The Registry and Follow-Up of Complex Pediatric Therapies Program of Western Canada: A Mechanism for Service, Audit, and Research after Life-Saving Therapies for Young Children

Newly emerging health technologies are being developed to care for children with complex cardiac defects. Neurodevelopmental and childhood school-related outcomes are of great interest to parents of children receiving this care, care providers, and healthcare administrators. Since the 1970s, neonatal follow-up clinics have provided service, audit, and research for preterm infants as care for these at-risk children evolved. We have chosen to present for this issue the mechanism for longitudinal follow-up of survivors that we have developed for western Canada patterned after neonatal follow-up. Our program provides registration for young children receiving complex cardiac surgery, heart transplantation, ventricular assist device support, and extracorporeal life support among others. The program includes multidisciplinary assessments with appropriate neurodevelopmental intervention, active quality improvement evaluations, and outcomes research. Through this mechanism, consistently high (96%) follow-up over two years is maintained.


Introduction
Neonatal and early childhood survivors of life-saving therapies including complex cardiac surgery and heart transplantation are at risk for neurodevelopmental sequelae. Possible determinants are brain injury [1,2], complications of early gestational age birth or associated genetic abnormalities [3][4][5][6][7], prolonged and recurrent hospitalizations [8], and in some cases, a void of experiences available to other children [9,10]. The risk factors associated with adverse outcomes are multifactorial [11][12][13][14]. Enhanced understanding is emerging about patterns of health and neurodevelopmental outcome with focus on cognition, attention, motor skills, behavior, adaptive abilities, functional outcomes, health-related quality of life, and parental stress research [9,10,[15][16][17][18][19][20]. This valuable information can assist in counselling parents of children needing specific procedures. In addition, combined with detailed acute care databases, this information can identify potentially modifiable predictors of outcome and evaluate part or all of the systems of care for a given therapy [21].
For more than 30 years, neonatal follow-up clinics have provided multidisciplinary assessments of high-risk neonates [22]. These clinics have the framework, experience, and expertise to complete long-term follow-up of at-risk children [23]. Most neonatal follow-up clinics provide neurodevelopmental assessments and have a mission of service to the child and family, as well as audit and research. Such service may be initiated by the program or through appropriate referrals. Individualized intervention may include nutritional therapy, physical therapy, occupational therapy, speech-language therapy, assessment and treatment of hearing loss, psychological interventions, social work support, and/or medical treatments. Group developmental intervention programs may be community based or tertiary care based. Key to the work of a neonatal follow-up clinic is the mission of audit [22]. Neurodevelopmental morbidity outcomes are linked to acute care variables to provide a basis for quality improvement as well as population trends. Various types of research facilitated through neonatal follow-up clinics include specific, local, and multicentered outcome studies as well as multicentered randomized controlled trials [22]. These clinics provide a setting for teaching trainees and interaction among all disciplines interested in the welfare of the high-risk infant [22]. Neonatal follow-up clinics are closely linked to neonatal intensive care but may be located in general or pediatric hospitals, or rehabilitation facilities. Many neonatal follow-up clinics now have extended entry criteria to include ill neonates beyond preterm children and some see children whose therapy occurred after the neonatal period [21]. Assessment guidelines for high-risk neonates have been published [24,25].
This paper presents the organization of the Registry and Follow-up of Complex Pediatric Therapies Program of Western Canada which is based on the neonatal follow-up clinic model. This registry has been successful in the followup of infants and young children surviving life-threatening illnesses treated with newly emerging health technologies, identifying modifiable variables at each level of care, building a quality improvement program, and publishing outcomes research while providing a range of services for survivors with neurodevelopmental sequelae.

Methods
This is an interprovincial longitudinal neurodevelopmental follow-up program that began in 1999 for children from the western provinces of Canada and corresponding northern territories who have had invasive complex therapies in Alberta, Canada. The majority of procedures including complex cardiac surgery are done at the Stollery Children's Hospital, Edmonton, Alberta, Canada. The program evolved over the period from 1996 to 1999 beginning with a specif-ic research study on neurodevelopmental follow-up after complex cardiac surgery done at ≤6 weeks of age and published in 2004 [26].

Organization.
Program members include representatives of acute care departments including, cardiology, cardiovascular surgery, nephrology, hepatology/gastroenterology, and neonatal and pediatric intensive care in the province of Alberta as well as members from these specialities from participating sites across western Canada. Follow-up occurs at six sites. Follow-up team members including paediatricians with expertise in neurodevelopmental follow-up, psychologists, and audiologists are part of the membership. There is a direct liaison with healthcare administration at each site. A steering committee consists of the cochairs (a neurodevelopmental pediatrician, pediatric intensivist, and neonatal developmental follow-up/intensivist) and membership from the acute care departments. There are three standing subcommittees within the steering committee: data entry, retrieval, and linkage; outcome measurements; evaluation. Enrolment criteria are determined through the work of the core steering committee and individual departments.
Following a research study from 1996 to 1999 [26], a project grant was obtained from the provincial government for the period from 1999 to 2006 to determine healthcare outcomes of emerging technologies. Since 2006, the program has been part of the global budget at all sites to varying degrees with respective accountability.

