Cerebral palsy (CP) is the leading cause of childhood-onset, lifelong physical disability in the Western world.Reference Hirtz, Thurman, Gwinn-Hardy, Mohamed, Chaudhuri and Zalutsky 1 CP is “a group of permanent disorders of the development of movement and posture, causing activity limitation, that are attributed to non-progressive disturbances that occurred in the developing fetal or infant brain. The motor disorders of cerebral palsy are often accompanied by disturbances of sensation, perception, cognition, communication, and behaviour; by epilepsy, and by secondary musculoskeletal problems.”Reference Rosenbaum, Paneth, Leviton, Goldstein and Bax 2 , Reference Smithers-Sheedy, Badawi and Blair 3 Motor sub-types of CP include spastic (bilateral, unilateral (hemiplegia, monoplegia)), dyskinetic (dystonic, choreoathetotic), ataxic, and mixed.Reference Smithers-Sheedy, Badawi and Blair 3 - Reference McIntyre, Morgan, Walker and Novak 5 Twenty-four months is accepted as the maximum age that an acquired injury to the developing brain may be considered as CP.Reference Shevell, Dagenais and Oskoui 6
A systematic review and meta-analysis of the median prevalence estimate of CP for birth years 1985 through 2004, with studies published by 2006, gives a pooled rate of 2.11 per 1000 live births (95% confidence interval [CI], 1.98-2.25). Reference Oskoui, Coutinho, Dykaman, Jette and Pringsheim 7 Rates in Western countries rose from the birth years of the 1960s to peak in the 1990s and dropped in the 2000s,Reference Odding, Roebroeck and Stam 8 - Reference Sellier, Platt, Andersen, Krageloh-Mann, De La Cruz and Cans 12 with a decrease in diplegia and an increase in hemiplegia over this timeReference Odding, Roebroeck and Stam 8 , Reference Ravn, Flachs and Uldall 9 ; risk factors included low-birth-weight survival and multiple birth.Reference Odding, Roebroeck and Stam 8 The increase in the mid-1990s prevalence of CP associated with an increase in survival of extremely preterm infants was followed by a stabilization and decline in this at-risk group.Reference Reid, Meehan, McIntyre, Goldsmith, Badawi and Reddihough 10 - Reference Robertson, Watt and Yasui 13 It is not known if the overall prevalence in northern Alberta has changed since the 1985 through 1988 birth years when the rate from administrative databases was 2.57 (95% CI, 2.41-2.79) per 1000 living children at age 8 years.Reference Robertson, Svenson and Joffres 14
In general, there are two approaches determining the prevalence of CP, administrative databases and population-based registries (e.g. data from patient registries), with no significant differences in rates as yet determined between the two methods.Reference Oskoui, Joseph, Dagenais and Shevell 15 The Canadian Cerebral Palsy Registry is a population-based registry, expanded from the Registre de la paralysie cerebrale du Quebec, as part of the NeuroDevNet National Networks of Centres of Excellence CP Demonstration Project.Reference Oskoui, Coutinho, Dykaman, Jette and Pringsheim 7 , Reference Oskoui, Joseph, Dagenais and Shevell 15 The first report from The Canadian CP Registry shows an overall rate per 1000 children alive at 9-10 years of 1.84 (95% CI, 1.6-2.08) using the frequentist approach. This study from northern Alberta gives the second report from the Canadian CP Cerebral Palsy Registry.
The primary study objective is to determine the overall population-based prevalence estimate of CP among 5-year-old children in northern Alberta for the birth years 2008 through 2010. Secondary objectives are to provide congenital and postneonatal CP rates and frequency among multiple births and by gestational age– and birth weight–specific subgroups and to describe the motor subtypes and function of these children at age 5 years.
