476SAW VOL 78 20 OCT 1990
Toxic cannabis psychosis
•IS a valid
K. SOLOMONS,V. M. NEPPE,
J. M. KUYL
One hundred black men admitted to hospital with acute
psychiatric symptoms were investigated for the presence of
urinary cannabis metabolltes in order to delineate the psy-
chiatric role played by 'dagga', the potent ~
cannabinol, in the study population and to determine the
diagnostic value of the entity 'toxic psychosis (dagga)'.
Cannabinoids were present in 29% of patients, and 31% were
discharged with a diagnosis of toxic psychosis (dagga).
Clinical and demographic material was gathered for all
patients and no consistent differences were found between
dagga-posItlve and dagga-negative patients or toxic dagga
psychotic patients and 'functional' psychotics other than a
history of recent dagga use and the dagga screening test
result. The latter measure was found to be both more sensitive
and more specific than the history of dagga use alone. The
findings support the routine use of a simple screening test for
dagga in the sample population studied. The study
demonstrated the heterogeneous nature of the toxic dagga
psychosis syndrome by documenting a variety of different
clinical presentations, which included schizophrenia (42%),
paranoia (26%), maniform psychosis (16%) and organic
S AIr Med J 1990; 78: 476-481.
Epidemiological studies on dagga, the potent southern African
cannabinoid, are very scanty: There are no reliable estimates
of the incidence of dagga use, intoxication or other associated
syndromes for southern Mrican populations.I-13A 1983 study
by NeIJ found measurable amounts of dagga in 83 of 192
violent deaths (43%) in the Durban area. The three youngest
people, aged 7 years, 12 years and 13 years, were pedestrians
with cannabinoids in their bodies. During the same period,
cannabinoids were detected in 15 of 23 cases (65%) where
people died suddenly from natural causes.
The ratio of dagga offences per head of population rose
from 81/100000 in 1945 to 198/100000 in 1970 and 135/
100000 in 1980. These figures compare favourably with figures
from Europe.14 The poor state of knowledge regarding dagga
use in South Africa is particularly noticeable with regard to
the indigenous black population -
South Africa have not included the black population. 13
the 12 reports published in
Department of Psychiatry, University of Witwatersrand,
Johannesburg and Sterkfontein Hospital, Krugersdorp, Tvl
K. SOLOMONS, M.B. CH.B., D.O.H., D.T.M. & H., MMED. (pSYCH.)
V. M. NEPPE, M.B. B.CH., D.P.M., B.A., F.F. PSYCH. (S.A.), M.MED. (pSYCH.),
PH.D. (MED.), F.R.CP.e., DIP. A.B.P.N. (Present address: Division of
Neuropsychiatry, Department of Psychiatry and Behavioral
Sciences, RP-IO, University of Washington, Seattle, Washington,
Department ofChemical Pathology, South African Institute
for Medical Research,Johannesburg
J. M. KUYL, M.B. CH.B., F.F. PATH. (S.A.), B.Se. (Present address:
Department of Chemical Pathology, University of the Orange
Free State, Bloemfontein)
Aaxptod 30 No. 1989.
Reprint r<qucstS to: Dr K. Solomons, 125 A1gernon Rd, Norwood, 2192.JlSA.
This study was designed to answer three questions:
1. Is there a value to routine screening for urinary canna-
binoids in the indigenous southern African male psychotic
2. Does the patient whose urine is positive for cannabinol
differ clinically from the patient whose urine is negative?
3. Is there a recognisable psychiatric entity of toxic dagga
psychosis, and do subgroups ofthis entity exist?
The sample included 110 consecutive black men admined to a
mental hospital with acute psychiatric symptoms. The 110
patients were evaluated to ensure that, after exclusions, at least
100 patients would be available for full study. Patients were
excluded if they failed to remain in the ward for the fIrst 7
days after admission for any reason, medical or otherwise.
Patient details were obtained on admission by an unstruc-
tured open-ended interview, which evaluated demographic
background, current psychiatric status, past medical and
_psychiatric history, family medical and psychiatric history, and
current physical and mental status. A differential diagnosis
was formulated and a plan of management determined. -The
differential diagnosis and management plan was reviewed- at
each subsequent interview. Each patient was seen by two
pnysicians - a registrar and a psychiatric consultant.
