Obesity clusters within the Empire Health Foundation Region Joe Campo, MPH Healthcare Research Group Office of Financial Management Highlights of November 17th, 2011, presentation Identifying clusters Obesity is on the rise nationally and locally – with broad consequence affecting quality of life and longevity as well as cost to the health care system as a whole. High & low obesity clusters brfss 2003-2010 Using cluster identification software from the National Cancer Institute, together with Behavioral Risk Factor Surveillance System (BRFSS) data for 2003 through 2010, three regions were identified within the Empire Health Foundation (EHF) Region as having higher than expected obesity rates. North Spokane City Lincoln/ Stevens South Spokane County These high obesity clusters include most of Adams County, portions of Lincoln and Stevens Counties, and the northern portion of Spokane City. Adams An area with lower than expected obesity rates was also identified; this area comprises most of the southern portion of Spokane County. Trends within each of these cluster areas, as well as for the EHF Region as a whole, were also assessed. For the Adams and Lincoln/Stevens clusters, no significant trends were identified, although the lines that best fit each of those areas’ obesity rates are moving upwards. For North Spokane City, there had been a significant upward trend (7% per year) from 2003 to 2008; however, in 2009 and 2010 the rates in that area are markedly lower. Surprisingly, within the low obesity cluster, South Spokane County, the rates are increasing over time at 4% per year. So, too, within the EHF Region as a whole the obesity rates are also increasing at 3% per year. Behavioral Risk Factor Surveillance System (2003-2010) Behavioral Risk Factor Surveillance System (2003-2010) percent obese within high & low clusters trends – percent obese from 2003-2010 40 EHF’s % obese w/ 95% ci 35 30 25 South Spokane County North Spokane City Lincoln/ Stevens Adams 45 45 45 40 40 40 35 35 45 40 ns 20 30 EHF trend + 2.6% per year 30 + 7.0% ns 35 35 30 30 25 ns 15 25 25 25 10 20 20 20 5 15 20 + 4.4% 15 2003 2004 2005 2006 2007 2008 2009 2010 15 2003 2004 2005 2006 2007 2008 2009 2010 15 2003 2004 2005 2006 2007 2008 2009 2010 2003 2004 2005 2006 2007 2008 2009 2010 0 So. Spokane County No. Spokane City Obesity clusters within the Empire Health Foundation Region Adams Lincoln/Stevens November 2011 Obesity clusters within the Empire Health Foundation Region Obesity clusters within the Empire Health Foundation Region November 2011 November 2011 Hospitalizations – Rates Age-adjusted inpatient day hospitalization rates, for 2005 to 2009 combined, were analyzed for conditions known to be associated with obesity. These rates were based upon the patients’ reported ZIP code of residence, not where the hospital was located. Rates for diabetes as the primary or secondary diagnosis within each of the high obesity clusters were higher than the EHF Region’s rate; in the low obesity cluster, they were lower. For coronary artery disease (CAD), rates within the high obesity Adams and Lincoln/Stevens clusters were higher than the EHF’s Region's rate; in the low obesity South Spokane County cluster they were lower. Somewhat surprisingly, the rates in the high obesity North Spokane City cluster were borderline low compared to the EHF Region. However, the rates in each of the high obesity clusters were higher than the rate in the low obesity one. Rates for stroke were markedly high in Lincoln/Stevens clusters. Similarly, rates for hypertension were markedly high in the Adams cluster. Washington & Oregon inpatient discharge data (2005-2009) Washington & Oregon inpatient discharge data (2005-2009) diabetes (any diagnoses) coronary artery disease 6000 1800 5500 1600 5000 1400 4500 4000 1200 3500 1000 3000 800 2500 2000 600 So. Spokane County No. Spokane City Adams Lincoln/Stevens So. Spokane County No. Spokane City age-adjusted patient day rates per 100,000 persons Obesity clusters within the Empire