Factors Affecting the Initiation of a Shared Decision Making Program in Obstetric Practices.

As healthcare systems progress toward initiatives that increase patient engagement, stakeholder hopes are that shared decision making (SDM) will become routine practice. Yet, there is limited empirical evidence to guide such SDM program implementations, particularly in obstetric practices. The first stage of any project implementation is the "initiation stage", in which project leaders define a project's purpose and stakeholders and structures are put in place to support the new initiative. Our study's objective was to identify factors affecting the initiation stage of an SDM program implementation project for TOLAC, trial of labor after Cesarean. We conducted a multiple-case study of an SDM program implementation in three obstetric settings in Washington State. The research design and analysis were guided by implementation science frameworks and project management literature. Data sources included interviews with key informants from the State, SDM tool vendors, and three project sites, as well as implementation documents. The study results provide insight into how the identified project implementation factors provide an essential foundation for informing project planning, execution, and reflection/evaluation. In this study, the State's decision aid certification program pressured the project sites to shape the project purpose and engage stakeholders that would meet immediate project requirements (specifically, state requirements). The study reveals that external demands may not be in perfect alignment with the internal necessities required for an SDM program's long-term viability and sustainability. Findings may be used by implementers and researchers to model and strategize the early stages of SDM program implementation projects, particularly in the obstetric setting.

State-Level Evaluation of Washington's State Innovation Models (SIM) Initiative.

The Washington State Innovation Models (SIM) $65 million Test Award from the Center for Medicare & Medicaid Services' Innovation Center is a statewide intervention expected to improve population health, quality of care, and cost growth through four initiatives: 1) regional accountable communities of health linking health and social services to address local needs; 2) a practice transformation support hub; 3) four value-based payment reform pilot projects mainly in state employee and Medicaid populations; and 4) data and analytic infrastructure development to support system transformation with common measures. We develop a conceptual model based on diffusion theory and apply the RE-AIM evaluation framework (Reach, Effectiveness, Adoption, Implementation, and Maintenance) to structure our evaluation. We find that in three years (2016-2018), SIM built the infrastructure for system transformation and increased Washington's readiness for health system change in the next decade. However, the initiatives have not spread statewide, which may take over 10 years.

Mixed-Methods Evaluation of the Washington State Practice Transformation Support Hub.

PURPOSE: Practice transformation initiatives have the potential to promote collaborations between public health, primary care, and behavioral health, but limited empirical evidence is available on how these programs affect participating clinical practices. OBJECTIVE: To report the findings from a mixed-methods program evaluation of the Washington Practice Transformation Support Hub (Hub), a publicly funded, multicomponent practice transformation initiative in Washington State. DESIGN: We used quantitative and qualitative methods to evaluate the impact of Hub activities on participating primary care and behavioral health practices. Pre- and posttest survey data were combined with administrative program data to understand the effect of program components. Qualitative interviews contextualized findings. SETTING: Urban and rural primary care and behavioral health practices in Washington State. PARTICIPANTS: One hundred seventy-five practices that were recruited to receive Hub coaching and facilitation from 8 coaches; of these, 13 practices and all coaches participated in key informant interviews. INTERVENTION: Practice coaching and facilitation supported by an online resource portal, from January 2017 through January 2019. MAIN OUTCOME MEASURES: Self-reported progress in specific activities in 3 practice-level domains: bidirectional integration of physical and behavioral health care (care integration); alignment with community-based services for whole-person care (clinical-community linkages); and value-based payment. RESULTS: Participation in Hub activities was associated with improvements in care integration and clinical-community linkages but not with progress toward value-based payment. Qualitative results indicated that practice progress was influenced by communication with practices, the culture of the practice, resource constraints (particularly in rural areas), and perceptions about sustainability. CONCLUSIONS: This statewide practice transformation initiative was successful in strengthening primary care and behavioral health integration and clinical-community linkages among participating practices but not value-based payment. Future practice transformation efforts may benefit from addressing barriers posed by communication, limited application of value-based payment, culture change, competing priorities, and resource limitations, particularly for rural communities.

RE-AIM Evaluation Plan for Washington State Innovation Models Project.

The State of Washington received a State Innovation Models (SIM) $65 million award from the federal Centers for Medicare & Medicaid Services to improve population health and quality of care and reduce the growth of health care costs in the entire state, which has over 7 million residents. SIM is a "complex intervention" that implements several interacting components in a complex, decentralized health system to achieve goals, which poses challenges for evaluation. Our purpose is to present the state-level evaluation methods for Washington's SIM, a 3-year intervention (2016-2018). We apply the RE-AIM (reach, effectiveness, adoption, implementation, and maintenance) evaluation framework to structure our evaluation. We create a conceptual model and a plan to use multiple and mixed methods to study SIM performance in the RE-AIM components from a statewide, population-based perspective.

