The abuse of barbiturates in the United Kingdom

Sections

I. Historical introduction: the "battle of the barbiturates"
II. Barbiturate consumption
III. Acute barbiturate poisoning
IV. Barbiturate habituation, addiction and chronic intoxication
V. Barbiturates and alcohol
Conclusions
Acknowledgements

Details

Author: M. M. Glatt
Pages: 19 to 38
Creation Date: 1962/01/01

The abuse of barbiturates in the United Kingdom

M. M. Glatt a

Recently so hot a controversy has raged around the barbiturates that the strife has been termed "the battle of the barbiturates".

E. W. Adams, 1937

Contents

   

Page

I.
Historical introduction: The "battle of the barbiturates
19
II.
Barbiturate consumption
22
III.
Acute barbiturate poisoning
 
1.
Suicides and accidental deaths
25
2.
Attempted suicides
28
IV.
Barbiturate habituation, addiction and chronic intoxication
32
V.
Barbiturates and alcohol
35
VI.
Conclusions
36
 
Acknowledgements
37
 
References
38

I. Historical introduction: the "battle of the barbiturates"

The modern era of the use of hypnotics dates back less than a hundred years. In 1869 Liebreich introduced chloral hydrate, and ever since, the pharmaceutic industry has worked overtime to produce more and better hypnotic drugs. Almost exactly a hundred years ago barbituric acid was synthesized, and forty years later Fischer & von Mering (1903) introduced the first barbituric acid derivative into clinical practice - veronal or barbitone. The new drug soon gained popularity and acclaim from the medical profession. At the seventy-third annual meeting of the British Medical Association, in 1905, a number of speakers (Brunton, Cushny, Hale-White) stated that it was "... an excellent drug ", the most effective of that group of hypnotics, allied in chemical composition to urea, that had been "... hitherto introduced into practice ", and that in spite of a few cases of idiosyncrasy "... nothing of serious impact has been brought against it yet ". However, the year of the introduction of veronal also brought the publication of the first case of non-fatal poisoning in the German medical press (Gerhartz), shortly afterwards followed by a similar report in Britain. Here a London general practitioner (Clarke, 1904) described the case of a nineteen-year-old married woman who was prescribed 8 gr of veronal for her insomnia; unable to sleep, she took 16 gr on the first night, and 24 gr on subsequent nights. She was first seen by her doctor in a semicoma on 13 December 1903, and she developed a rash on the 16th and a delirium on the 22nd; and though the drug had been officially stopped she was still found taking the drug secretly on 24 December. Between 10 and 24 December she had taken 128 gr. "On each of the subsequent visits," Clarke writes, "the patient implored me to give her some more of the drug, even going so far as to threaten if I did not comply with her wish." This early case is of some interest because it showed quite a number of the possible drawbacks of barbiturate medication and over-consumption, such as the rash, the development of a strong desire to take the drug in increased doses, and a psychotic state.

By courtesy of the honorary editors of the Proceedings of the Royal Society of Medicine.

Similar cases were soon described by others; in reporting one case showing idiosyncrasy by developing an erythematous rash after 8 gr, the practitioner wrote that "... as [veronal] was placed on the market as absolutely safe and without toxic effects, there is a probability that a reaction may set in against its use" (Wills, 1906). The first fatal cases of accidental poisoning were reported by the Registrar-General in 1906 and 1907 (table 1), but the first description in the medical press of a barbiturate death in this country did not appear until 1908 (Parsons), a woman (and the child) dying at the time of her confinement. Parsons complained when describing this case that it "... surely points a moral "; that it could "only be harmful in the end that the public should have such easy access to drugs that are often dangerous and undoubtedly many valuable lives are yearly sacrificed to the casual habit of drug-taking now so prevalent ...", the "vast majority" of people taking the drug without medical advice. Throughout the history of the barbiturates one comes across complaints and controversies later to be echoed with practically any new group of hypnotics (and other drugs affecting the central nervous system) coming on the market.

With increasing consumption of veronal and, later on, of other barbiturates fatalities too became less infrequent (tables 1 and 2), and in the years 1911-1913 veronal occupied the seventh place among causes of death from all poisons (Young), superseding sulphonal when the latter was placed on the Dangerous Drug Schedule. In 1913 an editorial in the British Medical Journal complained that veronal could be freely bought from anyone who chose to sell it, despite increasing numbers of cases of poisoning and despite the "great danger resulting from the acquisition of the drug habit"; shortly afterwards, in April 1913, barbiturates were placed under part II of the Poisons Schedule, and in 1918 under part I (Willcox, 1927). Another B.M.J. editorial in 1918 refuted the view that "there was no such thing as the veronal habit" and stated that it was possible to become addicted to the drug. The first case of addiction to medinal in the country was described in 1926 by Stolkind, in a man whom he had first seen in 1916 in an ataxic and drunken state. The patient refused to give up the drug, taking it for the next eight years in increasing doses three to five times a week, with very detrimental effects on his physical, mental, and moral state. Finally, in 1924, at the age of 68 years, he died from acute Medinal poisoning when, "exhausted and drowsy ", he had taken a large overdose.

TABLE 1

Number of deaths from barbiturate poisoning

 

Suicides

Accidents

 

Suicides

Accidents

 

Suicides

Accidents

Year

M.

F.

M.

F.

Year

M.

F.

M.

F.

Year

M.

F.

M.

F.

          1921 2 3 2 1 1941 18 27 17 21
          1922 1 2 2 1 1942 19 14 13 21
          1923 2 1 5
-
1943 13 22 12 17
          1924 2
-
2 4 1944 13 30 15 19
1905
-
-
-
-
1925 4 5 5 10 1945 20 35 18 22
1906
-
-
-
1 1926 1
-
2 2 1946 24 54 20 34
1907
-
-
1
-
1927 4 6 5 5 1947 39 70 30 47
1908 2
-
2
-
1928 4 7 7 9 1948 73 96 34 41
1909
-
2 6 5 1929 4 10
-
7 1949 81 110 36 51
1910 2 1 7 5 1930 4 4 10 15 1950 122 156 47 80
1911
-
2 8 9 1931 5 7 5 6 1951 100 148 40 77
1912 3 1 9 3 1932 8 11 4 7 1952 157 176 61 79
1913 8 3 10 9 1933 15 10 10 15 1953 124 197 84 85
1914 4 8 8 10 1934 13 16 4 11 1954 16 230 67 117
1915 5 2 9 9 1935 13 26 19 11 1955 167 260 77 115
1916 3 2 2 3 1936 11 21 10 10 1956 229 287 91 132
1917 4 4 4 4 1937 9 16 4 15 1957 215 289 96 112
1918 2 1 4 5 1938 18 21 7 7 1958 260 323 88 117
1919 2
-
1 4 1939 14 29 3 10 1959 234 342 96 136
1920 1 2 2 4 1940 17 35 9 21 1960a 257 388 115 178

Source: Registrar-General's Statistical Review for England and Wales.

Provisional figures.

It was in the late 'twenties that the controversy about the dangers of the barbiturates began to heat up, reaching its climax in the "battle of the barbiturates" in the mid-'thirties. The leading "prosecutor's" role was taken by Willcox, a Home Office toxicologist, whereas the" defence" was mainly led by the psychiatrists Gillespie and Craig. The controversy was fought out at meetings such as those of the Royal Society of Medicine (1927, 1934 a) and of the Society for the Study of Inebriety (1934) - and in the medical press, such as the correspondence column of the Lancet in 1934. The debates seem to have been bitter and passionate, reminding one of the highly emotional conflicts between the "drys" and the "wets" in regard to alcohol. For example, Willcox, opening a discussion in 1934 (c) referred to drugs of addiction other than the "dangerous drugs "

TABLE 2

Barbiturate fatalities recorded in British literature until 1932

Year

Cases

Dose Gr. G.

Suicide, therapeutic, accident or addiction

Drug

Author

1908 1 6O
S.
Veronal
Parsons
1909 1
100 (approx.)
S.
,,
Davies
1909 1
?
?S. ?Add.
,,
Anon
1909 2
?
?S.
,, ,,
1909 1 91
S.
,,
Walker
1912 1
?
-
,,
Anon
1912 1 85 5.40   ,,
McLean
1912 2
252 in 5 days
Add. S.
   
    177
S.
,,
McCrae & Coll
1913 1 150
2 S. 1 ?S. 1 Th. 3 ?Add.
,,
B.M.J. editorial
1913 8
? (1 case: 150 gr.)
S.
,,
Willcox
1913 1 125
Acc.
Medinal
,,
1914 1 120
-
Veronal
Fraser
1914 1 90
?S.
,,
Munro
1914 1 15.50
?
,,
Russell & Parker
1918 2
?
Add.
,,
B.M.J. editorial
1926 1
?
S in 1 case
Medinal
Stolkind
1927 3
200 gr. in 1 case
-
,,
Willcox
1927 1
50-100
Add., S., etc.
Veronal
Young
1927 19
?
?
,,
Willcox
1927 1
?
 
Luminal
,,

Source: R. D. Gillespie: Proc. Roy. Soc. Med. 1934, 27, 508 (with the kind permission of the honorary editors, Royal Society of Medicine).

". . . What I shall tell you is very controversial, but I stand by my statements. I know that what I am telling you is true, and it is not as a result of reading, but from tragedies which have come under my own notice, from many cases which I have seen and treated personally. There can be no doubt that the very large group of barbituric acid derivatives occupies the foremost place amongst the drugs of addiction. The actual danger to the public in this country at the present time from addiction to these drugs is greater than that from any other group of drugs, even including the dangerous drugs, which are controlled . . . by special Acts and regulations. . . . The continued daily use of these drugs in therapeutic doses may cause impairment of speech, ataxic gait, paralysis of the eye muscles, and other motor nerve affections . . . also mental disturbances, such as hallucinations of vision. . . . The need for care in the use . . . cannot be too strongly emphasized. The risk of suicide from accidental or purposeful overdosage is a very real one. . . . The members of the medical profession should exercise care in the prescription of these drugs. . . ."

Willcox asked for further restrictions - a demand which was strictly contradicted by others- but in fact the barbiturates were put on schedule IV by the Poisons Rules, 1935 (see below).

Gillespie 0934 a) disagreed with all points made by Willcox. Searching the world literature, he found, up to until 1932, 157 fatalities, but usually in suicidal cases and not from therapeutic doses as claimed by Willcox (table 2.) The proportion of barbiturate suicides in 1931 - 0.26% of all suicides (table 3) -was" surprisingly small ". Barbiturates were only eighth in the list of twelve poisons listed as means of suicide. As to addiction to barbiturates, he had never seen such a case (1934c). "Withdrawal of barbiturates is not accompanied by the distressing subjective results and objective manifestations that accompany withdrawal of alcohol or morphine" (1934b). Through a former Home Office toxicologist and colleague of Willcox described these drugs as" ... one of the great gifts . . . offered to suffering humanity" (Luff), Willcox (1934b) stopped prescribing barbiturates altogether because of the danger of their abuse "and also for the sake of example to those medical students who follow my teaching ". However, not all his students were impressed with his teaching on barbiturates. Thus Sargant (1958), who was trained at St. Mary's and in the school of Willcox, at a time when" . . . the giving of any barbiturates at all was considered a major therapeutic crime . . . in rebellion against some of the absurdities of my teaching on sedatives [I] was one of the first persons to advocate a much wider general use of the short-acting barbiturates as a routine sedative ". Sargant (1956)relates that in the early 'thirties "... mental hospitals were full with patients tranquil- lized and drugged with bromides in truly enormous doses ", bromide also being, in 1928, the fifth most commonly prescribed drug in general practice. By 1939, however, the recognition of the prevalence of bromide intoxication had helped to push the bromides out and the barbiturates in "... because it was obvious that the dangers [of the barbiturates] had been gready exaggerated by Willcox and others".

TABLE 3

Suicides in England and Wales, 1931

I. According to method used

Suicidal agent

Number

Solid and liquid poisons and corrosive substances
676
Poisonous gas (all but 12 coal gas)
1 451
Hanging and strangulation
736
Drowning
913
Firearms
267
Cutting and piercing instruments
611
Jumping from high places
167
Crushing
263
Other means
63
TOTAL
5147

II. Relative frequency of suicides from various poisons

Suicidal agent

Number

Suicidal agent

Number

Lysol
248
Barbituric acid
13
Carbolic acid
63
Aspirin
10
Potassium cyanide
60
Sulphuric acid
7
Hydrochloric acid
31
Strychnine
5
Prussic acid
29
Opium, laudanum,
 
Ammonia
28
morphia
3
Oxalic acid
27
Other poisons
106
   
TOTAL
630

III. Suicides by drugs belonging to the barbituric acid group

Suicidal agent

Number

Veronal
5
Medinal
2
Somnifaine
1
Dial
1
Allonal
1
Soneryl
1
Aspirin and medinal
1
Adalin8
1
TOTAL
13

Source: Registrar-General's Statistical Review /or England and Wales.

