Use of spatial epidemiology and hot spot analysis to target women eligible for prenatal women, infants, and children services

The spatial distribution mechanism of PM2.5 and NO2 on the eastern coast of China

The spatial distribution characteristics of multi-air pollutants and their impacts are difficult to quantify effectively. As PM2.5 and NO2 are the main air pollutants, it is of great significance to explore the spatial causes of their pollution and their interaction mechanism. This study used machine learning (LightGBM) and hot spot analysis to map the spatial distribution of PM2.5 and NO2 in Southwest Fujian (SWFJ) in 2018 and their key pollution areas. Then, the factors and interactive detection of geographical detectors were used to conduct a detailed analysis of the quantitative impact of potential factors such as human activities, terrain, air pollutants, and meteorology on PM2.5 and NO2 pollution. From this we can learn that 1. LightGBM has good stability for drawing the spatial distribution of PM2.5 and NO2. 2. The spatial mechanism of PM2.5 and NO2 can be effectively interpreted from a massive data and macro perspective. 3. A large amount of evidence shows that potential factors such as human activities, topography, air pollutants and meteorology have direct or indirect effects on PM2.5 and NO2 pollution in the SWFJ area. This includes the direct impact of local road traffic emissions on the distribution of PM2.5 and NO2 pollution, the digestion of both by vegetation, the mutual transformation of atmospheric pollutants themselves, and the impact of meteorological conditions. This study not only confirms the effectiveness of machine learning combined with geographical detectors to promote the study of regional air pollution mechanisms, but also confirms the feasibility of exploring the spatial distribution mechanisms of various air pollutants. Therefore, this study is of great significance for explaining the spatial distribution of PM2.5 and NO2, and can also provide reference for policy formulation to reduce regional PM2.5 and NO2 concentrations.

Who Are the Local Policy Innovators? Cluster Analysis of Municipal Tobacco Control Policies in Massachusetts

CONTEXT

Tobacco use is a leading cause of preventable death, yet it is challenging to establish public policy to reduce tobacco use. Massachusetts has been a national leader in tobacco control, and its policy-making patterns can be informative to the country.

OBJECTIVE

To identify factors associated with the adoption of tobacco policy within 351 Massachusetts municipalities.

DESIGN

We obtained the 2019 Massachusetts municipality-level tobacco control policy information from Massachusetts' Tobacco Automated Fact Sheet Information system and compiled it with data from American Community Survey, Massachusetts Municipal Association, and Massachusetts state government's Web sites. We used k -means clustering method to identify statistical clustering patterns and hotspot analysis (Getis-Ord Gi*) and Local Indicators of Spatial Association to identify geographic clustering patterns. We then performed multinomial logistic regression to identify factors associated with policy clusters.

SETTING

Massachusetts.

PARTICIPANTS

Three hundred fifty-one municipalities in Massachusetts.

MAIN OUTCOME MEASURE

Policy clusters-groups of municipalities with similar tobacco control policy behaviors.

RESULTS

Based on the k -means analyses, we identified 3 clusters in Massachusetts municipal tobacco control policy behaviors: 54% (N = 191) of municipalities were "Policy Leaders" with a high adoption rate of the 6 tobacco control policies; 18% (N = 63) were "Peer-Influenced Actors" focused on tobacco purchase restrictions for individuals younger than 21 years; and 28% (N = 97) were "Policy Non-Actors," with no tobacco control policies in place. Policy Leaders were geographically clustered in larger cities and the MetroWest region. Policy Non-Actors were clustered in rural areas of Western and Central Massachusetts. Larger municipal population size, higher municipal tax income, and higher percentages of residents voting Democratic were associated with higher policy adoption activities.

CONCLUSIONS

Local variation in the adoption of tobacco policies may exacerbate inequities in tobacco use and population health. Opportunities remain to implement additional tobacco control regulations at the local level to promote public health.

Impact of urbanisation and environmental factors on spatial distribution of COVID-19 cases during the early phase of epidemic in Singapore

Geographical weighted regression (GWR) can be used to explore the COVID-19 transmission pattern between cases. This study aimed to explore the influence from environmental and urbanisation factors, and the spatial relationship between epidemiologically-linked, unlinked and imported cases during the early phase of the epidemic in Singapore. Spatial relationships were evaluated with GWR modelling. Community COVID-19 cases with residential location reported from 21st January 2020 till 17th March 2020 were considered for analyses. Temperature, relative humidity, population density and urbanisation are the variables used as exploratory variables for analysis. ArcGIS was used to process the data and perform geospatial analyses. During the early phase of COVID-19 epidemic in Singapore, significant but weak correlation of temperature with COVID-19 incidence (significance 0.5-1.5) was observed in several sub-zones of Singapore. Correlations between humidity and incidence could not be established. Across sub-zones, high residential population density and high levels of urbanisation were associated with COVID-19 incidence. The incidence of COVID-19 case types (linked, unlinked and imported) within sub-zones varied differently, especially those in the western and north-eastern regions of Singapore. Areas with both high residential population density and high levels of urbanisation are potential risk factors for COVID-19 transmission. These findings provide further insights for directing appropriate resources to enhance infection prevention and control strategies to contain COVID-19 transmission.

