UHR Volume 14 (2009)_OCR

2009 UTAH'S HEALTH: AN ANNUAL REViEW Introduction and Editor's Note On behalf of this year's Editorial Board, I am pleased to present to you the fourteenth volume of Utah's Health: An Annual Review. Utah's Health is a peer reviewed journal and statistical update focusing on the issues timely to the health of Utah's population. Our objective is to provide readers with current and pertinent information regarding health and health care in Utah as compared to the nation, as well as to generate interest in and to facilitate discussion of health-related topics. Utah's Health: An Annual Review is also available online, at our new website: www.matheson.utah.edu/UHReview. As with previous editions, Volume XIV includes Original Research Articles and a Data Review section. The Original Research Articles cover topics related to public health, health policy, and clinical care provision. The articles represent the work of univer-sity faculty, students, physicians, statisticians, and other professionals from the community. The Data Review section follows the journal's tradition of providing a variety of current health statistical trends in Utah and the United States. Volume XIV of Utah's Health also includes a Legislative Review section and Health Directory. The Legislative Review covers all health-related bills addressed in the 2009 General Legislative Session, as well as some major budget changes observed in the Department of Health and Department of Human Services. The Health Directory is composed of a comprehensive list of health-related agencies in Utah, such as hospitals/medical centers, government resources, and research/education facilities. The successful publication of Utah's Health: An Annual Review involved the hard work and dedication of everyone involved. I thank all of the authors and contributors of the research articles. Utah's Health would not exist without your commitment to research focusing on health-related issues in Utah. I also extend a sincere thank you to the Advisory Board for their willingness to work along side us to ensure a quality publication. I thank Dr. Richard Sperry, our faculty advisor, for his support and guidance throughout the publication process. I extend a special thank you to the Editorial Board members for their diligent work on the Data Review section, and to our readers for their continued support. However, the greatest thanks of all goes to the Executive Editors: Cade Walker, Jeff Hebert, Sarah Watts, Donya Mohebali, Megan Bitner, Scott Coleman, and Cole Helm. These talented individuals went above and beyond what was expected of them to produce an exceptional publication I am extremely proud of. Safia Ahmed Editor-in-Chief Volume XIV, 2009 ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH'S HEALTH: AN ANNUAL REViEW Authors and Contributors ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw ©2009 The University o f Utah. All Rights Reserved. Table o f Contents Original Research Articles..................................................................................... 7 Arthritis and Co-Morbid Conditions among Adults in Utah.................................................................................................... 9 Nicole Bissonette, MPH, CHES; Michael Friedrichs, MS; Randy Tanner, MPA; Shelly Wagstaff, BS An Assessment of Community Access to Physical Activity Resources and Healthy Food Choices................................. 16 Phyllis Crowley, MS, RD, IBCLC; Shaheen Hossain, PhD; Angeni Marque, BS; Robert Satterfield, M.Stat; Lynda Blades, MPH, CHES; Heather Borski, MPH, CHES; Christopher D. Furner, MS, CHES; Patrice Isabella, MS,RD; Colleen Jenson, BS, LE; Tim Stempel, MSW; Nan Streeter, MS,RN; Rick Wardle, BS Patterns of Participant Satisfaction with Utah WIC Program................................................................................................21 Shaheen Hossain, PhD; Rob Satterfield, Mstat; Angeni Marque, BS; Phyllis Crowley, MS, RD, IBCLC; Nan Streeter, MS, RN; Chris D. Furner, MS, CHES Location of Adverse Event Documentation in Hospital Inpatient Medical Charts............................................................29 Carol J. Masheter, PhD; Paul Hougland, MD; Wu Xu, PhD Hormonal Contraception Use and Continuation by Formulation and Hormone Content in the Utah Medicaid Population.........................................................................................................................................................................37 Carrie McAdam-Marx RPh, MS; Diana Brixner, PhD, RPh; Patricia Murphy, CNM, DrPh Associations between Educational Attainment and Diabetes in Utah: The Behavioral Risk Factor Surveillance System, 1996-2007..................................................................................................................................................... 42 Eric N. Reither, PhD; Theresa M. Fedor, BS; Karin M. Abel, BS; Dan J. Hatch, BA The Evaluation of a Vaginal Self-Test Device for the Collection of Cervical and Endocervical Cells............................52 Susan M. Rose, MD; Andrew P. Soisson, MD; D. Yvette Lacoursiere, MD, MPH; Joel Bentz, MD Nutrient Quality of Competitive Foods in Two Utah Middle Schools....................................................................................56 Linda L. Tsai, MS, RD; Kristine C. Jordan, PhD, MPH, RD; Julie Metos, MPH, RD; Marilyn S. Nanney, PhD, MPH, RD A Pilot Study of Screening Outcomes in Patients at Risk for Atherosclerosis in the Utah Community Clinics.......62 Junhua Yu, PhD; Diana I. Brixner, PhD, RPh; Sameer Ghate, B.Pharm, MSPH; Ken Gondor Health Policy.........................................................................................................67 2009 Utah Legislative Review Megan Bitner 2009 Utah Health Data Review.......................................................................... 79 Health Services Directory.................................................................................167 Arthritis and Co-Morbid Conditions among Adults in Utah Authors: Nicole Bissonette, MPH, CHES Michael Friedrichs, MS Randy Tanner, MPA Shelly Wagstaff, BS 2009 UTAH's HEALTH: AN ANNUAL REviEw Key Words arthritis; co-morbid Conditions About the Authors Nicole Bissonette, MPH, CHES, is Director of the Utah Arthritis Program in the Utah Department of Health Michael Friedrichs, MS, is the Lead Epidemiologist for the Bureau of Health Promotion in the Utah Department of Health Brenda Ralls, PhD, is the Epidemiologist for the Utah Diabetes Prevention and Con-trol Program in the Utah Depart-ment of Health Randy Tanner, MPA, is the Epi-demiologist for the Utah Arthritis Program in the Utah Department of Health Shelly Wagstaff, BS, is an Epide-miologist for the Bureau of Health Promotion in the Utah Department of Health Correspondence Randy Tanner PO Box 142107 Salt Lake City, UT 84114-2107 Phone: [phone] [e-mail] Abstract Most public health reporting and intervention efforts focus on single conditions and rarely address the issue of co-morbid (coexisting) conditions that can have significant health implications. Arthritis has the highest prevalence of any chronic condition among adults, and, therefore, may be the most likely to coexist with at least one other chronic condition. The intent of this report is to present Utah-specific data about arthritis, its relationship with seven other chronic conditions, and the cumulative impact on health indicators like health status, physical and mental health, and activity limitation. Introduction Arthritis is highly prevalent among U.S. adults. It is the leading cause of disability and is associated with substantial activity limitation, work disability, and reduced quality of life. Findings from the National Health Interview Survey (2003, 2005) indicate that an estimated 22 percent of the adult U.S. population (46 million persons) reported they had doctor-diagnosed arthritis. Nearly 41% (19 million) of the 46 million adults with arthritis reported limiting their usual activities due to their arthritis. In addition to activity limita-tions, 31°% (8.2 million) of working-age adults ages 18-64 with arthritis reported they were limited in work activities due to arthritis. As the population ages, the prevalence of arthritis is expected to increase. By the year 2030, an estimated 67 million (25% of the projected adult population 18 years and older) will have arthritis, and 25 million adults will experience arthritis-attributable activity limitations. About 400,000 adults in Utah (22.2%) have been diagnosed with arthritis. Co-morbidity among Utahns with arthritis is common. About two-thirds (62.9%) of Utah residents with arthritis reported they had at least one additional chronic condition. Nearly one in three (29.3%) reported one other condition. Nearly one in five (19.5%) reported arthritis and two conditions, and nearly one in eight (14.1°%) reported arthritis and three or more conditions. Studying these relationships will be helpful in providing more comprehensive treatment options. ©2009 The University o f Utah. All Rights Reserved. Arthritis and Co-Morbid Conditions among Adults in Utah 9 2009 UTAH's HEALTH: AN ANNUAL REviEw Methods The 2005 and 2007 Utah Behavioral Risk Factor Surveillance System (BRFSS) data were combined for this report. The two years of pooled data generated a sample of 2,969 Utah adults who reported arthritis. The BRFSS is an ongoing random tele-phone survey of non-institutionalized adults age 18 and older that has been conducted continuously in Utah since 1984. Variables used in the study were self-reported and were defined as follows: Body Mass Index (BMI) - Weight in kilograms divided by height in meters squared (BMI=kg/m2). A person was consid-ered to be ideal weight if his or her BMI was less than 25, overweight if his or her BMI was between 25 and 30, and obese if his or her BMI was 30 or greater. Co-morbidity - The presence of one or more diseases in addi-tion to a primary disease. Arthritis - For this report, respondents who reported they had ever been told by a doctor or other health professional that they have some form of arthritis, rheumatoid arthritis, gout, lupus, or fibromyalgia, were considered to have arthritis. In other words, respondents who answered "yes" to the BRFSS question "Have you ever been told by a doctor or other health professional that you have some form of arthritis, rheumatoid arthritis, gout, lupus, or fibromyalgia?" were considered to have arthritis. Other chronic conditions: BRFSS respondents who answered "yes" to any of the following questions were considered to have the applicable condition: Angina or coronary artery disease - "Have you ever been told by a doctor, nurse, or other health professional that you have angina or coronary artery disease?" Asthma - "Have you ever been told by a doctor, nurse, or other health professional that you have asthma?" Diabetes - "Have you ever been told by a doctor that you have diabetes?" Heart attack - "Has a doctor or other health professional ever told you that you had a heart attack, or a myocardial infrac-tion? High cholesterol - "Have you ever been told by a doctor, nurse, or other health professional that your blood cholesterol is high?" Hypertension - "Have you ever been told by a doctor, nurse, or other health professional that you have high blood pressure?" Stroke - "Have you ever been told by a doctor, nurse, or other health professional that you have had a stroke?" Perceived healthy days - To assess the perceived number of healthy days among persons with arthritis and multiple con-ditions, four standard BRFSS questions were used. The first question about health status asks individuals to rank their health using a scale from poor through excellent. The next two questions ask individuals to report the number of days dur-ing the past 30 days when physical and mental health were not good. Finally, the activity-limitation question assesses whether a person's activity had been limited in any way because of physical, mental, or emotional problems. Each measure was assessed for persons with only arthritis, arthritis plus one con-dition, two conditions, and three or more conditions. All analyses were conducted using the SUDAAN statistical software package Version 9.0.1. Cross tabulations and logistic regression were used in the analyses. Logistic regression is a commonly used statistical technique used in public health that contrasts the likelihood (odds ratio) of the occurrence of a certain event or condition for one group compared to a reference group. In this study, the absence of arthritis was used as the reference group (odds ratio=1). An odds ratio greater than one implies the condition is more likely to occur in people with arthritis; an odds ratio of less than one implies it is less likely to occur. An odds ratio of one implies no difference between people with and without arthritis. A 95 percent Confidence Interval (CI) was used for all mea-sures in the figures, and all intervals are illustrated by the error bars contained on each bar in all figures. Results The Impact of Arthritis in Utah About 400,000 adults in Utah (22.2%) have been diagnosed with arthritis. The impact of arthritis as a public health problem in Utah is reflected across a variety of demographic measures. Prevalence of arthritis is higher among females and increased with age. One in four Utah women (25.1°%) reported having arthritis, compared to one in five Utah men (19.3%) (p<001). (See Fig-ure 1). Self-reported prevalence of arthritis ranged from 9.6 percent among Utah adults 18-44, compared to over half for those aged 65 and over (p<.001). Note: Because the 95 percent confidence intervals do not overlap for males and females, and for the age groups 18 to 74, the differences between age groups were not likely to have occurred by chance, except for ages 65-74 and 75+. Prevalence of arthritis was higher among those who were obese 10 Arthritis and Co-Morbid Conditions among Adults in Utah ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw Figure 1. Percentage of Utah Adults Reporting Doctor-diagnosed Arth ritis By G ender and Age Group, Utah BRFSS 2005 and 2007 G ender and Age Group Figure 2. Prevalence of Arthritis by Weight Category Utah BRFSS 2005 and 2007 Body M ass Index Figure 3. Percentage of Adults with Doctor-diagnosed Arth ritis and Other Chron ic C on d itions, Utah BRFSS 2005 and 2007 □ Arthritis Alone □ Arthritis and One Condition □ A rthritis and Two Conditions □ Arthritis and Three o r More C onditions ©2009 The University o f Utah. All Rights Reserved. Arthritis and Co-Morbid Conditions among Adults in Utah 11 2009 UTAH's HEALTH: AN ANNUAL REviEw (34.4%) or moderately overweight (22.2%) compared to those who were at an ideal weight (16.4%) (p<.001) (See Figure 2). Note: Because the 95 percent confidence intervals do not overlap, the differences were not likely to have occurred by chance. The Impact of Arthritis among Adults with Multiple Chronic Conditions in Utah As seen in Figure 3, co-morbidity is common among people with arthritis. About two-thirds (62.9%) of persons with arthritis reported they had at least one additional chronic condition. Nearly one in three individuals who reported arthritis (29.3%) reported one other condition. Nearly one in five (19.5%) reported arthritis and two conditions, and nearly one in eight (14.1°%) reported arthritis and three or more conditions. Arthritis is frequently associated with other health condi-tions like high blood pressure and other chronic diseases. In this