MANUEL C. F. PONTES, NANCY M. H. PONTES, and PHILLIP A. LEWIS

Transcription

1 Health Insurance Sources for Nonelderly Patient Visits to Physician Offices, Hospital Outpatient Departments, and Emergency Departments in the United States MANUEL C. F. PONTES, NANCY M. H. PONTES, and PHILLIP A. LEWIS Abstract. The authors used data from nationally representative surveys to estimate health insurance sources for nonelderly patient visits to U.S. physicians. Results show that hospital emergency departments attract a greater share of ambulatory care visits by uninsured patients than by patients with either Medicaid or private insurance. Results also show that hospital outpatient departments attract a greater share of visits by uninsured patients or patients with Medicaid than by patients with private insurance. The annual visit rate of uninsured individuals for nonemergency care is less than half of that for individuals with either private insurance or Medicaid. The proportion of uninsured emergency department visits by individuals between the ages of 0 and 64 years was significantly greater than the proportion of uninsured individuals between the ages of 0 and 64 years. In contrast, the proportion of uninsured physician office visits by individuals between the ages of 0 and 64 years was significantly less than the proportion of uninsured individuals between the ages of 0 and 64 years. Keywords: emergency department, National Ambulatory Care Survey, National Hospital Ambulatory Care Survey, uninsured Recently, there has been a growing concern regarding the increasing use of the emergency department by individuals without health insurance (Halfon et al. 1996; Richardson and Hwang 2001). A major goal of the present research was to examine whether hospital emergency departments attract a greater share of all uninsured ambulatory care visits than of all Medicaid visits or of all private insurance visits. A related goal of the present research was to examine whether the percentage of all emergency department visits by individuals with no health insurance between the ages of 0 and 64 years was greater than that of all individuals between the ages of 0 and 64 years. To that end, we used data from three nationally representative surveys to estimate the health insurance sources for patient visits for ambulatory care to physician offices, hospital outpatient departments, and hospital emergency departments in the United States. Literature Review Previous research has shown that uninsured children are nearly 4 times more likely to use the emergency department than are insured children (Johnson and Rimsza 2004). Also, research has shown that caregivers of uninsured children are more likely than are caregivers of insured children to identify the emergency department as the children s usual site of care (Newacheck, Pearl, et al. 1998). In addition, uninsured individuals are more likely to report frequent emergency department use, although the effect is only marginally significant (Hunt et al. 2006). A recent review of the literature found that increasing numbers of uninsured patients present to the emergency department (Newton et al. 2008). Manuel C. F. Pontes, PhD, is an associate professor in the Department of Marketing at the Rohrer College of Business at Rowan University. Nancy M. H. Pontes, DNSc, is the director of Student Health Services at Rowan University. Phillip A. Lewis, PhD, is an associate professor in the Department of Marketing at the Rohrer College of Business at Rowan University. Copyright 2009 Heldref Publications 19

2 20 Vol. 87, no. 3 Summer 2009 However, several studies have shown no relation between patients with no health insurance and more frequent emergency department use. One research study showed that health insurance status had no association with the use of the emergency department as a usual source of sick care (Halfon et al. 1996). Another research study showed that although uninsured individuals were more likely to report frequent emergency department use, the effect was only marginally significant (Hunt et al. 2006). Yet another study showed that privately insured and uninsured children showed no significant difference regarding the likelihood of having an emergency department visit (Luo et al. 2003). Research has shown that (1) uninsured and privately insured adults have the same risk of being frequent users of the emergency department (Zuckerman and Shen 2004), (2) uninsured individuals are no more likely to have an emergency department visit than are insured individuals (Weber et al. 2005), and (3) the rise in the number of emergency department visits between 1996 and 2003 was not solely because of more frequent visits by uninsured individuals (Weber et al. 2008). Relation between Patients with No Health Insurance and Use of Nonemergency Care Evidence has shown that patients without health insurance have less nonemergency care. Thus, children without health insurance are less likely than are children with health insurance to receive medically needed care from a physician (Stoddard, St Peter, and