Examining the spatial risk environment tied to the opioid crisis through a unique public health, EMS, and academic research collaborative: Lowell, Massachusetts, 2008–2018

Between 2015 and 2018, Lowell Massachusetts experienced outbreaks in opioid overdoses, HIV, and hepatitis C virus infections (HCV) among people who inject drugs. Through an innovative collaboration between emergency medical services (EMS), public health, and academic partners, we assessed the geographic distribution of opioid-related risks to inform intervention efforts. We analyzed data from three unique data sources for publicly discarded syringes, opioid-related incidents (ORIs), and fatal opioid overdoses in Lowell between 2008 and 2018. We assessed the risk environment over time using a geographic information system to identify and characterize hotspots and noted parallel trends within the syringe discard and ORI data. We identified two notable increases in ORIs per day: the first occurring between 2008 and 2010 (from 0.3 to 0.5), and the second between 2011 and 2014 (from 0.9 to 1.3), following the introduction of fentanyl within local drug markets. We also identified seasonal patterns in the syringe discard, ORI, and overdose data. Through our spatial analyses, we identified significant clusters of discarded syringes, ORIs, and fatal overdoses (p < 0.05), and neighborhoods where high densities of these outcomes overlapped. We found that areas with the highest densities shifted over time, expanding beyond the epicenter of the Downtown neighborhood. Data sharing and analyses among EMS, public health, and academic partners can foster better assessments of local risk environments. Our work, along with new public health efforts in Lowell, led to a city-funded position to improve pick-up and proper disposal of publicly discarded syringes, and better targeted harm reduction services.

Geospatial Analysis of Sodium and Potassium Intake: A Swiss Population-Based Study

Inadequate sodium and potassium dietary intakes are associated with major, yet preventable, health consequences. Local public health interventions can be facilitated and informed by fine-scale geospatial analyses. In this study, we assess the existence of spatial clustering (i.e., an unusual concentration of individuals with a specific outcome in space) of estimated sodium (Na), potassium (K) intakes, and Na:K ratio in the Bus Santé 1992–2018 annual population-based surveys, including 22,495 participants aged 20–74 years, residing in the canton of Geneva, using the local Moran’s I spatial statistics. We also investigate whether socio-demographic and food environment characteristics are associated with identified spatial clustering, using both global ordinary least squares (OLS) and local geographically weighted regression (GWR) modeling. We identified clear spatial clustering of Na:K ratio, Na, and K intakes. The GWR outperformed the OLS models and revealed spatial variations in the associations between explanatory and outcome variables. Older age, being a woman, higher education, and having a lower access to supermarkets were associated with higher Na:K ratio, while the opposite was seen for having the Swiss nationality. Socio-demographic characteristics explained a major part of the identified clusters. Socio-demographic and food environment characteristics significantly differed between individuals in spatial clusters of high and low Na:K ratio, Na, and K intakes. These findings could guide prioritized place-based interventions tailored to the characteristics of the identified populations.

Geographic Variation in the Prevalence of Rheumatoid Arthritis in Alberta, Canada

OBJECTIVES

Timely access to rheumatologists remains a challenge in Alberta, a Canadian province with vast rural areas, whereas rheumatologists are primarily clustered in metro areas. To address the goal of timely and equitable access to rheumatoid arthritis (RA) care, health planners require information at the regional and local level to determine the RA prevalence and the associated health care needs.

METHODS

Using Alberta Health administrative databases, we identified RA-prevalent cases (April 1, 2015-March 31, 2016) on the basis of a validated case definition. Age- and sex-standardized prevalence rates per 1000 population members and the standardized rates ratio (SRR) were calculated. We applied Global Moran's I and Gi* hotspot analysis using three different weight matrices to explore the geospatial pattern of RA prevalence in Alberta.

RESULTS

Among 38 350 RA cases (68% female; n = 26 236), the prevalence rate was 11.81 cases per 1000 population members (95% confidence interval [CI] 11.80-11.81) after age and sex standardization. Approximately 60% of RA cases resided in metro (Calgary and Edmonton) and moderate metro areas. The highest rate was observed in rural areas (14.46; 95% CI 14.45-14.47; SRR 1.28), compared with the lowest in metro areas (10.69; 95% CI 10.68-10.69; SRR 0.82). The RA prevalence across local geographic areas ranged from 4.7 to 30.6 cases. The Global Moran's I index was 0.15 using three different matrices (z-score 3.96-4.24). We identified 10 hotspots in the south and north rural areas and 18 cold spots in metro and moderate metro Calgary.

CONCLUSION

The findings highlight notable rural-urban variation in RA prevalence in Alberta. Our findings can inform strategies aimed at reducing geographic disparities by targeting areas with high health care needs.

Geospatial patterns of comorbidity prevalence among people with osteoarthritis in Alberta Canada

BACKGROUND

Knowledge of geospatial pattern in comorbidities prevalence is critical to an understanding of the local health needs among people with osteoarthritis (OA). It provides valuable information for targeting optimal OA treatment and management at the local level. However, there is, at present, limited evidence about the geospatial pattern of comorbidity prevalence in Alberta, Canada.