Use cases for genetic epidemiology in malaria elimination.

BACKGROUND: While traditional epidemiological approaches have supported significant reductions in malaria incidence across many countries, higher resolution information about local and regional malaria epidemiology will be needed to efficiently target interventions for elimination. The application of genetic epidemiological methods for the analysis of parasite genetics has, thus far, primarily been confined to research settings. To illustrate how these technical methods can be used to advance programmatic and operational needs of National Malaria Control Programmes (NMCPs), and accelerate global progress to eradication, this manuscript presents seven use cases for which genetic epidemiology approaches to parasite genetic data are informative to the decision-making of NMCPs. METHODS: The use cases were developed through a highly iterative process that included an extensive review of the literature and global guidance documents, including the 2017 World Health Organization's Framework for Malaria Elimination, and collection of stakeholder input. Semi-structured interviews were conducted with programmatic and technical experts about the needs and opportunities for genetic epidemiology methods in malaria elimination. RESULTS: Seven use cases were developed: Detect resistance, Assess drug resistance gene flow, Assess transmission intensity, Identify foci, Determine connectivity of parasite populations, Identify imported cases, and Characterize local transmission chains. The method currently used to provide the information sought, population unit for implementation, the pre-conditions for using these approaches, and post-conditions intended as a product of the use case were identified for each use case. DISCUSSION: This framework of use cases will prioritize research and development of genetic epidemiology methods that best achieve the goals of NMCPs, and ultimately, inform the establishment of normative policy guidance for their uses. With significant engagement of stakeholders from malaria endemic countries and collaboration with local programme experts to ensure strategic implementation, genetic epidemiological approaches have tremendous potential to accelerate global malaria elimination efforts.

Developing the Evidence for Public Health Systems to Battle Vaccine Preventable Disease at the Local Level: Data Challenges and Strategies for Advancing Research.

CONTEXT: Generally decreasing local health department (LHD) resources devoted to immunization programs and changes in LHD roles in immunization services represent major shifts in a core LHD service. OBJECTIVE: Within a rapidly changing immunization landscape and emerging vaccine preventable disease outbreaks, our objective was to examine how LHD immunization expenditures are related to county-level immunization coverage and pertussis rates. DESIGN: We used a practice-based approach in which we collaborated with practice partners and uniquely detailed LHD immunization expenditure data. Our analyses modeled the ecologic relationship between LHD immunization expenditures and LHD system performance and health outcomes. SETTING: This study was launched through a consortium of public health Practice-Based Research Network states as part of a suite of studies examining the relationship between various LHD service-related expenditures and health outcomes. PARTICIPANTS: We investigated and sought to include all LHDs in the states of Florida, New York (except New York City's LHD), and Washington. OUTCOME MEASURES: With LHD immunization expenditures as our independent variable, our outcomes were 1 year of jurisdiction-level rates of toddler immunization completeness, to measure immunization system performance, and 11 years of annual jurisdiction-level numbers of pertussis cases per 100 000 population, to measure related health outcomes. RESULTS: Immunization completeness and pertussis rates varied greatly, but our models did not produce significant results despite numerous analytic approaches and while controlling for other factors. CONCLUSION: While our study was part of a suite of studies using similar methods and producing significant results, this study was instead challenged by serious data limitations and highlighted the gap in consistent, standardized data that can support critically needed evidence regarding immunization rates and disease. With LHDs at the epicenter of reducing vaccine preventable disease, it is vital to utilize emerging opportunities to understand the nature of their efforts in immunization coverage and disease prevention.

Local health department food safety and sanitation expenditures and reductions in enteric disease, 2000-2010.

OBJECTIVES: In collaboration with Public Health Practice-Based Research Networks, we investigated relationships between local health department (LHD) food safety and sanitation expenditures and reported enteric disease rates. METHODS: We combined annual infection rates for the common notifiable enteric diseases with uniquely detailed, LHD-level food safety and sanitation annual expenditure data obtained from Washington and New York state health departments. We used a multivariate panel time-series design to examine ecologic relationships between 2000-2010 local food safety and sanitation expenditures and enteric diseases. Our study population consisted of 72 LHDs (mostly serving county-level jurisdictions) in Washington and New York. RESULTS: While controlling for other factors, we found significant associations between higher LHD food and sanitation spending and a lower incidence of salmonellosis in Washington and a lower incidence of cryptosporidiosis in New York. CONCLUSIONS: Local public health expenditures on food and sanitation services are important because of their association with certain health indicators. Our study supports the need for program-specific LHD service-related data to measure the cost, performance, and outcomes of prevention efforts to inform practice and policymaking.