Newacheck 1994). Also, children without health insurance have significantly worse access to primary care than do children with health insurance (Newacheck, Stoddard, et al. 1998), and uninsured adolescents are more likely to miss needed care than are insured adolescents (Klein et al. 1999). Relation between Medicaid and Patient Use of Hospital Emergency Department Findings from Chande, Krug, and Warm s (1996) study showed that children with Medicaid are more frequent users of emergency departments and are more likely to use emergency departments as their regular site of care. In their study, children with Medicaid had more frequent visits to hospital emergency departments than did children with private health insurance. Phelps et al. (2000) found that caretakers of children with Medicaid insurance are more likely to view the emergency department as the children s usual site of care. Findings of Luo et al. s (2003) study showed that publicly insured children are more likely to have an emergency department visit during the year than are privately insured and uninsured children. Research has also shown that publicly insured adults are more than 2 times more likely to be frequent users of the emergency department than are privately insured adults (Zuckerman and Shen, 2004) and that individuals with Medicaid make more than 2.5 times more visits to the emergency department for infectious diseases than do individuals with other forms of health insurance (Adekoya 2007). Although many studies have shown that individuals with Medicaid are more likely to use the emergency department, at least two studies have shown that individuals with Medicaid are not greater users of the emergency department. Halfon et al. (1996) found that Medicaid coverage had no association with use of the emergency department as a usual source of sick care. Further, Johnson and Rimsza (2004) found that insured children covered by Medicaid were 54% less likely to use the emergency department than were children with private insurance. Thus, several studies have investigated whether individuals without health insurance or with Medicaid are more likely to visit an emergency department than are individuals with private health insurance. Results of these studies are mixed: Some studies have shown that individuals with Medicaid or with no health insurance make more emergency department visits than do individuals with private health insurance. Yet, other studies have shown that individuals with Medicaid or without health insurance do not make more emergency department visits than do individuals with private health insurance. A review of recent research found that the available data show that an increasing number of uninsured patients present to the emergency department. However, the data do not show that uninsured patients (1) are a primary cause of emergency department overcrowding, (2) present with less acute conditions than do insured patients, or (3) seek emergency department care primarily for convenience (Newton et al. 2008). However, there is clear research that shows that individuals without health insurance have fewer nonemergency patient visits (Klein et al. 1999; Newacheck, Stoddard, et al. 1998; Stoddard et al. 1994). A recent review found that uninsured patients lack access to primary care (Newton et al.

3 HOSPITAL TOPICS: Research and Perspectives on Healthcare ). Thus, previous research suggests that hospital emergency departments attract a greater share of uninsured ambulatory care visits than private insurance visits. Researchers have generally not examined whether hospital outpatient departments also attract a greater share of all uninsured ambulatory care visits than all private insurance visits. METHOD The goal of the present study was to examine the relation between health insurance source and the practice setting at which patients receive ambulatory care in the United States (physician office, hospital outpatient department, or hospital emergency department). To that end, we used data from three nationally representative ambulatory medical care surveys conducted by the National Center for Health Statistics (NCHS): the National Ambulatory Care Survey (NAMCS); and two National Hospital Ambulatory Care Surveys (NHAMCS), one for hospital outpatient departments (NHAMCS OPD) and the other for hospital emergency departments (NHAMCS ED). The target population of the NAMCS is the set of visits made by patients to the offices of U.S. physicians (except anesthesiologists, pathologists, and radiologists) who are engaged in direct patient care in nonhospital settings. The sampling frame of NAMCS excludes visits to hospital-based outpatient clinics, federal facilities, home visits, and telephone contacts (NCHS 2007a). The target population of the NHAMCS OPD is the set of visits to U.S. hospital outpatient departments with physician services clinics. The target population of the NHAMCS ED is the set of visits to U.S. hospital emergency departments. The sampling frame of NHAMCS excludes federal, military, and veterans administration hospitals and