Urine specimens were collected on admission and after 7
days. These were screened for cannabinoids by the EMIT
semiquantitative enzyme immunoassay.15
Patients were rated on the Brief Psychiatric Rating Scale
(BPRS) of Overall and Gorham,16 at the initial interview and
on the 7th day. The BPRS has been used extensively in
research in many countries. No rating scale of symptoms has
been validated in the indigenous South Mrican population and
the use of the BPRS was felt to be justified as a way of
documenting the overall symptom severity and the change in
this over time. The 18 items of the BPRS were scored from 0
(no pathology) to 6 (extreme severity). One physician rated all
Diagnoses were made according to the diagnostic criteria
laid down in the Diagnostic and Statistical Manual ofMental
Disorders6of the American Psychiatric Association (DSM
III).17 For conditions diagnosed as toxic psychoses, which did
not adequately fit the clinical picture, the diagnosis was made
using ad hoc clinical criteria.
The diagnosis of toxic dagga psychosis was made when all
three ofthe following criteria were present:
1. The patient had either a history of recent dagga use or
the patient's urine tested positive for cannabinoid or both.
2. The patient had to be psychotic or have been psychotic
shortly before admission to hospital. Psychotic features
included: formal thought disorder; hallucinations; delusions;
behaviour disturbances; and lack ofinsight.
3. During hospitalisation the patient had to respond to
treatment reasonably rapidly, with clear evidence of return to
pre-episode personality functioning. An 8-week period was
used as an arbitrary cut-off point. Beyond this period, no-one
was diagnosed primarily as having toxic dagga psychosis.
SAMJVOL 7820 OKT 1990477
The primary diagnoses made at the time of discharge or at
the end of the study are shown in Table H. Toxic dagga
psychosis was diagnosed in 31% of cases. Other diagnoses
• Ufesly1e was defined asfollows: urban patients living in the majorcities; rural patients
living on farms; semi-rural patients living in small farming towns.
t Article refers to the Section of the Mental Health Act underwhic.', patients are admit-
tad to hospital. Article 4 provides for patients who are admitted voluntarily to hospital
by a family member. Article 9 provides for patients who are certified by two doctors
via the Courts as involuntary patients. Article 12 provides for patients who are urgent-
ly certified by one doctor in a manner which initially bypasses the Courts -
tients are admitted against their will, as with Section 9 patients.
BPRS = Brief Psychiatric Rating Scale
25,0 ± 5,7
1,68 ± 0,10
60,2 ± 9,4
6,2 ± 8,2
3 3 , ~
8,3 ± 13,2
27,9 ± 23,7 23,9 ± 23,016,2 ± 10,8
18,8 ± 12,116,6 ± 10,7 17,3 ± 11,9
Years at school
History of dagga
History of alcohol
BPRS initial score
taken (days be-
TABLE I. DEMOGRAPHIC AND CLINICAL RESULTS: COM-
PARISON BETWEEN TOTAL POPULATION, URINE-POSITIVE
PATIENTS AND TOXIC DAGGA PSYCHOSIS PATIENTS
(MEAN ± SO)
28,2 ± 8,7
1,65 ± 0,09 1,67 ± 0,10
57,1 ± 10,4 58,9 ± 10,3
24,9 ± 6,2
The data were analysed by the BMDP statistical software
package18using parametric and non-parametric techniques to
ascenain whether any statistically significant differences in any
of the parameters could be detected between the patients
whose urine tested positive for cannabinoids and those who
were negative, and between the toxic dagga psychosis group,
and the remaining non-toxic dagga psychosis group. Discrimi-
nant analysis was performed on all factors found to differ
significantly between both the urine-positive and -negative
groups and the toxic dagga psychosis and non-toxic dagga
psychosis groups. This was done to determine whether any of
the factors, either singly or in combination, could predict the
group into which each patient would be classified on the basis
of each factor and whether this predictive power was signifi-
cantly greater than the 50% prediction by simple guesswork.
The course which these symptoms ran was also noted,
particularly whether they returned to normal on recovery from
the acute episode.