Procedure
Referrals for follow-up are made by the attending physician within each acute care department based on predetermined entry criteria. Once survival is deemed likely, a nurse coordinator registers the child and discusses follow-up procedures with the parents. As parents become involved in the followup registration process, they come to understand the dual purpose of follow-up to include service for possible developmental concerns for their child and parental psychosocial support as well as an audit of outcomes of new complex therapies. Contact is made with the follow-up clinic at the tertiary site of origin. The project is approved by local health research ethics boards. All parents/guardians sign informed consent for follow-up and participation in the database.
Databases were established for children needing each of the listed therapies. Acute care variables were chosen and modified by the cochairs and representatives of each department, and data are recorded prospectively by a program assistant. Outcome variables are standardized age appropriate, and uniform at all sites. Once compiled, anonymous data are returned to the central site for checking with a hard copy of the assessment results and for keying. All data entry is checked by one of the cochairs. SPSS, version 17, is currently used for data storage and SPSS and SAS are used for analyses.

Assessments.
At hospital discharge, we record specific acute measures and any underlying conditions that may ultimately affect the outcome of the child. Birth head circumference and head circumference at discharge are recorded as has been previously published for children after hypoxic-ischemic encephalopathy [27]. Information on genetic abnormalities or dysmorphic features, antenatal infections, and maternal substance abuse are recorded. Any available cranial imaging information is recorded as has been recommended [28]. Birth prevalence for the various defects may vary by year, location, race, and socioeconomic levels [29] thus information on year, locale, race [30], and socioeconomic levels [31] are completed.
Prior to the first follow-up assessment, records concerning the visual awareness of the child, including ophthalmological reports, formal hearing test results as have been recommended for at-risk children, [32] and recorded observations of feeding and swallowing by an experienced nurse or therapist are obtained. The latter includes specific concerns for airway anomalies in children with congenital heart disease which may subsequently interfere with the child's health and development [33]; information on facial hypotonia, a common difficulty for ill children, often addressed with positioning and feeding suggestions; incoordination or absent swallowing requiring alternate feeding methods with sensory lip and mouth stimulation by a feeding therapist [34]. Any concerns about the vulnerable child syndrome and its consequences for the child's social, emotional, and behavioural growth are recorded [35].

Measures.
Timelines and approach to the long-term follow-up are similar to published guidelines [24,25]. Measures include assessment, rating, or screening tools. The focus of the assessment at each age is to assist in providing service for children to reduce impact on limitations of function, activities, and participation as recommended using the WHO International Classification of Function [36]. Outcome measures are standardized, and the latest edition is used. Overall, test batteries appropriate for age measure neurodevelopment, intellectual function, attention, memory and learning, language, behavior, motor function, and hearing, as well as neurological examination (Table 1).

Statistical Analysis
Kaplan-Meier survival curves with log-rank test are used to compare cumulative survival across several groups. Continuous outcomes of morbidity are presented as means (SD); categorical variables are presented as counts and percentages. To screen for variables associated with outcomes, univariate regression models are used. Multiple regression models are built based on variables significant at P < .10 in the corresponding univariate analysis, and after screening for multicollinearity. The type of outcome drives the choice of the regression model and presentation of results, for example, linear regression is used for continuous outcomes and results are presented as effect sizes, logistic regression is used for dichotomous variables and results are presented as odds ratios, and the Cox proportional hazard model is used for survival data, and results are presented as hazard ratios. All estimates are presented together with confidence intervals and two-sided P values.
When testing several outcomes, one should adjust for multiple comparisons. The larger the number of outcomes tested, the more inflated the Type I error. Bonferroni type adjustments are made in such situations.

Results
Since 1996, over 1000 children have been registered with this follow-up program. This includes 659 children following complex cardiac surgery done at age ≤ 6 weeks (Table 2), 11 children after chronic renal dialysis in the first year of life, 175 children after solid organ transplantation done at age 5 years or less (61 heart, 97 liver, 17 kidney), 153 children after extracorporeal life support given at age 5 years or less (109 cardiac, 44 pulmonary), 21 children with ventricular assist devices at age 5 years or less, and 16 children following cooling after in-hospital cardiopulmonary resuscitation done at age 5 years or less. Follow-up at two years is >96% (Figure 1).

