THE EPIDEMIOLOGY OF CEREBRAL PALSY IN BRITISH COLUMBIA by Leslie Smith B.Sc.O.T. (C), Queen’s University, 1991 THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in COMMUNITY HEALTH SCIENCE THE UNIVERSITY OF NORTHERN BRITISH COLUMBIA August 2005 © Les Smith, 2005 1^1 Library and Archives Canada Bibliothèque et Archives Canada Published Heritage Branch Direction du Patrimoine de l'édition 395 Wellington Street Ottawa ON K1A 0N4 Canada 395, rue Wellington Ottawa ON K1A 0N4 Canada Your file Votre référence ISBN: 978-0-494-28391-2 Our file Notre référence ISBN: 978-0-494-28391-2 NOTICE: The author has granted a non­ exclusive license allowing Library and Archives Canada to reproduce, publish, archive, preserve, conserve, communicate to the public by telecommunication or on the Internet, loan, distribute and sell theses worldwide, for commercial or non­ commercial purposes, in microform, paper, electronic and/or any other formats. AVIS: L'auteur a accordé une licence non exclusive permettant à la Bibliothèque et Archives Canada de reproduire, publier, archiver, sauvegarder, conserver, transmettre au public par télécommunication ou par l'Internet, prêter, distribuer et vendre des thèses partout dans le monde, à des fins commerciales ou autres, sur support microforme, papier, électronique et/ou autres formats. The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. L'auteur conserve la propriété du droit d'auteur et des droits moraux qui protège cette thèse. Ni la thèse ni des extraits substantiels de celle-ci ne doivent être imprimés ou autrement reproduits sans son autorisation. In compliance with the Canadian Privacy Act some supporting forms may have been removed from this thesis. Conformément à la loi canadienne sur la protection de la vie privée, quelques formulaires secondaires ont été enlevés de cette thèse. While these forms may be included in the document page count, their removal does not represent any loss of content from the thesis. Bien que ces formulaires aient inclus dans la pagination, il n'y aura aucun contenu manquant. Canada The Epidemiology of Cerebral Palsy in British Columbia Introduction: Cerebral palsy (CP) is a childhood condition related to a significant deficit in health for many individuals. Information provided by this research will assist health planners throughout British Columbia (EC) as resources are allocated for the treatment and care of children with CP. Objective: To quantify the incidence of CP in EC within a four-year birth cohort. Methods: The study is a population-based record linkage study of a birth cohort of EC children bom between April 1, 1991 and March 31, 1995. Cases will be identified by the presence of the International Classification of Diseases, Version 9 (lCD-9) diagnostic code “343” recorded at three years of age or older, by having the lCD-9 diagnostic code “343” recorded prior to the third birthday with two confirmatory diagnoses within the first three years of life, through a record search of the EC Medical Services Plan billing files for the fiscal years 1991/92, 1992/93, 1993/94, 1994/95. These data were linked longitudinally to all further EC Medical Services Plan activity until March 31, 2001. Results/Conclusion: This research has provided an estimate of the incidence of cerebral palsy in the four-year birth cohort 1991-1995 in British Columbia. An aggregate incidence rate of cerebral palsy was measured as 2.68 per 1000 live births, and a congenital rate was measured at 2.57 for the same population. Birth weight demonstrated a significant relationship with the development of cerebral palsy. As gestational age is highly correlated with birth weight, it may also have a significant relationship with the development of cerebral palsy. A significant relationship was not found between gender and area of residence (rural versus urban) and the development of cerebral palsy. Much research remains to be conducted in the epidemiology of cerebral palsy. 11 TABLE OF CONTENTS Abstract 11 Table of Contents 111 List of Tables Acknowledgement VI Chapter One Introduction Preamble Purpose(s) of the Study Rationale for the Study Objective(s) of the Study Research Question(s) 1 1 1 1 2 2 Chapter Two Literature Review What is Cerebral Palsy? The Epidemiology of Cerebral Palsy The Aetiology of Cerebral Palsy The Prognosis of Cerebral Palsy Conclusion 3 3 6 17 29 37 Chapter Three Methods Primary Objectives Choice of Study Design Study Population British Columbia Linked Health Database Data Linkage Linked Databases Data Analysis Incidence versus Prevalence Sample Size Data Handling and Analysis Ethical Considerations 38 38 38 39 40 41 41 43 43 44 44 45 Chapter Four Results Physician’s Codes Description of Sample Incidence of Cerebral Palsy in British Columbia Incidence of Cerebral Palsy by Birth Weight Incidence of Cerebral Palsy by Gestational Age Incidence of Cerebral Palsy per Health Region 46 46 46 47 49 50 51 111 Chapter Five Chapter Six Discussion Physician’s Diagnosing Cerebral Palsy Sample Population-Based Record Linkage Implications of the Results and Recommendations for Future Research 52 52 53 54 Conclusion 60 57 61 References IV LIST OF TABLES Table 2.1 Epidemiology of cerebral palsy by country 16 Table 2.2 Some recognized risk factors for cerebral palsy 28 Table 2.3 Aetiology of cerebral palsy of post-neonatal origin 29 Table 3.1 Cerebral palsy measures of occurrence 44 Table 4.1 Physicians making the diagnosis of cerebral palsy 46 Table 4.2 Number and proportion of children with cerebral palsy by age at first recorded diagnosis 47 Table 4.3 Number and percentage of children diagnosed with cerebral palsy by 3 years of age 47 Table 4.4 Incidence of childhood cerebral palsy by cohort inBritish Columbia 48 Table 4.5 Number and proportion of children with cerebral palsy in the cohort by birth weight 49 Table 4.6 Number and proportion of children with cerebral palsy by gestational age 50 Table 4.7 Number and proportion of children with cerebral palsy by British Columbian health region (1991-1995) 51 ACKNOWLEDGEMENT This study was made possible thanks to a grant received from the British Columbia Medical Services Foundation. Many thanks are offered to my thesis supervisor Dr. Donald Voaklander, and all members of my thesis committee. Special thanks go to the Prince George & District Child Development Centre for allowing me the time to attend classes and to adjust my schedule to allow this to occur. I would also like to thank my beloved wife, Leonor, for having the patience to put up with me, to listen to me and to support me. VI CHAPTER 1 - INTRODUCTION Preamble Cerebral palsy is considered to be the most common, most disabling chronic childhood disorder. It has been defined as an umbrella-like term referring to a group of chronic conditions characterized by non-progressive, but often changing impairments in the structure of the central nervous system that result in variable impairments in the development of a child’s motor and postural capabilities. Cerebral palsy is related to a significant deficit in health for many individuals. People with cerebral palsy also utilize a considerable number of health services. Limited research has been conducted in Canada regarding the epidemiology of cerebral palsy or health resource utilization for this population. Purpose(s) of the Study The purpose of this research is to describe the epidemiology of cerebral palsy in British Columbia. Rationale for the Study The epidemiology of cerebral palsy has not heen measured in British Columbia, based on a lack of published evidence to the contrary. The information provided by this research will assist health planners throughout this province in terms of how resourees are allocated for the treatment and care of children with cerebral palsy. The population-based approach will allow health planners to make decisions based on the most comprehensive data available in British Columbia. Objective of the Study The objective sought by this research is to quantify the incidence of cerebral palsy in British Columbia for a four-year birth cohort, dating from April 1, 1991 to March 30, 1995. Research Questions This research will attempt to answer three primary questions. First, is it possible to measure the incidence of cerebral palsy in a four-year birth cohort using linked health data? Second, what is the incidence of cerebral palsy in the four-year birth cohort from 1991-1995 in British Columbia? Third, are there any statistical differences among birth weight, gestational age and geographic area of residence with the occurrence of cerebral palsy? CHAPTER 2 - LITERATURE REVIEW This literature review provides an overview of the definitions of cerebral palsy and explores the aetiological, epidemiological and prognostic information related to cerebral palsy in the current literature. What is Cerebral Palsy? Cerebral palsy is not a single disorder with a single cause, but includes several different types of injury to a variety of areas of the developing brain (Goldstein, 2004). There have been and continue to be a number of definitions of cerebral palsy. Each definition is similar, but not exact. Cerebral palsy has been defined as an umbrella-like term referring to a group of cbronie conditions ebaraeterized by non-progressive, but often changing impairments in the structure of the central nervous system that result in variable impairments in the development of a child’s motor and postural capabilities. These impairments are often secondary to lesions or anomalies of the brain arising in the early stages of its development (Muteb, Alberman, Hagberg, Kodama, & Veliekovic Perat, 1992; Murphy, Allsopp, Deeoufle & Drews, 1993; and Koman, Smith & Shift, 2004). It has also been simply defined as a disorder of movement and posture due to a defect or lesion of the immature brain (Suzuki & Ito, 2002). A group calling itself the Surveillance of Cerebral Palsy in Europe (2000, p. 818-819), which constitutes a collaboration of 14 cerebral palsy surveys and registers, suggested that because of the variation in definitions, that it would be difficult to reach a consensual definition of what constitutes cerebral palsy. Rather they believe that any definition of cerebral palsy should include the following criteria: “(1) cerebral palsy is a group of disorders; (2) it is permanent but not unchanging; (3) it involves a disorder of movement and/or posture and motor function; (4) it is due to a non-progressive interference/lesion/abnormality; and (5) this interference/lesion/abnormality is in the developing/immature brain.” Jacobsson and Hagberg (2004) concur and suggest cerebral palsy should be thought of as a symptom complex rather than a disease. Cerebral palsy is also classified by subtype. While these subtypes have been relatively well defined throughout the years, some terms have been used synonymously. These subtypes follow the guidelines suggested by Hagberg, Hagberg and Olow (1993), and are defined as follows, according to the terms adopted by the Surveillance of Cerebral Palsy in Europe (2000): • Bilateral spastic cerebral palsy (a.k.a. quadriplegia, tetraplegia)', an abnormal pattern of posture and/or movement; increased muscle tone; an increase in pathological reflexes affeeting limbs on both sides of the body. • Spastic hemiplegic cerebral palsy: an abnormal pattern of posture and/or movement; increased muscle tone; an increase in the pathologieal reflexes affecting one side of the body. • Ataxic cerebral palsy: an abnormal pattern of posture and/or movement; a loss of orderly muscular coordination so that movements are performed with abnormal foree, rhythm and accuracy. • Dyskinetic cerebral palsy: an abnormal pattern of posture and/or movement with involuntary, uncontrolled, recurring and occasionally stereotyped movements; may be either dystonie (reduced musele aetivity with stiff movements and increased muscle tone) or choreo-athetotic (increased muscle activity with stormy movement and decreased muscle tone). • Unclassifiable: other. Some researchers have also attempted to elassify cerebral palsy based on the level of severity (Grether, Cummins & Nelson, 1992; Colver, Gibson, Hey, Jarvis, Mackie, & Richmond, 2000; Parkes, Dolk, Hill & Pattenden, 2001). These levels include very mild, mild, moderate and severe. However, there does not appear to be consensus on what constitutes these levels. Jacobsson and Hagberg (2004) believe that cerebral palsy should be classified or defined by the topographical distribution of the motor disorder (hemiplegia, diplegia, etc.) and the possible site of the lesion. Cerebral palsy occurs congenitally due to a variety of prenatal or perinatal risk factors, or it may be acquired during the post-natal period, i.e. 28 days or more after birth (Stanley et ak, 2000; and Cans, McManus, Crowley, Guillem, Platt, Johnson, and Arnaud, 2004). According to Palmer (2004), cerebral palsy is a disorder of movement and postural control that results in functional limitations, and it is these functional limitations and their effects on daily activities that become the disability that is cerebral palsy. The Epidemiology of Cerebral Palsy Cerebral palsy is considered to be one of the most common, most disabling chronic childhood disorders (Colver et ah, 2000; and Uldall, Michelsen, Topp and Madsen, 2001). It causes heavy demands on health, education and social service programs, as well as on families and the children themselves (Robertson, Svenson & Joffres, 1998; SCPE, 2000; and Uldall, Michelson, Topp and Madsen, 2001). The majority of studies cited in this paper report the prevalence of cerebral palsy as the number of cases per 1000 live births/neonatal survivors. The majority of research into the epidemiology of cerebral palsy has been based on cerebral palsy registries (Uldall et ah, 2001). Only one study has used a population-based record linkage (Robertson et ah, 1998). Two studies reported the incidence of cerebral palsy (Nottidge and Okogbo, 1991; and Meberg and Broch, 1995). Nottidge and Okogbo (1991) examined the number of referrals to a child neurology clinic in Ibadan, Nigeria from April 1980 to March 1983. They reported that 413 of 2053 referrals had a diagnosis of cerebral palsy. This yielded a percentage of 20.01% of all cases being referred to the clinic as having cerebral palsy. However, the authors inclusion criteria for what constituted cerebral palsy was very broad and these figures must be interpreted cautiously. Sciberras and Spencer (1999) measured the period prevalence of cerebral palsy in Malta over the ten-year period from January 1, 1981 to December 31, 1990. Potential children with cerebral palsy were identified using various population-health registers and education registers. In total, 134 children met their case definition of cerebral palsy. The period prevalence rate for cerebral palsy for the Maltese population was 2.4 per 1000 (Sciberras and Spencer, 1999). In 1995, Meberg and Broch published a study that traced a 20-year declining trend for cerebral palsy in Norway from 1970-89. They based their study on epidemiological data from the County of Vestfold, Norway for the period 19701989. Meberg and Broch (1995) split the group into two cohorts of children bom during the 10-year periods 1970-79 and 1980-89. They then attempted to extrapolate the total incidence of cerebral palsy. The authors determined the total incidence of cerebral palsy during this 