study, the co-existing prevalence of seven conditions was examined for adults with and without arthritis. For each of the seven conditions, the prevalence of other conditions was higher among adults with arthritis than those without it. In particular, the prevalence of high cholesterol, high blood pressure, and asthma was high among adults with arthritis. (See Figure 4). Note: Because the 95 percent confidence intervals do not over-lap when comparing persons with and without arthritis, the differences were not likely to have occurred by chance. Factors that generally increase the risk of arthritis are being female, aging, and being overweight or obese. Even after ad-justing for age, sex, and BMI, a diagnosis of arthritis signifi-cantly increased the odds for reporting other health conditions. Having arthritis doubled the odds of reporting asthma, high blood pressure, angina and coronary artery disease (CAD), and increased the odds of reporting they had a heart attack, stroke, high cholesterol or diabetes by roughly 50 percent, compared to people without arthritis. All odds ratios were statistically significant. Healthy Days Whether the measure was health status, unhealthy mental or physical health days, or activity limitation, adults who reported having three or more chronic conditions in addition to arthritis were more likely to report poorer health and activity limitation. Perceptions about health are very important and can serve as a proxy measure for the perceived burden of chronic health conditions. The impact of having conditions in addition to arthritis and reporting fair or poor health is evident in Figure 6. The percentage reporting fair or poor health increased with each additional condition, and was three times greater among adults who reported three or more conditions when compared to persons who report arthritis only. The most striking differ-ence is found between those who report having arthritis alone and those who report having arthritis and three or more other chronic conditions (15.5% vs. 50.8%; p<.001). A similar relationship is observed when the outcome measure is seven or more days of poor physical health within the past 30 days (See Figure 7). Nearly one in five adults (21.8%) with arthritis reported only seven or more days of poor physical health in the past 30 days, compared to nearly half (45.4%) who reported arthritis and three conditions (p<.001). The relationship between arthritis and the number of condi-tions and mental health is less dramatic; however, the pattern remains the same. One in six adults with arthritis only (16.4%) reported seven or more days of poor mental health within the past 30 days compared to one in five (22.4%) adults with arthri-tis and three or more conditions (See Figure 8). None of these differences were statistically significant. However, the differ-ence between those with arthritis alone and arthritis with three other chronic conditions approached statistical significance (p=.05). Activity limitation is another important measure of health. Not surprisingly, the percentage of persons reporting activity limi-tations increased with number of chronic conditions reported (See Figure 9). One-third (34.5%) of adults with arthritis only reported limiting their usual activities, compared to more than half (55.9%) of those with arthritis and three or more condi-tions (p<.001). Discussion There are a number of possible explanations for the number of conditions observed in persons with arthritis. Detection bias may have some effect. Individuals with one condition may have more contacts with the medical care system and may have a greater likelihood of being diagnosed with a second condi-tion. Individuals' response patterns to questionnaires may also play a role. Those who acknowledge having one disease may be more likely to acknowledge having other diseases. Finally, genetic and environmental factors may increase general sus-ceptibility to disease in some individuals, and increase the occurrence of multiple diseases during their later years of life. Implications The coexistence of multiple health conditions impacts medical outcomes and is important in determining effective treatment protocols. Therefore, managing the care of persons with arthritis who have multiple health conditions should be oriented toward a person's overall health rather than focusing on an individual disease. This requires: 12 Arthritis and Co-Morbid Conditions among Adults in Utah ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw 40 30 r<ou ra <5 20 OC<Lu 10 31.4 Figure 4. Age-adjusted* Percentage of Utah Adu lts with O ther Health Conditions By Arth ritis Status, Utah BRFSS 2005 and 2007 * Rates are adjusted to the U.S. 2000 standard population for comparison purposes. 23.2 H- 27.7 16.5 H- 14.4 6.1 -i- 9.1 5.7 -I- 4.3 2.9 •h- i r 4.3 -i- 2.4 1- i r 3.3 -i- 2.0 t- - i High Cholesterol High Blood Pressure Asthma Diabetes Heart Attack Angina or CAD Other Health Conditions Stroke 0 □ With Arthritis □ Without Arthritis =ra 2 O' Uu> O 1 2.3 2.2 Figure 5. Odds Ratios o f C h ro n ic D ise a se s for Ad u lts With A rthritis C om p a red to T h o se Without Arthritis, Utah BRFSS 2005 and 2007 * Data are adjusted for sex, age, and body mass index for comparison purposes. 2.0 ** The reference group is people without arthritis (indicated by horizontal line). 1.7 1.6 1.6 1.5 Asthma Angina or CAD High Blood Pressure Heart Attack Stroke High Cholesterol Diabetes Chronic Disease 3 0 Figure 6. Percentage o f P e rsons With A rthritis Who Reported Fair or Poor Health By Number of Conditions, Utah BRFSS 2005 and 2007 Number o f Conditions ©2009 The University o f Utah. All Rights Reserved. Arthritis and Co-Morbid Conditions among Adults in Utah 13 2009 UTAH's HEALTH: AN ANNUAL REviEw Figure 7. Percentage of Persons With Arthritis Who Reported Seven or More Days of Poor Physical Health By Number of Conditions, Utah BRFSS 2005 and 2007 Number of Conditions Figure 8. Percentage of Persons With Arthritis Who Reported Seven or More Days of Poor Mental Health By Number of Conditions, Utah, BRFSS 2005 and 2007 Number of Conditions Figure 9. Percentage of Persons With Arthritis Who Reported Activity Limitation By Number of Conditions, Utah BRFSS 2005 and 2007 Number of Conditions 14 Arthritis and Co-Morbid Conditions among Adults in Utah ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH'S HEALTH: AN ANNUAL REViEW • Increasing dialogue among public health professionals, medical providers, medical specialists, persons with arthri-tis, private organizations, and others, to better understand the high prevalence of arthritis and other health conditions. • Developing alliances among public and private health care systems to provide a more complete approach to treating arthritis and co-morbid conditions. • Using system resources more effectively to simultaneously address multiple conditions. • Collaborating with programs within state government, es-pecially the Bureau of Health Promotion, that have common goals such as increasing physical activity and/or reaching and maintaining a normal weight to help the common target audience. • Using cross-cutting, evidence-based self-management edu-cation programs (like the Chronic Disease Self-Management Program) and physical activity programs (like EnhanceFit-ness) to dramatically increase the reach of the Bureau of Health Promotion Programs. • Promoting and using arthritis-specific self-management education programs (like the Arthritis Self-Management Program) and physical activity programs (like the Arthri-tis Foundation Exercise Program and Arthritis Foundation Aquatic Program) to address the arthritis-specific concerns of people with arthritis and other chronic conditions. • Embedding arthritis information in other chronic disease program messages to reduce arthritis-specific barriers among people with other chronic conditions who also have arthritis. Conclusion Arthritis is a debilitating disease that may limit the indepen-dence of affected persons and disrupt the lives of their family members and caregivers. The impact can be especially dev-astating when multiple chronic health problems coexist with arthritis. Nearly two-thirds of adults who reported they had arthritis also reported they had at least one other condition. Increasing the awareness of arthritis and co-morbidity and collaboration with private and public partners will focus resources and help people with arthritis better meet their goals and needs and im-prove their overall quality of life. These findings also suggest more needs to be done to help people with arthritis and other coexisting chronic conditions get physically active to improve their health. Engaging in regu-lar physical activity and maintaining a healthy weight can help alleviate the pain and disability that often accompany arthritis. Disease self-management classes such as the Chronic Disease Self-Management Program, the Arthritis Foundation's Exercise Program, and EnhanceFitness may help adults with arthritis and other chronic conditions better manage their diseases. References Anderson, G., & Saudek, C. (2002). A cute Solutions fo r a Chronic Problem (John Hopkins University & The Robert Wood Johnson Foundation.) (Vol. 1 No. 1). Anderson, G., & Hovarth, J. (2003). Chronic conditions: m aking the case fo r ongoing care (C. Anderson, Ed.). www.partnershipforsolutions.org : John Hopkins University. Arthritis Foundation, Association o f State and Territorial Health Officials, & Centers for Disease Control and Prevention. (1999). National A r th r itis A c tio n Plan: A Public Heath S trategy. Centers for Disease Control and Prevention. (May 12, 2008). A r th r itis . Retrieved April 3, 2008, from Arthritis Program Web site: http://www.cdc. gov/arthritis/ Hoffman, C., Rice, D., & Sung, H. (1996). Persons with chronic conditions: their prevalence and costs. JAMA, 276, 1473-1479. Hootman, J., Helmick, C., & Bolen, J. (2006). Projections o f U.S. prevalence o f arthritis and associated activity limitations. Arthr itis R heum, 54, 226-229. Hootman, J., Helmick, C., Bolen, J., & Langmaid, G. (2006, October 13). Prevalence o f doctor-diagnosed arthritis and arthritis-attributable activity lim itation. MMWR, 55, 1089-1092. Lightsey, D. (2007, September 17). C D C 's A r th r itis Program an Overview. Address presented at Arthritis Program Coordinators Orientation, Atlanta, Georgia. Utah Department o f Health, Center for Health Data. 2003 and 2005 Behavioral Risk Factor Surveillance System. Unpublished raw data. Verbrugge, L. M. (1991). Risk factors for disability among US adults with arthritis. J C lin Epidemio l, 44, 167-82. ©2009 The University o f Utah. All Rights Reserved. Arthritis and Co-Morbid Conditions among Adults in Utah 15 2009 UTAH's HEALTH: AN ANNUAL REviEw Key Words WIC community needs assessment; healthy food choices; physical activity Correspondence Phyllis Crowley, MS, RD, IBCLC Utah Department of Health/WIC PO Box 141013 SLC, Utah 84114-1013 Phone: [phone] Fax: [phone] email: [e-mail] * Utah Department of Health **Weber Morgan Health Department *** University of Utah An Assessment of Community Access to Physical Activity Resources and Healthy Food Choices Authors: Phyllis Crowley*, MS, RD, IBCLC Shaheen Hossain*, PhD Angeni Marque*, BS Robert Satterfield*, M.Stat Lynda Blades*, MPH, CHES Heather Borski*, MPH, CHES Christopher D. Furner*, MS, CHES Patrice Isabella*, MS,RD Colleen Jenson**, BS, LE Tim Stempel***, MSW Nan Streeter*, MS,RN Rick Wardle*, BS Abstract The Women, Infants, and Children (WIC) program in Utah implemented a pilot study to assess current access to healthy food choices and the availability of safe physical activity among a Utah population who are at high-risk for obesity. A needs assessment tool entitled, WIC Healthy Living Survey, was developed and distributed to 650 WIC participants who were enrolled at the Ogden WIC Clinic during May of 2008. Overall, 69% said there were trails for walking, hiking, or bicycling near their homes. Among all respondents, 96% reported that there are grocery stores near their home that offer healthy foods. Examination of results of this pilot needs assessment suggests further study directions that would offer additional insight into the building blocks that make up a healthy lifestyle. Introduction Healthy lifestyle choices are much easier to make when neighborhoods, worksites, and schools support them. Generally, obesity is related to the development of serious chronic conditions such as heart disease and diabetes, increased health care costs, premature death, and a reduced quality of life. While obesity may directly result from the choices individuals make concerning physical activity and diet, strong environmental forces are at work that influence those choices. Every day individuals make food choices based on convenience and price. When communities have too many sources of inexpensive, low-nutrient, high-calorie foods, and limited access to affordable fresh fruits and veg-etables, meal choices reflect foods that are readily available. The prevalence of obesity varies by gender, age, socioeconomic status, and ethnicity and those most at risk come from low-income, ethnically/racially diverse communities that have the least access to parks, bike trails, and public pools. An objective of Healthy People 2010 is to reduce to 16 Community Access to Physical Activity Resources and Healthy Food Choices ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw Table 1. Existence of Community Recreation C enters Near Participant's H ouseholds Community recreation centers near households State # Percent English # Percent Spanish # Percent Yes 308 48.9 237 48.6 71 50.0 No 127 20.2 100 20.5 27 19.0 Not Sure 195 31.0 151 30.9 44 31.0 Total Response 630 100.0 488 100.0 142 100.0 Missing 18 11 7 Total 648 499 149 Figure 1. P ercent of Responden ts Who Reported that They Feel Safe Being Active Outdoors Q9. Do you feel safe being active o u td o o rs n e a r your hom e? co o Q. 100 90 80 70 60 50 40 30 20 10 0 87.85 88.98 84.03 6.31 6.33 6.25 5.84 4.69 9.72 Yes No Not Sure □ All □ English □ Spanish 15% the proportion of adults who are obese (U.S Department of Health and Human Services, 2000). In 2007, no state met this target as 25.6% of adults in the United States were obese (Center for Disease Control, 2006). During 2007 22.4% of Utah adults were obese (Center for Disease Control Behavioral Risk Factor Surveillance System, 2007). To reach the Healthy People 2010 target, increased national attention on actions that promote healthy eating and physical activity is essential. Prior-ity must be given to interventions that move beyond only in-creasing individual awareness about healthy choices, but rather toward those that provide tangible support for environmental changes that promote healthy lifestyles among those with the greatest need. In the Surgeon General's 2001 Call to Action to Prevent and Decrease Overweight and Obesity, many activities were identified that focus on increased access to healthy food choices and safe physical activity in settings such as worksites, communities, and schools (U.S. Department of Health and Hu-man Services, 2001). The purpose of this pilot study is to as-sess current access to healthy food choices and the availability of safe options for physical activity among a particular Utah population who are at high risk for obesity. Study Objective The Special Supplemental Nutrition Program for Women, In-fants, and Children - commonly known as the WIC Program - at the Utah Department of Health was the recipient of an Association of State and Territorial Public Health Nutrition Directors (ASTPHND) Blueprint Seed Grant in 2008. As a result, the Utah Department of Health implemented a project-- Cornerstones of a Healthy Lifestyle--designed to address both nutrition and physical activity as important components of healthy living. The grant workgroup determined that the most beneficial use of grant monies would be to develop a family-based community needs assessment tool to ascertain existing resources for physical activity and healthy eating in a com- ©2009 The University o f Utah. All Rights Reserved. Community Access to Physical Activity Resources and Healthy Food Choices 17 Table 2. B arriers to Engaging In More Physical Activity 2009 UTAH'S HEALTH: AN ANNUAL REViEW Reasons that keep participants from doing more physical activity (check all that apply) oR esp#o nodf en*t s Percent English # Percent Spanish # Percent a. Physical activities cost too much 108 16.7 97 19.4 11 7.4 b. I have an injury/health problem no physical activity 53 8.2 37 7.4 16 10.7 c. I do not feel safe outdoors in my community 47 7.3 33 6.6 14 9.4 d. There is no convenient place for physical activity 58 9.0 43 8.6 15 10.1 e. I do not enjoy physical activity 33 5.1 22 4.4 11 7.4 f. I have no time for physical activities. 