Upper limb deficiencies in Swedish children--a comparison between a population-based and a clinic-based register.

Objective: To validate information in the Swedish Register for Congenital Malformations (SRCM).

Methods: A comparison was made with a clinic-based register kept at the Limb Deficiency and Arm Prosthesis Centre (LDAPC). The report frequency and the quality of the information in SRCM were analysed. Cases were classified according to a detailed, clinically relevant classification, the ISO 8548-1:89 method.

Outcome Measures: The completeness of SRCM was first estimated. The Kappa statistic was then used to assess the agreement between the two registers regarding individual categories and across all categories.

Results: For the period 1973-1987, we found 125 cases of upper limb reduction deficiencies (ULRD) in the clinic-based register, of which 117 was found in the national register. The completeness of SRCM was thus estimated to be 94% (95% confidence interval 89-98%). The inter-register agreement varied from almost perfect agreement in laterality of deficiency (Kappa 0.98) to substantial agreement in type and level of deficiency (Kappa 0.72-0.79). For specific levels of transverse deficiency, however, the agreement varied between -0.05 and 0.66.

Conclusions: The results indicate that SRCM, with its calculated underestimation of 6%, can be used for studying the prevalence of ULRD in Sweden. However, as SRCM is a surveillance register, the quality of some information seems to be low, making detailed description of cases difficult. Use of the population register data for clinical purposes could therefore result in lower validity. Additional information and follow-up of specific cases are therefore recommended.