Drn
CLINICAL PRACTICE
Hospital Admissions of Adolescent Patients with Diabetes A.H. Challen“, A.G. Davies”, R.J.W. Williamsb, J.D. Baum’ ”Department of Child Health and bDepartment of Child and Family Psychiatry, lnstitute of Child Health, Royal Hospital for Sick Children, Bristol, UK
This study formed part of a psychological survey of young people aged 10-17 years attending three Bristol diabetes clinics. The aim was to examine the characteristics of those who were admitted to hospital with unstable diabetes during the period between two assessments, and analyse how they differed from the rest. Ten young people were admitted to hospital during the study period. These 10 individuals had greater emotional difficulty with diabetes as measured by diabetes specific psychometric scales (median scale score: 19) compared to the 89 adolescents who were not readmitted (median scale score: 14) ( p = 0.01). They did not differ in demographic characteristics or glycaemic control. Overall there was no relationship between psychological response to diabetes and glycaemic control as assessed by mean glycated haemoglobin. Five of those admitted presented with hyperglycaemia; they had greater emotional difficulty (median scale score: 31, p = 0.02) and a more negative attitude (median scale score: 22, p = 0.02) to diabetes than those presenting with hypoglycaemia (median scale scores: 16). Only those with hyperglycaemia differed from those who were not readmitted,having greater emotional difficulty ( p = 0.002) and a more negative attitude ( p = 0.01). The possibility of psychologicaldifficulties with diabetes should be sought following an admission, particularly for hyperglycaemia. KEY WORDS
Type 1 diabetes mellitus Adolescence Glycaemic control Hospital admission
Introduction
Methods
An association between emotional factors and admission to hospital with unstable diabetes in adult populations has been suggested by both anecdotal reports’-3 and retrospective ~ t u d i e s . Within ~,~ the paediatric age-group a number of researchers have noted a preponderance of psychosocial problems, particularly family stress, in association with recurrent ketoacid~sis.~-~ The aim of this prospective study was to determine whether young people aged 10-1 7 years who were admitted to hospital with unstable diabetes, differed from those who were not in their psychological response to diabetes. This study formed part of the previously reported research to develop a diabetes specific psychometric scale for adolescents.l o Two scores derived from principal components analysis have been used in this study: the first contained statements describing misery and anger about having diabetes and has been labelled ’emotional difficulty with diabetes’; the second contained statements describing a positive or negative approach towards having diabetes and its management, and has been labelled ’attitude to diabetes’. No attempt has been made to define a specific psychiatric disorder on the basis of score results, but only to express a variation in the psychological response of young people to their diabetes.
All young people (apart from two with another serious chronic disorder) aged between 10 and 17 years attending the diabetes clinics of three consultant paediatricians and living within Avon were asked to participate (n = 103). Approval for the study was obtained from the three district ethical committees. The three consultants varied in their management of diabetes to some extent: consultant A prescribed one insulin injection a day for some of his patients and consultant B was more likely to start his patients on more than two injections a day with an insulin pen, but all three consultants stressed the importance of home monitoring with blood glucose tests. Availability of Diabetes Liaison Health Visitor service (61 %) was dependent on living within certain health district boundaries rather than attending a particular clinic. The psychological survey involved two assessments 6-12 (median = 8.6) months apart; the results from the first assessment are presented in this report. Of the 103 adolescents who were asked to participate, 99 (46 boys) completed this assessment. Their median age was 14.9 (range: 10.6-17.5) years and their median duration of diabetes 4.8 (range: 0.3-14.7) years. The assessments were performed by one researcher who was not directly involved with their diabetes management. The majority were visited at home (n = 77) although 10 chose to be seen at hospital. During the
Correspondence to: Dr A. Challen, Royal Manchester Children’s Hospital, Hospital Road, Pendlebury, Manchester, M 2 7 1HA, UK.
850
0742-3071/92/090850-05$07.50
@ 1992 by John Wiley & Sons, Ltd.
Accepted 3 June 1992 DIABETIC MEDICINE, 1992; 9: 850-854
DT17 research visits they were asked to complete psychometric scales assessing emotional difficulties with, and attitude to, diabetes. The items of the scale were read aloud to the subjects, who responded independently on their own score sheets. Twelve young people did not wish to participate in a research visit; they completed the questionnaires at home and returned them by post. These 12 adolescents did not differ by age, age at diagnosis, duration of diabetes, sex ratio, parental occupational status or level of glycaemic control from the rest of the study group. Following the second assessment the medical case notes were examined to identify any hospital admissions due to hypo- or hyperglycaemia during the intervening period. None of the three consultant paediatricians had specific criteria for admission of their diabetic patients and thus the decision to admit was taken by the duty junior and middle grade medical staff. Admissions for other reasons, such as surgical procedures, and visits to casualty without admission, were excluded from analyses. The young people’s glycaemic control was assessed by their mean glycated haemoglobin taken routinely at clinic visits during a 12-month period, 6 months before and after the first assessment. The glycated haemoglobin values for the study group were measured in three different laboratories: the results from one laboratory were adjusted by a correction factor following a comparison study as previously reported.” Social class (nonmanual/manual) for each family was determined using Office of Population Censuses and Surveys classification,’* based on father’s occupation or mother’s if there was no father in the family.