Methods
The Study Region
Northern Alberta comprises about two-thirds of the geographic region of the Province of Alberta (661,185 km2) and for 2008 through 2010 had 49% of 151,405 live births. 16 , 17 Of live births, infant deaths occurred in 0.62%, preterm births in 8.7%, and multiple births in 3.5%. Alberta has universal-access government-funded health care with two regional perinatal/neonatal programs (north and south) for at-risk mothers and ill newborns. In northern Alberta, one centrally located rehabilitation center, the Glenrose Rehabilitation Hospital, a tertiary care center with four associated outreach clinics, assesses and treats almost all children with CP.Reference Robertson, Svenson and Joffres 14
Subjects
The study population includes 173 children (Figure 1), with a confirmed diagnosis of CP at age 5 years who reside in northern Alberta, were born between January 2008 and December 2010, and are registered for care at the Glenrose Rehabilitation Hospital. Children with suspect CP are referred from the community, other hospitals, and the Neonatal Follow-up Clinic. CP diagnoses are made by specialists in pediatric rehabilitation and child neurology. Motor subtype classifications follow standard practice.Reference Smithers-Sheedy, Badawi and Blair 3 - Reference McIntyre, Morgan, Walker and Novak 5 Exclusion criteria include any child whose home is not northern Alberta even if they receive Glenrose rehabilitation services and any child without a diagnosis of CP at age 5 years. Although children with CP from southern Alberta are now part of the Canadian CP Registry, they were added later and are not part of this study.Reference Shevell, Dagenais and Oskoui 6
Figure 1 Flow sheet describing 173 children with cerebral palsy, born 2008-2010.
Registration, Case Ascertainment, and Data Extraction
As part of the Canadian CP Registry, participants are ascertained after their second birthday and diagnosis confirmed after their fifth birthday. At the time of initial registration, data were collected from maternal medical and obstetrical records and the child’s neonatal, medical, and rehabilitation charts. Data collection forms were initially created for the Quebec CP Registry through a Delphi consensus process using a panel of provincial, national, and international experts to determine definitions, motor subtype classifications, and maternal and child-related variables to be collected.Reference Shevell, Dagenais and Oskoui 6 Subjects’ health records were reviewed again at 5 years of age for confirmatory evidence of the diagnosis of CP and functional classification. If confirmation of diagnosis was not available, the subject was evaluated by a child neurologist or pediatric rehabilitation specialist and the confirmed CP diagnosis recorded.
Ethics
This study was approved by the Health Research Ethics Board, Health Panel, University of Alberta, Edmonton, Alberta, with waiver of parental/guardian consent.
Measures
Population-Based Prevalence Estimate Study
Case Ascertainment
Through a long-standing referral system, diagnosis and care for children with CP in northern Alberta is regionalized. Case ascertainment is considered to be high, based in part, upon our previous documentation in the administration database studyReference Robertson, Svenson and Joffres 14 showing all but 1.1% of children with CP in this region were registered at the Glenrose Rehabilitation Hospital.
Prevalence
For the years 2008 to 2010, overall and yearly prevalence estimates of CP for 5-year-old children per 1000 children of all same age living in the same region regardless of place of birth were established as well as the rates per 1000 live births of congenital and postneonatal CP, frequency per among multiples, and gestational age– and birth weight–specific groups.
Denominator
A live birth is defined as “a complete expulsion or extraction from the mother, irrespective of the duration of the pregnancy, of a fetus in which, after expulsion or extraction, there is breathing, beating of the heart, pulsation of the umbilical cord or definite movement of a voluntary muscle.” 17 The number of live births in northern Alberta from 2008 to 2010 was 74,173, corresponding to regional referral patterns from Capital, East Central, Aspen, Peace Country, Northern Lights, and one-third of the David Thomson regions. 16 , 17 Born between 2008 and 2010, there were 77,779 5-year-old children living between 2013 and 2015 in northern Alberta, exceeding live births because of immigration numbers.
Numerators
The overall numerator for the birth years 2008 through 2010 is the sum of all the individually identified 5-year-old children living in northern Alberta with confirmed CP, whether born in or outside of northern Alberta. This was further defined by timing of possible cause of CP as congenital or postneonatal and by gestational-age and birth-weight groups.