METHODS

Five administrative health datasets were linked to identify OA cases and comorbidities using validated case definitions. We explored the geospatial pattern in comorbidity prevalence at two standard geographic areas levels defined by the Alberta Health Services: descriptive analysis at rural-urban continuum level; spatial analysis (global Moran's I, hot spot analysis, cluster and outlier analysis) at the local geographic area (LGA) level. We compared area-level indicators in comorbidities hotspots to those in the rest of Alberta (non-hotspots).

RESULTS

Among 359,638 OA cases in 2013, approximately 60% of people resided in Metro and Urban areas, compared to 2% in Rural Remote areas. All comorbidity groups exhibited statistically significant spatial autocorrelation (hypertension: Moran's I index 0.24, z score 4.61). Comorbidity hotspots, except depression, were located primarily in Rural and Rural Remote areas. Depression was more prevalent in Metro (Edmonton-Abbottsfield: 194 cases per 1000 population, 95%CI 192-195) and Urban LGAs (Lethbridge-North: 169, 95%CI 168-171) compared to Rural areas (Fox Creek: 65, 95%CI 63-68). Comorbidities hotspots included a higher percentage of First Nations or Inuit people. People with OA living in hotspots had lower socioeconomic status and less access to care compared to non-hotspots.

CONCLUSIONS

The findings highlight notable rural-urban disparities in comorbidities prevalence among people with OA in Alberta, Canada. Our study provides valuable evidence for policy and decision makers to design programs that ensure patients with OA receive optimal health management tailored to their local needs and a reduction in current OA health disparities.

Access to antenatal healthcare and the prevalence of oral clefts: a spatial analysis

Congenital anomalies are caused by a variety of risk factors, including individual and environmental ones. This study estimated the prevalence of oral clefts and examined their area-level risk indicators using data from The National Congenital Anomaly Survey in South Korea. Disease mapping and ecological regression were used to assess the geographical variation and potential risk indicators for oral clefts, such as living in a rural area, material deprivation, and limited access to antenatal healthcare. In South Korea, 1.15 (95% CI, 1.08-1.23) children out of every 1,000 births were born with oral clefts during 2005 and 2006. The most common oral cleft type was cleft palate, followed by cleft lip, then cleft lip and palate. Disease mapping showed that the proportion of areas with a higher risk than the national average was relatively higher for oral clefts than for all congenital abnormalities among rural and healthcare underserved areas. The relative risk of oral cleft prevalence was higher in rural areas and healthcare underserved areas than in urban areas and areas with better access to healthcare. The findings suggest that access to antenatal healthcare as well as area-level risk indicators should be considered a priority when developing measures to decrease the occurrence of oral clefts.

Contextual Factors and Spatial Patterns of Childhood Malnutrition in Provinces of Burkina Faso

BACKGROUND

Approximately 45% of all children's deaths are associated with malnutrition, and sub-Saharan Africa is hardest hit by this phenomenon. However, information on geographical variations of malnutrition in developing countries is limited. This study examined the geographical distribution and community characteristics associated with child malnutrition in Burkina Faso.

DESIGN

Data from the 2011 Burkina Faso Demographic Health Survey were analyzed. A general Kriging interpolation method was used to generate spatial malnutrition patterns. The global Moran's I test was used to identify significant malnutrition spatial patterns. Generalized estimating equations (GEEs) were fitted to examine the association between community level factors and malnutrition.

RESULTS

Average rates of stunting and wasting in the communities were 32.48% and 15.05%, respectively. Stunting hotspots were observed in the eastern and northeastern parts of Burkina Faso (i.e. Oudolan, Séno and Yagha, among others), while high rates of wasting were observed in the north-central part. The GEE results revealed lower stunting rates in communities with a higher percentage of households with improved sanitation. Communities with higher rates of professionally assisted births were associated with low wasting rates, while communities with higher rates of households with a low wealth index reported higher rates of wasting.

CONCLUSIONS

Spatial statistical models of malnutrition prevalence are useful for indicating hotspots over wide areas and hence, for guiding intervention strategies. This study revealed significant geographical patterns and community factors associated with childhood malnutrition. These factors should be considered in future programs aimed at reducing malnutrition in Burkina Faso.

Beyond mapping: a case for geospatial analytics in humanitarian health

The humanitarian sector is increasingly adopting geospatial data to support operations. However, the utilization of these data in the humanitarian health arena is predominantly in thematic map format, thereby limiting the full insight and utility of geospatial information.

Geospatial analytics, in contrast, including pattern analysis, interpolation, and predictive modeling, have tremendous potential within the field of humanitarian health. This paper explores a variety of historical and contemporary geospatial applications in the public health and humanitarian fields and argues for greater integration of geospatial analysis into humanitarian health research and programming. From remote sensing to create sampling frames, to spatial interpolation for environmental exposure analysis, and multi-objective optimization algorithms for humanitarian logistics, spatial analysis has transformed epistemological paradigms, research methods and programming landscapes across diverse disciplines.