The diagnosis of toxic dagga psychosis was further sub-
classified into: (z) maniform; (iz) paranoid; (iiz) schizophreni-
form; or (iv) organic; according to the predominant mode of
In the maniform mode, the patient was agitated and restless,
his speech pressured, his thoughts grandiose and he had either
circumstantiality, tangentiality or flight of ideas. Sleep was
diminished and the patient was overactive in a purposeless
The paranoid presentation was noted in patients whose most
prominent symptom was paranoid ideation with an appropriate
and congruent mood.
The schizophreniform presentation was characterised by
blunting of affect; withdrawal; poveny of ideation, volition
and movement; bizarre somatic or other delusions and formal
The features characterising the organic presentation were
clouding or fluctuation in consciousness, disorientation and
cognitive or intellectual deficit.
The population studied were mostly young men (mean age
28,2 years), single (81%) and childless (83%), who lived in
townships in and around Johannesburg (83%). In the main
they were poorly educated with a mean educational level of
standard 4, and a mean of 6,38 years of schooling. Only 39%
had progressed beyond junior school and only 5% had matricu-
lated. Some 7% had received no education at all and 9% had
not advanced beyond grade H. The majority of the sample
were unemployed (69%) and most of the rest (25%) were
employed as unskilled labourers. The study population sPoke
8 different languages with Sotho, Zulu, Tswana and Xhosa
being the most common. Most of the patients had been
committed by the Couns (72%) and a significant minority
were voluntary patients (28%). Readmissions accounted for
62% of all admissions with 5% having 5 or more previous
admissions. Just over one-third of the study population (38%)
had never been psychiatric inpatients before.
The study group tended to be physically small with a mean
height of 165 cm and.a mean weight of 57,14 kg; -as many as
60% weighed less than 50 kg. In spite of all this, the men
tended to be physically healthy with avitaminoses and poor
nutrition apparent in only 7% (Table I).
478SAMJVOL. 78 20 OCT 1990
commonly found included toxic alcohol psychosis (11%),
schizophrenia (9%), schizophreniform episodes (8%), affective
disorders (7%) and organic brain syndromes (9%). A secondary
discharge diagnosis was made in 55% of cases and included
chronic substance abuse (20%), toxic precipitation offunctional
disorders by dagga (3%) and alcohol (11%), mental retardation
(8%), personality disorders (4%) and 'other' in the remaining
The total BPRS score on admission varied between° and 52
with a mean of 18,8. This had dropped to 9,1 when rated 7
days later with 62% of the study population scoring less than
10. Whereas 25% of the population was initially assessed as
being extremely severely disturbed, only 7% were still found to
be as severely disturbed 7 days later.
A history of dagga use was obtained from 45% of the
patients. Dagga was reported to have been last taken within
the month preceding admission in 27% of cases, between 1 and
6 months preceding admission in 10% and more than 6 months
before admission in 8% of cases. No history of dagga use was
obtained in the remaining 55% ofthe study group.
A history of previous alcohol use was elicited from 60% of
the study group with 40% having used alcohol within the
month preceding admission. No one who was clinically intoxi-
Toxic dagga psychosis
The diagnosis oftoxic dagga psychosis was made in 31 cases
(31%) according to the criteria described above. A history of
recent dagga use in the month preceding admission was
obtained in 18 of the cases (58,1%), a further 8 patients
volunteered using dagga in the past but not in the month
preceding admission (25,8%) and the remaining 5 (16,1%)
denied any previous dagga use. In 26 of these 31 patients, the
urinary cannabinoid levels were positive (83,9%). Of the
remaining 5 urine-negative patients, 4 had spent 4 or more
weeks in a general hospital immediately before admission to
this hospital. The 5th person had spent 2 weeks in a general
hospital before admission to this hospital.
The role ofdagga
The test for cannabinoids in the urine was positive in 29
cases (29%). Of these 29, 26 were diagnosed as having toxic
dagga psychosis (89,6%) while oLthe remaining 3 patients, 2
were diagnosed as having schizophreniform episodes with
toxic precipitation by dagga and 1 had active neuros'yphilis as
well as cannabinoids in his urine. In 16 cases (55%) the urine
reverted to negative when tested 7 days later. The remaining
13 cases (45%) remained positive.
A history ofdagga use within the month preceding admission
was elicited in 19 cases (65,5%); 4 patients admitted to dagga
use, but not within the preceding month (13,8%), and 6
patients denied any previous dagga use (20,7%).