Descriptive Variables
Variables include hospital of birth; year of birth; estimated time of occurrence of CP, fetal, and ≤28 days of life (congenital) or postneonatal (with the cause of the final diagnosis of CP happening after 28 days and before 2 years of life); sex; maternal age; type of pregnancy, single or multiple birth; type of delivery, vaginal or caesarean section; birth weight (grams); and gestational age (completed weeks).
CP motor subtype and topography 4 , Reference McIntyre, Morgan, Walker and Novak 5 as well as motor function classification is reported. The definitions of the motor subtypes of CP used in this study include: (1) spastic diplegia, predominantly involving the two lower extremities where there may be some upper extremity involvement but any upper limb tone abnormality is less than in the lower limbs; (2) spastic triplegia, predominately involving three extremities, including both lower extremities and one upper extremity, or leg-dominant spasticity, or both upper extremities and one lower extremity, or arm-dominant spasticity; (3) spastic quadriplegia, involving all four extremities and including double hemiplegia in which the involvement of the arms is equal or more than of the legs; (4) spastic hemiplegia, which refers to only one hemibody part, including involvement of only one limb (monoplegia); (5) dyskinetic, which includes any combination of dystonia or involuntary (choreoathetotic) movement; and (6) ataxia, which includes loss of muscular coordination where movements have abnormal force, rhythm, and accuracy. 4 , Reference McIntyre, Morgan, Walker and Novak 5 Functional mobility is described according to The Gross Motor Function Classification System (GMFCS)Reference Palisano, Rosenbaum, Walter, Russell, Wood and Galuppi 18 and typical performance of hand use is described according to the Manual Ability Classification System (MACS).Reference Eliasson, Krumlinde-Sundholm and Rosblad 19
Statistical Analysis
Continuous variables are presented as means (standard deviation) and categorical variables as counts and percentages. The prevalence estimates of CP based on frequentist approachReference Oskoui, Joseph, Dagenais and Shevell 15 is given per 1000 children of the same age and per cohort-specific live birth using 95% CIs. Deidentified data were collected and managed using REDCapReference Harris, Taylor, Thielke, Payne, Gonzalez and Conde 20 electronic data capture tools hosted and supported by the Women and Children’s Health Research Institute at the University of Alberta and transferred to SPSS, version 21, for presentation and analysis.
Results
Prevalence Estimates of CP
For the birth years 2008 through 2010, there were 173 5-year-old children living in northern Alberta with confirmed CP diagnoses giving an overall rate for the 3-year period of 2.22 (95% CI, 2.12-2.32) per 1000 living children age 5 years. Table 1 gives the characteristics of 169 of the 173 children, but does not include four term-born children (one born in 2008, three in 2010) with confirmed CP but missing descriptive characteristics. Of the 169, 57% are boys, 15% were one of a multiple birth, 39% were <2500 grams at birth, and 38% were born prematurely. Two other children born in the same period and receiving rehabilitation services at Glenrose Rehabilitation Hospital are not included in these rates. One child had spastic right CP at 2 years of age, but the diagnosis was not confirmed at 5 years; another was diagnosed with a degenerative neurological disease. There were no known deaths from the time of registration until reporting at 5 years.
Table 1 Characteristics of 5-year-old children with CP born in 2008-2010 and living in northern Alberta in relation to location of birth
*Excludes four children born at term with confirmed CP with missing descriptive data.