20-year period to be 2.4 per 1000. They also noted that there was a steady decline in cerebral palsy births from 2.8 per 1000 in the first five-year period of this study (1970-74) to 2.0 per 1000 in the final five-year period (1985-89), but the decline did not achieve statistical significance. Meberg and Broch (2004) recently updated this study, extending the period investigated to 30 years, from 1970-1999. They identified 166 cases of cerebral palsy registered among 70, 824 children, arriving at a prevalence rate of 2.3 per 1000 live births. The prevalence of cerebral palsy was found to be 15 times higher in low birth weight infants than in normal birth weight infants (p.436). In a similar study, Kavcic and Veliekovic Perat (1998) examined the prevalenee of eerebral palsy in Slovenia during the birth years 1981 to 1990. Over the course of the study, 768 children with cerebral palsy were identified from Slovenia’s National Cerebral Palsy Register. The authors found a significant decreasing trend in cerebral palsy during the period 1981 to 1990 from 3.3 per 1000 live births in 1981 to 2.3 per 1000 live births in 1990. These figures are very similar to the study by Meberg and Broch (1995) in Norway. Bottos, Granato, Allibrio, Gioachin, and Puato (1999), in a study of the prevalence of eerebral palsy in two northeast Italian provinces, found that the prevalence of cerebral palsy increased from the mid-1960s to the mid-1980s, and then decreased between 1985-89. The authors identified 610 children with a diagnosis of cerebral palsy who were born in the provinces of Padua and Rovigo between 1965 and 1989. Bottos et al. (1999) recorded a crude prevalence rate for cerebral palsy of 1.82 per 1000 live births. They noted that this rate appeared to fluctuate from one year to the next during the study period. This study reported the prevalence rates in graphical form and expected the reader to extrapolate the information from the graphs provided. Thus, if one is to interpret the graphical information, the prevalence of cerebral palsy from 1965 to 1969 was about 1.5 per 1000 live births, from 1970 to 1974 was approximately 2.0 per 1000 live births, from 1975 to 1979 showed a sharp increase to roughly 2.7 per 1000 live births, and from 1980 to 1984 was equal to about 3.5 per 1000 live births. This rate then began to decrease in 1985 to 1989 to approximately 2.5 per 1000 live births. This was probably due to improved perinatal/neonatal care practices. Suzuki and Ito (2002) conducted a study looking at the overall prevalence of six-year old children with cerebral palsy in Shiga Prefecture, Japan. They identified 325 cases of six-year old children with cerebral palsy from 1977 to 1991. Suzuki and Ito (2002) found an overall prevalence of cerebral palsy of 1.34 per 1000 six-year old ehildren. The authors chose to use the age of six to identify the study population, as it is the mandatory age at which children must begin elementary sehool in Japan. This may be due to the fact that early diagnosis of cerebral palsy is sometimes diffieult to confirm prior to the age of three. Many epidemiological studies have been performed in the United Kingdom. All of these studies have been based on population-based registers of individuals with cerebral palsy. In 1995, MacGillivray and Campbell examined the changing pattern of cerebral palsy in the Avon area of southwest England during the 20year period from 1969 to 1988. They identified 489 cases of cerebral palsy during this time period, all of which were registered with the Avon Handicap Register. The authors recorded a relatively constant prevalence rate of cerebral palsy varying between 1.93 and 2.27 per 1000 births between 1969 and 1988. This yielded an overall prevalence rate for cerebral palsy of 2.06 per 1000 (MacGillivray and Campbell, 1995). Pharoah, Platt and Cooke (1996) disagree with MacGillivray and Campbell (1995) that the prevalenee rate of eerebral palsy is constant. In fact, Pharoah et al. (1996) believe that there was a sharp rise in the prevalence of cerebral palsy over the 25-year period between 1966 and 1989. The authors attributed this rise in the prevalence of eerebral palsy as being directly proportional to decreasing birth weights. Pharoah et al. (1996) agreed that the prevalenee of eerebral palsy among ehildren weighing greater than 2500 grams has remained steady at about 1.0-1.4 per 1000 neonatal survivors. However, for ehildren bom between 1500-2499 grams, there was an apparent threefold increase from approximately 4 per 1000 in the late 1960s to approximately 12 per 1000 neonatal survivors in the late 1970s. Coincidentally, ehildren weighing less than 1500 grams also showed a threefold increase in the prevalenee of cerebral palsy from 30 per 1000 to 90 per 1000 neonatal survivors during the 1970s. The prevalenee of cerebral palsy during the 1980s for this group (1000-1499 grams) remained constant. This study took place in the counties of Merseyside and Cheshire, England. The size of the study population for this 25-year period was 1612 cases of eerebral palsy, all of who were identified using a population-based register. However, Pharoah et al. do not cite an overall prevalenee of eerebral palsy for this study. In a study conducted by Sinha, Corry, Subesinghe, Wild and Levene (1997), the prevalenee of eerebral palsy was measured in an Asian community that was predominantly Pakistani Muslim, in the northeast of England. Cases of eerebral palsy were identified using the Yorkshire Regional Cerebral Palsy Register. Birth registrations were also recorded by ethnic subgroups, allowing Sinha et al. (1997) to determine the prevalenee of eerebral palsy in both the Asian and the non-Asian population within the Bradford District Health Authority (BDHA). The study was conducted between 1985 and 1987. Sinha et al. (1997) recorded an overall prevalenee of eerebral palsy of 3.87 per 1000 in the entire cohort within the Bradford District Health Authority. The researchers found that the Asian group 10 had a higher incidence of cerebral palsy (between 5.48 and 6.42 per 1000) as compared to the non-Asian group (3.18 per 1000). In 1998, Pharoah, Cooke, Johnson, King and Mutch, looked at the epidemiology of cerebral palsy in the Mersey and Oxford regions of England and in Scotland. All cases of cerebral palsy born between 1984 and 1989 were included in the study. These cases were identified using the Mersey Cerebral Palsy Register, the Oxford Region Register of Early Childhood Impairments, and the Scottish Register of Children with a Motor Deficit of Central Origin. Pharoah et al. (1998) located 1649 cases of cerebral palsy born between 1984 and 1989. They calculated a cerebral palsy prevalence of 2.1 per 1000 neonatal survivors. These results contradict that of Pharoah et al. (1996), and are similar to those of MacGillivray and Campbell (1995). Using data from the North of England Collaborative Cerebral Palsy Survey, Colver et al. (2000) ascertained 584 cases of cerebral palsy from all births between 1964 and 1993. The researchers analyzed the data in 5-year cohorts. They found a steadily increasing prevalence rate of cerebral palsy, rising from 1.68 per 1000 neonatal survivors in 1964-1968 to 2.45 per 1000 neonatal survivors in 1989-1993. Colver et al. (2000) attributed this increase in the overall rate of eerebral palsy to modern obstetric and neonatal care, citing the fact that many children bom at < 2500 grams who would have died now survive with cerebral palsy. 11 The prevalence of eerebral palsy in Northern Ireland was examined by Parkes and colleagues in 2001. They explored the epidemiology of cerebral palsy in Northern Ireland between 1981 and 1993, based on 784 cases identified using the Northern Ireland Cerebral Palsy Register. Parkes et al. (2001) estimated an overall rate of cerebral palsy between 1981 and 1993 in Northern Ireland to be approximately 2.24 per 1000 live births. This rate is similar to other reported studies (Colver et ak, 2000; Pharoah et ak, 1998; MacGillivray and Campbell 1995). Uldall, Topp and Langhoff-Roos (1997) reported an increase in the prevalence of cerebral palsy from 2.6 to 3.0 per 1000 from 1979-1982 to 1983-1986 in eastern Denmark. The data were gathered retrospectively from the Danish Cerebral Palsy Registry. In examining the data, Uldall et ak (1997) realized that the only rise in the rate of cerebral palsy was with those children born at less than 3 1 weeks gestation. The researchers concluded that the high prevalence rate of 3.0 per 1000 was due to broad inclusion criteria, higher completeness of registration, and possibly a higher risk for eerebral palsy in the Danish population. In 2001, Topp, Uldall and Greisen published another study based in eastern Denmark. They found that the prevalence rate of cerebral palsy decreased from 3.0 per 1000 in the birth year period I983-I986 to 2.4 per 1000 live births in the birth year period 1987-1990. The study population was again based upon children registered with the Danish Cerebral Palsy Register. Topp et ak (2001) identified 299 children with cerebral palsy during the study period. The researchers observed that between 1987-1990, the prevalenee of eerebral palsy decreased by 12 approximately 35% in infants bom at less than 38 weeks gestation. Thus, the rate of cerebral palsy in Denmark is similar to that in the United Kingdom. A number of studies related to the prevalenee of eerebral palsy have been reported from Sweden. In 1989, Hagberg, Hagberg and Zetterstrom expressed eoncem that the risk of cerebral palsy increased with a decrease in birth weight. They reported a decrease in the prevalence of cerebral palsy from 2.3 to 1.4 per 1000 live births during 1954-1970, and a subsequent increase in the prevalence of cerebral palsy after that time. This culminated in a prevalenee rate of 2.17 per 1000 live births in the latter part of the study spanning 1979-1982. This increase was attributed to an increase in the number of preterm births (Hagberg et ak, 1989). In a comparative study between southwest Germany and western Sweden covering a period from 1975 to 1986, the researchers found the prevalence of cerebral palsy per live births and per neonatal survivors did not differ significantly (Krageloh-Mann, Hagberg, Meisner, Sehelp, Haas, Eeg-Olofsson, Selbmann, Hagberg and Michaelis, 1994). A series of epidemiological studies were conducted in Sweden between 1993 and 2001 (Hagberg, Hagberg and Olow, 1993; Hagberg, Hagberg, Olow, and Wendt, 1996; and Hagberg, Hagberg, Beckung and Uvebrant, 2001). In the 1993 study, Hagberg et al. determined the prevalence of cerebral palsy to be 2.49 per 1000 live births for children born between 1983 and 1986. They reported this to represent an increase since 1970. In a follow-up study covering the period 1987- 13 1990, the prevalence decreased slightly to 2.36 per 1000 live births (Hagberg et ah, 1996). The latest Swedish study reported a further decrease in the prevalence of cerebral palsy of 2.12 per 1000 live births during the period 1991-1994 (Hagberg et al., 2001). This steady decrease was attributed to better obstetric and neonatal care. Two studies were found in the literature pertaining to the prevalence of cerebral palsy in the United States. In the first study, Grether, Cummins and Nelson (1992) reported on a population-based study of 192 children residing in four San Francisco Bay area counties between 1983 and 1985. These children had to he alive and residing in California at age 3 years. The authors reported a prevalence of 1.23 /1000 survivors at age 3. Inclusion criteria for this study were narrow in scope, which may explain the low prevalence rate. The second American study examined the prevalence of cerebral palsy among 10-year old children in metropolitan Atlanta from 1985 through 1987 (Murphy, Yeargin-Allsopp, Deeoufle and Drews, 1993). Murphy et al. (1993) used a record review approach to identify 204 10-year old children with cerebral palsy, resulting in a prevalence of 2.3 per 1000. Inclusion criteria was broader for this study than for other studies, as it included children who had either congenital cerebral palsy (1.9 per 1000) or acquired cerebral palsy (0.4 per 1000), i.e. the injury/ lesion/abnormality occurred after the neonatal period and was not due to a perinatal event. 14 Cerebral palsy prevalence has been recorded in Western Australia since 1956 to the present day using a cerebral palsy register (Stanley, Blair & Alberman, 2000). Rates of overall cerebral palsy have remained between 2.0-2.5 per 1000 live births (Stanley et ah, 2000). Cans et al. (2004), on behalf of the Surveillance of Cerebral Palsy in Europe (SCPE) collaborative group, reported the prevalence rate of eerebral palsy of postneonatal origin to be 1.26 per 10,000 live births. This study identified 347 cases of cerebral palsy of post-neonatal origin in children bom between 1976 and 1990 who were reported to the SCPE common database by seven different centres. Prevalence among the centres reporting ranged from 0.46 to 2.05 per 10,000. In Canada, the epidemiology of cerebral palsy has not been well documented. Robertson et al. (1998) conducted a population-based record linkage study in the province of Alberta. They identified 248 children living in the province of Alberta who were bom between April 1985 and March 1988 who were diagnosed with cerebral palsy. The researchers were able to link diagnostic codes for all fee-forservice physician claims, all hospital separations and individual birth data from the Department of Vital Statistics of the Government of Alberta in order to identify these children. This linkage enabled Robertson et al. (1998) to calculate a prevalenee rate for eerebral palsy of 2.57 per 1000 in the province of Alberta. This study included both congenital and acquired eases of cerebral palsy. The authors also calculated a prevalenee rate for congenital cerebral palsy of 2.37 per 1000. 