427 CONVULSION AND INSULIN TREATMENT By DALTON E. SANDS, M.R.C.P.ED., D.P.M. Deputy Physic-ian Superintendent, Banstead Hoipital; Psychiatric Consultant to the L.C.C.; Associate Psychiatrist to King's College Hospital Convulsion Therapy Over I2 years have passed since the first report on convulsion therapy was given by Von Meduna. In that period the treatment has been undergoing the cycle usually imposed on any promising remedy in medicine. Originally used for schizo- phrenia, it was subsequently tried in a wide field of psychiatric conditions. Now it is becoming recognized as a specific treatment for one form of mental illness, psychotic depression, and as an aid in a number of other conditions. It is significant that despite much controversy and the fundamental repulsiveness of inducing a fit, the method has more than held its place over a fair period of time. A radical change in the usual orientation was needed when for centuries the aim had been to cure rather than provoke fits. Yet the idea of using one potentially noxious agent for the relief of a greater menace is nothing new in other branches of medicine. The essential feature in the method still remains the artificial induction of a major convulsion. There have been many attempts to modify it but so far the only result of note has.been the finding that softening or even elimination of the associated muscular movements has had no untoward effect on the results. Indications Depressions. , Because of the inexactitudes of psychiatric classification, the particular form of depression intended needs a little further definition. The case may be certified or un- certified though it is amongst the former that the more dramatic triumphs are scored. These patients are consistently depressed and their attitude remains much the same whether or not they realize themselves to be observed. They show a diminishing state of awareness, a movement away from reality, and an increasing introspection with a narrowed range of thought. They tend to become completely preoccupied with a few ideas. So severe may the process be that eventually they may be totally inaccessible and in a state of stupor. Some show great agitation but this moves within the narrowed orbit of their thinking. They may be very tense but the tension is dependent on their own conflicts and not, as in the tension of the neuroses, readily influenced by adverse external influences. These depressions are mainly endo- genous or constitutional in type. They are seen in the depressed phase of a manic depressive psychosis and in involutional depression with hypochondrasis, nihilism and guilt. The mainly reactive depressions show a similar picture. Occasionally one meets cases precipitated by severe stress, the stress has passed but recovery seems delayed and in these a good treatment result is usually obtained provided it is reasonably certain that the initial stress was the main one. As a rule such exogenous aetiological factors as can be found are inadequate and differ little from common experience. In addition to the severe forms described, there are many milder depressions of the psychotic group who are incapacitated for work, feel ' run down,' tired, ineffective, having headaches or other bodily symptoms of ill-defined distribution and quality. They are rarely certified but none the less are often subject to much suffering and economic stress if allowed to persist unrecognized. They can present much difficulty in diagnosis since their real nature may be obscured by obsessional, anxious, or hysterical symptoms. Since such patients retain a certain capacity for activity they are liable to succumb to the most feared complication of depressive psychosis. They are responsible for a large proportion of those cases reported by the Coroner tss ' Suicide while the balance of the mind was disturbed.' Schizophrenias. This group of illnesses was the first treated by convulsion therapy, but it was soon apparent that results were inferior to those obtained with certain depressions as already described. In spite of this there are some cate- gories of schizophrenia in which the treatment has value. If insulin therapy is, not available,, it is worth a trial in cases of less than a year's duration irrespective of subtype, particularly if the cases are acute, of rapid onset, have definite pre- cipitating stresses and possessed of personalities that have been fairly well adjusted to life hitherto. Such patients would in any event carry a better prognosis than most, and sQme might suggest that. it would be wiser to await the expected remission than to subject a patient to treatment that is not quite free of risk, One has to remember that copyright. on May 24, 2021 by guest. Protected by http://pmj.bmj.com/ Postgrad Med J: first published as 10.1136/pgmj.24.274.427 on 1 August 1948. Downloaded from
Transcript
427
CONVULSION AND INSULIN TREATMENTBy DALTON E. SANDS, M.R.C.P.ED., D.P.M.
Deputy Physic-ian Superintendent, Banstead Hoipital; Psychiatric Consultant to the L.C.C.;Associate Psychiatrist to King's College Hospital
Convulsion TherapyOver I2 years have passed since the first report
on convulsion therapy was given by Von Meduna.In that period the treatment has been undergoingthe cycle usually imposed on any promisingremedy in medicine. Originally used for schizo-phrenia, it was subsequently tried in a wide fieldof psychiatric conditions. Now it is becomingrecognized as a specific treatment for one form ofmental illness, psychotic depression, and as anaid in a number of other conditions. It issignificant that despite much controversy and thefundamental repulsiveness of inducing a fit, themethod has more than held its place over a fairperiod of time. A radical change in the usualorientation was needed when for centuries theaim had been to cure rather than provoke fits.Yet the idea of using one potentially noxious agentfor the relief of a greater menace is nothing new inother branches of medicine. The essential featurein the method still remains the artificial inductionof a major convulsion. There have been manyattempts to modify it but so far the only result ofnote has.been the finding that softening or evenelimination of the associated muscular movementshas had no untoward effect on the results.
IndicationsDepressions. , Because of the inexactitudes of
psychiatric classification, the particular form ofdepression intended needs a little furtherdefinition. The case may be certified or un-certified though it is amongst the former that themore dramatic triumphs are scored. Thesepatients are consistently depressed and theirattitude remains much the same whether or notthey realize themselves to be observed. Theyshow a diminishing state of awareness, a movementaway from reality, and an increasing introspectionwith a narrowed range of thought. They tend tobecome completely preoccupied with a few ideas.So severe may the process be that eventually theymay be totally inaccessible and in a state of stupor.Some show great agitation but this moves withinthe narrowed orbit of their thinking. They maybe very tense but the tension is dependent on theirown conflicts and not, as in the tension of theneuroses, readily influenced by adverse external
influences. These depressions are mainly endo-genous or constitutional in type. They are seenin the depressed phase of a manic depressivepsychosis and in involutional depression withhypochondrasis, nihilism and guilt. The mainlyreactive depressions show a similar picture.Occasionally one meets cases precipitated bysevere stress, the stress has passed but recoveryseems delayed and in these a good treatment resultis usually obtained provided it is reasonablycertain that the initial stress was the main one. Asa rule such exogenous aetiological factors as canbe found are inadequate and differ little fromcommon experience.
In addition to the severe forms described, thereare many milder depressions of the psychoticgroup who are incapacitated for work, feel ' rundown,' tired, ineffective, having headaches orother bodily symptoms of ill-defined distributionand quality. They are rarely certified but none theless are often subject to much suffering andeconomic stress if allowed to persist unrecognized.They can present much difficulty in diagnosissince their real nature may be obscured byobsessional, anxious, or hysterical symptoms.Since such patients retain a certain capacity foractivity they are liable to succumb to the mostfeared complication of depressive psychosis. Theyare responsible for a large proportion of thosecases reported by the Coroner tss ' Suicide whilethe balance of the mind was disturbed.'
Schizophrenias. This group of illnesses was thefirst treated by convulsion therapy, but it wassoon apparent that results were inferior to thoseobtained with certain depressions as alreadydescribed. In spite of this there are some cate-gories of schizophrenia in which the treatmenthas value. If insulin therapy is, not available,, itis worth a trial in cases of less than a year'sduration irrespective of subtype, particularly if thecases are acute, of rapid onset, have definite pre-cipitating stresses and possessed of personalitiesthat have been fairly well adjusted to life hitherto.Such patients would in any event carry a betterprognosis than most, and sQme might suggest that.it would be wiser to await the expected remissionthan to subject a patient to treatment that is notquite free of risk, One has to remember that
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
while some features in the' schizophrenic are ofgood import, these are no absolute criteria for asuccessful outcomne. Therefore, to await re-mission or recovery means taking a considerablechance with the patients' health besides pro-longing an illness, that if not terminated in a shorttime, may have serious occupational, social andeconomic complications for the patient.
Schizophrenic stupors and confusional statesusually respond promptly to convulsion therapy,though liable to relapse if not later treated byinsulin comas. The more violent and impulsivephases of the catatonic type are benefited but theeffect is largely symptomatic. Some schizo-phrenias are mixed with depression. If con-vulsions are used a certain splitting of the psychosisoften results since the depression is improved'whilst the disordered thinking is little affected.
