Gut feelings: a psychologist's 20-year journey with paediatric surgeons

My journey with the paediatric surgeons at Great Ormond Street Hospital for Children (GOSH) started many years ago, and led to a question posed by the editor of Archives of Diseases in Childhood in his editorial comments on one of my recent publications¹: ‘should we be surprised that a psychologist's questions might probe more deeply than the surgeons fingers?’².

First, I must go back to the beginning of my association with the paediatric surgeons, way back in 1982 when I began developing a research study which formed the basis of my doctorate, and which was my introduction to the issues of psychology in paediatric surgery. The research on which this thesis was based, and which was initiated by Professor Lewis Spitz (Nuffield Professor of Paediatric Surgery) and Dr Richard Lansdown (then head of the Department of Psychological Medicine), was concerned with newborn babies who had a mixture of life-threatening surgically correctable congenital anomalies. Subsequent research studies moved in a bottom-up direction, along the gut, from anorectal malformations, through Hirschsprung's disease, to oesophageal atresia.

STUDY 1: MAJOR NEONATAL SURGICAL STUDY

This first piece of research was a prospective longitudinal study that aimed to examine the psychological effects of major neonatal surgery on infants and their families. I followed 72 babies and their families from birth to 3 years of age³, and recently went back to this study, following up the sample when the children had reached early adolescence⁴.

Major neonatal surgery and hospitalization soon after birth obviously go hand in hand, and involve separation from the parents. One aim of the study was to examine the effects of early hospitalization and mother—infant separation on the child and the parents. At that time (1982), following the setting-up of neonatal intensive care units and specialcare baby units in the 1950s/1960s, most studies of early hospitalization had been on populations of premature and low-birthweight babies. However, none of these investigations included newborn babies with congenital abnormalities needing surgical correction. So, ours was the first of its kind. Babies with these anomalies (for example, oesophageal atresia, diaphragmatic hernia, gastroschisis, Hirschsprung's disease, anorectal malformations, non-malignant tumours) are few in number, and because I only included those born at term the surgical group sample of 30 took just over 12 months to recruit. I matched these babies on six variables (gender, birth order, mother's age, social grouping, marital status, and geographical location) with healthy full-term newborns, and I also looked at a group of term babies admitted for intensive care for medical reasons—the idea being to see if those who required surgery had a different outcome from medically sick hospitalized neonates⁵ (healthy controls and some medical babies were recruited from Queen Charlotte's Maternity Hospital, London, and other controls from Luton and Dunstable Maternity Hospital).

Although longitudinal studies require a long-term commitment on the part of both the families and the researcher, there were clear benefits in the longitudinal design. For example, we expected to find that early hospitalization would have an adverse effect on the mother—child relationship. But when the 12-month-old hospitalized babies were compared with the healthy ones, contrary to expectations separation from the mother did not differentiate the case—control pairs. The surgical (and medical) mother—baby relationships were as secure, warm and loving as amongst the healthy pairs⁵.

However, by 3 years things had changed. A small but important minority of the surgical group mother—child relationships were troubled. This was shown by the fact that 30% of the surgical group children compared with just under 12% of controls had behaviour problems, and 27% of the surgical mothers had an overall rating of poor parenting, compared with just under 12% of the control mothers. The characteristics of poor parenting include insensitive, indulgent and inconsistent handling of a child's behaviour. Further evidence of deterioration in the surgical mother—child relationships was shown by the fact that 21% of the surgical group mother—child relationships that were secure at 12 months were problematic at 3 years, compared with only 4% among the control group. Parental factors such as the age and education levels of the mother, and repeat hospital admissions, influenced outcome at 3 years⁶.

I followed up the sample when they were aged between 11 and 13 years and found that, as judged by parent and teacher ratings, behaviour problems were more frequent among the surgical group than in the two comparison groups—parents' reports, surgical 30%, medical 22%, controls 17% (norm 18%); teachers' reports, surgical 21% medical 0%, control 11% (norm 18%). Not unexpectedly, the pattern of intellectual development also changed over time. At the 12-month stage, the developmental progress of the surgical and medical babies was significantly behind that of the healthy 1-year-olds⁵. At 3 years of age, composite IQ scores showed that the cognitive development of the medical group children was now similar to that of the healthy control children, whereas the scores of the surgical group were significantly behind those of the two comparison groups⁴. However, within the surgical group, the children whose conditions were resolved within the first few months of life were doing well, and their abilities were comparable to those of their healthy peers, while, not surprisingly, those who had repeat hospital admissions and persisting medical problems were functioning at a lower level than the other surgical and control children.