15 In a report published by the CanChild Centre for Disability Research in Hamilton, Ontario, a projected prevalenee rate for eerebral palsy in the province of Ontario is suggested to be approximately 2.5 per 1000 (Missiuna, Smits, Rosenbaum, Woodside, and Law, 2001). Methods for obtaining this figure were not explained. Table 2.1 - Epidemiology of cerebral palsy by country Author(s) Country Study Years Rate (per 1000) Data Source Sweden Hagberg et al. (1989) Registry 1954-1970 2.17 Grether et al. (1992) United States Registry 1983-1985 1.23 Hagberg et al. (1993) Sweden Registry 1983-1986 2.49 1985-1987 Murphy et al. (1993) United States Registry 2.30 Meberg & Broch (1995) 1970-1984 Norway Registry 2.40 Registry 1969-1988 MacGillivray & Campbell (1995) United Kingdom 2.06 Sweden Hagberg et al. (1996) Registry 1987-1990 2.36 United Kingdom Registry 1985-1987 3.87 Sinha et al. (1997) Denmark Registry 1979-1986 2.60-3.00 Uldall et al. (1997) Kavcic & Veliekovic Perat (1998) Slovenia Registry 1981-1990 2.30 Pharoah et al. (1998) United Kingdom Registry 1984-1989 2.10 1985-1988 2.57 Robertson et al. (1998) Canada Record linkage Italy Clinical record search 1965-1989 Bottos et al. (1999) 2.50 Colver et al. (2000) United Kingdom Registry 1964-1993 2.45 1956-2000 Stanley et al. (2000) Australia Registry 2.00-2.50 1991-1994 Hagberg et al. (2001) Sweden Registry 2.12 Northern Ireland Registry 1981-1993 2.24 Parkes et al. (2001) Denmark Registry 3.00 Topp et al. (2001) 1983-1986 Denmark 1987-1990 Topp et al. (2001) Registry 2.40 Japan Registry 1977-1991 1.34 Suzuki & Ito (2002) 1970-1999 Meberg & Broch (2004) Norway Registry 2.30 In summary, the prevalence rate of cerebral palsy in developed and developing nations appears to be between 2-3 per 1000 live births/neonatal survivors (Table 2.1), with a few exceptions. There seems to be some disagreement among researchers as to whether congenital and acquired eases of eerebral palsy should be considered together when calculating prevalence rates for eerebral palsy. Cerebral palsy registers exist in most European countries and some American States, but do not appear to exist in Canada. Only one study (Robertson et ah, 16 1998) appeared to use data linkages in order to calculate a prevalence rate of cerebral palsy. The Aetiology of Cerebral Palsy The aetiology for the majority of cases of cerebral palsy has not heen established (Coorssen, Msall, and Duffy, 1991; Stanley, Blair, Hockey, Petterson, and Watson, 1993; Ramin, 2000; Ramin and Gilstrap, 2000; Griffin, Fitch, and Griffin, 2002). It has heen estimated that for approximately 75% of cases of cerebral palsy, a single causative factor cannot he elucidated (Pschirrer and Yeomans, 2000). Stanley et al. (2000) believe that cerebral palsy is caused by multiple factors that follow a causal pathway, and that these causal factors may occur at any time during the early stages of the brain’s development. Di Tommaso and Tranquilli (2004) have developed a checklist to identify the origin of cerebral palsy, which if used appropriately they believe will reduce the number of cases ascribed to unknown aetiology. In 1998, Badawi and colleagues undertook the challenge of defining what constitutes cerebral palsy. They found that the inclusion criteria for many of the cerebral palsy registers were different and attempted to standardize it. Badawi et al. (1998) were able to compose a list of some conditions that definitely do not meet with the definition of cerebral palsy mentioned earlier in this paper. These included neurodegenerative conditions, neuromuscular disorders, neural tube defects of the spine, and tumours. Some disorders which may be excluded as cerebral palsy also include hypotonia, some genetic syndromes that are easily recognizable such as Down syndrome, metabolic disorders that are progressive in 17 nature, and syndromes with progressive vascular defects (Badawi et al., 1998; Badawi, 2000; and Gupta and Appleton, 2001). Until the late 1970s and early 1980s, cerebral palsy was thought to be primarily caused by asphyxia during the birthing process (Al-Rajeh, Bademosi, Awada, Ismail, Al-Shammasi, and Dawodu, 1991). Recent research has shown that asphyxia has been estimated to be responsible for only about 6% to 10% of all cases of cerebral palsy (Shields and Schifrin, 1988; Pschirrer and Yeomans, 2000; McDonald and McMenamin, 2001; Lawson and Badawi, 2003; and Shevell, Majnemer, and Morin, 2003). Pschirrer and Yeomans (2000) also concluded that the criteria needed to diagnose antepartum or intrapartum asphyxia are still imprecise. Moreover, they reported “most cases of perinatal asphyxia are not followed by cerebral palsy, and most cerebral palsy is not associated with severe intrapartum asphyxia” (p. 219). Contrarily, in a study conducted on children with a birth weight >2500 grams, intrauterine hypoxia/birth asphyxia was strongly associated with an increased risk of cerebral palsy in children with no congenital abnormalities (Dite, Bell, Reddihough, Bessell, Brennecke, and Sheedy, 1998). Shevell (2004) postulated “neonatal encephalopathy at a moderate or severe level is a necessary antecedent to later cerebral palsy if intrapartum asphyxia is causally responsible for the cerebral palsy” (p. 26). He concluded that cerebral palsy is but one possible outcome of intrapartum asphyxia. Related to asphyxia is the use of Apgar scores. Apgar scores had been used in the past and, continue to be used today, as an indication of lack of oxygen during the birthing process. However, Apgar scores are not indicative of an outcome 18 resulting in cerebral palsy (MacLennan, 1999). It has been reported that 93% to 95% of infants with 5-minute Apgar scores of 0 to 3 (Normal 1-minute Apgar scores are 7-10) will not have cerebral palsy (Paneth, 2001). Coorssen et al. (1991) attempted to identify the occurrence of multiple minor malformations as an indicator for the prenatal aetiology of cerebral palsy. The researchers identified 109 cases of cerebral palsy with a suspected prenatal origin and used a modified version of the Weighted Anomaly Score to examine for any birth anomalies in this group. This group was compared to a group of 28 children with a suspected postnatal cause of cerebral palsy. The results suggested that minor malformations/anomalies appeared to occur more often in children with an identified prenatal cause of cerebral palsy than in children with a postnatal cause (Coorssen et al., 1991). Thrombophilia, which is the presence of thrombi in maternal or foetal circulation, has also been implicated as a possible cause of cerebral palsy (McDonald and McMenamin, 2001). However, in a recent population-based study examining the relationship between thrombophilia and hemiplegic cerebral palsy, no conclusive association was found (Smith, Skelton, Howard, and Levene, 2001). Conversely, other epidemiological research is accumulating evidence that links thrombosis to the development and possible cause of cerebral palsy, as recently noted by Gibson and colleagues (2003). There have been a number of reports of the association between chorioamnionitis, an inflammation of the foetal membranes, and cerebral palsy 19 (Grether and Nelson, 1997; O’Shea, Klinepeter, and Dillard, 1998; Vigneswaran, 2000; Gilstrap III and Ramin, 2000; Wu and Colford, 2000; McDonald and McMenamin, 2001). Grether and Nelson (1997) reported that the presence of chorioamnionitis and a maternal fever of >38° Celsius were associated with an increased risk of cerebral palsy in children with birth weights greater than 2500 grams. Children with birth weights under 2500 grams were not included in the study population. Wu and Colford Jr. (2000) performed a meta-analysis on previous studies reporting an association between the presence of chorioamnionitis and the occurrence of cerebral palsy or cystic periventricular leukomalacia. They concluded that their meta-analysis indicated that chorioamnionitis is a risk factor for both cerebral palsy and cystic periventricular leukomalacia. This association is further supported by Vigneswaran (2000) and Gilstrap III and Ramin (2000). Wu and her colleagues later supported their conclusion with a case-control study to clarity the relationship between chorioamnionitis and the risk of cerebral palsy. Using a study population of 231, 582 singleton infants bom alive at greater than 36 weeks bom between 19911998, the researchers were able to identify 109 children with severe spastic or dyskinetic cerebral palsy. Chorioamnionitis was diagnosed in 14% of theses children, representing an odds ratio of 3.8. According to the researchers, this indicates that clinical chorioamnionitis is an independent risk factor for cerebral palsy in term and near-term infants (Wu, Escobar, Grether, Croen, Greene, and Newman, 2003). In a recent Swedish study, Jacobsson (2004) used a population-based series of 148 preterm infants with spastic cerebral palsy, born between 1983-1990, and 20 matched them with a control group in order to examine inflammatory meehanisms that eontribute to the development of cerebral palsy. He reported an assoeiation between antenatal infection/inflammation and cerebral palsy. Wheater and Rennie (2000) examined 69 very low birth weight infants with eerebral palsy at an 18-month follow up using eranial ultrasound in an attempt to detect brain abnormalities. They reported that 39 ehildren had cranial ultrasound abnormalities while 30 ehildren had normal cranial ultrasounds. Infection also appeared to play a prominent role in their findings, with the presence of sepsis inereasing the risk of cerebral palsy fourfold (Wheater and Rennie, 2000). There has been some question as to whether preeelampsia, a toxemia of late pregnancy, is associated with an inereased risk of cerebral palsy or redueed risk of eerebral palsy (Spinillo, Capuzzo, Cavallini, Stronati, De Santolo, and Fazzi, 1997; Collins and Paneth, 1998; and Gray, Jones, and O'Callaghan, 2001). Collins and Paneth (1998) suggested that there is insuffleient evidence to provide an explanation of the relationship between preeelampsia and eerebral palsy. Gray et al. (2001), studied a group of ehildren bom between 24 and 27 weeks’ gestation over an 8-year period in order to determine obstetric risk factors for cerebral palsy. They found that maternal preeelampsia was not assoeiated with a statistically significant reduction in the risk of cerebral palsy. However, Spinillo et al. (1997), in their investigation of the relationship between the presence of preeelampsia and cerebral palsy, indieated that preeelampsia provided a protective effect against the development of cerebral palsy in the preterm infant. 21 In 1997, Dammann and Leviton used annotated evidenee to hypothesize that maternal bacterial vaginosis increased the risk of intrauterine infection. This, in turn, inereased the risk of newborn white matter damage to the brain. As a result of this newborn white matter damage, Dammann and Leviton suggested that one could predict the development of cerebral palsy in ehildren born before 37 completed weeks of gestation. This hypothesis remains to be tested. Yoon and colleagues (2003) and Grether and colleagues (2003) recently examined the relationship of intrauterine infection and the development of cerebral palsy. Yoon et al. (2003) believe that intrauterine infection can cause a foetal inflammatory response that predisposes the foetus to develop cerebral palsy, but are quick to point out that it is unlikely that the infection or the inflammation is sufficient in itself to directly cause cerebral palsy. Grether et al. (2003), in a retrospective case-control study of 170 singleton children bom between 1988 and 1994 with spastic cerebral palsy who weighed less than 1999 grams and survived to age 2 years, found that exposure to intrauterine infection was not an independent risk factor for cerebral palsy in very preterm infants when gestational age and other confounders were tightly controlled. This supports Yoon et al. (2003) in presuming that intrauterine infection does not of itself cause eerebral palsy. Hughes and Newton (1992) reported that possibly as high as 2% of cases of cerebral palsy may have a genetic basis. They based this speculation on a review of the literature. 22 The use of dexamethasone to treat preterm infants with respiratory distress syndrome has been implicated by Shinwell and colleagues (2000) as a possible cause of cerebral palsy and developmental delay. Shinwell et al. (2000) conducted a randomized, double blind, placebo controlled study of early postnatal dexamethasone treatment for prevention of chronic lung disease. They found that a three-day course of dexamethasone treatment administered shortly after birth to preterm infants with respiratory distress syndrome was associated with a significantly increased incidence of cerebral palsy and developmental delay. Because cerebral palsy is thought to be due to a lesion or abnormality in the developing brain (SCPE, 2000), some research has been conducted using magnetic resonance imaging (MRI) to retrospectively analyze the brain images of people with cerebral palsy (Sugimoto, Woo, Nishida, Araki, Kara, Yasuhara, Kobayashi, and Yamanouchi, 1995; and Okumura, Kato, Kuno, Hayakawa, and Watanabe, 1997). In their 1995 retrospective study of 70 children with cerebral palsy, Sugimoto et al. found that cerebral palsy of term infants was likely due to prenatal factors as indicated by MRI findings showing neuronal migration disorders and cerebral infarct. They did caution that MRI on its own does not always reveal when the brain damage occurred. Okumura et al. (1997) studied MRIs of the brain in 152 children aged 1 to 19 years who had spastic cerebral palsy. From these imaging studies, the researchers found the primary lesion responsible for diplegia (bilateral spastic cerebral palsy affecting two limbs) was periventricular leukomalacia. In children with quadriplegia (bilateral spastic cerebral palsy affecting four limbs), periventricular leukomalacia, term-type brain injury and brain anomalies were the main types of lesions. Okumura et al. (1997) 23 believe that “MRI of the brain is helpful in understanding the pathogenesis of eerebral palsy, although further study is neeessary” (p.372). Aeeardo, Kammann, and Hoon (2004) noted that evidenee from imaging studies show that ehildren with spastic forms of cerebral palsy often demonstrate white matter injury, whereas children with extrapyramidal syndromes (dystonia, ataxia, etc.) frequently have damage to the basal ganglia, possibly caused by conditions such as genetic metabolic disorders, kernieterus, mitochondrial disorders or hypoxic ischemic encephalopathy. A number of studies have also implicated multiple pregnancies as being a risk of developing eerebral palsy (Laplaza, Root, Tassanawipas, and Cervera, 1992; Pharoah and Cooke, 1997; O’Shea at al., 1998; and Pharoah, 2001). Laplaza et al. (1992) found that of the 1217 eases of eerebral palsy in their study, 86 or 7.1% were the result of a twin pregnancy. They reported that cerebral palsy was about six times more frequent among children bom as a result of a twin pregnancy and was more likely to occur in monozygotic versus dizygotic pregnancies. Pharoah and Cooke (1997) proposed that spastic eerebral palsy of unknown aetiology might be the result of the death of a eo-twin early in the pregnancy and may result in impairing the neurological development of the surviving twin throughout gestation. In 2001, Pharoah tested this hypothesis by analyzing the birth and death certificates for same sex and different sex twins in England and Wales for the period 1993-1995 where both twins were live births. Pharoah (2001) found that the neonatal death rate in same sex twins was 25.4 per 1000 live births and in different sex twins was 18.0 per 1000 live births. This was thought to be attributable to the higher proportion of same sex twins with low birth weight. Blair 24 and Stanley (2002) suggest that multiple pregnancies bom pre-term and at term are 1-2 and 3 times, respectively, more likely to have cerebral palsy than single pregnancies bom at the same times. Using annotated information, Blickstein (2004) noted that the association of cerebral palsy and multiple pregnancies has been observed worldwide, with the data indicating that the higher the number of the foetuses, the greater is the prevalence of cerebral palsy. In a recent study from Denmark, Pinborg and colleagues (2004) found that twins born after assisted conception had a similar risk of neurological sequelae, including cerebral palsy, as naturally conceived twins and singleton births after invitro fertilization or intracytoplasmic sperm injection. The crude prevalence rates per 1000 for cerebral palsy in twins and singleton births after assisted conception and in naturally conceived twins were 3.2, 2.5 and 4.0 respectively. It has been well