Community Health Worker Interventions to Improve Glycemic Control in People with Diabetes: A Systematic Review and Meta-Analysis.

OBJECTIVES: We set out to review the efficacy of Community Health Worker (CHW) interventions to improve glycemia in people with diabetes. METHODS: Data sources included the Cochrane Central Register of Controlled Trials, Medline, clinicaltrials.gov, Google Scholar, and reference lists of previous publications. We reviewed randomized controlled trials (RCTs) that assessed the efficacy of CHW interventions, as compared to usual care, to lower hemoglobin A1c (A1c). Two investigators independently reviewed the RCTs and assessed their quality. Only RCTs with a follow-up of at least 12 months were meta-analyzed. A random effects model was used to estimate, from unadjusted within-group mean reductions, the standardized mean difference (SMD) in A1c achieved by the CHW intervention, beyond usual care. RESULTS: Thirteen RCTs were included in the narrative review, and nine of them, which had at least 12 months of follow-up, were included in the meta-analysis. Publication bias could not be ruled-out due to the small number of trials. Outcome heterogeneity was moderate (I(2)= 37%). The SMD in A1c (95% confidence interval) was 0.21 (0.11-0.32). Meta-regression showed an association between higher baseline A1c and a larger effect size. CONCLUSIONS: CHW interventions showed a modest reduction in A1c compared to usual care. A1c reduction was larger in studies with higher mean baseline A1c. Caution is warranted, given the small number of studies.

Discrimination and Depression among Urban Hispanics with Poorly Controlled Diabetes.

OBJECTIVES: We had three objectives for our study: 1) to describe the prevalence and burden of experiences of discrimination among Hispanics with poorly controlled diabetes; 2) to evaluate associations among discrimination experiences and their burden with comorbid depression among Hispanics with poorly controlled diabetes; and 3) to evaluate whether discrimination encountered in the health care context itself was associated with comorbid depression for Hispanic adults with diabetes. DESIGN: We conducted a cross-sectional analysis of baseline data of a randomized controlled trial (RCT). SETTING: We collected data in the context of an RCT in a clinical setting in New York City. PARTICIPANTS: Our sample comprised 221 urban-dwelling Hispanics, largely of Caribbean origin. MAIN OUTCOME MEASURES: The main outcome measure was major depression, measured by the Euro-D (score > 3). RESULTS: Of 221 participants, 58.8% reported at least one experience of everyday discrimination, and 42.5% reported at least one major experience of discrimination. Depression was associated significantly with counts of experiences of major discrimination (OR = 1.46, 95% CI = 1.09 - 1.94, P = .01), aggregate counts of everyday and major discrimination (OR = 1.13, 95% CI = 1.02 - 1.26, P = .02), and the experience of discrimination in getting care for physical health (OR = 6.30, 95% CI= 1.10-36.03). CONCLUSIONS: Discrimination may pose a barrier to getting health care and may be associated with depression among Hispanics with diabetes. Clinicians treating Caribbean-born Hispanics should be aware that disadvantage and discrimination likely complicate a presentation of diabetes.

Early results of an integrated maternal, newborn, and child health program, Northern Nigeria, 2009 to 2011.

BACKGROUND: This paper describes early results of an integrated maternal, newborn, and child health (MNCH) program in Northern Nigeria where child mortality rates are two to three times higher than in the southern states. The intervention model integrated critical health systems changes needed to reinvigorate MNCH health services, together with community-based activities aimed at mobilizing and enabling women to make changes in their MNCH practices. Control Local Government Areas received less-intense statewide policy changes. METHODS: The impact of the intervention was assessed using a quasi-experimental design, comparing MNCH behaviors and outcomes in the intervention and control areas, before and after implementation of the systems and community activities. Stratified random household surveys were conducted at baseline in 2009 (n = 2,129) and in 2011 at follow-up (n = 2310), with women with births in the five years prior to household surveys. Chi-square and t-tests were used to document presence of significant improvements in several MNCH outcomes. RESULTS: Between baseline and follow-up, anti-tetanus vaccination rates increased from 69.0% to 85.0%, and early breastfeeding also increased, from 42.9% to 57.5%. More newborns were checked by trained health workers (39.2% to 75.5%), and women were performing more of the critical newborn care activities at follow-up. Fewer women relied on the traditional birth attendant for health advice (48.4% to 11.0%, with corresponding increases in advice from trained health workers. At follow-up, most of these improvements were greater in the intervention than control communities. In the intervention communities, there was less use of anti-malarials for all symptoms, coupled with more use of other medications and traditional, herbal remedies. Infant and child mortality declined in both intervention and control communities, with the greatest declines in intervention communities. In the intervention communities, infant mortality rate declined from 90 at baseline to 59 at follow-up, while child mortality declined from 160 to 84. CONCLUSIONS: These results provide evidence that in the context of ongoing improvements to the primary health care system, the participatory and community-based interventions focusing on improved newborn and infant care were effective at changing infant care practices and outcomes in the intervention communities.