Adalin is today a non-barbiturate preparation.

With the waning popularity of the bromides, consumption of barbiturates rose rapidly." Perhaps the attack on [the barturates] was too intemperate; certainly the defence seems to have been successful," commented a Lancet leader in 1947, pointing out that in 1947 these drugs comprised 4.6% o National Health Insurance prescriptions, as against only 1.1% in 1934. The Lancet added the warning that because of the risk of accidental and suicidal poisoning barbiturates were ". . . clearly not the harmless sedatives that their present wholesale distribution might suggest. Constant watch must be kept for chronic intoxication." There was, however . . . ."fortunately . . . little risk of addiction ", a view which concurred with that of Sargant in the first edition of his book (with Slater) on Physical Methods of Treatment in Psychiatry (1944). The British Medical Journal in an answer in 1947 to a question about chronic barbiturate intoxication found ". . . it . . . remarkable how rare instances of chronic barbiturate intoxication are . . ." and made no mention of the risk of an abstinence syndrome.

In the 1950s, editorials in the leading medical journals began to return to the theme more frequently. In the Lancet a 1951 editorial pointed to the increase of deaths from barbiturate poisoning, proportionate to the rise of manufacture of the drugs, and an annotation in 1952 warned doctors against employing them ". . . indefinitely as placebos ". Again in 1953, the Lancet stressed ( a) that in spite of ". . . the freedom with which barbiturates are prescribed" even . . . in more or less normal doses these drugs are.., not . . . devoid of serious risk ", and ( b) that they were "perhaps used too lightly ". The Practitioner, in an editorial on "The Barbiturate Menace" (1953), warned that the barbiturates were ". . . much too dangerous to be used as a substitute for the careful search for the cause of the patient's insomnia and worry which used to be the hallmark of the old family doctor"

A year later the Lancet, in a leading article entitled" Shadow over the barbiturates" (1954), saw some ". . . evidence that the high noon of their popularity is passing ". Though there was ". . . little fear of toxic effects . . ." from therapeutic amounts, the barbiturates were ". . . true drugs of addiction ", a risk which was ". . . the least appreciated and ... most sinister ". Likewise an editorial in the British Medical Journal in 1954 - after a remark that "the sinister potentialities of the barbiturate hypnotics has been a perennial subject of debate in the medical press for the past thirty years" - agreed that apart from the relatively slight tendency to increase the dosage compared with the behaviour of morphine addicts, barbiturates otherwise fulfil all criteria for drugs of addiction ".

Yet all these denunciations failed to stem the rising tide, and in 1959 a British Medical Journal editorial, entitled" Barbiturate deaths increasing ", had to report a further "great increase of deaths in England and Wales from ... poisoning in the last few years ". Sixty years after the introduction of the first barbituric acid derivative into therapeutics and more than a quarter of a century after the" Battle of the barbiturates "was fought, they still seem to enjoy as great a popularity among the medical and the lay public as ever - as evident in rising consumption (cf. Chapter II, page 23), rising figures for suicides and fatal accidents as a consequence of barbiturate ingestion (cf. Chapter III, I, p. 25), a rise in suicide attempts (cf. Chapter III, 2, p. 28) and very likely also a rising prevalence of emotional and physical dependency on these drugs (Chapter IV, p. 32).

II. Barbiturate consumption

Until 1913 there were no restrictions on prescribing of barbiturates, anybody being able to buy and apparently also to sell barbiturates: in fact - as Willcox said in 1913 - veronal had for six years been commonly taken for insomnia without a doctor's advice or prescription. From 1913 onwards, the sale of barbiturates was subjected to more and more restrictions. Today, all barbiturates are schedule IV poisons; as such the prescription must state in every case the patient's name and address, the total quantity of the preparation, the date, and the prescriber's full signature. In spite of such restrictions, the consumption of these drugs over the years has risen steadily and considerably, even during the past decade and in the face of competition from tranquillizers and non-barbiturate hypnotics.

Barbiturate consumption was one of many questions discussed in the 1961 Report of the Interdepartmental Committee on Drug Addiction. Among other tasks, in 1958 this committee was asked by the Government "to consider whether any revised advice should also cover other drugs [than morphine and heroin] liable to produce addiction or to be habit-forming ". From a sample of prescriptions issued under the National Health Service (N.H.S.) the total quantity of barbiturates prescribed annually in the 1950s was estimated, showing it to have expanded "both progressively and substantially" (table 4). Thus the amount prescribed in 1959 (162,000 lb, or 73,000 kg) was almost twice as high as the 1951 figure (90,000 lb, or 41,000 kg). The Committee found this trend confirmed by production figures, though these naturally include exports as well.

The export of barbiturates in 1960 amounted in value to ?623,519, and in the first nine months of 1961 to ?465,243 (Davis, personal communication). The exported products are thus only a small percentage of the gross turnover. The 1958 census of production shows that during that year 855,000,000 barbiturate tablets were sold for a manufacturers' price of ?780,000.

An analysis of N.H.S. prescriptions according to therapeutic classification showed the proportion of barbiturates over the past few years to be roughly 7% of all prescriptions issued (tables 5 and 6); barbiturates, other sedatives and hypnotics, analgesics and antipyrotics (but excluding tranquillizers) taken together formed nearly 20% of the total of prescriptions (table 5). As for tranquillizers, the Committee believed that their prescription had also increased, although the scale of their use in general practice could not be ascertained. As to stimulants, amphetamines and phenmetrazine were found to account for over 5? million out of 214 million N.H.S. prescriptions - approximately 2?% - in 1959.

TABLE 4

Estimated total quantities of barbiturates prescribed by general practitioners in the National Health Service

Year

Total quantity per year (Thousands of lb)

1951 90
1953 81
1954 102
1955/6
108
1957 123
1959 162

Source: Report of the Interdepartmental Committee on Drug Addiction. London, 1961.

TABLE 5

Estimated prescribing, by general practitioners in the National Health Service, of certain drugs affecting the central nervous system

 

Percentage of total prescriptions

 

Year

Barbiturates

Other sedatives and hypnotics

Analgesics and antipyretics

Total

1957 7.01 3.17 10.16 20.43
1958 7.00 3.00 9.06 19.06
1959 6.65 2.80 9.38 18.83

Source: Report of the Interdepartmental Committee on Drug Addiction. London, 1961.

TABLE 6

Statistics obtained from an examination of a sample of prescriptions dispensed by a representative sample of chemists

   

Barbiturates (proprietaries and non-proprietaries)

Sedatives and hypnotics (other than barbiturates)

Year

Prescriptions dispensed in England and Wales (Millions)

Percentage of prescriptions

Percentage of ingredient cost

Percentage of prescriptions

Percentage of ingredient cost

1958 203.4 7.0 2.6 3.0 1.3
1959 214.0 6.7 2.5 2.8 1.4
1960 218.7 6.8 2.3 2.5 1.1

Source: H. Davis, personal communication, 1961.

The estimates obtained by the Committee confirm the findings of several investigations in recent years and show the marked increase in consumption of barbiturates. Thus in 1934 barbiturate prescriptions formed only 1.1% of the National Health Insurance prescriptions, rising to 4.6% in 1947; based on cost of ingredients of prescriptions the increase was from 1.5% in 1934 to 5.4% in 1947 ( Lancet, 1947). Before the Second World War, 40% of medicines prescribed to panel patients in Britain contained bromides (Mayer-Gross et al.) but with the growing recognition of the drawbacks of the bromides, the barbiturate box grew more and more in popularity, gradually ousting the bromide bottle. The manufacture of barbiturates in 1946 - when the total sales output of these drugs in Great Britain was 71,000 lb (Locket, 1957a) - was double that in 1938; it had become twice as high again by 1950 ( Lancet, 1951). Barbiturate consumption increased greatly after the introduction of the N.H.S. in 1948, from 45,000 lb sold in the home market in that year to 90,000 lb in 1951 (Hunter, 1957). During the following years, according to estimates made by the Ministry of Health, the amount of barbiturates prescribed annually varied between 80,000 and 100,000 lb (Brooke).

Surveys of N.H.S. prescriptions in Scotland and Great Britain in 1949 and 1951 respectively showed that sedatives and hypnotics made up about 15% of all drugs prescribed, barbiturates alone accounting for 9.4% in the 1949 sample (Dunlop et al. 1952) of over 17,000 N.H.S. prescriptions. In England and Wales analyses carried out by the Ministry of Health of monthly samples of over 100,000 N.H.S. prescriptions were carried out regularly; such a sample is less than 1% of the total monthly number of N.H.S. prescriptions, but is held to be fairly reliable (Davis), and the N.H.S. handles the prescriptions for over 90% of the population. In January 1953 barbiturates accounted for 6.30 %of all prescriptions, other sedatives and hypnotics adding another 2.53% (table 7). In October 1954, 8.8% of prescriptions were for barbiturates or preparations containing them, barbiturates being the sole or principal therapeutic agent in 6.4% (Brooke). In October 1955, out of 106,000 prescriptions over 10,000 (9.6% were for barbiturates or preparations containing them (Davis). The estimated cost of barbiturates for the domiciliary service only in the N.H.S. at the time was over ?l1/2 million a year, representing not much less than a thousand million doses of a barbiturate, or 100,000 lb in weight (Davis).

Recent analyses of N.H.S. prescriptions in Scotland and Northern Ireland gave similar proportions of barbiturate prescriptions as the ones in England and Wales. Thus a recent analysis of samples of over 30,000 N.H.S. prescriptions (out of a total of approximately 22 million per year) gave the percentage for barbiturates as being from 8 to 8.5% (Johnston). Dunlop's sample from Scotland in 1949 had given a slightly higher barbiturate proportion than other investigations, possibly influenced by the absence of a predominently rural area in his sample (Brooke); and it is interesting in this connexion to quote the findings recently obtained in Northern Ireland: a sample survey of prescriptions issued by general practitioners showed the barbiturate proportion to be about 9%, the figure being slightly higher in Belfast (9.6%) than in the rural areas (8.2%).

TABLE 7

Therapeutic classification of prescriptions dispensed in January 1953

Therapeutic category

Percentage of prescriptions

Antibiotics
4.19
Sulphonamides
4.14
Barbiturates
6.30
Sedatives and hypnotics (other than barbiturates)
2.53
Insulin
0.30
Sex hormones
0.79
Other hormones
0.32
Vitamins (except B 12 prescribed singly
1.70
Polyvitamin preparations
1.63
Haematinics (including vitamin B 12)
0.19
Tonics
5.25
Analgesics and antipyretics (excluding dangerous drugs)
9.89
Dangerous drugs
0.86
Medicaments for external use
11.73
Cardiac preparations
1.88
Anti-histaminics
1.28
Laxatives, purgatives and antacids
10.24
Cough preparations
17.20
Sera and vaccines
0.11
Asthma preparations
2.61
Miscellaneous
10.57
Unclassified
6.30

Source: "The Prescribing of Barbiturates" , by Harold Davis. Brit. J. Addiction, 1957, 53, 102.

Two monthly surveys taken early in 1959 both showed barbiturate prescriptions to number about 6% of the total of N.H.S. prescriptions. These surveys also showed that among individual barbiturates phenobarbitone was the most popular one, as shown in table 8. (This table has been prepared by calculating the average from both surveys, which, however, had very similar findings.) This corresponds to a 1955 estimate according to which 450 million doses of phenobarbitone were dispensed yearly, as against roughly equal amounts of butobarbitone and amylobarbitone (about 175 million doses each), followed by pentobarbitone and then by quinalbarbitone (Davis).

The finding that in Great Britain phenobarbitone prescriptions are two to three times more numerous than for each of the shorter-acting barbiturates is interesting from several aspects. On the one hand, phenobarbitone is used regularly in epilepsy; an example of what has sometimes been called the " correct " employment of these drugs (Locket, 1957a). The number of those epileptics in Great Britain who are receiving treatment for their condition has been estimated as 125,000 (Garland). Among barbiturates, furthermore, phenobarbitone is least addictive. On the other hand, " the practice of daytime sedating with phenobarbitone " has been described as "more controversial " than giving a hypnotic for a few nights (Dunlop, 1957). Phenobarbitone hypersensitivity ( Lancet, 1953b) is also more common than with other barbiturates and the percentage of patients who develop skin reactions with phenobarbitone has been estimated as 1% to 3% (Locket, 1957 b).

