BRITISH MEDICAL JOURNAL
LONDON, SATURDAY 7 OCTOBER 1978
Thyroid disease and pregnancy The complex changes in. thyroid function during pregnancy are still incompletely understood. In the mother, goitre is common, its frequency and severity depending in part on the background iodide state of the community.1 2 Pregnancy causes relative iodine deficiency in the mother from losses to the fetus and increased renal iodide clearance3; the uptake and turnover of iodine by the thyroid are increased.4 5 Furthermore, the placenta produces at least one polypeptide, human chorionic gonadotrophin (HCG), with some thyrotrophic activity owing to its common beta subunit with thyrotrophin (TSH).6 This thyrotrophic action is low, but early in pregnancy probably sufficient HCG is produced to stimulate the
maternal thyroid; but TSH concentrations fall complementarily in early pregnancy8 and thyrotoxicosis does not result. The rise in maternal serum concentrations of thyroxine (T4) and to a lesser extent triiodothyronine (T3) is principally due to oestrogens stimulating hepatic production of thyroxinebinding globulin (TBG)'0; free thyroid hormone concentrations are normal.1' The increased basal metabolic rate and hyperdynamic state of normal pregnancy result from the additional circulation to and metabolism by the fetoplacental unit and not from thyroid overactivity. In patients with hydatidiform mole or chorionic carcinoma, however, who have much higher circulating amounts of HCG, thyrotoxicosis is
frequent.9 The fetal thyroid gland, meanwhile, is developing in-
dependently.'213 Substantial secretion of T4 probably does not begin until the fetal pituitary starts to secrete TSH during the early part of the second trimester. Later, negative feedback control becomes established. Amounts of T3 are very low in the fetus, while concentrations of reverse T3, which is biologically inactive, are high.'4-'7 The fetal circulation also contains little TBG. There is little transplacental passage of thyroid hormones."13 The importance of intact fetal (as opposed to neonatal) thyroid function for ultimate development in man is uncertain, though development of the central nervous system and bone maturation are both affected in utero by thyroid function.'2 13 Immediately after birth there is an abrupt and transient rise in TSH secretion, in part owing to exposure to cold,'5 and as a result a surge in circulating concentrations of T4 and T3. In contrast, the concentration of reverse T3 falls while that of T3 remains three to four times that in utero.'4 1517 Normal thyroid function in the neonate is of great importance for its subsequent development, including C BRITISH MEDICAL JOURNAL 1978. All reproduction rights reserved.
maturation of the brain, as evidenced by the permanent retardation which may result in the hypothyroid infant when replacement is delayed beyond a few weeks. Maternal thyroid disease may influence fertility and pregnancy."'18 Autoimmune hypothyroidism often leads to menstrual disturbances and reversible infertility.'9 When pregnancy does occur in hypothyroid women the abortion rate is high,20 though successful pregnancies in untreated hypothyroidism have been recorded, in which case the fetus remains euthyroid.'8 19 21 Autoimmune thyrotoxicosis (Graves's disease) may also impair fertility. It may also arise during pregnancy, or become more obvious because of the superadded hypermetabolic state, or remit spontaneously.22 In untreated maternal hyperthyroidism the fetus usually remains euthyroid, but premature labour is common.22 Though the normal thyroid stimulator, TSH, does not cross the placenta, the abnormal thyroid stimulators responsible for thyrotoxic Graves's disease are antibodies to the TSH receptor- and may cross the placenta and cause congenital thyrotoxicosis. Pregnancies in which the fetus is particularly at risk of this complication are those where maternal concentrations of thyroid-stimulating antibodies are high: such patients may have pronounced exophthalmos and pretibial myxoedema."1 23 24 Management of such rare cases follows the lines' of the treatment of the thyrotoxic adult, with treatment continued for several months after birth until the amounts of thyroid-stimulating antibodies in the infant have fallen. Once on replacement therapy, hypothyroid women do not normally present much of a therapeutic problem in pregnancy unless they have high concentrations of thyroid-stimulating antibodies, when the fetus is at risk of neonatal thyrotoxicosis. Maternal thyrotoxicosis can be treated safely with thiourea drugs such as carbimazole or propylthiouracil,22 in the minimum doses necessary to make the patient euthyroid. These drugs readily cross the placenta and act on the fetal thyroid; fetal TSH concentrations rise, and goitre may result.22 If necessary beta-blockers may be given to the mother in small doses to control the sympathetic overactivity. Transient neonatal thyrotoxicosis may occur once exposure to the antithyroid drugs ceases at birth, but probably only where there has also been exposure to high concentrations of thyroidstimulating antibodies.22 25 Potassium perchlorate is not used during pregnancy, and iodides are inadvisable: they may lead to goitre with tracheal obstruction in the fetus. Maternal partial thyroidectomy can be carried out safely during NO 6143 PAGE 977