The field of humanitarian health, which is inextricably bounded by geography and resource limitations, should leverage the unique capacities of spatial methods and strategically integrate geospatial analytics into research and programming not only to fortify the academic legitimacy and professionalization of the field but also to improve operational efficiency and mitigation strategies.

Opioid overdose deaths and potentially inappropriate opioid prescribing practices (PIP): A spatial epidemiological study

INTRODUCTION

Opioid overdose deaths quintupled in Massachusetts between 2000 and 2016. Potentially inappropriate opioid prescribing practices (PIP) are associated with increases in overdoses. The purpose of this study was to conduct spatial epidemiological analyses of novel comprehensively linked data to identify overdose and PIP hotspots.

METHODS

Sixteen administrative datasets, including prescription monitoring, medical claims, vital statistics, and medical examiner data, covering >98% of Massachusetts residents between 2011-2015, were linked in 2017 to better investigate the opioid epidemic. PIP was defined by six measures: ≥100 morphine milligram equivalents (MMEs), co-prescription of benzodiazepines and opioids, cash purchases of opioid prescriptions, opioid prescriptions without a recorded pain diagnosis, and opioid prescriptions through multiple prescribers or pharmacies. Using spatial autocorrelation and cluster analyses, overdose and PIP hotspots were identified among 538 ZIP codes.

RESULTS

More than half of the adult population (n = 3,143,817, ages 18 and older) were prescribed opioids. Nearly all ZIP codes showed increasing rates of overdose over time. Overdose clusters were identified in Worcester, Northampton, Lee/Tyringham, Wareham/Bourne, Lynn, and Revere/Chelsea (Getis-Ord Gi*; p < 0.05). Large PIP clusters for ≥100 MMEs and prescription without pain diagnosis were identified in Western Massachusetts; and smaller clusters for multiple prescribers in Nantucket, Berkshire, and Hampden Counties (p < 0.05). Co-prescriptions and cash payment clusters were localized and nearly identical (p < 0.05). Overlap in PIP and overdose clusters was identified in Cape Cod and Berkshire County. However, we also found contradictory patterns in overdose and PIP hotspots.

CONCLUSIONS

Overdose and PIP hotspots were identified, as well as regions where the two overlapped, and where they diverged. Results indicate that PIP clustering alone does not explain overdose clustering patterns. Our findings can inform public health policy decisions at the local level, which include a focus on PIP and misuse of heroin and fentanyl that aim to curb opioid overdoses.

Shiga toxigenic Escherichia coli incidence is related to small area variation in cattle density in a region in Ireland

Shiga toxigenic Escherichia coli (STEC) are pathogenic E. coli that cause infectious diarrhoea. In some cases infection may be complicated by renal failure and death. The incidence of human infection with STEC in Ireland is the highest in Europe. The objective of the study was to examine the spatial incidence of human STEC infection in a region of Ireland with significantly higher rates of STEC incidence than the national average and to identify possible risk factors of STEC incidence at area level. Anonymised laboratory records (n = 379) from 2009 to 2015 were obtained from laboratories serving three counties in the West of Ireland. Data included location and sample date. Population and electoral division (ED) data were obtained from the Irish 2011 Census of Population. STEC incidence was calculated for each ED (n = 498) and used to map hotspots/coldspots using the Getis-Ord Gi* spatial statistic and significant spatial clustering using the Anselin's Local Moran's I statistic. Multivariable regression analysis was used to consider the importance of a number of potential predictors of STEC incidence. Incidence rates for the seven-year period ranged from 0 to 10.9 cases per 1000. A number of areas with significant local clustering of STEC incidence as well as variation in the spatial distribution of the two main serogroups associated with disease in the region i.e. O26 and O157 were identified. Cattle density was found to be a statistically significant predictor of STEC in the region. GIS analysis of routine data indicates that cattle density is associated STEC infection in this high incidence region. This finding points to the importance of agricultural practices for human health and the importance of a "one-health" approach to public policy in relation to agriculture, health and environment.

Unlocking pandemic potential: prevalence and spatial patterns of key substitutions in avian influenza H5N1 in Egyptian isolates

Background

Avian influenza H5N1 has a high human case fatality rate, but is not yet well-adapted to human hosts. Amino acid substitutions currently circulating in avian populations may enhance viral fitness in, and thus viral adaptation to, human hosts. Substitutions which could increase the risk of a human pandemic (through changes to host specificity, virulence, replication ability, transmissibility, or drug susceptibility) are termed key substitutions (KS). Egypt represents the epicenter of human H5N1 infections, with more confirmed cases than any other country. To date, however, there have not been any spatial analyses of KS in Egypt.

Methods

Using 925 viral samples of H5N1 from Egypt, we aligned protein sequences and scanned for KS. We geocoded isolates using dasymetric mapping, then carried out geospatial hot spot analyses to identify spatial clusters of high KS detection rates. KS prevalence and spatial clusters were evaluated for all detected KS, as well as when stratified by phenotypic consequence.