As judged by the National Curriculum Assessment taken at age 11 (in the three core subjects English, mathematics, science) and their teachers' report of academic performance, the adolescents in the surgical group were performing significantly less well than their peers on all measures of academic performance even when those with neurological deficits were excluded and the results were controlled for gender, social group and mothers' age at the time of birth. But the factors contributing to these differences between the groups differed from those in the first 3 years of life. Variables such as length of hospital stay, repeat admissions, and parental factors were no longer associated with outcome. The only medical factor shown to have long-term adverse effects was mechanical ventilation for ≥ 4 days in the neonatal period. (In addition, increased behaviour problems at 3 years were associated with lower performance levels in adolescence.) The babies who required ventilation for ≥ 4 days were obviously sicker than the other babies, but a measure of all the infants' medical condition during the neonatal period revealed that babies who were very sick but were not ventilated for more than 4 days were doing well academically⁴. The long-term effects of mechanical ventilation were most pronounced on the mathematics and science performance levels. These findings are similar to the evidence from long-term studies of premature and low birthweight adolescents and may indicate that early damage only becomes apparent when cognitive tasks require more complex skills⁴.

On completion of this study, Professor Spitz suggested a new line of research. He asked me to investigate the long-term psychological consequences after corrective surgery for anorectal malformations.

STUDY 2: ANORECTAL MALFORMATIONS

In 1991, we recalled all the patients who had definitive surgery for anorectal anomalies at GOSH between 1974 and in 1987. 160 patients aged between 6 and 18 and their parents participated (59% high/intermediate anomalies; 41% low anomalies). The surgeons assessed the patients clinically, and I interviewed the patients and parents separately—which in some instances took up to six hours—during an outpatient appointment at the hospital⁷.

This was another ground-breaking study. We were the first to examine, in depth, psychological and clinical outcomes in such a large group of patients with anorectal anomalies. The main clinical outcome variable was a continence score (Kelly score) obtained by the surgeons during their clinical examination of the patient. Apart from a diagnostic psychiatric interview with the child, the psychological outcome variables were a battery of questionnaires for assessing psychosocial adjustment of the child, as well as parental factors⁷.

Not unexpectedly, a sizeable proportion of the patients with high anomalies lacked full bowel control: only 13% were fully in control of their bowels compared with 73% of those with low anomalies. Figure 1 illustrates age at achieving continence in relation to the level of the anomaly, in an extended sample—that is, I added into the survival-time analyses data from children aged 4 to 6, bringing the total sample analysed to 256 children who had surgical repair at GOSH. It is evident that, by the age of 17, only 25% of those with high or intermediate anomalies were continent. In contrast, 95% of those with low lesions were continent by 16⁸.

Figure 1.

Age of achieving continence in relation to the level of the lesion. ▪ High/intermediate; • low

About 30% of the sample had emotional and behavioural adjustment problems—a significantly higher proportion than in normal populations. Much to my surprise, however, maladjustment was not in general related to poor bowel control. In the psychiatric interview, I found a high level of disturbance among both the continent and the incontinent adolescents, especially the boys (30% and 32%, respectively). Support for these findings came from the parents' and teachers' assessments. According to the parents' reports, the prevalences of disturbed behaviour were 33% in continent boys, 33% in continent girls, 24% in incontinent boys, 21% in incontinent girls; teachers indicated disturbed behaviour in 33% of continent boys, 18% of incontinent boys. The young girls (aged 6-11 years) were the only subgroup in which the incontinent children were more emotionally disturbed (27% versus 6%)⁷.

An important outcome was the insight I gained from the in-depth interviews with the children and their parents, about how they coped with this socially devastating physical disorder⁹. I was taken aback by reports of the unremitting misery many were experiencing. Fortunately, through presentations at surgical meetings and publications, together with supporting evidence from others who have since looked at psychological and functional outcomes amongst other groups of anorectal patients, there is I believe general acceptance that many of those patients experience serious difficulties after surgical repair of their anomalies.