Adaptation to Diabetes Scales These are two 8-item scales measuring ‘attitude to diabetes’ and ‘emotional difficulty with diabetes’, respectively. The scales were designed for young people with diabetes between the ages of 10 to 17 years to assess aspects of psychosocial adaptation to diabetes (no reference to metabolic control was made during test development). The items are statements derived from interviews with adolescents with diabetes, and respondents are asked how frequently their experience matches these statements. High scores on the scales indicate greater emotional difficulty and a more negative attitude to diabetes. For the study population the attitude and emotional difficulty scales were found to have internal consistency coefficients of 0.78 and 0.85, and test-retest coefficients over 6-12 months of 0.73 and 0.66, respectively. Validation of the test has been described.’O
Results During the period between the two assessments, 10 young people were admitted to hospital. Their demographic, psychological, and laboratory characteristics are shown HOSPITAL ADMISSIONS OF ADOLESCENTS WITH DIABETES
CLINICAL PRACTICE in Table 1, and details of their admissions are described in Table 2. A comparison of their demographic characteristics, psychological test scores, and glycated haemoglobin values with those of the remaining study group is shown in Table 3. (Four young people did not complete the second assessment; they were not admitted to hospital during a comparable period (i.e. the median interval between assessments of 8.6 months)). Of those who were admitted to hospital, five adolescents presented with hyperglycaemia and five with hypoglycaemia. A comparison of demographic characteristics, psychological test scores, and glycated haemoglobin values between these two subgroups is shown in Table 4. When the two subgroups were compared with those not admitted separately, only the hyperglycaemia group differed, they had both greater emotional difficulty (p = 0.002) and a more negative attitude (p = 0.01). There were no significant correlations between mean glycated haemoglobin results and scores on the emotional difficulty and attitude scales ( r = -0.02, p = 0.81; r = -0.1 1, p = 0.29, respectively).
Discussion The results of this study demonstrate that adolescents who were admitted to hospital with unstable control were more likely to have experienced emotional difficulties with diabetes, than those who were not. The study results should not be interpreted as demonstrating an association between unstable diabetes and emotional difficulties as we have not included those adolescents who had episodes of unstable diabetes who were managed at home or in the accident and emergency departments without an admission. Further analysis of the admission group indicated that it was those presenting with hyperglycaemia (of whom four had ketoacidosis on admission) who had the greatest psychological difficulties with their diabetes, concurring with previous descriptive r e p ~ r t s . ~It- ~is difficult to be certain whether those readmitted with hypoglycaemia have more than average difficulty coping with their diabetes, given the small sample size. Certainly there have been reports of factitious hyp~glycaemia~,’and suicide attempts14 from insulin overdosage. Social and demographic characteristics did not differ between the two groups. Glasgow et a1.15 have recently demonstrated increased socioeconomic deprivation in children admitted to hospital with acute diabetic complications, in comparison to outpatients and new admissions over a 5-year period, and suggested serious psychosocial dysfunction as a possible underlying cause. Glycaemic control (as reflected in glycated haemoglobin values) did not differ between the readmission group and the rest. In addition glycaemic control was not related either to emotional difficulties or attitude to diabetes. This result concurs with a number of studies that have shown no correlation between glycaemic control and self-report psychometric tests assessing gen-
85 1
DTfl
CLINICAL PRACTICE Table 1. Demographic, psychological and laboratory details of the ten adolescents who were admitted to hospital during the study period Case Sex, age Socioeconomic Mean HbAlc Emotional number and duration group in the 6 difficulty and months before attitude score (years) and after the 1st assessment
1. 2. 3. 4.
5.3, 5.9, M, 2.8, F, 5.6,
5. 6. 7.
F, 3.8, F; 11.3; M, 10.6,
8. 9. 10.
F, 16.7, M, 15.5, M, 15.8,
F,
F,
4.7 manual 4.4 manual 7.5 non-manual 2.9 both parents unemployed 7.6 non-manual 2.3 non-manual 4.2 both parents unemployed 7.9 manual 13.4 manual 5.8 in foster care
0.7 7.7 2.8 3.7
I10
31, 17, 12, 19,
29 17 16 16
2.3 10.6 13.4
16, 18 18, 1 7 35, 19
results 10.7 17.1
23, 22 16, 1 1 37, 30
Table 2. Clinical details of the ten admissions Case number 1
2 3
8
9 10
852
Details of admissions
