Sources of Clinician Burnout in Providing Care for Underserved Patients in a Safety-Net Healthcare System.

BACKGROUND: Burnout among clinicians is common and can undermine quality of care, patient outcomes, and workforce preservation, but sources of burnout or protective factors unique to clinicians working in safety-net settings are less well understood. Understanding these clinician experiences may inform interventions to reduce burnout. OBJECTIVE: To describe clinician perspectives on sources of burnout in a safety-net healthcare system. DESIGN: Semi-structured interviews were conducted with clinicians from multiple disciplines who provided care at a safety-net healthcare system from October 2018 to January 2019. Transcripts were analyzed using thematic analysis. PARTICIPANTS: Forty clinicians (25 female and 15 male; mean [SD] age, 41 [9.1]) participated, including physicians, social workers, advanced practice providers, nurses, psychologists, physical and occupational therapists, and other healthcare professionals. MAIN OUTCOMES AND MEASURES: Themes and subthemes that reflected clinician experiences, burnout, and workload expectations. KEY RESULTS: Five themes emerged: limited resources (entrenched social injustices, brokenness of the US healthcare system, precarious discharge options, and "revolving door" readmissions), barriers to building trust with patients (chasm of communication, addressing fear and mistrust, and being exposed to threats), administrative requirements (criticism hampering meaningful care, assuming extra workloads, bureaucratic burden), compassion fatigue (confronting traumatic situations, persistent worry about patient safety and welfare, witnessing mistreatment and stigmatization, overextending and overinvesting, withdrawing and shutting down, blaming self and career crisis), and advocacy as a counterbalance to burnout (solidarity when helping underserved communities, fervent advocacy, and patient gratitude). CONCLUSIONS: Among clinicians who provide care in a safety-net healthcare system, sources of burnout included limited resources, barriers to building trust with patients, administrative requirements, and compassion fatigue, but clinicians remained driven by their desire to advocate for patients. Policy-level interventions which increase funding to safety-net healthcare systems to bolster existing resources and staffing, create peer-support and wellness programs, and support advocacy efforts may mitigate burnout.

Social determinants of obesity in American Indian and Alaska Native peoples aged ≥ 50 years.

OBJECTIVE: American Indian and Alaska Native peoples (AI/ANs) have a disproportionately high rate of obesity, but little is known about the social determinants of obesity among older AI/ANs. Thus, our study assessed social determinants of obesity in AI/ANs aged ≥ 50 years. DESIGN: We conducted a cross-sectional analysis using multivariate generalized linear mixed models to identify social determinants associated with the risk of being classified as obese (BMI ≥ 30.0 kg/m2). Analyses were conducted for the total study population and stratified by median county poverty level. SETTING: Indian Health Service (IHS) data for AI/ANs who used IHS services in FY2013. PARTICIPANTS: 27,696 AI/ANs aged ≥ 50 years without diabetes. RESULTS: Mean BMI was 29.8 ± 6.6 with 43% classified as obese. Women were more likely to be obese than men, and younger ages were associated with higher obesity risk. While having Medicaid coverage was associated with lower odds of obesity, private health insurance was associated with higher odds. Living in areas with lower rates of educational attainment and longer drive times to primary care services were associated with higher odds of obesity. Those who lived in a county where a larger percentage of people had low access to a grocery store were significantly less likely to be obese. CONCLUSIONS: Our findings contribute to the understanding of social determinants of obesity among older AI/ANs and highlight the need to investigate AI/AN obesity, including longitudinal studies with a life course perspective to further examine social determinants of obesity in older AI/ANs.

Patient Pill Organization Strategies and Adherence Measured in a Cross-Sectional Study of Hypertension.