191 29.5 162 32.5 29 19.5 g. No transportation to do physical activities 71 11.0 51 10.2 20 13.4 h. Other 137 21.1 110 22.0 27 18.1 Figure 2. Percent of Responden ts Who Reported that G rocery S tores Offer Healthy Foods Q14. Are there grocery stores near your household that offer healthy foods that you are able to buy? a£) o Qa). 120 100 80 60 40 20 0 92.89 96 11 81.94 4 9 2.25 13.89 Yes No 2.21 1.64 4.2 Not Sure □ All □ English □ Spanish munity at high risk for overweight and obesity. The resulting findings of this needs assessment survey would allow WIC staff and others to develop strategies, consistent with available community resources, to target nutrition intervention and edu-cation in a way that would best support families in their efforts to practice healthy lifestyles. Methods Survey Instrument The ASTPHND grant workgroup hired a University of Utah graduate student in January of 2008 to con-duct a literature review of existing needs assessment tools and to develop an appropriate tool for the Utah WIC community. The grant workgroup, consisting of the graduate student and several public health pro-grams, developed an assessment tool entitled WIC Healthy Living Survey. The survey consisted of 31 questions divided into three distinct parts: individual and family physical activity levels; individual and family food and nutrition patterns, and demographic information. After pilot testing, this tool was ready for distribution by the end of April 2008. In order to accommodate the language preferences of WIC participants the Healthy Living Survey was prepared in both English and Spanish. Thus, survey findings were also reported by language version. 18 Community Access to Physical Activity Resources and Healthy Food Choices ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH'S HEALTH: AN ANNUAL REViEW Study Site Weber County was selected as the study site because it had been identified as an area with the highest risk for overweight and obesity based on the Utah Behavioral Risk Factor Surveil-lance System, Small Area Report, 2001-2005 (Utah Depart-ment of Health Office of Public Health Assessment, 2007). Thus, the WIC Healthy Living Survey was piloted only at the Ogden WIC clinic during May of 2008. There were 650 sur-veys distributed in both English and Spanish versions to a con-venience sample of Ogden WIC participants. The completed surveys were provided to the Utah Correctional Institute (UCI) for data entry. The UCI returned the completed data output in the form of an Excel spreadsheet. Univariate and bivariate analyses were performed. All analyses were conducted at the Utah Department of Health using SAS. Results Demographics Univariate analysis of the demographic data revealed that 83% of survey respondents were adult women between the ages of 18 and 35 years. Of all respondents, 47% identified their ethnicity as White while 46% identified their ethnicity as Mexican, Puerto Rican, or Other Hispanic/Latino. The aver-age household size reported by survey respondents was 4.3 persons. The average number of people per household among those completing the survey either in English or Spanish was 4.2 and 4.7, respectively. The average number of children per household was 2.3. The average number of children per house-hold was 2.2 among those taking the English version and was found to be 2.6 among Spanish language respondents. Of all survey respondents, 35% indicated that they use Food Stamps. Physical Activity Access Overall, 69% of survey respondents reported that there are trails for walking, hiking, or bicycling near their homes. Close to half (49%) reported that there are community recreation centers nearby (see Table 1). However, 16% of respondents said they do not use any community recreational resources in a given week. Sixty-four percent of all respondents indicated that their children do not play on sports teams or engage in after school physical activities (see Table 2). Additionally, 30% indicated that they have no time for physical activities. Feeling safe be-ing active outdoors near their homes was reported by 88% of all respondents (see Figure 1). Cost was a barrier to physical activity for 17% of those who responded to the survey. Healthy Food Access Overall, 93% of survey respondents indicated that there are nearby grocery stores offering healthy foods. However, this percentage varied among respondents who completed the survey in different languages (see Figure 2). In general, 60% said that there are restaurants nearby that offer healthy foods. Sixty-seven percent of WIC participants who took the survey in English reported that there are nearby restaurants offering healthy foods. Of those who took the survey in Spanish, only 38% reported similarly. Fifteen percent mentioned that they eat at restaurants or fast food three or more days per week. Fresh fruits and vegetables were eaten every day by only 19% of families. Eighty-six percent of survey respondents said that they eat meals cooked at home five or more days per week. Among the reasons provided by survey participants regarding why they were unable to provide more healthy meals for their families, the cost of healthy foods (38%) and lack of time to cook (14%) were the top two reasons. Overall, more than one in ten (15%) reported that they often worry about not having enough food for their families ("Always"- - 6.3% and "A lot"- - 8.2%). Conclusion The results of this community needs assessment survey piloted in the Ogden WIC clinic allow Utah WIC staff to now develop strategies, consistent with available community resources, which target nutrition interventions and education to best support families in their efforts to practice healthy lifestyles. The information ascertained about the influences of safe, af-fordable, and convenient opportunities for physical activity on healthy lifestyle choices provide the foundation from which evidence-based strategies may be implemented to improve physical activity levels for individuals and families. The find-ings of this survey focusing on the links between self-reported understanding of a nutritious diet and easy access to healthy food choices among a community at high risk for overweight and obesity offer valuable insight to direct the advance of strat-egies that better support healthy food choices. Examination of the results of this Utah WIC community needs assessment survey suggests further questions and study direc-tions that promise to offer additional insight into the building blocks that make up a healthy lifestyle. Further study may explore the impact of WIC nutrition and physical activity education on individual lifestyle choices through comparison of new WIC participants with those who have had the benefit of the ongoing educational opportunities provided by WIC professionals. By contrasting the healthy lifestyle choices of families participating in the WIC program with those in other demographic groups, additional pieces to the puzzle of how to best support and facilitate healthy living among the most vulnerable populations may become apparent. Such a compari-son might be made between families in the WIC program and families in the Medicaid program, for example. Refinements of the survey tool itself may provide additional clarification of the motivators for crucial choices regarding healthy eating and optimal physical activity. Other Utah WIC clinics, as well as WIC programs in several other states, have expressed interest in conducting studies similar to this in their own geographic areas. ©2009 The University o f Utah. All Rights Reserved. Community Access to Physical Activity Resources and Healthy Food Choices 19 2009 UTAH's HEALTH: AN ANNUAL REviEw The receipt of this Association of State and Territorial Public Health Nutrition Director's (ASTPHND) Blueprint Seed Grant in 2008 was the catalyst for enabling the Utah WIC program to make great strides toward identifying the barriers to a vital, healthy lifestyle for all Utah families and developing effective strategies for Utah's population. References ASTPHND. (2006). Cornerstones o f a healthy lifestyle: Blueprint for nutrition and physical activity. Retrieved from http://www.astphnd.org/resource_ files/42/42_resource_file1.pdf. Center for Disease Control and Prevention Behavioral R isk Factor Surveillance System. (2007). Prevalence and trend data, Utah-Utah, overweight and obesity (BMI). Retrieved March 27, 2007 from http://apps.nccd.cdc.gov/BRFSS/. Center for Disease Control and Prevention. (2005). State-specific prevalence o f obesity among adults- United States, 2005. Morbidity and Mortality Weekly Report, 55, 985-988. U.S. Department o f Health and Human Services. (2001). The Surgeon General's call to action to prevent and decrease obesity. Rockville, MD: U.S. Department o f Health and Human Services, U.S. Public Health Service, Office o f the Surgeon General. US Department o f Health and Human Services. (2000). Healthy people 2010: Understanding and improving health, 2nd edition. Washington D.C.: U.S. Department o f Health and Human Services. Utah Department o f Health Office o f Public Health Assessm ent. (2007). Utah Behavioral R isk Factor Surveillance System small area report, 2001-2005. Salt Lake City, Utah: Utah Department o f Health. 20 Community Access to Physical Activity Resources and Healthy Food Choices ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw Key words customer satisfaction; women in-fants and children program; nutri-tion; WIC outcomes; survey Correspondence Shaheen Hossain, PhD Program Manager Data Resources Program, MCH Bureau Utah Department of Health PO Box 142001 Salt Lake City, Utah 84114-2001 Phone: [phone] email: [e-mail] Acknowledgement The authors would like to thank all WIC Directors, clinics and State WIC staff for their continued support in conducting this important survey. We also extend our thanks to WIC partici-pants for taking the time to fill out the survey enabling the WIC program to improve services for Utah mothers and children. Patterns o f Participant Satisfaction with Utah WIC Program Authors: Shaheen Hossain, PhD Rob Satterfield, Mstat Angeni Marque, BS Phyllis Crowley, MS, RD, IBCLC Nan Streeter, MS, RN Chris D. Furner, MS, CHES Abstract The WIC program is a federally funded nutrition assistance program that provides nu-tritional assessment, education, and healthy food vouchers to lower-income pregnant, breastfeeding, and postpartum women, infants, and children up to age 5 who are at nutritional risk. The purpose of this study was to examine participant satisfaction with WIC services in Utah, as well as to determine participant understanding and need for those services through analyses of the 2008 Utah WIC Participant Satisfaction Survey. The survey examined the patterns of service utilization, nutrition education, behavioral change, voucher use, and food preferences among WIC participants. The knowledge gained will optimize WIC strategies and enhancements that will ensure that the WIC program can continue to improve the health of mothers and of young children during their most critical times of growth and development. Introduction Offering high quality customer service and satisfying customers are crucial aspects of any organization or business. Business management theories highlight the importance of customer satisfaction for the success of a business. When customers are unhappy and perceive poor quality service, they can jeopardize the business or services simply by word of mouth communication. It is imperative for any sound organization to be aware of customer expectations in order to improve product and service delivery practices. These same customer satisfaction principles apply to public health organizations that offer services to high-risk populations. The development and delivery of services, programs, and products provided by public health organizations can also benefit greatly from as-sessments of participant satisfaction (Christie et al., 2006). In order to achieve the vision set by Public Health In America -- Healthy People in Healthy Communities-- assessment of public health program performance and quality of service delivery is essential (Public Health Functions Steering Committee, 1995). Public health efforts continually strive for improvement in their services, programs, and outputs provided and are held accountable to the communities served. The WIC Program is a supplemental nutrition program for women, infants and children under five whose goal is to strengthen the nutritional health of participants through improved diet and access to health care. The program has an extraordinary record of preventing health problems and improving long-term health and development. ©2009 The University o f Utah. All Rights Reserved. Patterns of Participant Satisfaction with Utah W IC Program 21 2009 UTAH's HEALTH: AN ANNUAL REviEw The major goal of the study was to examine the extent of par-ticipant satisfaction with WIC services as well as to determine their understanding and need for those services. A brief de-scription of the WIC program and the services they offer are provided below. Background Overview of WIC Program The Special Supplemental Nutrition Program for Women, Infants and Children (WIC) is a federally funded nutrition assistance program administered by the U.S. Department of Agriculture's (USDA) Food and Nutrition Service (FNS). WIC is not an entitlement program. Congress does not set aside funds to allow every eligible individual to participate in the program. WIC is a federal grant program for which Congress authorizes a specific amount of funding each year for program operation (Nutrition Program Facts, 2006). The federal FNS provides these funds to State WIC agencies for administra-tion of the program at state and local levels. During fiscal year 2000, this federal program spent $4.1 billion to provide nutri-tional assistance to lower-income pregnant, breastfeeding, and postpartum women, infants, and children up to age 5 who are at nutritional risk throughout the country (GAO, 2001). More recent data shows that WIC was funded at a level of more than $5 billion in fiscal year 2005 (GAO, 2006). Program Benefits The WIC program offers three major benefits to all partici-pants: a. Vouchers for specific supplemental