documented in the literature that survival of newborn infants of low birth weight has increased the risk for the development of cerebral palsy (Skidmore, Rivers, and Hack, 1990; Stanley et al., 1993; Hagberg et al., 1993; Krageloh-Mann et al., 1995; Hagberg et al., 1996; Pharoah et al., 1996; O’Shea et al., 1998; Colver et al., 2000; Stanley et al., 2000; Dolk et al., 2001; and Hagberg et al., 2001). Hagberg et al. (1993) reported that the prevalence of cerebral palsy from 1983-1986 in Sweden was 49.8 per 1000 for a birth weight of <1500 grams compared to 14.4 for birth weight between 1500-2499 grams and 1.5 for birth weight >2500 grams. During the birth year period 1987-1990, Hagberg et al. (1996) found that infants with a birth weight <1000 grams had a cerebral palsy prevalence of 57 per 1000, 68 per 1000 for infants between 1000-1499 grams, 14 25 per 1000 for 1500-2499 grams and 1.4 per 1000 for birth weights >2500 grams. For the birth year period 1991-1994, the prevalence of cerebral palsy was 48.51 per 1000 for a birth weight <1000 grams, 80.97 for a birth weight of 1000-1499 grams, 11.42 for a birth weight of 1500-2499 grams, and 1.25 for a birth weight >2500 grams (Hagberg et al., 2001). Pharoah et al (1996) reported similar findings in the counties of Merseyside and Cheshire in England between 1966 and 1989, noting that the prevalence of cerebral palsy in low birth weight infants appeared to be increasing and tbat this apparent increase was also associated with an increase in the functional severities of the disability. Colver et al. (2000) ascertained that there was an increase in cerebral palsy in the north-east of England during the birth year period 1989-1993 for children with birth weights <1500 grams (29.8 to 74.2 per 1000 neonatal survivors) and between 1500-2499 grams (3.9 to 11.5 per 1000 neonatal survivors). Dolk et al. (2001) reported similar findings in Northern Ireland. The research indicates that because of improvements to obstetric and neonatal care over the past 30 years, children born at low and very low birth weights have an increased rate of survival, but may also have an increased risk of developing cerebral palsy (Skidmore et al., 1990; Hagberg et al., 1996; and Colver et al., 2000). It bas also been estimated that very low birth weight neonates now constitute approximately 26% to 28% of all new cases of cerebral palsy (O’Shea et al., 1998). The aetiology of congenital cerebral palsy remains largely speculative and is often not known. It would appear that the best indicator of risk for developing cerebral palsy is a newborn’s birth weight. Unfortunately, there does not appear to be a good screening test for cerebral palsy that is highly predictive. However, 26 research is ongoing in the areas of surveillance and early identification of cerebral palsy (Morgan and Aldag, 1996; and Lindstrom and Bremberg, 1997). For example, a recent Australian case-control study was able to identify neonatal risk factors associated with the development of cerebral palsy for children born at term and preterm (Walstab, Bell, Reddibougb, Brennecke, Bessell and Beiscber, 2004). The researchers found tbat for children bom at term 73% of cases and 2% of controls were likely to develop cerebral palsy if one of the following factors was present: seizures, congenital abnormalities of the brain and elsewhere, other lesions, abnormal muscle tone and meconium aspiration. For children bom preterm, the likelihood of the development of cerebral palsy for 68% of cases and 26% of controls were identified by the following factors: seizures, intraventricular haemorrhage, periventricular leukomalacia, other lesions and abnormal muscle tone. Table 2.2 provides a summary of recognized risk factors for developing cerebral palsy. 27 Table 2.2 - Some recognized risk factors for cerebral palsy Advanced maternal age Assisted reproduction Birth defects Catastrophic intrapartum events Child abuse Chromosomal anomalies Close child spacing Coagulation defects Consanguinity Cystic white matter damage Delayed childbearing Difficult delivery Domestic violence Environmental degradation Genital tract infections Infertility drugs Intrauterine growth restriction Maternal abdominal trauma Maternal mental retardation Maternal thyroid disease Multiple pregnancy Neonatal encephalopathy Nutritional deficiencies Poor education Poor hygiene Poor nutrition Poor obstetric history Poverty Preterm or very preterm birth Prolonged menstrual cycles Promiscuity Recessive genetic defects Sexually transmitted disease Social isolation Stress Substance abuse Teenage motherhood TORCH infections Toxic exposures Unemployment Reprinted with permission from Blair and Stanley (2002) (personal communieation May 31, 2004) The aetiology of cerebral palsy of post-neonatal origin is somewhat better understood, and is of interest to researchers due to it often being preventable (Cans et al., 2004). There seems to be little difference among developing and developed nations with reference to aetiology (Stanley et al., 2000). Table 2.3 provides a summary of conditions tbat have been known to cause cerebral palsy of post-neonatal origin (Stanley et al., 2000; and Cans et al., 2004). 28 Table 2.3 - Aetiology of cerebral palsy of post-neonatal origin Infection Meningitis or encephalitis Reye's syndrome Pertussis Post diphtheria-tetanus-pertussis immunization Gastroenteritis + dehydration Head injury Motor vehicle/road traffic accident Falls Traumatic head/brain injury Non-accidental injury Unspecified cause Vascular episode Post-heart surgery Post-other surgery Associated with congenital heart disease Cerebrovascular accident Miscellaneous Apparently life-threatening event Near-drowning Status epilepticus Suffocation Malnutrition Unknown Other miscellaneous events Adapted from Stanley et al. (2000) and Cans et al. (2004). The Prognosis of Cerebral Palsy When considering the prognosis of eerebral palsy, life expectancy, survival and mortality must be examined. However, there has been relatively little information published regarding the life expectancy, survival and mortality for persons with cerebral palsy (Hutton, Cooke, and Pharoah, 1994; Crichton, Mackirmon and White, 1995; and Strauss and Shavelle, 1998). Hutton et al. (1994) reported on the survival of ehildren with cerebral palsy born between 1966-1984 to mothers who were residents of the Mersey region of 29 England. They examined the effect of cognitive and motor disability, birth weight, and gestational age on life expectancy. The researchers found that the 20-year survival rate for the entire cohort was 89.3% for females and 86.9% for males. Subjects who were severely disabled had a 20-year survival of 50% as compared to those with no severe functional limitations (99%). Interestingly, Hutton et al. (1994) found that children with cerebral palsy who were of normal birth weight and who were of full term gestation were more likely to have a severe form of cerebral palsy and thus a reduced life expectancy, as compared to children with cerebral palsy who were of low birth weight and low gestational age. They also reported that birth weight and gestational age were less predictive of survival than the level of functional disability. Crichton et al. (1995) supported Hutton and colleagues’ suggestion stating that, “the most important factors influencing survival are the presence of severe mental disability and reduced mobility” (p. 567). Crichton et al. (1995) conducted a study in British Columbia on life expectancy of a cohort of all cases of cerebral palsy registered with the Health Surveillance Registry (now the Health Status Register, BC Department of Vital Statistics) who were born between 1952 and 1989. The researchers found that there was no apparent difference in the survival times between males and females, reporting a 30-year survival rate of approximately 87%. They found that there was an association of reduced survival time if the subject had profound mental retardation, a severe form of cerebral palsy, and epilepsy. 30 Strauss and Shavelle (1998) examined data on 24,768 individuals with cerebral palsy aged 15 years and over who received services in California between January 1980 and December 1995 to determine the predictors of mortality and life expectancy. The two most frequent risk predictors were lack of mobility and inability to feed oneself. They also found that the type, severity and location of cerebral palsy had a lesser effect on survival. Life expectancies for adults with high levels of functioning, i.e. full motor and feeding abilities, were approximately only 5 years less than those for the general population (Strauss and Shavelle, 1998). Strauss, Shavelle and Anderson (1998) authored a second study that examined risk factors for mortality of young children with cerebral palsy. Their study consisted of a sample of 12,709 children between the ages of 6 months to 3 years, 6 months with cerebral palsy who had received services from the state of California between 1980 and 1985. They again reported that the children with the best prognosis for survival were those with good mobility and feeding skills. Strauss et al. (1998) reported that the survival rate for children with fair motor and eating skills approached 90 % or better for reaching adulthood. Children who were unable to lift their heads when lying on their stomachs had approximately 8 times greater mortality risk than children who were able to roll/sit and who were able to walk 10 feet or more unaided. The authors also found that children who were tube fed were at a much higher risk for mortality, ranging from 3.85 at 2-3 years to 5.15 at 1 year, than children who had some self-feeding skills. Like Hutton et al. (1994), Strauss et al. (1998) found that ehildren with low birth 31 weight and gestation had better survival rates than those of normal birth weight and gestation. In 1998, a study examining the trends in mortality and cerebral palsy in a geographically-based cohort of infants weighing between 500 grams to 1500 grams bom in North Carolina between July 1982 through June 1994 was conducted (O’Shea, Preissner, Klinepeter, and Dillard, 1998). O’Shea et al. (1998) reported that mortality did not change significantly up to 1990, but then began to decrease between 1990 and 1994 for this group. They found that mortality decreased from approximately 37% from 1982 to approximately 14% in 1994. The prevalence of cerebral palsy among children of very low birth weight also decreased throughout the study period. The authors concluded that the decline in mortality among very low birth weight infants during the 1990s has not resulted in an increase in the prevalence of eerebral palsy as was reported from studies in Western Australia, England and Sweden (O’Shea et al., 1998). Williams and Alberman (1998) examined the role of severity and diagnostic labels on survival in eerebral palsy in the southeast area of England from 1980 to 1986. They found that approximately 85% of children with severe four-limb cerebral palsy survived to 15 years of age, consistent with other studies (Hutton et al., 1994; Crichton et al., 1995; and Strauss et al., 1998). The authors concluded that 15% of children with severe cerebral palsy died in childhood, suggesting that children with bilateral spastic cerebral palsy involving all four limbs are at a higher risk of mortality before they reach adolescence. 32 Strauss, Cable and Shavelle (1999) reported on diseases that may be causally linked to excess mortality in persons with cerebral palsy. They used a sample of 45,292 individuals with cerebral palsy, 4028 of whom died during the study period from 1986 to 1995. The authors obtained a standardized mortality ratio for this population of 8.4 as compared to the general population. They found that breast cancer rates were three times that of the general population, and that brain cancer rates were 24 times that of the general population, particularly for children, suggesting poorer detection and/or treatment. Standardized mortality ratios were also higher for persons with eerebral palsy for respiratory diseases, circulatory diseases, digestive tract diseases, and deaths due to external causes such as drowning or heing hit by motor vehicles (Strauss et al., 1999). In a similar study, Maudsley, Hutton and Pharoah (1999) evaluated cause of death information in people known to have cerebral palsy who had died prior to June 30, 1998, and who were registered with the Mersey Cerebral Palsy Register, a register which serves the Merseyside and Cheshire regions of England. The authors recorded mortality information on 2102 registered cases. This information was obtained from the National Health Service Central Register that requires a four digit International Classification of Disease - 9*'' Revision (lCD-9) code related to cause of death. This ICD-9 code was then linked to the case in question. Cerebral palsy was listed as the underlying cause of death in approximately 34% of cases. Another 13% of cases had the underlying cause of death listed as pneumonia. Approximately 61% of cases had a severe form of cerebral palsy (Maudsley et al., 1999). 33 Hutton, Colver and Mackie (2000) investigated the effect of motor and cognitive disabilities on the survival of persons with cerebral palsy in the Northumberland, Newcastle and North Tyneside health districts of northeast England bom between 1960 and 1990. The researchers found that approximately one-third of those with a severe disability died before the age of 30 and that onethird of the deaths were attributed to cerebral palsy on the death certificates. Similar to other studies (Hutton et al., 1994; and Crichton et al, 1995), people with a severe form of cerebral palsy, including severe mobility restrictions and cognitive limitations, have an increased risk of dying as compared to those with a less severe form of cerebral palsy, approximately 60% as compared to approximately 98% (Hutton et al., 2000). Hutton et al. (2000) also reported that persons with cerebral palsy who have severe manual disability have a risk of dying that is 25 times higher than those without. Survival was reported to be poorest for children bom weighing approximately 3000 grams, was better for larger birth weights, and was best for infants weighing less than 2500 grams. Hutton et al. (2000) believed this was due to severely damaged premature infants being less likely to survive until diagnosis than full term infants. Predictors and causes of death in persons with cerebral palsy were reported for a cohort of 2014 in Western Australia who were born between 1958 and 1994. There were 225 people who had died before May 31, 1997 (Blair, Watson, Badawi, and Stanley, 2001). Blair et al. (2001) estimated crude and standardized mortality rates for the sample. The crude mortality rate was 6.23 deaths per 1000 person years. Standardized mortality rates ranged from 0 in persons over 40 to 45.27 for children between 5 and 10 years of age. The standardized mortality rates 34 were highest for children between 1 and 15, being between 4 and 5 times greater than for persons between 15 and 40. The underlying cause of death was listed as cerebral palsy in 78.8% of cases, with respiratory disease accounting for 59% of deaths (Blair et al., 2001). Blair et al. (2001) concurred with previous studies that severe motor impairment and intellectual disability contributed significantly to mortality (Hutton et al., 1994; Crichton et al., 1995; and Hutton et al., 2000). They also found that infants bom after more than 32 weeks’ gestation were at a significantly higher risk of mortality than children born very preterm (Blair et al., 2001). The researchers concluded, “with the exception of those with profound intellectual deficit, most persons with cerebral palsy survive to adulthood” (Blair et al., 2001, p. 514). Reddibougb, Baikie and Walstab (2001) conducted a similar study in Victoria, Australia, examining the causes of death and the characteristics of children with cerebral palsy who had died between 1970 and 1995. The authors identified 155 children who had died during this period and found the predominant cause of death to be pneumonia (61/155, or 39.4%). There was an unknown eause of death in approximately 41.9% of eases. Of the 155 children who died, 89.0% had severe cerebral palsy eompared to 4.5% with moderate and 5.2% with mild eerebral palsy. Of the children who died, the majority had a birth weight >2500 grams and were bom between 37-44 weeks gestation. Most of the children who died during the study period had severe bilateral spastic cerebral palsy affecting four limbs, intellectual disability, and epilepsy (Reddibougb et al., 2001). 