Admitted via accident and emergency department 3 days after starting on an insulin pen, with symptoms of polyuria and polydipsia and home results of glycosuria and 1 % ketones. Results on arrival: blood glucose: 16.4 mol I-', normal electrolytes, no H C 0 3 analysed, ketones negative. Admission while on holiday with nocturnal hypoglycaemic fit. Glucagon not given as parents panicked over the difficulty of assembling syringe. Taken semiconscious to local hospital, in-patient for several days. Two admissions for hypoglycaemia. First admission: semiconscious, home blood glucose < 1 mmol I-', admitted via GP, after IV glucose. Second admission: nocturnal hypoglycaemic convulsion, emergency IV glucose treatment from ambulance staff prior to admission via GP. Parents did not give glucagon as had not collected new supply. Self referral to accident and emergency department in hysterical state. Felt 'shaky' after taking evening insulin, drinking cider and having not eaten. Known to have poor parental support. Blood glucose 4 mmol I-' following exchanges on arrival. Overnight admission. Day admission after self referral to accident and emergency department following hypoglycaemic episode at school that was treated with glucagon injection. Occurred following administration of twice normal insulin dose. Blood glucose 6.5 mmol I-' on arrival at hospital. Admitted with symptoms of polyuria and vomiting, after discussion with consultant. Dehydrated but alert on examination. Results: blood glucose 27.2 mmol I-', H C 0 3 8 mmol I-' and ketonuria. Cause presumed due to change of treatment to insulin pen coupled with an intercurrent viral infection. Admitted via GP with symptoms of abdominal pain and vomiting. On examination: dehydrated with Kussmaul respirations, but alert. Results: blood glucose 55 mmol I-', H C 0 3 6.0 mmol I-' and ketonuria. Parents thought episode could be due to dietary indiscretions. Family vague about details of diabetes management. Two admissions with ketoacidosis via GP. Had recently moved in with aunt and uncle after having difficulties at home. First admission: admitted with nausea and vomiting. Dehydrated, kussmaul respirations but alert. Results: blood glucose 29.6 H C 0 3 4.7 mmol I-' and ketonuria. Episode secondary to missed insulin. Second admission: admitted with vomiting, diarrhoea, anorexia and chest pain. Omitted insulin because of anorexia. On examination unwell with kussmaul respirations. Results: blood glucose 41.6 mmol I-', H C 0 3 6.0 mmol I-' and ketonuria. Admitted with hypoglycaemic convulsion after giving double dose of insulin as he thought his blood glucose level was high (he refuses to do home monitoring). Treated with IV glucose with good recovery. Admission with ketoacidosis after running away from foster home and omitting insulin injections. Mild dehydration and drowsy on examination. Results: BM stick 44 mmol I-' on arrival: after subcutaneous insulin: blood glucose 15.5 mmol I-', H C 0 3 9 mmol I-', ketonuria.
A.H. CHALLEN ET AL.
Drn
CLINICAL PRACTICE Table 3. Differences in demographic characteristics, glycaemic control and psychological response to diabetes, between those who were and were not admitted to hospital Adolescents admitted Adolescents not to hospital admitted to hospital ~
Sex ratio (ma1e:female) Age (years) Duration of diabetes (years) Age at diagnosis (years) Socioeconomic group (non-manual manual) MGH (%) Emotional difficulty scores Attitude to diabetes scores
4:6
44:49
15.4 (10.6-16.7) 5.2 (2.3-13.4)
14.9 (10.6-17.0) 4.8 (0.3-14.7)
8.9 (2.0-12.7)
9.9 (1.3-15.5)
2:5
38:41
12.3 (7.65-17.1) 19 (12-37)
11.7 (5.8-1 8.0) 14 (8-36)"
18
16
(11-30)
(8-40)
=p = 0.01. Results as median (range) or as stated.
MGH: mean glycated haemoglobin in the 6 months before and after the first assessment.
Table 4. Differences in demographic characteristics, glycaemic control, and psychological response to diabetes, betwen those admitted with hyperglycaemia and hypoglycaemia
Sex ratio (ma1e:female) Age (years) Duration of diabetes (years) Age at diagnosis (years)
MGH (%) Emotional difficult scores Attitude to diabetes scores
Adolescents admitted with hyperglycaemia
Adolescents admitted with hypoglycaemia
2:3
2:3
15.3 (10.6-16.7) 4.6 (2.3-7.9)
15.5 (1 3.5-1 5.9) 7.5 (2.9-1 3.4)
8.9 (6.4-10.7)
6.2 (2.0-12.7)
12.0 (10.6-17.1) 31 (1%-37)
12.3 (7.65-1 3.7) 16 (12-19)"
22
16
(17-30)
(ll-18)a
"p = 0.02. Results as median (range) or as stated. MGH: mean glycated haemoglobin in the 6 months before and after the first assessment.
eralized psychological function.'8-20 Most of the studies that have shown a relationship between poor glycaemic control and psychological function are not comparable with this study. They have characterized a subgroup with marked social and psychological problems, using predominantly interview and school or parental report, who also have poor glycaemic ~ o n t r o l . ~ In ~ addition, ~'-~~ glycaemic control has been based on medical ratings, rather than glycated haemoglobin values. However, because admission to hospital with unstable diabetes has been used as a criterion in the assessment of poor HOSPITAL ADMISSIONS OF ADOLESCENTS WITH DIABETES
glycaemic control, these studies may well be reflecting the same conclusion as in this r e p ~ r t . ~ , ~ ~ , ~ ~ , ~ ~ We speculate that the relationship between emotional difficulties and admission with unstable diabetes could be due to a number of interrelating factors.
1. 2.
Neglect or deliberate manipulation of injection and diet (probably causal in cases 4, 5, 9, and 10). Young people and their families who cope better with diabetes are more likely to take effective action, such as insulin adjustment or administration of
853
DTT7
CLINICAL PRACTICE 3.
4.
glucagon, to prevent hospital admission (probably contributory in cases 2, 3, 4, 7, and 8). Hospital admission might exacerbate young people’s emotional problems by confronting them with the reality that they have a life-threatening disorder, and by increasing their frustration with diabetes and their sense of failure in managing their disease. Within the context of this study the decision to admit was based upon the subjective assessment of the duty doctors, who may have been more likely to admit from the accident and emergency department if they were aware of psychosocial problems.