foods b. Nutrition education and counseling c. Referrals to health care and social services a. Supplemental food vouchers WIC provides participants with supplemental foods that are high in the nutrients frequently lacking in the diets of low-in-come pregnant women and children (Oliveira, Racine, Olmst-ed, Ghelfi, 2002). Nutritional risk information collected during initial assessment and periodic clinic appointments identifies medical conditions and nutritional factors that could put WIC participants at risk for nutrient deficiencies and poor health outcomes. Food vouchers are then tailored to the specific needs of each participant. WIC state agencies provide food vouchers to participants who then exchange them for supplemental foods at authorized retail stores. b. Nutrition education and counseling Nutrition education in the WIC program focuses on the relationship between nutrition and health, and counsels par-ticipants to make positive changes in eating habits (Devaney, Ellwood, Love, 1997). During WIC nutrition classes, WIC par-ticipants learn about healthy eating and they are encouraged to adopt positive food-related attitudes and behaviors. Nutrition education also encompasses many other important topics such as breastfeeding promotion, extending the duration of breast-feeding, and the harmful consequences of alcohol, tobacco and other drugs. c. Referrals to health care and social services The WIC program assists participants with referrals to health care services and other community resources. The WIC staff regularly refer participants to appropriate health care and social services. The USDA encourages and promotes greater coordi-nation among public health agencies to offer more systematic and comprehensive care. Such care coordination enables clients to access primary and preventive health care. Program Participation The number of individuals participating in WIC has increased significantly since the program's inception in 1972. Nationally, WIC now serves almost 9 million participants in all 50 states, the District of Columbia, 5 territories and 34 Indian Tribal Organizations (Bell, Ledsky, Silva, Anthony, 2007). The Utah WIC program, which is administered by the Utah Department of Health, Division of Community and Family Health Services, has also experienced an increase in program enrollment over the past years (see Figure 1). WIC services are provided at pub-lic health clinics located in 29 Utah counties (except Daggett) within the boundaries of 12 designated health districts. It is important to note that WIC is not an entitlement program and the number of people served by the program may be lim-ited by funding levels established by Congress. The numbers of participants that can be served each year depends upon the annual appropriation of funds and the cost of operating the program. Of primary concern are low participation rates in the WIC program among eligible women and children because eligibility for WIC services by definition includes a population possessing a greater risk for nutrition-related morbidity and mortality (Christie et al., 2006). Such high-risk populations are often in need of nutritional guidance. Program Outcomes The primary goal of the WIC program is to improve the health of mothers and children during their most critical times of growth and development. Much evaluative research has been conducted to assess WIC program benefits. The literature shows evidence that the WIC program has assisted families and promoted the health of mothers and children (Edozien, Switzer, Bryan, 1979, Lee et al., 2006). Low-income chil-dren enrolled in the WIC program have a lower prevalence of anemia than those who are not enrolled (Lee et al., 2006). Repeated measurements of the hematological status of partici-pants taken at certification provide fairly convincing evidence of WIC's effect on reducing the incidence of iron-deficiency anemia among infants (Devaney et al., 1997). Studies have found that WIC participation was linked to increased birth 22 Patterns of Participant Satisfaction with U tah W IC Program ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw weight and lower incidence of preterm births (Mathematica Policy Research Group, 1989). Also children receiving WIC benefits were more likely to receive regular preventive health care services than children not enrolled in WIC (Lee et al., 2000). Children enrolled in WIC were more likely to receive the recommended immunizations, and more likely to have a regular source of health care compared to non-WIC children (Devaney et al., 1997). Another study found that children on WIC were more likely to obtain preventive health care services through Early Periodic Screening, Diagnosis, and Treatment (EPSDT) services (Lee et al., 2000). Purpose of the study The purpose of the study was to examine the extent of par-ticipant satisfaction with WIC services as well as to determine their understanding and need of those services. The study ex-amined patterns of service utilization, nutrition education and behavioral change, breastfeeding, voucher use, and participant services. Associations between reported behavior change re-sulting from WIC participation and participant assessment of the quality of WIC services were also examined. Methods Sampling The study population consisted of participants enrolled in the Utah WIC program. The Participant Satisfaction Survey was designed to collect cross-sectional data from a proportional sample of WIC clinic participants during the months of March through June 2008. The State WIC office mailed a total of 3735 surveys and survey administration instructions to all WIC clinics. Each clinic was instructed to give the survey to partici-pants who entered the clinic on designated "regular service" days during the collection period. Surveys were obtained from participants who agreed to take part. To protect respondent's privacy and confidentiality, the survey did not request any unique personal identifiers. Survey instrument Although based on a survey instrument that has been used since 2001, the 2008 Participant Satisfaction Survey contained some questions that had been modified from those used in previous years. The purpose of revamping the instrument was to optimize the information garnered to best support im-provements in WIC programs and services at both state and local levels. The survey instrument consisted of 25 questions covering service awareness and utilization, nutrition educa-tion, clinic performance, voucher use, and food preferences. A draft of the survey tool was piloted to ensure adequate flow and comprehension by respondents before the instrument was finalized. Examination of the WIC program enrollment data Figure 1. Utah WIC Program Enrollments, 2005 - 2007 72000 -| 70000 s 68000 aM) ° 66000 o L. £ 64000 ED N 62000 60000 58000 - Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 1 1CY07 64757 64147 63994 63358 63429 62951 62749 63555 63787 65019 64917 64593 CY06 67675 66473 66145 64458 64513 64914 64914 63964 65107 66768 66065 64620 CY05 68156 67198 69517 68989 68710 68377 67067 67989 68283 69007 68984 67965 Months ] CY07 -^ CY06 - CY05 ©2009 The University o f Utah. All Rights Reserved. Patterns of Participant Satisfaction with Utah W IC Program 23 2009 UTAH's HEALTH: AN ANNUAL REviEw Table 1: Demographic C haracte ristics of WIC Participants Characteristic Number Percentage Age (yr) 15 - 19 20 - 29 > 30 Mean + SD Education < 8th Grade 9th - 11th Grade High School Some College College Graduate Time in WIC Program < 6 months 6 - 12 Months 1 - 2 Years 3 - 5 Years > 5 Years Type of Participant Pregnant Breastfeeding Postpartum Parent of Infant Parent of Child 333 1834 994 (27.0 + 6.4) 372 820 934 764 394 890 765 817 658 167 774 583 395 1203 1354 10.5 58.0 31.4 11.3 24.9 28.6 23.3 12.0 Results A total of 3,586 WIC participants com-pleted the survey with a response rate of 96.0%. Almost thirty four percent (33.6%) of the respondents completed the survey in Spanish. Overall, the average respondent age was 27.0 years (± SD = 6.4). Based on the distribution of responses, three age categories were created (see Table 1). The majority of survey respondents (58.0%) were between 20 and 30 years of age. 27.0 23.2 24.8 20.0 5.1 21.6 16.3 11.0 33.5 37.8 revealed that 39.0% of participants identified their ethnicity as Hispanic. In order to accommodate the language preferences of this rapidly growing participant group as well as to obtain culturally appropriate information, the survey instrument was administered in both English and Spanish. Statistical Analysis Analyses for this study included descriptive statistics includ-ing frequencies, chi squares, means, t-tests, as well as logistic regression. The rating of participant satisfaction with WIC services was measured using a four-point Likert scale (poor = 1, fair = 2, good = 3, and excellent = 4). Logistic regression was used to examine the effect of participant satisfaction on behavioral or lifestyle changes made since participating in the WIC program. In the regression analysis, the rating scale was recoded to create a dichotomous variable (participants who rated services as ‘excellent' were coded 1 and those who rated services as ‘poor', ‘fair', and ‘good' were combined and coded 2). The statistical significance was accepted at p <.05 level. Since each participant was allowed to select the language ver-sion of the survey they preferred, survey findings were also analyzed by language version. All analyses were performed using SAS version 9.1. When asked the question, "which of the WIC requirements is hardest for you?" a large percentage of respondents (60.5%) indicated that none of the requirements are hard. However, close to fourteen per-cent (13.8%) mentioned that the hardest requirement was keeping appointments. Also, more than one in ten participants (11.6%) indicated that bringing children to appointments was the hardest WIC requirement (see Table 2). More than half (55.9%) of WIC participants indicated that the reason for their last appointment with the WIC clinic was certification. About thirty-five percent (34.9%) said the purpose of their last WIC appointment was to attend a class. Close to half of sur-vey respondents (48.7%) reported that it took Vi hour to 1 hour to complete their last appointment. Very few respondents took 2 hours or more to complete their last appointment. Slightly more than half (53.8%) of survey par-ticipants said they were able to spend up to 30 minutes in the clinic to get the required nutrition education. Almost one-fifth of respondents (19.5%) said they were able to spend up to 45 minutes in the clinic. WIC voucher use information is provided in Table 3. Almost all (96.4%) of the survey respondents reported that the WIC clinic staff explained to them how to use their WIC vouchers at the store. When asked the question, "please rate your understand-ing of how to use the WIC vouchers," a large majority of WIC participants (68.2%) rated their understanding as "excellent." Close to one-third (28.5%) said their understanding of how to use the vouchers was "good." However, there was a discrepancy between the different language versions of the survey on the percentages of respondents who rated their understanding of WIC voucher use as "excellent" (English: 75.2% vs. Spanish: 53.9%). There were 59.3% of survey participants who reported that a cashier sometimes said they picked the wrong WIC foods. Approximately one-third (33.5%) said a cashier never told them that they picked the wrong WIC foods. No statistical 24 Patterns of Participant Satisfaction with U tah W IC Program ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw Table 2: Service Utilization Patterns among WIC Participants Characteristic All (n=3586) % English (n=2382) % Spanish (n=1204) % P Value* Difficulty in Meeting WIC Requirements Keeping Appointments 13.8 14.7 12.1 0.0001 Completing Forms 2.3 0.9 5.1 Getting Samples 3.6 3.8 3.1 Proof of Income / Identity 4.6 4.0 5.8 Proof of Residency 0.7 0.9 0.4 Bringing Children 11.6 14.3 6.2 Other 2.9 3.2 2.3 None are Hard 60.5 58.2 65.1 Reason for Last Appointment Certification 55.9 56.3 55.0 0.0174 Class 34.9 33.7 37.4 Other 9.2 10.0 7.6 Length of Last Appointment < / Hour 30.8 28.8 34.8 0.0017 / - 1 Hour 48.7 50.8 44.5 1 / Hours 13.1 13.3 12.7 2 Hours 5.0 4.9 5.1 > 2 Hours 2.5 2.3 3.0 Able to Spend Time at Clinic for Nutrition Education < 30 Minutes 53.8 54.7 51.9 0.0324 30 - 45 Minutes 19.5 19.8 19.8 45 - 1 Hour 20.1 19.8 20.9 > 1 Hour 6.5 5.7 8.1 Numbers may not sum to total due to missing values * Chi Square Distribution, p values of <0.05 were considered statistically significant differences were observed between those who completed the survey in Spanish and English. More than half of all respondents (54.8%) said they eat more fruits and vegetables since they started participating in the WIC program. About 43.4% mentioned that they drink less sodas and sweetened drinks since they started coming to WIC. More than one-third (35.1%) indicated that they have changed their behavior from giving their babies juice in a bottle to giv-ing juice to their babies in a cup. Close to one-third (31.6%) said that because of coming to WIC they participate in more physical activities. One in five (20.7%) reported that they spend less time watching TV and playing video games. When asked about preferred methods of receiving nutrition educa-tion (where participants were allowed to select more than one preference), "taking a packet of information home to read" and "attending WIC classes" were the favorite methods (51.0% and 46.1%, respectively). Close to one-third of respondents ©2009 The University o f Utah. All Rights Reserved. Patterns of Participant Satisfaction with Utah W IC Program 25 2009 UTAH's HEALTH: AN ANNUAL REviEw Table 3: Knowledge of Voucher Use Among WIC Participants Characteristic All (n=3586) % English (n=2382) % Spanish (n=1204) % P Value* WIC Staff Provided Instruction on Voucher Use Yes 96.4 95.9 97.3 0.0474 No 3.6 4.1 2.7 Participant Understanding of Voucher Use Excellent 68.2 75.2 53.9 0.0001 Good 28.5 22.8 40.2 Fair 3.1 2.0 5.4 Poor 0.3 0.1 0.6 Participant Told by Cashier They Picked the Wrong Food Item Almost Every Time 7.2 7.0 7.6 0.5746 Sometimes 59.3 59.9 58.1 Never 33.5 33.1 34.4 Numbers may not sum to total due to missing values * Chi Square Distribution, p values of <0.05 were considered statistically significant 26 Patterns of Participant Satisfaction with U tah W IC Program ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw Table 4: Rating of Services by WIC Participants WIC Service Rating All (n=3586) % English (n=2382) % Spanish (n=1204) % P Value* Rating Options 0.7438 Excellent 62.6 62.2 63.3 Good 32.3 32.7 31.3 Fair 4.8 4.8 4.9 Poor 0.4 0.3 0.5 Numbers may not sum to total due to missing values * Chi Square Distribution, p values of <0.05 were considered statistically significant (27.8%) expressed an interest in obtaining nutrition informa-tion via Internet classes or through the use of books and videos to be used at home (29.6%, see Figure 2). When asked to rate the services received from WIC, 62.6% of respondents said that the services were "excellent" (see Table 4). Only five percent rated services "fair," or "poor." No significant differences exist between those who completed the survey in English and Spanish in terms of service rating. It was of interest to note that participants who had been enrolled in the WIC program for less than one year rated the services more favorably than those who had been in the program longer