Service.
The Registry and Follow-up of Complex Pediatric Therapies Program has benefited children receiving emerging health technologies in our region. Assessments and interventions can make a significant improvement in the child's ultimate quality of life [52,53]. With early recognition of developmental delay and family support to promote the child's motivation and learning, the child's long-term outcome is improved. There are few children undergoing these rare-event therapies: in our region, they would typically not receive early intervention and early education without the follow-up clinic outcome assessments directing the family to the appropriate program (Table 3). Our follow-up program has a strong advocacy role for children and provides an added value service to the child and family while obtaining outcome data. The educational system benefits from having the children more prepared for learning at school entry and having knowledge about specific learning patterns of the survivors.
Knowledge of the risk, safety, and outcome of care for life-threatening illness enables acute care practitioners to discuss, from a stronger evidence base, the potential outcomes for children undergoing these therapies. Parents become more familiar with the process and begin to understand the life-saving procedures as only one step in the overall system  [39]. Visual Motor Integration-5th Edition [40]. Behavior Assessment System for Children-2nd Edition [41]. Adaptive-Behavior Assessment System-2nd Edition [42]. Pediatric Quality of Life Inventory, Generic Core Scales [43], and Cardiac Module [44]. Clinical Evaluation of Language Fundamentals-2nd Edition [45].
8 years ± 6 months (range 6.5 years to 16 y 11 months) selectively for children with single ventricle defects, heart transplantation, extracorporeal membrane oxygenation Psychologist Audiologist * * Physician/Nurse * * * Wechsler Intelligence Scale for Children-4th Edition [46]. Visual Motor Integration-5th Edition [40]. Wechsler Individual Achievement Test-2nd Edition [47]. NEPSY-2nd Edition [48]. Adaptive Behavior Assessment System-2nd Edition [42]. Behavior Rating Inventory of Executive Function Parent Form [49]. Child and Adolescent Scale of Participation [50]. Strength and Difficulties Questionnaire [51]. Pediatric Quality of Life Inventory, Generic Core Scales [43]. * Clinical assessment, no data collection. Measures individually determined. * * Standard audiological assessment for at-risk children [32]. * * * At each visit, the pediatrician and nurse complete a predetermined questionnaire about illnesses and hospitalizations as well as diet, feeding, sleep, immunizations, and parent support; record growth; complete a general and neurological examination. If required, other disciplines such as occupational therapist, dietician, or social worker assesses the child/family.  allows a proactive and dynamic approach to preparing parents with the challenges of raising children who have been very ill and gear interventions toward individual needs (Table 3). Modifications in the type of assessments we administer and the focus of counselling have evolved based on our results (Table 3). In some cases, these results may not be generalizable but apply to our populations. By grouping outcome data and analyzing predictor acute care variables, we have identified potentially modifiable variables resulting in specific quality improvement evaluations (examples are provided in Table 3).
Our program has contributed to the outcomes literature following complex cardiac surgery for very young children in a number of ways (Tables 4-7). Serial lactate determination has been shown to predict survival [64]. Building on our early study [26], we completed a longitudinal study and confirmed that neurodevelopmental outcomes at two years of age for children after complex cardiac surgery are predictive of the neurocognitive outcomes at age five years [65]. The sensitivity and specificity of the mental developmental index scores under 85 for predicting the full-scale intelligence score at preschool were 85.7% and 89.7%, respectively, indicating the value of testing children at younger ages.
We have published on the early childhood outcome following total anomalous pulmonary venous correction [60]. As an unexpected finding, we found that aboriginal children were markedly over represented in this group compared with the rest of our population. We have also published on the outcomes of children with interrupted aortic arch without chromosomal anomalies [66] and children with complex cardiac surgery in the neonatal period who have deletion 22q11.2 [7]. The latter are at a high risk for adverse neurodevelopmental outcomes. A recent paper of the twoyear neurodevelopmental outcomes of children following the Norwood right ventricle-to-pulmonary artery shunt surgical repair demonstrated improved survival and psychomotor developmental outcome over the previous Blalock-Taussig surgical era [58].
The findings of the follow-up program have justified the investment in several ways. We are one of few groups to describe the long-term neurodevelopmental outcome and general morbidity of many groups of cardiac diseases treated in the neonatal period (Tables 4 and 7). We have Therapy and early developmental intervention as needed is available from time of hospital discharge at each site. Children's attending physicians are more aware of the developmental needs of survivors and the importance and benefit of early developmental intervention.

