Cerebral Palsy—Don't Delay

DEVELOPMENTAL DISABILITIES RESEARCH REVIEWS 17:114–129 (2011)

CEREBRAL PALSY—DON’T DELAY

Sarah McIntyre,1,2,3* Cathy Morgan,1,3 Karen Walker,2,4 and Iona Novak1,3 1Cerebral Palsy Alliance, Research Institute, New South Wales, Australia 2The University of Sydney, School of Paediatrics and Child Health, New South Wales, Sydney, Australia 3The University of Notre Dame, School of Medicine, New South Wales, Sydney, Australia 4Grace Center for Newborn Care, The Children’s Hospital at Westmead, New South Wales, Australia

Cerebral palsy (CP) is the most severe physical disability within the 5. Present an evidence based algorithmic approach to recogniz- spectrum of developmental delay. CP is an umbrella term describing a ing CP and developing intervention plans. group of motor disorders, accompanied by many associated impairments. The disability is a result of injuries to the developing brain occurring any time from the first trimester of pregnancy through to early childhood. In the early months of life, global developmental delay However, for the great majority, their full etiological causal pathway and CP present similarly, if delayed, acquisition of develop- remains unclear. It is important to discriminate as early as possible mental milestones is the only comparator. It is the movement between: (a) mild or nonspecific motor delay, (b) developmental coordination disorder, (c) syndromes, (d) metabolic and progressive conditions, disorders (e.g., spasticity and dystonia), the level of functional and (e) CP with its various motor types and distributions. The most prom- impairment, and the associated impairments that set CP apart ising predictive tool for CP is the general movements assessment, which from other milder motor disorders or learning disorders such assesses the quality of spontaneous movements of infants in the first 4 as developmental coordination disorder (DCD). DCD is less months of life. We propose a change in diagnostic practice. We recommend a shift away from referral for intervention following a formal (most severe and 25 times more common than CP affecting 5–6% often late) description of CP, to one of referral for intervention the population and current practice is not to diagnose before which occurs immediately once an infant is considered “at risk” of CP. the age of 5. As a result, the diagnosis of CP is often delayed VC 2013 Wiley Periodicals, Inc. Dev Disabil Res Rev 2011;17:114–129. while the possibility of DCD is explored. DCD is primarily a learning problem where children Key words: cerebral palsy; early diagnosis; general movements; can achieve normal movement patterns and skills but have perinatal risk factors; neonatal risk factors; brain injury problems with learning and planning the movements. CP conversely is a physical disorder, where children are not able to achieve the normal movement patterns and the primary problem is motoric not learning, although deficits in learning may compound the motor problem. INTRODUCTION DCD is used to refer to children who fulfill a certain lobal developmental delay is an umbrella term that criteria; poor motor performance which significantly interferes describes two or more delays in the area of speech with activities of daily living which are not explained by any Gand language, social and emotional, cognitive and medical, neurological, or psychosocial condition. Thus a child motor development. Children with cerebral palsy (CP) often with CP whose motor disability is neurological cannot have a fall under the umbrella of global developmental delay, but CP diagnosis of DCD [Blank et al., 2011]. The physical disability cannot be considered “delay,” as children do not “grow out of CP is life-long whilst DCD is more apparent in the win- of it.” Health professionals need to understand what clinical dow where the child is learning key motor skills for example, features distinguish CP from other motor disorders, so the catching a ball, dressing independently, and handwriting. most effective interventions can be commenced earlier. The American Academy of Pediatrics have developed a policy for WHAT IS CEREBRAL PALSY? the surveillance and screening of developmental disorders CP is an umbrella term which “describes a group of dis- (Council on Children with disabilities et al., 2006), however orders of the development of movement and posture, causing this paper focusses specifically on CP. The objectives of this activity limitations, which are attributed to nonprogressive dis- review are fivefold: turbances that occurred in the developing fetal or infant brain.

1. Describe the nature of CP and what makes it different to other motor or learning disorders. *Correspondence to: Sarah McIntyre, Cerebral Palsy Alliance, The University of Sydney, The University of Notre Dame, Australia. E-mail: smcintyre@cerebralpalsy. 2. Outline the prevalence of CP. org.au 3. Determine who is at high risk of CP, what are the predictors Received 3 September 2012; accepted 5 October 2012 View this article online in Wiley Online Library (wileyonlinelibrary.com) and early signs? DOI: 10.1002/ddrr.1106 4. Identify tools that help clinicians to accurately predict CP. ' 2013 Wiley Periodicals, Inc. and contracture management. The ASAS Table 1. Classiﬁcation by Motor Type is currently undergoing further reliability

a studies, but it is freely available for ACPR 1 use along with the description of CP Reid, 2011a form: http://www.kemh.health.wa.gov.au/ Spasticty: Overactive muscles that display a velocity-dependent 85 – 91% services/register_developmental_anomalies/ resistance to stretch. Spasticity can cause secondary documents/CP%20Description%20Form% impairments such as loss of muscle length, joint 20-%20WARDA%20website.pdf. dislocation and pain. Dyskinesia: Dyskinesia is either athetosis or dystonia. Athetoid CP is 4–7% The gold standard tool for reliably hypotonic with hyperkinesia characterized by involun- describing motor function in CP is the tary writhing-stormy movement and can gross motor function classification sys- co-occur with chorea. In contrast, dystonic CP is tem (GMFCS) [Palisano et al., 1997]. hypokinetic, involving involuntary, abnormal twisting GMFCS provides a common language postures or repetitive movements with hypertonia. Tone is typically fluctuating. that conjures up a “picture” of a child Ataxia: Ataxia results in tremors with a shaky quality. Ataxic 4–6% with CP. GMFCS is a five level classifi- CP involves a loss of muscular coordination where cation system of gross motor function movements have abnormal force, rhythm, and accuracy. in people with CP. The classification is Hypotonia: Pure, generalized hypotonia (decreased muscle tone) is the 2% least common CP motor-type. Some argue that pure based on the person’s ability to self ini- hypotonia should not even be considered a cerebral palsy tiate movement with a focus on sitting, sub-type. transferring, and mobilizing [Palisano et al., 1997]. Different classification

aAustralian Cerebral Palsy Register. descriptions exist at different age groups. Table 3 summarizes the system for 2–4-year olds, to coincide with the The motor disorders of CP are often when identifying topographical subtype, most common time of recognition and accompanied by disturbances of sensa- the Surveillance of Cerebral Palsy Europe the proportion in a CP population with tion, cognition, communication, [SCPE, 2000] has recommended that tra- each level of GMFCS. perception, and/or behavior, and/or by ditional topographies be combined into It should be noted that whilst the a seizure disorder” [Bax et al., 2005]. two easily definable topographies: Unilat- GMFCS classification can be applied to This most recent definition acknowl- eral (one side of the body), Bilateral (both infants, about 40% change classification edges the complexity of the condition sides of the body). The ACPR instead levels by age 2. After 2 years, the classifi- and the impact of the associated applies a limb by limb coding using the cation system is stable and thus GMFCS impairments. Australian Spasticity Assessment Scale reassessment is recommended after age 2 (ASAS) [Love, 2007]. The ASAS scores [Gorter et al., 2008]. This is clinically the muscles’ response to rapid passive and diagnostically very important, What are the Fundamental Facts movement without the subjectivity and because parents are anxious to learn early We Know About Cerebral Palsy? wording ambiguities of the modified Tar- about the severity of their child’s condi- dieu and Ashworth scales [Mutlu et al., tion for future planning but in reality the Classification of cerebral palsy guides inter- 2008]. Nonspastic motor types are also most accurate description of function and vention decision making coded, resulting in a “stick figure dia- severity can only be given at 2 years. CP is a heterogeneous condition, gram” of motor impairment, which The presence of associated impairments and and to elucidate prognosis and guide provides an objective picture of the CP. selection of the most appropriate inter- functional limitations affects the child’s Figure 1 presents the CP description outcome ventions (e.g., constraint induced form. The descriptive form is also clini- movement therapy for hemiplegia and For many children with CP, it is cally useful for treatment decision- not just a physical disability. When selective dorsal rhizotomy for diplegia) making, such as pharmacological options three major classifications are applied; seeking to prognosticate the severity of motor-type, topography, and function. Clinicians often remark that a child may have two or three different descriptions Table 2. Classification by Topography of their CP within one medical file, evi- dencing the poor reliability of these ACPRa traditional classification systems. Tables 1 and 2 outline the traditional motor Hemiplegia: Hemiplegia/monoplegia is the involvement of one side of the 38% body. The upper limb is usually more affected than the types and topographies of CP and the lower limb. Strong early hand preference or hand disregard proportions of a CP population with is sometimes the first sign of a problem. each type. In this paper, we refer to the Diplegia: Diplegia is where both the legs are affected and are more 36% Australian Cerebral Palsy Register affected than the upper limbs. (ACPR) when reporting rates and for Quadriplegia Quadriplegia refers to the presence of spasticity in all four 26% (Tetraplegia) limbs; where the affect on the arms is equal or more than international comparisons the Swedish the legs. Trunk and oro-facial involvement is also to be Register and a study by Reid et al. expected. In rare cases, one limb is spared and this is [2011a] where registers throughout the referred to as triplegia. world are compared. To solve the problem of low inter- aAustralian Cerebral Palsy Register. rater (and sometimes intra-rater) reliability

DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL 115 technology systems such as speech Table 3. Classiﬁcation by Gross Motor Function at 2-4 Years generating devices, may be used to

a communicate. It is a fundamental ACPR human right to have the opportunity to Level I: Floor sits independently, hands-free. Walks without 32% communicate; however, high technol- assistive devices. ogy AAC systems are expensive, Level II: Floor sits independently, hands-free with balance 27% requiring wait listing and for some indi- affected. Walks using an assistive mobility device. viduals will mean that they are unable Level III: Floor sits using w-sitting. Walks short distances indoors 12% using a hand-held mobility device with assistance. to access systems that would support Level IV: Floor sits when placed, uses hands for balance. Rolls, 14% them to communicate. creeps or crawls for short distances. Vision. Vision impairments can range Level V: Unable to sit independently. No form of independent 15% from mild requiring glasses, to func- mobility. tionally blind. About 5–12% of individuals with CP have a severe aProportion in Australia with each level of GMFCS. impairment, or are functionally blind [McManus et al., 2006; ACPR Group, 2009]. Another 30% will have a mild to moderate vision impairment. CP and determine intervention plans, at term (48%) compared with preterm Hearing. Hearing impairments can also assessment of associated impairments (28%); bilateral CP (34–87%) compared range from a mild impairment to bilat- must also occur. The likelihood and se- with unilateral (23%); and those with eral deafness. Bilateral deafness occurs verity of associated impairments increase intellectual impairment (61%) compared in 2% of people with CP while other with the severity of motor impairment with no intellectual impairment (19%) hearing impairments occur in a further [Himmelmann et al., 2006; Odding [Carlsson et al., 2003; Wichers et al., 10% [Surman et al., 2006; ACPR et al., 2006]. Some have reported that 2005; Himmelmann et al., 2006]. Group, 2009]. Assessment of vision and associated impairments impact more on Intellectual impairment. Intellectual hearing in children with CP should be function and quality of life than the impairment can be defined by low gen- thorough and done early, as it can motor impairment [Himmelmann and eral intellectual functioning as measured impact greatly on their ability to learn Uvebrant, 2011]. A meta-analysis of CP by IQ scores, in combination with diffi- and achieve milestones. registers calculated the overall rates of culties with adaptive behavior, all Other. Other impairments strongly associated impairments and functional manifesting before the age of 18. Practi- associated with CP are hip dislocation limitations in the CP population to be: cally, this means that people with an (8%), displacement (27–35%) [Hagglund three in four are in pain; one in two intellectual impairment have memory et al., 2005; Soo et al., 2006] and spine have an intellectual disability; one in deficits, difficulty reasoning, learning deformities, sleep disorders (23%) three cannot walk; one in three have a new skills, attending and organizing in- [Newman et al., 2006], pain (70%) hip displacement; one in four cannot formation. 50% of individuals with CP [Jahnsen et al., 2004; Arnaud et al., talk; one in four have epilepsy; one in have an intellectual impairment and 2008], eating (8% tube fed) [Shevell four have a behavior disorder; one in between 20 and 30% [Jarvis et al., 2005; et al., 2009; Sigurdardottir and Vik, four have bladder control problems; one McManus et al., 2006] have a severe in- 2011], excessive drooling (22%) [Parkes in five have a sleep disorder; one in five tellectual impairment. Formal assessment et al., 2010], bladder and bowel control dribble; 1 in 10 are blind; 1 in 15 are of intellect is essential (but at times diffi- complaints (24%) [Roijen et al., 2001], tube fed; and 1 in 25 are deaf [Novak cult) for an individual with CP. and behavior difficulties (26%) [Parkes et al., in press]. Many will have a num- Communication. Communication dis- et al., 2008]. These less well-understood ber of these impairments, and the ability can have a major impact on the impairments are more likely to occur presence of these impairments compli- individual with CP and their family. with bilateral CP and intellectual cates therapy, decreases health status and Impairment in this domain can impact impairment. quality of life for the individual and on both understanding of language and their family, and increases costs for the expression. For individuals who have CP is the most common physical disability family and to society. The associated severe communication impairment, in childhood with prevalence unchanged for impairments of CP will now be dis- social isolation and poor self-esteem can 60 years cussed briefly. result. Between 20 and 30% of people The overall prevalence of CP is Epilepsy. Epilepsy can potentially with CP are nonverbal which means 0.2% of the population (i.e., 1 in 500) severely limit the quality of life for the that systems to support other forms of in developed countries. As can be seen person with CP and their family, and communication are required [Arnaud by a projected age distribution of one adults with CP and epilepsy are less likely et al., 2008; ACPR Group, 2009; state in Australia (Fig. 2), even though to find employment [Michelsen et al., Andersen et al., 2010; Parkes et al., the injury responsible for CP occurs in 2005]. Epilepsy occurs in 30% of individ- 2010]. They are more likely to be non- the developing brain, it is a lifelong uals with CP [Arnaud et al., 2008; verbal if they are non-ambulatory condition, with most patients having a ACPR Group, 2009]. In 2% of individu- (GMFCS IV-V, 57%) compared to normal life expectancy. In reality, CP is als with CP, their epilepsy will be those who are able to walk (GMFCS I- not just a condition of childhood. resolved by the time they turn 5 years of III, 4%) [Shevell et al., 2009]. Augment- The true incidence of CP cannot age [ACPR Group, 2009]. For those ative and alternative communication be estimated as there are a proportion of whose seizures are not resolved, epilepsy (AAC) systems, which can range from infants who die in the intrapartum, neo- is a lifelong condition. Rates of epilepsy low/light technology systems such as natal and infant period, who had brain are higher in those with: spasticity born signing or use of alphabet charts to high lesions that may or may not have met

116 DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL Figure 1 Cerebral palsy description form. [Color ﬁgure can be viewed in the online issue, which is available at wileyonlinelibrary.com.].

the criteria for CP. It has been suggested live births 51 (95% CI 24–79) [Watson, birth prevalence of 2/1,000 live births. therefore that the closest rate to inci- 2012, personal communication]. If neo- In developing countries, it is thought dence (for CP) is prevalence of neonatal natal deaths are not taken into account, that incidence is higher as the public survivors (NNS). Western Australia live births give a misleading lower rate. health measures that help prevent some (WA) is one register that reports in this In term births (371 weeks), where the CP cases are not freely available in manner, and is also one of the longest rate of intrapartum/neonatal death is developing countries [Blair and Watson, running CP Registers in the world. CP proportionally much less, the difference 2006]. All data sets across the world is mandatorily reported in WA, there- between NNS 1.7 (95% CI 1.4–2.1) agree there is a higher proportion of fore it is assumed that this register has as and live births 1.7 (95% CI 1.4–2.0) boys diagnosed with CP. Although CP close to a total population cohort as is becomes inconsequential. Despite this is found across all socio-economic possible. WA’s CP rates reported in denominator being the most accurate, classes, there is a clear association 2006 are 2.78/1,000 NNS increasing to for comparison live births are the most between low birth weight and low 3.9/1,000 when post-neonatal CP is widely used denominator. socio-economic status, and in normal taken into account [Blair and Watson, Estimates of prevalence through- birth weight ranges, rates of CP are 2006; Watson et al., 2006]. NNS are out the world vary depending on the 2.42/1,000 live births for those in the important when rates are reported by methodology of “count,” percentage lowest socio-economic groups, com- gestational age stratum. The lower the ascertained and variations in selection pared to 1.29/1,000 for the most gestational age stratum, the more rates criteria. CP Registers have identiﬁed afﬂuent groups. differ between NNS and live births. It is rates ranging between 1.4 and 2.77/ The overall rate of 2/1,000 has particularly important for those at the 1,000 live births; surveillance programs been fairly stable over the last 60 years youngest gestational ages. When report- range between 2.1 and 3.6/1,000 live in contrast to the dramatic falls in peri- ing rates in the birth years 2005 and births; and cross-sectional surveys range natal mortality rates. However, there 2006 for those born between 20 and 27 between 1.05 and 4.1/1,000 live births. have been some trends in gestational weeks in WA, the rate per 1,000 NNS The two largest data sets, the ACPR age stratum, shown in Figure 3. Rates was 72 (95% CI 32–110) compared to and the SCPE both have an overall in the extremely and very low

DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL 117 Figure 2 Estimated number of people living with CP in New South Wales, Australia. [Color figure can be viewed in the online issue, which is available at wileyonlinelibrary.com.] gestational groups rose during the example, intellectual impairment, epi- The rate of CP in neonatal survi- 1980s, but are now trending down. lepsy and low socioeconomic status. vors varies significantly with level of Moderately premature infants’ rates Early pregnancy: Infection, birth defects, risk at birth. To describe the risk of have decreased slightly, while in term multiple births, male gender, and other developing CP, infants have been sepa- infants the rates are unchanged [Blair genetic predispositions. rated into three distinct groups shown et al., 2001; Watson et al., 2006]. During pregnancy: Maternal disease, for in Figure 4: (1) premature infants (30– Because the majority (>73%) of infants example, thyroid disorders, pregnancy 40% of all CP); (2) term born infants are born over 32 weeks gestational age, complications, for example, preeclampsia who shortly after birth have neonatal the increases and decreases in the encephalopathy (NE), a clinically extremely and very preterm groups and bleeds in the second and third trimester, infection and inflammation, defined syndrome of disordered neona- have made little difference to the over- tal brain function (15–20% of all CP); intrauterine growth restriction (IUGR), all rate. and (3) term born “healthy” infants, placental abnormalities and other precur- who do not require special care in the sors to premature birth. neonatal period (40–50% of all CP) and Identification of infants “at-risk of cerebral Around the time of birth and the neonatal do not appear to have identifiable risk palsy” is possible; assessment and screening period: An acute intrapartum hypoxic factors at birth [Badawi et al., 2005; should follow event, stroke, seizures, hypoglycemia, jaun- Wu et al., 2006; McIntyre et al., 2011]. Since there are no identifiable dice, and infection. Premature infants. When considering biomarkers to accurately predict CP, Postnatal period: Infections, accidental which babies are at risk of CP, preterm and clinical risk factors only identify and nonaccidental injuries, stroke both infants commonly come to mind. The subpopulations of infants at risk [McA- spontaneous and following surgery. risk of CP increases as gestational age dams and Juul, 2011], understanding the term “causal pathways” is important. CP atiologies are described in terms of causal pathways, as there is very rarely one specific cause of brain damage severe enough to cause CP. Much research has been published that attempts to discern the risk factors that lie on one or more causal pathways to CP. What researchers are beginning to realize is how little is known about how these risk factors interact on causal pathways. Risk factors can be described according to when they occur or when they are identified. The following examples have been identified for CP: Prior to conception: Previous gynecolog- ical history of stillbirths/multiple miscar- riages/neonatal death/premature birth, Figure 3 Gestational age specific rates/1,000 live births in WA, 1980–2006. [Color figure can family history of CP and other genetic be viewed in the online issue, which is available at wileyonlinelibrary.com.] predispositions, maternal diagnoses, for

118 DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL decreases, therefore babies born at 36 maintained at between 5 and 7/1,000 receive a general movements (GM) weeks’ gestation are at much lower risk live births since the early 1980s [Wat- assessment before term equivalent age, than those born at 24 weeks. As a son et al., 2006]. and be referred to active surveillance result, rates in premature infants range Cerebral lesions in particular and early intervention when they leave between 3 and 80/1,000 neonatal survi- PVL, intraventricular hemorrhage the hospital. (see Pathway A Figure 5, vors, reflecting the wide variation in (IVH) and intracranial hemorrhage to be discussed in the following levels of risk across premature gesta- (ICH) grade III and IV, are the most section). tions. Premature infants constitute up to important predictors of CP in very pre- 40% of infants who develop CP [Kirby term infants [Tran et al., 2005; Beaino Term infants with and without neonatal et al., 2011]. So why are premature et al., 2010; Himpens et al., 2010]. In encephalopathy infants at increased risk of CP, and particular, PVL lesions in the corona The overall rate of CP for term which ones are at the highest risk? radiata above the posterior limb of the infants has been consistently 1.4–1.7/ The group of preterm infants can internal capsule (PLIC) observed in cor- 1,000 live births over the past 30 years be separated according to gestational onal sections have been used to [Watson et al., 2006; Himmelmann age, with the first subgroup being accurately predict motor prognosis et al., 2010]. Multiple births born at extreme prematurity, generally consid- [Nanba et al., 2007]. The presence of term are at four times the risk of CP ered less than 28 weeks’ gestation. lesions in this region was highly predic- than singletons born at term. The risk There is much data in the literature tive of CP (GMFCS 1 or higher) with rises again for surviving twins after the which depicts the outcomes of sensitivity 100% and specificity 97%. A death of a cotwin [Pharoah, 2006]. extremely premature infants and much study by Himpens et al. [2010] that Risk factors associated with the devel- research has been conducted in this age investigated the predictive value of opment of CP in the term population group [Hoon and Faria, 2010; Reid ultrasound in brain injury found that also include congenital malformations, et al., 2011b]. In the 1970s and 1980s, deep grey matter lesions are a signifi- maternal age over 35 years, chorioam- the frequency of CP in this gestational cant predictor for severe versus mild nionitis, preeclampsia, placental age group increased. This was attributed and moderate CP (OR 5 6), and that abnormalities, meconium aspiration to the increasing survival of extremely cerebral infarction and hemorrhage syndrome, IUGR, transient metabolic preterm infants and their predilection to grade IV are strong predictors of unilat- abnormalities, respiratory distress syn- germinal matrix hemorrhage and peri- eral spastic CP versus bilateral spastic drome, neonatal infections and seizures. ventricular leukomalacia (PVL) [Stanley CP (OR 5 49 and 24, respectively, P [Shankaran, 2008; McIntyre et al., and Watson, 1992; Hagberg et al., < 0.001). 2012]. One of the most well known 1996]. Evidence from population-based Recently, there has been increas- risk factors for term-born infants is NE. samples in Europe, Australia and the ing interest in and evidence regarding The second piece of the pie (Fig. United States, and analyses from CP the possible effects of intrauterine infec- 4), with a well-recognized predilection Registers in Australia and Europe tion or inflammation early in the to develop CP are term or near term describing trends in prevalence, sub- postnatal course, leading to CP. Carlo infants with NE. For term born infants types, and severity, suggest that this rise et al. [2011] recently argued that a late with NE, the rate of CP is between in frequency of CP in extremely pre- prenatal and/or early neonatal exposure 100 and 125/1,000 neonatal survivors, term infants has reached its peak and is to inflammation may predispose infants and those born with severe NE are at now decreasing [SCPE, 2000; Reid to neurodevelopmental impairment. the highest risk of CP of all infants. et al., 2011b; Watson, 2012, personal Wu and Colford [2000] also found that Infants with moderate to severe (Sarnat communication]. Up to 10% of clinical chorioamnionitis was associated Stage 2 or 3) NE account for one in extremely preterm infants (variations in with an increase in CP in preterm four cases of term CP [Badawi et al., reports exist from as low as 3–10%) and infants (OR 5 1.9) and term infants 2005]. Kurinczuk et al. [2010] report up to 5% of infants between 28 and 31 (OR 5 4.7). an incidence of NE between 2.5 and weeks gestation will be described as Transient hypothyroxinaemia, 3.5 per 1,000 live births and that 30% having CP [Himpens et al., 2008; Wat- bronchopulmonary dysplasia (BPD), of cases in developed countries are asso- son, 2012, personal communication]. and necrotizing enterocolitis have also ciated with evidence of an acute Practice point. Mothers whose labor is been associated with premature birth imminent (and prior to 30 weeks gesta- and a later description of CP. A recent intrapartum hypoxic event. These tion) should now be offered magnesium study of 1,047 preterm infants (<28 include sentinel birth events that are sulphate for neuroprotection of their weeks) demonstrated that while all also rare but important risk factors for child. Meta analyses have shown that infants with BPD had a higher risk of CP in term infants, such as placental CP can be reduced by 30% for infants CP those who were mechanically ven- abruption, cord prolapse, severe intra- under 30 weeks gestation [Crowther tilated until 36 weeks PMA had at least partum hemorrhage, severe shoulder et al., 2002]. a fourfold increased risk of CP [Van dystocia, and a tight nuchal cord. It is CP Registers in Europe report Marter et al., 2011]. In addition, pre- estimated that up to 8% of CP is attrib- that this trend for decreasing rates con- term infants who have had surgery to utable to an acute intrapartum event tinues into the group of late preterm repair a patent ductus arteriosus, or with moderate to severe NE [Blair and infants (32–36 weeks’ gestation or who required home oxygen have also Stanley, 1997]. 1,500–2,499 g) [Andersen et al., 2011]. been identified as at increased risk of Practice point. Infants with moderate to The overall prevalence of CP in these CP [Tran et al., 2005]. severe NE following an acute intrapar- children had dropped from 12.2 per Practice point. Infants born premature tum event benefit from hypothermia. 1,000 live births in 1983 to 4.5 per are at high risk of CP if they have This intervention prevents CP in one 1,000 in 1997. There is conflicting evi- abnormal cerebral imaging and a more out of eight of those treated [Jacobs and dence in Australia, with the rate being complex course. These infants should Tarnow-Mordi, 2010]. A number of

DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL 119 not have any of these neonatal factors, yet 60% of these infants had moderate to severe CP. This is not the first time a finding like this has been reported. The National Collaborative Perinatal Project reported that most children with CP did not derive from groups at high risk (low Apgar scores, or the presence of neonatal signs). About 43% were exam- ined and classified as “neurologically normal” in the neonatal period and concluded that a large proportion of CP cases remain unexplained [Nelson and Ellenberg, 1986; Ellenberg and Figure 4 Rate of CP in NeoNatal Survivors. [Color figure can be viewed in the online issue, Nelson, 1988]. Earlier still, in 1970, which is available at wileyonlinelibrary.com.] Eva Alberman attempted to model what were at that time the three most important risks around birth: (1) parity adjuvant therapies to help those that do of the cerebral peduncle should be >4; (2) abnormal method of delivery— not respond to cooling alone are cur- identified as “at risk” of CP and should breech, face or shoulder delivery, inter- rently in animal model and phase 1 join Pathway B (Figure 5) at nal version, or delivery by an untrained neonatal studies, for example, erythro- “assessment for CP.” person; and (3) neonatal illness in the poietin, melatonin, xenon, and The risk of developing CP in term 1st week of life—convulsions, cyanotic topiramate [Gonzalez and Ferriero, infants who have received routine care at 2009]. birth, the third group of infants who go attacks, cerebral signs, hypothermia, In term infants with moderate to on to develop CP, is 1/1,000 neonatal jaundice, Rh incompatibility, or serious severe NE, imaging showing basal gan- survivors and these infants are at the low- illness. Infants were at the highest risk glia/thalamus injury has a positive est risk. However, they represent 45% of of disability when all three of these risks predictive value for CP of 88% [de all infants with CP and numerically com- were apparent. They were only a small Vries et al., 2011]. In a study of 173 prise the largest group (Fig. 4). Why do group (0.1% of total births), but more term infants with NE, the basal ganglia/ these apparently “neurologically normal” importantly only 0.2% of those with a thalamus pattern of injury was associ- children at birth develop CP, and can we disability. When any combination of ated with the most severe motor and identify them earlier so they can have these three risks were used, 13.2% of all cognitive outcomes at 30 months access to active surveillance and early live births were classified as at risk, and [Miller et al., 2005]. intervention? this identified 26.3% of all those with a Practice point. Term infants with mod- From a total population case con- disability. A striking finding was that erate to severe NE and a basal ganglia/ trol study in Western Australia, 74% of all those with CP, severe mental thalamus injury should be automatically McIntyre et al. [2011] compared the handicap, hearing, and sight impair- described as “At high risk,” and go clinical descriptions of 295 term infants ments could not be identified using this straight to Pathway B (Figure 5). They with CP with 442 term control infants model. should receive a GMs Assessment, be none of which required special care. Very little has changed for those referred to active surveillance and early They identified six independent predic- born at term without any noticeable signs intervention when they leave the tors of CP in the neonatal period: during the neonatal period since the first hospital. abnormal fontanelle OR 4.4 (95% CI studies of these cohorts in the 1950s. For The remaining infants with NE 0.8–23); abnormal tone OR 7.3 (95% these infants, failure to reach major that go on to be described as having CI 2–26.8); birth defects identifiable in motor milestones, such as rolling, sitting CP have antenatal risks such as IUGR, the newborn period OR 5.2 (95% CI or standing, have often been the catalyst intrauterine infection, metabolic abnor- 2.4–10); ventilatory assistance restricted for the commencement of developmen- malities, syndromes, and birth defects to the labor room only OR 2.9 (95% tal assessments and interventions. Given [Badawi et al., 1998; Kurinczuk et al., CI 2.2–12); abnormal consciousness that the window for milestone attain- 2010]. Perinatal arterial stroke occurs in referred to irritability and lethargy, but ment in typically developing children is 1.7/100,000 live births. In the new- none were comatosed OR 3.7 (95% CI quite broad [WHO Multicenter Growth born period, it can also result in NE, 2–7); and in the small group with Reference Study Group, 2006], this usu- but the majority of these infants present abnormal temperature regulation tem- ally leads to a “wait and see” approach after the immediate neonatal period perature was down or fluctuating, not where infants receive no intervention with seizures or hemiparesis. Mothers high OR 4.1 (95% CI 1.2–14). A num- during their period of rapid neural devel- with preeclampsia and infants who have ber of these predictors are reminiscent opment. In view of the fact that every IUGR are at risk of perinatal arterial of criteria for mild NE, and the pres- second child with CP will be born at stroke [Shankaran, 2008]. Stroke with ence of two or more of these factors term and requires no special care in the abnormalities involving the cerebral yielded a high specificity (99%), but neonatal period, it is imperative that peduncle are also highly predictive of low sensitivity (14%) for CP. This is frontline health professionals such as CP PPV 78% [de Vries et al., 2011]. not surprising considering the unknown pediatricians, general practitioners and Practice point. Infants with a cerebral etiology of this group of infants. Of this allied health practitioners have a best birth defect, or stroke with involvement low risk group who had CP, 58% did practice pathway to follow when a parent

120 DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL Figure 5 Recommended assessment for identiﬁcation of infants at risk of CP. presents with a child who falls into this tions for parents. Consideration should CP) are essential to prevent, or at least category. also be given to risk factors during minimize, the handicapping effects of a Practice point. When parents bring pregnancy and signs of mild NE in the disability and to make the most of the their term born child (3 months to 3 neonatal period. When an abnormal assets a child possesses” [Alberman and years of age) that did not require special result is derived, Pathway B (Figure 5) Goldstein, 1970]. Yet, paradoxically, 40 care when born to a health professional should be followed to “assessment for years later families are not automatically with concerns regarding motor devel- CP” through standardized motor receiving early intervention while they opment or abnormal posturing they assessments. should go straight to Pathway B at “wait and see” whether their child will “screen for CP.” We propose that a The description of cerebral palsy is tradition- “catch up” from simply a slower motor tiered approach as developed by Rose- ally given late but can be given earlier developmental trajectory or if their nbaum et al. [2009] should be adopted. This review is timely as “it is child actually has CP or DCD or an in- They recommend using the ages and now universally accepted that the ear- tellectual impairment with associated stages questionnaire 1 three extra ques- liest possible diagnosis and treatment (of motor difﬁculties.

DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL 121 Figure 5 (Continued)

CP registers indicate the average while the brain is most plastic (i.e., in away. For those who are not at high age for a description of CP to be given the first 2 years of life), which is almost risk but have early signs, they should be is 19 months, but the range is wide. never the case when the family is taking regularly comprehensively assessed to For those with severe motor impair- part in “wait and see” monitoring prior ensure access to the most appropriate ment the description of CP can be to description. early intervention. given as early as 1 week but may take Good evidence shows that earlier up to 3 years, and less surprisingly for detection of CP is both possible and Why is Cerebral Palsy Missed and those with mild or moderate motor accurate and, more importantly, diag- Why is the Description so Difficult impairment the description of CP is nostic-specific early intervention is for Doctors to Make? given anywhere between 1 week and 5 therefore possible. Rather than waiting Health professionals hesitate to use years of age [Watson et al., 2006]. The for a formal description of CP to be the terminology CP early for a number burgeoning body of recent neuroplas- given, infants should be identified as “at of reasons, but importantly the condition ticity literature suggests that intensive, high risk of CP” when they are high is not a diagnosis; it is a “clinical repetitive, task-specific intervention for risk, and therefore commence diagnos- description.” There are no biological CP ought to commence very early tic-specific early intervention straight markers or definitive tests for CP. The

122 DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL Figure 5 (Continued)

term does not infer etiology, and it has bad news or provide time for the child conditions (e.g., Trisomy 18, Angelman no prognostic value as severity and asso- to grow out of it. Syndrome, Cornelia de Lange syn- ciated impairments are incredibly drome) [Badawi et al., 1998]. variable. However, 86% of parents know Rule out other diagnoses something is wrong with their child Doctors ﬁrst rule out other diag- Delay the delivery of bad news before a description of CP is given [Baird noses that may explain the symptoms. Doctors sometimes delay the et al., 2000]. Leading up to this point in This is an important step as there are delivery of bad news while exploring time, most parents experience being told other conditions that mimic the early the possibility of a less severe, more by their medical team that the plan is to signs of CP which can have important common disorder such as DCD. Differ- “wait and see.” When health professio- treatment implications, such as: neuro- ential diagnosis is critical as it informs nals use the term “wait and see,” the degenerative conditions (e.g., Ataxia the selection of intervention strategies intention is to use this time to rule out Telangiectasia); metabolic syndromes suited to the speciﬁc condition. For other diagnoses, delay the delivery of (e.g., Glutaric acidemia); and genetic example, effective intervention for

DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL 123 pediatrician who may lack access to the relevant maternal-fetal and/or neonatal medical history. The pediatrician may then be assessing a healthy baby that may just appear slightly “delayed,” and it is not until later in infancy that the gravity of the problem may be evident, precipitating a late diagnosis.