Hebephrenias, paranoid and simple dementingforms of schizophrenia are not benefited as a ruleeven by large numbers of fits. Under this headingshould certainly be included that sub-group' long'known as Dementia Praecox, a term formerlyapplied to all schizophrenias.Mania. Unfortunately convulsion treatment
'does not display the same specific action in maniaas it does in depression. It can mitigate anattack and achieve some degree of symptomaticcontrol that is indeed useful enough, but it doesnot habitually shorten the attack and relapse isfrequent. In any event whether the attack ismanic or depressive, convulsion therapy confersno immunity on the patient in respect of laterattacks.
Paranoia and paraphrenia can be improved if aconcentrated series of fits is given, but patientsrelapse when the confusion and memory dis-turbance, provoked by the fits, begins to recover.'Such types are not normally considered suitablefor the treatment and should not be confusedwith paranoid states that are part of a depressivepsychosis and which usually do well.
Generally '.speaking neuroses are a contra-indication to this form of treatment and many maybe made worse. No good case has been estab-lished for treating thet ordinary uncomplicatedneurotic patient by this means. The chiefdifficulty is in diagnosis, since a psychosis ofmild severity can be considerably masked byneurotic symptoms. Erroneous conclusions areapt to be drawn on the effect of the treatment inthe belief that a neurosis has been'benefited. Inanxiety states the e-ffect is usually to stimulatemore anxiety arid tension.
Hysterias and anxiety neurotic patients are aptto fix their neurotic symptoms on any physicaldiscomfort that may arise out of the'treatment andt6 complain strongly of the same. Temporary
dramatic relief can be obtained in hystericalstupors but such conditions can be relieved in lessdrastic ways and in any event psychologicaltreatment is required afterwards.
Obsessional neuroses often benefit while thetreatment is proceeding, but relapse when itceases. They are liable to complain early of thememory disturbance and forgetfulness in a wayrarely seen in psychotic patients.
Recently Milligan (1946) has treated chronicneuroses that have proved resistant to othermethods by concentrated fits up to three or four aday for four to six days followed by others morespaced out. The treatment was stated to havebeen successful but further experience is neededbefore conclusions can be drawn. A case ofchronic stammering that had proved very re-sistant to the more usual lines of treatment suchas speech re-education, was benefited by con-vulsion therapy given every five days though liableto relapse in times of stress (Owen and Stemmer-man, I947).
Contraindications. Depressions associated withorganic brain disease or cerebral trauma. Attemptshave been made to treat the depressed forms ofgeneral paralysis of the insane and even of pre-senile dementia. Though some transitory relief inthe depression may be achieved the dementingcourse is accelerated and incontinence may appearfor the first time. Similarly depressions that arelikely to have been precipitated by severe cerebralarteriosclerosis should not be treated. In patientswith a history of cerebral trauma, it is necessaryto know how far the trauma can be said to havecaused organic change. Where neurological andelectro-encephalographic investigation show thatthis has occurred, convulsion therapy is to beavoided, though psychotic depressions that havehad some 'definite blow on the head withoutsustaining cortical or other damage, may be'treated.
Contraindications are an individual matter inmost cases. The gain anticipated has to 'bebalanced against the risks involved with considera-tion for the hazards of leaving the patient un-treated.' For instance, a patient with mitralstenosis associated with strong suicida1 tendenciesmight justify one taking more risk in treatmentthan on the same patient not so 'complicated.There are few absolute contraindications butamo'ngst them should be placed coronary disease,aneurysm of the aorta, peptic ulcers in acutephase and decompensated diseased myocardia.
Systolic blood pressure over 2oo and diastolicpressure over i lo should be a bar to treatment ifdue to renal rather than emotional causes. In theabsence of renal lesions the presence of -theemotional' type of high pressure may be confirmed
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
August 1948 SANDS: Convulsion and Insulin Treatment 429
by the injection of o.z to 0.4 gm. of sodiumamytal intravenously, when a fall of from 50 to8o mgm. is usual, and contrasts with the slightchange seen in renal cases. Hypertensions of theemotional type can be safely treated by E.C.T.The liability to cerebral haemorrhage is muchless than might be supposed. Cases do notappear to have been reported with the possibleexception of one who was a poor candidate in anycase.
Pulmonary tuberculosis and pregnancy haveboth been associated with E.C.T., but are bestavoided. The former has been discussed. Thelatter appears to carry little risk if treatment isgiven before the end of the third month. Two ofthe author's patients were delivered of normalfull time infants after having an average number offits between the second and third months ofpregnancy.
Technique and Management ofConvulsion TherapyGenerally speaking pharmacological methods of
inducing a fit have been superseded by electrical.Von Meduna revived the suggestion of Weick-hardt, in 1798, to use camphor injections as thefit provoking agent, but it was uncertain and themore reliable cardiazol (or metrazol) held the fieldfrom I933 to 1937. Cardiazol is given by therapid intravenous injection of from 4 to 9 cc.according to tolerance. The drawbacks are thenecessity for repeated injections two to three timesa week, the highly unpleasant sensations ex-perienced by patients at the onset of. the fit andthe fact that at least half the patients vomit after-wards even when the treatment is given fasting.Intravenous injections can be avoided'by the useof triazol intramuscularly (Mayer-Gross andWalk, 1938), but the resulting fits are uncertain in'incidence and a succession of fits sometimesresults from 15 minutes to an hour after. injection.In practice this modification is difficult to control.Cardiazol is still advocated by some psychiatristsfor catatonic stupors or paranoid psychoses, andit is sometimes used in combination with insulincomas for schizophrenia.
Various attempts were made to lessen the un-pleasantness of the injections.*- Many sedativeshave been tried for premedication. Cook (I944)advocated the use of hyoscine, while even generalanaesthesia was used by Neustatter, et al. (I939).A moderate degree of hypoglycaemia, to the pointof stupor or light coma, was used by Sands (I939).Recently Tow (I947) has- descrnbed a methodihich involves the injection of sodium amytal assoon as respiration is re-established, with theproduction of an amnesia for the unpleasantconcomitants of cardiazol injections.
From IO to 30 seconds xafter the cardiazolinjections a grand mal type of epileptiform attackoccurs, and a cry often heralds the onset of thetonic phase. With the onset of this phase themouth usually opens for a few seconds, and if agag is not already in position it can be slipped inat that moment. From io to 15 seconds after thestart of the fit the clonic movements- begin, andfor about 30 seconds steadily increase in amplitudebut decrease in frequency. During this phase thepupils dilate and the conjunctivae become en-gorged. After a few seconds the patient takes along breath, having been apnoeic since the startof the fit, so that by the finish a fair degree ofcyanosis is common. As soon as the clonic stageceases the head should be tilted to one side and thejaw pushed forward by a finger placed behind theangle on each side. Occasionally t4e tongue fallsback despite this precaution, and tongue forcepsshould be at hand. If the resumption of respira-tion is delayed by more than 7 to 10 seconds, thereshould 'be no delay in applying artificial respira-tion, and oxygen with 5 per cent. carbon dioxidegiven through a B.L.B. mask. Respiration iseasily resumed after a short application of theseremedies. The use 'of oxygen and artificialrespiration need not be routine, but one shouldalways be on the alert to give it especially in thoseof more advanced years. Immediately after thefit ceases the pulse is often scarcely perceptible, butin most patients it rapidly regains volume thoughit may remain fast for several minutes. Tem-porary arrhythmia occurs occasionally but as suchhas no special significance.
Patients vary greatly in the time required toregain consciousness and so far' these variationshave not been associated with any specialpsychiatric types. Some are awake within aminute or two of the end' of the fit, while othersmay need ten minutes-to regain consciousness.While the majority recover quickly, some showpost-convulsive excitement and may leap out'ofbed. Most patients are cbnfused in varyingdegrees for about half an hour to an hour after afit. Nearly all ate easily startled for ten' minutesafterwards and -will jump violently if' anyoneu-nexpectedly appears; It is necessary for staff tobe at hand and 'alert in this period since in thepost-convulsive period impulsive movement' maybe "made very suddenly in one hitherto quiet.'Som'e become temporarily 'emotional and de-pendent, needing reassurance. The patient doesnot remem'ber the conv-ulsion excepting the onsetafter the injection'as noted. " '-
Electrical' Con"vu'lsive Therapy (E.C.T.) asdescribed by Cerletti and Bini (I937) has larelysuperseded the chemical method. Since loss ofconsciousness is instantaneous, the patient'feels
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
nothing of the fit at any stage till consciousness isregained, hence patients objections to the treat-ment are few. The objectors are chiefly those whoby reason of their mental state resist any kind ofmedicbl or routine nursing attention. Vomiting isvery unusual afterwards and a light meal can behad up to an hour before treatment without fear ofthis complication. It is much simpler and easierto apply than the pharmacological method.