than year (p <.01). Those respondents who indicated that a cashier either "sometimes" or "every time" said they picked the wrong WIC foods were also more likely to rate WIC services less fa-vorably than those who reported they had "never" been told by a cashier that they picked the wrong food item (p <.001). The logistic regression analysis showed that those who rated the services as "excellent" were more likely to also have modified their behavior toward a healthier lifestyle (see Table 5). They were significantly more likely to engage in more physical ac-tivity (OR = 1.8), to eat more fruits and vegetables (OR = 1.5), and to have had their children immunized (OR = 1.4) compared to those who made no changes in their behavior. Discussion and Conclusion The data gathered from the Participant Satisfaction Survey have provided valuable informa-tion to improve WIC program operation and service delivery. The insight into understanding the extent of service awareness and the service utilization pat-terns among WIC participants provides the knowledge base necessary for continued program and service improvement. Information gathered according to language version will allow WIC services to be tailored to the cultural preferences of each group. More than half (55.9%) of WIC participants indicated that the reason for their last appointment with the WIC clinic was cer-tification. This is a crucial aspect of the WIC program where participants can be counseled and empowered to make positive lifestyle changes. Federal regulations require height, weight, and blood sample measures be taken for WIC participants at certification visits. Such repeated measurements of the he-matological status of participants provide an opportunity to reduce the incidence of iron-deficiency anemia among infants (Devaney et al., 1997). During nutrition education classes participants can learn how to prepare well-balanced meals to improve the health of the entire family. More than half of all respondents said they eat more fruits and vegetables since they started WIC program. Close to one-third reported that because of coming to WIC they participate in more physical activities. The WIC program can benefit by offering additional educa-tional classes on the correct use of WIC vouchers since more Table 5: S ervice A s se s sm en t Ratings and Healthy Lifestyle M odifications Behavior Modification Number Odds Ratio (OR)¥ Confidence Interval No Changes Made 688 Eat Fruits & Vegetables 1966 1.467* 1.173 -- 1.834 Immunizations of Children 1029 1.394* 1.133 -- 1.715 Physical Activity 1132 1.761* 1.376 -- 2.253 Watch Television & Video Games 741 1.318 0.886 -- 1.960 ¥ Represents Unadjusted Odds Ratios - Referent Category * Statistically Significant ©2009 The University o f Utah. All Rights Reserved. Patterns of Participant Satisfaction with Utah W IC Program 27 2009 UTAH's HEALTH: AN ANNUAL REviEw than half of survey participants reported that a cashier "some-times" said they picked the wrong WIC foods. Mistakes in choosing the correct WIC foods appear to be associated with less satisfaction with WIC services. The program can serve as a vehicle to link participants with health care providers and sometimes operates as the gateway into the health care system (Bell et al., 2007). More than a quarter of WIC participants indicated that they immunized their children since they started coming to WIC program. Survey results identified several participation barriers that can now be improved. Both the effectiveness of nutrition education on family nutrition behaviors, as well as indications about the preferred methods of receiving nutrition education has pro-vided additional insight regarding the best directions to take in planning upcoming educational programs and for choosing the most effective information delivery methods. Close to one-third of respondents expressed an interest in obtaining nutrition information via internet classes. With the new Value Enhanced Nutrition Assessment (VENA) training, WIC staff can more effectively customize nutrition education plans for individual participants. The results of the Participant Satisfaction Survey suggest that the large majority of respondents rated WIC services as excel-lent and were highly satisfied. This bodes well for their future willingness to continue to participate in WIC services that offer such health benefits for them or their children. The findings of this survey will continue to enhance educational strategies that target healthy eating habits, nutritional equivalencies among food brands, awareness of eligibility requirements, and cor-rect voucher use. In the past, survey findings were utilized in decisions regarding the hiring of additional staff and improv-ing customer service through updated telephone service and expanded clinic hours. The combination of nutrition education, nutritious foods and access to other health services strengthens families long after their WIC eligibility has ended. As inherent in any surveys, some bias may exist. Since partici-pation in the survey was up to the discretion of each individual participant, there may be some self-selection bias. Although the standard administration instructions were given to all clin-ics, no oversight was provided to ensure uniformity between administration sites. All measures of behavior change are strictly based on participant self-report rather than objective measure. References Bell, L., Ledsky, R., Silva, S., & Anthony J. (2007). A n assessment o f the impact ofM ed ica id m anaged care on WIC program coord inatio n with p rim a ry care services (Contractor and Cooperator Report No. 33). USDA, Economic Research Service. Christie, C., Watkins, J.A., Martin, A., Jackson, H., Perkin, J.E., & Fraser, J. (2006). Assessment o f client satisfaction in six urban WIC clinics. Florida Pub lic H ea lth R e v iew , 3, 35-42. Devaney, B., Ellwood, M., & Love, J. (1997). Programs that mitigate the effects o f poverty on children. The F uture o f Child ren, 7 (2), 88-112. Edozien, J., Switzer, B., & Bryan, R. (1979). Medical evaluation o f the Special Supplemental Food Program for Women, Infants, and Children. The American Journal o f C linical N u tritio n , 32, 677-692. Lee, B. J., Mackey-Bilaver, L., & Chin, M. (2006). Effects o f WIC and Food Stamp p rogram particip a tio n on child outcomes (Contractor and Cooperator Report No. 27). USDA, Economic Research Service. Lee, B. J., M ackey-Bilaver, L., & Goerge, R. (2000). The p a tterns o f Food Stamp and WIC particip a tion and their effects on health o f low-income child ren (Working Paper, No. 129). Chicago, IL: Northwestern University/ University o f Chicago Joint Center for Poverty Research. Mathematica Policy Research Group. (1989). Savings in M edicaid costs fo r newborns and their mothers from p ren a ta l particip a tio n in the WIC program (Report submitted to the U.S. Department o f Agriculture, Food and Nutrition Service). Plainsboro, NJ: Mathematica Policy Research. Oliveira, V., R acine, E., Olmsted, J., & Ghelfi, L.M. (2002). The W IC program : Background, trends, and issues (Food Assistance and Nutrition Research Report No. 27, pp 44). USDA, Economic Research Service. United States General Accounting Office (GAO). (2001). Performance measures fo r assessing three WIC services (Food Assistance Report to Congressional Committees, GAO-01-339). United States General A ccounting Office. United States General Accounting Office (GAO). (2006). WIC P rogram: More de ta iled p ric e a nd q u a ntity data could enhance agr icu lture's assessment o f WIC program expenditures (Report to Congressional Requesters, GAO-06-664). United States General Accounting Office. WIC Nutrition Program Facts. (2006). The Special Supplemental Nutrition Program for Women, Infants and Children. Nutrition Program Facts. Food and Nutrition Service. United States Department o f Agriculture. Retrieved from: http://www.fns.usda.gov/wic/WIC-Fact-Sheet.pdf Public Health Funtions Steering Committee. (1995). Public Health in America. Public Health Functions Steering Committee, July 1995, Office o f Disease Prevention and Health Promotion, U.S. Department o f Health and Human Services. Retrieved from: http://web.health.gov/phfunctions/public.htm 28 Patterns of Participant Satisfaction with Utah W IC Program ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH'S HEALTH: AN ANNUAL REViEW Location o f Adverse Event Documentation in Hospital Inpatient Medical Charts Authors: Carol J. Masheter, PhD Paul Hougland, MD Wu Xu, PhD Key Words adverse event; adverse drug event; drug adverse event; medication adverse event; adverse drug reaction; medica-tion error; surgery-related adverse event; medical injury; medical chart review; hospital inpatient; iatrogenic Brief Bios Carol J. Masheter, PhD, epidemi-ologist, is affiliated with the Office of Health Care Statistics, Utah Depart-ment of Health, Salt Lake City, Utah. Paul Hougland, MD, consultant, is af-filiated with the Office of Health Care Statistics, Utah Department of Health, Salt Lake City, Utah. Wu Xu, PhD, data manager, is affili-ated with the Office of Public Health Informatics, Utah Department of Health, Salt Lake City, Utah. Acknowledgments This project is supported by grant num-ber U18 HS 11885 from the Agency for Healthcare Research and Quality (P.I. Scott D. Williams) and grant number P01 CD000284 from the Centers for Diseases Control and Prevention (P.I. Matthew H. Samore). Abstract Objectives: This paper reports where in hospital inpatient medical charts adverse events (AEs) are most frequently documented. Methods: Nurses, trained in the use of a structured chart review tool, reviewed 7,244 charts, selected from 493,256 inpatient stays at 41 Utah acute care hospitals in 2001 and 2003. Reviewers determined whether each undesired, unintended patient outcome was an AE and where in the chart it was documented. Results: Reviewers found 5,007 AE documentations for 2,454 AEs in 1,607 (22%) of the 7,244 reviewed charts. The chart location with the most AE documentations was progress notes (n=2,024), followed by the discharge summary (n=1,317). Percentage of AE documentations by chart location and level of harm to patients as well as percentage of AE documentations by chart location and AE type differed significantly. Conclusions: Progress notes and discharge summary included the most AE documenta-tions and highest percentage, regardless of harm level or AE type. Introduction Modern medications, surgeries and other treatments promise to cure patients, but they can cause harm if used inappropriately. Landmark publications on patient safety, such as "To Error Is Human" (Kohn et al., 2000) and "Crossing the Quality Chasm" (Institute of Medicine, 2001), have increased awareness, research funding and publications on medical injuries related to medical care, also known as adverse events (AEs). One major challenge to understanding and preventing AEs is developing systematic approaches to their detection. Review of patients' medical charts (chart review) has been a time-honored method of detection and study of AEs. While numerous studies have used chart review to study AEs, it is expensive and burdensome to conduct (Bates et al., 2003; Forster et al., 2005; Murff et al., 2003), particularly for larger studies, such as statewide or nationwide stud-ies. In fact, most studies that use chart review are smaller studies that have focused on ©2009 The University o f Utah. All Rights Reserved. Location o f Adverse Event Documentation 29 2009 UTAH's HEALTH: AN ANNUAL REviEw specific kinds of AEs, among certain kinds of patients, often at a single hospital unit, hospital or hospital system. Others have used chart review as a "gold standard" against which to evaluate other methods of AE detection, such as electronic AE monitoring tools and natural language processing (Bates et al., 2003), computerized surveillance (Evans et al., 1986; Jha et al., 1998), online incident reporting, telemetry check-lists, clinical engineering database and post-discharge patient survey (Samore et al., 2004). One approach to lessening the burden of chart review and making it more efficient is to start with portions of the chart (chart locations) that are most likely to include AE documentation. Past Research Computer assisted literature searches identified no published works that directly addressed which chart locations would be optimal "hunting grounds" for evidence of AEs. However, literature searches using the phrases "medical chart review", "medical record review", "chart review", "record review", "medical injury" and "adverse event" identified several studies that mention specific chart locations where researchers looked for AE evidence. Several studies focused on examining the discharge sum-mary or discharge abstract, the physician-generated, textual report that summarizes the patient's hospital stay (Forster et al., 2005). Kossovsky et al. (1999) used narrative data from discharge summaries to distinguish between planned and un-planned hospital readmissions. For their study of surgical or medical AEs, Weingarten et al. (2000) reviewed charts whose discharge abstracts included evidence of these types of AEs. Forster et al. (2005) developed an electronic method for screen-ing medical discharge summaries for AEs. They hypothesized that the discharge summaries would include information on complications, so they tested their electronic method on this portion of charts. These and other studies suggested that the discharge summary would be a likely chart location in which to find evidence of AEs. In a review of studies that have used computerized methods to detect AEs, Bates et al. (2003) cited several studies that scanned certain parts of patients' records and used chart review as a "gold standard". Examples included microbiology, labora-tory and pharmacy data for nosocomial infections (Evans et al., 1986; Hirschhorn et al., 1993; Rocha et al., 1994), ADEs (Honigman et al., 2001; Jha et al., 1998; Levy et al., 1999;) and surgery-related AEs (Weingart et al., 2000). In a more recent study of anesthesia-related AEs among Dutch children with mitochondrial defects, Driessen et al. (2007) performed a retrospective review of the anesthesia, surgical and medical records of children that underwent diagnostic surgical muscle biopsy. In a study of intravenous epinephrine for treatment of emergency department patients experiencing severe asthma, Putland et al. (2006) reviewed medical and nursing notes, drug orders and nursing observation charts on which vital signs were recorded. Brown et al. (2000) determined an ADE rate from evidence of ADEs in patient laboratory and pharmacy data. Investigators apparently thought these locations in pa-tient records or charts would be good "hunting grounds" for the kinds of AEs they were studying. This paper builds on these previous studies and examines directly the locations in hospital patient medical charts for evi-dence of AEs. Specifically this paper addresses the following questions: What chart locations are most likely to include AE documenta-tion? As previous research implies, is the discharge summary the best chart location to find evidence of AEs? Does the AE harm level differ by chart location? Is the dis-charge summary more likely to document AEs that cause patients serious harm than other chart locations? Does AE documentation vary by AE type? For example, are surgery-related AE documentations more likely to be found in anesthesia, operative and recovery room reports than other chart locations? Are ADE documentations more likely to be found in medication administration sheets and laboratory re-ports than other chart locations? Methods Contents of Medical Charts The following section briefly describes a patient medical chart and the information it contains. A patient's hospital medical chart starts when the patient is admitted to a hospital. It documents the patient's conditions and diagnoses, along with medications, surgeries and other treatments that the patient received during the hospital stay. The patient's chart is updated daily throughout the hospital stay. It ends with a discharge summary, which is a narrative written by one of the patient's physicians and summarizes the patient's hospital stay from admission to discharge. Each patient's hospital chart includes the following narrative sections: history and physical examination, progress notes and discharge summary from physicians and nursing notes from nurses. If the patient received medication during the hospital stay, the chart may include pharmacist notes. If the patient had surgery, the chart may include anesthesia, operative and recovery room notes. Depending on these and other kinds of treatment, the chart may include consultant notes, procedure notes and radiology reports. Each patient's chart also includes documentation sections, such as a registration form, the pa-tient's vital signs, physician orders, medication administration sheets and laboratory reports. 