Is the amount of barbiturates prescribed also the amount taken? Camps (1961), a Home Office pathologist stated that in the houses of people who committed suicide " fantastic quantities " were found; " it is nothing to have 1,000 gr of rapid-acting barbiturate in one household ". He estimated that at least 10% was lying around unused in people's houses, and he was in no doubt that promiscuous prescribing was an incentive to take overdoses. Locket (1957a), too, stated that many patients keep a large supply " in stock " leaving the tablets carelessly within reach of young children. In 1952 (b) the Lancet had stated that commonly such prescriptions were for large quantities; the average sizes of barbiturate prescriptions today are estimated in table 8.

TABLE 8

Barbiturate prescriptions, February-March 1959

Drug

Percentage

Tablets per prescription

Phenobarbitone
31 60
Soneryl
14 44
Amytal
20 49
Drinamyl
10 48
Nembutal
6 40
All others
19  
TOTAL
100  
(Approximately 6 % of all prescriptions)
   

Is there much " illegitimate " purchase of these drugs? The Interdepartmental Committee could find no definite information about this point. According to Locket (1957a) the supply of barbiturates to laymen comes with few exceptions from medical prescriptions. However, one occasionally encounters people who have managed to get such prescriptions from various doctors over a considerable length of time (Hunter & Greenberg, Armitage & Sim).

Are the vast amounts of barbiturates prescribed necessary for the patient's treatment or are they " excessive " ? Locket (1957a) has " no doubt ... that the barbiturates are misused on a vast scale. ... "According to Hunter (1957), " barbiturates are ... over-prescribed and over-consumed by patients with functional troubles "; they are used "... almost as a placebo, often to assuage the doctor's own anxieties ". Confronted with a large number of patients with functional troubles, insomnia (estimated by Pai to be the main or only symptom in at least one-third of patients in general practice), etc., all demanding relief, the harrassed practitioner lacking psychiatric education, . . . and time to inquire more fully into the patient's history often can do little else but to give drugs to sedate or stimulate. Various annotations in the Lancet in recent years complained that doctors seemed " sometimes too prone to employ [barbiturates] indefinitely as a placebo (1952), that they were ' nowadays ' perhaps used too lightly " and that such ". . . easy going practice is apt to lead to dependence and addiction " (1953 a), and that " perhaps too little account is taken of the risk of occasional disasters " (1953 b).

The general practitioner could argue that many of his neurotic patients referred to a psychiatric in-patient or out-patient department are probably sooner or later referred back to him with the suggestion to continue treatment with tranquillizers or barbiturates. A brief questionnaire relating to barbiturate prescribing was sent a few months ago to a few practitioners in and around London for the purposes of the present paper. Added to his replies was one practitioner's complaint that barbiturates were often started by his patients whilst they were in hospital; after discharge they were reluctant to discontinue the tablets and were resentful if a prescription was refused. Incidentally, three out of four doctors in general practice among the twenty-four questioned stated that the introduction of tranquillizers and non-barbiturate hypnotics had made little or no difference to barbiturate prescription in their practice, but the sample is clearly much too small to draw any conclusions. The general practice prescriptions' analysis early in 1959 which was referred to above (p. 24) found tranquillizers to form about 1?%.

In regard to mental hospital practice the Interdepartmental Committee found that the amount spent on one particular tranquillizer by nine selected mental hospitals had risen about ten-fold in the past five years. This could be expected to reduce at least the daytime sedation by barbiturates just as the introduction of a number of non-barbiturate hypnotics should influence the amount of barbiturates used at night-time. Inquiry at a large mental hospital in the London area with which the present author is associated showed that the amount of barbiturates used in 1960/61 was considerably less than in 1952/53 (because of its use in epileptics phenobarbitone was excluded from the comparison); the use of amylobarbitone had been halved. Paraldehyde, too, was employed much less often, the 1960/61 amount being one-sixth of the 1952/53 figure. On the other hand, non-barbiturate hypnotics were used a great deal at this hospital, which because of its interest in addicts cannot be regarded as representative in this respect.

Today, twenty-five years after the " battle of the barbiturates ", there still remains a wide divergence of views concerning these drugs. This is well illustrated, for example, in a recent issue of the Practicioner (1960). Of two contributors, both from Scotland, Wayne felt that, judged against the wide extent of their consumption, barbiturates were relatively non-toxic, had few side-effects and only a small incidence of addiction; they therefore ought not to be replaced by non-barbiturates and tranquillizers. The other author, Macgregor, preferred non-barbiturates, which, though less potent, were potentially less dangerous than the barbiturates. Many authors condemn the common use of barbiturates in functional troubles as merely " symptomatic " treatment (e.g., Burn). Others, however, point to their value "... in relief of suffering, restoring social adaptation, and as an adjuvant to specific therapy" (Mayer-Gross et al.); or in forming a kind of "... psychiatric first aid in preventing aggravation of an acute neurosis or an acute illness from drifting into chronicity " (Sargant 1956).

During the recent inquiries carried out by the Intedepartmental Committee on Drug Addiction most witnesses felt thet drugs acting on the central nervous system were nowadays used excessively without producing records to support their contention. The Committee concluded that there was "... scope and need for operational research into the prescribing pattern in this country, with particular reference to habit-formings drugs ", but aware of the difficulties in defining clearly the indications for their use or in estimating the benefit derived from central nervous system drugs - the question whether the amounts prescribed were " excessive " could not be answered. Among possible benefits, for example, "... the therapeutic use of barbiturates and tranquillizers may well have saved many people from suicide ".

What is the cause for the vast increase of the use of drugs affecting the central nervous system? This question has already been touched upon in the preceding discussion. In the views of the Committee, intensive advertising may be one important factor, whereas the alleged role often ascribed to the height- ened rush and anxiety of modern living is regarded as ". . . based on assumption more than fact ". A possibly very influential factor may be today's materialistic attitude" towards the rapeutics leading to a search for, and application of, a specific chemical corrective aimed not only at eradicating major and minor ills, but also if possible at positively enhancing health. A patient's " real need " may be a discussion of his psychological troubles with the doctor, but in its place he often leaves the surgery with a pill only.

III. Acute barbiturate poisoning

1. SUICIDES AND ACCIDENTAL DEATHS

In this country the barbiturates are today surpassed only by coal gas in the frequency of causing suicidal and accidental death. In the period of 1911-1913 veronal - the first barbiturate on the market - occupied the seventh place among causes of death from all poisons (Willcox, 1927). In 1931 barbiturates were eighth on the list of drugs causing fatal poisoning (table 3), all of which they have now left behind in this respect, including aspirin and lysol.

The continual increase of suicidal and accidental deaths caused by the barbiturates proceeded in some respects parallel to their rate of consumption. Thus the output of barbiturates in Great Britain increased four-fold between 1938 and 1950, whereas deaths from barbiturate poisoning multiplied by four and a half times from 1938 to 1948 (Lancet, 1951). As shown in table 1, for England and Wales, in 1910- seven years after their introduction into therapeutics - barbiturates caused 3 suicides and 12 accidental deaths (a total of 15), in 1920 3 suicides and 6 accidental deaths (total 9), in 1930 8 suicides and 25 accidental deaths (33), in 1940, 52 suicides and 30 accidents (82), in 1950 278 suicides and 127 accidents (405), and in 1960 645 suicides and 293 accidental deaths (a total of 938 deaths). Barbiturate suicides have increased by approximately twelve times in the past twenty years, accidental deaths about ten-fold. The total number of deaths from barbiturates is now more than twice what it was in 1950, more than eleven times that of 1940, nearly thirty times the figure of 1930 and more than a hundred times that of 1920. The total annual figure of barbiturate deaths having almost reached a thousand means that there are 18 people dying each week in England and Wales from barbiturate poisoning.

TABLE 9

Crude suicide-rates per million living for (1) all agents, (2) barbiturate poisoning, 1939-1960

 

All agents

Barbiturates

Year

M

F

M

F

1939 168 78 0.71 1.35
1940 159 75 0.93 1.62
1941 135 62 1.04 1.25
1942 125 62 1.13 0.65
1943 134 63 0.80 1.02
1944 135 58 0.80 1.39
1945 136 66 1.22 1.61
1946 144 75 1.28 2.46
1947 137 76 1.99 3.16
1948 145 79 3.58 4.29
1949 147 75 3.94 4.88
1950 136 70 5.76 6.88
1951 135 72 4.75 6.50
1952 132 68 7.44 7.70
1953 142 76 5.85 8.60
1954 149 81 7.56 10.01
1955 143 84 7.81 11.28
1956 149 90 10.64 12.40
1957 146 92 9.93 12.43
1958 146 91 11.96 13.82
1959 142 89 10.69 14.55
1960 139 87 11.64 16.38

Source: Registrar-General's Statistical Review for England and Wales.

Provisional figures.

TABLE 10

Crude suicide-rates per million living, according to the method used, 1912-1960

 

Poisoning: analgesics and soporifics

Poisoning: domestic gases

Poisoning: other solids and liquids

Hanging and strangulation

Drowning

Firearms and explosives

Cutting and piercing instruments

Jumping from high places

Other and unspecified means

Year

M

F

M

F

M

F

M

F

M

F

M

F

M

F

M

F

M

F

1912 3 1 5 2 17 12 42 8 28 16 15 1 30 5 3 1 10 3
1914 2 1 5 2 15 11 42 8 30 16 17 1 33 7 4 1 9 2
1916 1 0 4 2 9 6 34 8 22 17 10 0 28 5 4 2 7 1
1918 1 1 7 4 6 6 29 8 23 19 10 1 29 6 5 2 6 2
1920 1 0 9 6 10 6 32 8 27 20 13 1 32 5 3 2 8 2
1922 1 0 18 9 13 10 37 8 29 17 14 0 32 6 4 3 8 1
1924 0 0 20 10 11 10 31 7 27 16 10 0 29 5 4 2 9 2
1926 1 0 31 17 17 14 36 7 30 17 11 0 28 5 4 3 7 2
1928 1 1 35 20 19 13 37 8 32 17 12 1 30 5 4 3 9 2
1930 1 1 44 24 20 14 33 6 29 16 13 0 28 4 6 3 10 1
1932 1 1 60 30 21 16 38 6 35 16 13 0 28 4 5 4 9 1
1934 2 1 60 32 21 14 33 7 28 13 12 1 25 4 6 3 12 2
1936 1 2 54 32 15 10 34 8 25 13 12 0 20 3 5 3 9 2
1938 2 2 59 36 16 11 32 8 25 15 12 1 18 3 6 3 8 2
1940 2 4 52 35 11 9 30 7 20 11 14 1 17 3 6 3 5 2
1942 2 2 38 29 7 7 24 5 16 11 14 1 13 2 5 3 6 2
1944 2 3 46 28 7 5 24 5 19 11 12 1 15 2 4 2 6 2
1946 3 5 46 36 7 5 32 8 18 12 11 0 15 3 4 3 7 3
1948 7 10 54 40 6 3 29 6 17 11 10 1 9 2 5 3 7 3
1950 8 11 49 33 5 4 27 6 16 10 9 0 9 2 4 2 8 2
1952 11 13 52 34 5 3 25 5 13 8 8 0 7 1 3 2 7 1
1954 10 14 64 47 4 2 29 6 14 8 9 0 8 1 4 2 7 2
1955 11 15 59 47 4 2 27 5 13 9 10 0 8 1 3 2 6 2
1956 15 17 63 51 5 2 26 5 13 9 10 0 6 2 3 2 7 1
1957 14 17 65 52 4 2 26 6 12 9 9 1 6 1 3 2 5 2
1958 15 19 67 51 3 1 22 5 12 9 9 0 6 1 3 2 6 2
1959 15 19 65 50 3 2 20 5 11 8 10 0 5 1 3 2 7 1
1960
15 21 64 46 3 2 19 5 12 8 8 0 5 1 3 2 7 2

Source: Registrar-General's Statistical Review for England and Wales.

a

Provisional figures.

As far as the total suicide rate from all agents together is concerned (table 9) there has been relatively Little change over the years, apart from a recent slight increase among women. Thus the suicide rates in England and Wales per million living were in 1940, 159 for men, 75 for women; in 1950, 136 and 70 respectively and in 1960, 139 and 87. This is in contrast to the sharp rise of barbiturate suicide rates during the same period: in 1940, 0.93 for men and 1.62 for women; in 1950, 5.76 for men and 6.88 for women; and in 1960, 11.64 for men and 16.38 for women. Of 5,147 suicides in 1931, 12 (0.23%) were due to barbiturates (table 3); of 4,469 suicides in 1951, 248 (5.5%); of 5,262 in 1956, 516 (10.2%); and of 5,112 in 1960, 645 (12.6%).

As shown in table 10, over the years there has been a gradual reduction of the more violent methods for suicide and an increasing preference for the more passive ways (Booke). Not only deaths from barbiturates went up, but also those from aspirin: thus there were 20 suicidal and accidental deaths due to aspirin in 1930, 24 in 1936, 90 in 1946 and 145 in 1954.