The machines are designed to supply a certainvoltage, usually between 70 and I 50, for a setperiod of time from o.i to i second, using al-ternating current from electric light mains, or ifonly direct current is available, by the aid of a'transformer.
.More recent apparatus employs
monophasic or diphasic current measured inJoules. After the skin has been cleaned withmethylated ether, the current is applied by meansof an electrode to each temple just clear of thehair bearing area. The electrodes are covered bypads of gauze or lint soaked in 20 per cent. saline.There are several ways of actually applying theelectrodes but the simpler the method the better.With some apparatus the patient's resistance ismeasured before the fit is given. The patient'sresistance gives no indication of the strength of theshock needed to effect a convulsion since livingtissues do not obey Ohms' Law. Its chief valuewas to indicate the efficiency or otherwise of thecircuit and since this can be done in other ways,the practice of measuring patients' resistance haspractically dropped out. Resistance diminishesfor three to four successive shocks but thereafteris little changed.When the switch is pressed the patient gives a
start as the current passes, and the fit follows atonce. The course of the fit is similar to thatdescribed after chemical injection. If the currentis insufficient a ' sub-shock' occurs. Somepsychiatrists immediately give a further shockwhile others prefer to wait till breathing is re-established and all chance of a delayed fit afterI0 tp 15 seconds gone. The dosage can either beincreased or merely repeated according to one'sestimation of how nearly the fit was missed. Theinitial voltage and time, or number of Joules,varies with different types of apparatus. Onesats with the basal time and voltage recom-mended for the particular machine and if a sub-shock occurs the time may be increased by O.Isecond or the voltage by ten, or both. Patientswho have many sub-shocks do less well clinicallythan those who largely avoid them, so that it isadvi$able to be fairly generous with the dosage andsecure regular fits rather than give too little andhave to repeat- shocks. In a patient of averagephysical helth one can give as much as three to
fotir shocks in succession if the first three fail tocause a convulsion.
Dentures and hair clips must be removed beforetreatment, and the bladder emptied.
Complications(i) Fatalities are certainly rare. Kalinowsky
(I946) reported on 2,000 cases without a death.the author has known of one, due to coronarydisease, in over a thousand patients, in which deathcould be said to have been hastened by the fit.Kolb and Vogel in a survey of all Americanhospitals found a death rate of o.6 per cent. forE.C.T. and O.I per cent. for metrazol (cardiazol),while Impastato and Almansi found o.8 per cent.in a survey of the literature on E.C.T. Un-fortunately in many reports the actual connectionbetween the treatment and death is poorlydescribed and it is difficult to form an opinionfrom a statistical point of view of the risk entailed.One risk has to be balanced against another sincethe death rate from suicide is much higher amongstthose depressions not treated by E.C.T. thanamongst those so treated. Proper physicalexamination supported by X-rays and otherspecial methods if necessary will exclude thevast majority of conditions that can determine afatal outcome. For comparison it is well knownthat uncomplicated single fits in idiopathicepilepsy are very rarely fatal.
(2) Fractures and dislocations. Collectivelythese are probably the most troublesome of allcomplications. The most common are com-pression fractures of vertebrae bodies although inI2 years' experience of the treatment no lastingfunctional disablement appears to have been re-corded. The vertebrae most often affected are the5th to 8th dorsal with occasional fractures as highas the third dorsal and as low as the second lumbar.While the vertebral body may be uniformly com-pressed, a wedging is commoner with the narrowend of the wedge at the anterior border. Theextra support of the pedicles seems to prevent theposterior part of the body being appreciablyflattened as a rule. These fractures usually happenwithin the first five fits. The patient should bekept in bed for a fortnight and then treatment canbe resumed if preventive measures as describedbelow are used.These fractures may be associated with no pain
at all or with some soreness and tenderness topercussion over the vertebrae Involved. Veryrarely, if local swelling and oedema is sufficientto press on spinal nerves, there is referred painround the sides of the chest to the epigastrium. Ir.one of the author's patients it was severe enoughto need morphia for a few days. Such painsalways clear up with rest, with or without sedatives.
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
August I948' SANDS: Convulsion and Insulin Treatment 431
Many patients complain of pains in the back,especially after the first few fits, but unless theseconform to the types mentioned they have nodiagnostic value regarding the presence of afracture. Many complaints of pain in the backare really due to muscular soreness and crampfrom the unusually severe exercise involved in afit. Radiological diagnosis cannot be used to fulladvantage unless all patients are X-rayed beforetreatment, anteroposteriorly and laterally. Inpractice it is now considered unnecessary to dosuch routine X-rays before or after treatmentunless specially indicated. The incidence of thesefractures has been estimated variously by severalauthors but the average is represented by thereport of Cook and Sands (I94I) at I4 per cent.of male and I2 per cent. of female patients if noprophylactic measures are taken. The incidenceis reduced to approximately 4 per cent., and theseverity of the lesion is less if care is taken to givethe fits with the spine moderately extended. Withfracture boards under the mattress, a firmly rolledblanket is placed under the mid thoracic spine;one nurse exerts pressure on the shoulders overthe heads of the humeri, another fixes the pelvis.In this way sudden flexion at the passing of thecurrent, or during the fit, so liable to damage thearterior parts of the vertebral bodies, is avoided.Similar spinal fractures occur less frequently andwithout symptoms in epilepsy. None the less it isobviously desirable to minimize damage byextension of the spine as above.The most serious fractures are those of the head
of the femur. The author has not yet experiencedone but they occasionally occur, even in both bonessimultaneously. Elderly and very muscularyounger patients are the most liable. Except forkeeping the pelvis fixed, it is probably desirablenot to restrain leg movements during the fit.Fractures of the acetabulum rarely have been re-ported. Fractures of the humeral neck and dis-locations at this joint are recognized, and areespecially liable to happen if the patient flings thearms outwards at the start of the fit. Dislocationsare usually forward and if the -shoulders areproperly held during the fit, the head of thehumerus may be felt thumping against the hand ofthe restraining nurse. It is advisable, therefore,to fold the arms across the chest and keep thebedclothes well tucked in at the sides so as toavoid any sudden excursion, but not so as to fixthe arms completely.
Fractures of the scapula are rare. Dislocationof the jaw is perhaps the most frequent complica-tion but is easily prevented and remedied. Toprevent it, the jaw should be held firmly up frombelow. Some dislocations right themselves spon--taneously in the later stages or at the end of the
fit, and if not, the jaw can be easily replaced byhand, with the protection of a towel in the relaxedperiod at. the close of the fit before consciousnessis regained. Muscles, tendons and ligaments arerarely injured in convulsion therapy.
In practice, if care is used, these complicationshave not assumed an importance sufficient to~cause discontinuing the treatment. By the intra-venous injection of curare before each fit the risksof fracture or dislocation can be exluded, Bennett(1940). Such injections carry their own liabilities,so that most psychiatrists only use it in patientsknown to have a special risk of fracture or for theprotection of an already damaged limb, perhapsfrom a suicidal effort.
(3) Respiratory complications. The chief diffi-culties are the apnoea already described, aspira-tion pneumonia and activation of quiescent pul-monary tuberculosis. It may be added thatapnoea is more frequent after electrical thanchemically produced convulsions, and more usualafter sub-shocks than full fits. Aspirationpneumonia due to the inhalation of septic materialmore especially in the presence of dental sepsis israre. Activation of latent pulmonary tuberculosishas been reported and, generally speaking,patients with quiescent and active pulmonarytuberculosis should not be treated with convulsiontherapy. Any doubtful case should be X-rayedfirst and subsequent films taken during and aftertreatment. One cannot give an entirely hard andfast ruling on this point since there are times whena tuberculous patient who has become very de-pressed, and not taking food properly, may bebenefited by E.C.T. His neglect of himself byself-starvation or attempted suicide may prove agreater hazard than the tuberculosis. Clearly onerisk has to be weighed against another.