Background: The strategies patients use to organize medications (eg, pill dispenser) may be reflected in adherence measured at follow-up. We studied whether medication organization strategies patients use at home are associated with adherence measured using pharmacy-fills, self-report, and pill counts. Design: Secondary analysis of data from a prospective randomized clinical trial. Setting: Eleven US safety-net and community primary care clinics. Patients: Of the 960 enrolled self-identified non-Hispanic Black and White patients prescribed antihypertensive medications, 731 patients reported pill organization strategies and were included. Variable: Patients were asked if they use any of the following medication organization strategies: finish previous refills first; use a pill dispenser; combine same prescriptions; or combine dissimilar prescriptions. Outcomes: Adherence to antihypertensive medications using pill counts (range, 0.0-1.0% of the days covered), pharmacy-fill (proportion of days covered >90%), and self-report (adherent/non-adherent). Results: Of the 731 participants, 38.3% were men, 51.7% were age ≥65, 52.9% self-identified as Black or African American. Of the strategies studied, 51.7% finished previous refills first, 46.5% used a pill dispenser, 38.2% combined same prescriptions and 6.0% combined dissimilar prescriptions. Median (IQR) pill count adherence was 0.65 (0.40-0.87), pharmacy-fill adherence was 75.7%, and self-reported adherence was 63.2%. Those who combined same prescriptions had significantly lower measured pill count adherence than those who did not (0.56 (0.26-0.82) vs 0.70 (0.46-0.90), p<0.01) with no significant difference in pharmacy-fill (78.1% vs 74%, p=0.22) or self-reported adherence (63.0% vs 63.3%, p=0.93). Conclusion: Self-reported medication organization strategies were common. Combining same prescriptions was associated with lower adherence as measured using pill counts but not pharmacy-fills or self-report. Clinicians and researchers should identify the pill organization strategies used by their patients to understand how these strategies may influence measures of patient adherence. Trial Registration: ClinicalTrials.gov NCT03028597; https://clinicaltrials.gov/ct2/show/NCT03028597 (Archived by WebCite at http://www.webcitation.org/72vcZMzAB).

Comparing two durations of medically tailored meals posthospitalization: A randomized clinical trial.

BACKGROUND: Medically tailored meals (MTM) may be beneficial to patients after hospital discharge. OBJECTIVE: To determine if 2 versus 4 weeks of MTM posthospitalization will improve patient outcomes. DESIGN: Randomized unblinded trial. SETTINGS AND PARTICIPANTS: Six hundred and fifty patients pending hospital discharge with at least one chronic condition. INTERVENTION: One MTM a day for 2 versus 4 weeks. MAIN OUTCOME AND MEASURES: The primary outcome was a change from baseline to 60 days in the Hospital Anxiety Depression Scale (HADS). Secondary outcomes measured change in the Katz activities of daily living (ADLs), DETERMINE nutritional risk, and all-cause emergency department (ED) visits and rehospitalizations. RESULTS: From baseline to 60 days the HADS anxiety subscale changed 5.4-4.9 in the 2-week group (p = .03) and 5.4-5.3 in the 4-week group (p = .49); the difference in change between groups 0.4 (p = .25). HADS changed 5.4-4.8 in the 2-week group (p = .005) and 5.3-5.1 in the 4-week group (p = .34); the difference in change between groups 0.4 (p = .18). ADL score changed from 5.3 to 5.6 in the 2-week group (p ≤ .0001) and 5.2-5.5 in the 4-week group (p ≤ .0001); the difference in change between groups -0.01 (p = .90). The DETERMINE changed in the 2-week group from 7.2 to 6.4 (p = .0006) and from 7 to 6.7 in the 4-week group (p = .19); the difference in change between groups 0.5 (p = .13). There was no difference in ED visits and rehospitalizations between groups or time to rehospitalization. CONCLUSIONS: Different durations of short-term MTM did not affect patient-centered or utilization outcomes.

The spectrum of engagement in HIV care and its relevance to test-and-treat strategies for prevention of HIV infection.

For individuals with human immunodeficiency virus (HIV) infection to fully benefit from potent combination antiretroviral therapy, they need to know that they are HIV infected, be engaged in regular HIV care, and receive and adhere to effective antiretroviral therapy. Test-and-treat strategies for HIV prevention posit that expanded testing and earlier treatment of HIV infection could markedly decrease ongoing HIV transmission, stemming the HIV epidemic. However, poor engagement in care for HIV-infected individuals will substantially limit the effectiveness of test-and-treat strategies. We review the spectrum of engagement in care for HIV-infected individuals in the United States and apply this information to help understand the magnitude of the challenges that poor engagement in care will pose to test-and-treat strategies for HIV prevention.

Circulating adiponectin levels are lower in Latino versus non-Latino white patients at risk for cardiovascular disease, independent of adiposity measures.