30 Location o f Adverse Event Documentation ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH'S HEALTH: AN ANNUAL REViEW Definitions Similar to the Harvard study (Brennan et al., 1991; Leape et al., 1991)and the Utah and Colorado study (Thomas et al., 2000), this paper defines an AE as any undesired, unintended patient outcome that caused harm to the patient and is more likely due to medical care than the patient's disease (see Table 1). A chart location that included evidence of an AE is AE documentation. In this study, each AE must have at least one AE documentation and may have more than one AE docu-mentation. Any AE without at least one AE documentation was excluded from further analysis. Sample This study analyzed chart review findings from a larger study on detection of AEs among hospital inpatients (Hougland et al., 2006; Hougland et al., 2008). In the larger study, 7,260 charts were selected from 493,256 inpatients stays at 41 acute-care hospitals in Utah during the calendar years 2001 and 2003 (see Table 2). Each year's charts came from two samples, a population sample and a flagged sample. Of these charts, 7,244 were reviewed. The remaining 16 charts were unavailable or not sufficiently complete for adequate review. The population sample (2,065 charts from 2001 and 1,923 charts from 2003) was selected from all hospital stays in all 41 Utah acute care hospitals for these two years (Hougland et al., 2006; Hougland et al., 2008). It was stratified by hospital setting (rural vs. not rural) and patient length of stay (three days or less vs. four days or longer). The flagged sample (1,638 charts from 2001 and 1,634 charts from 2003) was a random sample of those charts having at least one of 1,003 ICD-9-CM AE flag codes as a secondary diagnosis or E-code. A national panel of experts had previously determined that the 1,003 flag codes were likely to indicate an AE as defined in this study. Chart Review The reviewers were nurses trained in the use of a structured chart review tool (Hougland et al., 2006; Hougland et al., 2008), similar to previously published tools (Brennan et al., 1991; Leape et al., 1991; Thomas et al., 2000). First, the reviewers reviewed each of the 7,244 charts looking for undesired, unintended patient outcomes or potential AEs. Second, the reviewers evaluated each potential AE in terms of whether the patient was harmed (see Table 1). Reviewers indicated the kind of harm for each potential AE from a list of 20 specific choices ranging from "no clinical change/no apparent injury" through "decreased levels of consciousness," "transfer to a higher level of care," and "death." For this study these 20 choices were collapsed into four harm levels: major, moderate, minor and none. If reviewers indicated more than one harm level for an undesired, unintended patient outcome, the highest of these became the harm level used for this study. For example, if a reviewer indicated that a potential AE resulted in "increased observation required for side effects" (minor harm) and "residual physical impairment" (major harm), the potential AE caused major harm. If the reviewer did not indicate a level of harm, the harm level was coded as "missing". Only potential AEs that caused major, moderate or minor harm were further analyzed; potential AEs that caused no harm or were missing were excluded from further analysis. Third, reviewers rated each potential, harmful AE using a six-point scale on the likelihood that medical treatment caused it rather than the patient's disease (see Table 1). All potential, harmful AEs for which reviewers indicated "case management causation more likely than disease", "moderate/strong evidence for management causation" or "virtually certain evidence for management causation" counted as AEs. Fourth, reviewers indicated where in the chart that the AE was documented. Reviewers could check one or more of the following 10 chart locations: anesthesia report, consultation reports/notes, dis-charge summary, laboratory reports, medication administra-tion sheets, physician order sheets, procedure notes, progress notes, operative report and recovery room. If reviewers wished to indicate a chart location not among these specified ones, they could write it in under "other". For example, an AE for a wrong dosage of a medication could have been documented in the discharge summary, progress notes, medication admin-istration sheets and other (nursing notes) for a total of four AE chart locations for this particular AE. Reviewers determined the AE type. They could indicate explic-itly whether an AE was an ADE and whether it was a surgery-related AE. Any AEs that reviewers considered to be both an ADE and a surgery-related AE were counted as an ADEs. Any AE that the reviewers did not indicate was an ADE or a surgery-related AE was counted as an "other AE". ADEs included (but were not limited to) wrong medication, wrong dosage, wrong means of administration or wrong patient. ADEs also included negative reactions or outcomes to medications such as rashes, itching, changes in mental status, or negative effects on renal, heart, digestive or other system functioning. Surgery-related AEs included (but were not limited to) drop in hematocrit after surgery, mucosal tears during surgery, postoperative bleeding and heart attack. Examples of AEs that were both ADEs and surgery-related AEs included (but were not limited to) adverse effects of anesthesia for surgery, such as temporary hypoxia, nausea and vomiting after surgery due to medication, postoperative drop in blood pressure, postopera-tive over-sedation and postoperative paralytic ileus. Other AEs that were neither ADEs nor surgery-related AEs included (but were not limited to) injuries during forceps delivery, drop in blood pressure, infections, dehydration or over hydration. Physicians, also trained in the use of the structured chart review tool, reviewed a subset of the nurse-reviewed charts for verification, as physicians usually conduct chart review ©2009 The University o f Utah. All Rights Reserved. Location o f Adverse Event Documentation 31 2009 UTAH's HEALTH: AN ANNUAL REviEw for research-based studies. The physician-reviewed charts included a random sample of charts in which the nurses found AEs and charts in which the nurses did not find AEs as well as any charts about which the nurses had questions. Analyses The purpose of the first analysis was to determine where in patients' medical charts AE documentation was most likely to be found. For this study, each chart location checked by the reviewer counted as one AE documentation. First, the total number of AE documentations was determined for all of the re-viewed charts. Second, the number and percentage of AE docu-mentations by chart location and harm level was determined. Third, the number and percentage of AE documentations by chart locations and AE type was determined. 95% Confidence Intervals were used as a test for statistical significance of dif-ferences among percentages of AE documentations. The second analysis found the number of AEs represented by the AE documentations. Because each AE could have more than one AE chart location, each AE was counted only once in a hierarchy of mutually exclusive chart locations. The order of chart locations in this hierarchy started with the discharge summary, because previous research indicated that it would be likely to include evidence of AEs. Any AE not detected in the discharge summary then was sought among the remain-ing chart locations, starting with the chart location with the highest number of AE documentations. Any AE not detected in the discharge summary and this chart location was then sought in the remaining chart locations in descending order by number of AE documentations. This process was repeated for all chart locations specified in the chart review tool plus a write-in, "other" chart location. Then the number of AEs was determined by harm level and AE type. Results AE Documentations The reviewers found 5,007 AE documentations representing 2,454 AEs in 1,607 (22%) of the 7,244 reviewed charts. The average number of AE documentations per AE was 2, with a range of one through 6. Of the 2,454 adverse events, 1,219 (50%) were ADEs, 392 (16%) were surgery-related AEs and 843 (34%) were other AEs. 735 (30%) AEs caused major harm, 1,507 (61%) AEs caused moderate harm and 212 (9%) AEs caused minor harm. Regarding chart location, 1,317 (26% of 5,007) AE documenta-tions were in the discharge summary. 2,024 (40%) AE docu-mentations were in the progress notes. 1,072 (53% of 2,024) AEs documented in progress notes were also documented in the discharge summary. Each of the remaining nine chart locations included a lower percentage of AE documentations. When combined, these other chart locations included 1,666 (33% of 5,007) AE documentations. Regarding harm level, the discharge summary included 489 (29% + 1.3% Exact 95% Confidence Interval of 1,697) major harm AE documentations, 749 (25% + 1.2% of 2,981) moderate harm AE documentations and 79 (24% + 1.2% of 329) minor harm AE documentations (see Figure 1). For the discharge sum-mary the percentage of major harm AE documentations was significantly higher than the percentages of moderate harm and minor harm AE documentations. Progress notes included 602 (36% + 1.3% of 1,697) major harm AE documentations, 1,261 (42% + 1.4% of 2,981) moderate harm AE documentations and 161 (49% + 1.4% of 329) minor harm AE documentations (see Figure 2). For progress notes the percentage of major harm AE documentations was significantly lower than the percentage of moderate harm AE documentations which in turn was signifi-cantly lower than minor harm AE documentations. Other chart locations included 606 (36% + 1.3% of 1,697) major harm AE documentations, 971 (33% + 1.3% of 2,981) moderate harm AE documentations and 89 (27% + 1.2% of 329) minor harm AE documentations. For other chart locations the percentage of major harm AE documentations was significantly higher than the percentage of moderate harm AE documentations which in turn was significantly higher than the percentage of minor harm AE documentations. Regarding AE type, the discharge summary included 640 (25% + 1.2% of 2,512) ADE documentations, 243 (30% + 1.3% of 823) surgery-related AE documentations and 434 (26% + 1.2% of 1,672) other AE documentations (see Figure 2). For the discharge summary the percentage of surgery-related AE documentations was significantly higher than the percentage of ADE and other AE documentations. Progress notes included 1,030 (41% + 1.4% of 2,512) ADE documentations, 297 (36% + 1.3% of 823) surgery-related AE documentations and 697 (42% + 1.4% of 1,672) other AE documentations. For progress notes the percentage of surgery-related AE documentations was significantly lower than the percentage of drug and other AE documentations. Other chart locations included 842 (34% + 1.3% of 2,512 ADE documentations, 283 (34% + 1.3% of 823) surgery-related AE documentations and 541 (32% + 1.3% of 1,672) other AE docu-mentations. For other AE documentations the differences in percentage by AE type were not significant. When analyzed separately, most of the other chart locations had significant differences, but the percentages of AE documentations were small. Specifically, medication administration sheets included 150 (6% + 0.8% of 2,512) ADE documentations, 20 (2% + 0.2% of 823) surgery-related AE documentations and 33 (2% + 0.2% of 1,672) other AE documentations. For medication admin-istration sheets the percentage of ADE documentations was significantly higher than the percentages of surgery-related 32 Location o f Adverse Event Documentation ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH'S HEALTH: AN ANNUAL REViEW and other AE documentations. Operative reports included 19 (1% + 0.1% of 2,512) ADE documentations, 62 (8% + 0.3% of 823) surgery-related AE documentations and 25 (1% + 0.2% of 1,672) other AE documentations. For operative reports the percentage of surgery-related AE documentations was significantly higher than the percentage of drug and other AE documentations. Anesthesia reports included 22 (1% + 0.2% of 2,512) ADE documentations, 9 (1% + 0.1% of 823) surgery-related AE documentations and 6 (0.4% + 0.1% of 1,672) other AE documentations. For anesthesia reports the percentage of other AE documentations was significantly lower than the percentages of surgery-related and ADE documentations. Phy-sician order sheets included 399 (16% + 0.8% of 2,512) ADE documentations, 76 (9% + 0.3% of 823) surgery-related AE documentations and 181 (11% + 0.5% of 1,672) other AE docu-mentations. For physician order sheets the percentage of ADE documentations was significantly higher than the percentages of surgery-related and other AE documentations. Discussion This study's results provide answers to the questions posed near its beginning. Regarding which chart locations were most likely to include evidence of AEs, progress notes included the most AE documentations, followed by the discharge summary. At first glance, these findings suggest that progress notes are the richest hunting ground for AE documentation. However, previ-ous studies (e.g., Forster et al., 2005) had implied that the dis-charge summary would document most, if not all, AEs because the discharge summary presumably captures the most impor-tant elements of the patient's hospital stay. Finding only about half (53%) of AEs documented in the discharge summary was surprising. If the discharge summary did not include most or all AEs, we expected it would capture the most "serious" AEs, or those that cause major harm. However, the percentage of major harm AE documentations in the discharge summary was significantly lower (29% + 1.3%) than the percentage of major harm AE documentations found in progress notes (36% + 1.3%) and other chart locations (36% + 1.3%). Significantly higher percentages of moderate harm and minor harm AE documentations also were found in progress notes and other chart locations than in the discharge summary (see Figure 1). Previous research implied that some chart locations would be