clinics that specialize in radiology, laboratory sciences, physical rehabilitation, and other ancillary services (NCHS 2007c). The NAMCS and the NHAMCS use three-stage and four-probability sample designs, respectively (NCHS 2007a, 2007c). For each of these surveys, data for a systematic random sample of visits were recorded by the physician or office hospital staff on standardized patient record forms. Although the NAMCS, NHAMCS OPD, and NHAMCS ED use slightly different forms, many of the items are common to the three forms. For all three surveys, data were recorded regarding demographic characteristics of patients (age, gender, race, ethnicity, geographic region, urban vs. rural location), health insurance source, and many other variables. Several hundred journal articles have been based on data from these surveys (NCHS 2007b). For the present study, we used a data set with 2 years of pooled data (2003, 2004) from the NAMCS and the NHAMCS (NHAMCS OPD and NHAMCS ED) to generate national estimates. We used 2 years of data to increase the sample size and power of the study. For the health insurance source variable, we assigned self-pay and charity care visits to the no health insurance group (Raofi and Schappert 2006). We combined the Medicare and other (e.g., CHAMPUS) sample visits into one group, labeled Medicare or other insurance. For these data sets, Medicaid visits included State Children s Health Insurance Program (SCHIP) visits (NCHS 2005). We calculated estimates of odds ratios (ORs) with 95% confidence intervals (CIs) and means or percentages with standard errors using the SAS (SAS Institute 2003) procedures, PROC SURVEYLO- GISTIC and PROC SURVEYMEANS, respectively. (These procedures are necessary for analyses of data from surveys with complex sample designs [Cherry 2004].) We also performed z tests for between-group differences (U.S. Census Bureau 2008). We obtained estimates of health insurance coverage for the population from previously published estimates based on data from National Health Interview Surveys conducted in 2003 and 2004 (Lethbridge-Çejku, Rose, and Vickerie 2006; Lethbridge-Çejku and Vickerie 2005). We obtained visit rates per 100 individuals, by health insurance source, by dividing the annual number of visits by the number of individuals in the population that were covered and multiplied the result by 100. Because health insurance coverage is almost universal for individuals aged 65 years or older (2003 4) we restricted our analyses to sample visits by individuals between the ages of 0 and 64 years that were covered either by private health insurance, Medicaid, or Medicare or other insurance (148,059 sample visits within this 2-year data set), as well as those with no health insurance. RESULTS Descriptive Statistics Selected demographic characteristics of visits by U.S. patients (between the ages of 0 and 64 years) are displayed in Table 1, by health insurance source.

4 22 Vol. 87, no. 3 Summer 2009 The mean patient age for Medicaid visits (21.80 years, SE = 0.68) was significantly less and the mean patient age for Medicare or other insurance visits (42.50 years, SE = 0.88) was significantly greater than the mean patient age for uninsured visits (35.70 years, SE = 0.44). Also, the proportion of uninsured visits made by males (40.6%) was significantly less than the proportion of Medicare or other visits made by males (44.6%). Finally, there were significant racial and ethnic differences in the share of patient visits across health insurance sources. For non-hispanic whites, the proportion of uninsured visits (63.3%) was significantly less than the proportion of private health insurance visits (76.9%) and significantly greater than the proportion of Medicaid visits (51.8%). In contrast, for non-hispanic blacks, the proportion of uninsured visits (15.4%) was significantly greater than the proportion of private health insurance visits (9.6%) and significantly less than the proportion of Medicaid visits (22.1%). Finally, for Hispanics, the proportion of uninsured visits (17.6%) was significantly greater than the proportion of private health insurance visits (9.4%). All other differences between uninsured and insured visits on these demographic variables were nonsignificant. Relation between Health Insurance and Probability of Emergency Department Visit We performed a logistic regression to analyze the relation between health insurance status and probability of emergency department visit, with health insurance payment source as the explanatory variable and whether the visit was an emergency department visit (yes or no) as the binary response variable (see Table 2). We performed the logistic regression analyses with and without the covariates of age, gender, and race or ethnicity of the patient. We used other demographic variables such as region of United States and urban versus rural location in a preliminary model but dropped them from the final model because the effects of those variables were nonsignificant. Results showed a significant relation between health insurance source and probability