Evidence-based communications strategies: NWPERLC response to training on effectively reaching limited English-speaking (LEP) populations in emergencies.

INTRODUCTION: Identifying and overcoming barriers to effective emergency preparedness and response is one of the objectives for the Centers for Disease Control and Prevention's network of 14 Preparedness and Emergency Response Learning Centers (PERLCs) and 9 Preparedness and Response Research Centers (PERRCs). This report describes how a PERLC and a PERRC colocated at the Northwest Center for Public Health Practice responded to Los Angeles County Department of Public Health's (DPH's) request to improve emergency communications with limited English-proficient (LEP) populations. METHODS/ACTIVITY: Activities included an assessment of training needs of the DPH preparedness workforce, a training series on social media and community engagement, and a toolkit of evidence-based findings to improve LEP populations' emergency communications and community resilience. RESULTS/OUTCOMES: Most respondents to the training needs assessment considered themselves essential personnel during an emergency and stated that they have received proper training. Respondents would like to receive further emergency preparedness training, including additional clarity on their role during an emergency. The majority of participants rated the training series as excellent/very good and agreed that they will be able to apply the course content to their work. The percentage of participants who reported confidence in their knowledge and skills related to each course learning objective increased from the precourse survey to the postcourse survey. DISCUSSION: This article discusses how the colocation of PERRC and PERLC offers efficiencies and expertise to accomplish multicomponent evidence-based requests. The ability to translate research findings quickly into evidence-based training and best practice resources is a strategic benefit to public health practice agencies working on emergency preparedness. LESSONS LEARNED/NEXT STEPS: LA County DPH was able to use knowledge and lessons learned gained from this work to design and prioritize education and training offerings to improve the capacity to effectively engage and communicate with LEP communities before and during public health emergencies.

Risk factors for infections in international travelers: an analysis of travel-related notifiable communicable diseases.

BACKGROUND: We sought to describe travel-related illness among our residents and gain insight into targeting pre-travel health advice to prevent travel-related illness. METHODS: A supplemental travel questionnaire was developed and administered for cases with a legally notifiable communicable disease reported in 2011-2012, who spent at least part of their exposure period outside the United States. RESULTS: Among 451 cases meeting the eligibility criteria, 259 were interviewed. Forty four percent reported receiving pre-travel advice. Two-thirds adhered fully with risk behavior recommendations; 94% followed immunization recommendations partially or fully; and 84% adhered fully with malaria prophylaxis recommendations. The primary reasons for not obtaining pre-travel advice were being unaware of the need (47.5%), or believing they already knew what to do (34.5%). Adults (OR = 2.8, 95% CI = 1.4-5.5), males (OR = 1.8, 95% CI = 1.1-3.0), those born outside the United States (OR = 2.0, 95% CI = 1.1-3.7), and those with planning time under two weeks (OR = 4.8, 95% CI = 1.5-15.9) or travel duration less than 7 days (OR = 7.9, 95% CI = 3.0-20.9) were more likely to travel without seeking pre-travel advice. CONCLUSIONS: The majority of cases reported not receiving pre-travel advice. Understanding the predictors of failure to receive pre-travel advice may help target public health prevention efforts.

2009 pandemic influenza a vaccination of pregnant women--King County, Washington State, 2009-2010.

OBJECTIVES: Our objectives were to estimate 2009 pandemic influenza A (pH1N1) vaccination coverage among pregnant women and identify associated factors. METHODS: We distributed a multimodal survey to 5341 women who gave birth between November 1, 2009, and January 31, 2010, identified by hospitals in King County, Washington State, with maternity services (n = 11). RESULTS: Of 4205 respondents, 3233 (76.9%) reported that they had received pH1N1 vaccine during pregnancy or within 2 weeks after delivery. Women whose prenatal care provider recommended vaccine had a higher vaccination prevalence than women whose provider did not (81.5% vs 29.6%; adjusted prevalence ratio = 2.1; 95% confidence interval = 1.72, 2.58). Vaccination prevalence was lower among women who had received prenatal care from a midwife only compared with women who had received care from other providers (62.9% vs 78.8%; adjusted prevalence ratio = 0.89; 95% confidence interval = 0.83, 0.96). CONCLUSIONS: Among pregnant women in King County, pH1N1 vaccination coverage was high. To improve coverage during nonpandemic seasons, influenza vaccine should be recommended routinely by prenatal care providers and vaccination provided where prenatal care is received. Barriers to midwives providing vaccination recommendations to patients should be explored.