(4) Cardiovascular complicatins are occasionallyencountered. Cardiac arrest with extreme pallorand widely dilated pupils has been seen and mayrecur several times in the same patient leading tocessation of treatment. It is thought to be due tocerebral vagal stimulation. Subconvulsive re-sponses seem especially liable to provoke thisphenomenon. Convulsions have at tinrtes ex-pedited coronary thrombosis in those so pre-disposed. On the whole one has to considerwhether the patient's cardiovascular system canstand the muscular exertion of the fit. If necessarydo exercise tolerance tests, and use electrocardio-raphy. Provided compensation is satisfactorythere are many patients with cardiac lesions whocan be treated successfully with E.C.T.
Blectro-cardiographic changes so far reported,have not been striking, nor specially useful indeciding fitness for treatment. The most fre-quent abnormalities are minor arrhythmias, though
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
they are absent in spontaneous epileptic seizures.Sinus tachycardia is followed by arrhythmia andextra systoles in the convulsive response, while insub-shocks a bradycardia is common. This isthought to be due mainly to vagus stimulationwhich in the major fit is counteracted by theassociated muscular effort. The author has seena sub-conjunctival haemorrhage and others havebeen noted. These absorb without difficulty andtreatment can be continued.
(5) The only neurological complication of noteconcerns the occurrence of spontaneous fits daysor months after E.C.T. There has been muchspeculation as to the possibility of the treatmentcausing epilepsy in those in whom the disease washitherto unknown. Occasional fits have been re-ported weeks or months after convulsion therapybut the view of the majority is that such fits onlyhappen in those patients possessed of a pre-disposition to epilepsy prior to treatment. Sucha predisposition can be shown by an electro-encephalogram (E.E.G.) and no patient has beenknown to have fits after a course of convulsiontherapy whose E.E.G. was formerly within normallimits.
(6) Psychiatric complications are few and are tobe distinguished from those phenomena habituallypresent in greater or lesser degree in any patienttreated by a course of fits. The latter are des-cribed under common forms of reaction to thetreatment. One well recognized complication isthe transformation of a manic depressive patientfrom depressed to manic phase by over frequent orlengthy treatment. Rarely activation of a psy-chosis may occur after a few fits, with an acuteflare up of psychotic symptoms. As a rule thepatient settles down again and after a week treat-ment can be continued. It is essential to differ-entiate this occurrence from the commoner organicreactions described later. Occasionally somepatients, more particularly men, become excitedduring the half hour immediately following thefit. They may ffing themselves out of bed or evenattack those near. The intravenous injection ofsodium amytal 0.2 to 0.3 gin. immediately beforethe induction of a fit generally prevents suchevents without stopping the fit.
(7) The electro-encephalogram shows changesresulting from E.C.T., though not if sub-shockswithout a convulsion are administered. A pro-gressive slowing of the normal alpha rhythmoccurs until considerable delta activity is evident.Such changes depend on the frequency andduration of the E.C.T. and also on individualsusceptibility. As a rule the more often fits aregiven and the longer treatment lasts, the greaterthe degree of dysrhythmia, though one patientmay be much slower than another to show E.E.G.
abnormalities. There is no correllation betweenthe onset of improvement clinically and thechanges in the E.E.G., though changes in memoryand confusion parallel the E.E.G. more closely.E.E.G. abnormalities, like the defect of memoryand confusion, are reversible, and disappear in twoto eight weeks in the vast majority of patients.Those who have had most fits are slowest inrecovery from dysrhythmia.
General management of the treatment(i) When to begin ? For a suicidally depressed
patient not under control it is necessary to begintreatment at once and repeat in a few hours ifconditions of safety are still lacking. If a patient isrefusing food, one to three fits on successive daysgenerally ensures a resumption of feeding, andtube feeding may be stopped. Sometimesnutrition is so bad that artificial methods offeeding have to be used for a few days before thepatient is strong enough to be treated. In theabsence of these complications, there is little pointin delaying treatment once examination is com-plete, since most patients have been ill for monthsbefore admission, and it is less than human to pro-long their suffering unreasonably. Delay is onlyindicated if the diagnosis is doubtful, if the onsetis very recent, e.g. a week or two, or if there is ahistory of recurring short attacks from whichspontaneous remission is likely. Successful treat-ment is possible even after three years of illness.
(2) How often and how long to give treatment ?Most depressions do well with two fits weekly,having perhaps three in the first week, with atotal of four to ten. Depressions complicated bydelusions tend to need more, and up to 15 aresometimes necessary. Schizophrenias requirethree fits per week and a total of 14 to 2o as a rule.Manias are best treated energetically at first withdaily fits or two fits a day for three days, laterspacing out to three a week as the acute phase iscontrolled.
(3) When to stop. When a depression is clearof symptoms two further treatments should begiven before treatment is terminated; similarly inschizophrenia, though more caution is needed instopping the treatment ofthis disease. Depressionswhich are part of a manic-depressive psychosisneed most careful observation for the earliestindication, usually failure to sleep, that mayherald the onset of a swing into mania. It shouldbe possible to prevent it by prompt termination oftreatment and the use of sedatives.
Relapses. On this point it should be em-phasized that convulsion therapy can cause re-mission of a psychosis but it confers no immunityto subsequent attack. For instance, those patientswhose depression is of the recurrent type, con-
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
August i948 SANDS: Convulsion and Insulin Treatment 433
tinue to have further attacks. Apart from thisthere are several factors liable to precipitate re-lapses. One of the commonest is a failure todiscover factors that have played a dominant partin causing breakdown. If the patient is too in-capacitated by depression to co-operate in a fullenquiry prior to treatment, attempts can be madeto complete the history and to search for thestresses or faults in the patient's way of life as soonas improvement begins to show itself.
Insufficient treatment. When improvement hasset in, patients sometimes wish to break off treat-ment. It is usually unwise to do so beforesymptomatic recovery, followed by two furtherfits. Relapses and indifferent results are morelikely in those depressions which occur in personsalready subject to chronic neurotic symptoms, orhaving psychopathic personality, organic braindisease or schizophrenic features. Schizophreniamay begin with symptoms of emotional qualityand when those are cleared by E.C.T. the essentialschizophrenic nature of the illness is then shown.The total relapse rate for the treatment is
approximately 28 per cent., but if transitory andminor relapses are exluded, the figure is not morethan 14 per cent. The majority of relapses occurwithin the first five weeks after the end of treat-ment, and observation should be maintained forthis period. In practice many patients have to betreated psychologically or may need varioussocial adjustments done for them in this time.
Out-Patient E.C.T. In view of the quick re-covery many patients make from a fit, out-patienttreatment is possible and often practical. It isindicated where the illness is mild and the patientfairly well controlled, and if it is desirable to keepthe patient at work. Some involutional depressionsand recurrent cases who have been treated, as in-patients on a former occasion, and whose reactionto E.C.T. is known, may be so treated. Manic-depressives in depressed phase should be excludedowing to the liability of a swing into mania.Because of the risk of confusional states especiallyin younger patients, not more than two fits a weekare advised, though if as commonly happens onlyone fit a week can be managed treatment mayprove inadequate. So far as possible, out-patientE.C.T. should be backed by the resources of ageneral hospital in case of complications. Patientsare allowed to rest for an hour after treatment andshould go home with relative or friend. On thewhole, out-patient treatment is something of amakeshift, but is better than none at all. Owing toslight tension before, and confusion after thetreatment, it is difficult to get a true estimate ofthe patient's condition if seen on the same day,and one misses the first hand observation andfuller interviews that go with in-patient care. It
is difficult to vary frequency or spacing of treat-ment to individual needs and the procedure tendsto become a very mechanical affair.