of hospital emergency department visit. Without adjustment for covariates, results showed that hospital emergency department visits accounted for a significantly smaller proportion of all privately insured visits in comparison with all uninsured visits (7.1% vs. 25.7%, respectively; unadjusted OR = 0.22; CI = ), with a significantly smaller proportion of all Medicaid visits in comparison with all uninsured visits (18.0% vs. 25.7%, respectively; unadjusted OR = 0.63; CI = ) and of Medicare or other insurance visits in comparison with uninsured visits (12.4% vs. 25.7%, respectively; unadjusted OR = 0.41; CI = ). Also, hospital emergency department visits accounted for a significantly greater proportion of all Medicaid visits than of all privately insured visits (18.0% vs. 7.1%, respectively; unadjusted OR = 2.88; CI = ) and a significantly greater proportion of all Medicare or other insurance visits than of all private insurance visits (12.4% vs. 7.1%, respectively; unadjusted OR = 1.87; CI = ). ORs estimated after adjustment for covariates were essentially similar to the corresponding unadjusted TABLE 1. Descriptive Statistics regarding Patient Visits, by Health Insurance Source No health insurance Private insurance Medicaid Medicare or other (n = 21,026) (n = 71,124) (n = 42,194) (n = 13,175) Variable M SE M SE M SE M SE Age (years) ** ** 0.88 Gender (%) Male * 1.14 Female * 1.14 Race/Ethnicity Non-Hispanic white ** ** Non-Hispanic black ** * Hispanic ** Other Note. Patients were between the ages of 0 and 64 years. * p <.05. ** p <.01.

5 HOSPITAL TOPICS: Research and Perspectives on Healthcare 23 TABLE 2. Percentage of All Ambulatory Visits to Hospital Emergency or Outpatient Departments, by Health Insurance Source Hospital emergency department Hospital outpatient department Group % OR CI Adj. OR CI % OR CI Adj. OR CI No health insurance as comparison group No health insurance Private Insurance Medicaid Medicare or other Private health insurance as comparison group Private insurance Medicaid Medicare or other Note. OR = odds ratio with no covariates; CI = 95% confidence interval; Adj. OR = adjusted odds ratio with covariates. ORs: These results show that the covariates do not explain all of the relations between health insurance and probability of emergency department visit (see Table 2). Relation between Health Insurance and Probability of Outpatient Department Visit Results showed a significant relation between health insurance source and probability of hospital outpatient department visit. Hospital outpatient department visits accounted for a significantly smaller proportion of all privately insured visits than of all uninsured visits (5.9% vs. 13.4%; unadjusted OR = 0.41; CI = ). In contrast, hospital outpatient department visits accounted for a significantly greater proportion of all Medicaid visits than of all uninsured visits (18.9% vs. 13.4%, respectively; unadjusted OR = 1.52; CI = ). The proportion of hospital outpatient department visits was not significantly smaller for Medicare or other insurance visits than for uninsured visits (10.8% vs. 13.4%, respectively; unadjusted OR = 0.79; CI = ). Results also showed that the proportion of hospital outpatient department visits was significantly greater for Medicaid visits than for privately insured visits (18.9% vs. 5.9%, respectively; unadjusted OR = 3.69; CI = ) and significantly greater for Medicare or other insurance visits than for privately insured visits (10.8% vs. 5.9%, respectively; unadjusted OR = 1.91; CI = ). Adjusted ORs computed with the addition of the covariates, age, gender, and race or ethnicity show similar results (Table 2). Effect of Covariates Results show that there was a significant negative relation between age and the probability of an emergency department visit (5-year OR = 0.94; CI = ; see Table 3). There were also significant relations between gender and the probability of an emergency department visit; relative to male visits, female visits had a significantly lower probability of being a hospital emergency department visit (OR = 0.79; CI = ). Finally, there were also significant relations between race or ethnicity and visit probabilities. Relative to non- Hispanic white visits, non-hispanic black visits were significantly more likely to be hospital emergency department visits (OR = 1.90; CI = ) or hospital outpatient department visits (OR = 1.63; CI = ). Also, relative to non-hispanic white visits, other race and ethnic groups visits were significantly less likely to be hospital emergency department visits (OR = 0.71; CI = ). Emergency and Nonemergency Department Visit Rates within Health Insurance Source Our results (see Table 4) show that the annual emergency department visit rate was lowest for individuals with private health insurance (22), followed by, in increasing order, those with no health insurance (43), Medicaid (80), and Medicare or other insurance (112). Thus, relative visit rates for emergency care (no insurance = 100) were 51 for private insurance, 186 for Medicaid, and 260 for Medicare or other insurance.