BACKGROUND: Latinos in the United States have a higher prevalence of type 2 diabetes than non-Latino whites, even after controlling for adiposity. Decreased adiponectin is associated with insulin resistance and predicts T2DM, and therefore may mediate this ethnic difference. We compared total and high-molecular-weight (HMW) adiponectin in Latino versus white individuals, identified factors associated with adiponectin in each ethnic group, and measured the contribution of adiponectin to ethnic differences in insulin resistance. METHODS: We utilized cross-sectional data from subjects in the Latinos Using Cardio Health Actions to reduce Risk study. Participants were Latino (n = 119) and non-Latino white (n = 60) men and women with hypertension and at least one other risk factor for CVD (age 61 ± 10 yrs, 49% with T2DM), seen at an integrated community health and hospital system in Denver, Colorado. Total and HMW adiponectin was measured by RIA and ELISA respectively. Fasting glucose and insulin were used to calculate the homeostasis model insulin resistance index (HOMA-IR). Variables independently associated with adiponectin levels were identified by linear regression analyses. Adiponectin's contribution to ethnic differences in insulin resistance was assessed in multivariate linear regression models of Latino ethnicity, with logHOMA-IR as a dependent variable, adjusting for possible confounders including age, gender, adiposity, and renal function. RESULTS: Mean adiponectin levels were lower in Latino than white patients (beta estimates: -4.5 (-6.4, -2.5), p < 0.001 and -1.6 (-2.7, -0.5), p < 0.005 for total and HMW adiponectin), independent of age, gender, BMI/waist circumference, thiazolidinedione use, diabetes status, and renal function. An expected negative association between adiponectin and waist circumference was seen among women and non-Latino white men, but no relationship between these two variables was observed among Latino men. Ethnic differences in logHOMA-IR were no longer observed after controlling for adiponectin levels. CONCLUSIONS: Among patients with CVD risk, total and HMW adiponectin is lower in Latinos, independent of adiposity and other known regulators of adiponectin. Ethnic differences in adiponectin regulation may exist and future research in this area is warranted. Adiponectin levels accounted for the observed variability in insulin resistance, suggesting a contribution of decreased adiponectin to insulin resistance in Latino populations.

Opportunities, Challenges, and Strategies for Engaging Family in Diabetes and Hypertension Management: A Qualitative Study.

Family engagement may improve disease management, yet little is known about this topic as it relates to underrepresented minorities who receive care in low-resource primary care settings. This study aimed to explore family engagement in diabetes and hypertension management at an Urban Indian Health Organization to identify opportunities and challenges, and inform care strategies. We employed semi-structured interviews, genograms, eco-maps, and timelines, among 23 English- and Spanish-speaking American Indian and Latino adults with a dual diagnosis of type 2 diabetes and hypertension and 13 family members. Using thematic analysis, we found that family support is not always available, patients have difficulty sharing medical information, and family often live far away. Conversely, opportunities to leverage included a desire for increased engagement, motivation from the younger generation, prevention within the family, outreach to family members with the same conditions, and learning from elders and ancestors. Implications for programs, clinical care, and research are discussed.

A randomized clinical trial of an interactive voice response and text message intervention for individuals with hypertension.

Interactive voice response and text message (IVR-T) technology may improve hypertension control in under-resourced settings. We conducted a randomized clinical trial to determine whether an IVR-T intervention would improve blood pressure (BP), medication adherence and visit keeping among adults with hypertension from multiple racial and ethnic groups in primary care at an Urban Indian Health Organization in Albuquerque, New Mexico. Two hundred and ninety-five participants were randomly assigned to IVR-T (N = 148) or to usual care (N = 147). The IVR-T arm received reminders for clinic visits, messages to reschedule missed clinic visits, monthly medication refill reminders, weekly motivational messages, and a blood pressure cuff. The usual care arm received no messages. The primary outcome was change in systolic BP (SBP) between baseline and 12 months. Secondary outcomes included change in SBP between baseline and 6 months, change in diastolic BP (DBP) at 6 and 12 months, self-reported adherence at 6 months, and the proportion of missed primary care clinic appointments. The intervention did not affect SBP or DBP at 6 or 12 months. The 12-month change in SBP/DBP was 1.66/1.10 mm Hg in usual care and 0.23/1.34 mm Hg in the intervention group (P values = .57 and .88, respectively). Self-reported medication adherence improved comparably in both groups, and there was no difference in percentage of kept visits. Several features of study design, clinic operations, and data transfer were barriers to demonstrating effectiveness.

