• N Ewen Wang, Michaela Kiernan, Mana Golzari, and Michael A Gisondi.
• Department of Surgery, Stanford University School of Medicine, Palo Alto, CA 94304, USA. ewen@stanford.edu
• Acad Emerg Med. 2006 Aug 1;13(8):840-7.

ObjectivesTo identify and characterize subgroups of a pediatric population at risk of poor emergency department (ED) aftercare compliance.MethodsThis was a prospective, cohort study conducted at a university hospital ED with a 2003 pediatric census of 11,040 patients. A convenience sample of 461 children was enrolled. The study follow-up rate was 97%. The primary outcomes were guardian compliance with instructions for physician follow-up appointment and with obtaining prescribed medications. Predictors of compliance outcomes were analyzed by using recursive partitioning to describe population subgroups at risk of poor compliance.ResultsOnly 60.4% of patient guardians followed up with instructions to see a physician. Children with private insurance were more likely to follow up than were children without private insurance (76.8% vs. 46.5%, p < 0.001). Of children with private insurance, those with high-acuity diagnoses were more likely to follow up than were patients with low-acuity diagnoses (80.0% vs. 38.5%, p < 0.001). Of children who were considered underinsured (defined as publicly insured or uninsured), those with English-speaking guardians were more likely to follow up than were those with non-English-speaking guardians (58.0% vs. 40.0%, p < 0.05). Only 63.3% of patient guardians obtained prescribed medications. Privately insured children were more likely to obtain medications than were underinsured children (71.0% vs. 58.0%, p < 0.05). Descriptive profiles of the subgroups revealed that those with lower socioeconomic status were at greatest risk of poor aftercare compliance.ConclusionsCompliance with ED aftercare instructions remains a challenge. Health insurance disparities are associated with poor ED aftercare compliance in our pediatric population. Interventions aimed at improving compliance could be targeted to specific subgroups on the basis of their descriptive profiles.

27 keywords...

hide…