cation (Table 2). With regard to suicidal intent, the
clinical assessors thought that 31 (39%) patients had
wanted to die, 20(25%) had not wanted to die, and
29(36%) did not mind eitherway. In 32(40%) cases,
Table 1
The motives chosen by the clinical assessors to explain
the overdoses
Liverdamage:no
yes.
(%) of patients
noStatisticSignificanceOverdose
>25 tablets19/24
(79)23/50 (46)Odds
ratio=4.46 (95% Cl 1.31 to17.41)P=0.01Abused
alcohol4/20
(20)18/49
(37)Odds
ratio=0.43 (95% Cl 0.09 to1.64)P=O.29Mean
(s.d.)MADRSscore21.5 (10)23.2
(11)(=0.52NSMean
(s.d.)SISscore13.0
(6)10.7 (6)(=0.12NS
46
HAWTON ET AL
Table3
Associations with liver damage following paracetamol self-poisoning
the suicidal intent scores recorded by the research
interviewers were in the high range (Table 1).
Liver dysfunction
Twenty-four (31%) out of 77 patients (notes
unobtainable in three cases) had evidence of acute
liver dysfunction), 17 having abnormal INR values
alone and seven abnormal INR and liver function
test values. Transfer to a specialised liver unit was
considered for four of the latter patients but
eventually proved unnecessary. Liver dysfunction
was associated with consumption of more than 25
tablets in the overdose (Table 3). No association was
found with evidenceof alcohol problems, depression
or suicidal intent scores.
Factors which might have deterred patients from
taking paracetamol overdoses
Danger. After being told that paracetamol overdoses
can cause death even when treatment has been
instigated, 53 (66%) patients said they would still
have taken paracetamol in spite of knowing this (ten
patients were uncertain).
De!ay in effects. When told that if a paracetamol
overdose causesharmthe personusuallyis not aware
of this for several days, during which time he or she
is likely to remain conscious, only 28(35%) patients
said they would have taken paracetamol despite this
(eight subjects were uncertain).
Warning on labe!. The rank order in which
patients allocated cardsshowing specific wording of
possible warning labels on paracetamol containers
indicated that the simple statement “¿Anoverdose
of paracetamol can kill―was likely to have been the
most off-putting, 58 of 74 patients (78°/a)having
given this warning the highest rank. However, only
20(25%) patients thought that any type of warning
on the container would have stopped them taking
an overdose.
B!ister packs. When the 32 patients who had not
used tablets from a blister-pack preparation for the
overdose were asked what they would have done if
only blister-pack preparations were available, 21
(660/a)said they would still have taken a paracetamol
overdose (five patients were uncertain).
Limited number of tab!ets. The 68 patients who
had taken more than 12 tablets of paracetamol in
their overdoses were asked what they would have
done if paracetamol had been available only in small
packets (of 12tablets). Forty-seven (69%) said they
would still have taken a paracetamol overdose,
although 12(180/a) said that they would have taken
a smaller amount. Thirteen (19%) said they would
not have taken an overdose at all, and five (70/,)that
they would have taken some other substance (three
were uncertain). Thus 25 (37%) out of 68 patients
said that they would either have taken a smaller
paracetamol overdose or no overdose at all.
Paracetamo!availableonly on prescription. The
patients were also asked what they would have done
if paracetamol had been available only on pre
scription from a general practitioner (n = 71;
excluding patients who obtained paracetamol from
general practitioners). Fourteen (20°!.)said they
would still have taken a paracetamol overdose, 24
(34%) that they would have taken something else in
overdose, 4 (60/a)that they would have used some
other means of self-harm, and 25 (350/a)that they
wouldnot havetakenanoverdose(sixwereuncertain).
Antidote preventing liver damage. The fmal 42
patients included in the study were asked if they
would still have taken a paracetamol overdose if
paracetamolcontained an antidote which would stop
the harmful effects of an overdose. Twenty-seven
(640/.) said that they would not have done and 13
(310/.) that they would have (two were uncertain).
Discussion
The prevention of paracetamol self-poisoning, or at
least of its consequences, is a complex issue.
Paracetamol is a useful minor analgesic and
antipyretic, and in considering prevention one must
consider how it can remain available, yet either be
used less for self-poisoning or made more safe if
taken in overdose. A methodological issue relevant
47
PARACETAMOL SELF-POISONING
to interpretation of our findings concerns the
representativeness of the sample. Unfortunately, we
were unable to interview approximately a third of
the potential sample. While the excluded patients did
not differ from those interviewedin termsof sex and
age, we cannot saythatthey might not havediffered
in termsof other factors relevantto the focus of the
inquiry.