978
pregnancy. The patient should first be made euthyroid by drugs and the vascularity of the gland then reduced by giving a short course of potassium iodide in the minimum effective dose (about 5 mg three times daily).26 After surgery, replacement thyroxine should be given to eliminate the risk of maternal hypothyroidism for the rest of the pregnancy. Radioactive iodine is absolutely contraindicated. Thiourea drugs are secreted in milk, so women taking them should not breast-feed. Finally, recent reports from Japan by Amino and colleagues27 28 point to an influence of pregnancy on the course of autoimmune thyroid disease and suggest that women predisposed to autoimmune thyroid disease may be particularly at risk after delivery. They described 25 episodes of primary hypothyroidism in 23 women in the postpartum period; in all but three it was transient and recovery was spontaneous 5-10 months later.26 Thyroid enlargement and the presence of antithyroid microsomal (not antithyroglobulin) antibodies were characteristic features, in conjunction with thyroidfunctiontest results typical of primary hypothyroidism. They also reported a recurrence of transient hyperthyroidism in four women with Graves's disease in the postpartum period,28 with evidence of increasing autoimmune disturbance. In one case the episode was followed by hypothyroidism, also transient. The immune state in the mother is altered during pregnancy,29 probably partly because of hormone changes, and other autoimmune diseases are known to remit during pregnancy and relapse post partum30 31. These findings indicate that women with a history of autoimmune thyroid disease should be watched closely post partum, but that even in those who become hypothyroid treatment may be required for only a short period. 1 Crooks, J, et al, Lancet, 1964, 2, 334. 2 Crooks, J, et al, Lancet, 1967, 2, 625. 3Aboul-Khair, S A, et al, Clinical Science, 1964, 27, 195. 4 Pochin, E E, Clinical Science, 1952, 11, 441. 5Halnan, K E, Clinical Science, 1958, 17, 281. 6 Kenimer, J G, Hersham, J M, and Higgins, M P, JIournal of Clinical Endocrinology and Metabolism, 1975, 40, 491. 7Hershman, J M, and Burrow, G N, Journal of Clinical Endocrinology and Metabolism, 1976, 42, 970. 8 Braunstein, G D, and Hershman, J M, Journal of Clinical Endocrinology and Metabolism, 1976, 42, 1123. 9 Higgins, H P, and Hershman, J M, Clinics in Endocrinology and Metabolism, 1978, 7, 167. 10 Dowling, J T, Freinkel, N, and Ingbar, S H,3Journal of Clinical Investigation, 1960, 39, 1119. 11 Tunbridge, W M G, and Hall, R, Clinics in Obstetrics and Gynaecology, 1975, 2, 381. 12 Fisher, D A, and Dussault, J H, Handbook of Physiology Sect 7, Endocrinology, vol 3, p 21. Washington, American Physiological Society, 1974. 13 Fisher, D A, et al, Recent Progress in Hormone Research, 1977, 33, 59. 14 Abuid, J L, et al, Journal of Clinical Endocrinology and Metabolism, 1974, 39, 263. 15 Fisher, D A, and Odell, W D, Journal of Clinical Investigation, 1969, 48, 1670. 16 Chopra, I J,3Journal of Clinical Investigation, 1974, 54, 583. 17 Chopra, I J, Sack, J, and Fisher, D A, Journal of Clinical Investigation, 1975, 55, 1137. 18 Montgomery, D A D, and Harley, J M G, Clinics in Obstetrics and Gynaecology, 1977, 4, 339. 19 Echt, C R, and Doss, J F, Obstetrics and Gynecology, 1963, 22, 615. 20 Man, E B, et al, Journal of Clinical Investigation, 1951, 30, 137. 21 Osorio, C, and Myant, N B, British Medical Bulletin, 1960, 16, 159. 22 Burrow, G N, Clinics in Endocrinology and Metabolism, 1978, 7, 115. 23 Rosenberg, D, Grand, M J H, and Silbert, D, New England Journal of Medicine, 1963, 268, 292. 24 Nutt, J, et al, British Medical3Journal, 1974, 4, 695. 25 Mutjaba, Q, and Burrow, G N, Clinics in Obstetrics and Gynaecology, 1975, 46,282. 26 Hall, R, et al, Fundamentals of Clinical Endocrinology, p 122. Tunbridge Wells, Pitman Medical, 1974. 27 Takai, S, et al, Annals of Internal Medicine, 1977, 87, 155. 28 Amino, N, et al, Journal of Clinical Endocrinology and Metabolism, 1977, 44, 130. 29 Beer, A E, and Billingham, R E, Advances in Immunology, 1971, 14, 1. 30 Oka, M, Annals of the Rheumatic Diseases, 1953, 12, 227. 31 Garsenstein, M, Pollak, V E, and Kark, R M, New England Journal of Medicine, 1962, 267, 165.