Results

A total of 39 distinct KS were detected in the wild, including 17 not previously reported in Egypt. KS were detected in 874 samples (94.5%). Detection rates varied by viral protein with most KS observed in the surface hemagglutinin (HA) and neuraminidase (NA) proteins, as well as the interior non-structural 1 (NS1) protein. The most frequently detected KS were associated with increased viral binding to mammalian cells and virulence. Samples with high overall detection rates of KS exhibited statistically significant spatial clustering in two governorates in the northwestern Nile delta, Alexandria and Beheira.

Conclusions

KS provide a possible mechanism by which avian influenza H5N1 could evolve into a pandemic candidate. With numerous KS circulating in Egypt, and non-random spatial clustering of KS detection rates, these findings suggest the need for increased surveillance in these areas.

Mapping pharmacy deserts and determining accessibility to community pharmacy services for elderly enrolled in a State Pharmaceutical Assistance Program

Objectives

Limited studies have investigated geographic accessibility to a nearby community pharmacy for elderly which is an essential determinant of the access to medications and pharmacy services. This research identified pharmacy deserts and investigated availability of different types of community pharmacies and their services for elderly enrolled in a State Pharmaceutical Assistance Program (SPAP).

Methods

The state of Pennsylvania in the US was used as a case to demonstrate the geographic accessibility to community pharmacy and services for elderly enrolled in SPAP. The locations of community pharmacies and households of elderly enrolled in SPAP were derived from Pharmaceutical Assistance Contract for the Elderly programs’ database. The street addresses were geocoded and the distance to a nearby community pharmacy was calculated for study sample using the haversine formula. The demographic and geographic data were aggregated to Census Tracts and pharmacy deserts were identified using the predefined criteria. Descriptive statistical analysis was used to determine whether there are statistical differences in the socio-demographic profiles and distribution of different types of community pharmacies and their services in pharmacy deserts and non-deserts. This research used hot spot analyses at county level to identify clusters of pharmacy deserts, areas with high concentration of different racial/ethnic groups and clusters of high densities of chain and independent pharmacies.

Results

The Spatial analysis revealed that 39% and 61% Census Tracts in Pennsylvania were pharmacy deserts and non-deserts respectively (p < 0.001). Pharmacy deserts were found to have significantly more females, married and white elderly and fewer blacks and Hispanics compared to pharmacy non-deserts. Pharmacy deserts had significantly fewer chain and independent pharmacies and less delivery and 24-hour services in pharmacies than pharmacy non-deserts. Hot spot analyses showed that clusters of pharmacy deserts were more concentrated in southcentral, northwest and northeast regions of the state which represent rural areas and overlapped with clusters of high concentration of white individuals.

Conclusions

The findings suggest that urban-rural inequality, racial/ethnic disparity and differences in availability of pharmacies and their services exist between pharmacy deserts and non-deserts. The methodological approach and analyses used in this study can also be applied to other public health programs to evaluate the coverage and breadth of public health services.

HIV Clustering in Mississippi: Spatial Epidemiological Study to Inform Implementation Science in the Deep South

BACKGROUND

In recent years, more than half of new HIV infections in the United States occur among African Americans in the Southeastern United States. Spatial epidemiological analyses can inform public health responses in the Deep South by identifying HIV hotspots and community-level factors associated with clustering.

OBJECTIVE

The goal of this study was to identify and characterize HIV clusters in Mississippi through analysis of state-level HIV surveillance data.

METHODS

We used a combination of spatial epidemiology and statistical modeling to identify and characterize HIV hotspots in Mississippi census tracts (n=658) from 2008 to 2014. We conducted spatial analyses of all HIV infections, infections among men who have sex with men (MSM), and infections among African Americans. Multivariable logistic regression analyses identified community-level sociodemographic factors associated with HIV hotspots considering all cases.

RESULTS

There were HIV hotspots for the entire population, MSM, and African American MSM identified in the Mississippi Delta region, Southern Mississippi, and in greater Jackson, including surrounding rural counties (P<.05). In multivariable models for all HIV cases, HIV hotspots were significantly more likely to include urban census tracts (adjusted odds ratio [AOR] 2.01, 95% CI 1.20-3.37) and census tracts that had a higher proportion of African Americans (AOR 3.85, 95% CI 2.23-6.65). The HIV hotspots were less likely to include census tracts with residents who had less than a high school education (AOR 0.95, 95% CI 0.92-0.98), census tracts with residents belonging to two or more racial/ethnic groups (AOR 0.46, 95% CI 0.30-0.70), and census tracts that had a higher percentage of the population living below the poverty level (AOR 0.51, 95% CI 0.28-0.92).

CONCLUSIONS

We used spatial epidemiology and statistical modeling to identify and characterize HIV hotspots for the general population, MSM, and African Americans. HIV clusters concentrated in Jackson and the Mississippi Delta. African American race and urban location were positively associated with clusters, whereas having less than a high school education and having a higher percentage of the population living below the poverty level were negatively associated with clusters. Spatial epidemiological analyses can inform implementation science and public health response strategies, including improved HIV testing, targeted prevention and risk reduction education, and tailored preexposure prophylaxis to address HIV disparities in the South.