Ofthe 71 patients whose urine were negative for cannabinoid
metabolites, 5 were diagnosed as having a toxic dagga psychosis
(7%) and 8 volunteered a history of dagga use within the
month before admission and urine testing (11,2%). Sixteen
volunteered a history of previous dagga use but not within the
month preceding admission (22,5%) and iD. 47 cases, no history
of previous dagga use was obtained.
Patients with cannabinoid metabolites in their urine differed
significantly from those without such signs in that they were 4
years younger (P 0,0053), had fewer children (P 0,02), had
fewer side-effects from medication (P 0,03) and were' more
conceptually disorganised on the_BPRS rating, (P 0,0034).
They also gave a history of previous dagga use. more frequently
(P<0,0001) and had been at home more recently before entry
into this study than the patients whose urine was negative (P
0,01). They also tended more often to be single (P 0,05), were
less likely to consume alcohol (P 0,07) and were more likely to
be admitted as urgent certiflcations than patients with negative
urine tests (P 0,07). The first four and latter three factors are
not useful in clinically differentiating indiv.idual patients despite
being statistically significant.
However, on discriminant analysis ofthese nine parameters,
the three variables of previous dagga use, previous alcohol use
and conceptual disorganisation taken together raise the possi-
bility ofcorrectly classifying each patient into the urine positive
or urine negative group from 50% by guessing to 76,3%. These
three variables therefore were useful predictors of whether the
patient's urine would test positive or negative.
cated with alcohol was present in the study population. When
the study ended, 90 days after the first patient entered the
study, 76% of the population had been discharged and were
still out of hospital ,whereas 9% had been discharged but
readmitted, 12% were still in hospital and the remaining 3%
had absconded from hospital without being discharged and
were lost to follow-up.
The mean duration of hospital stay was 27 days and most of
the population (72%) were discharged within 1 month of
TABLE 11. DIAGNOSES AT DISCHARGE AND AT END OF
Primary discharge diagnosis
No axis-I psychiatric disorder
Toxic dagga psychosis
Toxic alcohol psychosis
Major depressive episode
Organic brain syndromes
Post-ictal confusional state
Acute confusional state
Organic amnesic syndrome
Brief reactive psychosis
Temporal lobe epilepsy
Chronic alcohol abuse
Secondary discharge diagnosis
Chronic substance abuse
Toxic precipitation of functional
disorder by dagga
Toxic precipitation of functional
disorder by alcohol
S MJVOL 78 20 OKT 1990
TABLE IV. SENSITIVITY AND SPECIFICITY OF DAGGA
HISTORY AND URINARY CANNABINOID ASSAY FOR TOXIC
Discriminant analysis perform d on these variables produced
a classifIcation equation whereby history of dagga use and
presen e of urinary cannabinoids together signilicantly increase
the pos ibility of orrectly classifying a patient into the toxic
or non-toxic groups. History of dagga use alone had a predic-
dye power of 80% and the urine test result had a predictive
po er of 89%. The two variable
predictive power of 89,7%. The e results indicate that if a
person had a history of pre ious dagga use, there was an 80%
chance that he would be diagnosed as having toxi
psychosis. If his previous dagga history was not known, there
existed a 50% chance that he would be correctly classifIed into
either the toxic or non-toxic group. If the patient had
measurable cannabinoids in his urine, then there was an 89%
chance that he would be correctly classifIed as having toxic
dagga psychosi . Wherea the addition of the variable urine
cannabinoid result increased the chance by 9,7% over previous
dagga history alone, the addition of previous dagga history to
urinary cannabinoid result only raised the possibility ofcorrect
classifI ation by 0,7%. It follows that urinary cannabinoid
result were more valuable in diagnosing toxic dagga psychosis
than the history of previous dagga use.
The discriminant analysis results also confIrm d the finding
that in spite of demonsuable statistical signilicance of the
remaining factor , their clinical value for classifying a person
into the toxic or non-toxic group, and thus their utility, was
In thisrudy, a history of dagga use (less than 1 month
before admission to hospital), had a sensitivity of 74,0% for the
diagnosis of toxic dagga psychosis. A history of any previous
dagga use yielded a sensitiviry of 57,8%. The sensitivity of the
urinary cannabinoid assay for toxic dagga psycho is was 89,6%.