It is possible therefore that measures to reduce emotional stress and improve confidence with diabetes management may be as important as manipulating insulin and diet in preventing further admissions, particularly for those admitted with hyperglycaemia. The diabetes health visitor plays an important role in identifying emotional difficulties among young people with diabetes, and close liaison with child and adolescent psychiatrists may faci Iitate diagnosis. Intervention to reduce the emotional distress experienced by these young people might include the opportunity for further exploration of their problems with a member of the diabetes team, sharing experiences at a young people’s support group, or input from a child psychiatrist. Orr8 and Goldeng have demonstrated a dramatic reduction in the number of admissions for unstable diabetes following a programme of education and individual, group or family counselling. Further research is required to support the premise that psychological diagnosis and intervention can reduce readmissions and perhaps prevent the unfortunate recurrence of admissions with unstable diabetes with its disruptive and distressing affects on the lives of young people and their families.
Acknowledgements We are grateful to Dr D. Savage and Dr T.L. Chambers for their support, Dr C. Pennock for his advice on glycated haemoglobin results, and to the young people and their families who took part in this research. A.H. Challen was supported by Novo Laboratories.
5. 6. 7.
8.
9. 10. 11. 12. 13. 14.
15.
16. 17.
18. 19. 20. 21.
References
22.
1.
Rosen H, Lidz T. Emotional factors in the precipitation of diabetic acidosis. Psychosom Med 1949; 11 : 21 1-21 5. 2. Gale E, Tattersall R. Brittle diabetes. Br ) Hosp Med 1979; 21: 589-597. 3. Gill GV, Husband DJ, Walford S, Marshall SM, Home PD, Albertie KGMM. Clinical features of brittle diabetes. In: Pickup JC, ed. Brittle Diabetes Oxford: Blackwell, 1985: 29-40. 4. Flexner CW, Weiner JP, Saudek CD, Dans PE. Repeated
854
23. 24. 25.
hospitalization for diabetic ketoacidosis. The game of ’Sartoris’. Am I Med 1984; 76: 691-695. Schade DS, Drumm DA, Duckwroth WC, Eaton RP. The etiology of incapacitating, brittle diabetes. Diabetes Care 1985, 8: 12-20. Loughlin WC, Mosenthal HO. Study of the personalities Dis Child 1944; 68: of children with diabetes. Am 13-1 5. White K, Kolman ML, Wexler P, Polin G, Winter RJ. Unstable diabetes and unstable families: A psychosocial evaluation of diabetic children with recurrent ketoacidosis. Pediatrics 1984; 72: 749-755. Orr DP, Golden MP, Myers G, Marrero DG. Characteristics of adolescents with poorly controlled diabetes referred to a tertiary care centre. Diabetes Care 1983; 6: 170-1 75. Golden MP, Herrold AJ, Orr DP. An approach to prevention of recurrent ketoacidosis in the pediatric population. ) Pediatr 1985; 107: 195-200. Challen AH, Davies AG, Williams RJW, Haslum MN, Baum ID. Measuring psychosocial adaptation to diabetes in adolescence. Diabetic Med 1988; 5: 739-746. Challen AH, Davies AG, Williams RJW, Baum JD. Support of families with diabetic children: Parents views. Practical Diabetes 1990; 7: 26-31. Office of Population Censuses and Surveys. Classification of Occupations 1980. London: HMSO, 1980. Orr DP, Eccles T, Lawlor R, Golden M. Surreptitious insulin administration in adolescents with insulin-dependent diabetes mellitus. IAMA 1986; 256: 3227-3230. Potter J, Clarke P, Gale EAM, Dave SH, Tattersall RB. Insulin-induced hypoglycaemia in an accident and emergency department: the tip of an iceberg? Br Med 1982; 285: 1180-1 182. Glasgow AM, Weissberg-Benchell J, Tynan WD, Epstein SF, Driscoll C, Turek J, et a/. Readmissions of children with diabetes mellitus to a children’s hospital. Pediatrics 1891; 88: 98-104. Delbridge L. Educational and psychological factors in the management of diabetes in childhood. Med ) Aus 1975; 2: 737-739. Simonds J, Goldstein D, Walker B, Rawlings S. The relationship between psychological factors and blood glucose regulation in insu I in-dependent diabetic adolescents. Diabetes Care 1981; 4: 610-615. Gross AM, Delcher HK, Snitzer J, Bianchi B, Epstein S. Personality variables and metabolic control in children with diabetes. ) Genet fsychol 1985; 146: 19-26. Close H, Davies AG, Price DA, Goodyer IM. Emotional difficulties in diabetes mellitus. Arch Dis Child 1986; 61 : 337-340. Evans CL, Hughes IA. The relationship between diabetic control and individual and family characteristics. 1 Psychos Res 1987; 3: 367-374. Sterky G. Family background and the state of mental health in a group of diabetic children. Acta Paediatr Scand 1963; 52: 377-390. Swift CR, Seidman F, Stein H. Adjustment problems in juvenile diabetics. Psychosom Med 1967; 29: 555-571. Koski ML. The coping processes in childhood diabetes. Acta Paediatr Scand 1969; (suppl) 198: 1-56. Ludvigsson J. Socio-psychological factors and metabolic control in juvenile diabetes. Acta Paediatr Scand 1977; 66: 431-437. Gath N, Smith MA, Baum ID. Emotional, behavioural and educational disorders in diabetic children. Arch Dis Child 1980; 55: 371-375.
A.H. CHALLEN E l AL.