Psychiatric observations. Some phenomena arecharacteristic of all patients having convulsiontherapy. There are two main lines along whichimprovement or recovery develops. The mostdesirable and specific is that shown by the re-covering psychotic depression. In these patientsa change for the better may be discerned after onetreatment but more often after two or three. It isin the essentials of life that gain first appears, insleeping and eating. More notice is taken ofenvironment and there is less preocupation withself. Speech returns in those who were stuporose,and better attention is given to elementary hygiene.The patient becomes more accessible whether thedepression was originally of retarded or agitatedtype. Questions are asked showing a suitabledesire for reorientation, and the patient throws offthe appearance of misery. Disorders of thinkingusually recover last, namely delusions of wrong-doing, guilt, disordered bodily function and thelike. These changes occur at varying rates indifferent patients. In some it all appears tohappen overnight, in others over a week or two.It is most impressive to encounter the normalpersonality for the first time when relieved of thehavoc of a depressive psychosis. In some degreesuch a transition should be apparent in all wellselected depressions.While the foregoing picture is characteristic of
the treatment response in emotional or affectiveillness, a differept reaction is usual in illnesses thatare primarily disorders of thinking; the schizo-phrenias, paraphrenic and paranoid psychoses.Most of these show no benefit until a mild tem-porary organic confusional state has been pro-duced. To achieve this, it is necessary to givemore frequent fits, at least three a week, for thelonger periods customary in such cases. Con-fusion generally appears between the fifth andeighth treatments. Patients complain of loss ofmemory, of forgetfulness, have little capacity forcoinected thinking, are cheerful in a simple,easily satisfied way and apt to lose bits of property.Sometimes under the cloak of such confusion alasting recovery may develop, but too often in oneto three weeks after ceasing treatment the con-fusion clears and schizophrenic or paranoidsymptoms reappear. Consequently, though in-di-vidual patients do well, the results in the schizo-phrenias are little different from controls over largeseries of cases. It is characteristic of neuroses thatthey complain of memory loss much earlier thanthe psychoses, but with rare exceptions there is noevidence of lasting defect in memory or intellect inany type of patient.
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
Psychological treatment. While in numerousinstances convulsion therapy is clearly the chiefagent in recovery, and some depressed patients dowell with little or no other assistance,'there areothers in whom psychotherapy is the other half ofthe necessary treatment. Particularly is this so incases associated with illness precipitated by en-vironmental stresses (reactive types), and in thosecomplicated by neurotic features of hysteria oranxiety. It is not a question of a physical or apsychological approach to treatment. The methodsare complementary and their use is regulated bydiagnosis, and the progress of the patient. Mostof those suitable for E.C.T. are not accessible topsychotherapy, and proper rapport with thephysician is not obtained. After depression hasbeen eased by E.C.T. patients are able to benefitfrom psychotherapy provided this is not attemptedon the same day as a fit has been given, whenconcentration is temporarily impaired. In de-pressions secondary to neurosis, psychotherapy isthe treatment of choice. It is also indicated indepressed patients who have been adequatelytreated with E.C.T., shown a remission, and thenrelapsed owing to reactive factors or personaldifficulties hitherto unsuspected.
Results are similar whether the electrical orpharmacological method is employed. In de-pressions of manic-depressive psychosis, in in-volutional depressions and simple depressiveattacks, recovery rates of 75 per cent. to go percent. have been reported frequently. Investiga-tion has even been made to ascertain why fivepatients out of a series of a hundred did not im-prove. It is known that these patients mighteventually recover or improve spontaneously ifuntreated after from three months to a year orlonger. About a third do not spontaneouslyrecover. Nowadays mental hospitals contain fewchronic depressions where E.C.T. is employed.Such as are still hospitalized are mainly those un-treatable for physical reasons. Depressions thatare allowed to continue untreated often remainexposed to suicidal and other risks. They areliable to lose their employment and have muchdifficulty in rehabilitating themselves if theyrecover later.
In mania results do not approach those seen indepression. Because the condition is so much lesscommon than its depressive counterpart, largenumbers of cases have not been reported. It isdoubtful' if convulsion therapy shortens the totaltime'of the attack, but there is no doubt that aeoncentrated course of fits can bring the manicattacik'under control more safely and rapidly thannarcosis unless thie illness be of a mild typ'e.
Schiotphrenias. Cases of less than a year's'duration often show favourable results but in all
cases relapse is frequent, and over large series ofpatients the follow-up results are little differentfrom controls. It is in the sub-groups of catatonicexcitement and stupor that better results areobtained. The results in schizophrenia treated byconvulsion therapy are inferior to those obtainedwith insulin coma treatment, but if the latter is notpossible then E.C.T. should be given in the aboveclinical categories. Chronic schizophrenias canoften be helped through their more difficult phasesby repeated treatment over a long period.
Commentary. Now that a dozen years and morehave elapsed since the first trials of convulsiontherapy were made, some estimate of its place inpsychiatry is possible. The method has beenfreely criticized though mainly on theoreticalgrounds. When given at the usual rate of two tothree treatments weekly for a total of I2 or less,there is no evidence with the methods available atpresent, of irreversible pathological changes in thebrain, and with much larger numbers of fits it isremarkable how little can be found. In animalssubjected to convulsive states more severe thanthose given to patients, petechial haemorrhageswere discovered, but confirmation has been lack-ing. Such haemorrhages do not appear after themuch longer passage of a current (7j minutes) inelectronarcosis nor in the judicially electrocuted.Necropsy findings in deaths associated with con-vulsion therapy have been meagre. In this respectthe treatment differs strongly from the pathologynoted after insulin coma deaths. The brain isknown to be able to support an amount of currentfar in excess of that commonly used for E.C.T.In practice it is better to give ten volts too muchor a time of 0.2 second too long than have manysub-shocks. Technically the treatment is simpleand does not need any great knowledge of elec-tricity. Complications are few when it is care-fully administered. The skilful part of thistreatment is not in 'pressing the button' but inpainstaking examination preceding its use, inpsychiatric interviews between treatment, decidinghow often to treat and when to stop.The rationale of the treatment is not known.
Metabolic, biochemical, diencephalic, pavlovianand psychological mechanisms have theiradvocates.The interesting work of Masserman and Jacques
on electro shock and experimental neurosis in catssupports the view that the treatment produces acertain amount of disintegration so far as complexneurotic patterns, such as phobias, compulsionsand inhibitions are concerned. It is then possiblefor latent and more normal patterns to re-establish themselves, for new forms of adaptivebehaviour to arise, and for re-education to pro-
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
August 1948 SANDS: Convulsion and Insulin Treatment 435
ceed. They also remark that the ' price of re-covery" was a variable loss of certain higheradaptive capacities.' Some of their animals hadnot regained their notmal efficiency in complexskills of which they had formerly been capable bythe end of the post shock testing period.With patients, clinical observation suggests a
similar process goes on, but the recuperativepowers of the brain are such that the majority ofthose who are considered recovered reach a levelof efficiency which, to near relations, is the same asprevailed before breakdown. The Rorschach and-the Minnesota multiphasic personality inventorytests were employed by Pacella, et al., I947, toassess the effects of E.C.T. on some aspects ofpersonality in 75 psychiatric patients of schizo-phrenic, manic depressive and psychoneurotictypes. The Rorschach results of successful casesshowed that basic personality defects remainedafter treatment, but that thought, emotions andaction were under better control, and moreefficient use of remaining capacities was made.There was no gain in intellectual capacity inschizophrenias but improvement in intellectualefficiency wad notable. The Minnesota test wasnot specially useful except in determining ifimprovement in schizophrenia after treatmentwould be retained.
Insulin Coma TreatmentIt is I5 years since Sakel made his first report on
the treatment from the University Clinic ofVienna. The I939-45 War caused the closure ofnumerous insulin units in this country, but manyhave since resumed activity. Although long termresults are open to criticism, the method remainsunique in being the only one known to regularlycause schizophrenic patients to change for thebetter, if one excludes certain types treated byleucotomy. In view of the chronic nature of manyschizophrenic illnesses, this observation is ob-viously of cardinal importance even if it falls farshort of perfection. The treatment is particularlyhelpful in improving the disordered thinking ofthe schizophrenic patient. Any influence it mayhave on emotional disturbance is probablysecondary to this. It is not useful in manic de-pressive and other psychoses where emotionaldisorder assumes the primary role.
Selection of patients. Schizophrenias are theonly group of psychiatric disorders in whichinsulin coma therapy is widely used, and thedecision to give or withhold treatment is madeafter considering the following factors:
(I) Type of schizophrenia. The most suitablepatients are. the paranoid schizophrenias, followedby the catatonjas; while simple and hebephrenic
types are the least amenable to treatment. Manypatients are not easily classified in these maincategories and are termed ' atypical "because ofmanic, depressive, hysterical, obsessional or othercomplicating features. Many such cases haveresponded well. There is little evidence thatinsulin substantially helps those who can truly belabelled Dementia Praecox, where dementing pro-cesses set in during adolescence and progresssteadily to complete invalidism. Fortunately suchcases represent less than io per cent. of the wholeschizophrenic group, though formerly the termwas applied incorrectly to all schizophrenias.Patients having this malignant form of schizo-phrenia are most likely to be found amongst thesimple and hehephrenic types of more recentclassifications.