The specifIcity of these two parameters for correctly making
the diagnosis of toxic dagga psychosis was 58,I%for the dagga
history and 83,9% for the urine cannabinoid assay technique
over a history of dagga use alone for the diagnosis of toxic
dagga psychosis (Table IV).
together had a combined
In 8 cases, the discharge diagnosis
urine results were known, and the diagnoses were subsequently
changed to include toxic dagga psychosi after the urine te t
results were found to be positive. In 2 of the e cases, the initial
discharge diagnosis was atypical psychosis in mild mental
retardates, in another 2 ca es, toxi dagga psy hosis was added
to the diagnosis of post-uaumatic confusional episo e, atypical
psychosis (ofdissociative-hysterical kind), alcohol-related seizure
(with a normal EEG) and alcoholic hallucinosis was altered to
include toxic dagga psychosis. Thu urinary cannabinol result
A history of recent coexisting alcohol u
preceding admission to hospital was obtained in 10 cases of
toxic dagga psychoses (32,3%). No history of alcohol use at all
was obtained in 14 cases (45,2%) and 7 patients claimed to use
alcohol but not within the month preceding admission (22,6%).
It is possible that the 10 patients who used alcohol together
with dagga before the onset of their psychiatric symptoms and
admission to hospital were in fact suffering from a mixed
dagga and alcohol toxic' psychosi . The study design was not
able, however, to differentiate the influence of these two
factors from each other.
According to the criteria outlined abo e, the 31 toxic dagga
psychosis patients were subdivided into four subcategories as
follows: schizophreniform presentation (13 cases; 41,9%); mani-
form presentation (8 cases; 25,8%); paranoid presentation (5
cases; 16,1%); and organic presentation (5 cases; 16,1%).
This spread of different presentations in the toxic dagga
psychosis group confIrms the view that toxic dagga psychosis
is probably not a homogenous condition with a single mode of
Statistically signilicant differences between the toxic dagga
psychosis group and the control group were found with the
following variables reflected in Table III -
weight, number ofneuroleptic drugs and side-effects ipsofaceo
correlate highly with the toxic group. Clinically relevant fmd-
ings, however, are less neuroleptic agent, shoner duration
(half the length) of hospital stay, and less psychomotor
Differences which tended to ards signilicance included more
hostility (P 0,09), less overall severity (P 0,08) on the initial
BPRS rating, less emotional withdrawal (P 0,05) and less
conceptual disorganisation (P 0,07) on day 7 in the toxic group
'thin the month
• 01 no substance value because differences between 9rouPS were too small; statisti-
cally significant at P < 0,05 level.
•• P < 0,001; substantive differences of clinical value.
••• Tending to significance; 0,01 < P < 0,05.
History of dagge use
Positive urinary cannabinoid test
Duration of hospitalisation
No. of neuroleptics in treatment
Motor retardation (initial)
Motor retardation (d 7)
BPRS score (initial)
Emotional withdrawal (a 7)
Conceptual disorganisation (d 7)
TABLE Ill. DIFFERENCES BETWEEN THE TOXIC DAGGA
PSYCHOSIS AND CONTROL GROUPS
0,001 * * *
0,0001 * *
The profile of the typical dagga psychotic in this population
was a childless, single, urban, poorly educated, unskilled or
unemployed young man of normal weight and height who was
physically healthy. He had usually been committed (cenifIed)
but might be a first or a readmission. He provided a histoty of
dagga use and might or might not be a drinker. He was
psychiatrically ill with a lowish BPRS score and required neuro-
leptic treatment but was unlikely to develop side-effects. He
recovered rapidly and was discharged after 16 days. He was
unlikely to present with anxiety, tension, mannerisms, and
posruring, guilt or depressed mood and was more likely to
present with somatic complaints, emotional withdrawal, motor
retardation, excitement and pressured speech. He was likely to
be conceptually disorganised, unco-operative and hostile, sus-
picious, grandiose, hallucinated and disoriented with emotional
SAMJVOL 7820 OCT 1990
blunting and unusual thought content. He might present with
either a maniform, paranoid, schizophreniform or organic
picture which resolved rapidly.