Yearly rates and cohort-specific rates are found in Table 2. Excluding nine (5.2%) children considered to have postneonatal CP causation, the prevalence for congenital CP based on 164 children at age 5 years with detailed data is 2.1 (95% CI, 2.0-2.2) per 1000 children living in northern Alberta at age 5 years. The rate for congenital CP per 1000 live births is 1.99 (95% CI, 1.89-2.09). For those born as a twin or triplet, the multiple birth–specific rate is 8.99 (95% CI, 7.88-10.1) per 1000 multiple live births. The rate for unilateral congenital CP is 1.0 (95% CI, 0.64-1.36) per 1000 live births. Table 2 also summarizes the CP rates for Alberta for birth years 1985 through 1988 from an administrative database.Reference Robertson, Svenson and Joffres 14
Table 2 CP in northern Alberta: birth years, 2008-2010, and in Alberta, birth years 1985-1988: presentation of prevalence estimate by different study approaches
*Data from reference Reference Robertson, Svenson and Joffres14.
NA=not available.
Postneonatal Timing of Possible Causes of CP
Nine children had postneonatal causes for CP (Tables 1 and 2). Causes determined by neurology consultation included: accidental head injury, 2; nonaccidental head injury (shaken baby syndrome), 2; congenital complications of congenital heart defect/cerebrovascular accident secondary to cardiac arrest, 2; cerebrovascular accident (perisurgical), 1; postconvulsion anoxia, 1; and infection (meningitis), 1. Of these nine, five had direct neonatal risk factors for CP.
Motor Type and Topography
Table 3 gives the motor subtypes for 169 children with available characteristics in relation to gestational age and birth-weight subgroups. There were no differences in the proportion of CP subtypes in relation to birth within or outside of northern Alberta. Motor subtypes includes spasticity, 148 (87.6%) of which 31 (20.9%) had diplegia, 10 (6.8%) triplegia, 33 (22.2%) quadriplegia, 74 (50%) hemiplegia/monoplegia; 18 were dyskinetic (10.6%); and three were ataxic (1.8%). Term births were recorded for 10 (32.3%) of those with spastic diplegic CP, 25 (58.1%) with spastic tri- or quadriplegia, 52 (70.2%) with hemiplegia, and 14 (77.7%) with dyskinetic CP. Of the nine with postneonatal CP, three had spastic quadriplegia, two were dyskinetic, and four had hemiplegic CP. Unilateral CP comprises 43.8% of the 169 with documented characteristics, 46.3% of all congenital CP, and 51.3% of congenital CP born in this region.
Table 3 Gestational age groups in relation to motor type and topography for 169 5-year-old children with confirmed CP born 2008-2010*
* Excludes 4 children born at term with confirmed CP with missing descriptive data.
† 1 child had monoplegia.
Function
Functional mobility relative to CP type is shown on Table 4. Of 169 children with confirmed CP, 107 (63.3%) can walk without need for a hand-held mobility device indoors and outdoors (n=79) or for short distances on level surfaces outdoors (n=28) (GMFCS level I/II); 17 (10.1%) use a hand-held mobility device and need assistance for ambulation for uneven terrain and stairs (GMFCS III), whereas in 45 (26.6%) self-mobility is limited, predominately using a manual wheelchair or powered mobility (GMFCS IV/V). Of those 105 with spastic diplegic or hemiplegic CP, 95 (90.5%) have mobility within level I/II.
Table 4 Functional mobility of 169 5-year-old children with confirmed CP in relation to motor type and topography: birth years 2008-2010*
*Excludes 4 children born at term with confirmed CP with missing descriptive data.
† 1 child had monoplegia.
Hand function is given in Table 4. Of 169 children with confirmed CP, 111 (65.7%) could perform activities with their hands without the need for modification or preparation, although alternative ways of performance might be needed (MACS I/II); 20 (11.8%) handled objects with difficulty and needed activities to be adapted to their needs (MACS III); and 38 (22.5%) required continuous support or total assistance (MACS IV/V). Of the 105 with spastic diplegic or hemiplegic CP, 99 (94.2%) function within MACS level I/II.
Of the nine children with postneonatal CP, five function at level V for both the GMFCS and MACS.