BRITISH MEDICAL JOURNAL
7 OCTOBER 1978
Hypnosis in the NHS Hypnosis has had a chequered history since the days of Mesmerl and his theory of animal magnetism. Braid2 maintained that the phenomenon was a form of sleep. Janet3 and Bernheim4 advanced the ideas of dissociation and suggestion. Charcot5 discovered the psychogenic nature of hysteria, and subsequently the ideas of Breuer6 and of Freud6-psychodynamic exploration, catharsis, and the discovery of the transference-were derived from investigations into the hypnotic state. Nevertheless, when Freud stopped using hypnosis it was left to exploitation by charlatans and stage performers. Indeed, the lack of medical interest led to the Hypnotism Act of 1952,7 which placed the onus of permitting performances of hypnotism on local licensing authorities. Hypnosis came back into clinical use with the development of the behavioural school of psychotherapy,8 which added desensitisation to the Freudian methods. Well-defined techniques are now laid down for treatment.9 Hypnosis has been shown of value as an additional psychotherapeutic weapon in treating the neuroses; in relieving anxiety, phobic, obsessional, and hysterical symptoms; in psychosomatic illness; and in a large range of personality problems-as well as in obstetrics, dermatology, and dentistry, and for relieving physical pain. It is now 25 years since the Psychological Medicine Group Committee of the BMA appointed a subcommittee to consider the medical use of hypnosis.'0 The report confirmed that "hypnotism may be the treatment of choice in some cases of psychosomatic disorder and psychoneurosis. It may be of value for revealing unrecognised motives and conflicts in such conditions . . . it has proved its ability to remove symptoms and to alter morbid habits of thought and behaviour." The subcommittee recommended that a description of hypnotism and its psychotherapeutic possibilities, limitations, and dangers should be given to medical undergraduates and instruction in the clinical use be given to all postgraduate psychiatric trainees and possibly trainee anaesthetists and obstetricians. Yet a recent survey" of training in medical hypnosis in 33 medical schools and 18 dental schools in Britain showed that only two of each group provided official undergraduate training; and three of the medical schools but no dental schools provided official postgraduate training. Clearly little action has been taken to carry out the subcommittee's recommendation: the teaching of hypnosis within the psychiatric services of the NHS is very limited. Nevertheless, the British Society of Medical and Dental Hypnosis is open to membership to doctors and dentists both in the NHS and in the private sector, and a section of medical and dental hypnosis has now been formed within the Royal Society of Medicine. More doctors are now treating neurotic problems using hypnosis, though in hospital practice its use remains limited to psychiatrists who have the skill and to a few specially trained psychologists working in psychiatric units under the clinical responsibility of a consultant psychiatrist. This revival of medical interest in hypnosis is likely to be paralleled by an increase in treatment by lay therapists. The indiscriminate use of hypnosis can be harmful, and the 1952 Bill7 will need amendment if the public is to be protected from potentially dangerous non-medical hypnosis. 1 Mesmer, F A, Mesmerism, translated by V R Myers. London, Macdonald, 1958. 2 Braid, J, Neurypnology; or the Rationale of Nervous Sleep Considered in Relation With Animal Magnetism. London, Redway, 1899. 3 Janet, P, Psychological Healing, vols I and II. London, George Allen and Unwin, 1925.