Improving Access to Medical Lactation Support and Counseling: Building the Case for Medicaid Reimbursement

Objective While the Affordable Care Act improves access to lactation services for many women across the US, low-income mothers in states without Medicaid expansion lack coverage for lactation support. As these states consider individual Medicaid reimbursement policies, the availability, effectiveness, and cost-benefit of lactation services must be evaluated. We conducted such an analysis for low-income mothers in North Carolina (NC), providing a model for other states. Methods First, we analyzed the distribution of NC International Board Certified Lactation Consultants (IBCLCs) and county-level breastfeeding rates among low-income infants. Logistic regression was used to examine the association between IBCLC density and 6-week breastfeeding duration. Finally, state advocates collaborated on a cost-benefit analysis of Medicaid coverage of IBCLCs. Results Maps of the NC breastfeeding support landscape indicate that IBCLCs are available to provide services to low-income women across the state. Compared to counties with no IBCLCs, those with high IBCLC density were found to have a 6-week breastfeeding prevalence ratio of 1.20 (95 % CI 1.12, 1.28). Medicaid reimbursement of IBCLCs showed an estimated annual cost savings of $2.33 million. Conclusions for Practice In one state without Medicaid expansion, we found that breastfeeding support resources are available across the state, high density IBCLC support is associated with increased breastfeeding by low-income mothers, and services are cost-effective. Our model for Medicaid reimbursement in NC provides a framework for states to improve equity in access to optimal lactation support.

Reassessing the Association between WIC and Birth Outcomes Using a Fetuses-at-Risk Approach

Objectives Women with longer, healthier pregnancies have more time to enroll in the Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), biasing associations between WIC and birth outcomes. We examined the association between WIC and preterm birth (PTB), low birth weight (LBW), and perinatal death (PND) using a fetuses-at-risk approach to address this bias, termed gestational age bias. Methods We linked California Medi-Cal recipients with a singleton live birth or fetal death from the 2010 Birth Cohort to WIC participant data (n = 236,564). We implemented a fetuses-at-risk approach using survival analysis, which compared, in each week of gestation, women whose pregnancies reached the same length and who had the same opportunity to utilize WIC. In each gestational week, we assessed WIC enrollment and the number of food packages redeemed thus far and computed hazard ratios (HR) using survival models with time-varying exposures and effects. Results Adjusting for maternal socio-demographic and health characteristics, WIC enrollment was associated with a lower risk of PTB from week 29-36 (HR₂₉ = 0.71; HR₃₆ = 0.52); LBW from week 26-40 (HR₂₆ = 0.77; HR₄₀ = 0.64); and PND from week 29-43 (HR₂₉ = 0.78; HR₄₃ = 0.69) (p < 0.05). The number of food packages redeemed was associated with a lower risk of PTB from week 27-36 (HR₂₇ = 0.90; HR₃₆ = 0.84); LBW from week 25-42 (HR₂₅ = 0.93; HR₄₂ = 0.88); and PND from week 27-46 (HR₂₇ = 0.94; HR₄₆ = 0.91) (p < 0.05). Conclusions for Practice To our knowledge this is the first study to examine the association between WIC and birth outcomes using this approach. We found that beginning from about 29 weeks, WIC enrollment was associated with a reduced risk of PTB by 29-48 %, LBW by 23-36 %, and PND by 22-31 %.

Identifying and characterizing hepatitis C virus hotspots in Massachusetts: a spatial epidemiological approach

Background

Hepatitis C virus (HCV) infections have increased during the past decade but little is known about geographic clustering patterns.

Methods

We used a unique analytical approach, combining geographic information systems (GIS), spatial epidemiology, and statistical modeling to identify and characterize HCV hotspots, statistically significant clusters of census tracts with elevated HCV counts and rates. We compiled sociodemographic and HCV surveillance data (n = 99,780 cases) for Massachusetts census tracts (n = 1464) from 2002 to 2013. We used a five-step spatial epidemiological approach, calculating incremental spatial autocorrelations and Getis-Ord Gi* statistics to identify clusters. We conducted logistic regression analyses to determine factors associated with the HCV hotspots.

Results

We identified nine HCV clusters, with the largest in Boston, New Bedford/Fall River, Worcester, and Springfield (p < 0.05). In multivariable analyses, we found that HCV hotspots were independently and positively associated with the percent of the population that was Hispanic (adjusted odds ratio [AOR]: 1.07; 95% confidence interval [CI]: 1.04, 1.09) and the percent of households receiving food stamps (AOR: 1.83; 95% CI: 1.22, 2.74). HCV hotspots were independently and negatively associated with the percent of the population that were high school graduates or higher (AOR: 0.91; 95% CI: 0.89, 0.93) and the percent of the population in the “other” race/ethnicity category (AOR: 0.88; 95% CI: 0.85, 0.91).

Conclusion

We identified locations where HCV clusters were a concern, and where enhanced HCV prevention, treatment, and care can help combat the HCV epidemic in Massachusetts. GIS, spatial epidemiological and statistical analyses provided a rigorous approach to identify hotspot clusters of disease, which can inform public health policy and intervention targeting. Further studies that incorporate spatiotemporal cluster analyses, Bayesian spatial and geostatistical models, spatially weighted regression analyses, and assessment of associations between HCV clustering and the built environment are needed to expand upon our combined spatial epidemiological and statistical methods.