Enhancement of multidisciplinary assessment clinics
Therapists demonstrated the benefits of multidisciplinary assessments of survivors to their managers.
Funding for follow-up clinics has become part of the global budget for 4 of the 6 sites.
Identification of areas for focused developmental intervention.
Team assessments identified specific areas of developmental concerns.
Monitoring and developmental therapy interventions have become proactive and more focused to the specific needs of the child and family. Education and advocacy for developmental community supports have resulted in improved services for children and families.

Specific neurodevelopmental areas
Functional development Delays have been demonstrated [10].
Involvement of our program staff with bedside developmental care, discharge planning, early developmental intervention, and parent education.

Speech-language development
Language delay was noted after arterial-switch operation [54]. A high proportion of all survivors were noted to have delay on the vocabulary score of the Language Developmental Survey of the Child Behaviour Checklist [38] (unpublished data).
All survivors after early cardiac complex surgery are now assessed by a speech-language pathologist at the 2-year visit.
Social communication Impairments were shown after arterial-switch operation [55].
Children are monitored during the preschool period for language and social communication skills. Intervention is started as needed.
Preschool behaviour For our survivors, behavioural concerns do not exceed those in the normative population [56].
Counselling avoids suggestions of possible future adverse behaviour as this can become a self-fulfilling prophecy.
Cerebral palsy This is not a common complication among our survivors [57].
Counselling at the time of surgery based on imaging suggests monitoring without predicting cerebral palsy.

Gross motor abilities
Delays have been demonstrated [58]. Clinic plan to add a standardized motor measure [59].
Pulmonary complications Common following total anomalous pulmonary venous correction [60].
Consideration of additional routine referrals for pulmonary consultation.

Specific acute care areas
Mean arterial pressure after re-warming in the operating room Lower mean arterial pressure was associated with death [61].
Blood pressure control in the operating room has been changed.

Transfusion in neonates undergoing the Norwood operation
Transfusion did not improve outcomes [62]. Transfusion practices have been altered.
Extracorporeal life support Time for lactate to fall and inotrope score are associated with outcome [63].
There is increased focus on the optimal cannula size and initial blood flow rates when patients are placed on extracorporeal life support.  found several potentially modifiable acute care variables with the potential to improve outcomes of these neonates with each cardiac disease (Table 5). In addition, we have been able to examine specific potentially modifiable variables that have been of great interest in the literature, including blood pressure, transfusions, lactate, sedation, and CPR, for their association with outcome (Table 6). These findings have led to improved counselling of parents, improved long-term care, and changes in acute care that have great potential to improve outcomes (Table 3). Continued followup informs whether these outcomes are improving over time; for example, we found that the outcome of HLHS children has improved significantly between two surgical eras [58].

Discussion
We have chosen to present the organization of the Registry and Follow-up of Complex Pediatric Therapies Program to demonstrate a mechanism for longitudinal follow-up. Using the model of neonatal follow-up clinics, we have established a longitudinal follow-up program in western Canada for children receiving new and emerging therapies in Alberta and particularly for those receiving complex cardiac surgery as very young children at the Stollery Children's Hospital in Edmonton. By linking with the neonatal follow-up clinics, we have been able to use their expertise to maintain a high proportion of follow-up and establish an ongoing and interactive database. This database has been valuable to establish quality improvement studies and outcomes research as has been recommended [70]. Several areas of change have been initiated as a result of the findings of our follow-up program (detailed in Table 3). Examples in general areas include improved timely access to developmental intervention, enhanced funding for demonstrably needed multidisciplinary assessment clinics, and improved proactive advocacy for these children. Examples in specific areas include: improved early intervention and parent education regarding impaired functional development, standard speech-language pathologist assessment, and monitoring for language and social communication skills. Examples in acute care areas include: improved blood pressure control in the operating room, altered transfusion practices in single ventricle neonates, and improved management of neonates requiring extracorporeal life support. Ongoing follow-up will identify if these quality improvement interventions have been associated with improved outcomes in these neonates. In addition, the data from our follow-up group is now being used to inform design of prospective randomized trials of interventions.
We believe the key to the work we have done is based on the interaction between service for the child and family and the study of outcomes. This provides an opportunity to enhance the quality of care that we, as healthcare professionals, provide and enables us to minimize secondary developmental disabilities.
This longitudinal follow-up program is now established as part of the care of new and emerging therapies in Alberta and for children transferred to this province for this care. We are very grateful to the hospitals, clinics, and parents who cooperate to make this possible. Our attrition rate is low, under 4% over 2 years, and similar to other follow-up programs in western Canada. We have documented outcomes to aid in the counselling of parents of newly referred children needing these therapies. Some novel findings have been reported. Potentially modifiable variables have been identified that can influence acute care practices. Most importantly, we are providing a needed service to the children following these life-saving therapies.