Awareness, use, and unmet need for family planning in rural northern Nigeria.

Access to quality reproductive health and family planning services remain poor in Nigeria. We present results on family planning awareness and use from a survey of 3,080 women (age 15-49 years) in Jigawa, Katsina, Yobe, and Zamfara States. About 43.0% had heard of any method of contraception whereas 36.6% had heard of any modern method. Overall, 7.0% of all currently married women reported ever using a method of contraception; 4.4% used a modern method and 2.9% used a traditional method. Only 1.3% of women in union (currently married or cohabiting) used modern contraception methods at the time of the survey; 1.3% of women in union used traditional methods. Unmet need for family planning was 10.3%. Low family planning use in the presence of low awareness and low felt need suggests, among other things, a need to increase awareness and uptake and make family planning commodities available.

Building a consensus on community health workers' scope of practice: lessons from New York.

OBJECTIVES: We evaluated efforts in New York to build a consensus between community health workers (CHWs) and employers on CHWs' scope of practice, training standards, and certification procedures. METHODS: We conducted multiple-choice surveys in 2008 and 2010 with 226 CHWs and 44 employers. We compared CHWs' and employers' recommendations regarding 28 scope of practice elements. The participatory ranking method was used to identify consensus scope of practice recommendations. RESULTS: There was consensus on 5 scope of practice elements: outreach and community organizing, case management and care coordination, home visits, health education and coaching, and system navigation. For each element, 3 to 4 essential skills were identified, giving a total of 27 skills. These included all skills recommended in national CHW studies, along with 3 unique to New York: computer skills, participatory research methods, and time management. CONCLUSIONS: CHWs and employers in New York were in consensus on CHWs' scope of practice on virtually all of the detailed core competency skills. The CHW scope of practice recommendations of these groups can help other states refine their scope of practice elements.

Using community-based research to shape the design and delivery of maternal health services in Northern Nigeria.

Maternal mortality ratios in northern Nigeria are among the worst in the world, over 1,000 per 100,000 live births in 2008, with a very low level and quality of maternity services. In 2009, we carried out a study of the reasons for low utilisation of antenatal and delivery care among women with recent pregnancies, and the socio-cultural beliefs and practices that influenced them. The study included a quantitative survey of 6,882 married women, 119 interviews and 95 focus group discussions with community and local government leaders, traditional birth attendants, women who had attended maternity services and health care providers. Only 26% of the women surveyed had received any antenatal care and only 13% delivered in a facility with a skilled birth attendant for their most recent pregnancy. However, those who had had at least one antenatal consultation were 7.6 times more likely to deliver with a skilled birth attendant. Most pregnant women had little or no contact with the health care system for reasons of custom, lack of perceived need, distance, lack of transport, lack of permission, cost and/or unwillingness to see a male doctor. Based on these findings, we designed and implemented an integrated package of interventions that included upgrading antenatal, delivery and emergency obstetric care; providing training, supervision and support for new midwives in primary health centres and hospitals; and providing information to the community about safe pregnancy and delivery and the use of these services.

Determinants of influenza vaccination among young children in an inner-city community.

Few studies have examined potential factors that contribute to low influenza vaccination rates among minority children. This study aimed to assess the prevalence of early childhood influenza vaccination among young black and Latino children, living in inner-city neighborhoods, and examine the effects of child, caregiver and health system factors. Secondary data analysis was performed using a survey about medical home experiences conducted from May 2007-June 2008. The study sample was limited to children ≥6 months in any influenza season prior to the 2006-2007 influenza season. Bivariate analyses and multivariable logistic regression tested associations between influenza vaccination receipt and socio-demographic and health system characteristics. One-third of children received an influenza vaccination by the end of 2006-2007 season, while only 11% received a vaccination within their first season of eligibility. Black children were more likely than Latino children to have been vaccinated (50% vs. 31%, P<0.01) during their first few eligible seasons. Children whose mothers were older, proficient in English, and frequent users of healthcare were more likely to obtain vaccination. Child attendance at healthcare settings with immunization reminder systems was also positively correlated with influenza vaccination. Our findings suggest that initial vaccination receipt among minority children from inner-city communities might be improved by expanded influenza promotion activities targeting younger mothers or those with limited English proficiency. Strategies to increase the frequency of child's actual contact with the medical home, such as reminder systems, may be useful in improving uptake of influenza vaccination among inner-city, minority children.

