Early Peanut Introduction in Primary Care: Evaluation of a Multicomponent Intervention.

OBJECTIVE: To determine whether a multicomponent intervention focused on early peanut introduction was associated with a lower peanut allergy incidence in young children. METHODS: The study cohort comprised all children born January 1, 2013 through December 31, 2018 receiving care at a large health care organization. Intervention activities occurred over 16 months and included provider educational programs, electronic health record tools, and new patient instructions. We used an interrupted time series design to assess whether peanut allergy incidence differed across 3 time periods (preintervention, interim, postintervention) among high- and low-risk children. The primary outcome was incident peanut allergy by age 24 months, defined as peanut allergy in the allergy field or active problem list plus a positive supportive test. Severe eczema and/or egg allergy presence defined high-risk. Because the study was conducted as part of routine care, it was not feasible to measure what counseling clinicians provided, or how and when parents fed their children peanut-containing foods. RESULTS: In a cohort of 22,571 children, the percent with peanut allergy by age 24 months was 17.3% (116 of 671) among high-risk and 0.8% (181 of 21,900) among low-risk children. In multivariate analyses, the adjusted peanut allergy rate per 100 person-years was not significantly different across study periods among high-risk (9.6 preintervention, 11.7 interim, and 9.9 postintervention, P = .70) or low-risk (0.5 preintervention, 0.7 interim, and 0.5 postintervention, P = .17) children. CONCLUSIONS: In a community-based setting, the incidence of peanut allergy did not decline following a multicomponent intervention focused on early peanut introduction.

Commitment to Change Statements and Actual Practice Change After a Continuing Medical Education Intervention.

INTRODUCTION: Continuing medical education (CME) interventions often evaluate participant commitment to change (CTC) clinical practice. Evidence linking CTC to actual practice change is limited. METHODS: In an intervention that combined live CME with changes to the electronic health record to promote judicious antibiotic use for children with urinary tract infections (UTIs), we evaluated CTC and subsequent prescribing behavior in Kaiser Permanente Colorado, an integrated health care system. CTC was assessed immediately after the session using closed-ended questions about session learning objectives and open-ended questions to elicit specific practice changes. Perceived barriers to implementing recommended changes were also assessed. RESULTS: Among 179 participants, 80 (45%) completed postsession evaluations and treated one or more child with a UTI in the subsequent 17 months (856 UTIs in total). In closed-ended responses about session learning objectives, 45 clinicians (56%) committed to changing practice for antibiotic choice and duration, whereas 37 (46%) committed to implementing new practice guidelines. When asked open-ended questions to identify specific practice changes, 32 (40%) committed to antibiotic choice change and 29 (36%) committed to treatment duration change. Participants who made specific CTC statements had greater improvement in antibiotic choice (relative rate ratio 1.56, 95% CI 1.16-2.09) and duration (relative rate ratio 1.59, 95% CI 1.05-2.41) than participants who did not make specific commitments. Few perceived barriers affected subsequent prescribing. DISCUSSION: Commitments to changing specific clinical behaviors were associated with sustained changes in prescribing for children with UTIs. Linking self-evaluations with clinical data in integrated health care systems is an important tool for CME evaluators.

Improving Antibiotic Prescribing for Pediatric Urinary Tract Infections in Outpatient Settings.

OBJECTIVES: To determine if a multicomponent intervention was associated with increased use of first-line antibiotics (cephalexin or sulfamethoxazole and trimethoprim) among children with uncomplicated urinary tract infections (UTIs) in outpatient settings. METHODS: The study was conducted at Kaiser Permanente Colorado, a large health care organization with ∼127 000 members <18 years of age. After conducting a gap analysis, an intervention was developed to target key drivers of antibiotic prescribing for pediatric UTIs. Intervention activities included development of new local clinical guidelines, a live case-based educational session, pre- and postsession e-mailed knowledge assessments, and a new UTI-specific order set within the electronic health record. Most activities were implemented on April 26, 2017. The study design was an interrupted time series comparing antibiotic prescribing for UTIs before versus after the implementation date. Infants <60 days old and children with complex urologic or neurologic conditions were excluded. RESULTS: During January 2014 to September 2018, 2142 incident outpatient UTIs were identified (1636 preintervention and 506 postintervention). Pyelonephritis was diagnosed for 7.6% of cases. Adjusted for clustering of UTIs within clinicians, the proportion of UTIs treated with first-line antibiotics increased from 43.4% preintervention to 62.4% postintervention (P < .0001). The use of cephalexin (first-line, narrow spectrum) increased from 28.9% preintervention to 53.0% postintervention (P < .0001). The use of cefixime (second-line, broad spectrum) decreased from 17.3% preintervention to 2.6% postintervention (P < .0001). Changes in prescribing practices persisted through the end of the study period. CONCLUSIONS: A multicomponent intervention with educational and process-improvement elements was associated with a sustained change in antibiotic prescribing for uncomplicated pediatric UTIs.

Implementation of Recommendations for Long-Acting Contraception Among Women Aged 13 to 18 Years in Primary Care.

OBJECTIVE: Clinical specialty societies recommend long-acting reversible contraceptives (LARCs) as first-line contraception for adolescent women. We evaluated whether a combined educational and process improvement intervention enhanced LARC placement in primary care within an integrated health care system. METHODS: The intervention included journal clubs, live continuing education, point-of-care guidelines, and new patient materials. We conducted a retrospective cohort study across 3 time periods: baseline (January 2013-September 2015), early implementation (October 2015-March 2016), and full implementation (April 2016-June 2017). The primary outcome was the proportion of LARCs placed by primary care clinicians among women aged 13 to 18 years compared with gynecology clinicians. RESULTS: Kaiser Foundation Health Plan of Colorado cared for approximately 20,000 women aged 13 to 18 years in each calendar quarter between 2013 and 2017. Overall, LARC placement increased from 7.0 per 1000 members per quarter at baseline to 13.0 per 1000 during the full intervention. Primary care clinicians placed 6.2% of all LARCs in 2013, increasing to 32.1% by 2017 (P < .001), including 45.5% of contraceptive implants. Clinicians who attended educational sessions were more likely to adopt LARCs than those who did not (17.9% vs 6.4% respectively, P = .009). Neither overall LARC placement rates (relative risk, 1.9; 95% confidence interval, 0.7-5.6) nor contraceptive implant rates (relative risk, 3.0; 95% confidence interval, 0.9-9.8) increased significantly in clinicians who attended educational activities. CONCLUSIONS: This multimodal intervention was associated with increased LARC placement for adolescent women in primary care. The combination of education and process improvement is a promising strategy to promote clinician behavior change.