(2) Duration of illness is perhaps the most im-portant single consideration. All statistics agreethat cases of a year or less show substantiallybetter results than those of longer duration. Aftertwo years few patients can be benefited by thetreatment. Thus while several months delay doesnot of itself vitiate the result of a depressivepsychosis treated by E.C.T., a similar delay inschizophrenia is likely to be fatal to a patient'srecovery with insulin. For this reason Sargantand Slater remark that early cases of schizophreniashould have high priority on waiting lists. Ifcareful enquiry is made, it is generally found thatthe disorder has lasted longer than was firstapparent. The presence of previous attacks, es-pecially if followed by satisfactory remissions, isno bar to treatment and may be associated withgood prognosis.
(3) Type of onset. The slow insidious onset,lacking any precipitating cause, is unfavourable fortreatment and prognosis; while those whosuddenly fall ill, showing a wealth of symptoms,with a history of severe precipitating stress,generally react well.
(4) Prepsychotic personality. There is no doubtthat the chances of a successful outcome are in-creased in those patients known to have shown ahealthy adaptation to the difficulties of life beforebreakdown. Such patients are generally found tohave been reasonably successful in their occupa-tional, home, social and other personal relation-ships. By contrast, treatment is unfavourably in-fluenced in those whose history makes it clear thatthey have been shy, diffident types from an earlyage, prone to circumvent and avoid awkwardsituations when a solution is necessary. They donot mix well, form few friends, and are apt toachieve their ambitions in day dreaming ratherthan in reality. From a mental backgrQund of thisnature, schizophrenia is liable to develop i n-
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
sidiously, and the patient's illness appears as analmost logical development in a predisposed per-sonality. While insulin treatment may arrest theprocess towards a seclusive fantasy life, there islittle reason to suppose that constitutional trendsof this nature can be radically altered. Similarly,those hampered by intellectual defect, psycho-pathic or neurotic qualities in the personalty havea less favourable chance with treatment than inthe absence of such traits. In line with the fore-going is the view that the extravert personalityreacts better than the predominately introverted,and of physical constitutional types the pyknicand athletic habitus are more favourable than theasthenic or dysplastic.
(5) The factors of heredity, age and sex are ofless value in assessing the outcome. Hereditarytainting in near relations was thought to be a dis-advantage but a number of cases thus handicappedhave done quite well. Generally speaking, theresults are poor in patients under i8 years of ageand after 40. The former appear to have littleresistance to the schizophrenic process and viewsare conflicting regarding the late age groups.There is no substantial difference between thesexes as to results.
Risks and contraindications. In the hands of ex-perienced insulin therapists, risks are no greaterthan are accepted in other branches of medicine.In large series of cases collected from severalhundred hospitals in the U.S.A., mortality figuresvaried from 1.29 to o.85 per cent. (Kalinowski andHoch, 1946). Mortality figures are usually higherin surveys covering numerous hospitals and thefigure is not more than o.s per cent. in the mostskilfully treated series. The cause of half of alldeaths is hypoglycaemic encephalopathy after ir-reversible coma, which is largely an avoidablerisk. The remainder are mainly due to aspirationand other forms of pneumonia, and to cardio-vascular disturbances.
Despite these reports the risks of treatment areon the whole less than the risks of waiting for aspontaneous remission to occur. The figures ofNew York State hospitals show a higher mortalityamongst the untreated compared with thosetreated. The untreated not only carry the risk ofmental invalidism but of physical deterioration,proneness to tuberculosis and intercurrent illness.Twelve patients died in a control group comparedwith one in the treated series.
Mentally the chief risk is of a mild degree ofintellectual impairment where long series of deepcomas are required. Only a small number ofpatients ever -show this feature and it is probablyless than a similar impairment seen in convulsiontherapy. Greater degrees of intellectual deficitare seen-after irreversible comas, when a Korsakow
type of picture is common. It usually improvesthough recovery is not always complete.
Contraindications are on advanced degree ofcardiac disease, renal and liver diseases, diabetesmellitus and thyrotoxicosis. Treatment should besuspended for any fibrile illness.
Technique of treatment. Although suitableliterature is available (Sargant and Slater, I948;Kalinowski and Hoch, 1946), the acquisition ofsound technique is an essentially practical matter.On this account and in view of the necessarylimits of this article, it is proposed to deal mainlywith some observations which further experienceof insulin in both general and mental hospitalssuggest as important. From a purely technicalpoint of view insulin coma treatment demandsfar more skill from both medical and nursing staffthan does convulsion therapy. Preliminary ex-perience in an active insulin unit is desirablebefore embarking on treatment for the first time..It has been shown that the difference between theexpert and the indifferent insulin therapist canmean as much as 30 per cent. in the recoveryrate after treatment. Generally speaking, six casesis the maximum that one doctor can manage anda colleague must be on call. The situation issimilar to looking after six general anaesthetics atthe same time for five or six days a week. Thesame therapist should be in personal charge for sixmonths to a year or more and frequent changes ofmedical or nursing staff are to be avoided. Thenursing responsibility is a matter for a seniormember of the nursing staff experienced in thetreatment. She should not be liable for any dutiesoutside the insulin room during the morning. Twoother nurses in training assist her and a fourth isavailable if required. It is advisable for the seniornurse to be unchanged for nine to I2 months. Inwards with patients under insulin treatment, it isoccasionally necessary to deal with manifestationsof hypoglycaemia during the rest of the day ornight, so that staff and apparatus to combat signsof relapse must be available. For this reason,patients are not dispersed to other wards after thetreatment and are only allowed out for shortdistances if accompanied by a nurse carrying anemergency ration of sugar or sweets.The main routine of treatment has altered little
since Sakels' original outline was published. In-sulin is injected intramuscularly at 7 a.m., thepatient having fasted since supper time on theprevious evening. If intravenous insulin is used,the injection time is 9 a.m. The initial dose isusually 20 units, increasing by io units daily untilgo units has been reached, and then by 20 or even40 units a day until sopor occurs, when theincrease may be slowed. During the first twohours patients lie dozing, and towards the end of
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
August I948 SANDS Convulsion and Insulin Treatment 437
this time begin to flush and perspire. Some getvery hungry and may attempt to get food fromlockers. Between 9 and 9.30 a.m. the stage ofsopor is reached, and is recognized by considerable'impairment of consciousness, loss of orientation,of speech and co-ordinated thinking. Mostpatients sleep quietly through this phase thoughthey still react to painful stimuli and corneal re-flexes are preserved. Some become very restless,excited and noisy, and need restraining tem-porarily. Between I0.15 and i i a.m. sopor passesover into the,-true coma stage. This is recognizedby the patient being completely unconscious, bythe lack of any purposive response to stimuli, suchas pricking or pinching, by the absence of cornealreflexes and by a positive Babinski plantarresponse. While it is essential that loss of con-sciousness and purposive response be consideredcriteria of chiia, the other signs are apt to vary inthe time of their appearance and one should notnecessarily wait until all are present untildiagnosing coma. The phase of coma has to beachieved for the best therapeutic results. On thefirst day that coma occurs, it is interrupted at once.On succeeding days the comas may then be in-creased by five minutes a day until 20 minutecomas are reached. Longer coma periods arepossible, but at the expense of additional com-plications, 4ihd lacking additional benefit to thepatient. If the dose has been insufficient to causecoma, patients should be interrupted after ij hoursof the sopor phase.Once coma has been established, patients tend
to become "increasingly sensitive to insulin, andcoma comes on earlier. Dosage should be cutafter the first coma so that patients do not go intocoma earlier than 10.15 a.m., and the decrease iscontinued until it is clear that the minimumnecessary has been attained. In practice it iscommonly found that the patient's usual comadose is as little as half the initial coma producingdose. Neglect to allow for the phenomenon ofinsulin sensitivity is one of the chief causes of thecomplication of irreversible coma.