FluAlert: a qualitative evaluation of providers' desired characteristics and concerns regarding computerized influenza vaccination alerts.

OBJECTIVE: To explore pediatric providers' perceived barriers to influenza vaccine delivery, and desired characteristics and potential concerns regarding an influenza vaccine alert integrated into the electronic health record (EHR). METHODS: Four focus groups with providers (n=21) and five individual interviews with practice leaders in an urban, pediatric primary care network affiliated with an academic medical center in New York City were conducted. Data were collected during the 2009-10 influenza season and analyzed using thematic analysis. RESULTS: Participants identified several barriers to influenza vaccine delivery, including remembering to vaccinate during sick visits, need to review multiple sources of immunization information, time shortages and inadequate staffing. They felt that an alert could help many of these challenges. They desired the following alert characteristics: 1) alerting providers early in the visit, 2) accurately determining patients' vaccine status by merging multiple sources of immunization information, 3) facilitating vaccine ordering, and 4) generating appropriate documentation in the EHR when vaccines were refused or otherwise not given. Potential concerns regarding the alert included reliability and accuracy of alert, workflow interruptions and forced actions. CONCLUSIONS: This study highlights providers' interest in a well-integrated, accurate alert that streamlines assessment of vaccination eligibility, ordering and documentation without impeding work-flow.

A community-based strategy for improving asthma management and outcomes for preschoolers.

Although almost one in ten (8.6%) preschool children has been diagnosed with asthma, few asthma management programs are designed for parents of preschool children. The Asthma Basics for Children program addressed this need in 2003-2008 by implementing a multi-layered approach that offered educational activities to center staff, parents, and children and PACE training to physicians in 31 Northern Manhattan daycare centers. Following program participation, 85% of parents reported reducing their child's triggers, 89% said it was easier to talk to their child's physician, and 80% were confident in their ability to manage their child's asthma. Children's any daytime symptoms dropped from 78% to 42%, any nighttime symptoms from 81% to 49%, any daycare absences from 56% to 38%, any asthma-related emergency department (ED) visits from 74% to 47%, and any asthma-related hospitalizations from 24% to 11% (p < .001 for all differences). Outcomes varied by level of exposure. In the Center-Only group (no parent participation), the only reduction was from 19% to 10% (McNemar = 3.77, p = .052) in hospitalizations. Children whose parents participated in the program had significant reductions in daycare absences (62% to 38%, McNemar = 11.1, p < .001), ED visits (72% to 43%, McNemar = 19.2, p < .001), and hospitalizations (24% to 11%, McNemar = 5.54, p = .018). Children whose parents and healthcare provider participated had the greatest improvements with asthma-related daycare absences dropping from 62% to 32% (McNemar = 9.8, p = .001), ED visits from 72% to 37% (McNemar = 14.4, p < .001), and hospitalizations from 35% to 15% (McNemar = 8.33, p = .003). This study demonstrates that a multi-layered approach can improve asthma outcomes among preschoolers with a combination of parent and provider education having the greatest impact.

Change in parental reasons for use of an urban pediatric emergency department in the past decade.

OBJECTIVE: To assess changes over the past decade in parental reasons associated with nonurgent visits to pediatric emergency departments (PEDs) during regular primary care office hours. METHODS: Secondary analysis of cross-sectional surveys of families of children younger than 3 years visiting a PED in a low-socioeconomic area in New York City conducted in 1997 and 2006. We performed multivariable analyses to assess differences in parental reported reasons for PED use over the period, controlling for sociodemographic factors. RESULTS: Most children (95.6%) had a usual source of care across both periods. Compared with those seen in 1997, children seen in 2006 were far less likely to be brought to the PED during regular primary care office hours for parental perceived urgency (adjusted odds ratio [AOR], 0.076; 95% confidence interval [CI], 0.024-0.24; P < 0.001). At the same time, these children were more likely to be brought to the PED for limited access to their usual source of care (AOR, 3.35; 95% CI, 1.24-9.02; P < 0.05) and greater trust in the medical expertise of the PED (AOR, 5.95 95% CI, 1.20-29.45; P < 0.05). CONCLUSIONS: Over the last decade, despite the presence of a usual source of care, a greater number of parents report visiting this urban PED during regular office hours for reasons unrelated to parental perceived urgency. Limited access to care and greater trust in the medical expertise available in PEDs have played important roles. Approaches to decreasing nonurgent visits must take into account all of these factors.

