Validation of a cerebral palsy register.

OBJECTIVES

To analyse completeness and validity of data in the Cerebral Palsy Register in Denmark, 1979-1982.

METHODS

Completeness has been assessed by comparing data from The Danish National Patient Register (DNPR) with the cases included in the Cerebral Palsy Register (CPR). Agreement between 12 variables in the CPR and obstetrical medical records has been analysed using kappa-statistics.

RESULTS

Of 468 children in the DNPR, only 237 fulfilled the inclusion criteria of the CPR; and 35 (15%) of these cases had not been reported to the CPR. Data agreement was generally good, but gestational age was subject to a systematic error, and urinary infections in pregnancy (kappa = 0.43) and placental abruption (kappa = 0.52) were seriously under-reported in the CPR.

CONCLUSIONS

Completeness of the Cerebral Palsy Register in Denmark, 1979-1982, has been assessed to maximal 85%, emphasizing the impact of using supplementary case ascertainment sources in the future. Validity of data varied according to definition and significance of the specific variable.