Danger signs. During the coma period,twitching is often seen, usually' round the mouthat'first. In a few patients it increases to involve thewhole body and precedes a full convulsion. Ifthere is a fit, the patient may be partly wakened upby it. Fits may also occur before the coma stage,but are easily managed and subsequently con-trolled by the use of anticonvulsants. In anyevent, treatment should be interrupted. Waves ofextensor tonus occur in deep coma, with increasedtension of all muscles, extension and pronationof the arms, increased rate and depth of respiration,and dilation of the pupils. Though separated by
intervals of relative placidity, they tend to increasein frequency as coma deepens, and are mostexhausting. Circulatory failure is indicated byfall in blood pressure and by discoloration of thefinger tips together with extra systoles and pulseirregularities. Pulse rates below 55 make in-terruption advisable, as does persistent tachy-cardia above io or a blood pressure below ioo.Occasionally great restlessness and excitement inearly coma make treatment difficult and may be'avoided by the use of intravenous insulin or pre-medication with sodium amytal, gr. 6, at 9 a.m-Excessive salivation sometimes embarrassesrespiration. The head should be turned to oneside and atropine, gr. i/Ioo, given at 9 a.m. forsubsequent comas.Most patients have their own distinctive pattern
of sopor and coma. Although fits, extensor tonusand spasms, and circulatory failure are indications-for interruption in any patient, there are manylesser individual variations that are significant insome patients and not ini others, or are onlysignificant at certain stages. The insulin therapistwho knows the individual patterns for each casegets to know when a change from the usual coursemeans danger. It is necessary to deal with com-plications as early as possible otherwise poorgastric absorption or collapsed veins may render'interruption difficult.
Interruption of coma is accomplished by givingglucose either through a nasal tube or intra-venously. The nasal tube is inserted, shortlybefore coma should cease and its passage to thestomach confirmed by testing gastric juice againstlitmus paper. As much gastric juice as possible issucked out to avoid dilution of the sugar solutionor over distension and vomiting. At the in-terruption time 6oo cc. of 33 per cent. glucose,warmed and flavoured with lemon or tea, ispoured down. Slightly less may be used forfemales. The flavouring is for the benefit ofpatients who may need a glucose drink when stillin a conscious state. Many comas are terminatedregularly by this technique, but if after 15 to 20minutes the depth of hypoglycaemia is not sub-stantially diminished, then intravenous interrup-tion is given at once.
This method of termination is always used bysome psychiatrists as being more certain andtherefore safer than the nasal route. Others preferto employ it only in the event of failure with the'other method. From I00 to zoo cc. are given,and the patient usually recovers consciousness in afew minutes. It is important to note the timetaken f9r recovery since if recovery is delayedafter I0o cc. of glucose bas been given, thelikelihood is that too much insulin- has been
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
administered, that an irreversible coma is near, orthat coma has been allowed to go on toolong. Thesimplest way to give the glucose is by a simplepressure apparatus' after the type originally usedat the Maudsley Hospital' by Sargant and Fraser.A graduated bottle of 500 cc. capacity is fitted witha two way cap. One opening is connected byrubber tubing with a pressure bulb, the other isconnected by tubing to a syringe with a sideopening situated one-third of the way from theneedle end. The pressure is raised and the systemfilled with glucose and, with the plunger closed,the vein is entered as usual. The plunger iswithdrawn and as it passes back past the sideopening, the pressure forces the glucose over, andall that is. needed therafter is'an occasional squeezeof the bulb. The system is very adaptable sincethe pressure bottle can be unscrewed and re-placed with the same or a different solution asrequired. After intravenous interruption alone,awakening is followed by a light meal of sugaredtea and sandwiches. For the rest of the day it isessential that the patient takes a full lunch, teaand supper, but no special loading of the diet withcarbohydrate or anything else is necessary. Togive extra carbohydrate is merely to increase*resistance to insulin next day. Ideally the carbo-hydrate intake should be kept constant, but forvarious reasons it is generally not possible toachieve this in psychiatric patients.
Patients who have had insufficient glucose at theinterruption time, or who are difficult with theirfood, or in whom large injections of insulin havebeen needed, are liable at times to show signs oflate hypoglycaemia in the afternoon. If properobservation is maintained so that a drink of glucoseor sugared tea is given in good time, there isgenerally little difficulty in restoring the situation.None the less, a tray equipped for intravenousinterruption must always be ready for use inemergency on the insulin ward.The total number of comas required depends
on how soon improvement begins, and may varyfrom 30 to 6o. It is unwise to exceed this figurebecause of risk of damage to the brain.
It is essential that the time of onset of eachphase, sopor, coma, etc., should be noted asthese occur, and a time for interruption ordered as:oon as coma is diagnosed. Only by accuratetiming can the pace at which hypoglycaemia isproceeding be ascertained, and on this the safelength of coma decided daily individually.
Irreversible comna. This is one of the mostdangerous complications of the whole procedure,though one which is nearly always avoidable.' Itis considered to be present when the patient failsto recover consciousness after 250 cc. of 33' percent. glucose intravenously. If a nasal feed is
followed by facial pallor, vomiting and non-absorption by the stomach, there seems to be aspecial liability to the irreversible state. Somepatients have a tendency to irreversibility from thestart of comas and are slow in reacting to glucoseintravenously. Most irreversible comas occur inthe sensitization period early in the coma seriesthrough failure to reduce the initial coma pro-ducing dose quickly enough. In the early stage ofirreversibility, the patient is usually restless, overbreathing, flushed after initial pallor, and in ashocked state. Treatment is dire,9ted towardsminimizing shock, maintaining nutrition and re-gaining consciousness. The patient is thereforekept warm, the foot of the bed a little raised and2 CC. of coramine given intramuscularly. Theintravenous glucose, 33 per cent., is continueduntil 400 cc. has been given. If there is muchrestlessness, from o.i to 0.3 gn. of sodium amytalis injected through the tubing of the intravenousapparatus. Cyanosis and hyperpnoea are relievedby the intermittent administration of oxygen viaa B.L.B. mask. Most cases tend to settle downafter the first hour, but unconsciou3ness maycontinue as long as four to five days. In earlyyears cases were lost through what is now realizedto have been inadequate treatment, subsequentlyit has been evident that the same result may followover enthusiastic efforts. Many 'cses reactedwell to 75O cc. of 33 per cent. and i,50o cc. of5 per cent. glucose spread over thle 24 hours, givenin three or four injections. In some patients thecirculation could hardly support this amount andthere was a tendency to pulmonary oedema.Latterly the author has dispensed with the use of5 per cent. glucose, and while giving the sameamount of 33 per cent. glucose as before, hasfollowed the glucose injections by IOO Cc. of 50per cent. sucrose, three times a day. This wassuccessfully injected despite its high concentration,and after each injection patients certainly appearedto improve clinically and the comas to run a lessmalignant course. Apart from the foregoing thereare few remedies that appear to have any decisiveeffect on the coma with the possible exception ofblood transfusion. The vitamin B complex istheoretically of value in facilitating cerebralcarbohydrate metabolism, and 20 mgm. ofvitamin Bi and IOO mgm. of nicotinic acid shouldbe given four hourly. Eucortone and calciumchloride have also been tried, as have many othersubstances.
After the first two hours 4 fluid oz. of 33per cent. glucose are given two hourly by stomachor nasal tube. When it is clear that these feedsare being retained and gastric absorption resumed,then less intravenous feeding is required, though itshould not be discontinued until consciousness is
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
August 1948 SANDS: Convulsion and Insulin Treatment 439
regained. On recovery, the patient is left with avarying degree of confusion and disorientation, anda mood that is cheerful, over placid and somewhatfatuous. They have few ideas and are easilypleased. They show little of their former illness.The confusion clears over the ensuing two to threeweeks, depending on the length of the irreversibleperiod. Comas may be built up again after aweek's interval.
Psychiatric changes induced by the treatment. Inpromising cases there is generally temporary im-provement for an hour or two after interruption.The patient achieves some. appearance of nor-mality, their mood is more natural and they aremore in touch with those about them. In thisperiod they are most amenable to psycho-therapeutic enquiry and treatment, and betterrapport is possible. At this time too some arevery dependent, seek reassurance, and are at painsto identify familiar faces and surroundings. Inimproving cases this easing of the psychosisbecomes more permanent as further comas aregiven. On the whole few patients do well who havenot shown some gain before 25 comas.
Convulsion therapy is sometimes combined withinsulin, and if carefully used in selected cases therecovery rate of any series of schizophrenias isbetter than with insulin alone. Convulsiontherapy may be applied two or three times aweek in late sopor or early coma followed byimmediate interruption, or given on a non-insulin day. The former is the most effective asa rule. The combined treatment is indicatedwhere emotional disorder is evident, where thereare severe catatonic features, or when a case hasfailed to improve after 20 or more comas.