Changes in childhood immunization disparities between central cities and their respective states, 2000 versus 2006.

Central cities have lower childhood immunization coverage rates than states in which they are located. We conducted a secondary analysis of the National Immunization Survey (NIS) 2000 and 2006 of children 19-35 months old for 26 NIS-defined central cities and the rest of their respective states in order to examine patterns in early childhood immunization disparities between central cities and their respective states and the contextual factors associated with these disparities. We examined three measures of immunization disparities (absolute, difference, and ratio of change) and the patterns of disparity change with regard to selected contextual factors derived from the census. In 2000, immunization coverage in central cities was 68.3% and 74.7% in the rest of their states, a 6.4% disparity (t = 3.82, p < 0.000). Between 2000 and 2006, the overall city/state disparity narrowed to 3.5%, with the central city coverage up to 78.7% vs. 82.5% for the rest of state (t = 2.48, p = 0.017). However, changes in immunization disparities were not uniform: six cities narrowed, 14 had minimal change, and six widened. Central cities with a larger share of Hispanics experienced less reduction in disparities than other cities (beta = -4.2, t = -2.11, p = 0.047). Despite overall progress in childhood immunization coverage, most central cities still show significant disparities with respect to the rest of their states. Cities with larger Hispanic populations may need extra help in narrowing their disparities.

Effective strategies for integrating immunization promotion into community programs.

Most immunization coalitions have originated with health care providers, potentially excluding families without medical homes. This study focused on a community-based approach to providing timely vaccinations. A coalition of 23 organizations developed an immunization program in a low-income community in New York City. Nearly 1,000 community health workers incorporated immunization promotion into social service and educational programs. Outcomes were coverage rates for the 4:3:1:3:3 series at 19 to 35 months, which were compared with national data by ethnicity, as reported in the National Immunization Survey 2002-2006. Parents (n = 10,251) of children <5 years received immunization education and reminders. The 2003-2007 rates of 80% equaled or exceeded the national rates for 19- to 35-month-olds, and the 2007 rate of 96.8% far surpassed the national average. Coalitions can effectively integrate immunization promotion activities into community programs. Immunization rate improvements maintained for a 5-year period, suggesting this approach to be sustainable.

Effectiveness of a community coalition for improving child vaccination rates in New York City.

We used a retrospective, matching, birth cohort design to evaluate a comprehensive, coalition-led childhood immunization program of outreach, education, and reminders in a Latino, urban community. After we controlled for Latino ethnicity and Medicaid, we found that children enrolled in the program were 53% more likely to be up-to-date (adjusted odds ratio = 1.53; 95% confidence interval = 1.33, 1.75) and to receive timely immunizations than were children in the control group (t = 3.91). The coalition-led, community-based immunization program was effective in improving on-time childhood immunization coverage.

The impact of community health worker training and programs in NYC.

The Northern Manhattan Community Voices Collaborative is committed to improving health care in Harlem, Washington Heights, Inwood, and low-income communities in New York City, large parts of which are home to many immigrants to the U.S. The collaborative developed a program to train and integrate community health workers (CHWs) into ongoing programs at partner community organizations. We report on our 2000-2005 experiences with CHWs for health insurance, child immunizations, and asthma management. A total of 1,504 CHWs were trained, with 16%-200% increase in CHW competency for selected skills. The CHWs facilitated health insurance enrollment for about 30,000 individuals, assisted 8,000 children to become completely immunized, and supported 4,000 families improving asthma management. Integration of CHW training into community programs is effective for empowering health promotion in underserved communities.

Community-based strategies to reduce childhood immunization disparities.