more likely to include evidence of certain types of AEs. For example, operative reports and anesthesia reports might be good hunting grounds for surgery-related AEs. In fact, these two chart locations did include significantly higher percent-ages of surgery-related AE documentations compared to drug and other AE documentations (see Figure 2). However, the percentages were low. Operative reports included 8% + 0.3% surgery-related AE documentations and anesthesia reports included 1% + 0.2% surgery-related AE documentations, com-pared to 36% + 1.3% surgery-related AE documentations for Figure 2. Percentage of Adverse Event (AE) Documentations by AE Type and Chart Location, Utah, 2001, 2003 60% n 50% 40% egatnecre P 30% 20% 10% 0% 30% 25% f a 26% i r 41% i 36% £ - 42% Discharge Summary Progress Notes 3344%/0 34T% 32% □ Drug AEs □ Surgery-related AEs □ Other AEs Other Chart Locations ©2009 The University o f Utah. All Rights Reserved. Location o f Adverse Event Documentation 33 2009 UTAH's HEALTH: AN ANNUAL REviEw Table 1. Patient Harm and Causality Criteria for Adverse Events If an unintended and undesired patient outcome occurred that harmed the patient, reviewers considered it a potential adverse event. Description of Patient Harm Potential Adverse Event? Harm Level 1 No clinical change/no apparent injury No None 2 No additional lab or diagnostic tests ordered No None 3 Minor change in condition, single lab or diagnostic test ordered Yes Minor harm 4 Increased observation required for side effects Yes Minor harm 5 Vital signs changed Yes Moderate harm 6 Additional medications. Diagnosis or treatment required Yes Moderate harm 7 Change in therapy Yes Moderate harm 8 Decreased level of consciousness Yes Moderate harm 9 Multiple lab or diagnostic tests needed for follow-up Yes Moderate harm 10 Cardiac changes requiring intervention Yes Major harm 11 Hospital acquired fracture Yes Major harm 12 Bleeding requiring intervention Yes Major harm 13 Transfer to higher level of care Yes Major harm 14 Lab values changed to critical values Yes Major harm 15 Unplanned surgical procedure due to complications Yes Major harm 16 Length of stay increased Yes Major harm 17 Residual physical impairment Yes Major harm 18 Cardiac/respiratory arrest/failure/placed on respirator Yes Major harm 19 Critical lab values became more critical Yes Major harm 20 Death Yes Major harm 21 Missing No Omitted Reviewers considered a potential adverse event to be an adverse event if it was more likely to be caused by medical care than the patient's disease. Description of Causality Adverse Event? 1 Virtually certain evidence for disease causation No 2 Moderate/strong evidence for disease causation No 3 Disease causation more likely than case management No 4 Case management more likely than disease Yes 5 Moderate/strong evidence for case management causation Yes 6 Virtually certain evidence for case management causation Yes 7 Unable to determine No progress notes and 30% + 1.3% surgery-related AE docu-mentations for discharges summary. The percentage of ADE documentations in medication administration sheets was sig-nificantly higher (6% + 0.8%) than the percentage of surgery-related AE (2% + 0.2%) and other AE documentations (2% + 0.2%), but it was significantly lower than the percentages of ADE documentations in the discharge summary (25% + 1.2%) and progress notes (41% + 1.4%). In short, this study's findings showed that progress notes had the highest percentage of the AE documentations, particularly for major harm AEs as well as for all three types of AEs (ADEs, surgery-related AEs and other AEs). While the discharge summary captured a lower percentage of all AEs and major harm AEs than expected, it captured higher percentages than other chart locations except progress notes. In this study, the discharge summary and progress notes combined documented 92% of the AEs that reviewers found. 34 Location o f Adverse Event Documentation ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw Table 2. Inpatient Chart Selection, Adverse Event Documentation and Adverse Events, Utah, 41 Acute Care Hospitals, Inpatient Charts, 2001 and 2003 7,260 charts were selected from the 493,256 inpatient stays at 41 Utah acute care hospitals for the calendar years 2001 and 2003. Of the 7,260 charts, 3,988 were a population sample (2,065 from 2001 and 1,923 from 2003) and a flagged sample (1,638 from 2001 and 1,634 from 2003). The population sample was stratified by hospital setting (rural vs. urban) and patient length of stay (fewer than 4 days vs. 4 days or more). The flagged sample was a random-selection sample of all charts with at least one of 1,003 AE flag ICD-9-CM codes. Nurses, trained in the use of a structured chart review tool, reviewed 7,244 of the 7,260 charts. The remaining 16 charts were unavailable or not sufficiently complete for adequate review. Physicians, also trained in the tool, reviewed a subset of the 7,244 charts. Reviewers could report a maximum of five adverse events per chart and could indicate more than one adverse event chart location for each adverse event. Reviewers found 2,454 adverse events in 5,007 adverse event chart locations in 1,607 (22%) of the 7,244 reviewed charts. The 2,454 adverse events included 1,219 (50%) adverse drug events, 392 (16%) surgery-related adverse events and 843 (34%) other adverse events. Of the 2,454 adverse events, 735 (30%) caused major harm, 1,507 (61%) caused moderate harm, and 212 (9%) caused minor harm. The chart location with the most AE documentations was progress notes (2,024 or 40% of 5,007 AE documentations), followed by the discharge summary (1,317 or 26% of 5,007 documentations). The remaining 1,666 AE documentations were found in nine other chart locations plus "other." 1,077 (53% of 2,024) adverse events found in progress notes also were found in the discharge summary. Strengths This study has several strengths. It addresses an area in the literature on AE detection by chart review about which little has been published: where in hospital patients' medical charts is evidence of AEs most likely to be found. Unlike many previous studies, this study used a relatively large sample of charts that represents all hospital inpatients from all acute care hospitals in the state of Utah, including both rural and urban acute care hospitals as well as inpatients with both longer and shorter hospital stays. The reviewers reviewed charts that included evidence of no AEs as well as charts that included one or more AEs. The nurse and physician review-ers were trained in the use of a structured chart review tool, which contributed to the consistency of their reviews. The tool and definitions used in this study have built on previous, prominent studies. The study considered several AE types as well as harm levels. All of these features contribute to the usefulness of this study's findings for a wide range of investi-gators interested in medical chart review and AEs. Limitations Along with its strengths, this study has several limitations. Some are characteristic of chart review in general. Chart review is vulnerable the subjectivity of its reviewers. Though training reviewers in the use of a structured chart review tool and AE definitions based on well-regarded previous research may have reduced reviewer subjectivity, the tool and train-ing in its use possibly did not eliminate review subjectivity entirely. Chart review is dependent on the completeness and accuracy of the information within the charts reviewed (Forster et al., 2005). While a section of the structured chart review tool asked reviewers to determine whether the chart was sufficiently com-plete to conduct the review, this determination could have been vulnerable to reviewer subjectivity. Some charts may have appeared to be sufficiently complete but were lacking crucial data about which the reviewers were unaware. Other charts may have not have included adequate documentation of AEs, so that the reviewers could not confirm them. This study's sample of charts reflects the population of Utah hospital inpatient stays in 2001 and 2003 with over-sampling of rural hospital inpatient stays and stays of four days or longer. However, charting practices may vary in other states and over time in Utah. Conclusions Progress notes emerged as the clear "winner" as the most likely location in hospital inpatients' medical charts to find evidence of AEs. Discharge summaries took an unexpected second place. Focusing on progress notes especially would be useful for both retrospective and real-time chart review, whether manual or by some form of electronic data scanning or natural language processing, while the patient is still in the hospital. Adding re-view of the discharge summary, which becomes available after the patient leaves the hospital, could increases the number of AEs detected. We do not recommend partial chart review for AE surveillance, ©2009 The University o f Utah. All Rights Reserved. Location o f Adverse Event Documentation 35 2009 UTAH's HEALTH: AN ANNUAL REviEw which needs to detect as many AEs as possible. However, this study's findings provide hospital quality improvement person-nel a useful starting point, in terms of the most promising chart locations, to look for AEs for improvement of patient safety in care processes. The findings may also guide researchers who are interested in particular kinds of AEs or are interested in finding the most harmful AEs. Focusing on specialized chart locations may be appropriate for studies with narrower focus on specific AEs. More specialized locations are more likely to include the AEs of interest and less likely to include other kinds of AEs, reducing the number of false positives. This study's findings could be helpful to other investigators, whether they use manual or electronic chart re-view methods and whether they are interested in specific kinds of AEs or AEs in general. References Bates DW, Evans RS, M urff H, et al. Detecting adverse events using information technology. J Am M ed Inform A sso c . 2003 Mar-Apr;10(2):115-28. Brennan TA, Leape LL, Laird NM, Hebert Let al. Harvard Medical Practice Study I. Incidence o f adverse events and negligence in hospitalized patients: results o f the Harvard Medical Practice Study I. 1991. N E n g l J M ed . 324:370- 376, 1991. Brown S, Black K, Mrochek S, Wood A, et al. RADARx: Recognizing, assessing, and documenting adverse Rx events. Proceedings o f the AMIA Annual Symposium 2000; 101-105. Committee on Quality o f Health Care in America, Institute o f Medicine. Crossing the quality chasm: a new health system fo r the 21st century. Washington, DC: National Academy Press; 2001. Driessen J, Willems S, Dercksen S, et al. Anesthesia-related morbidity and mortality after surgery for muscle biopsy in children with mitochondrial defects. Paediatr A naesth. 2007 Jan;17(1):16-21. Evans RS, Larsen RA, Burke JP, Gardner RM, et al. Computer Surveillance o f hospital-acquired infections and antibiotic use. JAMA 1986; 256: 1007-1011. Forster AJ, Andrade J, van Walraven C. Validation o f a discharge summary term search method to detect adverse events. J Am M ed Inform A sso c . 2005 Mar-Apr;12(2):200-6. Epub 2004 Nov 23. Hirschhorn LR, Currier JS, Platt R. Electronic surveillance o f antibiotic exposure and coded discharge diagnoses as indicators o f postoperative infection and other equality assurance measures. Infe c t C ontr H osp Epidemiol 1998; 14:21-28. Honigman B, Lee J, Rothschild J, Light P, et al. Using computerized data to identify adverse drug events in outpatients. J A m M e d Inform A ssoc 2001; 8: 254-266. Hougland P, Xu W, Pickard S, Masheter C, Williams, SD. Performance of International Classification o f Diseases, 9th Revision, Clinical Modification codes as an adverse drug event surveillance system. M ed Care 2006 Jul;44(7):629-36. Hougland P, Nebeker J, Pickard S, Van Tuien M, Masheter C, Elder S, Williams SD, Xu, W. Using ICD-9-CM codes in hospital claims data to detect adverse events in patient safety surveillance. 2008. In A dvances in P a tien t Safety: New D irections an d Alternative A pp roaches: Volume 1. Assessment, AHRQ Publication No. 08-0034-1. Rockville, MD: Agency for Healthcare Research and Quality. Jha AK, Kuperman GJ, Teich JM, Leap L, et al. Identifying adverse drug events: development o f a computer-based monitor and comparison w ith chart review and stimulated voluntary report. J Am M e d Inform A ssoc 1998; 5: 305- 314. Kessomboon P, Panarunothai S, Wongkanaratanakul P. Detecting adverse events in Thai hospitals using medical record reviews: agreement among reviewers. J M e d A s so c Thai. 2005 0ct;88(10):1412-8. Kohn LT, Corrigan JM, Donaldson, MS, editors. To e rr is human: building a safer health system. A rep o r t o f the C ommittee on Q uality o f Health C are in Am erica, Institute o f Medicine. Washington, DC: National Academy Press; 2000. Kossovsky MP, Sarasin FP, Bolla F, Gaspoz JM, Borst F. Distinction between planned and unplanned readmissions following discharge from a Department o f Internal Medicine. M e th Inform M e d 1999; 38:140-143. Leape LL, Brennan TA, Laird N, et al. The nature o f adverse events in hospitalized patients: results from the Harvard M edical Practice Study II. N Engl J M ed 1991;324:377-84. Levy M, A zaz-Livshits T, Sadan B, Shalit M, et al. Computerized surveillance o f adverse drug reactions in hospital: implementation. E u rn J Clin Pha rm a co l 1999; 54: 887-892. Murff HJ, Forster AJ, Peterson JF, Fiskio JM, Heiman HL, Bates DW. Electronically screening discharge summaries for adverse medical events. J Am M ed Inform A sso c. 2003; 10: 339-350. Putland M, Kerr D, Kelly AM. Adverse events associated with the use o f intravenous epinephrine in emergency department patients presenting with severe asthma. A nnE m ergM ed . 2006 Jun;47(6):564-6. Rocha BH, Christenson JC, Pavia A, Evans RS, Gardner RM. Computerized detection o f nosocomial infections in newborns. Proceedings o f the Annual Symposium on Computer Applications in M edical Care 1994; 684-688. Samore M, Evans RS, Lassen A, et al. Surveillance o f medical device-related hazards and adverse events in hospitalized patients. JAMA 2004; 291: 325- 370. Thomas EJ, Studdert DM, Burstin HR, et al. Incidence and types o f adverse events and negligent care in U tah and Colorado. M ed Care. 2000 Mar;38(3):261- 71. Weingarten SN, Iezzoni LI, Davis RB, Palmer RH, et al. U se o f administrative data to find substandard are: validation o f the complications screening program. M ed Care 2000; 38: 796-806. Yokoe DS, Noskin GA, Cunningham SM, et al. Enhanced identification o f postoperative infections among inpatients. Emerg ing Infectious D isease s. 2004 Nov;10(11):1924-30. 