35 In 2002, Hutton and Pharoah published a study that attempted to quantify the effects of motor, cognitive and sensory disabilities, year of birth, birth weight, and gestational age on survival in eerebral palsy. They studied a cohort of 1942 ehildren with eerebral palsy born between 1966 and 1989 to mothers who were resident of Merseyside and Cheshire, England. They found that severe motor disability was associated with a 30-year survival of 42% and that severe cognitive disability was associated with a 30-year survival of 62% (Hutton and Pharoah, 2002). Hutton and Pharoah (2002) also concluded that survival among ehildren of low birth weight declined steadily from 1966 to 1989 after allowing for disability. In a systematic literature review dealing with life expectancy in children with eerebral palsy, Katz (2003) concluded that life expectancy depended on presence and severity of cognitive impairment, severity of physical disability, tube feeding, incontinence, and presence and severity of seizures. He also noted that life expectancy for physically and mentally disabled persons has inereased slightly with time, although this does not seem to be the result of improvements in medical care or as a result of de-institutionalization. Strauss, Ojdana, Shavelle and Rosenbloom (2004) studied the decline in function and life expectancy of 904 subjects of age 60 with cerebral palsy. They found that the abilities of person at 60 were not markedly different than younger persons with eerebral palsy. However, they also point out that persons with the most severe disabilities assoeiated with cerebral palsy rarely survive to age 60. Strauss et al. (2004) concluded that survival rates of ambulatory older adults were only moderately worse than the general population, but that survival was much 36 poorer among those older adults who had lost their mobility and the ability to dress themselves. In summary, it appears that the best predictors of a poor prognosis for cerebral palsy are the presence of severe motor and cognitive impairments, a birth weight >2500 grams, and gestation between 37-44 weeks. Most children with cerebral palsy will survive until adulthood, although the death rate is higher than in children who do not have cerebral palsy (Liptak and Aeeardo, 2004). Conclusion This literature review has presented the epidemiology, aetiology and prognosis of eerebral palsy based on information available in the current literature. Cerebral palsy occurs in approximately 2-3 per 1000 live births/neonatal survivors. Its aetiology remains largely speculative and/or unknown. Persons with cerebral palsy are likely to survive until adulthood, but those with severe motor and intellectual impairments are at a greater risk of dying. 37 CHAPTER 3 - METHODS Primary Objectives This research was conducted in order to provide an overview of the epidemiology of cerebral palsy in British Columbia in a four-year birth cohort of children born between April 1, 1991 and March 31, 1995. The objective of this research was to quantify the incidence of cerebral palsy in the four-year birth cohort for children in British Columbia. Choice of Study Design This research was a population-based record linkage study of a birth cohort of British Columbian children bom between April 1, 1991 and March 31, 1995. It employed the use of systematically collected health data, rather than using a survey or cohort approach. Systematically collected health data may be less prone to underreporting as the submission of health information has been traditionally linked to health funding. According to Robertson et al. (1998), population-based record linkage for the purpose of recording the epidemiology of cerebral palsy allows the researcher to avoid recall bias, to avoid the need for direct subject contact, to provide cost effective measurement and analysis, to generate a large sample size, and to allow for the results to be generalized to other studies of cerebral palsy. Systematic error was likely redueed as the majority of healthrelated services provided to persons with cerebral palsy are by specialist practitioners. Also, systematically recorded health data may identify some of the less severe cases that might not be identified by surveillance. 38 Study Population Inclusion Criteria: Subjects were identified by documenting diagnostic codes used by physicians and hospitals during the years reviewed for this study. The diagnostic codes being used are from the International Classification of Diseases, Version 9 (ICD-9) (International Classification of Diseases, 1977) as this was the relevant source for diagnostic codes for the time of the cohort. Cases were identified by the presence of the lCD-9 diagnostic code “343” for eerebral palsy recorded at three years of age or older, or by having the lCD-9 diagnostic code “343” recorded prior to the third birthday with two confirmatory diagnoses within the first three years of life. Each child had a minimum of six years of post-birth follow-up information in order to confirm the presence of eerebral palsy. In addition to the above codes for eerebral palsy, the following codes were also used to clarify a child’s diagnosis: congenital abnormalities of the nervous system (742), other congenital abnormalities, (743-757, 759), and progressive neurological disorders (330, 331, 334, 335, 340, 341). The following codes were searched in order to determine whether ehildren with eerebral palsy may have acquired the disorder: malignant neoplasms of the brain (191), inflammatory disease of the nervous system (320-326), other conditions of the brain and nervous system (348, 349), intracranial haemorrhage (430-432), aborted sudden infant death syndrome (798), fracture of the skull (800840), other injuries (950-959, 990-995), intracranial injury excluding skull fracture (850-854), late effects of injuries (905-909), and poisoning (960-989). These additional codes were suggested by Robertson et al. (1998). 39 Exclusion Criteria: Children whose health insurance was not continuous of the six-year follow-up period were excluded. Also, children with the ICD-9 diagnostic codes identifying chromosomal abnormalities (758.x) and spina bifida (741.x) were not included in this research. Based on the work of Robertson et al. (1998), who completed a similar study in Alberta, Canada, the estimated number of cases available for this four-year birth cohort should be approximately 450. This was based on the average yearly population during the study-subject birth years for British Columbia of 3.57 million (BC Statistics, 2004). British Columbia Linked Health Database (BCLHD) The BCLHD is housed at the Centre for Health Services and Policy Research (CHSPR) at the University of British Columbia. It is an extensive data resource for research concerning applied health services and population health (Centre for Health Services and Policy Research, 2004). If possible, the data located in each file is linkable at the individual level, allowing researchers to explore and apply population-based trends to groups of individuals over time. The information contained in the BCLHD covers the entire population of British Columbia with the data being derived from numerous sources related to health services utilization, as well as utilization of social services, health and socioeconomic status (CHSPR, 2004). The BCLHD is based on a public utility model, meaning that data are provided on a cost-recovery basis for any approved research project. All administrative data received at CHSPR are encrypted by the agencies providing it, ensuring that the privacy of the individual is protected. Upon 40 receiving the data, CHSPR checks the data and ensures that it is made linkable using probabilistic linkage. Each record in a data file is then assigned a linkage identifier ensuring that the records in the data files belonging to one individual ean be traced across different files and through time (CHSPR, 2004). Data Linkage In order to identify cases, a record search of the British Columbia (BC) Medical Services Plan billing files and the BC Hospital Separation Abstracts for the fiscal years 1991/1992, 1992/1993, 1993/1994, and 1994/1995 was conducted. The data gathered from these sources was linked longitudinally at the individual level to all further BC Medical Service Plan activity continuing until March 31, 2001. These data were longitudinally linked to any subsequent inpatient treatment recorded in the BC Hospital Separation Abstracts for the same time period. Using information from BC Vital Statistics, the cohort was linked to birth and death records. The birth records provided information on birth weight and gestational age at delivery. The death records included causes and dates of death for any of the children in the four-year cohort. Co-morbid conditions, including congenital abnormalities/anomalies and common childhood diseases such as asthma or injury, were also tabulated. Linked Databases This research used four linked databases as follows; the BC Medical Services Plan (MSP), BC Hospital Separation Abstracts, Continuing Care, and BC Vital Statistics. Each database is briefly outlined below. 41 Information gathered from MSP was through the payment information master file. The master file provides information on the annual fiscal-year files of serviees provided to MSP-covered individuals by practitioners for which MSP was billed and have paid. As this is a payment file, it represents services that have been reimbursed during the fiscal year rather than serviees provided during the fiscal year. Those billing MSP are separated in three groups: physicians, supplementary benefit practitioners (physical therapist, massage therapists, etc), and out-of-province practitioners. In addition, lCD-9 diagnostic coding is available for the fiscal years 1991/1992 and beyond. Hospital Separation Abstracts A file of hospital separations such as discharges, transfers, and deaths of inpatients and day surgery patients from acute care hospitals in BC exists. Inclusive in this file are lCD-9 diagnostic and procedure codes. Continuing Care This file provides information on client assessment and service. It includes the number and types of services received, e.g. inpatient or outpatient care, homemaking services, in-home nursing care, etc. Vital Statistics Birth and death files are available within this database. The birth file includes information such as gender, birth weight and gestational age. The death file includes information such as cause and date of death. 42 Data Analysis The incidence of cerebral palsy was calculated for each year of the birth cohort using the year-specific population of British Columbian children with health insurance registration of at least six (6) years active to infancy as the denominator. The numerator for each of the birth cohorts was the sum of the individuals in each cohort meeting the inclusion criteria for eerebral palsy who have at least six (6) years of continuous health insurance registration active to infancy. An aggregate incidence rate was calculated for congenital and acquired (post-neonatal) cerebral palsy. Finally, incidence rates were stratified by birth weight (0-999 grams, 1000-1499 grams; 1500-2499 grams; 2500-3999 grams and >4000 grams), gestational age (<32 weeks; 33-37 weeks, and >37 weeks) and geographic health region. The results were compared using the chi-square statistic, assuming a significance level of <0.05. Incidence versus Prevalence Prevalence is defined as the proportion of a given population experiencing a condition at a given time, and is measured as an aid to planning service provision. Incidence is the rate at which a condition arises in a population and is measured by those interested in aetiology and prevention. Prevalence ean be estimated from incidence rates if mortality rates and duration of the disease are known (Stanley et al., 2000). Table 3.1 summarizes the differences between prevalence and incidence. Based on this information, and because this research is studying a birth cohort, incidence was the appropriate measure to use. 43 Table 3.1 - Cerebral palsy measures of occurrence Prevalence Definition Existing/current cases in a population Service planning Used for Relationship Prevalence is incidence minus mortality and migration How measured Surveys to identify cases as a proportion of specific population Problems with Measure Very large surveys required to establish precise estimates Incidence New cases arising in a population Aetiological studies Follow up a birth cohort Register cases of cerebral palsy and relate these to their birth cohorts True incidence of brain defect or damage likely to cause cerebral palsy is not measurable Delay in ascertainment (months or years from cerebral defect) Cases lost due to deaths, out migration and other losses from birth cohorts Reprinted with permission from Stanley et al. (2000) (personal communication May 31, 2004). Sample Size This study assumed an incidence rate of cerebral palsy of 2% and an annual birth rate of 10 per 1000 population (British Columbia Vital Statistics Agency, 2002). The 95% confidence limits on this rate were ± 0.1% (Kahn and Sempos, 1989). Additionally, using this population, an incidence rate difference of 0.4% is detectable for analyses comparing those low birth weight and short gestational age to normative values (Kahn and Sempos, 1989). Data Handling and Analysis All data were entered into and analyzed using SPSS, Version 12.0 (SPSS Inc., 2003). 44 Ethical Considerations Research ethics approval has been obtained from the UNBC Research Ethics Board. The research plan has also been approved by a committee with the Health Research Development Unit at UBC. All data gained through the record linkage process will be de-identified such that no individual can be found as a result of the data linkage. All data will be stored on a password-protected computer in a locked office. 