What are the Most Important Things that can be Done in Clinical Practice to Describe Cerebral Palsy Earlier? We propose a new clinical pathway that is designed to circumvent the existing screening and diagnostic barriers by tying together the relevant evidence needed to make an earlier diagnosis and commence earlier intervention (see Pathways A and B). These pathways have been developed using GRADE level evidence [Guyatt et al., 2008] and “traffic lights” to signify the effectiveness of the interventions [Novak and McIntyre, 2010]. Green equals “go,” (high quality evidence to support the use of the intervention, therefore use this approach). Yellow equals “measure” (low quality or conflicting evidence supporting the effectiveness of Figure 6 the intervention). Red equals “stop” (high quality evidence indicating inef- fective interventions) [Novak and McIntyre, 2010]. The serious nature of these stand- DCD involves cognitive approaches into strange positions, I especially notice ard care limitations has led us to best suited to school-aged children, it each time I pick him up. I went to the conclude that “waiting and seeing” is whereas CP intervention uses a variety GP, who agreed and thought I should potentially harmful to children with CP of pharmacological, motor, social and see a pediatrician. I went to the pediatri- and their families. We therefore have cognitive intervention approaches that cian who agreed they were unusual and identified solutions to three of the can commence early in life. It is there- said let us see how he is when he is 10 major problems relating to the late di- fore important that children with CP months old. That is too long to wait! So I agnosis of CP, which are timely and are differentiated earlier in order to get went to another pediatrician who agreed possible for the health system to redress: the right interventions early. again, it was abnormal, so now I am booked to go to a physiotherapist for fur- New clinical diagnostic and intervention Provide opportunity to grow out of it ther tests, and after that they will decide pathways Doctors sometimes delay the what to do” but I do not know what to When the system fails to recog- delivery of bad news to provide enough do now...” (Personal communication, nize a child with CP very early due to time for the possibility that the child February 4, 2012, parent discussion with using the “wait and see” monitoring may “grow out of it.” However for first author over the phone). mode, this decision essentially ensures those few whose motor signs resolve, System barriers to description are that infants receive limited or no diag- commonly they transpire to have an in- also potentially at work. For example, nostic-specific intervention within the tellectual impairment or behavioral for any mother and her newborn, critical window of brain development. problems [Nelson and Ellenberg, 1981]. obstetricians hold vital information The window of brain development, The brain injury responsible for about maternal-fetal health. If the baby where the brain is actively sprouting CP may be suspected or even con- is premature or ill, care is immediately and pruning in response to activity, is firmed in the neonatal period, but the transferred to neonatal specialists, where often misspent in children with CP. In diagnosis for many does not occur until the primary patient is now the infant, Pathways A and B, we review the evi- the motor impairments and activity lim- not the mother, and some of the rele- dence for early intervention possibilities itations inherent in the definition are vant preconception and pregnancy in CP. The evidence tells us quite observable. This lag time is not useful history about risk factors for CP may clearly that general early intervention to families or to the child. not be passed on. When the infant is and parent interventions, designed to “...... I am very worried about my well and discharged from hospital, care enhance in-home care characterized by son, he is 5 months old, and over the last is likely to be transferred to a commu- positive interactions, categorically month I have noticed he seems to go nity based general practitioner or improve a child’s cognition with the

124 DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL best effect seen in children of low tiveness. It was deemed not practicable Newer techniques and technolo- socio-economic status. However, more to have universal screening of all chil- gies are being developed which are recent neuroplasticity evidence suggests dren, but it was felt essential that all likely to advance the role of imaging in that a skill-based, high-intensity practice children at risk be monitored. In a letter the diagnostic process and treatment approach to early intervention is to the Lancet in 1967 defending the selection process. Advanced neuroimag- required to impact on motor outcomes, concept, Dr Ronald Mac Keith and col- ing techniques such as diffusion as is the case in most adult brain inju- leagues wrote, “by the criterion of weighted imaging (DWI) and diffusion ries. These newer types of motor identifying handicaps which are in some tensor imaging (DTI) have been uti- learning approaches, which are effective cases undoubtedly, and in other cases lized to more specifically identify in older children with CP, require probably, benefited by having treatment diffuse or subtle white matter injuries urgent study within the CP infant pop- started without delay, developmental [Hoon and Faria, 2010]. Magnetic reso- ulation. It is therefore the responsibility and neurological assessment from the nance spectroscopy (MRS), provides of the health professional who observes age of 5 months is neither difficult nor measures of brain biochemistry and is major risk factors or a motor delay to inefficient” [Mac Keith et al., 1967]. proving an effective tool in understand- investigate further, diagnose “at risk of The concept itself was deemed by most ing prognosis in NE and preterm CP” early, and refer to early interven- to be a sound one. The problem at this infants [Ancora et al., 2010; Van Kooij tion at a minimum to optimize their time was the “at risk” criteria used was et al., 2012]. Large deformation diffeo- cognitive function. We outline a way identifying up to 60% of all live births morphic metric mapping (LDDMM), to do this via systematic use of risk fac- in an area. The goal of these programs where a 3D atlas of the brain is pro- tor history taking, neurobehavioral was to screen 10–20% of all births to duced, shows great promise for predictive tools, in addition to MRI identify the majority of the invisible illuminating the structural brain abnor- (Pathways A and B). handicaps that is, those that would oth- malities that occur in CP with the erwise not be identified until the 4th potential for informing selection, Promotion of a climate for new research that and 5th years of life. We recommend design, and measurement of rehabilita- will improve outcomes that the “wait and see” period is tion interventions [Faria et al., 2011]. Late description of CP is creating reframed to the “wait and be” period, a major problem for recruitment of where children are diagnosed “at risk of General neuromotor and developmental infants to promising early rehabilitative CP” early and are immediately referred assessments and potentially curative studies. Lack of to diagnostic-specific early intervention. Many neuromotor and develop- diagnosis is impeding the advancement mental assessments with sound of regenerative medicine, early inter- psychometric properties exist for infants vention and other well-recognized What Tools can be Used to and young children. For diagnostic pur- treatments for CP yet to be tested in Accurately Predict and Identify poses, tools with predictive properties the earlier years, for example, medical Early Signs of Cerebral Palsy? are the most worthwhile. However, interventions for tone management, there has been a historical preference by reflux, and epilepsy. When a health Imaging pediatricians and neonatal follow-up professional identifies an infant at high teams to use discriminative tools that risk for CP, coupled with referral to Practice point. All children with a pre- assess a combination of: abnormal mus- early intervention trials, it will help to sumed or suspected brain injury should cle tone of the trunk and extremities; accelerate future discoveries for these have magnetic resonance imaging (MRI). the presence of primitive reflexes; the children and change the landscape of Neuroimaging is used as an inte- quality and quantity of voluntary move- the diagnosis and prognosis. gral part of the diagnostic process ment (e.g., milestone acquisition); and [Krageloh-Mann and Horber, 2007]. the presence of involuntary movement. Promotion of good family mental health and MRI is the gold-standard neuroimaging The problem with this persistent prac- resilience for the long-term technique for elucidating the pathoge- tice is that these tools are only useful If late description is not helping nesis of CP: white matter damage of for discriminating between infants who infants or research, are we helping immaturity (WMDI) including PVL, are developing typically from those parents by sheltering them from bad lesions of the deep grey matter, malfor- who are not. Determining who is typi- news? A population study conducted in mations, focal infarcts, and cortical and cally developing and who is not is even Britain found that parental dissatisfac- subcortical lesions [Bax et al., 2006]. more complicated in premature infants tion with delayed diagnosis of CP is Cranial ultrasound (CUS) is a safe and because they have their own develop- associated with higher rates of parental inexpensive alternative used in the neo- mental trajectory [Heineman and depression [Baird et al., 2000]. So it natal intensive care unit (NICU) to Hadders-Algra, 2008; Spittle et al., would appear that sparing parents from detect structural changes in the new- 2008a]. Routinely used neuro observa- bad news is unhelpful. Therefore early born brain. However, MRI has higher tions and standardized developmental recognition and provision of early pre- sensitivity and specificity than CUS as a tests were not designed to specifically ventative mental health support for predictor of CP in very low birth detect the presence of CP and thus fur- families may help parents manage the weight (VLBW) infants [Mirmiran ther compound the complexity of the inevitable stress, which could help et al., 2004]. Despite strong correlations CP diagnostic process. They may be improve family outcomes long-term. between clinical findings and MRI, 12– helpful to some diagnosticians but will The concept of “at risk” is not a 14% of children with CP will have nor- lack adequate specificity for most. new one. During the 1960s in the mal MRIs [Bax, 2006; Krageloh-Mann Ideally the aim of monitoring United Kingdom, there were “at risk” and Horber, 2007] and therefore MRI ought to be to differentiate why some registers, with the usual accompanying should not be used in isolation for mak- children are not developing normally, debate over their value and cost effec- ing the description of CP. to enable diagnostic-appropriate best-

DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL 125 available evidence-based intervention to When correlated with MRI findings, four-months, to enable medical teams be provided. This paper will now focus namely white matter injury, the GMs to use the best predictive tools to help on the evidence for the best available assessment (specifically “absent fidgety”) make the description of CP earlier. tools for predicting and recognizing has been shown to accurately predict CP, distinct from tools better suited to CP 100% of the time in very preterm Practice point. When examining infants, suspecting global developmental delay infants [Spittle et al., 2008a] Evidence do not discount CP when spasticity or (GDD). Clinometric reviews indicate of the predictive value of GMs in full dyskinesia is not identified. A period of that different tools need to be used at term infants with hypoxic ischemic time lapses between the original damage different ages to describe and detect CP encephalopathy (HIE) has also been to the developing brain, whether in and that a combination of tools is best demonstrated [Prechtl et al., 1993]. utero or during early infancy/child- practice [Heineman and Hadders-Algra, Importantly, the GMs assessment has hood, and the appearance of 2008; Spittle et al., 2008a]. good clinical utility because it is quick, impairments. It is well known that the Practice point. A combination of risk inexpensive, and noninvasive. Rater brain, which begins development in factor history taking, neurological ex- training is provided by the GMs trust. utero, continues to develop during amination that includes assessment of Hammersmith infant neurological assessment childhood. Thus a child’s neural devel- quality of movement, volitional move- [Haataja et al., 1999]. The Hammer- opment is “age-specific,” so brain ment and neuroimaging are required. A smith assessment is based on the dysfunction will manifest according to health professional with clinical exper- Dubowitz and Dubowitz [1981] assess- the brain’s development at that age tise and experience in motor ment of the newborn and is a simple [Hadders-Algra, 2004]. Compared with development should interpret and eval- method of examining infants between 2 a mature brain which responds to injury uate the findings generated by these and 24 months of age. There are three with specific and localized signs, a assessments (Figure 6). parts to the examination: neurologic young infant may present with general- signs, developmental milestones, and ized and nonspecific signs (e.g., Tools predictive of cerebral palsy behavior. In the first section, the neu- hypotonia) [Kuban and Leviton, 1994; rologic exam, an optimality score is Hadders-Algra, 2004]. It is proposed Qualitative assessment of general movements obtained from the assessment of cranial that further brain development in an [Einspieler et al., 2004)]. Of all the nerve function, posture, quality and infant, including myelination of axons tools available to predict CP, GMs is quantity of movement, tone, and and maturation of basal ganglia neurons, consistently the most predictive, with reflexes and reactions. The second and must occur before spasticity and dyski- specificity and sensitivity rates higher third sections do not form part of the nesia can manifest [Kuban and Leviton, than MRI [Burger and Louw, 2009]. overall score but give important addi- 1994]. The infant with hypotonia may The GMs assessment measures the qual- tional information regarding thus “develop” spasticity and dyskinesia ity of spontaneous movements with the developmental progress. Recent studies by the age of 1 or 2 years, as the com- infant lying supine. Scoring is done by have demonstrated the predictive value plexity of neural functions increases trained assessors via observation of of the Hammersmith infant neurological [Kuban and Leviton, 1994; Hadders- video footage and can be used from the assessment (HINE) for CP. A large Algra, 2004]. preterm period until 20 weeks post study [Pizzardi et al., 2008] of 658 Movement assessment of infants [Chandler term age (PTA). Two distinct time infants who were either preterm or et al., 1980]. The movement assess- periods for assessment exist; the writh- term with NE were prospectively stud- ment of infants (MAI) is a criterion- ing period (up to 9 weeks PTA) and ied from birth until 12 months referenced scale that evaluates neuro- the fidgety period (from 9 to 20 weeks corrected age. ROC curve analysis was motor dysfunction in high risk infants PTA). In both periods, the infant is used to test the predictive power of the at 4, 6, 8, and 12 months of age. The scored with “normal” or “abnormal” HINE. Global HINE scores showed assessment is carried out by a therapist GMs. Abnormal GMs are then further high prediction of CP at all ages (ROC and takes 30–60 min to complete, classified. In the writhing period, curve areas above 0.9), but most impor- requiring a manual but no specialized abnormal GMs known as “cramped tantly movement quality and quantity equipment. The MAI assesses tone, synchronized” have been shown to be test items had even higher predictive primitive reflexes, equilibrium reactions, highly predictive of CP (sensitivity power. and volitional movement. The test has 5100%; specificity 5 40%; PPV59.4%; A retrospective study of 70 infants been shown to be twice as sensitive as NPV5 100% [Spittle et al., 2009]. If diagnosed at 2 years with CP observed the Bayley scales of infant development the abnormal GM of “cramped syn- a strong (r 5282) negative correlation in detecting early signs of CP [Harris, chronized” is followed by the abnormal between HINE scores at 3–6 months of 1987]. Studies of predictive values at 4 GM “absent fidgety” (in the fidgety pe- age and levels of GMFCS [Romeo and 8 months of age report sensitivity riod) this has consistently shown the et al., 2008a]. Infants in GMFCS levels rates ranging from 73.5 to 96.0 and highest predictive value for CP (Darsa- 3–5 scored below 40, whereas those in specificity of 62.7–78.2 [Spittle et al., klis and Snider, 2011). levels 1–2 scored between 40 and 60. 2008b]. A recent investigation of the A recent systematic review of 17 Combined use of the HINE and GMs predictive validity of the MAI at 6 studies demonstrated the accuracy of at 3 months PTA can be used to months of age demonstrated a signifi- the GMs assessment in predicting neu- describe an infant as at “high risk” of cant correlation between MAI scores rodevelopmental outcomes in infants up CP [Romeo et al., 2008b]. and Bayley scales of infant development to 2 years with a sensitivity 92% and at 12 months, although sensitivity and specificity 82% [Burger, 2009]. The Practice point. Routine follow-up for specificity for CP were not reported GMs assessment has been found to be preterm and sick infants should be [Metgud et al., 2011]. superior to ultrasound findings in pre- scheduled at three-months and six- Other useful assessments. Several other dicting CP [Einspieler et al., 2004] months corrected, not the conventional neuromotor assessments, such as the test