6 24 Vol. 87, no. 3 Summer 2009 Our results also show that the nonemergency department visit rate was lowest for individuals with no health insurance (123), followed by, in increasing order, those with private health insurance (294), Medicaid (363), and Medicare or other insurance (783). Thus, relative visit rates for nonemergency care (no insurance = 100) were 239 for private insurance, 295 for Medicaid, and 637 for Medicare or other insurance. Comparison of the Share of Individuals with Share of Ambulatory Visits by Insurance Source Results from the 2003 and 2004 National Health Interview Surveys show that 16.5% (SE = 0.18) of individuals between the ages of 0 and 64 years had no health insurance, 68.9% (SE = 0.28) had private health insurance, 11.9% (SE = 0.17) had Medicaid or SCHIP, and 2.8% had Medicare or other health insurance (Lethbridge-Çejku, Rose, and Vickerie 2006; Lethbridge-Çejku and Vickerie 2005; see Table 5). The percentage of individuals without health insurance (16.5%; SE = 0.18) was significantly less than the share of all hospital emergency department visits by individuals without health insurance (20.1%; SE = 0.72), z = 4.85, p <.01, but significantly greater than the share of all outpatient department (12.5%; SE = 1.34), z = 2.96, p <.01, and physician office visits by individuals without health insurance (6.4%; SE = 0.44), z = 21.25, p <.01. In contrast, the percentage of individuals covered by private health insurance (68.9%; SE = 0.28) was significantly greater than the private health insurance share of all emergency department visits (44.1%; SE = 0.97), z = 24.56, p <.01, and all outpatient department visits (44.3%; SE = 2.20), z = 11.09, p <.01, but significantly less than the privately insured share of all physician office visits (73.3%; SE = 0.93), z = 4.53, p <.01. Also, the percentage of individuals covered by Medicaid or SCHIP was significantly less than the Medicaid or SCHIP share of all hospital emergency TABLE 3. Effects of Covariates on Percentage of Visit Type Hospital emergency department Hospital outpatient department Variable Adj. OR CI Adj. OR CI Age (per increase of 5 years) Gender Male Female Race/Ethnicity Non-Hispanic white Non-Hispanic black Hispanic Other Note. Patients were between the ages of 0 and 64 years. Adj. OR = adjusted odds ratio; CI = 95% confidence interval. TABLE 4. Annual Visit Rates, by Health Insurance Source Emergency Nonemergency Variable Visit rate Relative visit rate Visit rate Relative visit rate No health insurance Private Medicaid Medicare or other Note. Patients were between the ages of 0 and 64 years. Visit rate = 100 number of annual of visits paid by health insurance source per 100 persons covered; Relative visit rate = visit rate for persons with no health insurance divided by corresponding visit rate for persons with specified insurance.