An Interactive Voice Response and Text Message Intervention to Improve Blood Pressure Control Among Individuals With Hypertension Receiving Care at an Urban Indian Health Organization: Protocol and Baseline Characteristics of a Pragmatic Randomized Controlled Trial.

BACKGROUND: Efficient and effective strategies for treating chronic health conditions such as hypertension are particularly needed for under-resourced clinics such as Urban Indian Health Organizations (UIHOs). OBJECTIVE: The objective of the Controlling Blood Pressure Trial is to assess the impact of an interactive voice response and text message (IVR-T) intervention compared with usual care among individuals with hypertension receiving care at a UIHO in Albuquerque, New Mexico. This manuscript presents the baseline characteristics of individuals enrolled in the trial and compares their characteristics with those in the hypertension registry who did not enroll in the trial. METHODS: A hypertension registry developed from the clinic's electronic health record was used for recruitment. Potentially eligible participants were contacted by letter and then by phone. Those who expressed interest completed an in-person baseline visit that included a baseline survey and blood pressure measurement using standardized procedures. Individuals randomized to the intervention group could opt to receive either automated text messages or automated phone calls in either English or Spanish. The messages include reminders of upcoming appointments at First Nations Community HealthSource, requests to reschedule recently missed appointments, monthly reminders to refill medications, and weekly motivational messages to encourage self-care, appointment keeping, and medication taking for hypertension. Individuals in the IVR-T arm could opt to nominate a care partner to also receive notices of upcoming and missed appointments. Individuals in the IVR-T arm were also offered a home blood pressure monitor. Follow-up visits will be conducted at 6 months and 12 months. RESULTS: Over a 9.5-month period from April 2017 to January 2018, 295 participants were enrolled from a recruitment list of 1497 individuals. The enrolled cohort had a mean age of 53 years, was 25.1% (74/295) American Indian or Alaska Native and 51.9% (153/295) Hispanic, and 39.0% (115/295) had a baseline blood pressure greater than or equal to 140/90 mmHg. Overall, the differences between those enrolled in the trial and patients with hypertension who were ineligible, those who could not be reached, or those who chose not to enroll were minimal. Enrolled individuals had a slightly lower blood pressure (129/77 mmHg vs 132/79 mmHg; P=.04 for systolic blood pressure and P=.01 for diastolic blood pressure), were more likely to self-pay for their care (26% vs 10%; P<.001), and had a more recent primary care visit (164 days vs 231 days; P<.001). The enrolled cohort reported a high prevalence of poor health, low socioeconomic status, and high levels of basic material needs. CONCLUSIONS: The Controlling Blood Pressure Trial has successfully enrolled a representative sample of individuals receiving health care at a UIHO. Trial follow-up will conclude in February 2019. TRIAL REGISTRATION: ClinicalTrials.gov NCT03135405; http://clinicaltrials.gov/ct2/show/NCT03135405 (Archived by WebCite http://www.webcitation.org/76H2B4SO6). INTERNATIONAL REGISTERED REPORT IDENTIFIER (IRRID): DERR1-10.2196/11794.

Processes of care desired by elderly patients with multimorbidities.

BACKGROUND: Most recommended care for chronic diseases is based on the research of single conditions. There is limited information on 'best' processes of care for persons with multiple morbidities. Our objective was to explore processes of care desired by elderly patients who have multimorbidities that may present competing demands for patients and providers. METHODS: Qualitative investigation using one-on-one interviews of 26 community-dwelling HMO members aged 65-84 (50% male) who had, at a minimum, the combined conditions of diabetes, depression and osteoarthritis. Participants were chosen from a stratified random sample to have a range of 4-16 chronic medical conditions. RESULTS: Participants' desired processes of care included: the need for convenient access to providers (telephone, internet or in person), clear communication of individualized care plans, support from a single coordinator of care who could help prioritize their competing demands and continuity of relationships. They also desired providers who would listen to and acknowledge their needs, appreciate that these' needs were unique and fluctuating and have a caring attitude. CONCLUSIONS: These respondents describe an ideal process of care that is patient centered and individualized and that supports their unique constellations of problems, shifting priorities and multidimensional decision making. Individual and ongoing care coordination managed by a primary contact person may meet some of these needs. Achieving these goals will require developing efficient methods of assessing patient care needs and flexible care management support systems that can respond to patients' needs for different levels of support at different times.

Improving patient care through health-promotion outreach.

