Using Linked Data to Assess Patterns of Early Intervention (EI) Referral among Very Low Birth Weight Infants

Abstract

Access to Early Intervention (EI) services may improve cognitive and behavioral outcomes in very low birth weight infants, but few states have population-based data to evaluate EI outreach efforts. We analyzed Massachusetts (MA) infants born weighing <1,200 g to identify maternal and birth characteristics that predicted EI referral and timing of referral. MA birth and hospital discharge records (Jan. 1998-Sept. 2000) were linked to EI referral records (Jan. 1998-Sept. 2003) via probabilistic and deterministic methods (88% linkage). Timing of EI referral among infants weighing <1,200 g was examined by infant and maternal characteristics using categorical (0-12 months, 12-36 months, or no referral) time comparisons in the crude analysis. Survival functions calculating median time to referral, and adjusted hazard ratios (HR) with 95% confidence intervals (CI) were calculated for continuous time comparisons of EI referral from birth to 36 months. Of 1,233 infants weighing <1,200 g, 93.2% were referred to EI. After risk adjustment, referral was more likely among multiple-birth infants (HR = 1.17, 95%CI 1.06-1.30) and less likely among infants <28 weeks (HR = 0.70; 95%CI 0.64-0.77) or with low Apgar scores (<5 at 5 min; HR = 0.75; 95%CI 0.62-0.92). EI referrals were lower for infants of black non-Hispanic mothers, and mothers without private insurance (HR = 0.85; 95%CI 0.74-0.98 and HR = 0.77; 95%CI 0.68-0.86, respectively). In MA, most infants born <1,200 g are referred to EI, but disparities exist. Analysis of linked population-based health and developmental services can inform programs in order to reduce disparities and improve access for all high-risk infants.