Results from large series of cases show a threeto one ratio of recovery and improvement infavour of insulin for those treated with two yearsonset. Statistics for New York State hospitalsshowed that this was maintained for two years,and 69 per cent. of treated patients were still at-home in a recovered or improved condition. Asafter spontaneous recovery, relapses occur andwith the passing of years treated and control casestend to approximate, though according to Hinkoand Lipschutz, after five years the remission rateis still i8 per cent. higher for insulin treatedpatients than in those allowed to remit spon-taneously. Relapses can be re-treated, and theyears spent out of hospital are a great gain for thepatient, not to mention economic benefits topatient and community.
For patients treated in the first,year a saving of422 hospital days per patient was obtained, andremission permitting parole occurred in one-thirdthe time required for spontaneous remission.,
Modified insuiin treatment. Insulin has on
other useful place in psychiatry although indica-tions and technique differ entirely from insulincomas. The principles underlying modified in-sulin therapy were embodied in the ideas of WeirMitchell, but the present modified insulin methodwas first described by Sargant and Craske on thetreatment of War Neurosis in 194I, and by theauthor in I944 in civilian neurotic states.Many neuroses and some psychotic depressions
deteriorate physically and lose weight on accountof prolonged mental tension arising out of theconflicts of their illness. The decrease in physicalwell being renders patients even less capable ofadjusting to their neurosis, and a vicious circle isset up. Restoration of weight in these neuroticcases enables many to renew their efforts atadjustment and to benefit more rapidly frompsychotherapy. The treatment also has a mildsedative effect. Itis therefore not specific for anyone psychiatric state but frequently proves to bathe turning point in a down hill course of illnessand facilitates progress with the main psycho-therapeutic procedure. It emphasizes the value oftreating the whole patient and -not being ex-clusively concerned with one aspect of their con-stitutions. That the good effect is not merely theresult of suggestion has been shown by the lackof response to control injections of sterile water.
Insulin is injected intramucsularly at 7 a.m.,starting with IO units and increasing by io unitsdaily until a state of mild hypoglycaemia is evidentbetween 9 and 9.30 a.m. The hypoglycaemiashould not be allowed to go further than the stageof flushing, sweating and slight drowsiness, notamounting to light coma. To attain this phasemay entail a dosage of from iO to IOO unitsaccording to sensitivity. At 9.30 patients areroused for breakfast consisting ofthe menu for theday together with 8 to 12 OZ. of mashed potatoes.Such is their appetite that this large meal is takenwell as a rule. Any patient threatening to slipinto sopor before 9.30 is aroused at once with aglucose drink and given breakfast. Thoughrarely needed, a stomach or nasal tube should beat hand for the administration of glucose inemergency. The increase in appetite often per-sists through the rest of the day. The patient isweighed before treatment and at weekly intervalsafterwards, each time in similar clothing. Treat-ment is continued six days a week until definitegain in weight ceases, usually after four to sevenweeks. How much weight is gained depends onthe quantity originally lost. Gain at the rate of apound per day is not uncommon. The treatmentwill not force patients above their normal weight,and those who are chronically thin are not fattened.
Indications. In general terms these are:-(i)To accelerate recovery in anxiety states, in hysteria
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
and particularly in neurasthenic conditions of ex-haustion or irritable weakness. At times thesecases also show suspicious paranoid, mildlyhostile trends, and it is difficult to make adequaterapport with the patient. It is remarkable how thesense of well being acquired under two weeksinsulin can swing this attitude towards one of co-operation and subsequent progress with purelypsychological methods. (2) To aid convulsiontherapy in the treatment of depression and incombination with prolonged sleep for acute anxietyneurosis. The treatment is only of value wherethe foregoing clinical states are associated withmetabolic change producing weight loss. Sevento ten pounds loss is the least change worthconsidering as significant.
BIBLIOGRAPHYBENNETT, A. E. (I940), J.A.M.A., 1I4, 322.CERLETTI, U., BINI, L. (1938), L'ELETTROSHOCK, Arch.
Gen. di neurol. psichiat., e Psicoanal, 19, 266.
COOK, L. C. and SANDS, D. E. (I941), Jour. Ment. Sc., Vol. 87,No. 367, 230.
COOK, L. C. (i944), 'Convulsion Therapy, Recent Progress inPsychiatry.' Ed. by G. W. T. H. Fleming. London.
HINKO, E. N. and LIPSCHUTZ, L. S. (I947), Amer. Jour.Psychiat., Vol. I04, No. 6, p,387.
KALINOWSKY, L. B. and HOCH, P. H. (1946), 'Shock Treat-ment in Psychiatry.'
KOLB, L. and VOGEL, V. H. (1942), Amer. J7our. Psychiat,, 9g,.90.
MASSERMAN, J. H. and JACQUES, M. G. (1947), Amer. Jour.Psychiat., Vol. 104, No. 2, 92.
MAYER GROSS, W. and WALK, A. (1938), Lancet, I, p. I324.MAYER GROSS, W. and WALK, A. (1939), Lancet, I, p. I324.MILLIGAN, W. L. (1946), Lancet, 2, 5I6, Oct. 12.NEUSTATTER, W. L., LOND, M. B and FREEMAN, H..
(I939), Lancet, 2, 1071.OWEN T. V. and STEMMERMANN, M. G., Amer. Your.
Psychiat., Vol. I04, No. 6, p. 410.PACELLA, B. L., PIOTROWSKI, Z. and LEWIS, N. D. C.
(I947), Amer. Jour. Psychiat., Vol. I04, No. 2, 83.SANDS, D. E. (I939), Lancet, II, 250.SANDS, D. E. (I944), Jour. Ment. Sci., Vol. 90, No. 380, p. 767.SARGANT, W. and SLATER, E. (1944), 'Physical Methods of
Treatment in Psychiatry,' Edinburgh.SARGANT, W. and CRASKE, N. (I94I), Lancet, 2, 2I2.TOW, P. M. (I947), Jour. Meat. Sci., Vol. 92, No. 392, 644.
BOOK REVIEWSINTRODUCTION TO MEDICAL PSYCHOLOGYBy L. ERWIN WEXBERG, M.D. William Heine-mann. 194g. Pp. I7I. Price 17s. 6d.The problem of teaching psychology to medical
students has come to the fore of recent years. Theinterim report of the Royal College of Physicians(1943) stressed the need for instruction in nornalpsychology in the pre-clinical years, indeed ' theobject of training at this stage should be to producea good general practitioner and not a specialist inpsychological medicine.' It was recommended thatthis course should be followed by an introductorycourse on the psychiatric aspects of clinical workat the beginning of the student's first clinical year.Ebaugh and Rymer in their comprehensive
treatise on Psychiatry in Medical Education (i942)are fully alive to the difficulties in teaching: ' It istoo little recognized that educators often do notknow what to teach, how to teach, or even why it isnecessary to teach such a course.'The recent report by the B.M.A. on' The Train-
ing of a Doctor ' advocates the need for instructionin normal psychology for the pre-clinical student,but deplores the absence of suitable textbooks inpsychology for medical students. Wexberg's littlebook ' Introduction to Medical Psychology' cer-tainly aims at meeting such a demand. The bookis the result of the author's trial and error in methodsof teaching psychology to medical students over aperiod of eight years.
While admittedly the book is 'an introduction,'.it is a matter for regret that there is not more of it.The subject matter is conveniently divided into sixchapters, at the end of which there is a useful list ofreferences, to enable the student to undertake widerreading. The book is thoroughly readable and fillsa useful gap not only for the medical student, but forgeneral practitioners, many of whom never had theopportunities provided today in the medicalcurriculum.
W.D.N.
ELEMENTS OF SURGICAL DIAGNOSIS
By. SIR ALFRED PEARCE GOULD, revised by SIRCECIL P. G. WAKELEY, K.B.E., C.B., F.R.C.S.gth Edition. Cassell. I948. Pp. xv + 7I8Price I5s.This excellent little book appears once again
having now reached its ninth edition. Practicallyeverything which the surgeon need know hindiagnosis is to be found within the covers of thispocket volume and though there are few illustra-tions, the text is so clear that even these could beomitted.' This manual of signs and symptoms hasalways remained quite distinctive, its appeal to thestudent being its completeness and accuracy. It is,in fact, the ideal reference book for ward and out-patient department and should be carried in thestudent's pocket at. all times. If it could be printedon India paper it would indeed be perfect.
copyright. on M
ay 24, 2021 by guest. Protected by
http://pmj.bm
j.com/
Postgrad M
ed J: first published as 10.1136/pgmj.24.274.427 on 1 A