Health Foundation Region November 2011 Adams Obesity clusters within the Empire Health Foundation Region November 2011 Washington & Oregon inpatient discharge data (2005-2009) Washington & Oregon inpatient discharge data (2005-2009) stroke hypertension 1800 Lincoln/Stevens age-adjusted patient day rates per 100,000 persons 100 90 1600 80 1400 70 60 1200 50 1000 40 30 800 20 600 10 400 0 So. Spokane County No. Spokane City Adams Lincoln/Stevens So. Spokane County No. Spokane City age-adjusted patient day rates per 100,000 persons Obesity clusters within the Empire Health Foundation Region November 2011 Obesity clusters within the Empire Health Foundation Region Adams Lincoln/Stevens age-adjusted patient day rates per 100,000 persons Obesity clusters within the Empire Health Foundation Region November 2011 November 2011 Hospitalizations – Trends Trends for these hospitalization rates were also assessed. For diabetes, however, no significant trends were identified for any of the clusters or the EHF region as a whole; the rates are essentially remaining flat over time. Conversely, for CAD, trends were seen in each cluster and for the region as a whole. In all instances those rates were decreasing, although at varying rates. Within the EHF Region, CAD rate are decreasing at close to -6% per year. In the South Spokane County, they are decreasing at a lower rate, approximately -4% per year. The rates are decreasing at -3% per year within the North Spokane City cluster, -6% per year in Adams, and about -5% per year in Lincoln/Stevens. For stroke, the rates within the EHF Region are decreasing at -6% per year. They are decreasing in Lincoln/Stevens by -21% per year. The rates in North Spokane City , however, have been increasing by 12% per year starting in 2005. Hypertension is increasing in South Spokane County – at 10% per year. No trends are seen elsewhere. Washington & Oregon inpatient discharge data (2000-2009) Washington & Oregon inpatient discharge data (2000-2009) trends – diabetes in primary or secondary trends – coronary artery disease South Spokane County South Spokane County North Spokane City 10000 10000 10000 8000 8000 Lincoln/ Stevens Adams 8000 8000 6000 6000 6000 4000 4000 4000 4000 2000 2000 2000 2000 0 0 0 0 Obesity clusters within the Empire Health Foundation Region 3500 3500 3000 3000 3000 2500 2500 2500 2500 2000 2000 2000 1500 1500 1500 1500 1000 1000 1000 1000 500 500 500 0 0 0 2000 500 - 5.6% - 4.4% 0 November 2011 3500 - 3.2% - 6.2% trends – hypertension Lincoln/ Stevens Adams 4000 4000 4000 3500 3500 3500 3500 3000 3000 2500 3000 2500 2000 2000 - 6.2% 2500 +12.1% (2005-09) 1500 1000 North Spokane City Lincoln/ Stevens Adams 100 100 100 100 90 90 90 90 80 80 80 80 70 70 70 70 60 60 60 50 50 50 40 40 40 40 30 30 30 30 20 20 20 20 3000 2500 2000 2000 1500 1500 1000 South Spokane County 60 50 1000 9.6% 500 500 500 500 10 10 10 10 0 0 0 0 0 0 0 0 Obesity clusters within the Empire Health Foundation Region - 4.6% November 2011 Washington & Oregon inpatient discharge data (2000-2009) North Spokane City 3000 Obesity clusters within the Empire Health Foundation Region trends – stroke 4000 1000 3500 Washington & Oregon inpatient discharge data (2000-2009) South Spokane County Lincoln/ Stevens Adams 10000 6000 1500 North Spokane City November 2011 Obesity clusters within the Empire Health Foundation Region Obesity clusters within the Empire Health Foundation Region November 2011 November 2011 Hospitalizations – Costs Excess costs or savings were also estimated for hospitalizations associated with obesity. These were derived by converting all reported hospitalization charges to “costs” using the Agency for Healthcare Research and Quality’s (AHRQ) cost-to-charge ratios, estimating the number of patient days that would be generated in each cluster if the populations living those clusters had been hospitalized at the same rate they were within the EHF Region, comparing those “expected” patient days to the actual days generated, multiplying any differences in days