Why was there such a high level of psychological disturbance among this large sample, continent and incontinent? What was contributing to the increased vulnerability for emotional disturbance? From my research findings I hypothesized that, as with the neonatal study, this could be traced to difficulties the parents had in dealing effectively and consistently with their child's behaviour, both in everyday living and in relation to their physical health problems. Most of the parents I had interviewed found it impossible to ignore or forget that their child had been born with a life-threatening abnormality. Even if the child or adolescent was functioning ‘as normal’, he or she continued to be very special and tended to be treated differently from other children in the family. So, from infancy onwards, sympathy for the child may lead parents to treat the child as especially vulnerable. As the child grows up, this results in overindulgence and over-involvement of the parents⁸, leading to difficulties for the mother in setting limits and boundaries to the child's behaviour; the child remains overly dependent on his or her parents.

To test this hypothesis I suggested that we should set up a study to assess the long-term effectiveness of therapeutic intervention, for families with children with impaired bowel function as a consequence of organic disease treated surgically.

STUDY 3: INTERVENTION STUDY

There were three main groups—patients with anorectal anomalies, patients with Hirschsprung's disease and a non-surgical comparison group with intractable constipation. The patients were referred by the paediatric surgeons, gastroenterologists, neurosurgeons, and by Dr Graham Clayden (the expert on childhood constipation). I aimed to provide for a minimum of four and a maximum of six therapeutic sessions, and a follow-up assessment some 6-12 months after completion of treatment. Therapeutic intervention involved cognitive behavioural techniques, operant conditioning and social learning techniques, and behavioural management and parenting skills advice for the parents.

The main aims of the therapeutic intervention were to help the child and the parents understand the nature and consequences of the child's condition; to encourage the child to comply with a toileting regimen and any therapy prescribed; and to help them deal with social problems associated with faecal incontinence. Outcome was measured with a continence score and my usual battery of psychological tests at or before the first appointment, at the end of treatment, and again, 6 to 12 months after completion of treatment.

I followed up 88 patients aged between 18 months and 15 years, 6 months or more after completion of the treatment sessions. The data showed that the intervention strategies resulted in considerable improvements in the children's bowel control, and this improvement was maintained over time. For example, among the Hirschsprung's disease patients, 14 of the 17 (82%) children achieved continence or showed a substantial improvement in bowel control. This was coupled with increased self-esteem, fewer behaviour problems, increased school attendance and social interaction with peers, and better parent—child relationships. Similarly, for the anorectal anomalies and intractable constipation patients the numbers improved were 23 out of 37 (62%) and 24 out of 34 (71%), respectively.

The intervention sessions were in great demand, but as a piece of research the study had some problems. I was generally able to keep to the study design and protocol, but some patients needed fewer than four sessions and others many more than six. Moreover several families, though desperately wanting advice and guidance, were reluctant to complete the questionnaires and other measures specified by the protocol. Although the referring consultants generally remembered to tell their patients they would be taking part in a research project, some families thought their referral was just part of NHS care. Over 4 years, I saw and treated around 200 patients and became convinced that many patients benefit from this kind of focused psychosocial support.

STUDY 4: HIRSCHSPRUNG'S DISEASE

In 1997 Professor Agostino Pierro asked if I would do a study with him to investigate faecal continence in patients with Hirschsprung's disease. We looked at patients with two types of Hirschsprung's disease—a group with total colonic aganglionosis, and a second group with disease confined to the rectosigmoid region. Professor Pierro had not been involved in the treatment of any of these patients. They had been operated on by surgeons on the staff in the years 1979 to 1990.

Functional and psychological outcomes of 15 total colonic aganglionosis (TCA) patients were compared with those of a group of patients with disease confined to the rectosigmoid (RS) carefully matched for age and gender. The 30 patients were aged between 7 and 18 years. (The numbers are small because TCA accounts for only a small proportion of patients with Hirschsprung's disease.) Professor Pierro examined the patients clinically. A continence score, the Kelly score, was the main medical outcome variable. Dr Hisato Tsuji, one of Professor Spitz's research fellows from Japan, used anal endosonography to assess the status of the internal and external sphincters, and I interviewed the patients and parents separately, in addition to my usual battery of tests¹.