36 Location o f Adverse Event Documentation ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw Brief Bios Carrie M cAdam Marx, RPh, MS is a research associate at the University o f Utah Pharmacotherapy Outcomes Research Center (PORC), where she participates in research studies that evaluate the value of pharmaceutical therapy. Previously she held roles focused on the development and delivery of disease management and appropriate use o f medicines programs for private and public third party payers. Ms McAdam-Marx research experience has been in the retrospective analysis o f large pharmacy/medical claims and medical records databases with a current emphasis in the areas o f postmenopausal osteoporosis and type 2 diabetes. Diana Brixner, PhD, RPh, is an associ-ate professor and chair o f the departm ent of Pharmacotherapy at the University of Utah College of Pharmacy. She serves as executive director o f the Pharmacotherapy Outcomes Research Center. She spent 15 years in the pharmaceutical industry in outcomes research and managed care, including a position as vice president of Health Care Management at Novartis Pharmaceuticals Corporation where she de-signed and conducted outcomes research to support the positioning of pharmaceutical products in the marketplace and to support formulary acceptance in managed markets. She is the 07-08 president of the Interna-tional Society of Pharmacoeconomics and Outcomes Research (ISPOR). Patricia Murphy, CNM, DrPH, is an associate professor and licensed clinical practitioner who brings 25 years experience in managing contraception to the table. She has been involved for several years in re-search on contraception, including studies of initiation and improving compliance, as well as clinical trials and pharmacokinetic/ pharmacodynamic studies o f drug-herb interactions w ith hormonal contraceptives. She has currently funded research investi-gating drug-herb interactions th at could diminish efficacy o f hormonal contraceptives. Correspondence Carrie McAdam Marx, RPh, MS University of Utah Pharmacotherapy Outcomes Research Center 421 Wakara Way, Suite 208 Salt Lake City, UT 84105 Phone: [phone] email: carrie.mcadam-marx@pharm. utah.edu Hormonal Contraception Use and Continuation by Formulation and Hormone Content in the Utah Medicaid Population Authors: Carrie McAdam-Marx RPh, MS Diana Brixner, PhD, RPh Patricia Murphy, CNM, DrPh Abstract In Utah, 12% of women with unintended pregnancies report having discontinued con-traception due to delivery method issues. As 47.6% of pregnancies in Utah women on Medicaid are unintended, it is important to understand hormonal contraception continu-ation patterns in this population. A retrospective utilization analysis was conducted in women in Utah Medicaid with a new hormonal contraception prescription. Refills at 3, 6 and 12 month were identified. A total of 49.3%, 43.0% and 31.2% of subjects had refills at 3, 6 and 12 months respectively. A greater proportion of oral contraceptive users received refills than those using a transdermal contraceptive at any time and for users of injectable contraceptives at 3 and 12 months. Ethinyl estradiol dose and progestin content also im-pacted continuation. Thus, hormonal contraceptive continuation in Utah Medicaid is low. Product selection and side effect counseling are important for contraceptive continuation and may help reduce the occurrence of unintended pregnancies in this population. Background Contraceptive related issues contribute to a large number of unintended pregnancies in the United States with one million pregnancies per year estimated to occur after misuse or discontinuation of oral contraceptives (Rosenberg, Waugh, & Long, 1995; Rosenberg, Waugh, & Meehan, 1995; Rosenberg & Waugh, 1998; Rosenberg, Waugh, & Burnhill, 1998). Thus, it is not surprising that over 20% of women using oral contraception dis-continue use during the first 6 months, while continuation rates fall over time ranging from 40% - 68% after 12 months (Rosenberg & Waugh, 1998; Trussell et al., 1995). Fur-thermore, only 16% to 34% of hormonal contraceptive users consistently obtain timely contraception refills over a 12 month period (Nelson, Westhoff, & Schnare, 2008). In Utah, 34% of births are unintended (Baksh, 2007; Baksh et al., 2007). Of the women in Utah with an unintended pregnancy, 58% report using contraception at the time of conception. While contraception failure appears to be prevalent, discontinuation of con-traception is also an issue. Approximately 12% of Utah women with unintended preg-nancies were not using contraception due to inability to tolerate the birth control method they were using (Baksh et al., 2007). Thus, discontinuation of hormonal contraception is contributing to unplanned pregnancies in this state. While 29% of all pregnancies in Utah are unplanned, the proportion of pregnancies that are unintended are highest in women who have less than a high school education ©2009 The University o f Utah. All Rights Reserved. Hormonal Contraception Use and Continuation 37 2009 UTAH'S HEALTH: AN ANNUAL REViEW monal contraception during the study period. A majority of the utilization in this population was for an oral form of delivery (85.1%), followed by injection (6.3%), transdermal (5.7%), and vaginal (2.9%). Overall, the percentage of women who received refills for a hor-monal contraception was 49.3% at three months, 43.0% at six months and 31.2% at 12 months (Figure 1). Continuation with the same form of birth control was higher at 3, 6 and 12 months among users of oral contraception than for transdermal formu-lations (p<.001). Continuation was also higher at 6 months for oral contraceptives than for injectables (p=0.01). No difference was detected between oral and injectable formulations at 3 and 12 months or vaginal formulations at any time. Continu-ation did not differ by ethinyl estradiol dose at 3 months, but continuation was significantly less at 6 and 12 months with lower dose products (p=0.03 and p>.001 respectively) (Figure 2). Continuation varied by progestins relative to drospirenone, with drospirenone having significantly higher continuation than norethindrone at all three time periods (p<.001). Continu-ation was not significantly different between drospirenone and other progestin groups (Figure 3). Continuation also did not differ significantly at any point in time between monophasic and triphasic products (data not shown). Discussion In a cohort of women continuously enrolled in Utah Medicaid using hormonal contraceptives, continuation with contracep-tion was low. Over half of all contraceptors did not obtain a refill for a hormonal contraceptive at three months, and at twelve months, 32% of the women obtained a refill. A national survey has reported continuation rates at 40% to 68% at three months (Rosenberg & Waugh, 1998), and a study specific to high risk Medicaid teenagers reported that 20% of this popula-tion had continued using contraception after 12 months (Zink, Shireman, Ho, & Buchanan, 2002). Thus, these data appear to provide a reasonable estimation of hormonal contraception continuation in Utah Medicaid. Continuation rates differed by formulation. The proportion of women remaining on contraception was less for those using a transdermal method at 3, 6, and 12 months and for those using injectable formulations at 6 months relative to those on oral contraception. Continuation was significantly less for oral contraceptives having lower doses of ethinyl estradiol dose and for those containing norethindrone than for products with desogestrel. There were also noted trends for reduced continu-ation with vaginal contraceptives at 12 months relative to oral contraception, for oral contraceptives containing ethynodiol ©2009 The University o f Utah. All Rights Reserved. Hormonal Contraception Use and Continuation 39 2009 UTAH's HEALTH: AN ANNUAL REviEw Figure 3. Oral Contraception Continuation by Progestin Content 3 Months 6 Months 12 Months -----•-----Norgestrel/Levonorgestrel (n=342) ---- ■---- Ethynodiol (n=33) - - - A- - - Norethindrone (n=891) - - M - - Desogestrel (n=190) - - Norgestimate (n=664) Drospirenone (n=284) *p<0.001 for norethindrone vs. drospirenone (p-value should be less than 0.012 with Bonferroni adjustment for an overall significance level of p=0.05) products at 12 months, and for oral contraceptives containing desogestrel at 6 months relative to drospirenone. While these trends likely represent clinically meaningful differences, they did not reach statistical significance due to the small number of women in the comparison groups. Product tolerability may contribute to the rates of hormonal contraceptive continuation in the Utah Medicaid population. This potential association is supported by the literature with one study finding that over 60% of female users of oral con-traception discontinue use due to side effects (Moreau, Cle-land, & Trussell, 2007). Furthermore, there are documented differences in tolerability between agents. For example, drospirenone, has been reported to have fewer adverse effects than other progestins such as weight gain and mood swings (Foidart, 2005). Similarly, the lowest doses of ethinyl estradiol are associated with more problems with spotting and hypom-enorrhea or amenorrhea than higher doses (Cianci & De Leo, 2007). Thus, selection of hormonal contraceptives with higher ethinyl estradiol doses and with a better tolerated progestin such as drospirenone, combined with counseling on managing side effects, may be important strategies for helping Medicaid women wishing to avoid pregnancy continue on contracep-tion. However, failure to continue contraceptive represents a relatively small portion of unintended pregnancies in Utah. While this is an important and potentially modifiable driver of unplanned pregnancies, socioeconomic and demographic factors as well as sexual activity likely contribute to this is-sue. Similarly, it is possible that contraceptive continuation in a Medicaid population differs than for the privately insured or the uninsured. However, this study was approached as utiliza-tion analysis, and investigators did not receive patient level data to further evaluate how demographics impact contraceptive continuation. Furthermore, socioeconomic and sexual activity data are not captured in Medicaid claims data and could not be evaluated as drivers of contraceptive continuation. Nonetheless, this pilot analysis confirms that hormonal contra-ception continuation in the Utah Medicaid population is low and suggests that continuation may be influenced by contra-ceptive tolerance. However, there are several limitations not previously discussed that warrant mention. First, an a priori power analysis was not conducted for this pilot study. Such an 40 Hormonal Contraception Use and Continuation ©2009 The University o f Utah. All Rights Reserved. 2009 UTAH's HEALTH: AN ANNUAL REviEw analysis would have predicted that that sample size would be an issue so that the study could be modified to optimize cohort size. In addition, it was not possible from the pharmacy utiliza-tion data to determine if the decision to discontinue hormonal contraception was due to change in sexual activity, desire to become pregnant, or to other reasons such as the ability to ac-cess care for a renewal prescription. It was also not possible to determine from this data if women who discontinued hor-monal contraception subsequently became pregnant. As such, additional research is needed to further assess the associations between product selection and contraceptive continuation con-trolling for demographic and socioeconomic factors, as well as the association between discontinuation and pregnancy rates in the Medicaid population. In conclusion, similar to other non-Utah Medicaid populations (Zink et al., 2002) and national studies (Rosenberg & Waugh, 1998), discontinuation of oral contraceptives in women covered by the Utah Medicaid program is high. These data suggest that appropriate product selection may help to prevent contracep-tion discontinuation, and thus, may help reduce unintended pregnancies in the Medicaid population. References Baksh L: Unintended Pregnancy. Salt Lake City, UT, University o f Utah, Utah Department o f Health, 2007. Baksh L, Bloebaum L, McGarry J, et al.: Births from unintended pregnancies and contraception use in Utah. Salt Lake City, UT, Utah Department o f Health, 2007. Cianci A, De Leo V: Personalisation of low-dose oral contraceptives. Pharmacological principles and practical indications for oral contraceptives. Minerva G inecol 59:415-425, 2007. Foidart JM: Added benefits o f drospirenone for compliance. Clim acteric 8 Suppl 3:28-34, 2005. Moreau C, Cleland K, Trussell J: Contraceptive discontinuation attributed to method dissatisfaction in the United States. Contracep tion 76:267-272, 2007. Nelson AL, W esthoff C, Schnare SM: Real-World Patterns o f Prescription Refills for Branded Hormonal Contraceptives: A Reflection o f Contraceptive Discontinuation. Obstet G ynecol 112:782-787, 2008. Rosenberg MJ, Waugh MS, Long S: Unintended pregnancies and use, misuse and discontinuation o f oral contraceptives. J R e p ro dM e d 40:355-360, 1995. Rosenberg MJ, Waugh M S, Meehan TE: Use and misuse o f oral contraceptives: risk indicators for poor p ill taking and discontinuation. Contracep tion 51:283- 288, 1995. Rosenberg MJ, Waugh MS: Oral contraceptive discontinuation: a prospective evaluation o f frequency and reasons. Am J O bstet G ynecol 179:577-582, 1998. Rosenberg MJ, Waugh MS, Burnhill MS: Compliance, counseling and satisfaction with oral contraceptives: a prospective evaluation. Fam Plann Perspect 30:89-92, 104, 1998. Trussell J, Leveque JA, Koenig JD, et al.: The economic value o f contraception: a comparison o f 15 methods. Am J Public H ea lth 85:494-503, 1995. Zink TM, Shireman TI, Ho M, et al.: High-risk teen compliance with prescription contraception: an analysis o f Ohio Medicaid claims. J P ed ia tr Adolesc G ynecol 15:15-21, 2002. ©2009 The University o f Utah. All Rights Reserved. Hormonal Contraception Use and Continuation 41 2009 UTAH's HEALTH: AN ANNUAL REviEw Associations between Educational Attainment and Diabetes in Utah: The Behavioral Risk Factor Surveillance System, 1996-2007 Authors: Eric N. Reither, PhD Theresa M. Fedor, BS Karin M. Abel, BS Dan J. Hatch, BA Key words Diabetes; obesity; socioeconomic status (SES); education; health disparities Brief Bios Dr. Reither is Assistant Professor of Sociology and affiliated with the Population Research Laboratory in the Department of Sociology, Social Work and Anthropology, Utah State University. Ms. Fedor is a M.S. student in Sociol-ogy and is affiliated with the Popula-tion Research Laboratory at Utah State University. Ms. Abel is a M.S. student in Sociol-ogy at Utah State University. Mr. Hatch is a Ph.D. student in Ex-perimental and Applied Psychological Science at Utah State University. Correspondence Eric N. Reither Utah State University Department of SSW&A 0730 Old Main Hill Logan, UT 84322-0730 office phone [phone] fax [phone] e-mail [e-mail] Abstract Background: Diabetes-now the sixth leading cause of death in Utah-has become more prevalent in recent decades. Diabetes is especially common among disadvantaged groups in Utah, such as persons without a high school diploma. To achieve stated public health goals of eliminating health disparities and increasing years of healthy life, steps must be taken to reverse recent trends in diabetes. Objectives: The main goal of this study is to examine how educational attainment is related to diabetes trends in the state of Utah over the period 1996-2007. Methods: We used Utah data from the Behavioral Risk Factor Surveillance System (BRFSS) to calculate the prevalence of diabetes by sex and educational attainment for three time periods (1996-1999, 2000-2003, and 2004-2007). To investigate how educa-tional attainment and other possible determinants influence the odds of diabetes, we examined a series of logistic regression models that were stratified by sex. Results: The prevalence of diabetes among adults in Utah increased by 44% during the period of observation in this study (from 4.37% in 1996-1999 to 6.30% in 2004-2007). In models controlling for age and educational status, the odds of diabetes were 60% higher among men and 44% higher among women in 2004-2007 than in 19