CLINICAL PRACTICE
Hospital Admissions of Adolescent Patients with Diabetes A.H. Challen“, A.G. Davies”, R.J.W. Williamsb, J.D. Baum’ ”Department of Child Health and bDepartment of Child and Family Psychiatry, lnstitute of Child Health, Royal Hospital for Sick Children, Bristol, UK
This study formed part of a psychological survey of young people aged 10-17 years attending three Bristol diabetes clinics. The aim was to examine the characteristics of those who were admitted to hospital with unstable diabetes during the period between two assessments, and analyse how they differed from the rest. Ten young people were admitted to hospital during the study period. These 10 individuals had greater emotional difficulty with diabetes as measured by diabetes specific psychometric scales (median scale score: 19) compared to the 89 adolescents who were not readmitted (median scale score: 14) ( p = 0.01). They did not differ in demographic characteristics or glycaemic control. Overall there was no relationship between psychological response to diabetes and glycaemic control as assessed by mean glycated haemoglobin. Five of those admitted presented with hyperglycaemia; they had greater emotional difficulty (median scale score: 31, p = 0.02) and a more negative attitude (median scale score: 22, p = 0.02) to diabetes than those presenting with hypoglycaemia (median scale scores: 16). Only those with hyperglycaemia differed from those who were not readmitted,having greater emotional difficulty ( p = 0.002) and a more negative attitude ( p = 0.01). The possibility of psychologicaldifficulties with diabetes should be sought following an admission, particularly for hyperglycaemia. KEY WORDS
Type 1 diabetes mellitus Adolescence Glycaemic control Hospital admission
Introduction
Methods
An association between emotional factors and admission to hospital with unstable diabetes in adult populations has been suggested by both anecdotal reports’-3 and retrospective ~ t u d i e s . Within ~,~ the paediatric age-group a number of researchers have noted a preponderance of psychosocial problems, particularly family stress, in association with recurrent ketoacid~sis.~-~ The aim of this prospective study was to determine whether young people aged 10-1 7 years who were admitted to hospital with unstable diabetes, differed from those who were not in their psychological response to diabetes. This study formed part of the previously reported research to develop a diabetes specific psychometric scale for adolescents.l o Two scores derived from principal components analysis have been used in this study: the first contained statements describing misery and anger about having diabetes and has been labelled ’emotional difficulty with diabetes’; the second contained statements describing a positive or negative approach towards having diabetes and its management, and has been labelled ’attitude to diabetes’. No attempt has been made to define a specific psychiatric disorder on the basis of score results, but only to express a variation in the psychological response of young people to their diabetes.
All young people (apart from two with another serious chronic disorder) aged between 10 and 17 years attending the diabetes clinics of three consultant paediatricians and living within Avon were asked to participate (n = 103). Approval for the study was obtained from the three district ethical committees. The three consultants varied in their management of diabetes to some extent: consultant A prescribed one insulin injection a day for some of his patients and consultant B was more likely to start his patients on more than two injections a day with an insulin pen, but all three consultants stressed the importance of home monitoring with blood glucose tests. Availability of Diabetes Liaison Health Visitor service (61 %) was dependent on living within certain health district boundaries rather than attending a particular clinic. The psychological survey involved two assessments 6-12 (median = 8.6) months apart; the results from the first assessment are presented in this report. Of the 103 adolescents who were asked to participate, 99 (46 boys) completed this assessment. Their median age was 14.9 (range: 10.6-17.5) years and their median duration of diabetes 4.8 (range: 0.3-14.7) years. The assessments were performed by one researcher who was not directly involved with their diabetes management. The majority were visited at home (n = 77) although 10 chose to be seen at hospital. During the
Correspondence to: Dr A. Challen, Royal Manchester Children’s Hospital, Hospital Road, Pendlebury, Manchester, M 2 7 1HA, UK.
850
0742-3071/92/090850-05$07.50
@ 1992 by John Wiley & Sons, Ltd.
Accepted 3 June 1992 DIABETIC MEDICINE, 1992; 9: 850-854
DT17 research visits they were asked to complete psychometric scales assessing emotional difficulties with, and attitude to, diabetes. The items of the scale were read aloud to the subjects, who responded independently on their own score sheets. Twelve young people did not wish to participate in a research visit; they completed the questionnaires at home and returned them by post. These 12 adolescents did not differ by age, age at diagnosis, duration of diabetes, sex ratio, parental occupational status or level of glycaemic control from the rest of the study group. Following the second assessment the medical case notes were examined to identify any hospital admissions due to hypo- or hyperglycaemia during the intervening period. None of the three consultant paediatricians had specific criteria for admission of their diabetic patients and thus the decision to admit was taken by the duty junior and middle grade medical staff. Admissions for other reasons, such as surgical procedures, and visits to casualty without admission, were excluded from analyses. The young people’s glycaemic control was assessed by their mean glycated haemoglobin taken routinely at clinic visits during a 12-month period, 6 months before and after the first assessment. The glycated haemoglobin values for the study group were measured in three different laboratories: the results from one laboratory were adjusted by a correction factor following a comparison study as previously reported.” Social class (nonmanual/manual) for each family was determined using Office of Population Censuses and Surveys classification,’* based on father’s occupation or mother’s if there was no father in the family.