2009 pandemic influenza A vaccination of pregnant women: King County, Washington State, 2009-2010.

PURPOSE: The objectives were to estimate 2009 pandemic influenza A (pH1N1) vaccination coverage among pregnant women and identify associated factors. METHODS: A multimodal survey was distributed to 5341 women who gave birth between November 1, 2009, and January 31, 2010, identified by hospitals in King County, Washington State, with maternity services (n=11). RESULTS: Of 4205 respondents, 3233 (76.9%) reported that they had received pH1N1 vaccine during pregnancy or within 2 weeks after delivery. Women whose prenatal care provider recommended vaccine had a higher vaccination prevalence than women whose provider did not (81.5% vs 29.6%; adjusted prevalence ratio=2.1; 95% CI=1.72, 2.58). Vaccination prevalence was lower among women who had received prenatal care from a midwife only compared with women who had received care from other providers (62.9% vs 78.8%; adjusted prevalence ratio=0.89; 95% CI=0.83, 0.96). CONCLUSIONS: Among pregnant women in King County, pH1N1 vaccination coverage was high. To improve coverage during nonpandemic seasons, influenza vaccine should be recommended routinely by prenatal care providers and vaccination provided where prenatal care is received. Barriers to midwives providing vaccination recommendations to patients should be explored.

Infant pertussis epidemiology and implications for tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis (Tdap) vaccination: King County, Washington, 2002 through 2007.

OBJECTIVES: To describe the epidemiology of infant pertussis in King County, Washington, and to better understand the implications for tetanus toxoid, reduced diphtheria toxoid, and acellular pertussis (Tdap) vaccination among older children, adolescents, and adults. DESIGN: Retrospective analysis of reported pertussis cases among infants younger than 1 year, January 1, 2002, through December 31, 2007. SETTING: King County, Washington. PARTICIPANTS: Reported pertussis cases among infants younger than 1 year between 2002 and 2007. MAIN OUTCOME MEASURES: Bordetella pertussis from a household member or close contact was the primary exposure. The main outcome measures were age and vaccination status, incidence by race/ethnicity, suspected exposure, and Tdap eligibility of household members and close contacts. RESULTS: Among 176 confirmed cases of infants with pertussis, the median age was 3 months (age range, 0-11 months); 80.1% were younger than 6 months. Seventy-seven percent were age-appropriately vaccinated. Between 2002 and 2007, the overall mean annual incidence was 136 cases per 100,000 infant population. Compared with a mean annual incidence of 73 cases per 100,000 infant population among whites, the incidence was 246 cases per 100,000 infant population among blacks (rate ratio [RR], 3.37; 95% confidence interval [CI], 2.59-4.44) and 194 cases per 100,000 infant population among Hispanics (RR, 2.66; 95% CI, 2.02-3.53). Households were the suspected exposure location for 70.0% of cases. Case households had a median of 3 (range, 1-15) Tdap-eligible persons. CONCLUSIONS: The burden of infant pertussis in King County, Washington, was high between 2002 and 2007, especially among racial/ethnic minorities. Tdap vaccination of eligible household members and close contacts should be promoted as an additional means of protecting infants from pertussis.

Shedding of pandemic (H1N1) 2009 virus among health care personnel, Seattle, Washington, USA.

The Centers for Disease Control and Prevention (CDC) recommends that health care personnel (HCP) infected with pandemic influenza (H1N1) 2009 virus not work until 24 hours after fever subsides without the use of antipyretics. During an influenza outbreak, we examined the association between viral shedding and fever among infected HCP. Participants recorded temperatures daily and provided nasal wash specimens for 2 weeks after symptom onset. Specimens were tested by using PCR and culture. When they met CDC criteria for returning to work, 12 of 16 HCP (75%) (95% confidence interval 48%-93%) had virus detected by PCR, and 9 (56%) (95% confidence interval 30%-80%) had virus detected by culture. Fever was not associated with shedding duration (p = 0.65). HCP might shed virus even when meeting CDC exclusion guidelines. Further research is needed to clarify the association between viral shedding, symptoms, and infectiousness.