7 HOSPITAL TOPICS: Research and Perspectives on Healthcare 25 department visits (27.1%; SE = 0.82), z = 18.15, p <.01, and all hospital outpatient department visits (34.1%; SE = 1.85), z = 11.95, p <.01. Finally, the percentage of individuals covered by Medicare or other insurance was significantly less than the Medicare or other insurance share of all hospital emergency department visits (8.8%; SE = 0.40), z = 14.55, p <.01, and of all hospital outpatient department visits (9.1%; SE = 0.67), z = 9.30, p <.01, and all physician office visits (7.3%; SE = 0.39), z = 11.18, p <.01. DISCUSSION Results indicate that hospital emergency department visits accounted for more than 1 out of every 4 uninsured ambulatory care visits, more than 1 out of every 6 Medicaid visits, and more than 1 out of every 8 Medicare or other insurance visits, but less than 1 out of every 14 private health insurance visits (Table 2). Thus, results clearly show that individuals with no insurance had a greater preference for hospital emergency departments when seeking ambulatory care (Table 2). This preference for emergency care by individuals with no health insurance might result from reduced visits for nonemergency care and increased visits for emergency care. Results show that individuals with no health insurance had substantially lower visit rates for nonemergency care. In comparison with individuals with no health insurance, those with private health insurance made about 2.4 times as many physician visits, individuals with Medicaid made almost 3 times as many physician visits, and individuals with Medicare or other insurance made more than 6 times as many physician visits per year for nonemergency care (Table 4). In contrast, an examination of visit rates to emergency departments showed that individuals with no health insurance had higher visit rates than did individuals with private health insurance, but lower visit rates than did individuals with Medicaid or Medicare or other insurance (Table 4). Compared with individuals with no health insurance, individuals with private health insurance made about half as many emergency department visits per year, individuals with Medicaid made almost twice as many emergency department visits per year, and individuals with Medicare or other insurance made about 2.6 times as many emergency department visits per year (Table 4). Results displayed in Table 5 indicate that in , the share of emergency department visits that are covered by health insurance (79.9%) was significantly less than the proportion of the population covered by health insurance (83.5%). However, emergency department visit rates are dependent on the type of health insurance. Almost 7 out of every 10 individuals between the ages of 0 and 64 years are covered by private health insurance. Yet, private insurers cover less than 9 out of every 20 emergency department visits. In contrast, fewer than 1 out of every 8 individuals are covered by Medicaid, yet Medicaid covers more than 1 out of every 4 emergency department visits. Previous research has often shown that there is no relation between health insurance and frequency of emergency department visit (Halfon et al. 1996; Hunt et al. 2006; Luo et al. 2003; Weber et al. 2005; Zuckerman and Shen 2004). These research studies used selfreported health status and household income as covariates, which were significantly related to emergency department use (Halfon et al. 1996; TABLE 5. Percentage Share of Persons and Patient Visits Covered by Health Insurance Sources, by Visit Type Visit type Hospital emergency Hospital outpatient Physician Coverage department department office Insurance source % SE % SE z % SE z % SE z No health insurance Private Medicaid Medicare or other Note. Patients were between the ages of 0 and 64 years. Patient visits included all patient visits except workers compensation.

8 26 Vol. 87, no. 3 Summer 2009 Hunt et al. 2006; Luo et al. 2003; Weber et al. 2005; Zuckerman and Shen 2004). Previous research has shown that individuals with private health insurance have higher family incomes than do individuals with no health insurance or Medicaid coverage (Luo et al. 2003), and individuals with higher household incomes have better health status (Feinstein 1993; McLeod et al. 2003). Thus, it is plausible that the relation between health insurance and number of emergency department visits is largely explained by the relation between household income and number of emergency department visits. Unfortunately, these ambulatory data sets do not have information about family income or self-reported health status. Finally, this research shows that individuals with no health insurance or Medicaid are more likely than individuals with private health insurance to choose hospital-based care. Thus, 39.1% of uninsured visits and 36.9% of Medicaid visits but only 13.0% of private health insurance visits were to hospital-based ambulatory care facilities. Similarly, results from the present study show that hospital-based ambulatory care facilities treat a greater number of underserved patients. Thus, 47.1% of emergency department visits and 46.6% of outpatient department visits