45 CHAPTER 4 - RESULTS Physician’s Codes On reviewing the physician’s codes and the accompanying ICD-9 diagnostic code for cerebral palsy, it appeared that paediatricians were responsible for recording the diagnosis in 39.2% of cases and orthopaedic surgeons in 35.6 % of cases, whereas general practitioners recorded the diagnosis 18.3 % of the time (Table 4.1). Table 4.1 - Physicians making the diagnosis of cerebral palsy____________________________ Specialty___________________________________________ # Making Diagnosis Percentage 2 Anaesthesiologist 0.4 General Practitioner 91 18.3 1 0.2 General Surgeon 3 Neurologist 0.6 1 0.2 Neurosurgeon 1 0.2 Ophthalmologist 177 35.6 Orthopaedic Surgeon 39.2 Paediatrician 195 25 5 Physical Therapist 1 0.2 Podiatrist 497 99.9 Total Description of Sample The number and proportion of children diagnosed with cerebral palsy in each of the birth cohorts is recorded in Table 4.2. Cerebral palsy was diagnosed in 497 of the children over this four-year cohort, yielding an aggregate incidence rate of 2.68 per 1000 (Table 4.4). Although there appears to be a decreasing trend in incidence rates among the cohorts, chi-square analysis for trend was not significant (%^ = 0.200, dj=\,p=Q.655\). Approximately two-fifths (41.0%) of the ehildren were diagnosed before their third birthday (Table 4.3). Twelve children (2.4%) died prior to their tenth birthday. Twenty (4.1%) children of the 485 46 confirmed living cases up to their tenth birthday had a definite post-natal event that could be considered the probable cause of the cerebral palsy. Table 4.2 - Number and proportion of children with cerebral palsy* by age at first recorded diagnosis Totai individuais Cohort (birth year) Children's Age at Diagnosis Diagnosed Cohort 1 (1991-1992) Cohort 2 (1992-1993) Cohort 3 (1993-1994) Cohort 4 (1994-1995) Total % of Total 0-1 8 13 20 25 66 13.3 1 21 26 25 12 84 16.9 2 20 10 14 8 52 10.5 3 18 21 23 21 88 16.7 4 12 14 17 19 62 12.5 5 12 15 16 14 57 11.5 6 16 9 6 7 38 7.6 7 8 9 10 2 9 9 5 9 10 1 0 3 2 0 0 5 0 0 0 26 14 10 5 5.2 2.8 2 1 132 128 126 111 497 100 includes 20 children with cerebral palsy of probable post-natal cause Table 4.3 - Number and percentage of children diagnosed with cerebral palsy by 3 years of age Age at Diagnosis_________________________________ Number__________ Percentage Below age 3 years Greater than age 3 years Total 204 293 497 41 59 100 The probable causes for motor disorder of post-natal origin were determined from the ICD-9 codes. These included traumatic brain injury (3), meningitis (2), malignant neoplasm of the brain (2), acquired paralysis (1), acquired hemiparesis (1), progressive neurological disorder (2), anomalies of the circulatory system (1), occlusion of the cerebral artery (1), and other condition of the brain and nervous system (7). Incidence of Cerebral Palsy in British Columbia An aggregate (congenital and acquired eases) incidence rate for eerebral palsy in this four-year birth cohort was calculated to be 2.68 per 1000 (95 % Cl = 2.44, 2.92). Using chi-square analysis, no significant statistical difference was found 47 amongst the cohorts = 2.68, df= 3), although the number of children diagnosed with cerebral palsy varied among the cohorts. The incidence rate of congenital cerebral palsy for this birth cohort was 2.57 per 1000 (95% Cl = 2.34, 2.80). Approximately 44% of the children diagnosed with cerebral palsy were female, and approximately 56 % were male. Table 4.4 - Incidence of childhood cerebral palsy* by cohort in British Columbia Number of Number in Cohort (birth year) Cases Cohort 132 Cohort 1 (1991-1992) 46,064 128 Cohort 2 (1992-1993) 46,175 126 Cohort 3 (1993-1994) 46,296 111 Cohort 4 (1994-1995) 47,211 497 Total 185,746 Incidence 95% Cl 2.87 2.77 2.72 2.35 2.68 2.38, 3.36 2.29, 3.25 2.25, 3.19 1.91, 2.78 2.44. 2.92 ‘ includes 20 children with probable acquired cerebral palsy. Cl = confidence interval Upon reviewing the ICD-9 codes for all contacts with the 477 children with congenital cerebral palsy, 25 (5.2%) demonstrated the code “742” for other congenital anomalies. Code “742” details conditions such as encephalocele, reduction deformities of the brain, congenital hydrocephalus, other specific abnormalities of the brain, spinal cord or nervous system, and unspecified anomalies of the brain spinal cord or nervous system (ICD-9, 1977). Fourteen children (2.9%) had one or more codes identifying other congenital anomalies, codes 743-747 and 759. Therefore, 39 children (8.2%) diagnosed with cerebral palsy thought to have occurred congenitally were coded as demonstrating congenital anomalies. Of those children with cerebral palsy thought to be due to a post-natal cause, 4 (20 %) coded 742, while 2 (10%) eoded 743-747 or 759. Children with the diagnoses spina bifida (code 741) or chromosomal 48 abnormalities (code 758) were excluded from this cohort, and were therefore not identified. Incidence of Cerebral Palsy by Birth Weight The occurrence of congenital cerebral palsy in British Columbia for the period 1991-1995 falls within the range of previously published reports (Hagberg, 1993; Hagberg et al., 1996; Pharoah et al., 1996; Colver et al., 2000; Dolk at al., 2001; and Hagberg et al., 2001). Table 4.5 illustrates the number and proportion of children with congenital cerebral palsy in this cohort by birth weight. Children born with a birth weight of less than 1000 grams in this study had an incidence rate of 66.5 per 1000 for this cohort. Children with a birth weight between 10001499 grams had an incidence of 53.1 per 1000, while those weighing between 1500-2499 grams had an incidence rate of 10.0 per 1000. For children with birth weights greater than 2500 grams, the incidence was calculated as 1.79 per 1000. Chi-square analysis was statistically significant (%^= 895.92, df= A) demonstrating the likelihood of being born with cerebral palsy may have a causal relationship with low birth weight. Table 4.5 - Number and proportion of children with cerebral palsy in the cohort by birth weight* Number of Incidence by Children with Birth Weight Birth Weight Groups Congenital in the Cohort (grams)_____________________________ Cohort SizeCerebrai Paisy Group per 1000 95% Ci 0-999 617 41 66.5 46.8, 66.2 1000-1499 829 44 53.1 37.8,68.4 1500-2499 7,782 78 10.0 7.8, 12.2 2500-3999 150,470 273 1.81 1.61,2.01 >4000_________________________________ 25,409________ £1__________ 1.61 1.11,2.11 *639 children had an unknown birth weight_______________________________________________ 49 Incidence of Cerebral Palsy by Gestational Age The number and proportion of children with congenital cerebral palsy are recorded in Table 4.6. These numbers are similar to those found in other published reports mentioned previously in this thesis. Children born at less than 26 weeks (extremely premature) in this study had an incidence rate of congenital cerebral palsy of 52.3 per 1000 in this cohort. Those bom between 26-30 weeks (very premature) had the highest incidence of cerebral palsy, measured at 67.5 per 1000. The incidence rate decreased in subsequent gestational periods from 10.8 for children born between 31-35 weeks (premature), to 1.84 for children bom between 36-40 weeks (normal gestation), to 1.75 for children born after 41 weeks gestation (longer than expected gestation). Chi-square analysis was statistically significant (%^=1,912.22, df= 4) illustrating that gestational age may be causally related to the development of cerebral palsy. Table 4.6 - Number and proportion of children with cerebral palsy by gestational age* Gestational Age (weeks) <26 26-30 31-35 36-40 >41 Total *3 cases had an unknown gestational period Cohort Size 382 992 5,470 142,188 37,529 186,561 Number of incidence Rate Children with by Gestational Age per 1000 Congenital Cerebral Paisy in Cohort 95% Cl 20 52.3 30.0, 74.6 67 67.5 51.9,83.1 10.8 8.1, 13.5 59 262 1.84 1.62,2.06 1.75 1.33,2.17 66 474 2.54 2.31,2.87 50 Incidence of Cerebral Palsy per Health Region Table 4.7 reports the number and proportion of children with cerebral paisy in British Columbia in each health region. The Interior Health Region recorded the lowest incidence rate of 2.14 per 1000 (95% Cl = 1.60, 2.68), while the Fraser Health Region recorded the highest incidence rate of 2.89 per 1000 (95% Cl = 2.48, 3.30) in this four-year birth cohort. No statistically significant difference was found among the health regions using the chi-square statistic (x^=4.97, dj^5). Therefore, no differences in the proportion of children with cerebral palsy were observed among geographic regions in the province of British Columbia. Table 4.7 - Number and proportion of children with cerebrai paisy by British Columbian health region (1991-1995)* Health Region Interior Fraser Vancouver Coastal Vancouver Island Northern Unknown * 967 British Columbian residents were born out-of-province_______ Number of Cases of Cerebral Palsy per Number of Births per Region in Incidence of Cerebral Palsy Health Region Cohort per Region 95% Cl 60 191 108 82 28,054 65,989 42,046 2.14 2.89 2.57 29,412 2.79 53 3 18,640 638 2.84 4.7 1.60, 2.68 2.48, 3.30 2.09, 3.05 2.48, 3.10 2.45, 3.23 -0.6, 10.0 51 CHAPTER 5 - DISCUSSION This thesis has presented a population-based record linkage study from administrative data, similar to that of Robertson et al. (1998). It has provided epidemiological information on cerebral palsy for children bom in the early 1990s, which can be compared to other epidemiological studies conducted in different geographic areas throughout the world (Al-Rajeh et al., 1991; Nottidge and Okogbo, 1991; Grether et al., 1992; Hagberg et al., 1993; Murphy et al., 1993; Krageloh-Mann et al., 1994; MacGillivray and Campbell, 1995; Meberg and Broch, 1995; Hagberg et al., 1996; Pharoah et al., 1996; Sinha et al., 1997; Topp et al., 1997; Kavcic and Perat, 1998; Pharoah et al., 1998; Robertson et al., 1998; Bottos et al., 1999; Sciberras and Spencer, 1999; Colver et al., 2000; Stanley et al., 2000; Surveillance of Cerebral Palsy in Europe, 2000; Hagberg et ah, 2001; Parkes et ah, 2001; Topp et ah, 2001; Suzuki and Ito, 2002; and Cans et al., 2004), although it’s specific focus is British Columbia, Canada. Physician’s Diagnosing Cerebral Palsy In contrast to Robertson et al. (1998), the number of general practitioners making or confirming the diagnosis of cerebral palsy in this study was quite high (18.3% versus 1.1%). This could occur for a number of reasons. Unlike in Alberta (Robertson et ah, 1998), children with eerebral palsy are not normally referred to a tertiary care centre for a multidisciplinary assessment due to distance to the tertiary care facilities and cost involved for travel. Therefore, it is difficult for the diagnosis of eerebral palsy to be made by specialists such as developmental paediatricians, paediatric neurologists, or paediatric neurosurgeons unless children and their families live in Vancouver or Victoria where the tertiary care centres are 52 located. Rather, it is more common to have multidiseiplinary rehabilitation assessments for children with cerebral palsy conducted by rehabilitation professionals through a Child Development Centre, located in many communities throughout British Columbia, or at one of the tertiary care facilities. In 57.5% of the eases in British Columbia, cerebral palsy has been diagnosed or confirmed by either a paediatrician or a general practitioner. Due to the complex geography of British Columbia, it is likely that these physicians are located in the local community, rather than a tertiary care facility. Interestingly, in over 35% of eases, paediatric orthopaedic surgeons either made or confirmed the diagnosis of eerebral palsy (Table 4.1). This did not occur in the study conducted by Robertson et al. (1998). These surgeons only practiced at the tertiary care facilities during the period of this study, and are likely confirming the diagnosis at the time of surgery, rather than initially identitying it. Sample The aggregate rate of cerebral palsy in British Columbian children born between 1991 and 1995 was measured at 2.68 per 1000. The congenital rate was measured at 2.57 per 1000 for the same cohort. These rates fall within the reported range of 2.0-3.0 per 1000 from previous studies throughout the world (Al-Rajeh et al., 1991; Nottidge and Okogbo, 1991; Grether et al., 1992; Hagberg et al., 1993; Murphy et al., 1993; Krageloh-Mann et al., 1994; MacGillivray and Campbell, 1995; Meberg and Broch, 1995; Hagberg et al., 1996; Pharoah et al., 1996; Sinha et al., 1997; Topp et al., 1997; Kavcic and Perat, 1998; Pharoah et al., 1998; Robertson et al., 1998; Bottos et al., 1999; Sciberras and Spencer, 1999; Colver et al., 2000; Stanley et al., 2000; Surveillance of Cerebral Palsy in Europe, 2000; 53 Hagberg et al., 2001; Parkes et al., 2001; Topp et al., 2001; Suzuki and Ito, 2002; and Cans et al., 2004). The incidence rates reported for children with cerebral palsy in this cohort by birth weight (<1000 grams = 66.5, 1000-1499 grams = 53.1, and 1500-2499 grams = 10.0) were similar to those reported by Hagberg et al. (2001) who studied a cohort bom between 1991 and 1994 in Sweden. More than 65% of children in this study had a birth weight of 2500 grams or greater, similar to that found in the other Canadian study (Robertson et al., 1998). Population-Based Record Linkage At the time of writing, this study was only the second one to use populationbased record linkage to present epidemiological information about the incidence of childhood cerebral palsy (Robertson et al., 1998). It was also the second Canadian-based study to accomplish this. All previous studies were based on information gathered from cerebral palsy registries or exhaustive record searches (Al-Rajeh et al., 1991; Nottidge and Okogbo, 1991; Grether et al., 1992; Hagberg et al., 1993; Murphy et al., 1993; Krageloh-Mann et al., 1994; MacGillivray and Campbell, 1995; Meberg and Broch, 1995; Hagberg et al., 1996; Pharoah et al., 1996; Sinha et al., 1997; Topp et al., 1997; Kavcic and Perat, 1998; Pharoah et al., 1998; Robertson et al., 1998; Bottos et al., 1999; Sciberras and Spencer, 1999; Colver et al., 2000; Stanley et al., 2000; Surveillance of Cerebral Palsy in Europe, 2000; Hagberg et al., 2001; Parkes et al., 2001; Topp et al., 2001; Suzuki and Ito, 2002; and Cans et al., 2004). 54 The use of population-based record linkage has a number of advantages over the other methods. These include the ability to provide large sample sizes, reduce the likelihood of recall bias, allow researchers to use the data for various types of studies like outcomes research and cohort studies, and eliminate the need for direct subject contact. Record linkage of administrative data is also cost effective and permits the researcher to generalize results, as administrative databases encompass large groups of people(Roos, Nicol and Cageorge, 1987; Lohr, 1990; Roos, Sharp and Cohen, 1991; Romano, Roos, Luft, Jollis, Doliszny, 1994; Stanley, Croft, Gibbins and Read, 1994; Ellenbecker, Wagner, and Cloutterbuck, 1997; Steinwachs, Stuart, Scholle, Starfield, Fox, and Weiner, 1998; Fleming and Rohrs, 1998; Melfi, 2001; Best, Khuri, Phelan, Hur, Flenderson, Demakis, and Daly, 2002; lezzoni, 2002; and Price, Estrada, and Thompson, 2003). Like all data sources, limitations also exist with record linkage of administrative data such as inaccurate coding of the diagnosis or procedures made by multiple physicians, underreporting of comorbid conditions, difficulty in recording the timing of events and causal inferences, and limited information about health status and clinical outcomes (Lloyd and Rissing, 1985; Roos et al., 1987; lezzoni, 1990; Fisher, Baron, Malenka, Barrett, and Bubolz, 1990; Lohr, 1990; Deyo, Cherkin and Ciol, 1992; Fisher, Whaley, Krushat, Malenka, Fleming, Baron, and FIsia, 1992; Harman, Kilburn, Lindsey, and Lewis, 1992; Quam, Ellis, Venus, Clouse, Taylor, and Leatherman, 1993; lezzoni, Daley, Heeren, Foley, Hughes, Fisher, Duncan, and Coffman, 1994; Romano et al., 1994; Stanley et al., 1994; Cohen, Roos, DeCoster, Black and Decker, 1995; Ellenbecker et al., 1997; Hawker, Coyte, Wright, Paul and Bombardier, 1997; lezzoni, 1997; Llixhauser, 55 Steiner, Harris and Coffey, 1998; Fleming and Kohrs, 1998; Steinwachs et a l, 1998; Powell, Lim and Heller, 2001; lezzoni, 2002; Twiggs, Fifleld, Apter, Jackson and Cushman, 2002; Price, Estrada and Thompson, 2003; and Romano, 2003). This was avoided in this study by using a unique identifier for each case. These databases may not accurately reflect or record the prevalence or incidence of disease in a population (Cans, Surman, McManus, Coghlan, Hensey, and Johson, 2004b). For this study, data on this childhood cohort was linked from the BC Vital Statistics database, the BC Medical Services Plan (MSP), and the BC Hospital Separation Abstracts. It is known that approximately 86% of the standard ICD-9 codes are covered in the MSP claims database, demonstrating that the information in the MSP database is relatively accurate and reliable. Also, greater than 96% of paid services and paid amounts to physicians are associated with the lCD-9 codes contained in the MSP database (Hu, 1996). A population-based record linkage study on childhood cerebral palsy has been conducted before in Canada (Robertson et al., 1998), and is not without its difficulties. This study attempted to overcome such challenges as early diagnosis, early death of children, and the lack of universally accepted diagnostic criteria by ensuring that children in this cohort would be followed for six years, active to infancy, for any claims against their health insurance registration. This study was able to identify the physician’s specialty via a unique identifier, allowing one to determine the number and percentage of physician’s diagnosing cerebral palsy. In the same respect, by using the ICD-9 diagnostic code “343” for fee-for-service 56 physician visits and hospital separations, it was possible to identify ehildren with eerebral palsy, in partieular, ehildren with more obvious forms of eerebral palsy as this population is more likely to reeeive the eare of a physieian. However, it is likely that the diagnosis eould have been made at an earlier time, but was not recorded until the ehildren were admitted to hospital, and subsequently discharged. This difficulty in affirming the diagnosis may have been averted if there existed in British Columbia a speeific eerebral palsy register, with the aim of identifying all cases of cerebral palsy within the geographieal population of British Columbia regardless of the need for hospital treatment (Cans et al., 2004b). Implications of the Results and Recommendations for Future Researeh Now that the ineidenee of eerebral palsy has been estimated for congenital cerebral palsy as 2.57 per 1000 live births and 2.68 for the aggregate rate in this eohort, important epidemiological evidence has been provided to add to the ineidenee and/or prevalence rates of this childhood disorder throughout Canada and the rest of North Ameriea. It is still evident that eerebral palsy affeets a large number of ehildren in the provinee of British Columbia, even with advanees in neonatal and perinatal care. Neonatal intensive eare units are loeated within regional and provineial health regions, yet the rate of eerebral palsy remains the same as it did 30 years ago. This study ean now be used to eompare with other population-based record linkage studies in Canada and throughout other parts of the world. The results of this study ean also contribute valuable information for future researeh. It has 57 provided a baseline from which one can now begin to conduct research such as measuring the use of continuing care services by children with cerebral palsy. It could also lead researchers to explore the relationship of variables such as maternal age, parity, gestational age, birth weight, rural residence, socioeconomic status, gender and maternal health status to the development of cerebral palsy. All of this information could be important to health planners throughout British Columbia in determining how to allocate health and social service funding for ehildren with cerebral palsy. This research has also provided a basis for the exploration of such things as measuring health and social outcomes of this population, specifically looking at quality of life issues, life expectancy and prognosis. The information provided by this research may also lend credence to the establishment of a cerebral palsy register in the province of British Columbia. This would allow Canadian researchers the same advantages as those in Europe and the United Kingdom where cerebral palsy registers have been established for many years, specifically, the ability to investigate the origins of cerebral palsy using a case-control design because a register would provide an unbiased sample of cases from a geographically defined population (Cans et ah, 2004b). A register would also allow researchers to monitor trends related to cerebral palsy, monitor quality of life of children with cerebral palsy, and provide longitudinal outcome data related to the costs of caring for children with cerebral palsy (Cans et ah, 2004b; Liptak and Accardo, 2004: and Majnemer and Mazer, 2004). Alternatively, this information could be combined with that contained in the 58 British Columbia Health Status Registry to provide a more comprehensive report of the occurrence of cerebral palsy in British Columbia, as has already been done with other chronic childhood conditions such as Down syndrome and autism. 59 CHAPTER 6 - CONCLUSION This research has provided an estimate of the incidence of cerebral palsy in the four-year birth cohort 1991-1995 in British Columbia. An aggregate ineidenee rate of cerebral palsy was measured as 2.68 per 1000 live births, and a congenital rate was measured at 2.57 for the same population. Birth weight demonstrated a signifieant relationship with the development of cerebral palsy. As gestational age is highly correlated with birth weight, it may also have a significant relationship with the development of eerebral palsy. A significant relationship was not found between gender and geographic area of residence and the development of cerebral palsy. Much research remains to be conducted in the epidemiology of cerebral palsy. 60 REFERENCES Accardo, J., Kammann, H. and Hoon Jr., A. H. (2004) Neuroimaging in cerebral palsy. The Journal o f Pediatrics, 145, p. S19-S27. Al-Rajeh, S., Bademosi, O., Awada, A., Ismail, H., Al-Shammasi, S. and Dawodu, A. (1991) Cerebral palsy in Saudi Arabia: A case-control study of risk factors. Developmental M edicine and Child Neurology, 33, p. 1048-1052. Badawi, N. (2000) The international consensus statement on cerebral palsy causation. M edicalJournal o f Australia, 172, p. 199-200. Badawi, N., Watson, L., Petterson, B., Blair, E., Slee, J., Haan, E. and Stanley, F. (1998) What constitutes cerebral palsy? Developmental Medicine and Child Neurology, 40, p. 520-527. Best, W. R., Khuri, S. P., Phelan, M., Hur, K., Henderson, W. G., Demakis, J. G., and Daley, J. (2002) Identifying patient preoperative risk factors and postoperative adverse events in administrative databases: results from the Department of Veterans Affairs National Surgical Quality Improvement Program. Journal o f the American College o f Surgeons, 194 (3), p. 257- 266. Blair, E. and Stanley, F. (2003) Causal pathways to cerebral palsy. Current Paediatrics, 12, p. 179-185. 61 Blair, E., Watson, L., Badawi, N., and Stanley, F. J. (2001) Life expectancy among people with cerebral palsy in Western Australia. Developmental M edicine and Child Neurology, 43, p. 508-515. Blickstein, I. (2004) Do multiple gestations raise the risk of cerebral palsy? Clinics in Perinatology, 31, p. 395-408 Bottos, M., Granato, T., Allibrio, G., Gioachin, C., and Puato, M. L. (1999) Prevalence of cerebral palsy in northeast Italy from 1965 to 1989. Developmental Medicine and Child Neurology, 41, p. 26-39. British Columbia Vital Statistics Agency (2002). Government o f British Columbia Information Bulletin, Issue 042, http://os8150.pb.gov.bc.ca/4dcgi/nritem75535. Cans, C., McManus, V., Crowley, M., Guillem, P., Platt, M.-J., Johnson, A., and Arnaud, C. (2004) Cerebral palsy of post-neonatal origin: characteristics and risk factors. Paediatric and Perinatal Epidemiology, 18, p. 214-220. Cans, C., Surman, G., McManus, V., Coghlan, D., Hensey, O., and Johnson, A. (2004b) Cerebral palsy registries. Seminars in Pediatric Neurology, 11, p. 18-23. Centre for Health Services and Policy Research (2004) BC Linked Health Database, http ://www.chspr.ubc.ca/B clhd/. 62 Cohen, M. M., Roos, N. P., DeCoster, C., Black, C., and Decker, K. M. (1995) Manitoba’s population-based databases and long-term planning: Beyond the hospital databases. Healthcare M anagement Forum, 8 (3), p. 5 - 13. Collins, M. and Paneth, N. (1998) Preeclampsia and cerebral palsy: Are they related? Developmental M edicine and Child Neurology, 40, p. 207-211. Colver, A. P., Gibson, M., Hey, E. N., Jarvis, S. N., Mackie, P. C., and Richmond, S. (2000) Increasing rates of cerebral palsy across the severity spectrum in north­ east England \96A-\99?). Archives o f D isease in Childhood Fetal and Neonatal Edition, 83, p. F7-F12. Coorssen, E. A., Msall, M. E. and Duffy, L. C. (1991) Multiple minor malformations as a marker for prenatal etiology of cerebral palsy. Developm ental Medicine and Child Neurology, 33, p. 730-736. Crichton, J. U., Mackinnon, M., and White, C. P. (1995) The life expectancy of persons with cerebral palsy. Developmental Medicine and Child Neurology, 37, p. 567-576. Dammann, O. and Leviton, A. (1997) Does prepregnaney bacterial vaginosis increase a mother’s risk of having a preterm infant with cerebral palsy? Developmental M edicine and Child Neurology, 39, p. 836-840. 63 Deyo, R. A., Cherkin, D. C., and Ciol, M. A. (1992) Adapting a clinieal comorbidity index for use with ICD-9-CM administrative databases. Journal o f Clinical Epidemiology, 45 (6), p. 613 - 619. Dite, G. S., Bell, R., Reddihough, D. S., Bessell, C., Brennecke, S., and Sheedy, M. (1998) Antenatal and perinatal antecedents of moderate and severe spastic cerebral palsy. Australian and New Zealand Journal o f Obstetrics and Gynaecology, 38, p. 377-383. Di Tommaso, M. and Tranquilli, A. L. (2004) A checklist to identify the origin of cerebral palsy. Journal o f M aternal-Fetal and Neonatal Medicine, 15, p 281-286. Elixbauser, A., Steiner, C., Harris, R., and Coffey, R. M. (1998) Comorbidity measures for use with administrative data. M edical care, 36 (1), p. 8 - 27. Ellenbecker, C. H., Wagner, E., and Cloutterbuck, J. (1997) Using insurance claims data and medical record reviews to assess the quality of medical care. Journal o f Healthcare Quality, 19 (3), p. 21 - 31. Fisher, E. S., Baron, J. A., Malenka, D. J., Barrett, J., and Bubolz, T. A. (1990) Overcoming potential pitfalls in the use of Medicare data for epidemiologic research. American Journal o f Public Health, 80 (12), p. 1487 - 1490. Fisher, E. S., Whaley, F. S., Krushat, W. M., Malenka, D. J., Fleming, C., Baron, J. A., and Hsia, D. C. (1992) The aceuracy of Medieare’s hospital claims data: 64 Progress has been made, but problems remain. American Journal o f Public Health, 82 (2), p. 243 - 248. Fleming, S. T. and Kobrs, F. P. (1998) Linking claims data and cancer registry data: Is it worth the effort? Clinical Performance and Quality Health Care, 6 (2), p. 88 —96. Gibson, C. S., MacLennan, A. H., Goldwater, P. N., and Dekker, G. A. (2003) Antenatal causes of cerebral palsy: Associations between inherited thrombophilias, viral and bacterial infection and inherited susceptibility to infection. O bstetrical and G ynecological Survey, 58, p. 209-220. Goldstein, M. (2004) The treatment of cerebral palsy: what we know, what we don’t know. The Journal o f Pediatrics, 145, p. S42-S46. Government of British Columbia (2004) B C Stats: Population & Demographics: H istorical Population Estimates - BC 1971-2003 (Age & Sex). http://www.bcstats.sov.be.ca/data/Dop/poD/BCPoDage.htm. Gray, P. H., Jones, P., and O’Callaghan, M. J. (2001) Maternal antecedents for cerebral palsy in extremely preterm babies: a case-control study. Developmental Medicine and Child Neurology, 43, p. 580-585. Grether, J. K., Cummins, S. K., and Nelson, K. B. (1992) The California cerebral palsy project. Paediatric and Perinatal Epidemiology, 6, p. 339-351. 65 Grether, J. K. and Nelson, K. B. (1997) Maternai infection and cerebral palsy in infants of normal birth weight. Journal o f the American M edical Association, 278, p. 207-211. Grether, J. K., Nelson, K. B., Walsh, E., Willoughby, R. E., and Redline, R. W. (2003) Intrauterine exposure to infection and risk of cerebral palsy in very preterm infants. Archives o f Pediatric and Adolescent Medicine, 157, p. 26-32. Griffin, H. C., Fitch, C. L. and Griffin, E. W. (2000) Causes and interventions in the area of cerebral palsy. Infants and Young Children, 14, p. 18-23. Gupta, R. and Appleton, R. E. (2001) Cerebral palsy: Not always what it seems. Archives o f D isease in Childhood, 85, p. 356-360. Hagberg, B., Hagberg, G., Beckung, E., and Uvebrant, P. (2001) Changing panorama of cerebral palsy in Sweden. VIII. Prevalence and origin in the birth year period 1991-94. Acta Paediatrica, 90, p. 271-277. Hagberg, B., Hagberg, G., and Glow, I. (1993) The changing panorama of cerebral palsy in Sweden. VI. Prevalence and origin during the birth year period 1983-1986. Acta Paediatrica, 82, p. 387-393. 66 Hagberg, B., Hagberg, G., Olow, L, and Wendt, L. (1996) The changing panorama of cerebral palsy in Sweden. VII. Prevalence and origin in the birth year period 1987-90. A cta Paediatrica, 85, p. 954-960. Hagberg, B., Hagberg, G., and Zetterstrom, R. (1989) Decreasing perinatal mortality - Increase in cerebral palsy morbidity? Acta Paediatrica, 78, p. 664670. Hannan, E. L., Kilburn, Jr., H., Lindsey, M. L., and Lewis, R. (1992) Clinical versus administrative databases for CAB G surgery. M edical Care, 30 (10), p. 892 -9 0 7 . Hawker, G. A., Coyte, P. C., Wright, J. G., Paul, J. L., Bombardier, C. (1997) Accuracy of administrative data for assessing outcomes after knee replacement surgery. Journal o f Clinical Epidemiology, 50 (3), p. 265 - 273. Hu, W. (1996) Diagnostic codes in MSP claim data: Summary report. Program Monitoring and Information Management Branch, resource Management Division, Medical Services Plan. Hughes, I. and Newton, R. (1992) Genetic aspects of cerebral palsy. Developmental M edicine and Child Neurology, 34, p. 80-86. 