Spring 2009 H1N1 influenza outbreak in King County, Washington.

BACKGROUND: In April 2009, King County, Washington, experienced a sustained outbreak of 2009 H1N1 influenza A. This report describes the epidemiology of that outbreak in King County, home to a diverse population of 1.9 million people. METHODS: The 2 primary sources of data are case investigations of reported laboratory-confirmed 2009 H1N1 influenza A and a population-based syndromic surveillance system that captures data from emergency departments (EDs). A syndromic category for influenza-like illness was defined based on chief complaint and diagnosis. RESULTS: ED visits for influenza-like illness peaked quickly in the first week of the outbreak and remained high for approximately 6 weeks, with school-age children accounting for the greater number of ED visits, followed by young adults. Children ages 0 to 4 years had the highest rate of hospitalization. Among reported cases, blacks, Asians, and Hispanics were more likely to be hospitalized. Predisposing factors associated with admission were immune compromise, chronic lung disease, chronic heart disease, pregnancy, diabetes, and asthma. Of people receiving antiviral treatment, 34% started their medication more than 2 calendar days after the onset of illness. Mean days between illness onset and antiviral treatment were greater for blacks, Hispanics, and foreign language speakers. CONCLUSIONS: The spring 2009 influenza A H1N1 outbreak disproportionately affected children, young adults, and racial and ethnic minorities. Opportunities exist to improve the timeliness of antiviral treatment. Potential barriers to care for racial and ethnic minorities should be proactively addressed to ensure prompt evaluation and treatment.

Association between school closure and subsequent absenteeism during a seasonal influenza epidemic.

BACKGROUND: Mathematical models suggest that social distancing measures, such as school closures, may mitigate community transmission during an influenza pandemic. Because closures are disruptive to schools and families, they are rarely employed during seasonal influenza outbreaks. A rare circumstance enabled us to examine the association between school closure and absenteeism during a seasonal influenza outbreak when half of King County, Washington public schools closed for a winter recess 19-23 February 2007, while half remained open for all or part of the week. METHODS: Using absenteeism as a proxy for influenza activity, we tested the hypothesis that schools on break would experience lower rates of post-break absenteeism than schools remaining open. We conducted daily retrospective and prospective surveillance from 5 February-9 March 2007 in schools on break (n = 256) and in session (n = 205). We use generalized estimating equations with Poisson distribution to evaluate whether mean absenteeism after the break differed between schools on break and those in session, adjusting for baseline absenteeism and repeated measurements by schools over time. RESULTS: Results indicate no difference in post-break absenteeism in schools on break compared with schools that remained in session (relative risk = 1.07 [95% confidence interval = 0.96-1.20]). This result held in elementary schools (1.00 [0.91-1.10]), where absenteeism patterns are thought to be most representative of community influenza activity. CONCLUSION: We did not find that school closure during a seasonal influenza outbreak reduced subsequent absenteeism. However, limitations in this "natural experiment" hampered our ability to detect a benefit if one truly was present.

Carbon monoxide epidemic among immigrant populations: King County, Washington, 2006.

OBJECTIVES: We investigated an outbreak of carbon monoxide (CO) poisoning after a power outage to determine its extent, identify risk factors, and develop prevention measures. METHODS: We reviewed medical records and medical examiner reports of patients with CO poisoning or related symptoms during December 15 to 24, 2006. We grouped patients into households exposed concurrently to a single source of CO. RESULTS: Among 259 patients with CO poisoning, 204 cases were laboratory confirmed, 37 were probable, 10 were suspected, and 8 were fatal. Of 86 households studied, 58% (n = 50) were immigrant households from Africa (n = 21), Asia (n = 15), Latin America (n = 10), and the Middle East (n = 4); 34% (n = 29) were US-born households. One percent of households was European (n = 1), and the origin for 7% (n = 6) was unknown. Charcoal was the most common fuel source used among immigrant households (82%), whereas liquid fuel was predominant among US-born households (34%). CONCLUSIONS: Educational campaigns to prevent CO poisoning should consider immigrants' cultural practices and languages and specifically warn against burning charcoal indoors and incorrect ventilation of gasoline- or propane-powered electric generators.

Task analysis in action: the role of information systems in communicable disease reporting.