by the EHF Region’s cost per day, dividing that product by the population in each cluster, and then multiplying the results by 1,000 to generate an estimated excess cost or saving per 1,000 person within each cluster. For instance, because the hospitalization rate for diabetes in the South Spokane County cluster is lower than the rate in the EHF Region, the saving there equals about $35,000 per 1,000 persons per year compared to what it would have cost had they been hospitalized at the same rate as seen for the region as a whole. In Lincoln/Stevens the excess hospitalization rates equal an excess cost of $40,000 per 1,000 persons per year for diabetes. Washington & Oregon inpatient discharge data (2006-2010) Washington & Oregon inpatient discharge data (2006-2010) costs - diabetes (any diagnoses) Savings costs - coronary artery disease Savings Excess costs So. Spokane County So. Spokane County No. Spokane City No. Spokane City Adams Adams Lincoln/Stevens Lincoln/Stevens -$40,000 -$20,000 $0 $20,000 $40,000 -$25,000 -$15,000 -$5,000 per 1,000 persons November 2011 Washington & Oregon inpatient discharge data (2006-2010) $15,000 $25,000 Obesity clusters within the Empire Health Foundation Region November 2011 Washington & Oregon inpatient discharge data (2006-2010) costs - stroke costs - hypertension Savings Excess costs So. Spokane County So. Spokane County No. Spokane City No. Spokane City Adams Adams Lincoln/Stevens Lincoln/Stevens -$5,000 $5,000 per 1,000 persons Obesity clusters within the Empire Health Foundation Region Savings Excess costs $0 $5,000 $10,000 $15,000 $20,000 per 1,000 persons Obesity clusters within the Empire Health Foundation Region -$200 Excess costs $0 $200 $400 $600 $800 $1,000 per 1,000 persons November 2011 Obesity clusters within the Empire Health Foundation Region Obesity clusters within the Empire Health Foundation Region November 2011 November 2011 Characteristics of the population Various behavioral and demographic characteristics of each cluster area were also assessed. Included, below, are measures pertaining to exercise, diet, insurance coverage, education, income and race. Behavioral Risk Factor Surveillance System (2003-2010) Behavioral Risk Factor Surveillance System (2005, 2007 & 2009) no exercise outside of work less than 5 fruits or vegetables per day 40 84 35 82 80 30 78 25 76 20 74 15 72 10 70 5 68 0 66 So. Spokane County No. Spokane City Adams Obesity clusters within the Empire Health Foundation Region Lincoln/Stevens November 2011 So. Spokane County No. Spokane City Adams Lincoln/Stevens Obesity clusters within the Empire Health Foundation Region November 2011 Behavioral Risk Factor Surveillance System (2003-2010) Behavioral Risk Factor Surveillance System (2003-2010) no health care coverage education: college or tech school grad 40 60 35 50 30 40 25 20 30 15 20 10 10 5 0 0 So. Spokane County No. Spokane City Adams Obesity clusters within the Empire Health Foundation Region Lincoln/Stevens November 2011 So. Spokane County No. Spokane City Adams Lincoln/Stevens Obesity clusters within the Empire Health Foundation Region November 2011 Behavioral Risk Factor Surveillance System (2003-2010) Behavioral Risk Factor Surveillance System (2003-2010) income above 200% federal poverty level american indian/alaska native 80 6 70 5 60 4 50 40 3 30 2 20 1 10 0 0 So. Spokane County No. Spokane City Obesity clusters within the Empire Health Foundation Region Adams Lincoln/Stevens November 2011 Obesity clusters within the Empire Health Foundation Region So. Spokane County No. Spokane City Obesity clusters within the Empire Health Foundation Region Adams Lincoln/Stevens November 2011 November 2011 Characteristics of the environment A community’s access to healthy food is one factor to consider in assessing the role environment plays in promoting or preventing obesity. The map below is based upon data provided by the CDC showing the percent of health food retailer in a tract relative to all food retailers in that tract including quickie-marts