LONDON, SATURDAY 7 OCTOBER 1978
Thyroid disease and pregnancy The complex changes in. thyroid function during pregnancy are still incompletely understood. In the mother, goitre is common, its frequency and severity depending in part on the background iodide state of the community.1 2 Pregnancy causes relative iodine deficiency in the mother from losses to the fetus and increased renal iodide clearance3; the uptake and turnover of iodine by the thyroid are increased.4 5 Furthermore, the placenta produces at least one polypeptide, human chorionic gonadotrophin (HCG), with some thyrotrophic activity owing to its common beta subunit with thyrotrophin (TSH).6 This thyrotrophic action is low, but early in pregnancy probably sufficient HCG is produced to stimulate the
maternal thyroid; but TSH concentrations fall complementarily in early pregnancy8 and thyrotoxicosis does not result. The rise in maternal serum concentrations of thyroxine (T4) and to a lesser extent triiodothyronine (T3) is principally due to oestrogens stimulating hepatic production of thyroxinebinding globulin (TBG)'0; free thyroid hormone concentrations are normal.1' The increased basal metabolic rate and hyperdynamic state of normal pregnancy result from the additional circulation to and metabolism by the fetoplacental unit and not from thyroid overactivity. In patients with hydatidiform mole or chorionic carcinoma, however, who have much higher circulating amounts of HCG, thyrotoxicosis is
frequent.9 The fetal thyroid gland, meanwhile, is developing in-
dependently.'213 Substantial secretion of T4 probably does not begin until the fetal pituitary starts to secrete TSH during the early part of the second trimester. Later, negative feedback control becomes established. Amounts of T3 are very low in the fetus, while concentrations of reverse T3, which is biologically inactive, are high.'4-'7 The fetal circulation also contains little TBG. There is little transplacental passage of thyroid hormones."13 The importance of intact fetal (as opposed to neonatal) thyroid function for ultimate development in man is uncertain, though development of the central nervous system and bone maturation are both affected in utero by thyroid function.'2 13 Immediately after birth there is an abrupt and transient rise in TSH secretion, in part owing to exposure to cold,'5 and as a result a surge in circulating concentrations of T4 and T3. In contrast, the concentration of reverse T3 falls while that of T3 remains three to four times that in utero.'4 1517 Normal thyroid function in the neonate is of great importance for its subsequent development, including C BRITISH MEDICAL JOURNAL 1978. All reproduction rights reserved.
maturation of the brain, as evidenced by the permanent retardation which may result in the hypothyroid infant when replacement is delayed beyond a few weeks. Maternal thyroid disease may influence fertility and pregnancy."'18 Autoimmune hypothyroidism often leads to menstrual disturbances and reversible infertility.'9 When pregnancy does occur in hypothyroid women the abortion rate is high,20 though successful pregnancies in untreated hypothyroidism have been recorded, in which case the fetus remains euthyroid.'8 19 21 Autoimmune thyrotoxicosis (Graves's disease) may also impair fertility. It may also arise during pregnancy, or become more obvious because of the superadded hypermetabolic state, or remit spontaneously.22 In untreated maternal hyperthyroidism the fetus usually remains euthyroid, but premature labour is common.22 Though the normal thyroid stimulator, TSH, does not cross the placenta, the abnormal thyroid stimulators responsible for thyrotoxic Graves's disease are antibodies to the TSH receptor- and may cross the placenta and cause congenital thyrotoxicosis. Pregnancies in which the fetus is particularly at risk of this complication are those where maternal concentrations of thyroid-stimulating antibodies are high: such patients may have pronounced exophthalmos and pretibial myxoedema."1 23 24 Management of such rare cases follows the lines' of the treatment of the thyrotoxic adult, with treatment continued for several months after birth until the amounts of thyroid-stimulating antibodies in the infant have fallen. Once on replacement therapy, hypothyroid women do not normally present much of a therapeutic problem in pregnancy unless they have high concentrations of thyroid-stimulating antibodies, when the fetus is at risk of neonatal thyrotoxicosis. Maternal thyrotoxicosis can be treated safely with thiourea drugs such as carbimazole or propylthiouracil,22 in the minimum doses necessary to make the patient euthyroid. These drugs readily cross the placenta and act on the fetal thyroid; fetal TSH concentrations rise, and goitre may result.22 If necessary beta-blockers may be given to the mother in small doses to control the sympathetic overactivity. Transient neonatal thyrotoxicosis may occur once exposure to the antithyroid drugs ceases at birth, but probably only where there has also been exposure to high concentrations of thyroidstimulating antibodies.22 25 Potassium perchlorate is not used during pregnancy, and iodides are inadvisable: they may lead to goitre with tracheal obstruction in the fetus. Maternal partial thyroidectomy can be carried out safely during NO 6143 PAGE 977