are by individuals with Medicaid or no health insurance. In contrast, only 19.3% of physician office visits are by individuals with Medicaid or no health insurance. These results clearly show the important role played by hospital ambulatory care settings for the treatment of underserved individuals. REFERENCES Adekoya, N Medicaid/State Children s Health Insurance Program patients and infectious diseases treated in emergency departments: U.S., Public Health Reports 122: Chande, V. T., S. E. Krug, and E. F. Warm Pediatric emergency department utilization habits: A consumer survey. Pediatric Emergency Care 12 (1): Cherry, D Understanding and using NAMCS and NHAMCS data: A hands-on workshop (Part II Using raw data files). Washington, DC: U.S. Department of Health and Human Services. Sess_44_NAMCS&NHAMCS_hands-on_CHERRY.ppt (accessed July 20, 2004). Feinstein, J. S The relationship between socioeconomic status and health: A review of the literature. Milbank Quarterly 71: Halfon, N., P. W. Newacheck, D. L. Wood, and R. F. St Peter Routine emergency department use for sick care by children in the United States. Pediatrics 98: Hunt, K. A., E. J. Weber, J. A. Showstack, D. C. Colby, and M. L. Callaham Characteristics of frequent users of emergency departments. Annals of Emergency Medicine 48 (1): 1 8. Johnson, W. G., and M. E. Rimsza The effects of access to pediatric care and insurance coverage on emergency department utilization. Pediatrics 113: Klein, J. D., K. M. Wilson, M. McNulty, C. Kapphahn, and K. S. Collins Access to medical care for adolescents: results from the 1997 Commonwealth Fund Survey of the Health of Adolescent Girls. Journal of Adolescent Health 25 (2): Lethbridge-Çejku, M., D. Rose, and J. Vickerie Vital and health statistics, series 10, number 228. Summary health statistics for U.S. Adults: National Health Interview Survey, Washington, DC: U.S. Department of Health Statistics. Lethbridge-Çejku, M., and J. Vickerie Vital and health statistics, series 10, number 225. Summary health statistics for U.S. adults: National Health Interview Survey, Washington, DC: U.S. Department of Health Statistics. Luo, X., G. Liu, K. Frush, and L. A. Hey Children s health insurance status and emergency department utilization in the United States. Pediatrics 112: McLeod, C. B., J. N. Lavis, C. A. Mustard, and G. L. Stoddart Income inequality, household income, and health status in Canada: A prospective cohort study. American Journal of Public Health 93: National Center for Health Statistics Instructions for patient record items, Washington, DC: National Center for Health Statistics a. NAMCS description. Washington, DC: National Center for Health Statistics. about/major/ahcd/namcsdes.htm (accessed May 18, 2008) b. National Ambulatory Care Survey (NAMCS) and National Hospital Ambulatory Care Survey (NHAMCS). List of publications (updated 12/03/2007). Washington, DC: National Center for Health Statistics. (accessed May 18, 2008) c. NHAMCS description. Washington, DC: National Center for Health Statistics. nchs/about/major/ahcd/nhamcsds.htm (accessed May 18, 2008). Newacheck, P. W., M. Pearl, D. C. Hughes, and N. Halfon The role of Medicaid in ensuring children s access to care. JAMA 280: Newacheck, P. W., J. J. Stoddard, D. C. Hughes, and M. Pearl Health insurance and access to primary care for children. New England Journal of Medicine 338: Newton, M. F., C. C. Keirns, R. Cunningham, R. A. Hayward, and R. Stanley Uninsured adults presenting to U.S. emergency departments: Assumptions vs data. JAMA 300: Phelps, K., C. Taylor, S. Kimmel, R. Nagel, W. Klein, and S. Puczynski Factors associated with emergency department utilization for nonurgent pediatric problems. Archives of Family Medicine 9: Raofi, S., and S. M. Schappert Vital and health statistics, series 13, number 163. Medication therapy in ambulatory medical care: United States, Washington, DC: U.S. Department of Health and Human Services. (accessed April 20, 2008). Richardson, L. D., and U. Hwang America s health care safety net: Intact or unraveling? Academic Emergency Medicine 8: SAS Institute SAS version for Windows. Computer software. Cary, NC: SAS Institute.

9 HOSPITAL TOPICS: Research and Perspectives on Healthcare 27 Stoddard, J. J., R. F. St Peter, and P. W. Newacheck Health insurance status and ambulatory care for children. New England Journal of Medicine 330: U.S. Census Bureau Instructions for applying statistical testing to ACS 2007 data. Washington, DC: U.S. Census Bureau. Weber, E. J., J. A. Showstack, K. A. Hunt, D. C. Colby, and M. L. Callaham Does lack of a usual source of care or health insurance increase the likelihood of an emergency department visit? Results of a national population-based study. Annals of Emergency Medicine 45 (1): Weber, E. J., J. A. Showstack, K. A. Hunt, D. C. Colby, B. Grimes, P. Bacchetti, and M. L. Callaham Are the uninsured responsible for the increase in emergency department visits in the United States? Annals of Emergency Medicine 52 (2): Zuckerman, S., and Y. C. Shen Characteristics of occasional and frequent emergency department users: do insurance coverage and access to care matter? Medical Care 42: