Paediatric and Perinatal Epidemiology 1992,6,339-351
The California Cerebral Palsy Project Judith K. Grether*, Susan K. Cumminst and Karin 8.Nelson$ *California Birth Defects Monitoring Program and tEnvironmenta1 Epidemiology and Toxicology Branch, California Department of Health Smices, Emeryoille, California, and $National Institute of Neurological Disorders and Stroke, Neuroepidemiology Branch, Bethesda, Mayland, USA
Summary. The California Cerebral Palsy Project (CACP) is a population-based study of 192 children with moderate or severe congenital cerebral palsy who were born between 1983 and 1985 in four San Francisco Bay area counties and who were alive and residing in California at age 3 years. Initial ascertainment of cases was based on records of two agencies known to enrol virtually all CACP-eligible children. Final case status was established by standardised clinical examination in 67% of cases and extensive record review in 33%.The 192 cases gave a prevalence at age 3 of 1.23/10oOsurvivors. Twins were 10%of the cases with a prevalence of 6.711000. Overall, 53% of the cases had birthweight 3 2,500 g and 28% had birthweight < 15oog. There was no association between birthweight and severity of functional impairment and no consistent association between birthweight and the presence of associated disabilities. The CACP prevalence is lower than that reported in other studies and is believed to be due to the more stringent case inclusion criteria employed for this research data base.
Introduction The California Cerebral Palsy Project (CACP) is a population-based study of children with moderate or severe congenital cerebral palsy (CP) who were born between 1983 and 1985 in four San Francisco Bay area counties. This report describes the methods used to ascertain and diagnose cases and reports the Address for correspondence: Dr J.K.Grether, California Birth Defects Monitoring Program, 5900 Hours Street, Suite A, Emeryvllle, California 94608, USA.
339
340
1.K. Grether et al.
prevalence of CP as observed at 3 years of age. It also reports the dominant CP subtypes in relation to birthweight and the presence of associated disabilities.
Methods
Dejinitions Cerebral palsy was defrned as a chronic disability of central nervous system origin characterised by aberrant control of movement or posture, appearing early in life and not the result of a progressive disease. Excluded were transient motor disabilities, motor disabilities due to meningomyelocele or other spinal cord lesions, isolated hypotonia, and motor abnormalities due solely to mental deficiency. To be defined as a case, a child must have met all of the following criteria: born between 1983 and 1985; maternal residence as verified on the child's birth certificate in one of four San Francisco Bay area counties (Contra Costa, San Francisco, San Mateo, Santa Clara); child alive and residing in California at age 3 years; moderate or severe congenital CP at age 3 years as determined by a standardised examination or record review conducted by a single CACP physician. The criterion age of 3 years was chosen as the earliest age at which reliable ascertainment was feasible in these moderately and severely affected children. Excluded were children whose CP was acquired through infection, trauma, or other known adverse event after the first 28 days of life or through documented severe head trauma in the first month of life. Assessment of severity of CP was based on functional ability of the most affected limb. CP was classified as severe if that limb had no function, and as moderate if some function was preserved although assistive devices were usually required. Mild CP, with abnormalities oftone and reflexes but no functional impairment, was excluded because it could not be reliably ascertained. CP was categorised as spastic hemiplegia (armand leg of one side of the body affected), spastic diplegia (legs more affected than arms), spastic quadriplegia (involvement of all four limbs, arm involvement at least as severe as legs), dyskinesia (abnormal involuntary movements), ataxia (lack of balance and uncoordinated movement) or mixed (spastic CP with ataxia and/or dyskinesia). The descriptive data used to assign CP subtype are compatible with the Standard Recording of Central Motor Deficit reported by Evans et al.' Information on associated disabilities was obtained through parental interview and classified as follows: vision disability ('some difficulty' or 'blind), hearing disability ('some difficulty' or 'deaf'), speech disability ('difficult to understand, 'non-verbal communication only' or 'no communication'), chronic seizure disorder (parental report of physician diagnosis) and mental retardation (parental report of diagnosis by physician or other professional). In addition, an informal
California Cerebral Palsy Project
341
impression of intellectual functioning was recorded by the CACP physician as part of the clinical examination. No other validation of associated disabilities was conducted.
Case ascertainment procedures Children with definite or possible motor disabilities were identified through review of service records of two state-funded agencies: the California Children ServiceslDepartment of Health Services (CCS) and Regional Centers for the Developmentally Disabled/Department of Developmental Services (DDS). Both programmes provide services to children with CP without regard to financial or citizenship status in accordance with statewide guidelines. Community surveys conducted by the California Birth Defects Monitoring Program prior to establishment of the CACP had demonstrated that, in the four study counties, virtually all children with moderate or severe congenital CP were CCS clients, many having dual enroIment with DDS. A very small minority of children with severe disabilities were exclusively DDS clients. Regulatory restrictions on access to confidential data prohibited inclusion of these few DDS clients in the data base. However, as explained below, limited non-confidential information was obtained for these children. Within CCS, the $year birth cohort was tracked until all members were at least 4% years old. Using very broad criteria applied to computerised service records, 1110 possible cases were selected for manual chart review, including children who had moved to other counties within the state and children who were not currently enrolled with CCS (Figure 1). Chart review was conducted by CACP field staff to determine presumptive eligibility. Records in which the diagnosis of CP was ambiguous underwent a second review by a CACP physician. Children with ambiguous aetiological information were retained as presumptive cases, pending review of medical records by the CACP physician. All presumptive cases were then linked to vital statistics records to venfy birthdate, residence and survival to age 3 (Figure 1). To establish the final diagnosis of CP, all CCS children who remained as presumptive cases were invited to undergo a standardised neurological and motor assessment. For children who could not be examined, the final diagnosis was established through extensive review of CCS records, other medical records and, in many cases, personal discussion with therapists or physicians familiar with the child. Physical examinations and record review were performed by a CACP physician (S.K.C.) who was blinded to the previous medical history of the children and who had received special training for this project. Prior to the examination, a brief interview was conducted with a parent or guardian, usually on the telephone, by a CACP field staff member. The interview was used to obtain
342
1.K. Grether et al.
CCS records of 1110 children initially reviewed I
I
-Excluded
-
18 children died or moved before age 3 years
369 children had M inellgibie birthdate or birth residence
n -
71 children had another condition that was not CP
51 children had CP acquired after 28 days of age or from head trauma after birth
339 had no evidence of CP I
262 children with possible CP matched to their birth certiflcate
i
-Excluded
-Excluded
-
14 children had mild CP
192 children with moderate or severe
Figure 1. Case ascertainment process at CCS: CACP 1983-1985.
1
Culifarniu Cerebral Palsy Project
343
information on medical and surgical treatment which might affect examination findingsand on associated disabilities. Intensive procedures were used to locate parents or guardians and solicit consent for participation in the examination. Bilingual translators .were used as needed. Examinations were conducted in CCS clinic facilities or the child’s home or school. Identical procedures were followed for all children residing in California during the study period, whether their residence was inside or outside the study area. Within DDS, clients with possible CP (n=322) were selected from computerised client records which contained non-confidential diagnostic data. Staff at the local DDS Regional Centers then reviewed confidential records to determine birth residence and dual enrolment with CCS. For the few presumptive cases who were exclusively DDS clients, DDS staff attempted to obtain parental consent for CACP staff to review confidential DDS records. This effort was generally unsuccessful because of inability to locate parents of institutionalised children. DDS medical staff then reviewed the complete DDS record on each presumptive case and completed a non-confidential data form querying residency, birthdate, survival and residency in California at age 3, type of motor condition, whether acquired or congenital, and severity. Because identdymg data were not available to link these children with vital records, they were not included in the CACP data base.
Evaluation of cuse ascertainment procedures To evaluate CCS and DDS as adequate sources for initial identification of cases, telephone interviews were conducted with 22 private physical and occupational therapists and with selected public school special education programmes to enquire about children with moderate or severe CP who were born between 1983 and 1985in the study area and who had never been enrolled with CCS and DDS. No such children were identifed. To check the selection of possible cases from the computerised CCS files, all open and closed charts between January 1983 and March 1990 were manually reviewed at two of the four local CCS programmes; all therapy records for the same period were reviewed at a third local CCS programme; all referral diagnoses were reviewed at the remaininglocal CCS programme. Two additional cases were found through these procedures and were added to the data base. To evaluate the exclusion of possible cases through the manual review of records, a second review was conducted by a different reviewer on a 40% sample (n = 100) of the subset of CCS clients who had been excluded because there was no evidence of motor problems at age 3, they had mild CP, developmental delay and hypotonia, or an apparently progressive condition with no specific diagnosis. Inter-rater agreement was 100%.
344
J.K. Grether et al.
To evaluate the assignment of final case status based on physical examination or physician review of records, a sample of cases underwent an informal review by a second CACP physician (K.B.N.). This review indicated that primary diagnoses were reliably established but that the assessment of secondary CP subtypes and degree of severity across subtypes may not be robust.
Statistical methods A chi-square test of association with sigruhcance level of 0.05 was used. A risk ratio (RR) measure with 95%confidence levels was used to obtain a point estimate of the relative risk of CP among subgroups of survivors to age 3.
.
I
Results The final CACP case series includes 192 children with moderate or severe CP as established by physical examination in 129 cases (67%)and record review in 63 cases (33%).Examined cases averaged 4.9 years of age at the time of examination, with a range from 2.6 to 7 years. Of the 63 cases whose final diagnosis was established through record review, 19 had refused consent or had repeatedly failed to keep appoinbnents for examinations. Nine had died after their third birthday, and three were known to have moved out of the state. The remaining32 cases could not be located and examined within the time frame of the study. The examined cases and record review cases were similar with respect to sex, maternal race, birth type (single or twin) and severity of CP. Examined cases were more often of very low birthweight (