978
pregnancy. The patient should first be made euthyroid by drugs and the vascularity of the gland then reduced by giving a short course of potassium iodide in the minimum effective dose (about 5 mg three times daily).26 After surgery, replacement thyroxine should be given to eliminate the risk of maternal hypothyroidism for the rest of the pregnancy. Radioactive iodine is absolutely contraindicated. Thiourea drugs are secreted in milk, so women taking them should not breast-feed. Finally, recent reports from Japan by Amino and colleagues27 28 point to an influence of pregnancy on the course of autoimmune thyroid disease and suggest that women predisposed to autoimmune thyroid disease may be particularly at risk after delivery. They described 25 episodes of primary hypothyroidism in 23 women in the postpartum period; in all but three it was transient and recovery was spontaneous 5-10 months later.26 Thyroid enlargement and the presence of antithyroid microsomal (not antithyroglobulin) antibodies were characteristic features, in conjunction with thyroidfunctiontest results typical of primary hypothyroidism. They also reported a recurrence of transient hyperthyroidism in four women with Graves's disease in the postpartum period,28 with evidence of increasing autoimmune disturbance. In one case the episode was followed by hypothyroidism, also transient. The immune state in the mother is altered during pregnancy,29 probably partly because of hormone changes, and other autoimmune diseases are known to remit during pregnancy and relapse post partum30 31. These findings indicate that women with a history of autoimmune thyroid disease should be watched closely post partum, but that even in those who become hypothyroid treatment may be required for only a short period. 1 Crooks, J, et al, Lancet, 1964, 2, 334. 2 Crooks, J, et al, Lancet, 1967, 2, 625. 3Aboul-Khair, S A, et al, Clinical Science, 1964, 27, 195. 4 Pochin, E E, Clinical Science, 1952, 11, 441. 5Halnan, K E, Clinical Science, 1958, 17, 281. 6 Kenimer, J G, Hersham, J M, and Higgins, M P, JIournal of Clinical Endocrinology and Metabolism, 1975, 40, 491. 7Hershman, J M, and Burrow, G N, Journal of Clinical Endocrinology and Metabolism, 1976, 42, 970. 8 Braunstein, G D, and Hershman, J M, Journal of Clinical Endocrinology and Metabolism, 1976, 42, 1123. 9 Higgins, H P, and Hershman, J M, Clinics in Endocrinology and Metabolism, 1978, 7, 167. 10 Dowling, J T, Freinkel, N, and Ingbar, S H,3Journal of Clinical Investigation, 1960, 39, 1119. 11 Tunbridge, W M G, and Hall, R, Clinics in Obstetrics and Gynaecology, 1975, 2, 381. 12 Fisher, D A, and Dussault, J H, Handbook of Physiology Sect 7, Endocrinology, vol 3, p 21. Washington, American Physiological Society, 1974. 13 Fisher, D A, et al, Recent Progress in Hormone Research, 1977, 33, 59. 14 Abuid, J L, et al, Journal of Clinical Endocrinology and Metabolism, 1974, 39, 263. 15 Fisher, D A, and Odell, W D, Journal of Clinical Investigation, 1969, 48, 1670. 16 Chopra, I J,3Journal of Clinical Investigation, 1974, 54, 583. 17 Chopra, I J, Sack, J, and Fisher, D A, Journal of Clinical Investigation, 1975, 55, 1137. 18 Montgomery, D A D, and Harley, J M G, Clinics in Obstetrics and Gynaecology, 1977, 4, 339. 19 Echt, C R, and Doss, J F, Obstetrics and Gynecology, 1963, 22, 615. 20 Man, E B, et al, Journal of Clinical Investigation, 1951, 30, 137. 21 Osorio, C, and Myant, N B, British Medical Bulletin, 1960, 16, 159. 22 Burrow, G N, Clinics in Endocrinology and Metabolism, 1978, 7, 115. 23 Rosenberg, D, Grand, M J H, and Silbert, D, New England Journal of Medicine, 1963, 268, 292. 24 Nutt, J, et al, British Medical3Journal, 1974, 4, 695. 25 Mutjaba, Q, and Burrow, G N, Clinics in Obstetrics and Gynaecology, 1975, 46,282. 26 Hall, R, et al, Fundamentals of Clinical Endocrinology, p 122. Tunbridge Wells, Pitman Medical, 1974. 27 Takai, S, et al, Annals of Internal Medicine, 1977, 87, 155. 28 Amino, N, et al, Journal of Clinical Endocrinology and Metabolism, 1977, 44, 130. 29 Beer, A E, and Billingham, R E, Advances in Immunology, 1971, 14, 1. 30 Oka, M, Annals of the Rheumatic Diseases, 1953, 12, 227. 31 Garsenstein, M, Pollak, V E, and Kark, R M, New England Journal of Medicine, 1962, 267, 165.