and fast food vendors. Overlaid on the map are selected cities within the EHF Region. Also overlaid on the map are the high obesity clusters identified through BRFSS. While no rigorous assessment has been done on the correlation between the obesity clusters and the regions with poor access to healthy food retailers, a scan of the map suggests that both Springdale and Ritzville differ from the other cities shown in that they both are within regions with relatively poor access to healthy food retailers. healthy food stores as a percent of all food stores Colville by census tract less than 4% Springdale 4% to 12% 12% to 20% 20% to 33% Davenport Deer Park Spokane City Cheney 33% or more Washington average = 12% National average = 10% Ritzville Othello Obesity clusters within the Empire Health Foundation Region November 2011 What the future may hold Factors that lead to obesity may be traced back to early youth – or younger. In that context, we examined hospitals birth data, as well as the school-based Healthy Youth Survey data. For hospital-based births, a concerning harbinger was seen in assessing macrosomia, or large-baby births; these newborns may be at greater risk for adult obesity than normal weight ones. Surprisingly, the low obesity South Spokane County had the highest macrosomia birth rate. Washington & Oregon inpatient discharge data (2006-2010) macrosomia (large baby syndrome) 90 80 70 60 50 Healthy Youth Survey data were also assessed, with clusters being identified for selected risk-factors and overlaid on the obesity cluster areas. These overlays suggest a degree of concordance between risk factors and the obesity clusters identified. 40 30 20 10 0 So. Spokane County No. Spokane City Adams Lincoln/Stevens Obesity clusters within the Empire Health Foundation Region November 2011 Healthy Youth Survey (2006, 2008 & 2010) Healthy Youth Survey (2006, 2008 & 2010) obese drinks two or more sodas per day 20 20 Obese 18 18 16 16 14 14 12 10 8 6 Drinks two or more sodas per day 12 Adams/ Lincoln School Districts 4 10 8 Spokane School Districts 6 EHF Region 2 0 0 November 2011 EHF Region 4 2 Obesity clusters within the Empire Health Foundation Region Adams/ Lincoln School Districts Cluster Obesity clusters within the Empire Health Foundation Region November 2011 Healthy Youth Survey (2006, 2008 & 2010) Healthy Youth Survey (2006, 2008 & 2010) 3+ hours of TV viewing per day physically active less than 5 days/week 40 54 Three or more hours of TV viewing per day 35 Less than five days per week of being physically active for one hour or more 52 30 50 25 48 20 15 Spokane School Districts Adams School Districts 46 EHF Region 44 10 Obesity clusters within the Empire Health Foundation Region Spokane School Districts 5 42 0 40 November 2011 Obesity clusters within the Empire Health Foundation Region Obesity clusters within the Empire Health Foundation Region EHF Region November 2011 November 2011 Additional findings Behavioral Risk Factor Surveillance System (2003-2010) Behavioral Risk Factor Surveillance System (2003-2010) background – large area comparisons background – county comparisons 45 45 40 40 normal overweight obese normal 35 normal overweight normal overweight normal overweight overweight 35 30 30 obese 25 obese obese obese 25 20 20 15 15 10 10 5 5 0 0 Nation (2010) Washington State Eastern Washington Obesity clusters within the Empire Health Foundation Region Empire Health Foundation Region November 2011 Adams Ferry Lincoln Pend Oreille Spokane Stevens Obesity clusters within the Empire Health Foundation Region November 2011 Washington & Oregon inpatient discharge data (2005-2009) Washington & Oregon inpatient discharge data (2005-2009) low-extremity amputations w/ diabetes diabetes long-term complications 250 Whitman 600 500 200 400 150 300 100 200 50 100 0 0 So. Spokane County No. Spokane City Adams Lincoln/Stevens So. Spokane County No. Spokane City age-adjusted patient day rates per 100,000 persons Obesity clusters within the Empire Health Foundation Region