We found that the proportion of patients with faecal incontinence 7-17 years after definitive surgery was high in both groups, whatever the extent of the disease, but once again there was no association between incontinence and psychological adjustment¹. This study highlighted an important pitfall. There was difficulty obtaining full disclosure about faecal incontinence from patients and their parents. For example, according to Professor Pierro's clinical examination and review, 40% of the TCA patients had full bowel control. However, in answer to my questions, only 13% had full control, suggesting that a psychologist (experienced in interviewing patients with these abnormalities) does probe more deeply than the surgeon's fingers². A psychologist has more time, but I believe many patients with these kinds of conditions are very reluctant to own up to a surgeon about any physical problems, even if the surgeon was not involved in their surgery or care, in case they may be required to have further invasive investigations¹.

This study also demonstrated a large shift in patient cooperation with clinical research studies. In the anorectal study very few, if any, refused to allow the surgeon to examine them physically. That was in the early 1990s. Towards the end of the 1990s, patient attitudes had changed, and they were prepared to assert their rights and refuse any form of internal examination, despite having signed a detailed informed consent form. 20% of the patients, principally the RS patients, would not allow Professor Pierro to carry out a rectal examination, and Dr Tsuji was unable to persuade a large proportion of the patients (43%) to undergo anal endosonography. So, not only is it difficult to find funding for some forms of clinical research, but also there is increasing difficulty in gaining a fully representative sample of patients who will agree to clinical research protocols.

I now come to the study that marks the end of my association with the surgeons—at the top of the gut, the oesophagus. Professor Spitz asked me to assess the quality of life of his patients who had had gastric transposition (GT) for oesophageal atresia—patients in whom the gap between the two blind ends of the oesophagus was too great to be joined together. He has done hundreds of these operations since 1981, but a high proportion of patients are resident overseas and have conditions other than congenital oesophageal atresia (OA).

STUDY 5: GASTRIC TRANSPOSITION FOR CONGENITAL OESOPHAGEAL ATRESIA

Our study group was small since we included only patients with congenital OA resident in the UK. The 28 patients, aged between 2 and 22 years, fell into two groups—those treated by gastric transposition as the primary reconstructive surgical procedure (n=13, mean age 13 years, mean time since GT 12 years), and those who had had previous, unsuccessful attempts at repair or replacement of their oesophagus (n=15, mean age 13 years, mean time since GT 8 years). I modified the Gastrointestinal Quality of Life Index (Eypasch) to measure the health-related quality of life of the OA patients. The index assesses eating behaviours, disease-specific symptoms, and psychological, physical and social functioning. I interviewed the patients and their families and they completed my usual questionnaires.

I found that, in general, the patients who had had primary reconstructive surgery experienced fewer eating and disease-specific troubles than patients who had undergone previous unsuccessful attempts at repair or replacement of the oesophagus. The patients without debilitating associated anomalies were living normal lives, although they tended to be less socially and emotionally independent than their peers.

Psychological assessments revealed a high proportion of patients with externalizing behaviour problems, such as ‘acting out’ and antisocial behaviours—a similar finding to that in adolescents who had had neonatal surgery. The picture differed from that in the anorectal patients, for whom internalizing problems such as depression and withdrawal symptoms predominated.

The paper on this research will be published in the Journal of Pediatric Surgery. Suffice it for me to say that these patients were the most grateful I have ever met. They all regarded Professor Spitz as their saviour, and it was gratifying for me that my final study involved such satisfied patients and their families.

CONCLUSION

My journey with the paediatric surgeons has been a fascinating experience, and I am fortunate to have worked with them for so long. I have become increasingly persuaded that there is a place for a full-time close association between a psychologist and surgeons working in this field—not only for research but also to provide support for these families. Difficult or disturbed behaviour might be avoided, and the quality of life for the child and the family might be improved both in the short and the longer term, if informed psychological guidance and support were available as part of the continuing care following surgery. In the 21st century these considerations are particularly important since many more survivors of major neonatal surgery reach adult life.