Discussion
Because of differences in measurements of the numerator and denominator, the age at diagnosis, the inclusion of those with postneonatal causes, statistical approaches, and possible death before diagnosis, comparisons of the prevalence of CP remain difficult.Reference McIntyre, Morgan, Walker and Novak 5 , Reference Blair and Watson 21 For northern Alberta, we have shown an overall population-based prevalence estimate of CP, based on confirmed diagnoses after the children reached 5 years of age, with the denominator per 1000 same-age children from the same period to be within the usually quoted prevalence rate of 2 to 2.5 per 1000 live births. Our rate is similar to that reported in the recently determined pooled rateReference Oskoui, Coutinho, Dykaman, Jette and Pringsheim 7 and by the Australian Cerebral Palsy Register.Reference Smithers-Sheedy, McIntyre and Gibson 11 Our study has a number of similarities with the Australian report, including rate of postneonatally determined CP and proportion of boys, preterm births, and plurality. Reference Smithers-Sheedy, McIntyre and Gibson 11
Although comparison of rates remains problematic, Oskoui et al,Reference Oskoui, Coutinho, Dykaman, Jette and Pringsheim 7 in a large systematic review, has shown no significant difference in overall prevalence between population-based data and administrative data.Reference Oskoui, Coutinho, Dykaman, Jette and Pringsheim 7 Our presentation of the current results in Table 2, beside our administrative database results of more than 20 years ago, show no overlap in the 95% CIs for the studies. This suggests improvement at least in the rates of CP among the low-birth-weight children; for these children, we are confident in complete case ascertainment because of the complete follow-up by the Neonatal Follow-up Program in northern Alberta.Reference Robertson, Watt and Yasui 13 This study also suggests a trend toward lower rates than the British Columbia study using administrative data.Reference Smith, Kelly, Prkachin and Voaklander 27 Our prior linkage study used administrative diagnostic codes that may not be comparable; on the other hand, the denominator for both studies included children of the same age as those with diagnosed CP, rather than live births. The prevalence of this present study may be an underestimate. The final year of this study shows a lower rate, 1.89 (95% CI, 1.73-2.05), suggesting possible incomplete ascertainment relative to the first 2 years. We have no explanation for this apparent improvement other than year-to-year variability.
Our study supports the current reported change in the idea of an inverse relationship between gestational age and CP rates with proportionately fewer extremely low-birth-weight children diagnosed with CP. 4 , Reference Odding, Roebroeck and Stam 8 , Reference Reid, Meehan, McIntyre, Goldsmith, Badawi and Reddihough 10 - Reference Sellier, Platt, Andersen, Krageloh-Mann, De La Cruz and Cans 12 , Reference Himpens, Van den Broeck, Oostra, Calders and Vanhaesebrouck 22 Similar to others,Reference Oskoui, Coutinho, Dykaman, Jette and Pringsheim 7 , Reference Reid, Meehan, McIntyre, Goldsmith, Badawi and Reddihough 10 we have shown a rate of CP within the extremely preterm children that is lower than those of 28 to 31 weeks’ gestation. This reduction is unlikely the result of chance.Reference Sellier, Platt, Andersen, Krageloh-Mann, De La Cruz and Cans 12 The birth weight–specific rates presented here are lower than those we gave for births in 1985 through 1988,Reference Robertson, Svenson and Joffres 14 but higher than we reported for children of 20 to 27 weeks’ gestation born between 2001 and 2003.Reference Robertson, Watt and Yasui 13 Variability in the recording of live births for the lowest gestational age and birth weight subgroupsReference Oskoui, Joseph, Dagenais and Shevell 15 presents difficulty for the comparison of gestation- and birth weight–specific prevalence rates.