Correlates of county-level nonviral sexually transmitted infection hot spots in the US: application of hot spot analysis and spatial logistic regression

PURPOSE

We used a combination of hot spot analysis (HSA) and spatial regression to examine county-level hot spot correlates for the most commonly reported nonviral sexually transmitted infections (STIs) in the 48 contiguous states in the United States (US).

METHODS

We obtained reported county-level total case rates of chlamydia, gonorrhea, and primary and secondary (P&S) syphilis in all counties in the 48 contiguous states from national surveillance data and computed temporally smoothed rates using 2008-2012 data. Covariates were obtained from county-level multiyear (2008-2012) American Community Surveys from the US census. We conducted HSA to identify hot spot counties for all three STIs. We then applied spatial logistic regression with the spatial error model to determine the association between the identified hot spots and the covariates.

RESULTS

HSA indicated that ≥84% of hot spots for each STI were in the South. Spatial regression results indicated that, a 10-unit increase in the percentage of Black non-Hispanics was associated with ≈42% (P < 0.01) [≈22% (P < 0.01), for Hispanics] increase in the odds of being a hot spot county for chlamydia and gonorrhea, and ≈27% (P < 0.01) [≈11% (P < 0.01) for Hispanics] for P&S syphilis. Compared with the other regions (West, Midwest, and Northeast), counties in the South were 6.5 (P < 0.01; chlamydia), 9.6 (P < 0.01; gonorrhea), and 4.7 (P < 0.01; P&S syphilis) times more likely to be hot spots.

CONCLUSION

Our study provides important information on hot spot clusters of nonviral STIs in the entire United States, including associations between hot spot counties and sociodemographic factors.

Nonprescription naloxone and syringe sales in the midst of opioid overdose and hepatitis C virus epidemics: Massachusetts, 2015

OBJECTIVES

To determine the prevalence of nonprescription naloxone and sterile syringe sales, factors associated with nonprescription sales, geospatial access to nonprescription naloxone and syringe-selling pharmacies, and targets for potential interventions.

DESIGN

Cross-sectional study.

SETTING AND PARTICIPANTS

Massachusetts has experienced steep increases in reported opioid overdoses and hepatitis C virus cases in the past decade. Pharmacists have the potential to play a substantial role in increasing access to nonprescription naloxone and sterile syringes, which can reverse opioid overdoses and decrease hepatitis C virus transmission, respectively. We completed brief telephone surveys with 809 of 1042 retail pharmacies across Massachusetts (response rate = 77.6%) during 2015 to assess experience with nonprescription sales of naloxone and sterile syringes.

OUTCOME MEASURES

Our primary outcomes were the stocking and selling of naloxone in the pharmacy (yes or no) for nonprescription sales and nonprescription syringe sales (yes or no). We conducted multivariable regression analyses and created maps using a geographic information system to identify factors associated with nonprescription sales of naloxone and sterile syringes, and to improve our understanding of geospatial access to pharmacy-based naloxone and syringe sales.

RESULTS

More than 97% of pharmacies reported selling sterile syringes without requiring a prescription, and 45% of pharmacies reported stocking and selling naloxone. Factors associated with nonprescription sales included hours of operation, experience with and interest in harm reduction activities, and presence in an opioid overdose hotspot. Geographic access to nonprescription sale of sterile syringes is widespread, whereas geospatial access to naloxone is limited. Training to understand the benefits, applications, and distribution needs of naloxone is of interest to surveyed pharmacists.

CONCLUSION

Access to sterile syringes through nonprescription sales is strong across Massachusetts, and although more than 350 pharmacies (45%) reported stocking and selling naloxone to prevent opioid overdose deaths, there is much room for improvement in access and training among pharmacy staff members.

Small-area Variation in Hypertension Prevalence among Black and White Medicaid Enrollees

OBJECTIVES

To examine within-state geographic heterogeneity in hypertension prevalence and evaluate associations between hypertension prevalence and small-area contextual characteristics for Black and White South Carolina Medicaid enrollees in urban vs rural areas.

DESIGN

Ecological.

SETTING

South Carolina, United States.

MAIN OUTCOME MEASURES

Hypertension prevalence.

METHODS

Data representing adult South Carolina Medicaid recipients enrolled in fiscal year 2013 (N=409,907) and ZIP Code Tabulation Area (ZCTA)-level contextual measures (racial segregation, rurality, poverty, educational attainment, unemployment and primary care physician adequacy) were linked in a spatially referenced database. Optimized Getis-Ord hotspot mapping was used to visualize geographic clustering of hypertension prevalence. Spatial regression was performed to examine the association between hypertension prevalence and small-area contextual indicators.

RESULTS

Significant (alpha=.05) hotspot spatial clustering patterns were similar for Blacks and Whites. Black isolation was significantly associated with hypertension among Blacks and Whites in both urban (Black, b=1.34, P<.01; White, b=.66, P<.01) and rural settings (Black, b=.71, P=.02; White, b=.70, P<.01). Primary care physician adequacy was associated with hypertension among urban Blacks (b=-2.14, P<.01) and Whites (b=-1.74, P<.01).