In order to improve the design of information systems for notifiable conditions reporting, it is essential to understand the role of such systems in public health practice. Using qualitative techniques, we performed a task analysis of the activities associated with notifiable conditions reporting at a large urban health department. We identified seventeen primary tasks associated with the use of the department's information system. The results of this investigation suggest that communicable disease information management takes place in a dynamic environment where changing needs may require new solutions for system users. Additionally, our work suggests that task analysis, when applied to document the use of a legacy information system, can provide local public health agencies with valuable information about their information management needs, and can become a part of their information management decision-making strategy.

Health services utilization by low-income limited English proficient adults.

OBJECTIVES: We evaluated the health care utilization of limited English proficiency (LEP) compared to English proficient (EP) adults with the same health insurance (Medicaid managed care) and full access to professional medical interpreters. METHODS: Health care utilization over two years was compared for 567 LEP and 1162 EP adults. Multivariate analysis controlled for age, gender, months enrolled in Medicaid and morbidity. RESULTS: LEP compared to EP subjects were enrolled longer and more continuously in Medicaid, were 94% more likely to use primary care and 78% less likely to use the emergency department. Specialty visits and hospitalization did not differ. CONCLUSIONS: When language barriers are reduced and health insurance coverage is the same, LEP patients show ambulatory health care utilization associated with lower cost and more access to preventive care through establishing a primary care home.

Toxic shock associated with Clostridium sordellii and Clostridium perfringens after medical and spontaneous abortion.

OBJECTIVE: To better understand the risk of fatal toxic shock caused by Clostridium sordellii in women who had a recent medical abortion with mifepristone and misoprostol. METHODS: We performed active and passive surveillance for cases of toxic shock associated with medical or spontaneous abortion. To identify the cause of toxic shock, immunohistochemical assays for multiple bacteria were performed on formalin-fixed surgical and autopsy tissues. We extracted DNA from tissues, performed Clostridium species-specific polymerase chain reaction assays, and sequenced amplified products for confirmation of Clostridium species. RESULTS: We report four patients with toxic shock associated with Clostridium species infection after medical or spontaneous abortion. Two women had fatal Clostridium perfringens infections after medically induced abortions: one with laminaria and misoprostol and one with the regimen of mifepristone and misoprostol. One woman had a nonfatal Clostridium sordellii infection after spontaneous abortion. Another woman had a fatal C sordellii infection after abortion with mifepristone and misoprostol. All four patients had a rapidly progressive illness with necrotizing endomyometritis. CONCLUSION: Toxic shock after abortion can be caused by C perfringens as well as C sordellii, can be nonfatal, and can occur after spontaneous abortion and abortion induced by medical regimens other than mifepristone and misoprostol. LEVEL OF EVIDENCE: III.

Effectiveness of a parental educational intervention in reducing antibiotic use in children: a randomized controlled trial.

OBJECTIVE: To determine whether an educational intervention aimed at parents leads to fewer antibiotic prescriptions for their children. DESIGN: Placebo-controlled, randomized controlled trial. SETTING: Offices of primary care pediatricians who are members of a regional practice-based research network. PARTICIPANTS: Healthy children younger than 24 months old enrolled at the time of an office visit. INTERVENTIONS: Parents of study children were randomized to receive either a pamphlet and videotape (featuring one of their child's pediatricians) promoting the judicious use of antibiotics (intervention group) or brochures about injury prevention (control group). A total of 499 eligible children were enrolled, and data on outpatient visits during a 12-month observation period were collected. MAIN OUTCOME MEASURES: We compared the number of visits for upper respiratory tract infections (URIs), number of diagnoses and antibiotic prescriptions for otitis media and/or sinusitis and total number of antibiotics per patient among children in the intervention and control groups using Poisson regression analysis, adjusted for clustering into different practices. RESULTS: : Data on 4924 visits were reviewed; 28.8% of these visits were because of URI symptoms. The mean number of visits per study patient for URI symptoms was 2.8. Including all visits, the mean number of diagnoses of otitis media in study children was 2.1, mean number of diagnoses of otitis media and/or sinusitis was 2.3 and mean number of antibiotic prescriptions was 2.4; there were no significant differences between children in the intervention and control groups for any of these outcomes. Overall physicians prescribed 1 or more antibiotics during 45.9% of visits for a chief complaint of URI symptoms; 92% of antibiotic usage in children presenting with URI symptoms was for a diagnosis of otitis media and/or sinusitis. CONCLUSIONS: An educational intervention aimed at parents did not result in a decrease in the number of antibiotic prescriptions in their children. The use of antibiotics among children with URI symptoms was common; other interventions promoting the judicious use of these medications are needed.