Postneonatal CP rates are lower in developed countries, decreasing over the past three decades.Reference Blair and Watson 21 , Reference Germany, Ehlinger and Klapouszczak 23 We report these rates as lower than our previous report,Reference Robertson, Svenson and Joffres 14 similar to the mean European rate, but higher than other current rates.Reference Reid, Meehan, McIntyre, Goldsmith, Badawi and Reddihough 10 , Reference Germany, Ehlinger and Klapouszczak 23
Heterogeneity of reported motor CP subtypes has resulted in calls to make this classification more reliable.Reference Reid, Carlin and Reddihough 24 Compared with the Victorian study from birth years 1970 through 2003,Reference Reid, Carlin and Reddihough 24 our study shows the proportion of children with dyskinesia to be more than double, with lower rates of ataxia and slightly lower rates of spasticity. For some time, there have been concerns about an increasing rate of unilateral CP.Reference Odding, Roebroeck and Stam 8 , Reference Ravn, Flachs and Uldall 9 Although this increase has not been linear,Reference Sellier, Platt, Andersen, Krageloh-Mann, De La Cruz and Cans 12 the proportion of unilateral spastic CP in relation to all subtypes is increasing.Reference Sellier, Platt, Andersen, Krageloh-Mann, De La Cruz and Cans 12 We are reporting one of the highest rates of unilateral spastic CP, which makes up half of all spastic CP in this study. Similar findings of other investigators are leading to increased investigation of perinatal stroke.Reference Golomb, Garg, Saha, Azzouz and Williams 25
Neurological CP subtyping has been found to be a strong predictor of functional status related to ambulation.Reference Shevell, Dagenais and Hall 26 Compared with the Quebec study,Reference Shevell, Dagenais and Hall 26 fewer children in our study have GMFCS levels of IV/V. Differences in GMFCS levels for similar subtypes may be explained in part by other comorbidities not addressed by this report.Reference Shevell, Dagenais and Hall 26 We have shown that 65.7% of all children and 94.3% of those with spastic diplegia and hemiplegia/monoplegia have MACS levels of I/II.
A strength of this study is case ascertainment, with a confirmed diagnosis for children older than age 5 years giving high true positives. However, because complete case ascertainment is critical to the numerator, this is also a limitation of the study as only those referred to the Glenrose Rehabilitation Hospital and its satellites are included. Previous documentation reported that all but 1.1% of children with CP for this region were assessed at GlenroseReference Robertson, Watt and Yasui 13 ; however, our results may still underestimate the true rates in this geographically large area. Although this study uses the methodology for one of the two approaches to the study of CP prevalence estimates recently outlined,Reference Oskoui, Joseph, Dagenais and Shevell 15 case ascertainment would be stronger if multiple sources had been used rather than information from a single tertiary center and associated satellites. This region’s population is smaller with fewer cases relative to larger registries,Reference Reid, Meehan, McIntyre, Goldsmith, Badawi and Reddihough 10 - Reference Sellier, Platt, Andersen, Krageloh-Mann, De La Cruz and Cans 12 reducing the generalizability of results, particularly for subgroups. A strength is the denominator of the number of children of the same age as the children with CP, allowing for the most appropriate denominator for the overall prevalence estimate. A limitation is the denominator of live births; using the current definition, the number of live births is larger than the previous definition that included only live births with gestational ages ≥20 weeks and birth weights ≥500 grams. The present denominator is larger and hence the prevalence rates are slightly lower than if previous cutoff levels were used. Similarly, denominators based upon neonatal survivors, if available, would give somewhat higher prevalence rates and assist with comparison with other studies. Limitations of this study also include having no information on deceased children with likely CP before age 2 years of age and omission of children leaving the area. We have not addressed comorbidities. Comparison with current administrative data from the same years would be an asset.
Conclusion
The population-based prevalence estimate of CP in northern Alberta is within the current levels for the Western world and suggests a reduction in congenital CP prevalence relative to our previous study from the 1980s birth years, especially for the low-birth-weight children. Rates for unilateral spastic CP make up the most common subtype. The gestational age–specific rate for those with CP of <28 weeks is similar to that of those of 28 through 31 weeks. The results provide needed regional information.
