Cecil G. Sheps Center for Health Services Research, University of North Carolina, Chapel Hill, NC 27599, USA.
Health Informatics J. 2010 Jun;16(2):115-28. doi: 10.1177/1460458210364036.
For program evaluation purposes, the feasibility of matching Medicaid claims with physician-completed structured checklists (encounter forms, EFs) was assessed in a pediatric office-based preventive dental program. We examined agreement on visits (weighted kappa) and predictors of a match between EFs and claims (multinomial logit model with practice-level clustering). In total, 34,171 matches occurred between 41,252 EFs and 40,909 claims, representing 82.8 per cent of EFs and 83.5 per cent of claims. Agreement on visits was 56 per cent (weighted kappa = 0.66). Pediatric practices provided the majority of visits (82.4%) and matches. Increasing age of child and residence in same county as the medical practice increased the likelihood of a match. Structured checklists can be combined with claims to better assess provision of preventive dental services in pediatric primary care. However, future research should examine strategies to improve the completion of structured checklists by primary care providers if data beyond claims are to be used for program evaluation.
为了进行项目评估,我们评估了在一家儿科基础预防牙科项目中,将医疗补助索赔与医生填写的结构化检查表(就诊表,EF)相匹配的可行性。我们检查了就诊记录(加权 Kappa)和 EF 与索赔之间匹配的预测因素(带有实践水平聚类的多项逻辑回归模型)。在总共 41,252 份 EF 和 40,909 份索赔中,有 34,171 次匹配,占 EF 的 82.8%和索赔的 83.5%。就诊记录的一致性为 56%(加权 Kappa=0.66)。儿科诊所提供了大部分就诊记录(82.4%)和匹配记录。儿童年龄增长和与医疗实践所在的同一县居住增加了匹配的可能性。结构化检查表可以与索赔相结合,以更好地评估儿科初级保健中预防牙科服务的提供情况。但是,如果要使用索赔以外的数据进行项目评估,则未来的研究应该检查提高初级保健提供者完成结构化检查表的策略。