Thus no more people commit suicide nowadays than formerly, but there has been a change in fashion in regard to methods preferred. Barbiturates are a much less lethal suicidal tool than many of their competitors, including coal gas; the barbiturate mortality rate among hospital admissions has been calculated by Locket (1957) as up to 7% or 8%. One might therefore argue that the preference for the use of barbiturates as suicidal agents has in fact saved many people from death, as they would otherwise have employed more lethal tools (Stenge, 1958). According to the statistical branch of the Metropolitan Police (1950) in London, drugs proved to be the least effective method of committing suicide, and it was felt that "... the steady post-war reduction of the percentage of successful suicides (in London) was due to the increasing commonness of the use [of barbiturates]" in preference to other methods. On the other hand, barbiturates, readily available everywhere, may offer an easy and painless way out of difficulties, tempting people during depressive episodes, etc., or even to a sudden gesture made on the spur of the moment. Accidental deaths would be cut down, were such drugs less widely available, but people who are intent on committing suicide will probably not be kept off it by the lack of a particular agent; many patients seen in hospital after a suicide attempt give a history of previous attempts by other means, or may in some cases commit suicide later using a different agent. It is noteworthy that whereas more men commit suicide than women - e.g., in 1960, 3,058 men as against 2,054 women - the reverse is true in the case of barbiturate suicides, which in 1960 numbered 288 among women and 257 among men. The ratio of barbiturate suicides to total suicides for 1960 was 8.4% among men and 18.4% among women - i.e., more than twice as high. On the other hand, coal gas has over the years been the more popular method of committing suicide among men. As shown in table 10, the rates for suicidal gas poisoning were - per million living - in 1938, 59 and 36 for men and women respectively; 64 (men) and 46 (women) in 1946; 46 (men) and 36 (women) in 1954, and 64 for men and 47 for women in 1960. However, Locket (1957 a), who strongly disagrees with the suggestion that the use of barbiturates for suicidal purposes was advantageous because of the avoidance of more lethal agents, thinks that it would be "... even more correct ... to suggest that the high (and rising) incidence of suicide in women is because of their preference for the comfort of the barbiturates as against more unpleasant methods ".

The sex-age suicide rates from barbiturate poisoning (table 11) show over the years an increase in all age groups; e.g., in young adults from 1.7 among men and 0.5 among women in 1942, to 9.6 and 10.4 respectively in 1960; in the middle-aged and the elderly from 1.6 and 0.6 respectively in men in 1942 and 21.9 in 1960; and in women from 1.5 and 0.4 in 1942 to 35.4 and 24.1 among the middle-aged and the elderly respectively in 1960. Brooke (1956) points out that the increase in rates is thus more pronounced from the age of the climacteric onwards, at a time where one might find cases of insomnia and depression for which barbiturates might be prescribed. Anyhow, barbiturate poisoning carries a much higher mortality rate in both men and women beyond the age of 50 (Locket, 1957 b); this elder age-group, though accounting for no more than one third of hospital admissions, provided more than half of the deaths.

TABLE 11

Suicide-rates per million living, from barbiturate poisoning, by sex and age, 1942, 1948, 1954, 1959 and 1960

 

15 - 44

45 - 64

65 and over

Year

M

F

M

F

M

F

1942 1.7 0.5 1.6 1.5 0.6 0.4
1948 3.2 3.3 7.3 8.6 5.1 6.3
1954 6.3 5.4 14.9 18.1 12.8 24.4
1959 6.6 9.0 23.6 28.4 21.2 26.5
1960a 9.6 10.4 21.9 34.5 21.9 24.1

Source: Registrar-General's Statistical Review for England and Wales.a Provisional figures.

The proportions of suicides and accidental deaths from barbiturate poisoning in recent years are - according to the Registrar-General's figures for England and Wales - both among men and women more than two suicides for every accidental death (table 1). Until 1930 more accidental deaths were recorded from barbiturate poisoning than suicides, which began to preponderate only from that time on. In Scotland over the past fifteen years as a whole (table 12) accidental deaths were still slightly more numerous both among men and women than suicides (women outnumbering men both among suicidal and accidental deaths from barbiturate poisoning, as in England and Wales). England's total rate of accidental poisoning has been stated to be "disgracefully high" and Scotland's rate to be" the highest in the world" (Simpson, 1960). (As seen in table 13, in 1953 the largest proportion of such accidental deaths in Scotland was caused by carbon monoxide and domestic gas and to a much lesser extent by barbiturates.) Suicide rates in England have been described as" comparatively low" and as" even lower" in Scotland by Stengel, on the basis of suicide rates in 1952 which were in England and Wales 13.2 and 6.8 for men and women respectively and in Scotland 7.7 and 3.4 per 100,000. However, it is often very difficult to differentiate between cases of deliberate and cases of accidental poison-ing; thus in "automatism" a previous dose of barbiturates may have put the individual into an intoxicated state in which he then continued to take further tablets. In coroners' courts it was found that in such cases the initial doses were taken to procure sleep and not with suicidal intent (Camps, 1957). Locket (1957 a) believes automatism to be rare (in our experience it may not be uncommon in alcoholics who also take barbiturates), and from his experience with hospital admissions for barbiturate poisoning he estimates the incidence of accidental non-lethal poisoning as less than 15%. He therefore feels that the Registrar-General's figures underestimate the true proportional contribution of (deliberate) suicides in fatal cases of barbiturate poisoning.

TABLE 12

Accidental and suicidal deaths in Scotland

 

Accidental

Suicidal or self-inflicted

Year

M

F

Total

M

F

Total

1946 1 5 6 1 5 6
1947 2 1 3 3 3 6
1948 3 5 8 4 6 10
1949 8 11 19 3 2 5
1950 11 9 20 6 8 14
1951 8 14 22 8 3 11
1952 13 17 30 11 10 21
1953 11 24 35 9 12 21
1954 22 27 49 6 11 17
1955 8 20 28 13 30 43
1956 13 18 31 11 18 29
1957 23 18 41 15 16 31
1958 8 25 33 23 32 55
1959 17 23 40 15 26 41
TOTAL
148 217 365 127 182 309

Source: Records of the Registrar-General. There is a small element of approximation because it is occasionally difficult precisely to classify a reported death where drugs are involved.

TABLE 13

Deaths from poisoning in Scotland, 1953

Agent

Accidental

Suicidal

Barbiturate
35 21
Salicylates
3 2
Ferrous sulphate
2
-
Carbon monoxide and domestic gas
221 132

Source: Registrar-General's report.

Pills left within easy reach of children may lead to accidental overdosage. Thus Ormiston (table 14) found among 240 children who were admitted to two general hospitals between 1949 and 1958 (and of whom 80% were aged three years or under) because of accidental swallowing of noxious substances, fourteen cases of barbiturate poisoning. Among the total there were two deaths, neither of them among the barbiturate cases. In England and Wales barbiturates caused seven deaths in children under five years of age between 1940 and 1949, aspirin and salicylates being responsible for five times this number. More recently, Locket (1957 b) found among 2,300 hospital admissions for barbiturate intoxication, 1948-1951, forty-eight cases of children up to the age of four years, of whom two died.

TABLE 14

Poisonous substances ingested-by 240 children (1949-58)

Type of poison

Number of cases

Medicaments
 
Salicylates
52
Barbiturates
14
Iron pills and mixtures
14
Others
46
TOTAL
126
Industrial and household products
 
Paraffin
35
Turpentine
15
Others
41
TOTAL
91
Toadstools and berries
23

Source: The Practitioner, April 1959, vol. 182, p. 410, " Emergencies in Infancy and Childhood ", by G. Ormiston, M.D., F.R.C.P.Ed., by kind permission of author and publishers.

In Northern Ireland, suicide is less common than in England and Scotland, with a rate of 3 to 4 per 100,000 of the population per year. Barbiturate deaths are relatively uncommon; e.g., only two deaths were recorded on certain returns in 1961 until November. According to Locket's investigations (1957 b) the barbiturate mortality rate in Northern Ireland in 1948/1951 was 2%, compared to approximately 7% in Great Britain; one barbiturate death occurred among 59,000 general admissions to hospitals in Northern Ireland, compared with 1 death among approximately 12,000 general admissions in Great Britain.

Though more commonly suicidal or accidental, barbiturate deaths may rarely be also homicidal in nature (Simpson & Molony). In the view of Camps (1957) the use of barbiturates for homicidal purposes in England is" probably not very common. It is certainly not often detected, although this may be because it is not suspected ", but even when there are strong grounds for suspicion proof can often not be obtained, "because the rapid acting preparations are broken down and may not be recovered on analysis ".

2. ATTEMPTED SUICIDES

Until a few months ago, attempted suicide in England and Wales constituted a criminal offence which it has ceased to be under the provisions of the Suicide Act, 1961. In practice, however, most attempted suicides were never reported to the police; e.g., of 55 attempted suicides arriving at Guy's Hospital in London in 1957 no more than 13 were " known " to the police (Wood side). There are thus no reliable official statistics about the number of attempted suicides. A Ministry of Health memorandum in 1961 estimated the annual figure to be 30,000, whereas Home Office statistics in 1959 reported 4,980 cases of attempted suicide as" known "to the police (Bleg, 1960 b). Certain official statistics, such as those of the London Metropolitan Police (table 15) do not show a marked difference between the numbers of suicides, and officially "known" attempted suicides, but clearly the latter are several times as common as the suicides. How much more common can only be guessed: the two figures quoted above would give a ratio of 6:1, which would correspond to American estimates quoted by Stengel. Should a similar relationship exist in the case of barbiturate suicides and attempted suicides, the barbiturates would have been responsible for nearly 4,000 suicidal attempts in England and Wales in 1960, for nearly 3,500 in 1959 and 1958, for about 3,000 in 1957 and 1956, and for about 2,500 in 1955.

TABLE 15

Suicides and attempted suicides, 1933-1937 and 1952-1954, London Metropolitan Police area

 

Suicides

Attempted suicides

 

Year

Male

Female

Total

Male

Female

Total

Total suicides and attempts

1933 697 326 1023 413 393 806 1829
1934 705 315 1020 394 391 785 1805
1935 568 343 911 351 359 710 1621
1936 591 311 902 338 366 704 1606
1937 617 373 990 362 379 741 1731
1952 451 250 701 437 524 961 1662
1953 469 314 783 483 555 1038 1821
1954 526 329 855 482 562 1044 1899

Source: Metropolitan Police Statistical Department.

TABLE 16

Poisoning by barbiturates

Cases discharged from or dying in N.H.S. hospitals in England and Wales

Year

Estimated number

Year

Estimated number

1949 2 800 1956 6 060
1951 3 950 1957 6 160
1953 4 200 1958 6 620
1955 5 640 1959 7 520

Source:Registrar-General's hospital in-patient inquiry.

However, compared with other agents frequently used for suicidal purposes, barbiturates are not particular lethal. Locket (1956) reported a mortality rate of 1.8% for a series of 1,000 cases of barbiturate poisoning treated in his specially equipped centre for the treatment of such cases. In such special centres one would expect barbiturate poisoning cases to do better than when treated in the average hospital, but Locket (1959) states that only 1.5% to 8% of cases of barbiturate poisoning seen in hospital are fatal; to get the total incidence of poisoning one would thus have to multiply the number of fatalities by a factor lying between 10 and 50.

From samples statistics of hospital patients (treated for barbiturate poisoning in N.H.S. hospitals) made by the General Register Office, estimates have been arrived at (table 16)which are about twice as high as the guesses made above, but include the cases of" accidental poisoning ", which in Locket's view are only a small proportion of the total. To these cases one would have to add those which have never reached hospital, being either treated at home or not recognized. Including the barbiturate fatalities which occurred outside hospitals a figure of 10,000 cases of barbiturate poisoning in 1959 would probably be an under-estimate (approximately 200 per million of population living).

Barbiturate poisoning is not equally prevalent in different parts of the United Kingdom. Locket (1957 b) gives the number of cases of acute barbiturate poisoning per 1,000 general hospital admissions in Great Britain as a whole as 1.21 (England 1.14, Scotland 1.65, Wales 0.69) and in Northern Ireland as only 0.81.

Mention was made above of the statistics of suicides and attempted suicides in London, which show a clear rise of the proportion of attempted suicides by drugs since 1948, the year of the introduction of the N.H.S. (table 17). This also caused a rise in the number of known attempted suicides in the metropolitan area - chiefly owing to a rise in the female cases - whereas there was no such rise in the number of suicides (table 15). In the views of the statistical branch of the Metropolitan Police (1950) and of Stengel, the N.H.S., by making drugs widely available, contributed to their increasing consumption and possibly also to their increased use for suicidal attempts, but by indirectly reducing the use of other more lethal methods of suicidal attempts it may have also lessened the rate of successful suicides, as "... drugs proved the least effective method ..."