Packaging
It appears that warningson paracetamol containers
might deteronly a minority of patients from taking
paracetamol in overdose, with a blunt warning that
“¿Anoverdose of paracetamol can kill―being the
most off-putting of those shown to the patients in
this study. The impulsive nature of many of the
overdoses also suggests that warningson containers
might have little effect. Thereis the furtherriskthat
they might make paracetamol more attractive to
patients with high suicidal intent.
Ithasbeensuggestedthatblister-packpreparations
might reduce the likelihood that would-be self
poisoners would choose paracetamol for self
poisoning (Gazzard eta!, 1976). Our fmdings do not
support this hypothesis, as 600/aof patients had used
tablets from blister-pack preparations in their
overdoses, a figure which is similar to the current
salesof blisterpacks (54.90/a)as a percentageof sales
of all preparations of paracetamol in the UK
(Paracetamol Information Centre). Secondly, of
those who took tablets from a loose preparation,
two-thirds said they would still have taken an
overdose of paracetamol even if it was available
only in blister-pack form. Blister packs were,
however, associatedwith fewertabletsbeingtakenin
overdose. Thus significantly fewer of the patients
in this study who used this type of preparation in
contrast to those who used pills from a loose
container took more than 25 tablets. This was not
due to any difference in premeditation or suicidal
intent. Nor did it reflect a tendency for those using
tablets from a blister pack to consume a smaller
proportion of the availabletablets; therewas in fact
a tendencyfor the reverseto be true.The blisterpack
itself seems unlikely to be the key factor, but more
the fact that blister packs in general contain fewer
tablets. On average they contain less than half the
number of tablets found in loose preparations
(Paracetamol Information Centre). In patients who
took 25 tablets or less, one might have expected that
those who used a blister-pack preparation would
have had somewhat higher suicidal intent than
those who used a loose preparation, but this was
not found.
Limiting availability
It appears that limiting the overall number of
paracetamol tablets available for sale in each pack
mightpreventsome patientstakingan overdose, and
could mean that others would still make a para
cetamol overdose but use fewer tablets. This strategy
might, therefore, reducethe morbidityand mortality
associated with paracetamol self-poisoning. In order
to avoid the risk of patients turning to other over
the-counter medication for self-poisoning, which can
be at leastas dangerousas paracetamol(e.g. aspirin),
it would be sensible if this strategy were applied to
similar preparations.
Ifparacetamolwereavailableonlyon prescription,
it appears that this would be likely to prevent a
sizeable proportion of patients from taking an
overdose, although an equally large number said
they would have taken something else instead of
paracetamol. It would also mean that paracetamol
would be less readily available for its intended
use as a minor analgesic. Were all paracetamol
preparations to contain an antidote to liver damage
(e.g. methionine) this might reduce the incidence of
hepatic toxicity following paracetamol self-poisoning.
However, lackof knowledgeabout whetherthereare
any adverse effects of long-term use of methionine
in regularusers, possible toxicity of a large dose of
methionine, fears that knowledge of the antidote
might deter regular users of paracetamol, and
increased costs, limit the relevance of this strategy
at this stage. The majority of patients indicated that
knowledge of the presence of an antidote in
paracetamol tablets would put them off taking them
in overdose but, paradoxically, most patients said
that warnings of any kind on containers would not
prevent them taking an overdose.
Highlighting the delay before the effects of a
paracetamol overdose become apparent seems to be
important, since the majority of patients said they
would not havetakenparacetamolif theyhadknown
this. Perhaps this information could be imparted
through the media. It could also be given to patients
recovering from paracetamol self-poisoning, perhaps
supplemented by an information sheet, in the hope
of preventing repetition. Any method of reducing
the frequency of paracetamol self-poisoning is
associated with the problematic question of what
patients might do instead. One risk is that some
might use another dangerous method of self-harm,
particularly with an increasing range of drugs being
available in over-the-counter preparations.
Our overall conclusion is that limiting the
number of tablets available in all over-the-counter
paracetamol (and other similar) preparations (to 25
48
HAWTON ET AL
or perhaps fewer tablets) is the most pragmatic
initial approach to the problem of paracetamol
self-poisoning. Other approaches, except making
paracetamol available only on prescription, are
unlikely to succeed in view of the considerable
impulsiveness associated with most paracetamol
overdoses, the low likelihood that warnings on the
container will influence the behaviour and possible
effects on regular users of paracetamol. It should
become apparent relatively soon whether or not a
reduction in the number of tablets per container
would have the desired effect in terms of reduced
morbidity and mortality from paracetamol self
poisoning. If it did, this could contribute to the
achievement of the government's Health of the
Nation general suicide target (Secretary of State for
Health, 1992), especially in young people.