Adaptation to Diabetes Scales These are two 8-item scales measuring ‘attitude to diabetes’ and ‘emotional difficulty with diabetes’, respectively. The scales were designed for young people with diabetes between the ages of 10 to 17 years to assess aspects of psychosocial adaptation to diabetes (no reference to metabolic control was made during test development). The items are statements derived from interviews with adolescents with diabetes, and respondents are asked how frequently their experience matches these statements. High scores on the scales indicate greater emotional difficulty and a more negative attitude to diabetes. For the study population the attitude and emotional difficulty scales were found to have internal consistency coefficients of 0.78 and 0.85, and test-retest coefficients over 6-12 months of 0.73 and 0.66, respectively. Validation of the test has been described.’O
Results During the period between the two assessments, 10 young people were admitted to hospital. Their demographic, psychological, and laboratory characteristics are shown HOSPITAL ADMISSIONS OF ADOLESCENTS WITH DIABETES
CLINICAL PRACTICE in Table 1, and details of their admissions are described in Table 2. A comparison of their demographic characteristics, psychological test scores, and glycated haemoglobin values with those of the remaining study group is shown in Table 3. (Four young people did not complete the second assessment; they were not admitted to hospital during a comparable period (i.e. the median interval between assessments of 8.6 months)). Of those who were admitted to hospital, five adolescents presented with hyperglycaemia and five with hypoglycaemia. A comparison of demographic characteristics, psychological test scores, and glycated haemoglobin values between these two subgroups is shown in Table 4. When the two subgroups were compared with those not admitted separately, only the hyperglycaemia group differed, they had both greater emotional difficulty (p = 0.002) and a more negative attitude (p = 0.01). There were no significant correlations between mean glycated haemoglobin results and scores on the emotional difficulty and attitude scales ( r = -0.02, p = 0.81; r = -0.1 1, p = 0.29, respectively).
Discussion The results of this study demonstrate that adolescents who were admitted to hospital with unstable control were more likely to have experienced emotional difficulties with diabetes, than those who were not. The study results should not be interpreted as demonstrating an association between unstable diabetes and emotional difficulties as we have not included those adolescents who had episodes of unstable diabetes who were managed at home or in the accident and emergency departments without an admission. Further analysis of the admission group indicated that it was those presenting with hyperglycaemia (of whom four had ketoacidosis on admission) who had the greatest psychological difficulties with their diabetes, concurring with previous descriptive r e p ~ r t s . ~It- ~is difficult to be certain whether those readmitted with hypoglycaemia have more than average difficulty coping with their diabetes, given the small sample size. Certainly there have been reports of factitious hyp~glycaemia~,’and suicide attempts14 from insulin overdosage. Social and demographic characteristics did not differ between the two groups. Glasgow et a1.15 have recently demonstrated increased socioeconomic deprivation in children admitted to hospital with acute diabetic complications, in comparison to outpatients and new admissions over a 5-year period, and suggested serious psychosocial dysfunction as a possible underlying cause. Glycaemic control (as reflected in glycated haemoglobin values) did not differ between the readmission group and the rest. In addition glycaemic control was not related either to emotional difficulties or attitude to diabetes. This result concurs with a number of studies that have shown no correlation between glycaemic control and self-report psychometric tests assessing gen-
85 1
DTfl
CLINICAL PRACTICE Table 1. Demographic, psychological and laboratory details of the ten adolescents who were admitted to hospital during the study period Case Sex, age Socioeconomic Mean HbAlc Emotional number and duration group in the 6 difficulty and months before attitude score (years) and after the 1st assessment
1. 2. 3. 4.
5.3, 5.9, M, 2.8, F, 5.6,
5. 6. 7.
F, 3.8, F; 11.3; M, 10.6,
8. 9. 10.
F, 16.7, M, 15.5, M, 15.8,
F,
F,
4.7 manual 4.4 manual 7.5 non-manual 2.9 both parents unemployed 7.6 non-manual 2.3 non-manual 4.2 both parents unemployed 7.9 manual 13.4 manual 5.8 in foster care
0.7 7.7 2.8 3.7
I10
31, 17, 12, 19,
29 17 16 16
2.3 10.6 13.4
16, 18 18, 1 7 35, 19
results 10.7 17.1
23, 22 16, 1 1 37, 30
Table 2. Clinical details of the ten admissions Case number 1
2 3
8
9 10
852
Details of admissions