TABLE 17

Attempted suicides, 1933-1953 (London Metropolitan Police area)

 

Coal gas

Drugs

All means (including coal gas and drugs)

Year

M

F

Total

M

F

Total

M

F

Total

1933 150 151 301 12 14 26 413 393 806
1947 136 143 279 72 107 179 378 394 772
1948 145 153 298 112 261 373 426 554 980
1953 140 150 290 162 301 463 483 555 1038

Source:Statistics of the Metropolitan Police Statistical Department.

Among successful suicides coal gas is the agent most often responsible. In many ways the individual attempting (and failing) suicide is quite different from the one who commits it; in Stengel's view the failed attempted suicide has not been bungled but was meant as an appeal for help. What methods are used by people attempting suicide without achieving it: Tables 18-20 give the proportions in cases recently treated in a large mental hospital and two general hospitals in southern England. The percentages in the two general hospitals are very similar, barbiturates having been responsible for about two-third of the cases; barbiturates, the aspirin-codein group and coal gas accounting for 85% to 90%. Taking both samples together, phenobarbitone was the most popular agent but was employed much less frequently than the combined number of the shorter-acting barbiturates. By and large, aspirin was used more commonly by the younger age group, the barbiturates by the middle aged. Locket (1957 b) gives the peak of barbiturate poisoning at the 25-29 and 40-44 agegroups, of aspirin poisoning at 20-24 years, which corresponds to these more recent findings.

TABLE 18

Women admitted to a mental hospital after attempted suicide, near London, 1960/61

Method

Number of cases

Percentage

Drugs
18
60
Barbiturates
12 40
"Sleeping tablets"
4 13
Equanil
1  
"Asthma tablets"
1  
Other methods
12
40
Coal gas
5 17
Cutting wrist or throat
4 13
Train
2  
Drowning
1  
Types of barbiturate
   
Sodium amytal
5  
Tuinal
2  
Seconal
2  
Soneryl
1  
Unspecified
2  

NOTE. - Only the main method involved has been named. Tranquillizers were mentioned by two patients (Librium and Stelazine), in addition to barbiturates.

TABLE 19

Cases of poisoning admitted to a Berkshire general hospital, 1955-61

(Suicidal attempts in the great majority of cases)

Method

Men

Women

Total and approximate percentage

Barbiturates
15 45 60
Aspirin group
4 12 16
Coal gas
4 7 11
Industrial solvents
4 4 8
Carbromal group
-
4 4
Quinine
-
2 2
Tranquillizers
-
1 1
  27 75 102
 
Both sexes
 
Types of barbiturate (25 cases)
   
Phenobarbitone
7  
Sodium amytal
6  
Soneryl
4  
Nembutal
3  
Seconal
2  
Unspecified
3  
Mean age of patients (years)
   
Barbiturates
41  
Aspirin group
29  

NOTE. - Only the main method involved has been counted.

In the mental hospital sample (table 18) drugs are still the most commonly used group, but relatively less common than in the general hospital, and barbiturates account for roughly half the attempts. More" violent" methods were less uncommon, possibly because among these patients there were more psychotics, but the sample is very small anyhow.

Turning to the findings of other investigators we find similar proportions. Thus the Metropolitan Police figures, which showed up to 1947 coal gas to be the most commonly used agent for suicidal attempts, give from 1948-1953 (the last year for which these figures are available) drugs as the most widely used group; e.g., in 1953 among 1,038 "known" attempts, 463 (45%) were due to drugs, 290 (29%) to coal gas. The finding that in these statistics" violent "methods accounted for nearly a fourth of suicide attempts may perhaps be taken as a reflection of the fact that such methods were brought to the attention of the police more frequently than the more "passive" methods of using drugs.

Another set of figures relating to London has been published by Stengel, who studied attempted suicide in five hospitals, including general and mental hospitals and observation wards. One of his groups were patients admitted to an observation ward in 1946/47 - i.e., before the introduction of the N.H.S.; all the others referred to periods after 1948. There was a marked increase in the use of drugs in the post-1948 groups and a decline in other methods such as coal gas, wounding and hanging. In the 1946/47 sample coal gas was the agent used by 16% of the men and 21.9% of men and women respectively; "wounding "was attempted by 30% of the men. In the same observation ward in 1953 the coal-gas percentages were 13.4% men and 16.1% women, "narcotics" accounting for 32.7% and 31.1% respectively; "wounding" occurred in 23.8 % of the men. (The sizes of these 1946/47 and 1953 samples were 74 men and 64 women in 1947/47, and 66 men and 101 women in 1953.) In two other groups in the early 1950s "narcotics" accounted for about 31% and 58% of suicide attempts among the men, and 47% among the female samples; "wounding" occurred in 18% and 4% of the men. In general, suicidal attempts were more common among women, as they were among the more recent samples described above. The greater frequency of attempts among women contrasts with the greater number of men among (successful) suicides, illustrated by Stengel in a sample from a London coroner's court in 1953 (73 men, as against 44 women). As amongst suicides generally, coal gas was the most commonly used agent (almost 50% amongst both women and men), the" narcotics "trading with a proportion of approximately 10% among men and 20% among women.

Among another sample described by Locket & Angus (1952), in more than 80% of suicide attempts barbiturates were the chosen method. In all but two of these 64 cases the drug had been prescribed for the patient by a doctor, and in more than 90% it was given for insomnia as the only or the major complaint. The ratio man: woman was 4: 1, and of the 64 cases 49 were definite suicidal attempts. Nearly two-thirds of the barbiturate attempts were carried out with phenobarbitone (40 cases), almost one-sixth with amylobarbitone (11 cases).

TABLE 20

Admissions for suicidal attempts to a general hospital in Middle sex from January to October 1161

Method

Men

Women

Total

Percentage

Barbiturate
1 25 26 67
Aspirin/codeine
-
6 6 15
Coal gas
1 3 4 10
Other methods
-
3 3 8
  2 37 39 100

Types of babiturate

Men

Women

Phenobarbiton
  4
Nembutal (including 5 carbrital)
  6
Soneryl
  3
Tuinal
  1
Sodium amytal
1  
Unspecified
  11

Others

Tofranil plus largactil plus doriden

Stabbing

Mist. Expect.

Mean Age

Aspirin/codeine 24 years

Barbiturates 40 years

Source:H. R. George, personal communication.

A more recent London sample - admissions to a London teaching hospital (casualty, general wards and a psychiatric unit) in 1956/57 - was described by Woodside. Among a total of 55 suicidal attempts barbiturates were the chosen agent in 63% (35 cases); aspirin accounted for 4 cases, coal gas for 3.

An investigation comprising 100 women admitted to a London prison for attempted suicide from the southern parts of England between 1954 and 1956 was carried out by Epps. An overdose of drugs had been taken by 54 women: of these, 34 had consumed barbiturates and "sleeping tablets", and 18 aspirin or codein. Gas was the method chosen by 17. Not surprising in view of the selection of this sample is the relatively high proportion of "violent" methods employed, "cutting" and drowning having been attempted by 20 women. As is clear from the figures, not infrequently several methods had been combined in the same attempt.

In apaper published in December 1961 Middleton et al. present an analysis of attempted suicide in an English urban industrial district. Between 1953 and 1959, 219 cases of attempted suicide were admitted to their hospital group, which serves a population of about 150,000. Among the agents used, barbiturates headed the list (91 cases) followed by salicylates (34) and carbon monoxide (30). Thirty-three other patients had used "a mixture of barbiturate, salicylate and carbon monoxide", and preparations containing a barbiturate were also named among those employing "other agents ". Barbiturates were also named among those employing "other agents ". Barbiturates were thus apparently taken in about 50% of these attempts. Among all these attempted suicides women outnumbered men by 2 to 1, but whereas as many men used CO as women (15 of each sex), only 27 men took barbiturates, as against 64 women, "possibly because (the women) were already in possession of them on account of poor sleeping habits ". This urban industrial area is naturally not representative of the country as a whole, but it is interesting to note that a ratio of about 20 suicidal attempts with barbiturates per year in a population of 150,000 would lead to a much higher figure for the country than the estimates made above; but in rural and non-industrial districts such attempts may be much more infrequent.

To mention a sample from Scotland, Batchelor (1954 a) - studying 200 cases of attempted suicide admitted to a general hospital in Edinburgh - found among them 50 "psychopaths" of whom 90% (38) had used poisoning, including 5 cases of coal gas (4 men, 1 woman) and 5 of aspirin (4 men, 1 woman); 23 patients - 13 women and 10 men - had taken barbiturates (46%). Batchelor feels that the paramount factor deciding on the choice of the method in these psychopaths lies in availability: "barbiturates can now be found in a great many homes, coal gas is usually on tap, and aspirin may be bought for the purpose ".

As shown by all these samples, drugs, and in particular the barbiturates - are now clearly the most popular agent chosen by people attempting suicide. Coal gas is nevertheless the method responsible for most suicides, being a more lethal weapon than the barbiturates. In Locket's words (1959) "Barbiturate poisoning is important because of its frequency and not because it is highly lethal." Yet because of their availability in so many households the barbiturates still take heavy toll of human lives. Not infrequently one encounters among individuals who have attempted suicide some who are not "fussy" as to the method they adopt. Thus one woman in the mental hospital series had on previous occasions tried to kill herself by taking 24 Soneryl tablets and 10 Stelazine tablets, cutting her wrists, by taking 13 Stelazine tablets and finally by gas. There are thus those who - if there were no barbiturates about - would use any other means available if they were really intent on committing suicide. But others who may make an attempt on a sudden impulse only, and even more so those who poison themselves accidentally - e.g., children - may never risk their lives again once they have survived that attempt. In order to lower the risk involved in ingesting large numbers of barbiturate tablets, attempts have been made to incorporate an emetic ( Lancet, 1953 c) into the barbiturate tablet, or more recently to combine it with bemegride, which has been claimed to be a functional barbiturate antagonist. In some cases overdosage of such "megimated barbiturates" was indeed followed by recovery (Heffernan, Skinner); but, on the other hand, a recent death following the consumption of 120 hoarded Phenaglate capsules (containing quinalbarbitone, phenobarbitone and bemegride) (McGuinness & Roberts) shows that this method is not infallible.

The frequency of suicidal attempts with barbiturates will to some extent depend on the extent of their practical replacement in everyday practice by tranquillizers and non-barbiturate hypnotics; their relative popularity sometimes seems reflected among suicide attempts. Thus tranquillizers, Doriden, Distaval, Persomnia and others have recently been employed in such attempts, and their toxicity relative to that of the barbiturates is in this connexion a very important factor. That they are not free from the risk of causing fatalities is illustrated in table 21, showing therapeutic misadventures. As regards the value of bemegride in the treatment of acute barbiturate poisoning, opinions are divided, Shaw quoting 200 papers favourable to its use for that purpose, whereas Locket (1959 b) has abandoned the bemegride treatment in acute barbiturate poisoning altogether after a trial period of fifteen months.

TABLE 21

Therapeutic misadventures, 1954 to 1960, in England and Wales

(Wrong doses of a prescribed medicine)

Agent

1954-56

1957-58

1959

1960a

Barbiturates
217 157 99 84
Tranquillizers
6 19 13 5
Drug and alcohol
7 15 10 10
Preludin
-
-
-
-
Persomnia
-
4 1 1

Source: Registrar-General.

a

Provisional figures.

The risk of accidental and suicidal overdosage, particularly of barbiturates - in the present author's experience as well as according to certain statements in the literature - seems much higher than in the general population among barbiturate addicts and among alcoholics, who are often prone also to the habitual and sometimes excessive ingestion of barbiturates.

IV. Barbiturate habituation, addiction and chronic intoxication

The risks of habit-formation and of accidental or deliberate overdosage with the barbiturates have been known for a long time; the danger of true "addiction" (physical dependence) and a definite abstinence syndrome has not been recognized until fairly recently. Almost fifty years ago, Willcox (1913), reading a paper at the seventeenth International Congress of Medicine on the "veronal habit" described how this drug taken in repeated small doses could lead to chronic veronal poisoning characterized by features such as tremor, ataxia, etc., and often an abnormal mental condition and disorganization of the moral sense. Tolerance (he stated) "undoubtedly occurs in some cases" though (1934 a) it was only slight in degree. He repeatedly stressed the risk of fatalities from suicide especially in these taking the drug habitually: "Persons who have been taking the barbituric acid derivatives daily for a long period very commonly take a large overdose when they are faced with mental stress and worry and often death results thereby." Rather surprisingly, on the other hand, he stated (1934 a) that barbiturate addiction differed from morphine and heroin addiction "... in that sudden discontinuance is not followed by severe withdrawal symptoms". Willcox's main adversary in the "battle of the barbiturates" in the 1930s, Gillespie, also stressed the "safety of their sudden withdrawal in habituated patients" (1934 b).