Patient receipt of recommended preventive and chronic disease care is suboptimal, partly because of poorly organized clinical information and time-limited and sporadic appointments with medical care providers. To overcome these problems, we describe the features of a novel health-promotion outreach system based on electronic data queries, active patient-centered outreach, and minimizing patient barriers to adherence. We describe advantages of this system over traditional preventive care delivery systems and chronic disease management programs, and explain how it complements and helps to promote the type of clinical practice change contemplated by the chronic care model.

Optimizing number and timing of appointment reminders: a randomized trial.

OBJECTIVES: To compare missed appointment rates for patients receiving a single reminder either 3 days prior to a primary care visit, 1 day prior to the visit, or both 3 days and 1 day prior to the visit. STUDY DESIGN: Three-armed randomized controlled trial. METHODS: Text messages or interactive voice response calls were sent to patients with appointments at 25 primary care clinics in an integrated delivery system. A multivariable prediction model was developed to identify patients at high risk of missing appointments, based on prior appointment-keeping history and other variables from electronic health records. RESULTS: Among 54,066 randomized patients, those who received reminders both 3 days and 1 day prior to the visit were less likely to miss their appointment than those who received only a 3-day or 1-day reminder (4.4% vs 5.8% vs 5.3%, respectively; P <.001). In patients at high risk, 20.5% of those who received 2 reminders missed their visit, compared with 25.0% and 24.2% of those with only 3-day or 1-day reminders, respectively (P <.001). Visit satisfaction was unaffected by providing an additional reminder. CONCLUSIONS: Two automated reminders were more effective than 1 in reducing missed appointments and did not reduce visit satisfaction. A predictive model based on clinical characteristics and prior appointment history can accurately identify patients who are at highest risk of missing appointments. These individuals may benefit more from multiple reminders, but additional strategies are necessary to further reduce their rates of missed appointments.

Methods for engaging stakeholders in comparative effectiveness research: a patient-centered approach to improving diabetes care.

Engaging stakeholders in the research process has the potential to improve quality of care and the patient care experience. Online patient community surveys can elicit important topic areas for comparative effectiveness research. Stakeholder meetings with substantial patient representation, as well as representation from health care delivery systems and research funding agencies, are a valuable tool for selecting and refining pilot research and quality improvement projects. Giving patient stakeholders a deciding vote in selecting pilot research topics helps ensure their 'voice' is heard. Researchers and health care leaders should continue to develop best-practices and strategies for increasing patient involvement in comparative effectiveness and delivery science research.

Interactive behavior change technology. A partial solution to the competing demands of primary care.

BACKGROUND: Primary care practices are faced with the challenge of having too much to do in too little time. As a result, behavioral counseling is often overlooked, especially for patients with multiple health behaviors in need of change. METHODS: This paper describes recent examples of the application of interactive behavior change technologies (IBCTs) to deliver health behavior change counseling before, during, and after the office visit to inform and enhance patient-clinician interactions around these issues. The 5A's framework (assess, advise, agree, assist, arrange follow-up) is used to consider how interactive technology can be used to implement behavior change counseling more consistently. RESULTS: A variety of IBCTs, including the Internet, clinic-based CD-ROMs, and interactive voice-response telephone calls have been shown to be feasible and potentially valuable adjuncts to clinic-based behavioral counseling. These technologies can both increase the effectiveness of behavioral counseling and extend the reach of these services to patients with barriers to face-to-face interactions. CONCLUSIONS: If appropriately developed with the context of primary care in mind and integrated as part of a systems approach to intervention, IBCT can be a feasible and appropriate aid for primary care. Recommendations are made for the types of IBCT aids and research that are needed to realize this potential.

The case for interdepartmental research in primary care.

Research problems in human biology, clinical medicine, and health services delivery are complex, and increasingly require collaborative approaches. Despite the commitment of general internists, general pediatricians, and family physicians to comprehensive, coordinated, and longitudinal care, their substantial overlap in research topics, methods, and data sources, and their shared emphasis on research conducted in community-based settings, the three primary care disciplines rarely collaborate in research. The research enterprises of most primary care departments and divisions in the United States are small "cottage industries," while interdepartmental research units are rare. Interdepartmental research units can develop and maintain the critical mass of investigators, trainees, and staff that is necessary for an intellectually vibrant and financially sound primary care research enterprise. The University of Colorado Health Sciences Center School of Medicine has developed a successful interdepartmental research effort in primary care that includes joint fellowship training and faculty development programs and a Primary Care Research Unit that supports the analysis of secondary databases, practice-based research networks, and interdisciplinary research projects. Key elements of this collaboration include shared projects among faculty and trainees, proactive negotiation about resources, and shared research space, staff, and faculty. Such a collaboration provides the breadth of perspectives necessary to address complex health care problems, and the pragmatic infrastructure necessary to sustain research themes and careers.

Managed care for uninsured adults: the rise and fall of a university-based program.

OBJECTIVE: To assess the impact of CU CARE, a managed care program for medically indigent adults developed by University Hospital (UH) in Denver and Kaiser Permanente, on outpatient and inpatient utilization. STUDY DESIGN: Pre-post study with concurrent comparison groups. PATIENTS AND METHODS: Administrative claims from 1994-1996 were analyzed for all enrollees in a state-funded medically indigent program (intervention group) compared with Medicaid patients and uninsured adults rated as "self-pay" who were ineligible for the medically indigent program. RESULTS: In 1994, before initiation of CU CARE, UH provided care to 10,118 medically indigent, 5330 Medicaid, and 7626 self-pay patients; similar numbers received care in 1995-1996, but only 12% of medically indigent patients received care in both time periods. The proportion of medically indigent patients with 1 or more primary care visits increased by 185% (from 10.9% in 1994 to 31.1% in 1995-1996). Medically indigent patients had relative declines of 36% in specialty clinic visits, 25% in emergency department visits, 40% in hospital visits, and 31% in visit costs between 1994 and 1995-1996. All these changes were significant compared with Medicaid and self-pay patients. The impact on acute care utilization was greater for medically indigent patients who used UH in both 1994 and 1995-1996. CONCLUSIONS: This managed care program increased utilization of primary care and reduced specialty and acute care utilization. However, the program was scaled back in 1997 and terminated in 2000 because of problems with care coordination across institutions, increasing costs (particularly pharmacy costs), and competitive pressures.

Disenrollment from a state child health insurance plan: are families jumping S(c)HIP?

BACKGROUND: Colorado's Child Health Plan Plus (CHP+) is a non-Medicaid state child health plan that began enrollment in April 1998. Families are disenrolled 12 months after enrollment if they fail to re-enroll. OBJECTIVE: To assess insurance coverage before and 1 year after initial enrollment in CHP+; reasons for disenrollment; and factors associated with re-enrollment. DESIGN/METHODS: We interviewed 480 randomly selected families 2 months after initial enrollment into CHP+ (September 1999 through January 2000) and 1 year later. RESULTS: Prior to CHP+, 38% of families had Medicaid (MK), 35% were privately insured (PI), 6% were uninsured (UI), and 20% had other/unknown insurance. After the 12 months, 34% were re-enrolled, 16% got other insurance (6% MK, 10% PI/other), 4% had children older than 18 years, and 46% were UI (9% had intentionally and 37% had unintentionally disenrolled from CHP+). All unintentionally disenrolled families were planning to re-enroll and 90% still appeared eligible. In multivariate analysis, having a primary care provider prior to enrollment was associated with re-enrollment (odds ratio [OR] 1.7, 95% confidence interval [CI] 1.1 to 2.6), but having problems with the application process impeded re-enrollment (OR 0.7, 95% CI 0.6 to 0.9). CONCLUSIONS: Only about a third of families eligible for State Children's Health Insurance Program successfully re-enrolled before their termination date. Institution of a passive renewal process would decrease unnecessary disenrollment in eligible families.

Can paraprofessional home visitation enhance early intervention service delivery?

A 1-year randomized trial compared intensive case management (ICM) versus basic case management (BCM) in facilitating early intervention (EI) service use among children in an urban health system. Of 159 participating families with delayed or at-risk preschool-aged children, 88 received ICM from paraprofessionals versus 71 families who received less comprehensive BCM from a nurse. In the ICM versus BCM group, a shorter interval to assessment (98 vs 140 d, p =.05) but similar assessment rate (86% vs 80%, p =.29) was observed. The ICM group had more services recommended per child (1.64 vs 1.16, p < .004) and initiated (1.20 vs 0.85, p < .04). There was no difference in median time to EI program initiation for ICM versus BCM (228 vs 200 d, p = .88) or initiation and visit compliance rate for EI services. Specific efforts to improve outcomes (e.g., decrease initiation time and increase use of EI services) are still needed.