Admitted via accident and emergency department 3 days after starting on an insulin pen, with symptoms of polyuria and polydipsia and home results of glycosuria and 1 % ketones. Results on arrival: blood glucose: 16.4 mol I-', normal electrolytes, no H C 0 3 analysed, ketones negative. Admission while on holiday with nocturnal hypoglycaemic fit. Glucagon not given as parents panicked over the difficulty of assembling syringe. Taken semiconscious to local hospital, in-patient for several days. Two admissions for hypoglycaemia. First admission: semiconscious, home blood glucose < 1 mmol I-', admitted via GP, after IV glucose. Second admission: nocturnal hypoglycaemic convulsion, emergency IV glucose treatment from ambulance staff prior to admission via GP. Parents did not give glucagon as had not collected new supply. Self referral to accident and emergency department in hysterical state. Felt 'shaky' after taking evening insulin, drinking cider and having not eaten. Known to have poor parental support. Blood glucose 4 mmol I-' following exchanges on arrival. Overnight admission. Day admission after self referral to accident and emergency department following hypoglycaemic episode at school that was treated with glucagon injection. Occurred following administration of twice normal insulin dose. Blood glucose 6.5 mmol I-' on arrival at hospital. Admitted with symptoms of polyuria and vomiting, after discussion with consultant. Dehydrated but alert on examination. Results: blood glucose 27.2 mmol I-', H C 0 3 8 mmol I-' and ketonuria. Cause presumed due to change of treatment to insulin pen coupled with an intercurrent viral infection. Admitted via GP with symptoms of abdominal pain and vomiting. On examination: dehydrated with Kussmaul respirations, but alert. Results: blood glucose 55 mmol I-', H C 0 3 6.0 mmol I-' and ketonuria. Parents thought episode could be due to dietary indiscretions. Family vague about details of diabetes management. Two admissions with ketoacidosis via GP. Had recently moved in with aunt and uncle after having difficulties at home. First admission: admitted with nausea and vomiting. Dehydrated, kussmaul respirations but alert. Results: blood glucose 29.6 H C 0 3 4.7 mmol I-' and ketonuria. Episode secondary to missed insulin. Second admission: admitted with vomiting, diarrhoea, anorexia and chest pain. Omitted insulin because of anorexia. On examination unwell with kussmaul respirations. Results: blood glucose 41.6 mmol I-', H C 0 3 6.0 mmol I-' and ketonuria. Admitted with hypoglycaemic convulsion after giving double dose of insulin as he thought his blood glucose level was high (he refuses to do home monitoring). Treated with IV glucose with good recovery. Admission with ketoacidosis after running away from foster home and omitting insulin injections. Mild dehydration and drowsy on examination. Results: BM stick 44 mmol I-' on arrival: after subcutaneous insulin: blood glucose 15.5 mmol I-', H C 0 3 9 mmol I-', ketonuria.
A.H. CHALLEN ET AL.
Drn
CLINICAL PRACTICE Table 3. Differences in demographic characteristics, glycaemic control and psychological response to diabetes, between those who were and were not admitted to hospital Adolescents admitted Adolescents not to hospital admitted to hospital ~
Sex ratio (ma1e:female) Age (years) Duration of diabetes (years) Age at diagnosis (years) Socioeconomic group (non-manual manual) MGH (%) Emotional difficulty scores Attitude to diabetes scores
4:6
44:49
15.4 (10.6-16.7) 5.2 (2.3-13.4)
14.9 (10.6-17.0) 4.8 (0.3-14.7)
8.9 (2.0-12.7)
9.9 (1.3-15.5)
2:5
38:41
12.3 (7.65-17.1) 19 (12-37)
11.7 (5.8-1 8.0) 14 (8-36)"
18
16
(11-30)
(8-40)
=p = 0.01. Results as median (range) or as stated.
MGH: mean glycated haemoglobin in the 6 months before and after the first assessment.
Table 4. Differences in demographic characteristics, glycaemic control, and psychological response to diabetes, betwen those admitted with hyperglycaemia and hypoglycaemia
Sex ratio (ma1e:female) Age (years) Duration of diabetes (years) Age at diagnosis (years)
MGH (%) Emotional difficult scores Attitude to diabetes scores
Adolescents admitted with hyperglycaemia
Adolescents admitted with hypoglycaemia
2:3
2:3
15.3 (10.6-16.7) 4.6 (2.3-7.9)
15.5 (1 3.5-1 5.9) 7.5 (2.9-1 3.4)
8.9 (6.4-10.7)
6.2 (2.0-12.7)
12.0 (10.6-17.1) 31 (1%-37)
12.3 (7.65-1 3.7) 16 (12-19)"
22
16
(17-30)
(ll-18)a
"p = 0.02. Results as median (range) or as stated. MGH: mean glycated haemoglobin in the 6 months before and after the first assessment.
eralized psychological function.'8-20 Most of the studies that have shown a relationship between poor glycaemic control and psychological function are not comparable with this study. They have characterized a subgroup with marked social and psychological problems, using predominantly interview and school or parental report, who also have poor glycaemic ~ o n t r o l . ~ In ~ addition, ~'-~~ glycaemic control has been based on medical ratings, rather than glycated haemoglobin values. However, because admission to hospital with unstable diabetes has been used as a criterion in the assessment of poor HOSPITAL ADMISSIONS OF ADOLESCENTS WITH DIABETES
glycaemic control, these studies may well be reflecting the same conclusion as in this r e p ~ r t . ~ , ~ ~ , ~ ~ , ~ ~ We speculate that the relationship between emotional difficulties and admission with unstable diabetes could be due to a number of interrelating factors.
1. 2.
Neglect or deliberate manipulation of injection and diet (probably causal in cases 4, 5, 9, and 10). Young people and their families who cope better with diabetes are more likely to take effective action, such as insulin adjustment or administration of
853
DTT7
CLINICAL PRACTICE 3.
4.
glucagon, to prevent hospital admission (probably contributory in cases 2, 3, 4, 7, and 8). Hospital admission might exacerbate young people’s emotional problems by confronting them with the reality that they have a life-threatening disorder, and by increasing their frustration with diabetes and their sense of failure in managing their disease. Within the context of this study the decision to admit was based upon the subjective assessment of the duty doctors, who may have been more likely to admit from the accident and emergency department if they were aware of psychosocial problems.