126 DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL of infant motor performance (TIMP) of all CP falls into this category. There- Bax M, Goldstein M, Rosenbaum P, et al. 2005. [Campbell, 2005], The neuro-sensory fore we recommend prompt response to Proposed definition and classification of cerebral palsy, April 2005. Dev Med Child motor development assessment parental concerns with screening and Neurol 47:571–576. (NSMDA) [Burns et al., 1989], and the assessments as outlined, followed by im- Bax M, Tydeman C, Flodmark O. 2006. Clinical Alberta infant motor scale (AIMS) mediate referral for intervention for and MRI correlates of cerebral palsy. JAMA [Piper and Darrah, 1994], are appropri- those infants then considered “at risk.” 296:1602–1608. ately used to discriminate infants with Premature and term infants with Beaino G, Khoshnood B, Kaminski M, et al. 2010. Predictors of cerebral palsy in very abnormal motor function from those brain injury identified on MRI are at preterm infants: the EPIPAGE prospective typically developing. All have sound high risk of CP. We have identified population-based cohort study. Dev Med psychometrics. Of these tools, the pathways which make recognizing “at Child Neurol 52:e119–125. TIMP has been shown to be sensitive high risk” of CP easier for health pro- Blair E, Stanley FJ. 1997. Issues in the classifica- to change in response to intervention fessionals. We propose a change in tion and epidemiology of cerebral palsy. Ment Retard Dev Disabil Res Rev 3: [Campbell et al., 1995]. diagnostic practice, a shift away from 184–193. Assessment summary. referral for intervention following a for- Blair E, Watson L. 2006. Epidemiology of cere- mal (most often late) description to one bral palsy. Semin Fetal Neonatal Med High risk infants should be routinely of referral when an infant is “at high 11:117–125. assessed using the GMs preferably three risk” of CP. This will provide the op- Blair E, Watson L, Badawi N, et al. 2001. Life expectancy among people with cerebral times; during early admission, around portunity for targeted research in early palsy in Western Australia. Dev Med Child term corrected (if preterm) and at 9–14 intervention, thus providing optimal Neurol 43:508–515. weeks (corrected for gestational age). outcomes for children with CP. Blank R, Smits-englesman B, Polatajko H, et al. “High risk of CP” designation should 2011. Eurpoean Academy for Childhood Disability (EACD): recommendations on be given to infants at 9–14 weeks (cor- ACKNOWLEDGMENTS the definition, diagnosis and intervention rected) with a combination of absent Many thanks to the families that of developmental coordination disorder fidgety GMs and white matter injury participate in CP Registers and research (long version). Dev Med Child Neurol throughout the world, the clinicians 54:e1–7 on MRI. Burger M, Louw Q. 2009. The predictive validity After 20 weeks (corrected), use the who work in this important area, and Dr. Monique Hines for her fine edito- of general movements—a review. Eur J Pae- HINE or MAI. diatr Neurol 13:408–420. rial skills. This research was conducted MRI is the best imaging tool to eluci- Burns Y, Ensby R, Norrie M. 1989. The neuro- at Cerebral Palsy Alliance Research sensory motor development assessment. Part date the pathogenesis of CP and should Institute, The University of Notre 1: development and administration of the be offered to all infants who have Dame, Australia. test. Aust J Physiother 35:141–157. abnormal findings. Campbell SK. 2005. The test of infant motor per- Use the CP description form to REFERENCES formance. Test user’s manual version 2.0. Chicago: LLC. describe motor type and severity to ACPR Group. 2009. Report of the Australian Campbell SK, Kolobe TH, Osten ET, et al. inform intervention planning. Cerebral Palsy Register, birth years 1993– 1995. Construct validity of the test of 2003, www.cpregister-aus.com.au. infant motor performance. Phys Ther 75: CONCLUSION Alberman ED, Goldstein H. 1970. The “at risk” 585–596. Until recently, CP was considered register: a statistical evaluation. Br J Prev Carlo WA, McDonald SA, Tyson JE, et al. 2011. Soc Med 24:129–135. Cytokines and neurodevelopmental out- unpreventable, incurable, and almost Ancora G, Soffritti S, Lodi R, et al. 2010. A comes in extremely low birth weight infants. untreatable. However, preventive efforts combined a-EEG and MR spectroscopy J Pediatr 159:919–925, e913. including: rubella vaccination, iodine study in term newborns with hypoxic-ische- Carlsson M, Hagberg G, Olsson I. 2003. Clinical supplementation in areas of severe iron mic encephalopathy. Brain Dev 32:835–842. and aetiological aspects of epilepsy in chil- Andersen G, Mjoen T, Vik T. 2010. Prevalence dren with cerebral palsy. Dev Med Child deficiency, anti-D vaccination, prevent- Neurol 45:371–376. ing methyl–mercury contamination, of speech problems and the use of augment- ative and alternative communication in chil- Chandler L, Andrews M, Swanson M, et al. reducing the number of embryos trans- dren with cerebral palsy: a registry-based 1980. Movement assessment of infants: a ferred in invitro fertilization (IVF) (in study in Norway. Perspect on Augment manual. Washington: Rolling Bay. Australia), and enforcing laws for seat Altern Commun 19:12–20. Council on Children With Disabilities, Section belts and fencing around swimming Andersen GL, Romundstad P, De LaCruz J, et al. on Developmental Behavioral Pediatrics, 2011. Cerebral palsy among children born Bright Futures Steering Committee and pools have been successful prevention Medical Home Initiatives for Children With strategies. Recently, magnesium sulfate moderately preterm or at moderately low birthweight between 1980 and 1998: a Eu- Special Needs Project Advisory Committee. and hypothermic intervention have also ropean register-based study. Dev Med Child 2006. Identifying Infants and Young Chil- started to prevent a small proportion of Neurol 53:913–919. dren With Developmental Disorders in the Medical Home: An Algorithm for Develop- CP. Both of these interventions occur Arnaud C, White-Koning M, Michelsen SI, et al. mental Surveillance and Screening Pedia- very early and require health professio- 2008. Parent-reported quality of life of chil- trics;118:405–420. dren with cerebral palsy in Europe. Pedia- nals to be mindful of CP as a potential Crowther CA, Hiller JE, Doyle LW. 2002. Mag- trics 121:54–64. outcome that could be prevented or nesium sulphate for preventing preterm birth Badawi N, Felix JF, Kurinczuk JJ, et al. 2005. cured. With advances in medical, public in threatened preterm labor. Cochrane Data- Cerebral palsy following term newborn base Syst Rev CD001060. health, and allied health research, the encephalopathy: a population-based study. Darsaklis V, Snider L, Majnemer A, et al. 2011. likelihood of further breakthroughs are Dev Med Child Neurol 47:293–298. Predictive validity of Prechtl’s method on probable. Badawi N, Watson L, Petterson B, et al. 1998. the qualitative assessment of general move- Further research is required to What constitutes cerebral palsy? Dev Med ments: a systematic review of the evidence. determine why infants born at term, not Child Neurol 40:520–527. Dev Med Child Neurol 53:896–906. Baird G, McConachie H, Scrutton D. 2000. de Vries LS, vanHaastert I, Benders M, et al. at “high risk” of CP in the newborn pe- Parents’ perceptions of disclosure of the di- 2011. Myth: cerebral palsy cannot be pre- riod go on to develop CP. Health agnosis of cerebral palsy. Arch Dis Child dicted by neonatal brain imaging. Semin Fe- professionals need to be aware that 45% 83:475–480. tal Neonatal Med 16:279–287.

DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL 127 Dubowitz L, Dubowitz V. 1981. The neurologi- Jacobs SE, Tarnow-Mordi WO. 2010. Therapeu- Nelson KB, Ellenberg JH. 1981. Apgar scores as cal assessment of the preterm and full-term tic hypothermia for newborn infants with predictors of chronic neurologic disability. newborn infant. London: Heinemann. hypoxic–ischaemic encephalopathy. J Pae- Pediatrics 68:36–44. Einspieler C, Prechtl HF, Bos AF, et al., editors. diatr Child H 46:568–576. Nelson KB, Ellenberg JH. 1986. Antecedents of 2004. Prechtl’s method on the qualitative Jahnsen R, Villien L, Aamodt G, et al. 2004. cerebral palsy. Multivariate analysis of risk. assessment of general movements in preterm, Musculoskeletal pain in adults with cerebral N Engl J Med 315:81–86. term and young infants. Clinics in develop- palsy compared with the general population. Newman CJ, O’Regan M, Hensey O. 2006. mental medicine. London: Mac Keith Press. J Rehabil Med 36:78–84. Sleep disorders in children with cerebral Ellenberg JH, Nelson KB. 1988. Cluster of perina- Jarvis S, Glinianaia SV, Arnaud C, et al. 2005. palsy. Dev Med Child Neurol 48:564–568. tal events identifying infants at high risk for Case gender and severity in cerebral palsy Novak I, Hines M, Goldsmith S, et al. Clinical death or disability. J Pediatr 113:546–552. varies with intrauterine growth. Arch Dis prognostic messages from a systematic Faria AV, Hoon A, Stashinko E, et al. 2011. Quan- Child 90:474–479. review about cerebral palsy. Pediatrics titative analysis of brain pathology based on Kirby RS, Wingate MS, VanNaarden Braun K, 130:e1285–1312. MRI and brain atlases—applications for cere- et al. 2011. Prevalence and functioning of Novak I, McIntyre S. 2010. The effect of educa- bral palsy. Neuroimage 54:1854–1861. children with cerebral palsy in four areas of tion with workplace supports on practi- Gonzalez FF, Ferriero DM. 2009. Neuroprotec- the United States in 2006: a report from the tioners’ evidence-based practice knowledge tion in the newborn infant. Clin Perinatol Autism and Developmental Disabilities and implementation behaviors. Aust Occup 36:859–880, vii. Monitoring Network. Res Dev Disabil Ther J 57:386–393. Gorter J, Ketelaar M, Rosenbaum P, et al. 2008. 32:462–469. Odding E, Roebroeck ME, Stam HJ. 2006. The Use of the GMFCS in infants with CP: the Krageloh-Mann I, Horber V. 2007. The role of epidemiology of cerebral palsy: incidence, need for reclassification at age 2 years or magnetic resonance imaging in elucidating impairments and risk factors. Disabil Rehabil older. Dev Med Child Neurol 51:46–52. the pathogenesis of cerebral palsy: a system- 28:183–191. Guyatt GH, Oxman AD, Vist GE, et al. 2008. atic review. Dev Med Child Neurol Palisano R, Rosenbaum P, Walter S, et al. 1997. GRADE: an emerging consensus on rating 49:144–151. Development and reliability of a system to quality of evidence and strength of recom- Kuban KC, Leviton A. 1994. Cerebral palsy. N classify gross motor function in children mendations. BMJ 336:924–926. Engl J Med 330:188–195. with cerebral palsy. Dev Med Child Neurol Haataja L, Mercuri E, Regev R, et al. 1999. Kurinczuk JJ, White-Koning M, Badawi N. 39:214–223. Optimality score for the neurologic exami- 2010. Epidemiology of neonatal encephal- Parkes J, Hill N, Platt MJ, et al. 2010. Oromotor nation of the infant at 12 and 18 months of opathy and hypoxic–ischaemic encephalop- dysfunction and communication impairments age. J Pediatr 135:153–161. athy. Early Hum Dev 86:329–338. in children with cerebral palsy: a register Hadders-Algra M. 2004. General movements: a Love SC. 2007. Better description of spastic cere- study. Dev Med Child Neurol 52:1113–1119. window for early identification of children bral palsy for reliable classification. Dev Med Parkes J, White-Koning M, Dickinson HO, et al. at high risk for developmental disorders. J Child Neurol 49:24–25. 2008. Psychological problems in children Pediatr 145:S12–S18. Mac Keith R, Robson P, Howorth I. 1967. The with cerebral palsy: a cross-sectional Euro- Hagberg B, Hagberg G, Olow I, et al. 1996. The "at risk" infant. Lancet 290:886–887. pean study. J Child Psychol Psychiatry changing panorama of cerebral palsy in Sweden. McAdams R, Juul S. 2011. Cerebral palsy: preva- 49:405–413. VII. Prevalence and origin in the birth year pe- lence, predictability, and parental counseling. Pharoah PO. 2006. Risk of cerebral palsy in multi- riod 1987-90. Acta Paediatr 85:954–960. Neoreviews 12:e564. ple pregnancies. Clin Perinatol 33:301–313. Hagglund G, Andersson S, Duppe H, et al. 2005. McIntyre S, Badawi N, Brown C, et al. 2011. Piper M, Darrah J. 1994. Motor assessment of the Prevention of dislocation of the hip in chil- Population case-control study of cerebral developing infant. Phildaelphia: WB Saunders. dren with cerebral palsy. The first ten years palsy: neonatal predictors for low-risk term Pizzardi A, Romeo DMM, Cioni M, et al. 2008. of a population-based prevention pro- singletons. Pediatrics 127:e667–e673. Infant neurological examination from 3 to gramme. J Bone Joint Surg Br 87:95–101. McIntyre, S., Taitz, D., Keogh, J. et al. 2012. A 12 months: predictive value of the single Harris SR. 1987. Early detection of cerebral palsy: systematic review of risk factors for cerebral items. Neuropediatrics 39:344–346. sensitivity and specificity of two motor palsy in children born at term in developed Prechtl H, Ferrari F, Cioni G. 1993. Predictive assessment tools. Bayley motor scale and the countries. Dev Med Child Neurol Nov 26 value of general movements in asphyxiated movement assessment of infants (MAI). J DOI: 10.1111/dmcn.12017. fullterm infants. Early Hum Dev 35:91–120. Perinatol 7:11–15. McManus V, Guillem P, Surman G, et al. 2006. Reid S, Carlin J, Reddihough D. 2011a. Distri- Heineman K, Hadders-Algra M. 2008. Evaluation SCPE work, standardization and defini- bution of motor types in cerebral palsy: how of neuromotor function in infancy—a sys- tion—an overview of the activities of SCPE: do registry data compare? Dev Med Child tematic review. J Dev Behav Pediatr a collaboration of European CP registers. Neurol 53:233–238. 29:315–323. Zhongguo Dang Dai Er Ke Za Zhi 8: Reid SM, Carlin JB, Reddihough DS. 2011b. Himmelmann K, Uvebrant P. 2011. Function and 261–265. Rates of cerebral palsy in Victoria, Australia, neuroimaging in cerebral palsy: a popula- Metgud D, Patil V, Dhaded S. 2011. Predictive 1970 to 2004: has there been a change? Dev tion-based study. Dev Med Child Neurol validity of the movement assessment of Med Child Neurol 53:907–912. 53:516–521. infants (MAI) for six-month-old very low Roijen LE, Postema K, Limbeek VJ, et al. 2001. Himmelmann K, Beckung E, Hagberg G, et al. birth weight infants. J Phys Ther 3:19–23. Development of bladder control in children 2006. Gross and fine motor function and Michelsen SI, Uldall P, Kejs AM, et al. 2005. and adolescents with cerebral palsy. Dev accompanying impairments in cerebral palsy. Education and employment prospects in cer- Med Child Neurol 43:103–107. Dev Med Child Neurol 48:417–423. ebral palsy. Dev Med Child Neurol 47: Romeo DMM, Cioni M, Scoto M, et al. 2008a. Himmelmann K, Hagberg G, Uvebrant P. 2010. 511–517. Neuromotor development in infants with cer- The changing panorama of cerebral palsy in Miller S, Ramaswamy V, Michelson D, et al. 2005. ebral palsy investigated by the Hammersmith Sweden. X. Prevalence and origin in the Patterns of brain injury in term neonatal infant neurological examination during the first birth-year period 1999-2002. Acta Paediatr encephalopathy. J Pediatr 146:453–460. year of age. Eur J Paediatr Neurol 12:24–31. 99:1337–1343. Mirmiran M, Barnes P, Keller K, et al. 2004. Romeo DMM, Guzzetta A, Scoto M, et al. 2008b. Himpens E, Oostra A, Franki I, et al. 2010. Pre- Neonatal brain magnetic resonance imaging Early neurologic assessment in preterm-infants: dictability of cerebral palsy and its character- before discharge is better than serial cranial integration of traditional neurologic examina- istics through neonatal cranial ultrasound in ultrasound in predicting cerebral palsy in tion and observation of general movements. a high-risk NICU population. Eur J Pediatr very low birth weight preterm infants. Eur J Paediatr Neurol 12:183–189. 169:1213–1219. Pediatrics 114:992–998. Rosenbaum PL, Missiuna C, Echeverria D, et al. Himpens E, Van denBroeck C, Oostra A, et al. Mutlu A, Livanelioglu A, Gunel MK. 2008. Reli- 2009. Proposed motor development assess- 2008. Prevalence, type, distribution, and se- ability of Ashworth and modified Ashworth ment protocol for epidemiological studies in verity of cerebral palsy in relation to gesta- scales in children with spastic cerebral palsy. children. J Epidemiol Community Health tional age: a meta-analytic review. Dev Med BMC Musculoskelet Disord 9:44. 63(Suppl 1):i27–i36. Child Neurol 50:334–340. Nanba Y, Matsui K, Aida N, et al. 2007. Mag- SCPE. 2000. A collaboration of cerebral palsy regis- Hoon AH, Faria AV. 2010. Pathogenesis, neuroi- netic resonance imaging regional T1 abnor- ters. Dev Med Child Neurol 42:816–824. maging and management in children with malities at term accurately predict motor Shankaran S. 2008. Prevention, diagnosis, and treat- cerebral palsy born preterm. Dev Disabil outcome in preterm infants. Pediatrics ment of cerebral palsy in near-term and term Res Rev 16:302–312. 120:e10–e19. infants. Clin Obstet Gynecol 51:829–839.

128 DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL Shevell MI, Dagenais L, Hall N. 2009. Comor- Spittle AJ, Doyle LW, Boyd RN. 2008b. A sys- Van Marter L, Kuban K, Allred E, et al. 2011. bidities in cerebral palsy and their relation- tematic review of the clinimetric properties Does bronchopulmonary dysplasia contribute ship to neurologic subtype and GMFCS of neuromotor assessments for preterm to the occurrence of cerebral palsy among level. Neurology 72:2090–2096. infants during the ﬁrst year of life. Dev Med infants born before 28 weeks of gestation? Child Neurol 50:254–266. Arch Dis Child Fetal Neonat Ed 96:F20–F29. Sigurdardottir S, Vik T. 2011. Speech, expressive Watson L, Blair E, Stanley F. 2006. Report of the language, and verbal cognition of preschool Stanley FJ, Watson L. 1992. Trends in perinatal Western Australian Cerebral Palsy Register to children with cerebral palsy in Iceland. Dev mortality and cerebral palsy in Western Aus- birth year 1999. West Perth, Australia: Tele- Med Child Neurol 53:74–80. tralia, 1967 to 1985. BMJ 304:1658–1663. thon Institute for Child Health Research. Soo B, Howard JJ, Boyd RN, et al. 2006. Hip Surman G, Bonellie S, Chalmers J, et al. 2006. WHO Multicenter Growth Reference Study Group. displacement in cerebral palsy. J Bone Joint UKCP: a collaborative network of cerebral 2006. WHO motor development study: win- Surg Am 88:121–129. palsy registers in the United Kingdom. J dows of achievement for six gross motor develop- Public Health (Oxf) 28:148–156. Spittle A, Boyd R, Inder T, et al. 2009. Predicting ment milestones. Acta Paediatr Suppl 450:86–95. motor development in very preterm infants at Tran U, Gray PH, O’Callaghan MJ. 2005. Neona- Wichers MJ, Odding E, Stam HJ, et al. 2005. Clinical 12 Months’ corrected age: the role of qualita- tal antecedents for cerebral palsy in extremely presentation, associated disorders and aetiologi- tive magnetic resonance imaging and general preterm babies and interaction with maternal cal moments in cerebral palsy: a Dutch popula- movements assessments. Pediatrics 123:512– factors. Early Hum Dev 81:555–561. tion-based study. Disabil Rehabil 27:583–589. 517. Wu YW, Colford JM, Jr. 2000. Chorioamnionitis Van Kooij BJM, Benders MJNL, Anbeek P, et al. as a risk factor for cerebral palsy: a meta- Spittle A, Brown N, Doyle L, et al. 2008a. Qual- 2012. Cerebellar volume and proton magnetic analysis. JAMA 284:1417–1424. ity of general movements is related to white resonance spectroscopy at term, and neurode- Wu YW, Croen LA, Shah SJ, et al. 2006. Cerebral matter pathology in very preterm infants. velopment at 2 years of age in preterm infants. palsy in a term population: risk factors and Pediatrics 121:e1184–e1189. Dev Med Child Neurol 54:260–266. neuroimaging ﬁndings. Pediatrics 118:690–697.

DEV DISABIL RES REV CEREBRAL PALSY—DON’T DELAY MCINTYRE ET AL 129