67 Hutton, J. L., Colver, A. F., and Mackie, P. C. (2000) Effect of severity of disability on survival in northeast England cerebral palsy cohort. Archives o f D isability in Childhood, 83, p. 468-474. Hutton, J. L., Cooke., and Pharoah, P. O. D. (1994) Eife expectancy in children with cerebral palsy. British M edical Journal, 309, p. 431-435. Hutton, J. E. and Pharoah, P. O. D. (2002) Effects of cognitive, motor, and sensory disabilities on survival in cerebral palsy. Archives o f D isease in Childhood, 86, p. 84-90. lezzoni, E. I. (1990) Using administrative diagnostic data to assess the quality of care: Pitfalls and potential of ICD-9-CM. International Journal o f Technology Assessment in Health Care, 6, p. 272 - 281. lezzoni, E.I. (1997) Assessing quality using administrative data. Annals o f Internal M edicine, 127 (8, Part 2), p. 666 - 674. lezzoni, E. I. (2002) Using administrative data to study persons with disabilities. The Milbank Quarterly, 80 (2), p. 347 - 379. lezzoni, L. 1., Daley, J. Heeren, T., Foley, S. M., Hughes, J. S., Fisher, E. S., Duncan, C. C., and Coffman, G. A. (1994) Using administrative data to screen hospitals for high complication rates. Inquiry, 31, p. 40 - 55. 68 International Classification of Diseases (1977) Manual of the International Statistical Classification of Diseases, Injuries and Causes of Death, 9*^ Revision. Jacobsson, B. (2004) Infectious and inflammatory mechanisms in preterm birth and cerebral palsy. European Journal o f Obstetrics & Gynecology and Reproductive Biology, 115, p. 159-160. Jacobsson, B. and Hagberg, G. (2004) Antenatal risk factors for cerebral palsy. Best Practice & Research Clinical Obstetrics and Gynaecology, 18, p 425-436. Kahn, H. A. and Sempos, C. T. (1989) Statistical Methods in Epidemiology. Oxford University Press, Toronto. Katz, R. T. (2003) Life expectancy of children with cerebral palsy and mental retardation: Implications for life care planning. NeuroRehabilitation, 18, pp. 261270. Kavcic, A and Velickovic Perat, M. (1998) Prevalence of cerebral palsy in Slovenia: birth years 1981 to 1990. D evelopmental Medicine and Child Neurology, 40, p. 459-463. Kragelob-Mann, I., Hagberg, G., Meisner, C., Haas, G., Eeg-Olofsson, K. E., Selbmann, H. K., Hagberg, B., and Micbaelis, R. (1994) Bilateral spastic cerebral palsy - A comparative study between south-west Germany and western Sweden. II: epidemiology. Developmental Medicine and Child Neurology, 36, p. 473-483. 69 Koman, L. A., Smith, B. P., and Shilt, J. S. (2004) Cerebral palsy. The Lancet, 363, p. 1619-1631. Laplaza, F. J., Root, L., Tassanawipas, A., Cervera, P. (1992) Cerebral palsy in twins. Developmental Medicine and Child Neurology, 34, p. 1053-1063. Lawson, R. D. and Badawi, N. (2003) Etiology of cerebral palsy. H and Clinics, 19, p. 547-556. Lindstrom, K. and Bremberg, S. (1997) The contribution of developmental surveillance to early detection of cerebral palsy. Acta Paediatrica, 86, p. 736-739. Liptak, G. S. and Accardo, P. J. (2004) Health and social outcomes of children with cerebral palsy. The Journal o f Pediatrics, 145, p. S36-S41). Lloyd, S. S. and Hissing, J. P. (1985) Physician and coding errors in patient records. Journal o f the American M edical Association, 254 (10), p. 1330 - 1336. Lobr, K. N. (1990) Use of insurance claims data in measuring quality of care. International Journal o f Technology Assessment in Health Care, 6, p. 263 - 271. MacGillivray, 1. and Campbell, D. M. (1995) The changing pattern of cerebral palsy in Avon. Paediatric and Perinatal Epidemiology, 9, p. 146-155. 70 MacLennan, A. (2001) A template for defining a causal relation between acute intrapartum events and cerebral palsy: International consensus statement. British M edicalJournal, 319, p. 1054-1059. Majnemer, A., and Mazer, B., (2004) New directions in the outcome evaluation of children with cerebral palsy. Seminars in Pediatric Neurology, 11, p. 11-17. Maudsley, G., Hutton, J. L., and Pharoah, P. O. D. (1999) Cause of death in cerebral palsy: A descriptive study. Archives o f Disease in Childhood, 81, p. 390394. McDonald, D. G. M. and McMenamin, J. B. (2001) Moving beyond birth asphyxia as the cause of cerebral palsy. Irish M edicalJournal, 94, p. 68-69. Meberg, A. and Broch, H. (1995) A changing pattern of cerebral palsy. Declining trend for incidence of cerebral palsy in the 20-year period 1970-89. Journal o f Perinatal M edicine, 23, p. 395-402. Meberg, A. and Broch, H. (2004) Etiology of cerebral palsy. Journal o f Perinatal Medicine, 32, p. 434-439. Missiuna, C., Smits, C., Rosenbaum, P., Woodside, J., and Law, M. (2001) Cerebral palsy. In Report to the Ministry of Health on the Incidence and Prevalence of Childhood Disability. Hamilton: CanChild. 71 Morgan, A. M. and Aldag, J. C. (1996) Early identification of cerebral palsy using a profile of abnormal motor patterns. Pediatrics, 98, p. 692-697. Murphy, C. C., Yeargin-Allsopp, M., Decoufle, P., and Drews, C. D. (1993) Prevalence of cerebral palsy among ten-year-old children in metropolitan Atlanta, 1985 through 1987. Journal o f Pediatrics, 123, p. S13-S20. Mutch, L., Alberman, E., Hagberg, B., Kodama, K., and Velickovic Perat, M. (1992) Cerebral palsy epidemiology: Where are we now and where are we going? Developmental M edicine and Child Neurology, 34, p. 547-555. Nottidge, V. A. and Okogbo, M. E. (1991) Cerebral palsy in Ibadan, Nigeria. Developmental M edicine and Child Neurology, 33, p. 241-245. Okumura, A., Kato, T., Kuno, K., Hayakawa, F., and Watanabe, K. (1997) MRI findings in patients with spastic cerebral palsy. 11: Correlation with type of cerebral palsy. D evelopmental M edicine and Child Neurology, 39, p. 369-372. O’Shea, T. M., Klinepeter, K. E., and Dillard, R. G. (1998) Prenatal events and the risk of cerebral palsy in very low birth weight infants. American Journal o f Epidemiology, 147, p. 362-369. O’Shea, T. M., Preissner, J. S., Klinepeter, K. L., and Dillard, R. G. (1998) Trends in mortality and cerebral palsy in a geographically based cohort of very low birth weight neonates bom between 1982 to 1994. Pediatrics, 101, p. 642-647. 72 Palmer, F. B. (2004) Strategies for the early diagnosis of cerebral palsy. The Journal o f Pediatrics, 145, p. S8-S11. Paneth, N. (2001) Cerebral palsy in term infants - Birth or before birth? Journal o f Pediatrics, 138, p. 791-792. Parkes, J., Dolk, H., Hill, N., and Pattenden, S. (2001) Cerebral palsy in Northern Ireland: 1981-93. Paediatric and Perinatal Epidemiology, 15, p. 278-286. Pharoah, P. O. D. (2001) Cerebral palsy in the surviving twin associated with infant death of the co-twin. Archives o f D isease in Childhood Fetal and Neonatal Edition, 84, p. FI 11-FI 15. Pharoah, P. O. D. and Cooke, R. W. 1. (1997) A hypothesis for the aetiology of spastic cerebral palsy - The vanishing twin. Developmental Medicine and Child Neurology, 39, p. 292-296. Pharoah, P. O. D., Cooke, T., Johnson, M. A., King, R., and Mutch, L. (1998) Epidemiology of cerebral palsy in England and Scotland, 1984-89. Archives o f D isease in Childhood Fetal and Neonatal Edition, 79, p. F21-F25. Pharoah, P. O. D., Platt, M. J., and Cooke, T. (1996) The changing epidemiology of cerebral palsy. Archives o f D isease in Childhood Fetal and Neonatal Edition, 75,p. F169-F173. 73 Pinborg, A., Loft, A., Schmidt, L., Greisen, G., Rasmussen, S. & Andersen, A. N. (2004) Neurological sequelae in twins born after assisted coneeption: eontrolled national cohort study. British M edical Journal, BMJ, doi: 10.1136/bmj.38156.715694.3A (published 15 July 2004), p 1-6. Powell, H., Lim, L. L.-Y., and Heller, R. F. (2001) Accuracy of administrative data to assess comorbidity in patients with heart disease: An Australian perspective. Journal o f Clinical Epidemiology, 54, 687 - 693. Price, J. P., Estrada, C. A., Thompson, D. (2003) Administrative data versus eorreeted administrative data. American Journal o f M edical Quality, 18 (1), p. 38 -4 5 . Pschirrer, E. R. and Yeomans, E. R. (2000) Does asphyxia cause cerebral palsy? Seminars in Perinatology, 24, p. 215-220. Quam, L., Ellis, L. B. M., Venus, P., Clouse, J., Taylor, C. G., and Leatherman, S. (1993) Using claims data for epidemiologic research: the concordanee with claims-based eriteria with the medical record and patient survey for identifying a hypertensive population. M edical Care, 31 (6), p. 498 - 507. Ramin, S. M. (2000) Introduetion. Seminars in Perinatology, 24, p. 195. 74 Ramin, S. M. and Gilstrap III, L. C. (2000) Other factors/conditions associated with cerebral palsy. Seminars in Perinatology, 24, p. 196-199. Reddihough, D. S., Baikie, G., and Walstab, J. E. (2001) Cerebral palsy in Victoria, Australia: Mortality and causes of death. Journal o f Paediatric and Child Health, 37, p. 183-186. Robertson, C. M. T., Svenson, L. W., and Joffres, M. R. (1998) Prevalence of cerebral palsy in Alberta. Canadian Journal o f Neurological Sciences, 25, p. 117122 . Romano, P. S. (2003) Asking too much of administrative data? Journal o f the American College o f Surgeons, 195, p. 337 - 338. Romano, P. S., Roos, L. L., Luft, H. S., Jollis, J. G., Doliszny, K., and the Ischemic Heart Disease Patient Outcomes research Team (1994) Journal o f Clinical Epidemiology, 47 (3), p. 249 - 260. Roos, L. L., Nicol, J. P., and Cageorge, S. M. (1987) Using administrative data for longitudinal research: Comparisons with primary data collection. Journal o f Chronic D isease, 40 (1), p. 41 - 49. Roos, L. L., Sharp, S. M., Cohen, M. M., and Wajda, A. (1989) Risk adjustment in claims-based research: The search for efficient approaches. Journal o f Clinical Epidemiology, 42 (12), p. 1193 - 1206. 75 Sciberras, C. and Spencer, N. (1999) Cerebral palsy in Malta 1981 to 1990. Developmental M edicine & Child Neurology, 41, p. 508-511. Shevell, M. I. (2004) The “Bermuda Triangle” of neonatal neurology: cerebral palsy, neonatal encephalopathy, and intrapartum asphyxia. Seminars in Pediatric Neurology, 11, p. 24-30. Shevell, M. L, Majnemer, A., and Morin, 1. (2003) Etiologic yield of cerebral palsy: A contemporary case study. Pediatric Neurology, 28, p. 352-359. Shields, J. R. and Schifrin, B. S. (1988) Perinatal antecedents of cerebral palsy. Obstetrics & Gynecology, 71, p, 899-905. Shinwell, E.S., Karplus, M., Reich, D., Weintraub, Z., Blazer, S., Bader, D., Turman, S., Dolfin, T., Kogan, A., Dollberg, S., Arbel, E., Goldberg, M., Gur, I., Naor, N., Sirota, L., Zaritsky, A., Barak, M., and Gottfried, E. (2000) Early postnatal dexamethasone treatment and increased incidence of cerebral palsy. Archives o f D isease in Childhood, 83, p. F177-F180. Sinha, G., Corry, P., Subesinghe, D., Wild, J., and Levene, M. I. (1997) Prevalence and type of cerebral palsy in a British ethnic community: the role of consanguinity. Developm ental M edicine and Child Neurology, 39, p. 259-262. 76 Skidmore, M. D., Rivers, A., and Hock, M. (1990) Increased risk of cerebral palsy among very low birth weight infants with chronic lung disease. Developm ental Medicine and Child Neurology, 32, p. 325-332. Smith, R. A., Skelton, M., Howard, M., and Levene, M. (2001) Is thrombophilia a factor in the development of hemiplegic cerebral palsy? Developmental Medicine and Child Neurology, 43, p. 724-730. Spinillo, A., Capuzzo, E., Cavallini, A., Stronati, M., De Santolo, A., and Fazzi, E. (1997) Preeclampsia, preterm delivery and infant cerebral palsy. European Journal o f Obstetrics and Gynecology, 77, p. 151-155. SPSS Inc. (2003) SPSS 12.0. Chicago: SPSS 2003. Stanley, P., Blair, E., and Alberman, E. (2000) Cerebral Palsies: Epidemiology & Causal Pathways. Clinics in Developmental Medicine No. 151. London: Mac Keith Press. Stanley, F. J., Blair, E., Hockey, A., Petterson, B. and Watson, L. (1993) Spastic quadriplegia in Western Australia: A genetic epidemiological study. I: Case population and perinatal risk factors. Developmental Medicine and Child Neurology, 35, p. 191-201. 77 Stanley, F. J., Croft, M. L., Gibbins, J., and Read, A. W. (1994) A population database for maternai and child health research in Western Australia using record linkage. Paediatric and Perinatal Epidemiology, 8, p. 433 - 447. Steinwachs, D. M., Stuart, M. E., Scholle, S., Starfield, B., Fox, M. H., Weiner, J. P. (1998) A comparison of ambulatory Medicaid claims to medical records: A reliability assessment. American Journal o f M edical Quality, 13 (2), p. 63 - 69. Strauss, D., Cable, W., and Shavelle, R. (1999) Causes of excess mortality in cerebral palsy. Developmental M edicine and Child Neurology, 41, p. 580-585. Strauss, D., Ojdana, K., Shavelle, R., and Rosenbloom, L. (2004) Decline in function and life expectancy of older persons with cerebral palsy. NeuroRehabilitation, 19, pp. 69-78. Strauss, D. and Shavelle, R. (1998) Life expectance of adults with cerebral palsy. Developmental M edicine and Child Neurology, 40, p. 369-375. Strauss, D. J., Shavelle, R. M., and Anderson, T. W. (1998) Life expectancy of children with cerebral palsy. Pediatric Neurology, 18, p. 143-149. Sugimoto, T., Woo, M., Nishida, N., Araki, A., Kara, T., Yasuhara, A., Kobayashi, Y., and Yamanouchi, Y. (1995) When do brain abnormalities in cerebral palsy occur? An MRI study. D evelopmental Medicine and Child Neurology, 37, p. 285-292. 78 Surveillance of Cerebral Palsy in Europe (2000) Surveillance of cerebral palsy in Europe; a collaboration of cerebral palsy surveys and registers. Developmental Medicine and Child Neurology, 42, p. 816-824. Suzuki, J. and Ito, M. (2002) Incidence patterns of cerebral palsy in Shiga Prefecture, Japan, 1977-1991. Brain & Development, 24, p. 39-48. Topp, M., Uldall, P., and Greisen, G. (2001) Cerebral palsy in eastern Denmark, 1987-90: Implications for neonatal care. Paediatric and Perinatal Epidemiology, 15, p. 271-277. Topp, M., Uldall, P., and Langhoff-Roos, J. (1997) Trend in cerebral palsy birth prevalence in eastern Denmark: Birth-year period 1979-86. Paediatric and Perinatal Epidemiology, 11, p. 451-460. Twiggs, J. E., Fifield, J., Apter, A. J., Jackson, E. A., and Cushman, R. A. (2002) Stratifying medical and pharmaceutical administrative claims as a method to identify pediatric asthma patients in a Medicaid managed care organization. Journal o f Clinical Epidemiology, 55, p. 938 - 944. Uldall, P., Michelsen, S. I., Topp, M., Madsen, M. (2001) The Danish cerebral palsy registry. Danish M edical Bulletin, 48, p. 161-163. 79 Vigneswaran, R. (2000) Infection and preterm birth: Evidence of a common causal relationship with bronchopulmonary dysplasia and cerebral palsy. Journal o f Paediatric and Child Health, 36, p. 293-296. Walstab, J. E., Bell, R. J., Reddihough, D. S., Brenneeke, S. P., Bessell, C. K., and Beiseher, N. A. (2004) Factors identified during the neonatal period associated with risk of cerebral palsy. Australian and New Zealand Journal o f Obstetrics and Gynaecology, 44, p. 342-346. Wheater, M. and Rennie, J. M. (2000) Perinatal infection is an important risk factor for cerebral palsy in very low birth weight infants. Developmental Medicine and Child Neurology, 42, p. 364-367. Williams, K. and Alberman, E. (1998) Survival in cerebral palsy: the role of severity and diagnostic labels. Developmental M edicine and Child Neurology, 40, p. 376-379. Wu, Y. W. and Colford Jr., J. M. (2000) Chorioamnionitis as a risk factor for cerebral palsy: A meta-analysis. Journal o f the American M edical Association, 284, p. 1417-1424. Wu, Y. W., Escobar, G. J., Grether, J. K., Croen, L. A., Greene, J. D., and Newman, T. B. (2003) Chorioamnionitis and cerebral palsy in term and near-term infants. Journal o f the American M edical Association, 290, p. 2677-2684. 80 Yoon, B. H., Park, C-W., and Chaiworapongsa, T. (2003) Intrauterine infection and the development of cerebral palsy. British Journal o f Obstetrics and Gynaecology, 110 (Suppl 20), p. 124-127. 81