The study was limited by a number of factors. Corroborative
information about the patient's current and past psychiatric
history from family members, employers and other profes-
sionals was unavailable or scanty in most cases. This was due
to the socio-economic circumstances of the patients as well as
insufficient staffmg to permit satisfactory eliciting of details.
The reliability of clinical assessments was constrained by
factors such as the understaffing and overcrowding of ward
where the study was carried out and that it had a very high
turnover of patients. An average monthly admission and dis-
charge rate for the BO-bed ward of 100 - 120 was not
uncommon. This shoncoming was exacerbated by the lack of
other evaluation sources, since no social workers, psychologists
or occupational therapists worked on the black admission
ward. The nursing staff were as overworked as the psychiatric
medical staffand their clinical input was similarly sub-optimal.
Cannabis was the only toxic substance measured in body
fluids. Neither alcohol nor other toxins, e.g. methaqualone
(Mandrax) was tested for strategic reasons, despite histories of
recent alcohol and Mandrax use in 40 and 2 cases, respectively.
The imponant variable ofset and setting, which is imporrant
in psychological responses to daggal9,20 was not and could not
be evaluated and controlled for in the study design. The
groups of patients with urine positive and negative for canna-
binoids, and toxic dagga psychosis and control patients were
not prospectively matched for demographic variables but this
was compensated for by the finding that the groups did not
The results were analysed as though no false-positives
occurred with the EMIT assay technique, whereas false-
positives and false-negatives do occur. No more accurate
method ofconfmning the presence ofcannabinoid metabolites,
speerrometry or radio-lDlIDunoassay, was used,-
limiting an accurate assessment. Previous studies comparing
the EMIT with more sensitive techniques have found false-
negatives to be more of a problem than false-positives/4-26a
2,9% occurrence of false-positives and lB% of false-negatives
was found in one study.26 The bias introduced by false-
positive results was unlikely to have influenced the results in
this study, and an underestimate of cannabis use was more
likely than an overestimate.
In spite of these drawbacks, this study sheds light on a
number of previously undocumented concerns. Demographic
similarity between the toxic dagga psychosis and positive
urinary cannabinoid groups was evident, as was similarity
between urine-positive and urine-negative groups. Where
differences did occur, they were of little clinical value. For
example, the age difference was only 4 years and the mean age
of all groups was below 30 years. These demographic features
and profIles accord with those found in other communities
where the toxic dagga psychosIs is encountered frequently.27
Dagga was demonstrated to be a factor in a significant
proportion of the study population. This substantial frequency
correlates well with the frequency with which the diagnosis
was made over the same 4-week period the year before and for
the entire year during which the study was performed. In the
4-week period of the previous year, 21 out of68 patients (30%)
were discharged with the diagnosis of toxic dagga psychosis
and 237 of 720 patients discharged (33%) for the full year were
diagnosed as having toxic dagga psychosis. The principal
difference between the diagnoses made previously and those
c h r o r : n a ~ o g r a p h y , gas c h r o m a t o & f ~ R h y / m a s s
made in this Study was that the previous diagnoses were made
without any objective evidence that dagga had, in fact, been
The clinical presentation of psychotic reactions in association
with dagga use in this study population was similar to the
variety of psychotic reactions to dagga that have been exten-
sively reponed.28-23The heterogenous nature of the toxic
dagga psychosis was ably demonstrated in this study where
maniform, schizophreniform, paranoid and organic presenta-
tions were all encountered. This study demonstrated the vali-
dity and applicability ofusing the postulated diagnostic criteria
for toxic dagga psychosis.
This fmding is well supponed by the large number of
physicians who have, anecdotally, clinically described this
entity in areas such as southern Mrica where the potency of
cannabinol is 20-3O-fold greater than American marijuana.9,27
In general, marijuana has been perceived as a non-psychoto-
genic substance in the USA and the clinical demonstration of
a real cannabinol psychosis is therefore imporrant.34,35
This study also confirmed the value of routine screening of
urine for dagga metabolites in this population. This is for four
reasons: (1) dagga metabolites are present in a high percentage
of cases (29%); (il) these fmdings affect the fmal discharge
diagnosis in a number of patients (35%); (iil) no reliable
criteria distinguish demographically or clinically between
dagga-using patients and the r ~ t of the population; and (iv)
the diagnosis of toxic dagga psychosis is made almost exclu-
sively in urine-positive patients. Theoretically this is almost a
. tautology but even theoretical tautologies require scientific
validation in practice. Thus, the value of a simple, reliable
screening procedure that is able to differentiate toxic dagga
psychotic patients from the rest is an invaluable aid to patient
management, panicularly in this clinical setting where so little
information is readily available..