CONCLUSIONS

The significant geographic overlap of hypertension prevalence hotspots for Black and White Medicaid enrollees provides an opportunity for targeted health intervention. Provider adequacy findings suggest the value of ACA network adequacy standards for Medicaid managed care plans in ensuring health care accessibility for persons with hypertension and related chronic conditions.

Identifying areas at risk of low birth weight using spatial epidemiology: A small area surveillance study

OBJECTIVES

To assess the geographic distribution of Low Birth Weight (LBW) in New York State among singleton births using a spatial regression approach in order to identify priority areas for public health actions.

METHODS

LBW was defined as birth weight less than 2500g. Geocoded data from 562,586 birth certificates in New York State (years 2008-2012) were merged with 2010 census data at the tract level. To provide stable estimates and maintain confidentiality, data were aggregated to yield 1268 areas of analysis. LBW prevalence among singleton births was related with area-level behavioral, socioeconomic and demographic characteristics using a Poisson mixed effects spatial error regression model.

RESULTS

Observed low birth weight showed statistically significant auto-correlation in our study area (Moran's I 0.16 p value 0.0005). After over-dispersion correction and accounting for fixed effects for selected social determinants, spatial autocorrelation was fully accounted for (Moran's I-0.007 p value 0.241). The proportion of LBW was higher in areas with larger Hispanic or Black populations and high smoking prevalence. Smoothed maps with predicted prevalence were developed to identify areas at high risk of LBW. Spatial patterns of residual variation were analyzed to identify unique risk factors.

CONCLUSION

Neighborhood racial composition contributes to disparities in LBW prevalence beyond differences in behavioral and socioeconomic factors. Small-area analyses of LBW can identify areas for targeted interventions and display unique local patterns that should be accounted for in prevention strategies.

Spatial analysis of myocardial infarction in Iran: National report from the Iranian myocardial infarction registry

Background:

Myocardial infarction (MI) is a leading cause of mortality and morbidity in Iran. No spatial analysis of MI has been conducted to date. The present study was conducted to determine the pattern of MI incidence and to identify the associated factors in Iran by province.

Materials and Methods:

This study has two parts. One part is prospective and hospital-based, and the other part is an ecological study. In this study, the data of 20,750 new MI cases registered in Iranian Myocardial Infarction Registry in 2012 were used. For spatial analysis in global and local, spatial autocorrelation, Moran's I, Getis-Ord, and logistic regression models were used. Data were analyzed by Stata software and ArcGIS 9.3.

Results:

Based on autocorrelation coefficient, a specific pattern was observed in the distribution of MI incidence in different provinces (Moran's I: 0.75, P < 0.001). Spatial pattern of incidence was approximately the same in men and women. MI incidence was clustering in six provinces (North Khorasan, Yazd, Kerman, Semnan, Golestan, and Mazandaran). Out of the associated factors with clustered MI in six provinces, temperature, humidity, hypertension, smoking, and body mass index (BMI) could be mentioned. Hypertension, smoking, and BMI contributed to clustering with, respectively, 2.36, 1.31, and 1.31 odds ratio.

Conclusion:

Addressing the place-based pattern of incidence and clarifying their epidemiologic dimension, including spatial analysis, has not yet been implemented in Iran. Report on MI incidence rate by place and formal borders is useful and is used in the planning and prioritization in different levels of health system.

HIV and hepatitis C mortality in Massachusetts, 2002-2011: spatial cluster and trend analysis of HIV and HCV using multiple cause of death

Background

Infectious diseases, while associated with a much smaller proportion of deaths than they were 50 years ago, still play a significant role in mortality across the state of Massachusetts. Most analysis of infectious disease mortality in the state only take into account the underlying cause of death, rather than contributing causes of death, which may not capture the full extent of mortality trends for infectious diseases such as HIV and the Hepatitis C virus (HCV).

Methods

In this study we sought to evaluate current trends in infectious disease mortality across the state using a multiple cause of death methodology. We performed a mortality trend analysis, identified spatial clusters of disease using a 5-step geoprocessing approach and examined spatial-temporal clustering trends in infectious disease mortality in Massachusetts from 2002–2011, with a focus on HIV/AIDS and HCV.

Results

Significant clusters of high infectious disease mortality in space and time throughout the state were detected through both spatial and space time cluster analysis. The most significant clusters occurred in Springfield, Worcester, South Boston, the Merrimack Valley, and New Bedford with other smaller clusters detected across the state. Multiple cause of death mortality rates were much higher than underlying cause mortality alone, and significant disparities existed across race and age groups.

Conclusions

We found that our multi-method analyses, which focused on contributing causes of death, were more robust than analyses that focused on underlying cause of death alone. Our results may be used to inform public health resource allocation for infectious disease prevention and treatment programs, provide novel insight into the current state of infectious disease mortality throughout the state, and benefited from approaches that may more accurately document mortality trends.