Acknowledgements and Funding
We thank the Canadian Cerebral Palsy Registry for funding, direction, and outline for data collection that has made this study possible. We thank the referring doctors of Northern Alberta for their continued early referral of young children with suspect cerebral palsy to pediatric rehabilitation at the Glenrose Rehabilitation Hospital and for the devoted disciplines at Glenrose that care for these children and make these data available for study.
The Canadian Cerebral Palsy Registry has been funded by the NeuroDevNet Networks of Centres of Excellence, Ottawa, Ontario, Canada, and the generous support of the Stollery Children’s Hospital Foundation through the Women and Children’s Health Research Institute and the Glenrose Rehabilitation Hospital Foundation, Edmonton, Alberta, Canada. These funding agencies have not been part of the design, data collection, data analysis, manuscript preparation, or publications decisions for this manuscript.
Disclosures
MO has the following disclosures: funding recipient, Kids Brain Health Network (Networks of Centres of Excellence), Ottawa, ON, Canada; consultant and consulting fees: Biogen. JCA has the following disclosures: principal investigator, research grant, Kids Brain Health Network (NeuroDevNet) NCE; Women’s and Children Health Research Institute; funding recipient, Glenrose Rehabilitation Hospital Foundation, Edmonton, Alberta, Canada. CMTR, MFR, KO, HG, and JYY do not have anything to disclose.
Statement of Authorship
CMTR completed a literature search; wrote and revised the manuscript; participated in the concept, design, writing the main document, revising, and approval of the final version; and had access to the original database. MFR prepared a report for one of the funding agencies, the Women and Children’s Health Research Institute, University of Alberta, Edmonton, Alberta, Canada, outlining the progress of the study; she also submitted an abstract to the NeuroDevNet fall meeting; and had access to the data and participated in the interpretation of analysis and the revision and final versions of the manuscript. KO collected and entered the data to REDCap and hence had access to all data and was part of the original design of the study, revising and approving of the final version. MO represents the Canadian Cerebral Palsy Registry on this manuscript and she revised and approved the final version. HG was part of the original study design, revising and approval of the final manuscript. JYY was part of the original study design, reading and approving the manuscript. JCC completed all aspects of the study including concept/design, supervision of collection of outcomes data, interpretation of analysis, and revisions of the original and final versions of the manuscript; he represents the Canadian Cerebral Palsy Registry on this article.
Following is the list of site primary investigators of the Canadian Cerebral Palsy Registry.
Michael I. Shevell, MD, and Maryam Oskoui, MD, MSc (Department of Neurology/Neurosurgery, McGill University; Division of Pediatric Neurology, McGill University Health Centre; Department of Pediatrics, McGill University); John Andersen, MD (Developmental Pediatrics, University of Alberta; Glenrose Rehabilitation Hospital); David Buckley, MD (Department of Pediatrics, Memorial University); Darcy Fehlings, MD, MSc (Institute of Health Policy, Management and Evaluation, University of Toronto; Dalla Lana School of Public Health, University of Toronto; Department of Pediatrics, University of Toronto; Division of Developmental Pediatrics, Holland Bloorview Kids Rehabilitation Hospital and Bloorview Research Institute); Adam Kirton, MD, MSc (Departments of Pediatrics and Clinical Neurosciences, University of Calgary; Alberta Children’s Hospital Research Institute); Alison J. Moore, MD, MSc (Department of Pediatrics, University of Calgary; Developmental Pediatrics and Neurology Alberta Children’s Hospital); Nicole Pigeon, MD (Department of Pediatrics, Centre hospitalier universitaire de Sherbrooke; Centre de réadaptation Estrie); Louise Koclas, MD (Division of Neonatology, McGill University Health Centre; Centre de Réadaptation Marie Enfant, CHU Sainte Justine); Esias van Rensburg, MD (Department of Pediatrics, University of British Columbia; Neuromotor Program and Division of Developmental Pediatrics, Sunny Hill Health Centre for Children); Ellen Wood, MD MSc (Faculty of Medicine, Dalhousie University; Department of Pediatrics, IWK Health Centre).