1. Nel JP. The prevalence of cannabinoids and alcohol in body fluids of
persons who died violently in the Durban area. Souzh African Conferma on
Dagga. Pretoria: Department ofHealth and Welfare, 1983: 41-44.
2. Simon AM. A study of drug abuse in a group of South African univ=ity
students. S AfrMed] 1982; 61: 666-668.
3. Herr P, Motley JE. Drug use patterns among South Mrican undergraduates.
S AfrMed] 1972; 46: 1404-1407.
4. Levin SM, Berman C, Cobb H, Mcllraith I. Dagga (cannabis) usage among
medical students in Johannesburg. S AfrMed] 1983; 63: 507-609.
5. Du Toit BM. Drug Use ami South African Studems. (papers in International
Studies, African Series No. 35.) Athens, Oh.: Ohio University Center for
International Studies, 1978.
6. Levin A. The pattern of drug taking among drug-dependent South African
national servicemen. S AfrMed] 1972; 46: 1690-1694.
7. Levin A. 'n Ontleiding van die georuik van divelsmasmiddels en sekere
gevolge daarvan, met klem op Cannabis satroq,'by 'n monster ioJ!gtD8IlS
opgeroep vir militere diensplig. Thesis for the degree M.D., University of
8. Le Roux PHduP, Botha EM. Dagga use in the Cape Peninsula. Souzh
African Cqnferma on Dagga. Pretoria: Department of Health and Welfare,
9. Van der Burgh C. Some epidemiological aspects ofdagga use. SouzhAfrican
Conferma on Dagga. Pretoria: Department of Health and Welfare, 1983:
10. Botha EM, Le Roux PJ, Du Pre PI. Daggagehruik in die Kaapse SkiereiJand.
Bellville: Institute for Social Development, University ofthe Western Cape,
11. Logie P, Motley JE, Bensusan AP. The dagga smoker: a survey. S AfrMed
] 1972; 46: 1400-1403.
12. Rottanburg D, Robins AH, Ben-Arie 0, Tegin A, Elk.R. Cannabis-
associated psychosis with hypnomanic features. Lanat 1982; 2: 1364-1366.
13. Neethling LP. The enent of the dagga problem in the Republic of South
Africa. South African Conference on Dagga. Pretoria: Department of Health
and Welfare, 1983: 8-11.
14. Solomons K, Neppe VM. Cannabis: its clinical effects. S Afr MedJ 1989;
15. Package insert, EMIT d.a.u. Cannabinoid 20 Assay. Palo Alto, Calif.: Syva
16. Overall JE, Gorham DR. The Brief Psychiatric Raring Scale. Psychol Rep
1962; 10: 799-812.
17. American Psychiatric Association. Diagnostic amiStaristicalManualofMmud
Disorders. 3rd ed. Washington, DC: APA, 1980: 59-89.
18. Dixon WJ, Brown MB, Enlteiman C et al. BMDP Staristical SojtfJJare.
Berkeley, Calif.: UniversityorCalifornia Press, 1961.
19. Smith DE. The acute and chronic £ODciry of marijuana. PsycMthlic Drugs
1968; 2: 37-47.
20. Jones RT. Marijuana-induced 'high': influence of expectation, setting and
previous drug experience. Plttzrmaml Rro 1971; 23: 359-369.
21. Whiting 10, Manders WW. Confirmation of a tetrahydrocannibol metabo-
lite in urine by gas chromatography.] Anal Toricoll982; 6: 49-52.
22. Whiting 10, Manders WW. The confirmation of 9-<:arboxy-THC in urine
by gas chromatography/mass spectrometry. ATliat Spaa EntJiron Med 1983;
23. Schwanz RH, Hawks RI. Laboratory detection of marijuana use. ]AMA
1985; 254: 788-792.
24. Sutheimer CA, Yarborough R, Hepler BR et al. Detection and confUlIllltion
of urinary cannabinoids.] Anal Toricoll985; 9: 156-160.