Acknowledgements
ThisstudywassupportedbyagrantfromOxfordRegionalHealth
Authority.The MonitoringSystemforAttemptedSuicideis
currently funded by a grant from the Department of Health. We
thankAnnDayandJoanFaggfortheirassistancewiththeanalysis
of the results, Dr Louise Sell and the staff at the Barnes Unit,
DepartmentofPsychologicalMedicine,JohnRadcliffeHospital,
fortheirassistancewiththisstudy,Dr JoanTrowellforadvice,
and Dr John Spoonerand the Paracetamol Information Centre for
adviceand provisionof furtherinformationaboutthegeneral
availability of paracetamol preparations.
References
AUGLJSTIN, C. & SCHMOLDT, A. (1991) Zunahme der
Paracetamolintoxikationen. Beitrag Gerichiliche Medizin, 49,
127—131.
BANCROFT, J., HAWTON, K., SIMKIN, S., el a! (1979) The reasons
people give for taking overdoses: a further enquiry. British
Journal of Medical Psychology,52, 353—365.
BECK, A. T., SCHUYLER, D. & HERMAN, I. (1974) Development of
suicidal intent scales.In ThePrediction of Suicide(edsA. T.
Beck, H. L. P. Resnick & D. J. Lettieri), pp. 45—56.
Philadelphia: CharlesPress.
BRAY, G. P. (1993) Liver failure induced by paracetamol. British
MedicalJournal,306, 157-158.
DEAN, A. 0., DEAN, J. A., BURTON, A. H., et a! (1990) Epi Info,
Version 5: A Word Processing, Database, and Statistics Program
for Epidemiologyon Microcomputers.Georgia:USD.
GAZZARD, B. G., DAVIS, M., SPOONER, J., et al (1976) Why do
peopleuseparacetamol for suicide?British Medical Journal, 1,
212—213.
HAWTON,K. & CATALAN,J. (1987) Attempted Suicide: A Practical
Guide to itsNature and Management (2nd edn). Oxford: Oxford
UniversityPress.
& FAGG,J. (1992a)Trendsin deliberateself-poisoningand
self-injury in Oxford, 1976—1990.British Medical Journal, 304,
1409—1411.
& (1992b) Deliberate self-poisoning and self-injury
in adolescents: a study of characteristics and trends in Oxford,
1976—1989.British Journal of Psychiatry, 161, 816—823.
WARE, C., MISTRY, H., et al (1995) Why patients choose
paracetamol for self-poisoning and their knowledge of its
dangers. British Medical Journal, 310, 164.
MITCHELL, J. R., THORGEIRSSON, S. S., POTrER, W. Z., et a!
(1974) Acetaminophen-induced hepatic injury. Protective role
of glutathione in man and rational for therapy. Clinical
Pharmacologyand Therapeutics,16,676—684.
MONTGOMERY, S. A. & ASBERG, M. (1979) A new depression scale
designed to be sensitive to change. British Journal of Psychiatry,
134, 382—389.
O'GRADY, J. 0., WENDON, J., TAN, K. C., et a! (1991) Liver
transplantation after paracetamol overdose. British Medical
Journal,303,221—223.
Oi-i-,P., DALHOFF,K., HANSEN,P. B., et a! (1990)Consumption,
overdose and death from analgesics during a period of over-the
counter availability of paracetamol in Denmark. Journal of
InternalMedicine,277,423—428.
PRESCOTT, L. F. (1983) Paracetamol overdosage: pharmacological
considerations and clinical management. Drugs, 25, 290—314.
SECRETARYOF STATE FOR HEALTH (1992) The Health of the Nation:
A Strategyfor HealthinEngland.London: HMSO.
SPOONER, J. B. & HARVEY, J. G. (1993) Paracetamol overdosage —¿
factsnot misconceptions.PharmaceuticalJournal, 250, 706-707.
Keith Hawton, DM,Christopher Ware, MRCPsych,Stephen Kingsbury, MRCPsych,University Department
of Psychiatry, Warneford Hospital, Oxford; Hamant Mistry, MRCGP, Jonathan Hewitt, MRCPsych, Dave
Roberts, RMN, Heather Weitzel, RMN, Barnes Unit, Department of Psychological Medicine, John Radcliffe
Hospital, Headington, Oxford
Correspondence: Dr K. Hawton, University Department of Psychiatry, Warneford Hospital, Oxford 0X3 7JX
(First received 23 January 1995, final revision 18 May 1995, accepted 19 May 1995)