BRITISH MEDICAL JOURNAL
7 OCTOBER 1978
Hypnosis in the NHS Hypnosis has had a chequered history since the days of Mesmerl and his theory of animal magnetism. Braid2 maintained that the phenomenon was a form of sleep. Janet3 and Bernheim4 advanced the ideas of dissociation and suggestion. Charcot5 discovered the psychogenic nature of hysteria, and subsequently the ideas of Breuer6 and of Freud6-psychodynamic exploration, catharsis, and the discovery of the transference-were derived from investigations into the hypnotic state. Nevertheless, when Freud stopped using hypnosis it was left to exploitation by charlatans and stage performers. Indeed, the lack of medical interest led to the Hypnotism Act of 1952,7 which placed the onus of permitting performances of hypnotism on local licensing authorities. Hypnosis came back into clinical use with the development of the behavioural school of psychotherapy,8 which added desensitisation to the Freudian methods. Well-defined techniques are now laid down for treatment.9 Hypnosis has been shown of value as an additional psychotherapeutic weapon in treating the neuroses; in relieving anxiety, phobic, obsessional, and hysterical symptoms; in psychosomatic illness; and in a large range of personality problems-as well as in obstetrics, dermatology, and dentistry, and for relieving physical pain. It is now 25 years since the Psychological Medicine Group Committee of the BMA appointed a subcommittee to consider the medical use of hypnosis.'0 The report confirmed that "hypnotism may be the treatment of choice in some cases of psychosomatic disorder and psychoneurosis. It may be of value for revealing unrecognised motives and conflicts in such conditions . . . it has proved its ability to remove symptoms and to alter morbid habits of thought and behaviour." The subcommittee recommended that a description of hypnotism and its psychotherapeutic possibilities, limitations, and dangers should be given to medical undergraduates and instruction in the clinical use be given to all postgraduate psychiatric trainees and possibly trainee anaesthetists and obstetricians. Yet a recent survey" of training in medical hypnosis in 33 medical schools and 18 dental schools in Britain showed that only two of each group provided official undergraduate training; and three of the medical schools but no dental schools provided official postgraduate training. Clearly little action has been taken to carry out the subcommittee's recommendation: the teaching of hypnosis within the psychiatric services of the NHS is very limited. Nevertheless, the British Society of Medical and Dental Hypnosis is open to membership to doctors and dentists both in the NHS and in the private sector, and a section of medical and dental hypnosis has now been formed within the Royal Society of Medicine. More doctors are now treating neurotic problems using hypnosis, though in hospital practice its use remains limited to psychiatrists who have the skill and to a few specially trained psychologists working in psychiatric units under the clinical responsibility of a consultant psychiatrist. This revival of medical interest in hypnosis is likely to be paralleled by an increase in treatment by lay therapists. The indiscriminate use of hypnosis can be harmful, and the 1952 Bill7 will need amendment if the public is to be protected from potentially dangerous non-medical hypnosis. 1 Mesmer, F A, Mesmerism, translated by V R Myers. London, Macdonald, 1958. 2 Braid, J, Neurypnology; or the Rationale of Nervous Sleep Considered in Relation With Animal Magnetism. London, Redway, 1899. 3 Janet, P, Psychological Healing, vols I and II. London, George Allen and Unwin, 1925.