25. SilbetTJ, GetsOn P, Ridley S, losefschn M, Hicks JM. Adolescem marijuana
use -concotdance berween questionnaire and immunoassay for cannabinoid
merabolites.] PtdUlrr 1987; 111: 299-302.
26. SchwarIZ RH, Willene RE, Hayden GF, Bogema S, Thome MM, Hicks J.
Urinary cannabinoids in monitoring abstinence in a drug abuse treatment
program. Arch Parhol Lab Med 1987; 111: 708-711.
SAMJVOL 7820 OKT 1990
27.•'ahas GG. Cannabis: £ODcological properties and epidemiological aspectS.
Med] Awt 1986; 145: 82-87.
28. Bensusan AD. Drug pollution -the problem of abuse. S Afr MedJ 1971;
29. Chopra IS, Smith ]W. Psychotic reactions following cannabis use in East
Indians. Arch Gm Psychiarry 1974; 30: 24-27.
30. Thacore VR, Shukla DRP. Cannabis psychosis and paranoid schizophrenia.
, Arch Gm Psychiatry 1976; 33: 383-386.
31. Knight F. Role of cannabis in psychiatric disturbance. AM NY Acad Sci
1976; 282: 64-71.
32. Rorranbutg 0, Robins AH, Ben-Arie 0, Teggin A, Elk R. Cannabis-
.associated psychosis with hypomanic features. Laneer 1983; 2: 1364-1366.
33. Harding T, Knight F. Marijuana-modified mania. Arch Gm Psychiarry
1973; 29: 635-637.
34. Negrete]e. What's happened £0 the cannabis debate? BrJ Addict 1988; 83:
35. Taschner KL Psychopathology and differential diagnosis of so-called can-
nabis psychoses. Foruchr Neurol Psycmalr 1983; 51: 235-248.
Prenatal ultrasonographic diagnosis and
successful management of mediastinal teratoma
A case report
H. R. J. DUMBELL,
A. C. COLEMAN,
J. M. PUDIFIN, W. S. WINSHIP
A case of mediastinal teratoma, diagnosed in utero by real-
time u"rasonography during a late 3rd trimester evaluation of
polyhydramnios, is described. Prompt respiratory assistance
to the infant at birth and early surgical intervention led to a
S Air Med J 1990; 78: 481-483.
A 25-year-old primigravida was referred for investigation of
rapidly developing polyhydramnios at 35 weeks' gestation; this
was conf"rrmed by real-time ultrasonography. The thorax of
the fetus was seen to contain multiple cystic areas, which
displaced the hean to the right (Fig. 1). The stomach was
identified in the abdomen and appeared normal (Fig. 2).
There were no ascites and the kidneys were normal; there was
a small amount of skin oedema. A provisional diagnosis of
either cystic malformation of the lung or a cystic mediastinal
mass was made.
Labour was induced at 36 weeks' gestation following spon-
·aneous rupture of membranes. A female infant weighing
~ e p a r t m e n t s of Paediatrics and Radiology, University of
'";atal and Addington Hospital, Durban
-I. R. J. DUMBELL, M.B. CH.B., D. OBST. R.C.O.G., M.R.C.P., D.C.H.
. C. COLEMAN, M.B. CH.B., D.C.H.
.M. PUDIFIN, B.A. HONS, D.D.R., c.R.U.
V. S. WINSHIP, M.B. CH.B., M.MED. (pAED.)
Fig. 1. Transverse ultrasonographic section through the thorax of
the fetus showing the heart (small arrow) and the cystic masses
of the teratoma (large arrow).
2900 g was delivered. The baby required immediate active
resuscitation but despite interminent positive pressure venti-
lation, air entry could not be detected clinically.
Apgar scores were 2, 3 and 5 at 1, 5 and 10 minutes,
respectively. The infant was put onto a Bourne's ventilator in the
neonatal unit. Onexamination she was cyanosed and oedematous
with a distended abdomen and 5 cm hepatomegaly. There was
bradycardia and the apex beat was displaced to the right.
Chest radiography showed opacity ofboth lung fields (Fig. 3).
Chest drains were insened, and 30 ml of straw-eoloured fluid
was aspirated from the right pleural space after which some
aerated lung was seen (Fig. 4).