This study demonstrates how community-based immunization promotion reduced immunization disparities. In 2002 to 2004, the coalition enrolled 3,748 children younger than 5, with 1,502 aged 19 to 35 months in April 2004. Disparity reduction was assessed by comparing coalition immunization coverage rates (4:3:1:3:3) to the National Immunization Survey 2003 rates. Logistic regression was used to assess factors contributing to up-to-date immunizations. Coverage increased from 46.0% at enrollment to 80.5%, matching nationwide rates for all (t = 0.87) or White (t = 1.99) children. The 78% for African Americans was higher than 73% for U.S. African American children (t = 2.90); 84% for Latinos was higher than 77% for U.S. Latinos (t = 2.32). Being current with age-appropriate immunizations at enrollment (OR = 9.8), being Latino (OR = 1.6), and participating through child health insurance enrollment (OR = 4.9), Women, Infants, and Children (OR = 3.1), or child care or parenting (OR = 1.9) programs increased immunization coverage. Embedding immunization promotion into existing community programs was successful in eliminating immunization disparities. Most effective programs were those with direct linkages to health care systems or that targeted young children.

Female health workers at the doorstep: a pilot of community-based maternal, newborn, and child health service delivery in northern Nigeria.

INTRODUCTION: Nigeria has one of the highest maternal mortality ratios in the world. Poor health outcomes are linked to weak health infrastructure, barriers to service access, and consequent low rates of service utilization. In the northern state of Jigawa, a pilot study was conducted to explore the feasibility of deploying resident female Community Health Extension Workers (CHEWs) to rural areas to provide essential maternal, newborn, and child health services. METHODS: Between February and August 2011, a quasi-experimental design compared service utilization in the pilot community of Kadawawa, which deployed female resident CHEWs to provide health post services, 24/7 emergency access, and home visits, with the control community of Kafin Baka. In addition, we analyzed data from the preceding year in Kadawawa, and also compared service utilization data in Kadawawa from 2008-2010 (before introduction of the pilot) with data from 2011-2013 (during and after the pilot) to gauge sustainability of the model. RESULTS: Following deployment of female CHEWs to Kadawawa in 2011, there was more than a 500% increase in rates of health post visits compared with 2010, from about 1.5 monthly visits per 100 population to about 8 monthly visits per 100. Health post visit rates were between 1.4 and 5.5 times higher in the intervention community than in the control community. Monthly antenatal care coverage in Kadawawa during the pilot period ranged from 11.9% to 21.3%, up from 0.9% to 5.8% in the preceding year. Coverage in Kafin Baka ranged from 0% to 3%. Facility-based deliveries by a skilled birth attendant more than doubled in Kadawawa compared with the preceding year (105 vs. 43 deliveries total, respectively). There was evidence of sustainability of these changes over the 2 subsequent years. CONCLUSION: Community-based service delivery through a resident female community health worker can increase health service utilization in rural, hard-to-reach areas.

Reinvigorating health systems and community-based services to improve maternal health outcomes: case study from northern Nigeria.

BACKGROUND: Maternal health outcomes in Nigeria, the most populous African nation, are among the worst in the world, and urgent efforts to improve the situation are critical as the deadline (2015) for achieving the Millennium Development Goals draws near. OBJECTIVE: To evaluate the results of an integrated maternal, newborn, and child health (MNCH) program to improve maternal health outcomes in Northern Nigeria. DESIGN: The intervention model integrated critical health system and community-based improvements aimed at encouraging sustainable MNCH behavior change. Control Local Government Areas received less intense statewide policy changes. METHODS: We assessed the impact of the intervention on maternal health outcomes in 3 northern Nigerian states by comparing data from 2360 women in 2009 and 4628 women in 2013 who had a birth or pregnancy in the 5 years prior to the survey. RESULTS: From 2009 to 2013, women with standing permission from their husband to go to the health center doubled (from 40.2% to 82.7%), and health care utilization increased. The proportions of women who delivered with a skilled birth attendant increased from 11.2% to 23.9%, and the proportion of women having at least 1 antenatal care (ANC) visit doubled from 24.9% to 48.8%. ANC was increasingly provided by trained community health extension workers at the primary health center, who provided ANC to 34% of all women with recent pregnancies in 2013. In 2013, 22% of women knew at least 4 maternal danger signs compared with 10% in 2009. Improvements were significantly greater in the intervention communities that received the additional demand-side interventions. CONCLUSIONS: The improvements between 2009 and 2013 demonstrate the measurable impact on maternal health outcomes of the program through local communities and primary health care services. The significant improvements in communities with the complete intervention show the importance of an integrated approach blending supply- and demand-side interventions.

Registry-linked electronic influenza vaccine provider reminders: a cluster-crossover trial.

OBJECTIVE: To determine the impact of a vaccination reminder in an electronic health record supplemented with data from an immunization information system (IIS). METHODS: A noninterruptive influenza vaccination reminder, based on a real-time query of hospital and city IIS, was used at 4 urban, academically affiliated clinics serving a low-income population. Using a randomized cluster-crossover design, each study site had "on" and "off" period during the fall and winter of 2011-2012. Influenza vaccination during a clinic visit was assessed for 6-month to 17-year-old patients. To assess sustainability, the reminder was active at all sites during the 2012-2013 season. RESULTS: In the 2011-2012 season, 8481 unique non-up-to-date children had visits. Slightly more non-up-to-date children seen when the reminder was 'on' were vaccinated than when 'off' (76.2% vs 73.8%; P = .027). Effects were seen in the winter (67.9% vs 62.2%; P = .005), not fall (76.8% vs 76.5%). The reminder also increased documentation of the reason for vaccine non-administration (68.1% vs 41.5%; P < .0001). During the 2011-2012 season, the reminder displayed for 8630 unique visits, and clinicians interacted with it in 83.1% of cases where patients required vaccination. During the 2012-2013 season, it displayed for 22 248 unique visits; clinicians interacted with it in 84.8% of cases. CONCLUSIONS: An IIS-linked influenza vaccination reminder increased vaccination later in the winter when fewer vaccine doses are usually given. Although the reminder did not require clinicians to interact with it, they frequently did; utilization did not wane over time.

Impact of the Advisory Committee on Immunization Practices' 4-day grace period in a low-income community.

BACKGROUND: In 2002, the Advisory Committee on Immunization Practices (ACIP) recommended vaccine doses administered < or = 4 days before the minimum age or interval be counted as valid. The study objective was to assess the impact of the 4-day grace period on the need for revaccination and associated costs in a low-income community, compared to standard practice (i.e., repeating all doses that fall outside current ACIP guidelines). METHODS: From 1999 to 2001, semi-annual immunization assessments of 8293 randomly selected children, aged 19-35 months, were conducted at a 16-practice network serving an underserved community in New York City. Outcome measures were rates of antigen-specific invalid doses and number of children needing revaccination, with and without the 4-day grace period. Revaccination costs were based on the Vaccines for Children (VFC) price list. RESULTS: The 4-day grace period reduced the number of children needing revaccination from 17.1% to 12.0%, a drop of 30%. The rates of invalid doses decreased from 1.9% to 1.3%, a drop of 33%. Invalid doses for hepatitis B (HepB)-2 decreased by two thirds (69.7%); for diphtheria-tetanus-acellular pertussis (DTaP)-1, Haemophilus influenzae type b (Hib)-1 and Hib-3 by half (44.9%-50.0%); for Polio-1, Polio-2, Polio-3, and measles-mumps-rubella (MMR) by one third (31.6%-33.3%); and for DTaP-2, DTaP-3, HepB-3, and varicella by nearly one quarter (20.0%-24.0%). At these rates, revaccinating 100,000 children younger than age 3 years would cost 213,588 dollars per year, compared to 152,539 dollars with the 4-day grace period, in vaccine costs alone. CONCLUSIONS: In a low-income community, ACIP's 4-day grace period made a significant impact on the number of children requiring revaccination and on revaccination costs. However, the number of children needing revaccination remains high.

Early continuity of care and immunization coverage.

OBJECTIVE: We examined the relationship between early and exclusive continuity of care at the initial source of care and immunization coverage. METHODS: We used a cohort study design with 641 randomly selected children initiating care before 3 months and making 2 or more visits to an inner-city practice network. We used 2 complementary data sources: medical records and the New York City Department of Health Citywide Immunization Registry. Immunization measures were cumulative age appropriate and up-to-date at 18 months (UTD18). RESULTS: There was a gradual attrition from the initial source of care. By 18 months, less than half the children (46%) remained in care. Regardless of continuity, nearly half (42%) had used other immunization providers. The initial source of care contributed most immunizations (89%-94%); however, across all levels of continuity, children who also used other providers had higher immunization rates. We found a threshold effect of continuity beginning at 12 months: children in care from 12 to 14 months were 17.5 times more likely to be UTD18 than those in care less than 6 months. Each additional period in care increased the time remaining current with immunizations. Among children UTD18, 88% were in care at 11 months compared with 38% among those not UTD18, a 50% difference. CONCLUSIONS: Continuity of care at the initial source of care had a significant and lasting impact on immunization coverage, even if not used exclusively. Interventions promoting continued use of the medical home over the first 2 years of life may help improve immunization coverage.