November 2011 Adams Lincoln/Stevens age-adjusted patient day rates per 100,000 persons Obesity clusters within the Empire Health Foundation Region November 2011 Behavioral Risk Factor Surveillance System (2009-2010) Behavioral Risk Factor Surveillance System (2009-2010) high adverse childhood experiences self-assessed health status: fair or poor 60 30 50 25 40 20 30 15 20 10 10 5 0 0 So. Spokane County No. Spokane City Obesity clusters within the Empire Health Foundation Region Adams Lincoln/Stevens November 2011 Obesity clusters within the Empire Health Foundation Region So. Spokane County No. Spokane City Obesity clusters within the Empire Health Foundation Region Adams Lincoln/Stevens November 2011 November 2011 Summaries Behavioral Risk Factor Surveillance System (2003-2010) summary of obesity rates There are pronounced geographic variations in obesity rates Compared to the nation, the EHF obesity rate is lower; compared to the state, it is higher Within EHF Region, Adams County has the highest obesity rate; Spokane and Whitman have the lowest Only Spokane &Whitman show a trend in their obesity rates: upwards Small area clusters may differ, but there is a general concordance For 2003-2010, three high and one low obesity clusters are seen Only the low obesity South Spokane area has a trend: upwards There is also a increasing trend seen in the EHF Region as a whole Obesity clusters within the Empire Health Foundation Region November 2011 Washington & Oregon inpatient discharge data (2006-2010) summary of hospitalization rates High diabetes rates for high obesity clusters, low for the low one Amputations: high in high obesity cluster, low in low Long-term comp.: high in two high clusters, low in low Short-term comp: high in No. Spokane, low in So. Spokane Uncontrolled : high in two high clusters, low in low & No. Spokane CAD – high in Adams & Lincoln/Stevens, low in S. Spokane & N. Spokane Stroke – high in No. Spokane & Lincoln/Stevens, low in So. Spokane & Washington & Oregon inpatient discharge data (2006-2010) Adams summary of hospitalization rates Hypertension – high in Adams No trends seen with diabetes hospitalization rates CAD – trending downwards with regions and each clusters area (smoking?) November 2011 Stroke – trending downwards in EHF region and Lincoln/Stevens; trending upwards in No. Spokane City For diabetes in general, nearly $40,000 in excess costs per 1,000 persons in Lincoln/Stevens; the opposite in So. Spokane County For CAD, nearly $20,000 in excess costs per 1,000 persons in Lincoln/Stevens; the opposite in So. Spokane County Obesity clusters within the Empire Health Foundation Region Obesity clusters within the Empire Health Foundation Region November 2011 Discussion and sources Obesity clusters within the Empire Health Foundation Region discussion Wide variations in obesity rates across the region High (and low) obesity clusters are identifiable The health care needs and costs associated with obesity disproportionately affect the high obesity regions The trends, in general, do not portend well Identifying and assessing geographic variations can, however, help to engage communities in addressing this condition Obesity clusters within the Empire Health Foundation Region November 2011 background – sources and methods data sources • brfss • inpatient hospitalizations • chars & oregon tools • head-banging weighted two-dimensional median-based smoothing algorithm • births and deaths • zip-code population ute 16 o t ate R S 16 ute16 o t ateR S oute t ateR S • cost-to-charge ratios • modified retail food ...environment index • spatial scan statistic • joinpoint regression • healthy youth survey Obesity clusters within the Empire Health Foundation Region Obesity clusters within the Empire Health Foundation Region analyzes varying models to test if a change in trend is statistically significant. November 2011 November 2011 Pend Oreille Ferry Stevens Lincoln Spokane Adams Whitman Obesity clusters within the Empire Health Foundation Region November 2011