It is of course impossible to say how prevalent emotional dependence is and how often physical dependence occurs among these psychologically dependent individuals. Certainly a certain proportion of people admitted to hospital after an acute overdosage are found to have been taking the drug habitually and occasionally excessively over a considerable period, the state leading to admission being no more than an episode in, and an acute exacerbation of, a chronic enough often unrecognized condition. Not infrequently the true state of affairs is not even recognized in hospital when the patient, as often occurs, gives a misleading history, and then a state of true addiction comes to light by the supervention of physical withdrawal symptoms. Thus a middle-aged man recently admitted in an abusive and violent state developed in quick succession three grand mail seizures three days after admission. It later came out that this man had been taking 12 gr of sodium amytal daily over the past six months in order to "steady the nerves" ; often he had found them "too weak" and unable to sleep he had been taking four or five capsules at a time. Another patient, a middle-aged female of an hysterical type, was fre- quently admitted to hospital after an overdosage or in a state of chronic barbiturate intoxication. On an earlier occasion she was said to have had two epileptiform convulsions two days after admission and an acute psychotic episode with auditory hallucinations on the fifth day. Her state of depression, confusion, irritability and slight unsteadiness usually cleared up fairly rapidly in hospital and when her continual demand for more tablets was not granted she would leave beleaguering her doctor (or doctors) at home with requests for them (usually Tuinal) and often threatening to throw herself into a river.

Locket (1957 b) estimates that 4% - 6% of all the cases of acute barbiturate poisoning admitted to his unit were".. barbiturate addicts accidentally taking overdoses or attempting suicide ". If this proportion should hold good all over the country there would be roughly 500 addicts alone among an estimated number of approximately 10,000 cases of acute barbiturate intoxication in England and Wales per year.

As the incidence of overdosage seems considerably higher in those taking the drug habitually and sometimes excessively than among the general population - reminding one in some way of the much more frequent occurrence of acute drunkenness among alcoholics than among other drinkers, though these too may sometimes imbibe too much - the high rate of barbiturate poisoning points rather to the likelihood that a state of (at least emotional) dependence on these drugs is not uncommon. Three barbiturate addicts described by Hunter & Greenberg all gave a history of frequent suicidal attempts or overdosage. Among three cases of barbiturate addiction described in detail by Armitage & Sim one had a history of frequent suicidal attempts, another had shown ... catastrophic results of(barbiturate) intoxication "on four occasions". There is, however, no unanimity as regards either the prevalence or the danger of addiction. Nowadays barbiturate addiction is probably by far the commonest form of drug addiction in Great Britain (Hunter & Greenberg), seeing that the numbers of known addicts to dangerous drugs over the past few years totalled between approximately 350 and 450 per year. Batchelor (1960) thinks that "there must be few general practices without a case of barbiturate addiction ", and Dunlop (1957) states that chronic barbiturate intoxication is still on the increase. Partridge, however, points out that although barbiturate addiction is undoubtedly commoner than addiction to other new drugs it is yet "remarkably rare in Britain considering the prodigal amounts that are prescribed ". Mayer-Gross, Slater & Roth in their textbook on Clinical Psychiatry (1960) write that "the dangers of poisoning or addiction, though they should not be forgotten, should not be exaggerated" and that "poisoning can be avoided by keeping an eye open for signs of danger, such as drowsiness, vertigo and ataxia ".

If chronic barbiturate intoxication is indeed not very uncommon, why does it seem relatively little known and diagnosed? Morgan (1953) believes the condition to be common, but not well recognized because it is unfamiliar, just as was bromide psychosis a quarter of a century ago. Sometimes the state resembles long-standing alcoholism, or, as pointed out by Willcox (1934 a) and Purvis-Stewart many years ago, it may mimic a great number of neurological conditions. However, once more widely known, mild chronic barbiturate intoxication may be diagnosed more often; in one London teaching hospital the "interesting psychiatric syndromes" resulting from the prolonged use of barbiturates as sedatives and hypnotics became known as "chronic subclinical barbiturate intoxication" (Strauss). As long as the addict may be able to satisfy his demands for the drug he may feel tolerably well in spite of his symptoms though "persistent symptoms, particulary those accompanied by a demand for drugs" (Hunter and Greenberg) might arouse suspicion. If this demand is not met the addict may suffer from symptoms such as anxiety, depression, insomnia; these temporary withdrawal symptoms may be thought to show an exacerbation of his original symptoms so that more barbiturates are prescribed perhaps more frequently and in increasing doses. A vicious circle may set in in this manner so that an originally "... mild psychiatric disturbance is converted into a serious condition" (Hunter, 1957) which has on rare occasions been known to have led to a leucotomy (Hunter), or to several such operations." Patients with chronic tension states "- as Sargant (1958) writes -" may continue to remain tense and anxious for many years, because they are suffering more from a self-induced or doctor-induced chronic barbiturate tension state than from a persisting anxiety neurosis per se. I am now seeing patients who have had one, two, or even three leucotomies performed for a chronic persisting tension, and who have then turned out to be, probably, cases of chronic barbiturate addiction."

In order to get a very rough idea of the view of general practitioners about the prevalence or otherwise of barbiturate dependence in their practices, the recent questionnaire mentioned above asked G.P.s whether such a state was common, rare or absent in their own practice and whether they could give an approximate estimate. It was clear from their replies that the views as to what could be regarded as "common" or "rare" differed greatly. From the examples given in the questionnaire of what was meant by" dependence" the emphasis was put on mild (psychological) "habituation" rather than on "addiction ". Of the twenty-four G.P.s who replied, two out of three found this state rare in their practice, one out of four thought it common, and one practitioner stated it to be absent. The percentages quoted as to the prevalence of such a state of "dependence" among their patients varied from 3% to 0.12%, the sizes of the practices varying between 1,000 and 3,600 patients. The sample was very small (25 doctors), the study being thought of as merely a "trial run" and as a pilot investigation, but the written replies corresponded to the views expressed to the writer by a much greater number of general practitioners with whom this problem has been discussed in recent months. Some, of course, stated that they were aware of the problem, but faced with the persistent demand by patients for sedative drugs, and full waiting rooms, they could see no alternative.

As one would expect, most barbiturate addictions are caused by the intermediate and short-acting preparations though more rarely (Morgan, Sargant, 1958) they also occur with the long-action ones. Among Armitage & Sim's recent series of twelve barbiturate addicts, amylobarbitone sodium had been taken by 8 patients, quinalbarbitone sodium by 3, phenobarbitone by 2 patients. These patients had taken the drugs for periods varying from 9 months to 9 years, the doses of the quick-action barbiturates having been 12 gr (0.8 g) daily; thus conforming with the conditions as to timing and dosage (at least 0.4 g daily) needed to produce true addiction (physical dependence on pentobarbital and secobarbital (H. F. Fraser et al (1958) J.A.M.A. 166, 126). Thus "addiction" occurs only with habitual ingestion of" far above therapeutic doses "and as tolerance is relatively small, mainly psychologically abnormal people are likely to increase the doses so far ( Lancet, 1954), except when barbiturate consumption is combined with that of amphetamines. This combination is quite popular; among the 12 barbiturate addicts of Armitage & Sim, in 2 cases the primary addiction was to amphetamines. The latter may produce restlessness and insomnia necessitating higher doses of barbiturates, which in turn by causing lethargy and drowsiness may lead to the pushing up of the amphetamine. Simultaneous consumption of alcohol may occur alongside habitual overdoses of barbiturates and amphetamines. "Cured" alcoholics as well as still "practicing" ones are very prone to take overdoses of other drugs, chiefly barbiturates (Glatt, 1954 a) and the present author saw his first cases of barbiturate addiction - with withdrawal convulsions and withdrawal delirium tremens - and chronic barbiturate intoxication (Glatt, 1956) in people who had been "cured" of their alcoholism. Barbiturates themselves taken in fairly high doses for some time may lead to a state of depression, and to counteract this effect the barbiturate addict may take amphetamines in a similar way as the opium addict may take cocaine (Mayer-Gross et al.). It may be of some interest here to mention that when recently it became more difficult - because of wider knowledge of the risks among doctors in the case of barbiturates and because of greater restrictions of the amphetamines - to get hold of these drugs, habituation became more common to another combination (Glatt, 1958, 1959 a, b.) - i.e. that to carbromal preparations and the stimulant Phenmetrazine, both of which were at the time still "unrestricted ".

Apart from causing depression directly, habitual excessive consumption of barbiturates like any other form of addiction may lead to deterioration in economic and social conditions, domestic friction, difficulties at work, etc., all factors which indirectly may heighten the depressive state; and as at the same time barbiturate tablets provide a handy agent for that purpose, the fact that suicidal attempts are not uncommon among barbiturate addicts becomes easily understandable (Parker).

Willcox (1934 b) went so far as to say that in his experience "most of the fatalities from poisoning by the barbiturates occur in people who have been taking the drugs in repeated daily doses "and that the barbiturates were" hardly ever used as a means of suicide except by persons accustomed to their daily use and effect ".

How are addicts able to procure such high doses? In the three Cases described by Hunter & Greenberg the supply came either from unknown sources or the addicts managed to conceal them from their psychiatrists. One of their patients, a medical auxiliary, managed to get his supplies over three years from twelve legitimate sources. In a case known to the present writer the addict, though continually in a state of chronic barbiturate intoxication, was able to carry on for years by getting his drugs from most doctors living in the district at different times -- as emerged later on -- at the time he claimed that he bought them at exorbitant prices on the local" black market ". Like many of these patients, he was a rather inadequate person with psychopathic features who simultaneously took amphetamines and at times also alcohol to excess. Such multiple addictions were also common among the patients of Armitage & Sim: among their total of nineteen drug addicts, fourteen were also dependent (usually only emotionally) on another drug apart from the primary addiction, and seven were moreover dependent on alcohol. Of these nineteen addicts, eleven obtained the drugs by prescription, often misrepresenting to their doctors the amounts which they were taking. Seven of their nineteen patients had professional access to drugs, and although "professional" addicts are much less common proportionally among barbiturate addicts than among those to dangerous drugs (where in the past few years about one addict in six was a "professional" addict) (Interdepartmental Committee on Drug Addiction), easier access produces a greater risk for doctors, nurses and similar professions also in the case of barbiturates. The present writer saw a number of "professional" addicts in various mental hospitals, and two of the patients of Hunter & Greenberg had easy access to drugs -- which led the British Medical Journal (1954) to the remark that one might hope "... that this deplorable state is rare among the general population ". The seven patients of Armitage & Sims with professional access to drugs initiated and maintained their own addiction.

The study of Armitage & Sim illustrates some other interesting points. Tolerance to barbiturates may develop to some extent, as shown, for example, by the observation (already mentioned) of Camps (1961) who saw people walking about who were taking 16 gr. of sodium amytal during the night; Armitage & Sim describe three examples illustrating the opposite phenomenon of "barbiturate sensitivity" - i.e. three patients who suffered from recurrent symptoms of intoxication following small and economical doses of amylo-barbitone or pentobarbitone. None of these patients had suspected the relationship between their periodically recurring symptoms and their taking of the drug. Even previous investigations in hospital failed to elucidate the cause and had arrived at diagnoses of neurogical or other psychiatric conditions.

In five other barbiturate-addicted patients, too, the magnitude of the habitual intake had not been appreciated and sudden withdrawal was carried out. Complications followed in all these patients: four had fits on the second or the third day or both, and the fifth developed a delirium from the second to the sixth day. Among the other seven barbiturate addicts in whom the drug was withdrawn slowly the only complication was a single epileptiform fit on the twelfth day of withdrawal.

Three interesting detailed case histories are presented by these authors, all of which describe an E.E.C. showing the "... generalized fast high voltage activity typical of barbiturate addiction".

One not infrequently encounters people who have been taking barbiturates in small doses for years without increasing the dose and who swear that they could not possibly sleep without their night-time drug. However, because of its prevalence, because of the danger of a supervening addiction and state of chronic intoxication and because of the risks of accidental or deliberate poisoning, barbiturate-dependency is a problem of considerable importance. A greater appreciation of the risks by all concerned would diminish the problem; Locket (1957 a) believes that in this way the problem might become restricted to a "hard core of recidivist psychopaths, psychotic depressives and a few severe psychoneurotics". Many people who misuse these drugs are emotionally unstable, and the view has been repeatedly expressed that barbiturate addiction is rare except in fundamentally very disturbed individuals ( Brit. Med. J., 1958; Lancet, 1954), such as psychopaths and hysterics who compulsively swallow anything, including barbiturates that they can lay their hands on and in truly enormous doses" (Sargant, 1956). If that is so the prognosis would be naturally very poor. But, on the other hand, the habit of taking a hypnotic at night time is so prevalent, regarded as harmless and accepted by society so that people who would "not dream of taking any well-known drugs of addiction ..." (Morgan 1953) or of ever getting drunk may gradually develop a state of dependency, in particular doctors, nurses and others who have got easy access to drugs. As we have seen, attempted suicides are more frequent in women than in men, and barbiturates are the most popular method; and in suicides -- which are in general more common in men -- women preponderate only among those taking tablets: both among barbiturate suicides and fatal accidents women have been outnumbering men consistently for the past thirty years (table 1). This probably reflects the more widespread habitual use of barbiturates and a greater prevalence of a state of dependency on barbiturates among women. Among men, on the other hand, alcohol abuse is much more common. One of the possible reasons why in general prognosis for women alcoholics may be worse is the greater social "taboo" on heavy drinking among females: female alcoholics must thus have broken many more taboos to arrive at their alcoholic state than their male counterparts, and may therefore originally already have been less "normal" and thus more ready to break those taboos (Glatt, 1955 a). With socially sanctioned barbiturates, however, the position is quite different, and one might think that many women, rather than risk censure for somewhat heavier drinking, may prefer the lesse conspicuous method of taking tablets when worried or in difficulties, without being necessarily or even commonly psychopathic or severely hysteric. Once such people have embarked on the pill-taking habit it may easily grow on them. It seems, therefore, by no means certain that the majority of barbiturate-addicted and even less of the barbiturate-habituated individuals are psychologically very abnormal, and specialized treatment should hold out quite a measure of success.

Just as in alcoholism, it is clear that treatment of barbiturate dependency must not stop short at gradual discontinuation of these tablets and disintoxication; there is need for long-term therapy so that the addict learns to cope with inner and environmental difficulties without recourse to drugs. Group therapy in special units forming therapeutic communities may be very suitable. The author is not aware that this has been systematically attempted; but barbiturate addicts (and other addicts to "minor" drugs) fitted in quite well in mainly alcoholic groups at St. Bernard's hospital (Glatt, 1960) as they were able to identify with alcoholics in spite of certain differences. The hypothetically "average" alcoholic may be less often a disturbed person basically than the average barbiturate addict, as after all his drug - alcohol - is obtainable without any restrictions, whereas the barbiturate addict has to go to some trouble to maintain an adequate supply of his tablets. Yet the argument that it is not the bottle or the pill that is at fault, but the individual who abuses them, seems in many aspects to apply both to alcoholics and to barbiturate addicts and often mainly environmental, cultural or traditional factors will decide which "drug" is chosen by the individual who feels in need of "escape" whatever his degree of psychological "abnormality".

V. Barbiturates and alcohol

Seven per cent of more than 200 million N.H.S. prescriptions in the country are for barbiturates; two in three adults take alcoholic drinks. There must be many people who take these two depressants of the central nervous system occasionally within a short space of time; and the question of their combined effect is therefore of great importance. Some observers believe in a potentiating effect, as, for example, Douthwaite (1958) and Dunlop (1957), who states that whilst in acute barbiturate poisoning the fatal dose is seldom less than twelve times the therapeutic dose, a smaller dose may prove fatal if taken in combination with alcohol "which greatly potentiates the effects of barbiturates ". Deaths from the combination of barbiturates and alcohol (though the time interval between their consumption is not given) occurred in 19 of 232 accidental barbiturate fatalities in 1959 and in 48 of 293 such deaths in 1960 ( Brit. Med. J., 1961 b). Some fatalities from the combination of drug and alcohol (the drug presumably being mostly a barbiturate) occur among the" therapeutic misadventures" (table 21). Sometimes fatalities from sublethal doses of barbiturates taken concurrently with drink give rise to comments at coroners' courts ( Brit. Med. J., 1953 a; Lancet, 1953 a). However, seen in relation to the population at risk, such fatal outcome does not seem unduly large ( Brit. Med. J., 1953 b; Lancet, 1953 a), and most authorities believe that alcohol and the barbiturates have merely an additive synergistic and not a potentiating effect (Buttle et al., Camps, 1957; Locket, 1957 a). A Lancet annotation (1953 a) also adheres to this view, but points out that chronic alcoholism by damaging the liver may delay the excretion of barbiturates by detoxication so that cumulative effects might be more likely.

Of greater practical importance may be this question of synergism in regard to smaller amounts of drug and alcohol taken concurrently. In Sweden Goldberg pointed out ten years ago that barbiturates have a deteriorating effect on driving, and in this country the experiments of Cohen et al. and of Drew et al. have shown that even a little alcohol taken reduces judgement and driving skill. As the British Medical Journal pointed out in 1953 b, the risks from combining the two drugs lie in this aspect rather than in the possibility of death arising from gross overdosage. The Government's new Road Traffic Bill published in March 1961 declares a person unfit to drive a motor-vehicle if his "... ability to drive properly is for the time being impaired" ( B.M.J., 1961 a). From the legal and the safety point of view, as well as from that of his and others' safety, for the motorist drink and barbiturates are certainly very dangerous fellow-travellers.

Any risks arising out of a combination of alcohol and barbiturates are naturally particulary great in the alcoholic. In the experience of the present writer over the past tenyears the dangers of acute and chronic barbiturate poisoning of" automatism ", of suicidal attempts, of fatal accidents and of suicide, and of habituation and addiction to barbiturates are all very much greater in the long-standing alcoholic than in the general population.

One could enumerate a number of reasons why alcoholics should be prone to develop suicidal ideas and why suicidal attempts should be carried out by means of barbiturates. Quite apart from the original psychological difficulties which in many cases may have put the future alcoholic on the road of excessive drinking, the latter in turn then creates secondary problems. The drinking histories of 150 male and 54 female alcoholics observed by the writer at Warlingham Park Hospital between 1952 and 1957 (Glatt, 1955, 1961 d) show, for example, the presence of frequent feelings of despondency and of remorse (in 90% of these alcoholics) at some time in the past, and of self-pity in about 55%. More than half among these men and over one-third of the women had contemplated suicide, and similar proportions admitted stupidly dangerous behaviour towards themselves or others. Many alcoholics seem to be prone also to take other drugs rather " promiscuously" to excess and become habituated to them, a risk which in this country seems to affect chiefly the middle-aged "loss of control" alcohol addict in whom emotional conflict may have led to excessive drinking, rather than the non-addictive alcoholics as seen among working-class drinkers or among those in whom chiefly continual temptation and easy access because of occupation seem to have been responsible for heavy drinking (Glatt, 1961 a, b). Alcoholics may take barbiturates either concurrently with drink or substitute them once they have given up drinking, and in such cases it is often difficult to know whether they have resumed drinking or are suffering from chronic barbiturate intoxication or from a mixture of this condition and alcoholism. Alcoholics state that they take drugs to get the desired effect more quickly, more profoundly and cheaper; moreover, their breath does not give them away if watched by suspicious relatives. On the other hand, it is rare to meet a barbiturate addict who later changed over to alcohol which is usually regarded by drug addicts as much less potent (" chicken-feed "), though occasionally a barbiturate addict explains that he takes drink to help him over the interval needed after having consumed all the tablets on the previous prescription before he can ask his doctor for another prescription. Alcoholics are often introduced to barbiturates early in their drinking career, insomnia being a symptom in 90% (Glatt, 1962).

The frequency of suicidal attempts by barbiturates among alcoholics is thus easily understandable. Batchelor (1954 b) found among 200 attempted suicides in Edinburgh 43 "excessive drinkers" whose" main method "in nearly half of them had been barbiturate poisoning, coal gas," cutting "and" other" methods accounting for the remainder in about equal proportions. Among Epps' 100 women who were in prison for attempted suicide were 24 long-standing alcoholics; among these 100 the commonest method had been an overdose of drugs (54 cases), in particular of barbiturates (34). Stengel had among 138 admissions for attempted suicide to a London observation ward in 1946/47 twelve alcohol addicts; and altogether eleven individuals were known to have taken alcohol at the time of the suicidal attempt. Our own investigations showed 25% attempted suicide in the history of about 250 middle-class alcohol addicts treated at Warlingham Park Hospital (Glatt 1954 b, 1962) and the same proportion among a smaller sample of alcoholics admitted to a London observation ward; in contrast to the first group, chiefly classes I and II out of the Registrar General's five social classes, the observation ward group were in the main class III-IV. Homeless alcoholics, also studied at about the same time (1950), had again a similar proportion (26 %) of attempts, which were in the main not serious (Glatt & Whiteley).

As regards habit-formation or addiction to drugs, this was found to be chiefly pronounced among the middle-class alcoholics (Registrar-General's classes I and II): 2 in 5 out of 68 women patients at Warlingham and one alcoholic in 4 out of 200 male patients (Glatt 1962). Of these 75 patients (28% of the total), almost all had taken barbiturates, sometimes in combination with amphetamines. Among the men, habituation was most common among those alcoholics who had been referred to the hospital by Alcoholics Anonymous; this subgroup had the "highest" social composition, and one in three of them also habitually took drugs to excess. The proportions of persons habituated or addicted to drugs were much smaller among the observation ward group (13%) and even more so among homeless alcoholics.

The state of affairs among women alcoholics treated this year at St. Bernard's Hospital, Middlesex, is shown in table 22. One has, however, to keep in mind that these figures were obtained from certain questions put to these patients in a long questionnaire and are in some ways probably under-estimates. Among 41 female alcoholics, more than one in three admitted a suicidal attempt, in the great majority (1 in 4 of the total) by barbiturates. All those who had attempted suicide with these drugs stated that they had taken them in the past either regularly or periodically and often in excess of the doses prescribed. Two patients in three denied suicidal attempts; among these only one-fifth had taken barbiturates in the past. Of the total of 41 patients, 1 in 4 had in the past taken barbiturates regularly or periodically. Among women alcoholics treated at St. Bernard's Hospital between 1958 and 1960 more than one in three had habitually taken drugs to excess, barbiturates alone being used by one-third of them (Glatt & Judge). However, by then the risks of barbiturates had become more widely appreciated, and other drugs, often "unrestricted" ones, had been employed relatively more frequently.

TABLE 22

Questionnaire on 41 women alcoholics, St. Bernard's Hospital, 1961

A. 15 admitted suicide attempts (36.6 %)

11 by barbiturates (26.8 %)

3 by gas (7.2%)

1 by drowning (2.4 %)

B. Out of 11 barbiturate attempts:

6 took them regularly a

5 took them periodically a

C. Out of 4 by other methods:

1 took barbiturates periodically

D. 26 admitted to no suicide attempts (63.4 %)

Of these, 5 admitted to taking barbiturates - 4 periodically, 1 regularly a

E. Number of patients taking barbiturates, 17 (41.5 %)

In view of the greatly increased risks attached to the use of barbiturates in alcoholics, we have over the past ten years replaced them by non-barbiturate hypnotics of which a few have been introduced in recent years. One must, of course, be very careful as many alcoholics who had been dependent on barbiturates do not divulge this fact during the first interview, paying for this omission with withdrawal convulsions when they receive no barbiturates. With this exception, however, we have been able to manage without barbiturates, by and large quite satisfactorily. As a group the non-barbiturate hypnotics seem somewhat less potent than barbiturates, but also less habit-forming or addictive. None of them has proved quite free from such risks, however (Glatt, 1955 b, c; 1959 a, b, c; 1960, 1961 b, c), though the opportunities to show toxicity and risks of dependency in their case extended over a few years only in contrast to 60 years' use of the barbiturates. Anyway it seems likely that in certain unstable, immature alcoholics and similar personality-types any drug with sedative or stimulating effects will sooner or later be "found out" to be habit-forming and possibly addictive.

Conclusions

A few years ago the British Medical Journal (1956) described an interesting cycle of events which could often be noticed with the appearance of any new drug. Initially a trickle and later a stream of favourable reports make the drug fashionable, but they are then replaced by quite different observations: "Therapeutic failures, undesirable side-effects, addictions, and accidents claim attention." The drug falls into disrepute and may be completely written off:" However, if it has some real value it will usually maintain . . . some . . . more limited . . . place in the medical armoury. What was once a cure-all, and later something useless or dangerous, ends by being moderately helpful in range of conditions."

Whilst conforming to this described cycle in some points, the story of the barbiturates has not been " running true to form ". More than half a century after their introduction they still enjoy a very widespread popularity, and having weathered the " battle " and many spasmodic attacks later on, in clinical practice, they still seem today often to be administered as a kind of " cure-of-almost-all " nervous conditions. Newcomers such as the tranquillizers and the non-barbiturate hypnotics have so far made few inroads into the popularity of the barbiturates.

Why in spite of all statistics about rising rates of poisoning, in spite of the fact that Willcox's old but little heeded warning of the possibility of addiction has been dearly vindicated - do barbiturates still today make up 7% of all prescriptions in the country. One of the reasons may be that the toxicologist, thc pathologist and the specialist interested in the field of addiction chiefly see the ill effects resulting from the use of such drugs; the general practitioner, on the other hand, watches the benefits which his patients derive from them. He sees what the barbiturates do for his patients, whereas others may observe more what they do to them. For the practitioner the barbiturates provide a conveniently wide range of drugs suitable for the varying requirements, drugs on which he can rely, and in regard to their abuse for suicidal purposes, etc., he may point out that relative to their widespread use cases of misuse are after all not very frequent and toxic manifestations are rare. It has also been argued that in the absence of such drugs the man wanting to attempt suicide would choose other and probably much more lethal agents.

However, even if relative to their widespread use fatalities and cases of addiction may not be very frequent, the enormous quantities consumed are indirectly responsible for the facts that the absolute figure of cases of accidental and deliberate poisoning from barbiturates is staggering, and that emotional and physical dependence is higher than with any other drug (except alcohol). And it may seem very doubtful that there are enough " medical " and " legitimate " psychiatric indications to justify this vast consumption.

What then could be done to decrease it? Education, a wider appreciation of the risks involved, of the need of " casual " rather than symptomatic " treatment, of the advisability to face problems rather than escape from them, etc., would all help to some extent. One has, however, to be realistic about this problem. Approximately one prescription in four in this country - if one includes tranquillizers and stimulating drugs - constitutes a state of vast " overconsumption " and even if, with Asher, one believes that people need "hope and not dope ", some such drugs will always be widely desired and to a much more limited extent also be very valuable. Thus whilst education in this respect is very necessary it will not eradicate the problem. Could one safeguard the use of barbiturates? The attempts hitherto made may hold out some hope, but do not seem to provide " the " answer; even " megimated " barbiturates have produced fatal poisoning.

Could one replace barbiturates by equally effective and dependable but even less toxic and less lethal drugs: From table 5 it may seem that the advent of the tranquillizers -which are not listed in that table - may have led in recent years to a slight decrease of the use of hypnotics, including barbiturates. However, tranquillizers too have their dangers, including - at least in the case of meprobamate - that of addiction, and for certain psychiatric conditions they may perhaps not be so effective as the barbiturates. The new non-barbiturate hypnotics, whilst fairly satisfactory in many respects and on the whole probably less toxic and less habit-forming, are not free from such dangers and seem also slightly less effective than the barbiturates. The drawbacks of the latter certainly justify further search for alternatives, but in the case of the newer drugs the testing time has been much too short to evaluate all the risks attending their more widespread use.

There is therefore no doubt about the need for education and research in this field: education of doctors and public towards using fewer hypnotics, research to produce equally but less toxic ones, so that in the future the man plagued with insomnia and anxiety may require fewer pills, but may have at his disposal " better " ones when he needs them. Until this happens the old " battle of the barbiturates " - which the Lancet described fifteen years ago as " short-lived " - may still linger on as a smouldering controversy, erupting occasionally into an acute flare-up. Undoubtedly the barbiturates are effective and valuable drugs; yet in view of their enormous consumption being responsible for many cases of habituation (of whom some are bound to become addicted) and yearly for several thousands of suicidal attempts (of which a not inconsiderable proportion are bound to be fatal), one can hardly any longer dismiss all attacks on the barbiturates as " intemperate " and regard the defence as completely adequate.

Acknowledgements

For their help in obtaining scattered data I am very much indebted to: members of the staff of the Ministry of Health, the Department of Health for Scotland, the Ministry of Health and Local Government (Northern Ireland); the Metropolitan Police (Statistical Department), and in particular to Miss E. M. Brooke and the Mortality Statistics Section of the Registrar-General's Office.

a

Provisional figures.

a

Provisional figures.

a

Some of these taking barbiturates frequently to excess,

References

ADAMS, E. W. Drug Addiction. 1937. Oxf. Univ., Pr.: London.

ARMITAGE, G. H. & SIM, M., Brit. J. med. Psychol., 1960, 33, 148.

ASHER,R., Lancet, 1958, I, 954.

BATCHELOR, I. R., Brit. Med. J., 1954, 1, 1342 (a).

BATCHELOR, I. R., J. Ment. Sci., 1954, 100, 451 (b).

BATCHELOR, I. R., Pract., 1960, 184, 718.

British Medical Journal, 1913, I, 566 (Edit.).

British Medical Journal, 1918, I, 182 (Edit.).

British Medical Journal, 1947, I, 167 (Any questions).

British Medical Journal, 1953, I, 846 (Any questions) (a).

British Medical Journal, 1953, I, 1269 (Edit.) (b).

British Medical Journal, 1954, II, 1534 (Edit.).

British Medical Journal, 1956, I, 969 (Edit.).

British Medical Journal, 1958, I, 954 (Any questions).

British Medical Journal, 1959, II, 751 (Annot.).

British Medical Journal, 1960, I, 1578 (a).

British Medical Journal, 1960, II, 1301 (b).

British Medical Journal, 1961, II, 973 (a).

British Medical Journal, 1961, I, 1262 (b).

BROOKE, E. M., Lancet, 1956, I, 150.

BRUNTON, L., Brit. Med. J., 1905, II, 1002.

BURN, J. H., Brit. J. Add., 1957, 53, 115.

BUTTLE, G. A. H., FEARN, H.J. & HODGES, J. R., Brit. med. J., 1953, II, 222.

CAMPS, F. E., Brit. J. Add., 1957, 53, 113.

CAMPS, F. E., ibid., 1961, 57, 97.

CLARKE, G. F. M., Lancet, 1904, I, 223.

COHEN, J., DEARNALEY, E.J. & HANSEL, C. E. M., Brit. med. J., 1958, I, 1438.

CRAIG, M., Proc. R. Soc. Med., 1934, 27, 517.

CRAIG, M., Lancet, 1934, I, 708.

DAVIS, H., Brit. J. Add., 1957, 53, 101.

DAVIS, H., 1961, personal communication.

DOUTHWAITE, A. H., Brit. Med. J., 1958, II, 1347.

DREW, G. C., COLQUHOUN, W. P. & LONG, H. A., ibid., 1958, I, 1438.

DUNLOP, D. M., Pract., 1957, 178, 26.

DUNLOP, D. M., HENDERSON, T. L. & INCH, R. S., Brit. Med. J., 1952, I, 292.

EPPS, P., Lancet, 1957, II, 182.

FISHER, E. & VON MERING, J., Ther. Mh., 1903, I7, 208.

GARLAND, H., Proc. R. Soc. Med., 1957, 50, 611.

GERHARTZ, B. klin. W., 1903, XI, 928.

GILLESPIE, R. D., Proc. R. Soc. Med., 1934, 27, 504 (a).

GILLESPIE, R. D., Lancet, 1934, I, 337 (b).

GILLESPIE, R. D., ibid., 1934, I, 482 (c).

GLATT, M. M., Lancet, 1954, II, 363 (a).

GLATT, M. M., Brit. Med. J., 1954, II, 303 (b).

GLATT, M. M., Brit. J. Add., 1955, 52, 55 (a).

GLATT, M. M., Lancet, 1955, I, 308 (b).

GLATT, M. M., ibid., 1955, II, 675 (c).

GLATT, M. M., ibid., 1956, I, 313.

GLATT, M. M., ibid., 1957, II, 387.

GLATT, M. M., Brit. Med. J., 1958, II, 1100.

GLATT, M. M., ibid., 1959, I, 50.

GLATT, M. M., Brit. J. Add., 1959, 56, 23 (a).

GLATT, M. M., Lancet, 1959, I, 887 (b).

GLATT, M. M., Brit. Med. J., 1960, I, 1891.

GLATT, M. M., Brit. J. Clin. Pract., 1961, 15, 153 (a).

GLATT, M. M., Brit. Med. J., 1961, I, 1246 (b).

GLATT, M. M., ibid., 1961, I, 1844 (c).

GLATT, M. M., Acta psychiat. Scandin., 1961 (in the press).

GLATT, M. M. & JUDGE, C. G., Med. J. Austral., 1961, I, 590.

GLATT, M. M. & WHITELEY, J. S., Mschr. Psychiat. Neurol., 1956, 132, 1.

GOLDBERG, L., Proc. 1st Internat. Confer. Alcohol and Road Traffic Stockholm, 1951.

HALE-WHITE, W., Brit. Med. J., 1905, II, 1006.

HEFFERNAN, D. T., ibid., 1959, II, 1097.

HUNTER, R. A., Lancet, 1955, I, 1265.

HUNTER, R., A., Brit. J. Add., 1957, 53, 93.

HUNTER, R. A. & GREENBERG, H. P., Lancet, 1954, II, 58.

Interdepartmental Committee Report, Drug Addiction, 1961, London, H.M.S.O.

ISBELL, H. et al., Arch. Neurol. Psychiat., 1960, 64, 1.

JOHNSTON, J. M., 1961, personal communication.

Lancet, 1947, II, 583 (Edit.).

Lancet, 1951, II, 297 (Edit.).

Lancet, 1952, I, 406 (Annot.) (a).

Lancet, 1952, I, 879 (b).

Lancet,1953, I, 1140 (Annot.) (a).

Lancet, 1953, II, 554 (Annot.) (b).

Lancet, 1953, II, 890 (c).

Lancet, 1954, II, 75 (Edit.).

LOCKET, S., Proc. R. Soc. Med., 1956, 49,585.

LOCKET, S., Brit. J. Add., 1957, 53, 105 (a).

LOCKET, S., Clinical Toxicology, 1957. Henry Kimpton: London (b).

LOCKET, S., Pract., 1959, 182, 457 (a).

LOCKET, S., Brit. Med. J., 1959, II, 1332 (b).

LOCKET, S. & ANGUS, J., Lancet, 1952, I, 580.

LUFF, A. P., ibid., 1934, I, 423.

MACGREGOR, A. G., Pract., 1960, 184, 15.

MATTS, S. G. F., ibid., 1959, 182, 732.

MAYER-GROSS, W., SLATER, E. & ROTH, M. Clinical Psychiatry, 1960, Cassell & Co. Ltd.: London.

McGUINNESS, B. W. & ROBERTS, F.J., Brit. med. J., 1960, II, 996.

Metropolitan Police (Statistical Department). Quoted from Stengel, E., and personal communication (1961).

MIDDLETON, G. D., ASHBY, D. W. & CLARK, F., Pract., 1961, 187, 776.

MORGAN, T. N., Pratt., 1953, 171, 196.

ORMISTON, G., ibid., 1959, 182, 457.

PAI, M. N., Med. Pr., 1961 (quoted from" Stethoscope ", 1961, No. 27).

PARKER, L. R. B., Brit. J. Add., 1961, 57, 115.

PARSONS, J. A., Brit. Med. J., 1908, II, 832.

PARTRIDGE, M., ibid., 1959, I, 850.

Practitioner, 1953, 171, 230 (Edit.).

PURVES-STEWART, J., Proc. R. Soc. Med., 1934, 27, 503.

Registrar-General, Statistical Reviews, England and Wales. H.M.S.O.: London.

SARGANT, W., Brit. Med. J., 1956, I, 939.

SARGANT, W., Proc. R. Soc. Med., 1958, 51, 353.

SARGANT, W. & SLATER, E., Physical Treatments in Psychiatry, 1944, Livingstone, Edinburgh.

SHAW, F. H., Brit. Med. J., 1960, I, 278.

SIMPSON, K., 1960; quoted from Parker.

SIMPSON, K. & MOLONY, J. T., Med. leg. J., 1957, 25, 53.

SKINNER, J. K., Brit. Med. J., 1960, I, 1058.

STENGEL, E., COOK, G. & KREEGER, I. S. "Attempted Suicide ", 1958, Institute of Psychiatry: London.

STOLKIND, E., Lancet, 1926, I, 391.

STRAUSS, E. B., ibid., 1956, I, 321.

WAYNE, E. J., Pract., 1960, 184, 10.

WEATHERALL, M., Proc. R. Soc. Med., 1957, 50, 617.

WILLCOX, W., Lancet, 1913, II, 1178.

WILLCOX, W., Proc. R. Soc. Med., 1927, 20, 1479.

WILLCOX, W., ibid., 1934, 27, 489 (a).

WILLCOX, W., Lancet, 1934, I, 370 (b).

WILLCOX, W., Brit. J. Inebr., 1934, 31, 131 (c).

WILLS, W. K., Brit. Med. J., 1906, I, 498.

WOODSIDE, M., ibid., 1958, II, 411.

YOUNG, H. M., Proc. R. Soc. Med., 1927, 20,1495.