It is possible therefore that measures to reduce emotional stress and improve confidence with diabetes management may be as important as manipulating insulin and diet in preventing further admissions, particularly for those admitted with hyperglycaemia. The diabetes health visitor plays an important role in identifying emotional difficulties among young people with diabetes, and close liaison with child and adolescent psychiatrists may faci Iitate diagnosis. Intervention to reduce the emotional distress experienced by these young people might include the opportunity for further exploration of their problems with a member of the diabetes team, sharing experiences at a young people’s support group, or input from a child psychiatrist. Orr8 and Goldeng have demonstrated a dramatic reduction in the number of admissions for unstable diabetes following a programme of education and individual, group or family counselling. Further research is required to support the premise that psychological diagnosis and intervention can reduce readmissions and perhaps prevent the unfortunate recurrence of admissions with unstable diabetes with its disruptive and distressing affects on the lives of young people and their families.
Acknowledgements We are grateful to Dr D. Savage and Dr T.L. Chambers for their support, Dr C. Pennock for his advice on glycated haemoglobin results, and to the young people and their families who took part in this research. A.H. Challen was supported by Novo Laboratories.
5. 6. 7.
8.
9. 10. 11. 12. 13. 14.
15.
16. 17.
18. 19. 20. 21.
References
22.
1.
Rosen H, Lidz T. Emotional factors in the precipitation of diabetic acidosis. Psychosom Med 1949; 11 : 21 1-21 5. 2. Gale E, Tattersall R. Brittle diabetes. Br ) Hosp Med 1979; 21: 589-597. 3. Gill GV, Husband DJ, Walford S, Marshall SM, Home PD, Albertie KGMM. Clinical features of brittle diabetes. In: Pickup JC, ed. Brittle Diabetes Oxford: Blackwell, 1985: 29-40. 4. Flexner CW, Weiner JP, Saudek CD, Dans PE. Repeated
854
23. 24. 25.
hospitalization for diabetic ketoacidosis. The game of ’Sartoris’. Am I Med 1984; 76: 691-695. Schade DS, Drumm DA, Duckwroth WC, Eaton RP. The etiology of incapacitating, brittle diabetes. Diabetes Care 1985, 8: 12-20. Loughlin WC, Mosenthal HO. Study of the personalities Dis Child 1944; 68: of children with diabetes. Am 13-1 5. White K, Kolman ML, Wexler P, Polin G, Winter RJ. Unstable diabetes and unstable families: A psychosocial evaluation of diabetic children with recurrent ketoacidosis. Pediatrics 1984; 72: 749-755. Orr DP, Golden MP, Myers G, Marrero DG. Characteristics of adolescents with poorly controlled diabetes referred to a tertiary care centre. Diabetes Care 1983; 6: 170-1 75. Golden MP, Herrold AJ, Orr DP. An approach to prevention of recurrent ketoacidosis in the pediatric population. ) Pediatr 1985; 107: 195-200. Challen AH, Davies AG, Williams RJW, Haslum MN, Baum ID. Measuring psychosocial adaptation to diabetes in adolescence. Diabetic Med 1988; 5: 739-746. Challen AH, Davies AG, Williams RJW, Baum JD. Support of families with diabetic children: Parents views. Practical Diabetes 1990; 7: 26-31. Office of Population Censuses and Surveys. Classification of Occupations 1980. London: HMSO, 1980. Orr DP, Eccles T, Lawlor R, Golden M. Surreptitious insulin administration in adolescents with insulin-dependent diabetes mellitus. IAMA 1986; 256: 3227-3230. Potter J, Clarke P, Gale EAM, Dave SH, Tattersall RB. Insulin-induced hypoglycaemia in an accident and emergency department: the tip of an iceberg? Br Med 1982; 285: 1180-1 182. Glasgow AM, Weissberg-Benchell J, Tynan WD, Epstein SF, Driscoll C, Turek J, et a/. Readmissions of children with diabetes mellitus to a children’s hospital. Pediatrics 1891; 88: 98-104. Delbridge L. Educational and psychological factors in the management of diabetes in childhood. Med ) Aus 1975; 2: 737-739. Simonds J, Goldstein D, Walker B, Rawlings S. The relationship between psychological factors and blood glucose regulation in insu I in-dependent diabetic adolescents. Diabetes Care 1981; 4: 610-615. Gross AM, Delcher HK, Snitzer J, Bianchi B, Epstein S. Personality variables and metabolic control in children with diabetes. ) Genet fsychol 1985; 146: 19-26. Close H, Davies AG, Price DA, Goodyer IM. Emotional difficulties in diabetes mellitus. Arch Dis Child 1986; 61 : 337-340. Evans CL, Hughes IA. The relationship between diabetic control and individual and family characteristics. 1 Psychos Res 1987; 3: 367-374. Sterky G. Family background and the state of mental health in a group of diabetic children. Acta Paediatr Scand 1963; 52: 377-390. Swift CR, Seidman F, Stein H. Adjustment problems in juvenile diabetics. Psychosom Med 1967; 29: 555-571. Koski ML. The coping processes in childhood diabetes. Acta Paediatr Scand 1969; (suppl) 198: 1-56. Ludvigsson J. Socio-psychological factors and metabolic control in juvenile diabetes. Acta Paediatr Scand 1977; 66: 431-437. Gath N, Smith MA, Baum ID. Emotional, behavioural and educational disorders in diabetic children. Arch Dis Child 1980; 55: 371-375.
A.H. CHALLEN E l AL.
